Fig 1.
The directed evolution process for the generation of DR3 mutants with improved TL1A affinity, stability and biological activity comprises: (A) library generation, (B) screening, and (C) characterization.
Fig 2.
Screening for improved DR3 mutants using yeast surface display.
(A) Yeast surface display of DR3; DR3 (turquoise) is displayed as an Aga2 (gray) fusion on the surface of the yeast. Expression and TL1A binding are detected by fluorescent antibodies (green) and streptavidin (yellow), respectively. (B) Flow cytometry histogram analysis of cell populations displaying native DR3 (blue), and DR3 mutant libraries after two rounds (R2, red) and three rounds of enrichment (R3, green) for TL1A binding. In each round of enrichment, tje cell population exhibiting high expression (FITC signal) and TL1A binding (APC signal) were sorted according to the fluorescent signals. A cell population that was not surface displayed is shown in black. Analysis of cells was performed following incubation with 0.2 μM TL1A.
Table 1.
List of mutations in the six selected DR3 variants.
Fig 3.
Cell-based assay for the analysis of DR3 variants.
(A) The H3 and O6 variants are more potent in inhibiting TL1A-induced secretion of IFN-γ in human CD4+ than native DR3. Cells were incubated for 72 hours with 100 ng/ml TL1A, 20 ng/ml IL-12 and 50 ng/ml IL-18 and different concentrations of soluble native DR3 or the H3 or O6 receptors. The 1:10 diluted cell supernatant was analyzed by ELISA to assess IFN-γ levels. The IFN-γ values were calculated according to a IFN-γ calibration curve. Black stars denote measurements that are statistically different from no receptor (DR3 = 0) (p<0.02) while red stars are measurements that are statistically different between the O6 and native version of the protein (p<0.05). (B) The H3 and O6 mutants are more potent in inhibiting TL1A-induced apoptosis in TF-1 cells than is native DR3. Cells were incubated for six hours with 8 μg/ml cyclohexamide (CHX), 75 ng/ml TL1A and the indicated concentrations of soluble native DR3 or the H3 or O6 receptors. Following six hours of incubation, lysis buffer containing the caspase-3 fluorescent substrate DEVD-AMC was added and caspase-3 activity was monitored for 10 minutes. Black stars denote measurements that are statistically different from no receptor (DR3 = 0) (p< 0.003) while red stars are measurements that are statistically different between the O6 and native versions of the protein (p < 0.05). The data presented in the experiments involving CD4+ and TF-1 cells is the average of three independent repeats of each experiment while the error bars represent the standard deviation from the average.
Table 2.
Kinetic rate constants of TL1A cytokine binding to native DR3-Fc (WT), H3 and O6 variants as determined by SPR.
Fig 4.
Stability of native DR3, and the H3 and O6 variants.
(A) Thermal denaturation of DR3 variants measured using CD spectroscopy. The DR3 variants, including the native protein (black), H3 (green) and O6 (red), were subjected to thermal denaturation at 30–70°C at a heating rate of 0.37°C/min, and changes in the ellipticity at 215 nm were monitored. Solid lines represent a logistical model fitting of the melting curves. The O6 and H3 variants exhibit improved melting temperature of 3.5°C and 5.5°C, respectively. (B) Stability of native DR3, and the H3 and O6 variants in human serum. The DR3 variants were incubated for 30 minutes in human serum at 0°C, 37°C and 42°C and the residual inhibitory activity of TL1A-induced apoptosis in TF-1 cells was determined (see also Fig 4B). All results were normalized to the DR3 activity following incubation at 0°C. High sensitivity of the native protein relative to mutant activity is observed following incubation at 42°C. Black stars denote measurements that are statistically different than the residual activity of the native protein following incubation at 42°C (p<0.05).