Tailored collagen binding of albumin-fused hyperactive coagulation factor IX dictates in vivo distribution and functional properties

Abstract

The efficacy of hemophilia B (HB) replacement therapy is evaluated by coagulation factor IX (FIX) activity in plasma, although FIX bound to extravascular type IV collagen (Col4) also contributes to efficient hemostasis. Here, we investigated the impact of engineering FIX for improved (K5R) or reduced (K5A) Col4 binding on the pharmacokinetic properties of FIX Padua, fused to human serum albumin (HSA_QMP) engineered for favorable neonatal Fc receptor (FcRn) engagement. Hyperactive features and extended plasma half-life in human FcRn expressing mice, attributed to FIX Padua and HSA_QMP engineering, respectively, was confirmed. In HB mice, Padua_KA-HSA_QMP exhibited negligible extravascular distribution and the highest plasma levels at early time points followed by the steepest decay. Conversely, Padua_KR-HSA_QMP showed increased extravascular distribution and a 3-fold longer functional half-life (80 hours). These findings support the use of Padua_KA-HSA_QMP and Padua_KR-HSA_QMP as hyperactive short- or long-term therapeutics, respectively, with opportunities for tailored HB replacement therapy.

Fig. 1. Design and activity profiles of FIX-HSA variants with modifications in both fusion partners.

a Four fusion proteins were designed with FIX genetically fused to HSA via a cleavable linker. FIX was modified in position 5 of the Gla domain (K5A or K5R) and in position 338 in the catalytic domain (R338L; Padua). HSA was either unmodified (WT) or engineered with three amino acid substitutions (E505Q/T527M/K573P; QMP) in DIII. Gla, γ-carboxyglutamic acid domain; EGF1/EGF2, epidermal growth factor like domains 1 and 2; AP, activation peptide. Coagulant activity of FIX measured by aPTT-based assays in FIX-deficient plasma supplemented with pure monomeric fractions of (b), FIX-HSA_WT or (c), Padua-HSA_QMP. Data represent the mean ± SD of technical duplicates. d Specific clotting activity of each FIX-HSA variant. The data represent the mean ± SD of technical quadruplets, where specific activity is calculated as the ratio between activity and protein levels (FIX-HSA_WT = 1). Statistical significance was tested by unpaired two-tailed Student’s t tests with 95% confidence level (n = 4; ns p = 0.8792; **p = 0.0010, ****p < 0.0001). FIX-HSA_WT, black; Padua-HSA_QMP, gray; FIX_KA-HSA_WT, blue; FIX_KR-HSA_WT, orange; Padua_KA-HSA_QMP, pink; Padua_KR-HSA_QMP, purple. Source data are provided as a Source Data file. a Created in BioRender. Hovden Aaen, K. (2025), https://BioRender.com/auh41ln.

Fig. 2. QMP engineering in HSA yields FIX-HSA fusion proteins with favorable FcRn binding properties and FcRn-mediated cellular handling.

a Illustration of the ELISA-based binding assay used to study binding between hFcRn or mFcRn and FIX-HSA variants at pH 5.5. Soluble truncated FcRn (light gray) was captured on IgG1 MST/HN (blue) coated in the well, before FIX-HSA (red/black) was added to the wells, and the HSA region of the fusion protein was detected by an ALP-conjugated anti-HSA antibody (dark gray). b, c Results from the ELISA-based hFcRn and mFcRn binding assays performed at pH 5.5. Data represent the mean ± SD of technical duplicates from one representative experiment. d Illustration of the SPR experiment used to study the interaction between FcRn (gray) and FIX-HSA (orange/black) performed by immobilizing FIX-HSA fusion proteins on a CM5 sensor chip before injecting a truncated form of soluble FcRn in a concentration gradient. Representative sensorgrams from SPR performed by injecting soluble truncated hFcRn (e–h) or mFcRn (i–l) over immobilized FIX-HSA variants at pH 5.5, showing one out of three independent experiments performed. The dotted lines show curves fitted by the 1:1 Langmuir binding model. m An illustration of the HERA setup designed to study hFcRn-mediated cellular (n), uptake and (o), recycling of FIX-HSA fusion proteins. Data are presented as relative to FIX-HSA_WT (=1, dotted line) and represent the group mean ± SD of three independent experiments performed in technical triplicates (uptake: n = 9, except n = 8 for FIX_KA-HSA_WT and FIX_KR-HSA_WT, n = 6 for Padua_KR-HSA_QMP; recycling: n = 9, except n = 7 for FIX_KA-HSA_WT). Statistical significance was tested by unpaired two-tailed Student’s t tests with 95% confidence level (precise p-values are given in Supplementary Table 1). FIX-HSA_WT, black; Padua-HSA_QMP, gray; FIX_KA-HSA_WT, blue; FIX_KR-HSA_WT, orange; Padua_KA-HSA_QMP, pink; Padua_KR-HSA_QMP, purple. Source data are provided as a Source Data file. a, d, m Created in BioRender. Hovden Aaen, K. (2025) https://BioRender.com/auh41ln.

