Design and analysis of an engineered human interleukin-10 monomer.

A monomeric form of human interleukin 10 (IL-10M1) has been engineered for detailed structure-function studies on IL-10 and its receptor complexes. Wild type IL-10 (wtIL-10) is a domain swapped dimer whose structural integrity depends on the intertwining of two peptide chains. wtIL-10 was converted to a monomeric isomer by inserting 6 amino acids into the loop connecting the swapped secondary structural elements. Characterization of IL-10M1 by mass spectroscopy, size exclusion chromatography, cross-linking, and circular dichroism shows that IL-10M1 is a stable alpha-helical monomer at physiological pH whose three-dimensional structure closely resembles one domain of wtIL-10. As previously reported, incubation of wtIL-10 with a soluble form of the IL-10Ralpha (sIL-10Ralpha) generates a complex that consists of 2 wtIL-10 molecules and 4 sIL-10Ralphas. In contrast, IL-10M1 forms a 1:1 complex with the sIL-10Ralpha. Characterization of the interaction using isothermal titration calorimetry confirmed the 1:1 stoichiometry and yielded a dissociation constant of 30 nm with an apparent binding enthalpy of -12.2 kcal/mol. Despite forming a 1:1 complex, IL-10M1 is biologically active in cellular proliferation assays. These results indicate that the 1:1 interaction between IL-10M1 and IL-10Ralpha is sufficient for recruiting the signal transducing receptor chain (IL-10Rbeta) into the signaling complex and eliciting IL-10 cellular responses.

Interleukin-10 (IL-10) 1 is a pleotropic cytokine that inhibits cell-mediated immune responses while enhancing humoral immunity (1,2). The immunosuppressive properties of IL-10 are largely derived from its ability to inactivate macrophages by suppressing the synthesis of pro-inflammatory cytokines (including IL-1, IL-6, IFN-␥, and tumor necrosis factor-␣) and by inhibiting the expression of cell surface molecules involved in antigen presentation and co-stimulation (3)(4)(5). IL-10 is also a powerful growth and differentiation factor for B-lymphocytes, mast cells, and thymocytes (6). These immunostimulatory functions may have negative effects in some instances, as IL-10 has been implicated as an autocrine growth factor in certain B-cell malignancies (7,8). Thus, antagonists of IL-10 may be useful in treating B-cell lymphomas as well as infections where cellmediated immune responses are suppressed by IL-10 (9).
Cellular responses to IL-10 require at least two cell surface receptors, IL-10R␣ and IL-10R␤ (10 -12). Both receptors are members of the class II cytokine receptor family (13,14). The extracellular domains of the receptors are responsible for binding to IL-10. The intracellular domains of IL-10R␣ and IL-10R␤ are associated with Jak1 and Tyk-2 kinases, respectively (15). Kinase activation and subsequent IL-10-induced biological activities are dependent on the formation of a complex between IL-10, IL-10R␣, and IL-10R␤ (11). It is believed that IL-10R␣ acts as the primary high affinity receptor for IL-10 and that signaling results from the recognition of this initial complex by IL-10R␤.
Characterization of the interaction between IL-10 and the IL-10R␣ has been facilitated by the production of large quantities of the extracellular domain of IL-10R␣ (sIL-10R␣). The apparent dissociation constant between IL-10 and immobilized sIL-10R␣ is approximately 500 pM (16). Since IL-10, like IFN-␥, contains 2 identical domains it was expected to bind 2 sIL-10R␣ molecules as observed in the crystal structure of the IFN-␥ receptor complex (17). However, incubation of the sIL-10R␣ with IL-10 generates a stable complex that consists of 2 IL-10 molecules and 4 sIL-10R␣s (16). The stoichiometry and affinity of the sIL-10R␤ for this complex as well as the significance of the 2:4 IL-10⅐sIL-10R␣ complex in IL-10 signaling is currently unknown.
The crystal structure of IL-10 revealed a symmetric homodimer composed of two ␣-helical domains oriented at 90°to one another (18,19). The structural integrity of each domain is dependent on the intertwining of ␣-helices from each peptide chain such that the first four helices of one chain (A-D) associate with the last two helices (E and F) of the other. The domains of IL-10 share structural similarity with the monomeric type-I (IFN-␣, IFN-␤, and IFN-) and dimeric type-II interferons (IFN-␥) that together form the class II cytokine family (20 -23).
