Temperature-sensitive Differential Affinity of TRAIL for Its Receptors

TRAIL is a member of the tumor necrosis factor (TNF) family of cytokines which induces apoptotic cell death in a variety of tumor cell lines. It mediates its apoptotic effects through one of two receptors, DR4 and DR5, which are members of of the TNF receptor family, and whose cytoplasmic regions contain death domains. In addition, TRAIL also binds to 3 “decoy” receptors, DcR2, a receptor with a truncated death domain, DcR1, a glycosylphosphatidylinositol-anchored receptor, and OPG a secreted protein which is also known to bind to another member of the TNF family, RANKL. However, although apoptosis depends on the expression of one or both of the death domain containing receptors DR4 and/or DR5, resistance to TRAIL-induced apoptosis does not correlate with the expression of the “decoy” receptors. Previously, TRAIL has been described to bind to all its receptors with equivalent high affinities. In the present work, we show, by isothermal titration calorimetry and competitive enzyme-linked immunosorbent assay, that the rank order of affinities of TRAIL for the recombinant soluble forms of its receptors is strongly temperature dependent. Although DR4, DR5, DcR1, and OPG show similar affinities for TRAIL at 4 °C, their rank-ordered affinities are substantially different at 37 °C, with DR5 having the highest affinity (K D ≤ 2 nm) and OPG having the weakest (K D = 400 nm). Preferentially enhanced binding of TRAIL to DR5 was also observed at the cell surface. These results reveal that the rank ordering of affinities for protein-protein interactions in general can be a strong function of temperature, and indicate that sizeable, but hitherto unobserved, TRAIL affinity differences exist at physiological temperature, and should be taken into account in order to understand the complex physiological and/or pathological roles of TRAIL.

TRAIL (1) (systematic name, TNFSF10; formally Apo-2L (2) or TL2(3)), a type II membrane protein, is a recently described member of the TNF 1 family of cytokines which induces apoptotic cell death in a variety of tumorigenic or transformed cell lines but not in normal cells (1). To date, five receptors, all members of the TNFR family, have been shown to bind to TRAIL. Of these, DR4 (4), DR5 (5)(6)(7)(8)(9)(10), and DcR2 (11,12) are membrane-anchored proteins with transmembrane and cytoplasmic domains, where DR4 and DR5 have functional death domains, while DcR2 contains a truncated, "non-functional" death domain. A fourth receptor, DcR1 (5,6,8,13), is glycosylphosphatidylinositol-anchored, which is unique among the TNFR family, and contains no transmembrane or cytoplasmic regions, while the fifth receptor, OPG, is a secreted protein with no known membrane anchor.
Four of the TRAIL receptors, DR4, DR5, DcR1, and DcR2, have all been mapped to a cluster within 8p22-p21 (10,11,14). OPG, the least related of the TRAIL receptor genes, by its sequence homology, structure, and ligand recognition patterns, including its high affinity for another ligand, RANKL, has been mapped to the other end of the same chromosome, to 8q24 (3). TRAIL itself has been mapped to 3q26 (1).
TRAIL-induced apoptosis appears to require expression of one or both of its death domain containing receptors DR4 or DR5 (15), and is mediated via a Fas-associated death domaindependent pathway (7). However, the expression patterns of DR4 and DR5 mRNA had suggested that DR4 and/or DR5 expression is necessary but not sufficient for TRAIL-induced apoptosis (15). The explanation for this latter observation was thought to be due to the expression of the "decoy" TRAIL receptors DcR1 and/or DcR2. Although ectopic expression, through transfection, of either of these two receptors into reporter cell lines (Refs. 5 and 12 for DcR1 and DcR2, respectively) has been shown to be protective against TRAIL-induced cell death, mRNA or protein (16) expression patterns of these two receptors did not support the hypothesis on their antiapoptotic roles with respect to transformed cell lines, since both sensitive and resistant cell lines have been shown to express DcR1 and/or DcR2 mRNA.
More recent data using mAbs for detection of TRAIL receptors on the cell surface corroborate the earlier mRNA data and show that the expression of the decoy receptors does not correlate with sensitivity or resistance of tumor cell lines to TRAILinduced cell death (16). Possible explanations which have been invoked to account for the discrepancy between the expression patterns of the TRAIL death and decoy receptors and sensitivity to TRAIL-induced cell death include intracellular regulation of caspase activation or the ability of DcR2 to provide intracellular anti-apoptotic signals, possibly through transcriptional regulation of other anti-apoptotic genes (15,17). However, despite these early indicators, other additional mechanisms are likely, and the protective mechanisms involved in the DR4 and DR5 apoptotic pathways, as well as the role of the decoy receptors in normal physiology or in pathological conditions remain to be elucidated.
