Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Fas ligand (FasL¹) and Fas receptor are members of the tumor necrosis factor receptor superfamily that play an important role in regulating apoptosis in normal physiology (Armitage, 1994; Smith et al., 1994; Torcia et al., 1996). DcR3 (TR6) is a novel member of the TNF receptor superfamily, and in agreement with previous literature reports (Kramer et al., 1994, Yu et al., 1999; Zhang et al., 2001), we have demonstrated that DcR3 binds both FasL and LIGHT and blocks both sFasL-mediated apoptosis of human Jurkat cells and the LIGHT-mediated inhibition of HT-29 cell growth, in a dose-dependent fashion (data not shown). Subsequent to these observations DcR3 has been shown to also bind TL1A, another member of this family, which appears to play a role in apoptosis and immune regulation (Migone et al., 2002).

While Fas/FasL-mediated apoptosis is believed to be important in the maintenance of normal physiological processes, excessive or inappropriate apoptosis induced by the Fas/FasL system has been proposed to be key factor in the pathogenesis of autoimmune disease, cancer, and diseases of the liver, lung, kidney, and central nervous systems (Maggi, 1998; Patel et al., 1998; O'Connell et al., 2001). The application of DcR3 to disrupt FasL or LIGHT-mediated signaling pathways may thus provide a novel pharmacological approach in the treatment of a number of human diseases in which FasL-induced apoptosis plays a significant role.

The processed, soluble form of DcR3 is a 271 amino acid polypeptide having one N-linked glycosylation site. We have previously reported that DcR3 is susceptible to proteolytic cleavage in vivo and in vitro at the Arg218-Ala219 peptide bond to yield a primary metabolite DcR3(1-218) (Wroblewski et al., 2003). DcR3(1-218) was able to bind LIGHT and block its activities in vitro but unlike the parent molecule, no longer bound sFasL or inhibited FasL mediated apoptosis in vitro. In an effort to create a molecule that was less susceptible to proteolysis, the primary sequence of DcR3 was engineered by changing the arginine at position 218 to a glutamine. The analog, DcR3(R218Q), which we have termed FLINT (LY498919), was more stable to proteolytic degradation in vitro and maintained its ability to bind both FasL and LIGHT (Wroblewski et al., 2003). FLINT also retained its capacity to block both sFasL-mediated apoptosis of human Jurkat cells and the LIGHT-mediated inhibition of HT-29 cell growth, in a dose-dependent fashion.

The efficacy of DcR3 administered as a therapeutic protein could be limited by the proteolysis occurring at position 218 since the metabolic fragment resulting from this cleavage no longer binds sFasL or inhibits sFasL-mediated apoptosis. Theoretically, an analog of DcR3, which was resistant to proteolysis at this site, could have more favorable pharmacologic and pharmacokinetic properties that may translate to the need for a lower dose to obtain a desired effect. The current study examined this possibility by comparing the pharmacokinetics and bioavailability of FLINT and DcR3 in cynomolgus monkeys and mice.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Production and Purification of DcR3 and FLINT. Production of DcR3 and FLINT was carried out by expressing these proteins in both AV12 RGT cells and DG44 cells media containing DcR3 or FLINT was adjusted to 0.1% CHAPS and concentrated in an Amicon ProFlux M12 tangential filtration system to 350 ml. The concentrated media was adjusted to pH 6.0 and passed over a SP Sepharose Fast Flow (50 ml; Pharmacia, Peapack, NJ), followed by fractionation of the DcR3 or FLINT containing pool by reverse phase-HPLC on a Vydac C4 column (1 × 15 cm). Fractions containing DcR3 were pooled, concentrated under vacuum, and applied to a Superdex 75 (Hi Load 16/60, Pharmacia) size exclusion column. Fractions containing DcR3 or FLINT were analyzed by SDS-PAGE and found to be greater than 95% pure. The N-terminal sequence of these proteins were confirmed on the purified polypeptide.

Production and Purification of the DcR3(1-218). DcR3 purified from AV12 RGT 18 cells was incubated with thrombin at an enzyme to substrate ratio of 1:100 (w/w) for 3 h at room temperature. The reaction solution was then dialyzed against 20 mM MOPS, 0.1% CHAPS, pH 6.5, and fractionated on a SP Sepharose column at a flow rate of 1 ml/min. The bound metabolite (amino acids 1 to 218) was eluted and fractions were analyzed by SDS-PAGE and mass spectrometry. Fractions containing only the DcR3(1-218) were pooled and concentrated in Millipore Ultrafree centrifugal filter (Millipore Corporation, Bedford, MA). The concentrated DcR3(1-218) was again analyzed by SDS-PAGE and mass spectrometry to assess purity. The N-terminal of the DcR3 metabolite was also confirmed by Edman sequencing.