The IL-10 dimer is thought to be the result of an evolutionary mechanism of protein oligomerization often referred to as 3D domain swapping (24). This suggests that IL-10 evolved from a monomeric protein by exchanging structural domains (␣-heli-ces E and F for IL-10) with another monomer to create the dimer. Stability studies with IL-10 have established that significant amounts of a monomeric form of IL-10 can be generated at low protein concentrations, increased temperature (including 37°C), or under extreme conditions such as low pH or high concentrations of guanidine HCl (25). However, under physiological conditions the monomeric form of IL-10 is not stable and does not have the same domain structure as the IL-10 dimer (25). Furthermore, biological assays of the denatured samples show a linear correlation between the percent dimer present and IL-10 biological response suggesting that the dimer is the active species in signaling.
Since IL-10 is likely to have evolved from a monomeric species, its quaternary structure must be very important for generating IL-10-induced cellular activities. To test this hypothesis, we engineered an IL-10 monomer (IL-10M1) by inserting six amino acids (GGGSGG) into the inter-domain linker region of IL-10 between residues Asn 116 and Lys 117 . Our studies show that IL-10M1 is stable under native conditions (pH 7.0, 25°C) and structurally similar to one domain of IL-10. IL-10M1 binds to sIL-10R␣ with a 1:1 stoichiometry, representing one-fourth of the 2 IL-10⅐4 sIL-10R␣ complex observed for the IL-10 dimer. Surprisingly, IL-10M1 was shown to be biologically active and was unable to antagonize IL-10 in cellular proliferation assays. Our studies show that IL-10M1 is suitable for x-ray crystallographic studies of 1:1 IL-10 receptor complexes, protein-folding studies, and for defining the energetic contribution of each residue in the IL-10/sIL-10R␣ binding interface.
IL-10M1 was engineered using the previously determined crystal structure of wtIL-10 (Protein Database entry 1INR). Possible insertion peptide sequences were modeled using the computer graphics program CHAIN (26). The expression plasmid for IL-10M1 was generated from the ligation of two DNA fragments into pET-32 to generate pET32-IL-10M1. DNA fragment 1 was PCR-amplified from the 5Ј primer 5Ј-AAAGAGTTCCTCGACGCTAACCTG-3Ј and the 3Ј primer 5Ј-ATTG-GATCCACCACCGTTTTCACAGGGAAGAAATCG-3Ј and digested with NdeI and BamHI. It codes for wtIL-10 residues 1-116 and IP residues GGS. DNA fragment 2 was PCR-amplified from the 5Ј primer 5Ј-AT-TGGATCCGGTGGTAAGAGCAAGAGCAAGGCCGTGGA-3Ј and the 3Ј primer 5Ј-CCGCTCGAGTCAGTTTCGTATCTTCATTGT-3Ј and digested with BamHI and XhoI. It codes for IP residues SGG and wtIL-10 residues 117-160. The sequences of pET-32-wtIL-10 and pET-32-IL-10M1 were verified by automated DNA sequencing. pET-32-wtIL-10 and pET-32-IL-10M1 were transformed into the Escherichia coli strain BL21 (DE3 lysogen) for expression under the control of the T7 promoter.
Expression and Purification of wtIL-10 and IL-10M1-Cells containing the pET-32-wtIL-10 or pET-32-IL-10M1 plasmids were grown in Luria Broth media to an optical density of 0.6 and then induced with 1 mM isopropyl-␤-D-thiogalactoside for 3 h. The cells were collected by centrifugation, and the inclusion bodies were purified by sonication in 100 mM Tris-HCl, 100 mM NaCl, 5 mM EDTA, 0.1 M phenylmethylsulfonyl fluoride, pH 8.0. IL-10M1 and wtIL-10 were refolded using conditions previously reported for wtIL-10 (19). Briefly, inclusion bodies were solubilized in 6 M guanidine HCl containing 5 mM DTT. Refolding was initiated by a 10-fold dilution of the solubilized inclusion bodies into a solution containing 50 mM Tris-HCl, pH 8.0, 50 mM NaCl, 5 mM EDTA, 2 mM reduced glutathione, and 0.2 M oxidized glutathione. Purification of the refolded material was achieved by sequential ion exchange chromatography on HS-20 and HQ-10 columns (PE Biosystems) followed by an additional gel filtration procedure. Each purification step was analyzed by SDS-PAGE to assist in pooling fractions.