One potential mechanism by which apoptotic responses to TRAIL may be regulated is by the expression levels of these receptors. In addition, the possibility that TRAIL may crosslink different receptors to form heteromeric complexes cannot be ruled out. In these cases, the affinity of TRAIL for the different receptors may be important to determine the cellular consequences of ligand binding. DR4, DR5, DcR1, and DcR2 have been shown to exhibit high affinity binding to TRAIL with comparable, subnanomolar K D (12,13,15,18). Similarly high affinity binding for OPG has also been reported (19), making it difficult to account for the known sequence and inferred structural diversity of this set of receptors that are recognized by the same ligand. As has been common practice for these types of studies, all these experiments were conducted at subphysiological temperatures. We have re-examined the affinities of recombinantly expressed DR4, DR5, DcR1, and DcR2 for recombinant TRAIL, using isothermal titration microcalorimetry (ITC) at 37°C, and by standard competition ELISA at 4 and 37°C. Here, we report, substantial temperature-sensitive differences in the relative affinities of these receptors for TRAIL which likely have important physiological implications. These findings are discussed in the context of the sequence and structural similarities and differences of these receptors.

Expression and Purification of Ig Fusion Proteins-
The cloning, expression, and purification of the Ig-fusion proteins DR5-Ig, DcR1-Ig, OPG-Ig, and HVEM-Ig has previously been described (19 -21). DR4-Ig was also similarly expressed. Briefly, the cDNA sequence coding for the leader and extracellular domains of the receptors were amplified by PCR and cloned upstream of a Factor Xa cleavage site and the Fc region of a human Ig␥1 heavy chain in plasmid COSFclink (22,23). The vector constructs were transfected into Chinese hamster ovary (CHO) cells and stable transfectants were selected in nucleotide-free media. For purification, the CHO expressed Ig-fusion proteins were captured from 10 or 30 liters of cell media on a 2.5 ϫ 11-cm Protein A column (Amersham Pharmacia Biotech) equilibrated in 20 mM sodium phosphate, 150 mM sodium chloride, pH 7 (PBS). The columns were eluted with 100 mM glycine, pH 2.5, and the eluates neutralized immediately with 1 M Tris, pH 8, yielding Ig-fusion proteins at Ͼ95% purity. The recovered proteins were analyzed by size exclusion chromatography on 1 ϫ 30-cm Superdex 200 analytical columns (Amersham Pharmacia Biotech), matrix-assisted laser desorption ionization-mass spectrometry, and N-terminal sequencing.
Construction and Expression of TRAIL in Pichia pastoris-Expression of the soluble form of TRAIL in CHO cells, with N-terminal peptide tags for affinity purification and detection, has previously been described (19). Briefly, the putative extracellular domain of human TRAIL (amino acids 95-281) (1) was constructed in the pCDN mammalian expression vector that had been modified to contain an in-frame tissue plasminogen activator signal sequence upstream of the FLAG-TRAIL coding region. For expression in P. pastoris, the region spanning amino acid residues 114 -281 was subcloned from this plasmid by PCR using the oligonucleotides 5Ј-AAAGAATTCCATCATCACCATCAT-CATATTGAAGGTAGAGTGAGAGAAAGAGGTCCTCAG-3Ј and 5Ј-AAAGGTACCTTATTAGCCAACTAAAAAGGCCCCG-3Ј which inserted an N-terminal His 6 tag (a tag consisting of 6 consecutive histidine residues) and Factor Xa cleavage site (IEGR) adjacent to residue 114 and 5Ј EcoRI and 3Ј Asp718I sites (underlined) for subcloning. The PCR product was subcloned into pCR2.1TOPO (Invitrogen), confirmed by sequence analysis, and subsequently subcloned into the EcoRI and Asp718I sites of pPICZ␣A (Invitrogen). The resulting plasmid, pPICZ␣A-TRAIL, has sTRAIL expression under the control of the methanol-inducible AOX1 promoter (24) and secretion under the control of the Saccharomyces cerevisiae ␣-factor signal sequence. This plasmid was digested with SstI, electroporated into P. pastoris strain KM71 (aox1⌬::SARG4, his4, arg4) (25), and transformants were selected according to protocols provided by Invitrogen. Strain MM49 was selected as the highest expressing strain as estimated by immunblot analysis with a monoclonal antibody specific for the hexa-histidine tag (CLON-TECH, Palo Alto, CA). Production was carried out in shaker flasks with a yield of about 25 mg/liter. TRAIL was then purified by capturing the protein from 1 liter of Pichia-conditioned media on a 10-ml Ni-NTA column (Qiagen) equilibrated in PBS, pH 7. The column was washed with 30 mM imidazole (Calbiochem) in PBS and eluted with 300 mM imidazole in PBS. 30 mg of His 6 TRAIL was recovered.
