Sylviane Komesli, Martine Page, and Patrick Dutartre
Dpartement dImmunologie, Laboratoires Fournier S.A., France
TGF- type I and type II receptors form ligand dependent heteromeric signaling complexes, in which transforming growth factor- receptor type II (TRII) tends to act as the primary receptor. In the present study, we investigated the feasibility of using the FlashPlate platform to assay a chimeric soluble type II receptor fused with the Fc regions of human immunoglobulin (TRIIs-Fc), in order to screen for potent agonists and antagonists of TGF.
Transforming growth factors 1, 2 and 3 (TGF-s) are multifunctional cytokines involved in the regulation of cell proliferation, differentiation and extracellular matrix production. In mammalian cells, responses to TGF- are mediated by types I and II cell surface receptors (TRI and TRII respectively) which are expressed in most cell types and tissues. TGF- binds directly to TRII, allowing this receptor to associate with and phosphorylate TRI, which then propagates the signal through activation of Smad2 and Smad4 heteromeric complexes1.
At this time, there are no known, clinically useful TGF- agonists or antagonists. A better understanding of the mechanism of activation of the TGF- receptor complexes may be useful, therefore, as a step towards the development of TGF- antagonist or agonist drugs. We have investigated the possibility of using the soluble extracellular domain of TRII in a nonseparation microplate receptor binding assay.
To facilitate this approach and increase the probability of success, we constructed a chimeric soluble receptor, by fusing the extracellular domain of TRII to the Fc regions of human immunoglobulin (TRIIs-Fc). Through the Fc region, the chimeric receptor, TRIIs-Fc, expressed in a transiently tran sfected Cos-7 cell line, was easily purified by one-step protein A affinity chromatography, then coated with high efficiency into the wells of a Protein A FlashPlate microplate (SMP102). This protein was biochemically characterized; then its ability to bind 125I-labeled TGF-1 was studied2.
Construction of vector: The cDNA encoding the extracellular domain of the human TGF- type II receptor (TRII) was amplified by PCR from the plasmid, which contained a full-length cDNA of the receptor. The resulting amplified PCR product was digested and ligated into the cloning sites of the vector pIg-Tail (R&D System).
Transient transfection of Cos-7 cells: Cos-7 cells (American Type Culture Collection CRL 1651) were transiently transfected with the expression plasmid pIg-Tail containing the cDNA encoding for the truncated TGF- type II receptor. The recombinant protein was expressed in the transfected cells and secreted into the medium.
Protein purification and analysis: The human recombinant protein TRIIs-Fc was purified by one-step protein A affinity chromatography. The eluted protein was dialyzed overnight against 0.1 x PBS and lyophilized.
Protein A FlashPlate binding assay: Protein A FlashPlate microplates are precoated with protein A. This was allowed to immobilize the chimeric protein (100 μl of affinity purified protein at 2.5 μg/ml in pH 7.2 NaCl/Pi) by the Fc portion for two hours at room temperature. After 2 washes with binding buffer (128 mM NaCl; 5 mM KCl; 5 mM MgSO4; 1.3 mM CaCl2; 50 mM Hepes, pH 7.6), the wells were treated for two hours at room temperature with blocking solution consisting of binding buffer containing 5% BSA. After incubation, the wells were washed three times with binding buffer containing 1% BSA.
For the binding studies, 125 I-labeled TGF-1 (Specific activity 3,000-4,500 Ci/mmole, NEX267) was diluted in binding buffer just prior to its addition into the wells. The final concentrations of 125I-labeled TGF-1 ranged from 50 pM to 2.5 nM. The plates were incubated at room temperature for two hours, then sealed and counted on a Packard Top Count Microplate Scintillation Counter. A one minute counting period was used. All determinations were carried out at least in duplicate. Nonspecific binding was determined with an excess of unlabeled TGF-1 (100-fold). The counting efficiency was determined through the use of 125I-labeled IgG directly coated into Protein A FlashPlates. The cpm obtained at equilibrium, divided by the dpm added to the well, gave a measure of the counting efficiency, which was evaluated at 5%. For the competition binding assay, the final concentration of radiolabeled ligand was 500 pM (specific activity 300-450 Ci/mmole), with final concentrations ranging from 0.05 to 200 nM for TGF-1, TGF-2 and TGF-3.
Binding of 125I-labeled TGF-1 to cell: Ligand binding of cell monolayers with 125I-labeled TGF-1 (20 pM to 1 nM) were carried out as previously described3, then scatchard plotted with values obtained after gamma scintillation counting. Scatchard experiments were performed on Cos-1 cells (American Type Culture Collection CRL1650) transiently transfected either with the TRII (pCMV5 TRII), or with both TRI and TRII containing plasmids (pCMV5 TRI + pCMV5 TRII), and compared to Cos-1 cells transfected with the empty pCMV5 (Invitrogen) in order to measure the nonspecific binding.
Comparative studies on binding of 125I-labeled TGF-1 to human TRIIs-Fc and wild type TRII: Ligand binding activity of the recombinant hTRIIs-Fc receptor was tested in a Protein A FlashPlate binding assay. Protein A precoated plastic surface s within the wells were coated with the Fc portion of the affinity purified recombinant chimeric receptor. The background level was measured in noncoated wells of Protein A FlashPlate, and shown to represent 50% of the nonspecific binding obtained upon TRIIs-Fc coating. The soluble chimeric receptor secreted by Cos-7 cells was able to effectively bind 125I-labeled TGF-1 (Figure 1). The Protein A FlashPlate assay showed a typical saturation curve of TGF-1 binding to hTRIIs-Fc (Figure 2). The kinetics of 125I-labeled TGF-1 binding, performed at room temperature, were relatively rapid, since equilibrium binding was usually achieved after just two to three hours of incubation.