Animal Experiments. DcR3 or FLINT was administered as a single bolus i.v. or single s.c. injection in phosphate-buffered saline (pH 7.4).

Primate Studies. FLINT or DcR3, derived from CHO-cells, was administered as a single intravenous bolus, via the saphenous vein, or single subcutaneous injection. Two male cynomolgus monkeys (Maccaca fasicularis, 2.5-3.5 kg) were used for each compound by each route of administration. Compounds were administered at a dose level of 0.5 or 1 mg/kg for the intravenous and subcutaneous routes, respectively. Blood samples (1 ml) were collected from the femoral vein into tubes containing potassium EDTA as anticoagulant prior to dosing and at 0.25, 0.5, 1, 2, 4, 8, 12, 24, and 30 h after s.c. administration and 0.083, 0.25, 0.5, 1, 3, 6, 12, 24, and 30 h after i.v. administration.

Mouse Studies. FLINT or DcR3, derived from AV12 cells, was administered as a single i.v. bolus, via the tail vein, or single subcutaneous injection to male CD-1 mice (25-30 g) at a dose level of 0.5 mg/kg. Blood samples were collected by cardiac puncture into tubes containing potassium EDTA as anticoagulant from two animals per treatment group per time point at 0.25, 0.5, 1, 3, 6, 10 h after administration.

Bioanalytical Assays. Plasma samples were analyzed for concentrations of DcR3, FLINT, or DcR3(1-218) using sandwich ELISA methods employing affinity purified rabbit polyclonal anti-DcR3 antibodies.

Specific ELISA. This method was used to selectively recognize intact DcR3 or FLINT with minimal cross-reactive interference by the fragment, DcR3(1-218). Polyclonal antisera were generated in rabbits using DcR3 in a standard hyperimmunization protocol. In this method, protein A purified polyclonal antibodies were affinity purified over a column generated using a synthetically prepared peptide corresponding to the C-terminal 50 amino acids in DcR3. The flow through from this column recognizes the N-terminal portion of DcR3 or FLINT and was used as the coating antibody (WQR-066A). The antibodies eluted from this column, which recognize the C-terminal 52 amino acids of FLINT or DcR3 were biotinylated and used as the sandwich antibody in this assay (WQR-066B). In brief, 0.1 ml of coating antibody (2 µg/ml) was added to the wells of an Immunlon 4 microtiter plate and allowed to bind at 4°C overnight. After washing and blocking steps, standards and samples were added to the wells in a volume of 0.05 ml and incubated for 1 h at room temperature. Standards (DcR3 or FLINT) were prepared in either cynomolgus monkey or CD-1 mouse plasma (EDTA), study samples were diluted in the appropriate matrix. After washing, biotinylated antibodies were added and incubated for 1 h at room temperature, followed by an additional 1-h incubation with conjugated streptavidin. The standard curve range was from 0.125 to 8 ng/ml. The lower limit of quantitation was defined as 0.5 ng/ml. As formatted, measurement in this ELISA relies on the presence of full-length DcR3, FLINT, or molecules with minimal degradation at the N- or C-terminal regions and does not cross-react significantly with DcR3(1-218) (Fig. 1A).

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Selectivity of ELISA methods used for plasma measurement of DcR3, FLINT, and DcR3(1-218). Panel A, assay for full-length DcR3 or FLINT, showing minimal cross-reactivity with DcR3(1-218). Panel B, assay for DcR3(1-218), showing minimal cross-reactivity with intact DcR3. Data are the mean ± standard deviation of responses from three standard curves.

Metabolite ELISA. This assay used the same coating antibody (WQR-066A) as used in the specific ELISA. For the detection reagent, polyclonal antisera were generated in rabbits using the C-terminal 100 amino acids of DcR3 in a standard hyperimmunization protocol. The sera were protein A purified and antibodies affinity purified over a column prepared using a synthetically produced peptide corresponding to the C-terminal 50 amino acids in DcR3. Interestingly, this batch of affinity purified antibodies had an unanticipated selectivity recognizing DcR3(1-218) but not full-length DcR3 by immunoblotting (not shown). The reason for this is not clear but may be related to unique conformational epitopes generated with the fragment immunogen compared with the intact protein. When this affinity purified antibody preparation was biotinylated and used in a sandwich ELISA, the assay showed selectivity toward DcR3(1-218) and had minimal reactivity with intact DcR3 or FLINT (Fig. 1B). This method was applied to selectively recognize DcR3(1-218) in plasma after administration of DcR3 or FLINT in vivo. DcR3(1-218), produced as previously described, was used as the standard in this assay with a range from 0.16 to 10 ng/ml. Standards were prepared in either cynomolgus monkey or CD-1 mouse plasma (EDTA); study samples were diluted in the appropriate matrix. The lower limit of quantitation was defined as 0.5 ng/ml.