ITC of TRAIL and Its Receptors-The solution affinities of TRAIL for various receptors were measured by ITC (26,27), which detects the extent of binding from the intrinsic binding heat of the interaction. ITC is a quantitative method for measuring binding constants because it enables unmodified, native forms of proteins to be characterized in solution phase. ITC also measures binding at equilibrium and does not involve separation steps to quantitate bound and unbound species. Reactant concentrations were determined from molar extinction coefficients calculated from amino acid sequences (28). Titrations were conducted in 10 mM sodium phosphate, 150 mM NaCl, pH 7.4, at 37°C. Phosphate buffer was chosen by virtue of its small ionization enthalpy change of approximately 1 kcal/mol (26). The observed TRAIL-receptor binding enthalpy changes therefore closely approximate the molecular binding enthalpy changes, regardless of whether binding is coupled to changes in protonation. Circular dichroism thermal melting experiments demonstrated that TRAIL and the receptor constructs were stable against thermal unfolding up to at least 50°C.
ELISA Binding of TRAIL to the Soluble Receptors-ELISAs were performed according to standard techniques. Nunc MaxiSorp Flat-bottomed plates were coated overnight at 4°C with either DR5-Ig or OPG-Ig (20 ng/well, 100 l/well) in 0.05 M carbonate/bicarbonate buffer, pH 9.5-9.6. After washing with PBS-Tween (0.05% Tween 20) four times and blocking with 1% bovine serum albumin in phosphate-buffered saline for 2 h, serial dilutions of DR5-Ig/OPG-Ig/DcR1-Ig/DR4-Ig (800 to 6.25 ng/well, 100 l/well) and TRAIL-biotin (25 ng/well, 100 l/well), in pre-equilibrated assay buffers (0.01% Tween 20 in PBS pre-equilibrated at 37 or 4°C) were added and incubated for 1 h at 37 or 4°C. The plates were washed four times with PBS-Tween buffer and incubated with alkaline phosphatase-conjugated streptavidin (1:1000, 100 l/well) for 30 min at room temperature. The ELISA plates were washed five times and p-nitrophenyl phosphate disodium salt (pNPP Tablets, Pierce, Rockford, IL) (100 l/well) was added according to the manufacturer's instructions for 15 min. The absorbance was measured at 405 nm on a microtiter plate reader (MRX, Dynatech Laboratories).
Cell Surface Expression of TRAIL Receptors-The full-length cDNA clones of DR5, DR4, without their death domains (DR5⌬DD residues 1 to 268, DR4⌬DD residues 86 to 351), and DcR2 (residues 55-386) were generated by PCR and cloned into modified pcDNA3 vector that allowed in-frame fusion with a Flag epitope tag (8). DcR1 (residues Ϫ63 to 217) was cloned into mammalian double expression vector without any tag (8). Flag-tagged Fas was constructed in pcDNA3 as described (29). All constructs were verified by sequence analysis prior to transfection. HEK293 human embryonic kidney cells were transiently transfected with the expression plasmids using the LipofectAMINE (Life Technologies, Inc.) method. Cell lysates and Western blotting were done as described (30). The level of receptor expression was measured by flow cytometry using biotin-labeled affinity purified goat antibodies to the TRAIL receptor (R&D Systems Inc., Minneapolis, MN) with streptavidin-FITC (Sigma) as the secondary staining reagent. Both 0.1 and 1 g/ml antibodies produced similar results and so only data with 1 g/ml is reported in this study. The data was analyzed as relative mean fluoresence intensity representing the receptor density of the positive cells (% positive ϫ mean fluoresence intensity of positive cells) as described previously (31).
