Specific sequence mutations in a long-lasting rhIFN-α2b version reduce in vitro and in vivo immunogenicity and increase in vitro protein stability

Abstract

For decades, recombinant human interferon alpha (rhIFN-α2b) has been used to treat emerging and chronic viral diseases. However, rhIFN-α2b is immunogenic and has a short in vivo half-life. To solve these limitations, two long-lasting hyperglycosylated proteins with reduced immunogenicity were developed and designated as 4N-IFN(VAR1) and 4N-IFN(VAR3). Here, we continue to study the relevant characteristics of these therapeutic candidates. Thus, we demonstrated that both de-immunized IFN versions elicited significantly lower neutralizing antibody responses than the original molecule in HLA-DR1 transgenic mice, confirming our previous in vitro protein immunogenicity data. Also, we found that these biobetters exhibited remarkable stability when exposed to different physical factors that the protein product may encounter during its production process and storage, such as low pH, thermal stress, and repeated freezing/thawing cycles. Taking into consideration our previous and present results, 4N-IFN(VAR1) and 4N-IFN-4N(VAR3) appear to be valuable candidates for the treatment of human viral diseases.

Keywords: Anti-viral therapy; De-immunization; EpiMatrix; IFN alpha; Immunogenicity; In vivo assays.

Fig. 1

4N-IFN de-immunized variants showed a markedly reduced in vivo immunogenicity in comparison with the original molecule. Six HLA-DR1 transgenic mice per group were inoculated with PBS (control) or the indicated IFN variant emulsified with CFA. Neutralizing antibody titers were determined by a neutralization test against 4N-IFN, 4N-IFN(VAR1), or 4N-IFN(VAR3). Serial dilutions (starting from 1:10 and increasing twofold) of sera from HLA-DR1 transgenic mice immunized with each IFN species or unimmunized controls were incubated at 37 °C with 20 IU/ml of each type of IFN-α2b, considering the specific antiviral activity previously reported for each protein. Then, the individual mixtures were added to monolayers of MDBK cells to assay the antiviral activity. Titers were expressed as the dilution of serum that reduces to 50% (IC50) the initial IFN antiviral activity (IFN proteins incubated with control serum). While sera from mice treated with PBS alone did not develop detecqtable neutralizing antibodies against IFN (data not shown) high Nab titers were observed for those mice inoculated intraperitoneally with 4N-IFN. In contrast, a marked reduction in Nab titers was detected in mice treated with both de-immunized variants. Horizontal bars represent mean ± SD. (***p < 0.001).

Fig. 2

De-immunized 4N-IFN variants retained the prolonged in vitro human plasma stability of 4N-IFN. IFN variants were incubated with human plasma for 7 days at 37 °C. After treatment, residual antiviral biological activity was evaluated at the indicated times. While a gradual loss of activity was observed for parental IFN, no significant differences were detected for all 4N-IFN proteins. This reflects the potential protective effect of N-glycans attached to the IFN molecule to proteases present in human plasma. Data points represent mean ± SD of triplicate samples.

Fig. 3

4N-IFN(VAR1) and 4N-IFN(VAR3) showed higher thermal stability than the original molecule. IFN variants were exposed to different temperatures for 10 min. Protein stability was assayed by measuring the residual antiviral activity in each sample. All 4N-IFN proteins retained nearly intact their biological function after incubation at 65 °C. However, when treated at 75 °C, 4N-IFN lost nearly 50% of the initial antiviral activity. Contrarily, 4N-IFN(VAR1) and 4N-IFN(VAR3) were not affected by this stress condition as both proteins maintained intact their original biological function. These results highlight the additional stabilizing effect of specific mutations introduced in the 4N-IFN sequence to reduce its immunogenicity. Experiments were performed in triplicate and each data point represents mean ± SD. (***p < 0.001).

Fig. 4

De-immunized 4N-IFN proteins are more stable than the original molecule after treatment of successive freezing/thawing cycles. IFN variants were exposed to 15 cycles consisting of freezing at −70 °C followed by thawing in a water bath at 37 °C. To evaluate the effect of successive freezing/thawing cycles on the stability of each protein, we measured the antiviral biological activity at the end of the treatment. After this stress condition, 4N-IFN protein was severely affected as only retained 26% of the initial antiviral activity. In contrast, 4N-IFN(VAR3) was less affected by this stress treatment, as this protein retained more than 70% of the initial biological antiviral activity. The magnitude of this effect was even more notorious for 4N-IFN(VAR1) which conserved full residual biological function. Experiments were carried out in triplicate and each bar represents mean ± SD. Statistical comparisons were assessed by One-way ANOVA followed by Tukey (**p < 0.01, ***p < 0.001).

Fig. 5

De-immunized IFN versions are more stable at acid pH treatment than the original molecule. IFN variants were exposed at low pH at the indicated times. After each time point, we measured the antiviral biological activity as an indicator of the stability of each protein. 4N-IFN retained the initial bioactivity for 30 min of incubation and then evidenced a gradual loss of antiviral activity until almost complete inactivation after 2 h of treatment. In contrast, both 4N-IFN(VAR1) and 4N-IFN(VAR3) retained full biological function even at the end of treatment. Assays were performed in triplicate and each data point represent mean ± SD. (***p < 0.001).