Pharmacokinetics. Pharmacokinetic parameters were calculated using WinNonlin Professional version 3.1 software package (Pharsight Corporation, Mountain View, CA). Intravenous plasma concentration data were modeled using a two compartment model with a first order elimination rate constant and a weighting function of Y². The parameters that were calculated include maximum plasma concentrations (C_max), AUC_0-, elimination half-life (t1/2;), clearance (CL), and steady-state volume of distribution (V_ss). Subcutaneous plasma concentration data were described using a noncompartment model to calculate the basic pharmacokinetic parameters of C_max, T_max, AUC_0-, and t1/2. Absolute bioavailability was calculated using the following relationship: (mean AUC_s.c.0-/mean AUC_i.v.0-) × (dose i.v./dose s.c.) × 100.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Pharmacokinetics in Primates. After i.v. administration to cynomolgus monkeys, FLINT and DcR3 were cleared from the plasma in a bi-phasic manner. The clearance of both compounds was characterized by a rapid distribution phase (Fig. 2), with approximately 20% of the 5-min levels remaining 1 h after administration. The clearance of FLINT from the plasma was approximately 2-fold slower after i.v. administration compared with DcR3 (Table 1; Fig. 2). This is reflected in the approximately 2-fold greater area under the plasma concentration curve (AUC_0-), or overall exposure to FLINT, compared with DcR3 at equivalent doses (Table 1; Fig. 2). The terminal phase half-life was slightly longer for FLINT than for DcR3 (12.9 versus 9 h, respectively).

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Kinetics of DcR3 and FLINT in cynomolgus monkeys after intravenous administration. Compounds were administered intravenously at a dose of 0.5 mg/kg. Immunoreactivity was determined using an ELISA that selectively recognizes full-length DcR3 and FLINT. Data are from individual animals.

0-

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Kinetics of DcR3 and FLINT in cynomolgus monkeys after subcutaneous administration. Compounds were administered as a single 1 mg/kg subcutaneous dose. Immunoreactivity was determined using an ELISA, which selectively recognizes the full-length molecules. Data are from individual animals.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Kinetics of DcR3(1-218) immunoreactivity after subcutaneous administration of DcR3 or FLINT to cynomolgus monkeys. DcR3 or FLINT were administered as a single 1 mg/kg subcutaneous dose. Immunoreactivity was determined using an ELISA, which selectively recognizes the N-terminal 218 amino acids of DcR3. Data are from individual animals.

Pharmacokinetics in Mice. FLINT and DcR3 were also very rapidly cleared after i.v. administration to CD-1 mice, however, the somewhat more rapid clearance of DcR3 was not evident in this species (Table 1; Fig. 5). The terminal phase half-life was longer for FLINT than for DcR3, (3.1 versus 1.2 h, respectively).

View larger version (14K): [in this window] [in a new window]   Fig. 5.   Kinetics of DcR3 and FLINT in CD-1 mice after intravenous administration. Compounds were administered intravenously at a dose of 0.5 mg/kg. Immunoreactivity was determined using an ELISA that selectively recognizes full-length DcR3 and FLINT. Data represent the mean from two animals per time point.

0-

View larger version (13K): [in this window] [in a new window]   Fig. 6.   Kinetics of DcR3 and FLINT in CD-1 mice after subcutaneous administration. Compounds were administered intravenously at a dose of 0.5 mg/kg. Immunoreactivity was determined using an ELISA that selectively recognizes full-length DcR3 and FLINT. Data represent the mean from two animals per time point.

View larger version (13K): [in this window] [in a new window]   Fig. 7.   Kinetics of DcR3(1-218) immunoreactivity after subcutaneous administration of DcR3 or FLINT to CD-1 mice. DcR3 or FLINT was administered as a single 0.5 mg/kg subcutaneous dose. Immunoreactivity was determined using an ELISA that selectively recognizes the N-terminal 218 amino acids of DcR3. Data represent the mean from two animals per time point.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

DcR3 binds to specifically to FasL and LIGHT and inhibits the activities of both of these TNF receptor family ligands. These ligand systems have been implicated in modulating cell survival, cell proliferation, and cell death (Armitage, 1994; Yu et al., 1999; Zhang et al., 2001). As such, regulation of the expression of mediator molecules such as the soluble forms of FasL, Fas, and antagonistic decoy receptors such as DcR3 may be involved in the tight control of these biological functions, which are integral to both normal physiology and pathophysiology. The application of DcR3 as a pharmacological agent designed to modulate the biological events triggered by either FasL or LIGHT may have therapeutic benefit in human diseases in which the pathogenesis is believed to be related to excessive or inappropriate apoptosis or inflammation (Maggi, 1998).

