Fig 2: Ex vivo immunogenicity test of EPO proteins. Mature DC cells were pulsed with EPO protein including EPO-1.2 (L70V, V74L, L102I), EPO-3.1 (T106A), EPO-3.2 (T106G), EPO-3.3 (T106H) or EPO-4.1 (L109A). Anti-CD3/anti-CD28, the known CD4+ T cell activator was used as a positive control. EPO-pulsed DCs and CD4+ T cell from the same individual were cocultured at a 1:20 ratio. CD4+ T cell activation was quantified by the measurement of IFN-? release using sandwich ELISA. The mean value from 3 HLA-DRB1*09-negative volunteers and 3 HLA-DRB1*09-positive volunteers were combined as a negative group or a positive group, respectively. Each group was relatively compared to the response to BRP and presented in the percentage of relative response of EPO proteins shown in (a). Bar graphs are shown as mean ± SEM calculated from 3 individuals in each group. Asterisks denote a significant difference by t-test (* p = 0.05, ** p = 0.01, *** p = 0.001, **** p = 0.0001). IFN-? level of each HLA-DRB1*09-positive volunteers shown in (b) and HLA-DRB1*09-negative volunteers shown in (c). Bar graphs are shown as mean ± SEM calculated from duplicate experiments. The similarity or difference (**** p = 0.0001) in IFN-gamma level among BPR, EPO-WT, EPO mutants and anti-CD3/anti-CD28 in both positive and negative group was analyzed by one-way ANOVA followed by Tukey’s multiple comparisons test for (b) and (c). The significant differences were only shown in comparison to EPO-WT in this figure. Graphs were generated using GraphPad Prism version 8.4.3 for Windows (https://www.graphpad.com/).

Fig 3: Anti-EPO VHH specificity and pH dependence. a Western blot of 0.1 µg purchased EPO (lane 1) and 10 µl of CHO cell supernatant transfected with a vector containing the human EPO gene (lane 2), a vector containing the Rituximab heavy and light chains (lane 3), and an empty vector (lane 4). Blots were developed using anti-EPO VHH (top panel), anti-EPO IgG (second panel), a VHH raised against an unrelated glycoprotein (third panel), and an anti-human IgG-HRP conjugate (lower panel). b Sensorgrams showing immobilized anti-EPO VHH binding of in-house purified EPO (10 µg/ml) at the indicated pH values

Fig 4: Coomassie blue stained gel (a) and Western immunoblotting (b) of the EPO proteins. EPO standard, wild type and mutant proteins were separated according to their molecular mass in 12% SDS polyacrylamide gel at a constant 175 voltages for 1.5 h. (a) The purity of separated proteins was visualized by Coomassie blue staining. The gel image was taken using HP LaserJet scanner. (b) The identity of EPO antigen was detected by Western immunoblotting. The chemiluminescence signals of protein samples were detected and the image was taken using ImageQuant LAS 4000 control software version 1.2 provided with ImageQuant LAS 4000 machine at the exposure time of 1 s. The signals were then overlaid with the marker using ImageQuant TL software version 7.0 (https://www.cytivalifesciences.com/en/us). The native gel and blot are presented in Supplementary Figs. S1 and S2, respectively. BRP, EPO standard; Marker, Precision Plus Protein Kaleidoscope Prestained Protein Standards; WT, EPO-WT; 1.2, EPO-1.2 (L70V, V74L, L102I); 3.1, EPO-3.1 (T106A); 3.2, EPO-3.2 (T106G); 3.3, EPO-3.3 (T106H); 4.1, EPO-4.1 (L109A).

Fig 5: Isoform analysis by IEF of EPO proteins. EPO standard, wild type and mutant proteins were separated according to their pI in IEF gel with a pH gradient of 2–6 followed by Western blot. The chemiluminescence signals of protein samples were detected and the image was taken using ImageQuant LAS 4000 control software version 1.2 provided with ImageQuant LAS 4000 machine (https://www.cytivalifesciences.com/en/us) at the exposure time of 5 s. The native blot with multiple exposures is presented in Supplementary Fig. S3. The separated bands of the proteins represent the number of their existing glycoforms. Distribution of different isoforms of the purified EPO protein was compared to BRP, an EPO standard. BRP, EPO standard; WT, EPO-WT; 1.2, EPO-1.2 (L70V, V74L, L102I); 3.1, EPO-3.1 (T106A); 3.2, EPO-3.2 (T106G); 3.3, EPO-3.3 (T106H); 4.1, EPO-4.1 (L109A).