Meaningful values for binding affinity constants could be derived from the obtained binding data. Scatchard analysis (Figure 2b) revealed a single class of binding sites giving a Kd value of 1370 + 363 pM. The data obtained from the scatchard analysis indicated that approximately 180-200 fmoles of recombinant hTRIIs-Fc were immobilized in each coated well.
In order to understand the relevance of our results, we compared the Kd value of the hTRIIs-Fc for TGF-1 with the Kd value obtained with recombinant hTRII expressed in the Cos-1 cell line. Previous data showed that hTRI was able to enhance the affinity of complexes formed with hTRII in comparison to TRII alone. Here, we have investigated the affinity of Cos-1 cells expressing TRII alone, or expressing both TRI and TRII. As shown in Figures 3 and 4, although the relative affinity of Cos-1 cells for 125I-labeled TGF-1 was low (Kd 1123 + 413 pM) when TRII was expressed alone (Figures 3a and 3b), co-expression of both TRI and TRII (Figures 4a and 4b) converted the system to a higher affinity model (470 + 32 pM).
Differential binding affinity of TGF- isoforms for soluble chimeric recept or TRIIs-Fc: It has previously been shown that TGF-1 and TGF-3 bind with much higher affinities than TGF-2 to TRII. To identify the selectivity of hTRIIs-Fc, TGF- isoforms were compared in competition experiments with 125I-labeled TGF-1 binding in the Protein A FlashPlate assay, where 125I-labeled TGF-1 (final concentration 500 pM) was mixed with a serial dilution of recombinant TGF- isoforms. Figure 5 shows the displacement curve of three different TGF- isoforms: TGF-1, TGF-2, and TGF-3. The relative inhibitory concentration (IC50) was determined as the TGF- concentration which inhibited 50% of 125I-labeled TGF-1 binding. The IC50 values obtained in this test were 3.3 + 0.06 nM and 3.9 + 0.09 nM for TGF-1 and TGF-3, respectively, whereas TGF-2 was unable to compete, even at a concentration of 200 nM, with the binding of 500 pM of 125I-labeled TGF-1 to recombinant TRIIs-Fc. These results indicated that the selectivity for binding TGF- isoforms was maintained by this chimeric recombinant receptor.
In this study, we have analyzed the binding properties of recombinant hTRIIs-Fc for TGF-1 using FlashPlate technology. For our experiments, the hTRIIs-Fc component was coated onto the wells of Protein A FlashPlates and incubated with 125I-labeled TGF-1. The key feature of this Protein A FlashPlate binding assay, as described here, is its simplicity comprising only the addition of the 125I-labeled TGF-1, incubation, and measurement. Unlike separation assays, this assay does not disturb the equilibrium, while the CPM obtained are directly correlated to the bound 125I-labeled TGF-1. In our assay, we have determined a counting efficiency of about 5%; this result may be explained by the increased distance between 125I-labeled TGF-1 and solid scintillant, due to the coating of protein A and hTRIIs-Fc on t he wells of the FlashPlate. Nevertheless, despite the low counting efficiency obtained, the assay is accurate and its correlation with a conventional approach suggests that it is suitable for hTRIIs-Fc binding studies.
Moreover, we compared the affinity of hTRIIs-Fc (1370 + 363 pM) to the affinity of Cos-1 cells expressing either TRII alone (1123 + 413 pM), or both TRI and TRII (470 + 32 pM). In these experiments, the affinity of the chimeric soluble receptor for TGF-1 was threefold lower than the affinity of the heterodimer receptor complex (TRI and TRII) on the cell surface, but was very similar to that of TRII when expressed alone in the Cos-1 cell line.
One significant result is that the chimeric protein shows the same selectivity for different TGF- isoforms as the native protein. TGF-2 is not able to compete with the binding of 125I-labeled TGF-1, although TGF-1 and TGF-3 both show an IC50 value in the 3-4 nM range. Previous data have shown that TRII binds TGF-2 with low affinity, and co-expression of TRIII is required.
In conclusion, in vitro binding assays demonstrate that hTRIIs-Fc is able to bind TGF-1 with a similar affinity to that of wild type hTRII when overexpressed alone in Cos-1 cells, and can be used accordingly in Protein A FlashPlate microplates to search for TGF- agonist and antagonist compounds.
1. Massague, J. (1996) TGFbeta signaling: receptors, transducers, and Mad proteins. Cell 85, 947-950.
2. Komesli, S., Vivien D., & Dutartre, P. (1998) Chimeric extracellular domain of type II transforming growth factor (TGF)- receptor fused to the Fc region of human immunoglobulin as a TGF- antagonist. Eur. J. Bioch. 234, 505-513.
3. D. Vivien and J. L. Wrana (1995) Ligand-induced recruitment and phosphorylation of reduced TGF-beta type 1 receptor, Exp. Cell Res. 221, 60-65.
Most of these results have been published (Komesli et al. 1998 Chimeric extracellular domain of type II transforming growth factor (TGF)- receptor fused to the Fc region of human immunoglobulin as a TGF- antagonist. Eur. J. Bioch. 234, 505-513). We thank European Journal of Biochemistry for giving us the permission to reproduce published material.