Protein Concentrations-Protein concentrations for all experiments were determined from the UV absorbance at 280 nm to obtain mg/ml using the appropriate ⑀ 0.1% values for wtIL-10 and sIL-10R␣. An ⑀ 0.1% for wtIL-10 and IL-10M1 of 0.47 was used as previously reported (27). The calculated ⑀ 0.1% value for sIL-10R␣ is 1.6.
MALDI-TOF Mass Spectroscopy-Mass analysis was carried out on a Voyager Elite mass spectrometer with delayed extraction technology (Perspective Biosystems, Framingham, MA) operating in the positive mode. The acceleration voltage was set at 25 kV, and 50 -100 laser shots were summed. Samples were mixed 1:10 with sinapinic acid dissolved in acetonitrile, 0.1% trichloroacetic acid (1:1). Apomyoglobin was used as an internal standard Cross-linking-Cross-linking reactions were carried out in a volume of 30 l with 150 g/ml wtIL-10 or IL-10M1 in 20 mM sodium phosphate, pH 7.0. Samples were cross-linked in 1% glutaraldehyde (final concentration) for 12 min at room temperature. Control reactions received an equal volume of phosphate buffer. Reactions were quenched for 20 min by the addition of 0.5 M NaBH 4 in 0.1 M NaOH to 25 mM. The pH of the quenched reactions were adjusted to approximately pH 8.0 with the addition of 0.1 N HCl. The quenched reactions were loaded onto a 15% SDS-PAGE gel with molecular weight markers (Novagen) followed by silver staining.
Size Exclusion Chromatography-Receptor complexes were formed by incubating IL-10M1 or wtIL-10 with sIL-10R␣ at different molar ratios for 1 h at 4°C. The complexes were purified by gel filtration using two Superdex 200 (10 ϫ 300 mm) columns (Amersham Pharmacia Biotech) linked in tandem that were equilibrated in 20 mM Tris-HCl, 150 mM NaCl, pH 8.0, at a flow rate of 0.35 ml/min. 100-l volumes were injected onto the column. Apparent molecular weights are calculated using a standard curve of K av versus the log of the molecular weights of the gel filtration standards (Bio-Rad). The extracellular fragment of the IL-10R␣ (sIL-10R␣) for gel filtration chromatography and ITC experiments and neutralization studies was produced in Drosophila S2 cells (Invitrogen and SmithKline Beecham). 2 Circular Dichroism-Circular dichroism spectra were measured using an AVIV model 62DS spectropolarimeter. Reported spectra were averaged from three scans base line subtracted from the buffer spectrum taken in an identical manner. Data collection was made every 0.5 nm using a 1-s averaging time and a 1.5-nm bandwidth. Sample concentrations were 2 mg/ml in 20 mM sodium phosphate buffer, pH 7.0, in an 80-l sample cell with a 0.l-mm path length. Melting experiments were done at 0.04 mg/ml in 20 mM sodium phosphate, pH 7.0, in a 3.5-ml stirred sample cell, with a 1-cm path length. The temperature was changed by 0.5°increments to 95°C and monitored by a sample cell temperature probe. At each set temperature the signal at 222 nm was averaged for 10 s using and 1.5-nm bandwidth.
Isothermal Titration Calorimetry-Titration calorimetry was carried out using a VP-ITC calorimeter (Microcal, Inc., Northampton, MA) at 25°C. Prior to ITC analysis protein solutions were extensively dialyzed against 150 mM NaCl and 20 mM disodium/Pipes (disodium Pipes) adjusted to pH 7.1 with 10 N HCl. A 65 M solution of IL-10M1 was titrated into a 1.39-ml cell containing a 6.5 M solution of Drosophila S2-derived sIL-10R␣. After an initial injection of 0.5 l, which was not used in data fitting, 12 injections of 20.5 l each were made at 240-s intervals. A blank titration was also performed in which ligand was injected into buffer. The data were fit to a single binding site model using Origin (version 5.0; Microcal, Inc.) after subtracting a linear regression of the integrated peak areas from the blank run (28).