TRAIL-Biotin Binding to Cell Surface Receptors by Europium Method-Twenty-four hours after transfection with the expression plasmids, the HEK293 transfectants were harvested, washed three times with PBS, and plated in presaturated (with PBS) 96-well U-bottom polypropylene plates (Costar, Corning Inc.) at 4 ϫ 10 5 cells/well. After blocking with PBS ϩ 1% bovine serum albumin, the cells were incubated with 0, 1, and 10 g/ml TRAIL-biotin in PBS at 37°C for 1 h. After incubation, the plates were washed three times with PBS and incubated with europium-labeled streptavidin (100 dilution) (DELFIA, Wallac Oy, Fin-land) for 30 min at room temperature. The cells were washed three times with PBS and incubated with 100 l of enhancement solution (DELFIA, Wallac Oy, Finland) according to the manufacturer's instruction. A parallel experiment with alkaline phosphatase-labeled streptavidin, instead of europium, produced similar results.

Biochemical Characterization of Recombinant TRAIL Receptor-Ig Fusion
Proteins-In the present investigations, we used the CHO expression system to produce the soluble TRAIL receptors DR5, DR4, DcR1, and OPG as Ig-fusion proteins (19,20). The purified proteins were characterized by SDS-PAGE (Fig. 1), analytical size exclusion chromatography, matrix-assisted laser desorption ionization-mass spectrometry, and Nterminal sequence analysis. The key properties are summa-rized in Table I. The heterogeneous appearance of SDS-PAGE bands and the disparity between the observed molecular weights and those calculated from the amino acid sequences ( Table I) are indicative of glycoforms of the expressed proteins. N-terminal sequencing gave single sequences for DR5-Ig, DR4-Ig, DcR1-Ig, and OPG-Ig, thus confirming the identity and high purity of the recombinant proteins. The N termini for DR5, DR4, and DcR1 aligned well with each other but less well with OPG. In conjunction with the overall identity for these proteins, where DR5 shows 59, 58, and 52% identity within the first 126 amino acids of the extracellular domains of DR5, DcR1, and DcR2, respectively, this data provides further evidence for the closer evolutionary relationship of these 4 TRAIL receptors. In contrast, OPG, which has a different N terminus and shows only 26% identity to DR5 within this region (see Table II), appears to have a more distant relationship to the other TRAIL receptors.
Biochemical Characterization of Recombinant TRAIL from P. pastoris-N-terminal sequencing of purified TRAIL confirmed the predominant presence of His 6 -tagged TRAIL, although two minor species (about 15%) with truncations within the His 6 tag were also present in the purified material. The purity of TRAIL expressed in Pichia was superior to that produced using the CHO expression system (19), as determined by SDS-PAGE and N-terminal sequence analyses. The purified TRAIL from P. pastoris was found to be a stable trimer, down to at least micromolar concentrations, by analytical ultracentrifugation, as was also shown by crystallography of Escherichia coli produced TRAIL (32,33).

DR5 Shows Highest Reactivity for TRAIL in Competition
ELISAs-Biotinylated TRAIL binds, in a concentration dependent manner, to DR5-Ig immobilized on microtiter plates ( Fig. 2A). Binding was more efficient at 37°C than at 4 or 25°C (Fig. 2B). Immobilized OPG-Ig also bound to biotinylated TRAIL but, under the same conditions, OPG-Ig was less efficient than DR5-Ig.
DR5-Ig in solution was very efficient at competing with TRAIL binding to immobilized DR5-Ig or OPG-Ig in competitive ELISAs, both at 4 and 37°C (Fig. 3). In contrast, OPG-Ig was unable to compete with TRAIL binding to immobilized DR5-Ig or OPG-Ig at 37°C, but was able to compete at 4°C, although less efficiently than DR5-Ig (Fig. 3). This indicates that at 37°C, OPG-Ig has a weaker affinity for TRAIL than does DR5-Ig. DcR1-Ig and DR4-Ig partially blocked TRAIL binding to immobilized DR5-Ig, and required higher solution concentrations when compared with inhibition by DR5-Ig (Fig.  4). However, they were almost equivalent to DR5-Ig at blocking TRAIL binding to immobilized OPG-Ig, suggesting that DcR1-Ig and DR4-Ig have intermediate affinities for TRAIL when compared with the high affinity of DR5-Ig and low affinity of OPG-Ig.