In an earlier report, we demonstrated that DcR3 is proteolytically cleaved in a specific manner to yield a major metabolic fragment, DcR3(1-218), which does not inhibit sFasL-mediated apoptosis of human Jurkat cells but maintains its ability to inhibit the activities of LIGHT (Wroblewski et al., 2003). In vitro and in vivo studies indicated that the R218-A219 bond was uniquely sensitive to proteolysis and is one of the primary sites of degradation of DcR3 when administered in vivo by the subcutaneous route. Since the proteolysis of DcR3 resulted in a fragment, DcR3(1-218), which did not bind FasL, cleavage at position R218 would be considered detrimental to the pharmacologic potency of DcR3 if efficacy required the interruption of the interaction of both FasL and LIGHT with their cognate receptors or relied upon inhibition of events elicited specifically by FasL. In the present study, we demonstrated that after s.c. administration of DcR3, exposure in both rodents and primates was accounted for almost exclusively by the DcR3(1-218) fragment. Modification of DcR3 to FLINT resulted in a 6-fold increase in exposure after s.c. administration, accounted for by the fully active intact form molecule. This benefit is most likely a result of the stability afforded to the molecule by the R218Q mutation. The increase in the absolute s.c. bioavailability of FLINT may translate to a requirement for a lower s.c. dose to obtain a therapeutic effect. Interestingly, while intravenous administration of a DcR3-Fc fusion protein provided survival benefit in a murine model of acute fulminant hepatic apoptosis, induced specifically by FasL; no benefit was observed when DcR3-Fc was administered by the subcutaneous route even at a 4-fold higher dose (Connolly et al., 2001). It is likely that the lack of subcutaneous efficacy observed in this model was related to the s.c. proteolysis of DcR3 as described in this report, and is consistent with our observation that the proteolytic fragment does not bind FasL or inhibit FasL mediated apoptosis.

The serine protease responsible for the proteolysis of DcR3 at position R218 has not been identified but appears to be acting at the s.c. site of administration. We have not been able to clearly demonstrate the specific proteolysis of DcR3 after intravenous administration to normal animals, although we have demonstrated that the R218-A219 bond was specifically cleaved by purified bovine thrombin in vitro (Wroblewski et al., 2003). This observation suggests that cleavage and inactivation of DcR3 could occur after i.v. administration via the expression/release of proteases in conditions of unregulated inflammatory response, which are potential therapeutic indications for intervention at the Fas-FasL system (Papathanassoglou et al., 2000; Dhainaut et al., 2001; Esmon, 2001).

The targeted modification of a single amino acid residue has been an approach successfully applied to generate stable peptide analogs of growth hormone releasing hormone and glucagon-like intestinal peptide, the biological and functional half lives of which are negatively affected after cleavage by dipeptidyl aminopeptidase (Frohman et al., 1989; Burcelin et al., 1999; Sharpe and DeMeester, 2001). Knowing the specificity and nature of the exopeptidase involved, combined with the relatively simple structure of these smaller peptides, makes the strategy for directed engineering for increased stability somewhat more obvious. On the other hand, successful identification of a single cleavage event important to the in vivo metabolic stability of a large glycoprotein such as DcR3 is not intuitively obvious. Therapeutically administered proteins are often extensively degraded either locally or at peripheral organs of clearance leading to inactivation via hydrolysis of a multitude of peptide bonds. DcR3 is also extensively degraded upon s.c. administration, but we were able to identify a highly sensitive cleavage site that appears to be a major initial event in the degradation of this molecule, at least when injected into the subcutaneous space. Characterizing the mechanism of the limited proteolysis of DcR3 involved the application of a number of analytical approaches along with a correlation of data obtained in vivo with that from in vitro models (Wroblewski et al., 1991, 1993a, 2003). Physiologically, limited proteolysis may lead to inactivation of the parent protein, but has been shown to yield variants having altered receptor affinities and enhanced or selective biological activities (Korc and Finman, 1989; Powers and Hatala, 1990; Wroblewski et al., 1993b; Angelloz-Nicoud et al., 1998; Bergsten et al., 2001; Sharpe and DeMeester, 2001). In addition to the biological relevance, characterizing the preferred cleavage bonds and the sequence of proteolytic events involved in protein degradation may help in the engineering of molecules having improved pharmacological properties (Bryant et al., 1996; Authier et al., 1999).

In conclusion, the findings in this report indicate that the molecularly engineered version of DcR3, FLINT (LY498919), is more stable to proteolytic degradation upon s.c. administration than DcR3 itself in both rodents and primates. The stability afforded to FLINT results in higher bioavailability from the s.c. site, which may translate to lower doses and a superior pharmacological effect in the therapeutic intervention of diseases in which the pathogenesis is linked to FasL mediated apoptotic or inflammatory events.