Proliferation Assays-IL-10M1 or wtIL-10 was diluted into assay medium consisting of RPMI 1640, 1% fetal bovine serum, 2 mM Glu-taMax-1, 100 units/ml penicillin, 100 g/ml streptomycin, 0.1 mM 2-mercaptoethanol. Samples were dispensed into 96-well microtiter plates (Becton Dickinson) in duplicate wells and serially diluted 3-fold across the plates. Ba8.1c1 or BaMr 29␣-1 cells were then added to the assay plates at a final density of 1.5 ϫ 10 4 cells/well. After incubation for 48 h at 37°C, 5% CO 2 , proliferation was assessed by measuring cell concentration via mitochondrial reduction of the tetrazolium salt MTS in the presence of 5% phenazine methosulfate over a 4-h incubation period. The colored reaction product formazan is soluble in the assay medium and can be measured spectrophotometrically at a wavelength of 490 nm with a reference wavelength of 570 nm. For determining specific activities, the concentration of cytokine that induced half-maximal response was defined as 1 unit/ml. Neutralization of wtIL-10 or IL-10M1 was performed by addition of various concentrations of sIL-10R␣.

RESULTS
Engineering of IL-10M1-The crystal structure of IL-10 revealed a homodimer comprised of two intertwined peptide chains that contain six ␣-helices labeled A-F (18,19). The two chains tightly associate with one another such that each domain of the dimer consists of helices A-D from one chain and helices E and F from the other. Helices D and E are connected by a 12-residue linker of which the final four residues (Glu 115 -Asn 116 -Lys 117 -Ser 118 ) extend across the dimer interface. Graphical analysis of wtIL-10 showed that without a large structural change, this linker was too short to allow helices E and F to insert into the hydrophobic cleft formed from helices A-D of the same chain (Fig. 1).
Computer modeling was used to identify the location and length of an insertion peptide (IP) that would allow the formation of a stable IL-10 monomer. These studies led to the design of IL-10M1 that contained a 6-residue insertion peptide (IP) Gly-Gly-Gly-Ser-Gly-Gly between Asn 116 and Lys 117 . The amino acid sequence of the IP ensured that the linker would have sufficient flexibility and the proper length required to allow the helices E and F to fold into the A-D cleft of the same peptide chain to form the monomeric isomer (Fig. 1).
Expression and Purification of IL-10M1 and wtIL-10 -IL-10M1 and wtIL-10 are expressed at equal levels in E. coli where they form insoluble inclusion bodies. Both polypeptides are refolded using the conditions previously described for wtIL-10 (19). Purified IL-10M1 and wtIL-10 proteins were estimated to be greater than 98% pure by SDS-PAGE. IL-10M1 and wtIL-10 peptide chains were analyzed by MALDI-TOF mass spectrometry. The observed masses of 19,151.6 (calculated 19150.8) for IL-10M1 and 18,778.4 (calculated 18, 778.5) for wtIL-10 are consistent with the addition of an initiating Nterminal methionine to each sequence.
IL-10M1 Adopts a Stable Monomeric Structure at Physiological pH-The oligomeric state of IL-10M1 was analyzed by cross-linking and size exclusion chromatography (Fig. 2). As expected for a monomer, IL-10M1 elutes with an apparent molecular weight of 18,600 (calculated 19,150), whereas the apparent molecular weight for wtIL-10 dimer is 44,600 (calculated 37,500). Protein fractions corresponding to the peaks in the gel filtration experiments were cross-linked with 1% glutaraldehyde at pH 7.0 in phosphate buffer. Cross-linked IL-10M1 ran at a molecular weight of ϳ17,000, whereas cross-linked wtIL-10 ran with a molecular weight of 34,000. Only bands corresponding to the IL-10M1 monomer and wtIL-10 dimer species were detected on the gel.
Based on previous biochemical analysis of wtIL-10, IL-10M1 should contain two intrachain disulfide bonds linking Cys 12 : Cys 108 and Cys 62 :Cys 114 (27). To confirm that two disulfide bonds were formed, IL-10M1 was incubated with 10 mM iodoacetic acid for 1 h at room temperature in the dark with or without DTT as described previously (29). MALDI-TOF mass spectrometry showed the predicted addition of 236 mass units (59 per cysteine residue) to samples containing DTT and no mass change for samples without DTT. Masses corresponding to IL-10M1 with unpaired cysteines were not detected in the preparations confirming that the refolded monomer preparations are homogeneous and contain two disulfide bonds.