DR5 Is a High Affinity Receptor for TRAIL-The TRAIL affinities of the receptor-Ig constructs in solution were measured directly at 37°C by isothermal titration calorimetry (26). Fig. 5 shows an example of titration calorimetry data for the titration of TRAIL with DR4-Ig. The equilibrium K D of 70 nM measured from the data in Fig. 5 is much weaker than values measured previously at low temperatures (12,13,15,18,19). However, the data in Fig. 5 also reveal that the binding enthalpy change for DR4-Ig is very large in comparison to typical protein-protein interactions (34). Consequently, the binding affinity is a strong function of temperature, according to the Gibbs-Helmholz relationship which is described elsewhere (35). TRAIL binding experiments were also conducted with DR5-Ig, DcR1-Ig, and OPG-Ig (Fig. 6) from which the K D and binding enthalpy change values were determined (Table II). Of the four TRAIL receptors evaluated, DR5-Ig very distinctly has the highest affinity at 37°C, as can be judged by the steep transition of the binding data in Fig. 6. The TRAIL K D of DR5-Ig is at least as tight as 2 nM at 37°C, and is more than 30-, 100-, and 200-fold tighter than DR4-Ig, DcR1-Ig, and OPG-Ig, respectively. The observed K D (2 nM) is very close to the instrumental lower limit and the true affinity could be tighter. Similar high affinity binding of DR5-Ig was also found for TRAIL expressed in CHO cells (K D Յ 1 nM), indicating that the high affinity binding to DR5-Ig is not unique to P. pastoris expressed TRAIL, and that TRAIL expressed in mammalian and yeast expression systems retains the appropriate folding for receptor recognition.
The temperature dependence of binding affinities for protein-protein interactions can be highly variable. The magnitude and direction of the change in affinity with temperature is governed by the binding enthalpy change (35). A key strength of isothermal titration calorimetry is that it measures directly the binding enthalpy change, along with the binding affinity, at a given temperature. In the course of measuring the TRAIL affinities of the various receptors we found that their binding enthalpy changes were considerably variable. Observed binding enthalpy changes ranged from Ϫ35 to Ϫ96 kcal/mol (Table  II), reflecting 61 kilocalories difference in binding enthalpy. The affinities of some of these interactions are therefore much stronger functions of temperature than others. Consequently, the rank order of TRAIL affinities of the receptors at low temperatures (4 or 25°C) is very different than the rank order at 37°C. While all the receptors have similar TRAIL affinities at 25°C, the affinities of DR4-Ig and OPG-Ig are strong functions of temperature, and they are greatly weakened at 37°C. The affinity of DcR1-Ig is only moderately weaker at 37°C, and the affinity of DR5-Ig exhibits the shallowest temperature dependence.
Preferentially Enhanced Binding of TRAIL to Cell Surface DR5-To determine whether the high affinity binding of TRAIL seen with the recombinantly expressed soluble DR5 also translates to enhanced binding on the cell surface, fulllength DR5, DR4, DcR1, or DcR2 were transfected for cell surface expression on HEK293 cells. In the case of the death receptors DR5 and DR4, to prevent triggering of downstream apoptotic events upon binding of TRAIL, death domain truncated versions of DR5 (DR5⌬DD) and DR4 (DR4⌬DD) were also transfected into these cells. OPG was not included in these experiments because it is a naturally secreted protein lacking a member anchor. In these experiments, Fas was used as a control for transfection and TRAIL binding. The level of surface expression of the receptors was evaluated by flow cytometry using antibodies specific for each receptor. The transfections resulted in very similar levels of expression of the various receptors. Roughly 50% (ranges from 47 to 51%) of all the transfectants were positive for the relevant transfected receptor with broad intensity of staining for each receptor ranging over 3 orders of magnitude on the intensity scale. The level of expression for each receptor (represented in Fig. 7A as relative mean intensity) was comparable for each transfected receptor. Similarly comparable expression levels were also observed by Western blot analysis.
b Value is at most 2 nM but may be tighter and cannot be determined with certainty by the ITC method for higher affinity interactions.