Far-UV CD studies were performed to compare the ␣-helical content of IL-10M1 and wtIL-10. As shown in Fig. 3, CD spectra for IL-10M1 and wtIL-10 are essentially identical. Consistent with the crystal structure of wtIL-10, the spectra reveal a high ␣-helical content for both IL-10M1 and wtIL-10. This suggests that elimination of the small domain interface in the wtIL-10 dimer does not distort the ␣-helical content of IL-10M1.
Far-UV CD was also used to study the thermal stability of wtIL-10 and IL-10M1 (Fig. 4). At a concentration of 40 g/ml (ϳ1 M wtIL-10), temperature-induced changes in ellipticity are similar for IL-10M1 and wtIL-10 up to 37°C. Above 37°C, IL-10M1 retains more of its ␣-helical secondary structure in comparison to wtIL-10. For both IL-10M1 and wtIL-10, complete unfolding was not reached even at 95°C. However, the data clearly show that at equivalent concentrations, IL-10M1 displays greater thermal stability than wtIL-10.
IL-10M1 Forms a 1:1 Complex with sIL-10R␣-IL-10M1 and wtIL-10 receptor complexes were analyzed by size exclusion chromatography to determine the stoichiometry of the IL-10M1⅐sIL-10R␣ complex. Free IL-10M1 eluted at a volume of 29.7 ml with a calculated molecular weight of 18,600. Free sIL-10R␣ eluted at a volume of 26.62 ml with a calculated molecular weight of 42,600. Fig. 5 shows the chromatograms of IL-10M1 and sIL-10R␣ mixed at 1:1 or 1:4 molar ratios. IL-10M1 and sIL-10R␣ mixed at a 1:1 ratio elutes from the column at 24.92 ml and has an apparent molecular weight of ϳ69,000. When the proteins are mixed in this ratio, essentially all of the individual reactants are used to form the receptor complex. Incubation of a 1:4 molar ratio of IL-10M1 and sIL-10R␣ results in the same size complex as observed for the 1:1 molar ratio plus an additional peak corresponding to excess sIL-10R␣.
Characterization of the IL-10M1-sIL-10R␣ Interaction-Isothermal titration calorimetry (ITC) was used to define further the stoichiometry and dissociation constant between IL-10M1 and sIL-10R␣. Aliquots of 20.5 l of a 65 M solution of IL-10M1 were titrated into a 6.5 M solution of sIL-10R␣ at 25°C. The titration data revealed a stoichiometry of 1.15 mol of IL-10M1 per mol of sIL-10R␣. A dissociation constant (K d ) for the complex of 30 nM with an apparent ⌬H of binding of Ϫ12.2 kcal/mol was derived from least squares fitting of the binding isotherm (28). The small deviation of the stoichiometry from 1:1 is believed to be due to experimental errors in determining the concentrations of the reactants.
IL-10M1 Stimulates Cellular Proliferation-IL-10M1 was tested for its ability to induce proliferation of murine Ba/F3 cells transfected with either the human (Ba8 cells) or murine (BaMr cells) IL-10R␣ (Fig. 6). The concentrations of wtIL-10 required for half-maximal responses (EC 50 ) on BaMr and Ba8 cells were ϳ0.05 and ϳ0.8 ng/ml, respectively. These concentrations translate to specific activities for wtIL-10 of ϳ2 ϫ 10 7 units/mg on BaMr cells and ϳ1.25 ϫ 10 6 units/mg on Ba8 cells as previously reported (30). Surprisingly, IL-10M1 also displayed considerable biological activity with EC 50 values of ϳ0.9 ng/ml on BaMr cells and ϳ7 ng/ml on Ba8 cells. Based on the EC 50 values reported above, wtIL-10 is approximately 18-fold more active than IL-10M1 on BaMr cells and about 9-fold more active than IL-10M1 on Ba8 cells. The specific activity of wtIL-10 was approximately 16-fold lower on Ba8 cells than BaMr cells. The same trend was also observed for IL-10M1 although the activity difference between the cell lines was only 8-fold. Despite the lower activity of IL-10M1, higher concentrations of IL-10M1 resulted in maximal activity equivalent to wtIL-10 on both cell lines. IL-10M1 did not antagonize wtIL-10 on either cell line (data not shown).
IL-10M1 Activity Is Antagonized by sIL-10R␣-IL-10M1 or wtIL-10 induced cellular proliferation was neutralized by addition of sIL-10R␣ to the culture media (Fig. 7). Neutralization of IL-10M1 activity required an ϳ3.8-fold molar excess of sIL-10R␣ on Ba8 cells and a 35-fold molar excess of sIL-10R␣ on BaMr cells. A 38-and 343-fold molar excess of sIL-10R␣ was required to inhibit wtIL-10 induced proliferation on Ba8 cells and BaMr cells, respectively. These studies show that IL-10M1 biological activity requires the specific interaction with the IL-10R␣. Furthermore, IL-10M1 activity on either cell line is neutralized with approximately 10-fold less sIL-10R␣ than required to neutralize wtIL-10. For both IL-10M1 and wtIL-10, 10-fold higher concentrations of sIL-10R␣ are required to neutralize their activities on BaMr cells compared with Ba8 cells. This result is consistent with the higher affinity of the murine IL-10R␣ for IL-10 than the human IL-10R␣. DISCUSSION PCR mutagenesis has been used to insert the sequence Gly-Gly-Ser-Gly-Gly-Gly between Asn 116 and Lys 117 of the wtIL-10 polypeptide chain. Refolding this mutant under conditions identical to wtIL-10 resulted in a monomeric protein (IL-10M1) rather than the wtIL-10 dimer. Computer models of IL-10M1 indicate that it can adopt a three-dimensional structure that closely resembles one domain of the wtIL-10 dimer shown in Fig. 1. Several lines of experimental evidence are consistent with this proposal. First, the secondary structure of each protein is essentially identical as demonstrated by far-UV CD shown in Fig. 3. In addition to similar ␣-helical content, MALDI-TOF mass spectroscopy analysis of IL-10M1 is consistent with the presence of two disulfide bonds that are observed in wtIL-10 (27). Furthermore, IL-10M1 binds to the sIL-10R␣ that recognizes a three-dimensional binding site on wtIL-10. These experiments taken together suggest that despite eliminating the wtIL-10 dimer interface, IL-10M1 has a structure similar to one domain of the wtIL-10 dimer.
The thermal unfolding data show that IL-10M1 is more stable that wtIL-10. This suggests that the wtIL-10 dimer has evolved to carry out a specific biological function at physiological temperature rather than to achieve optimal thermal stability. Currently it is not clear why wtIL-10 has not evolved a more extensive dimer interface such as observed for IFN-␥ or interleukin-5 (29,31,32). This is especially curious since wtIL-10 begins to dissociate in vitro at a concentration of ϳ1 M but is biologically active at concentrations ϳ10 6 -fold (ϳ1 pM) below where dimer dissociation is observed (25). This suggests that the stability characteristics of the wtIL-10 dimer may play an important and yet undefined role in regulating wtIL-10 biological activity.
The interaction between IL-10M1 and sIL-10R␣ expressed in Drosophila S2 cells was characterized by gel filtration chromatography and ITC studies. The data reveal the binding stoichiometry for the IL-10M1-sIL-10R␣ interaction is 1:1 at all concentrations studied (6 -25 M). As a control, we repeated the efforts of Tan et al. (16) who showed that incubation of wtIL-10 and sIL-10R␣ generates a complex of 2 wtIL-10 molecules and 4 sIL-10R␣. Since IL-10M1 is essentially identical to one domain of wtIL-10, we expected that incubation of IL-10M1 with sIL-10R␣ might form complexes that contained 2 or 4 IL-10M1 molecules and 2 or 4 sIL-10R␣s. However, these larger complexes were not observed over the concentrations studied. Thus, the interaction between IL-10M1 and the sIL-10R␣ is not sufficient to form the 2 IL-10⅐4 sIL-10R␣ complex observed with the IL-10 dimer.