FIG. 2. Binding of biotinylated TRAIL to microtiter plate immobilized DR5-Ig. A, DR5-Ig was immobilized at different concentrations and TRAIL-biotin was titrated in for binding, as described under "Materials and Methods." B, comparison of TRAIL binding to immobilized DR5-Ig (20 ng/well) at different temperatures.
higher binding to the DR5 (DR5⌬DD) than to the DR4 (DR4⌬DD), DcR1, or DcR2 transfectants. Similar results were also obtained with the full-length (death domain untruncated) DR5 and DR4 (not shown). This assay has sufficient sensitivity to distinguish between the large (Ͼ70-fold) affinity differences observed in the ITC determinations between DR5 and the other TRAIL receptors (DR4 and DcR2), but not the smaller (2-5fold) differences observed (or expected in the case of DcR2) among the lower affinity receptors. No binding of TRAIL was observed in the Fas-transfected cells. Thus, this data confirms that similar rank order of affinities govern the binding of TRAIL to its receptors on the cell surface.
Two groups have recently described the crystal structure of the TRAIL-DR5 complex (36,37). As had been predicted from the known TNF-␤-TNFR-I complex, the trimeric TRAIL binds three DR5 molecules, one each in the cleft formed at the interface of the TRAIL subunits. DR5 has one truncated, and two full-length pseudorepeats of cysteine-rich domains (CRD1, CRD2, and CRD3, respectively, with CRD3 as the membrane proximal domain). In its interaction with TRAIL, CRD2 and CRD3 straddle the TRAIL interface thus forming two major contact surfaces whereby patch A within CRD3 primarily contributes to the specificity of the interaction with TRAIL, while patch B within CRD2 contributes to more general hydrophobic interactions (36). Alignment of the TRAIL receptors shows, with the exception of the less related OPG, there is a high degree of sequence conservation within the region represented by patch A, which drives the specificity of the interactions with TRAIL. Thus, these receptors have, for the most part, retained identical key contact residues or have conservative or semicon-servative substitutions in this region. Although there are some differences within this region among the 4 closely related receptors, interestingly, the region represented by patch B within CRD2 displays the most divergence between DR5 on one side and DR4, DcR1, and DcR2 on the other. Within the putative TRAIL recognition regions, DR4 and DcR1 are more related to each other than either one is to DR5. Overall within the extracellular region of these receptors, DcR1 and DcR2 are very closely related with 93 identities within the first 126 residues, compared with 65/126 and 73/126 for DcR1 and DcR2, respectively, and DR5. Thus, based on sequence identity or similarity, DcR1, DcR2, and DR4 would be predicted to have closely related recognition elements in binding to TRAIL. Therefore, although precise affinity determination by the ITC method was not conducted for TRAIL binding to DcR2, based on the cell binding experiments, and the sequence identities and presumed structural similarities with the lower affinity receptors, DcR2 would be predicted to be closer to DcR1 and DR4 than to DR5.
In addition to affinity differences among the TRAIL receptors discussed in this report, other mechanisms may contribute to the functional differences of the receptors. These include their unique expression patterns in different tissues and cells types, receptor regulation in response to activation or cell differentiation, the relative receptor densities at the cell surface, ability to recognize other ligands (as is the case for OPG which binds to RANKL with high affinity, and which may also be the case for the other TRAIL receptors, with the possibility that there may be other unidentified ligands), and the important differences in their cytoplasmic domains which contribute to their ability or inability to couple to different signaling pathways. The relative contribution of these factors remains to be elucidated.
FIG. 6. Isothermal titration calorimetry data for DR5-Ig, DR4-Ig, DcR1-Ig, and OPG-Ig binding to TRAIL as labeled. TRAIL binding heats, normalized per mole of receptor-Ig injected, are shown versus molar ratio of receptor-Ig added per mole of TRAIL. Best-fit curves shown yield binding enthalpy changes and affinities in Table II. FIG. 7. A represents expression of TRAIL receptors on HEK293 transient transfectants measured by flow cytometry. Data is represented as relative mean fluorescence intensity (% positive cells X mean fluorescence intensity of those cells). B represents binding of biotinylated TRAIL to HEK293 cells transfected with TRAIL receptors: DR5⌬DD, DR4⌬DD, DcR1, and DcR2. In addition to the DD truncated death receptors, similar data was also obtained for the full-length transfectants of DR5 and DR4 (not shown). Fas transfected and unstransfected cells were used as controls for TRAIL binding and transfection, respectively.