Since IL-10M1 forms a 1:1 complex with the sIL-10R␣, we hypothesized that IL-10M1 would not be biologically active. However, the ability of IL-10M1 to induce short term proliferative responses was only about 9-fold lower than wtIL-10 on murine BaF3 cell line expressing the human IL-10R␣ (Ba8 cells) and 18-fold less active than wtIL-10 on BaF3 cells expressing the murine IL-10R␣ (BaMr cells). Differences in activity between these cell lines is at least partly the result of different receptor components (human IL-10R␣/murine IL-10R␤ on Ba8 or murine IL-10R␣/murine IL-10R␤ on BaMr) displayed on their cell surfaces (11,30). However, on both cell lines maximal proliferative responses equivalent to wtIL-10 could be achieved with higher concentrations of IL-10M1 (Fig.  6). Our experiments do not rule out the possibility that in the biological assay IL-10M1 might dimerize on the cell surface. A fixed concentration of IL-10M1 or wtIL-10 was mixed with increasing amounts of sIL-10R␣. Cell proliferation was monitored after 48 h by optical density measurement following MTS staining. Data points from five dilution experiments for each cytokine on each cell type were normalized, and a best fit line was drawn using the program TableCurve 2D (Janel Scientific). Data are plotted as percent activity of the IL-10M1 or wtIL-10 as a function of sIL-10R␣ concentration. IL-10M1 concentrations were 51 pM on BaMr cells and 459 pM on Ba8 cells. wtIL-10 concentrations were 5.2 pM on BaMr cells and 47 pM on Ba8 cells. At the above concentrations of cytokine, all curves converged to an IC 50 concentration for sIL-10R␣ of 50 ng/ml. However, this seems unlikely since IL-10M1 is observed as a monomer at concentrations much higher (ϳ13 M) than required for IL-10M1 activity (ϳ0.5 nM).
This leads to the question of the importance of the wtIL-10 dimer for generating wtIL-10 biological responses. One advantage for the wtIL-10 dimer is its ability to simultaneously bind two IL-10R␣s. Based on the crystal structure of wtIL-10 and the IFN-␥ receptor complex, the orientation of the wtIL-10 dimer domains (90°) is optimized for this 2-fold interaction that is lost in IL-10M1. As a result, wtIL-10 binds to the IL-10R␣ ϳ60-fold more avidly than IL-10M1 (0.5 versus 30 nM). Although this is an important difference between wtIL-10 and IL-10M1, it is interesting that viral IL-10 with an apparent dissociation constant of ϳ500 nM (ϳ16-fold higher than IL-10M1) has the same specific activity as wtIL-10 (30). The decreased biological activity of IL-10M1 versus wtIL-10 may be the result of the different receptor complexes they form with the IL-10R␣. The 2IL-10⅐4IL-10R␣ complex generated by the dimer provides a way to cluster four IL-10R␣s on the cell surface. The ability of the dimer to cluster four IL-10R␣s may enhance IL-10R␤ binding and subsequently wtIL-10 biological activity. Since IL-10M1 only forms a 1:1 interaction with IL-10R␣, the ability to recruit IL-10R␤ should be significantly diminished.
Our studies strongly argue that the interaction of one domain of wtIL-10 with one IL-10R␣ and one IL-10R␤ is sufficient for cellular proliferative responses. Thus, this ternary complex is all that is required for activation of the intracellular kinases, Jak1 and Tyk2. From our present data, we predict that each domain of wtIL-10 in the 2:4 wtIL-10 receptor complex independently activates the Jak kinases responsible for signal transduction. Said another way, phosphorylation events occur only between receptors associated with the same domain of wtIL-10 instead of across the 2-fold axis of the dimer. Currently, our predictions are only for wtIL-10 proliferative responses. This is because Riley et al. (33) recently reported that different portions of the IL-10R␣ intracellular domain are responsible for wtIL-10-induced proliferative and cytokine inhibitory activities. Experiments to determine the ability of IL-10M1 to suppress cytokine synthesis are currently under way.
IL-10M1 provides an important tool for studying the structural, stoichiometric, and energetic requirements necessary for IL-10-induced receptor activation. Since IL-10M1 forms a 1:1 interaction with sIL-10R␣ it provides a simplified complex, compared with the 2:4 wtIL-10⅐sIL-10R␣ complex, for detailed biophysical studies. The 1:1 complex will be most helpful to our efforts to obtain crystals of the IL-10 receptor complex for high resolution x-ray diffraction studies and to define the energetic contributions of residues in the IL-10/sIL-10R␣ binding interface without the complications of the larger 2:4 complex. Additional protein folding and NMR studies on IL-10M1 are also under way. The design and characterization of IL-10M1 represents our first efforts to antagonize wtIL-10 activity by changing its quaternary structure. Our current and future studies outlined above should provide new insights for modulating the IL-10 signal transduction pathway.