Fig. 3. QMP engineering extends the plasma half-life of FIX-HSA.

a Illustration of the extra- and intra-vascular storages of FIX. Infused and endogenous FIX may relocate from the circulation to the extravascular space where it associates with Col4. The K5A and K5R substitutions in FIX weakens or strengthens its interaction with Col4, respectively. b Illustration of the half-life study in Tg32 mice where 2 mg/kg of FIX-HSA were administered by I.V. injection. Protein quantification was performed by ELISA. The results are presented as (c), protein concentration 24 h after administration and (d), percentage remaining in plasma at each time point compared to day 1 (= 100%), with plasma half-life values to the right, and represent the group mean ± SD of biological replicates (n = 5). e Illustration of the half-life study in FIX^plus Balb/c mice where 2.5 mg/kg of FIX-HSA was administered by I.V. injection. Protein quantification was performed by ELISA. The results are presented as (f), protein concentration 24 h after administration and (g), percentage remaining in plasma at each time point compared to day 1 (= 100%), with plasma half-life values to the right, and represent the group mean ± SD of biological replicates (n = 5). Statistical significance was tested by unpaired two-tailed Student’s t tests with 95% confidence level (n = 5; precise p-values are given in Supplementary Tables 2–5). FIX-HSA_WT, black; FIX_KA-HSA_WT, blue; FIX_KR-HSA_WT, orange; Padua-HSA_QMP, gray; Padua_KA-HSA_QMP, pink; Padua_KR-HSA_QMP, purple. Source data are provided as a Source Data file. a, b, e Created in BioRender. Hovden Aaen, K. (2025) https://BioRender.com/auh41ln.

Fig. 4. K5R engineering of FIX-HSA increases plasma half-life and extravascular presence in HB mice.

a Illustration of the study in HB mice where 2.5 mg/kg of FIX-HSA was administered by I.V. injection. At termination (day 4), lungs, kidneys, liver, and knee joints were harvested and processed to homogenates, or paraffin embedded and prepared on slides before IF or IHC staining. Blood processed to plasma and tissue homogenates were analyzed by ELISA for protein quantification, and coagulant activity was measured in plasma by an aPTT-based assay. b–d Protein concentration in plasma 1 h, 24 h, and 96 h post-administration. Data are presented as the group mean ± SD of biological replicates (n = 4–5). Elimination curves showing (e), protein concentration remaining in plasma over time and (f), percentage of protein remaining in plasma at each time point compared to day 1 (= 100%), presented as the mean ± SD of biological replicates (n = 4–5). g Plasma half-life values shown as the group mean ± SD of biological replicates (n = 4–5). h MRT values in FIX^plus Balb/c mice (x-axis) and HB mice (y-axis) analyzed in gPKPDsim. i–l Homogenates of livers, kidneys, lungs, and knee joints sampled at termination were analyzed by ELISA for protein quantification. Data represent the group mean ± SD of biological replicates (n = 4–5). m IF stained sections of liver, kidney, lung, and knee joint collected from HB mice 4 days after administration of Padua_KA-HSA_QMP or Padua_KR-HSA_QMP (green, anti-FIX; blue, DAPI for nuclei). Images show one representative animal per test article (scale bar: 100 µm). The fluorescence signal at ten different fields per image was quantified. The data represent the group mean ± SD of technical replicates (n = 10). Statistical significance was tested by unpaired two-tailed Student’s t tests with 95% confidence level (b–g, n = 4 except n = 5 for Padua_KR-HSA_QMP, i-l, n = 4; precise p-values are given in Supplementary Tables 6–8; m ****p < 0.0001 of n = 10). FIX-HSA_WT, black; Padua-HSA_QMP, gray; Padua_KA-HSA_QMP, pink; Padua_KR-HSA_QMP, purple. Source data are provided as a Source Data file. a Created in BioRender. Hovden Aaen, K. (2025) https://BioRender.com/auh41ln.

Fig. 5. The K5R amino acid substitution grants FIX-HSA enhanced activity and extended functional half-life in HB mice.

FIX activity in plasma of HB mice after (a), 1 h, (b), 24 h, and (c), 96 h post-administration of a FIX-HSA fusion, compared to the activity in plasma of mice that received FIX-HSA_WT. Data represent the group mean ± SD of biological replicates (n = 4–5). d Change in FIX clotting activity in plasma over time. Data represent the group mean ± SD of biological replicates (n = 4–5) and are presented as the percentage of that of FIX-HSA_WT (= 100%). e Specific activity measured in mice over time, calculated as the ratio between FIX clotting activity and protein concentrations relative to that of FIX-HSA_WT (= 1). Data represent the group mean ± SD at each time point (n = 5). f Change in FIX activity in plasma over time. Data represent the group mean ± SD of biological replicates (n = 4–5) and are presented as the level of clotting activity over time compared to day 1 (= 100%). g Functional half-life of the FIX-HSA variants measured by gPKPDsim. Data represent the group mean ± SD of biological replicates (n = 4–5). h Correlation between MRT values obtained by applying the FIX plasma concentration (x-axis) and the FIX clotting activity (y-axis) to gPKPDsim. Data represent the group mean ± SD of biological replicates (n = 4–5) and were applied to a simple linear regression. i Inverse correlation between Padua_KR-HSA_QMP activity and hemoglobin loss upon tail clipping at day 4 post-administration. Data represent the group mean ± SD of biological replicates (n = 5) and were applied to a simple linear regression. Statistical significance was tested by unpaired two-tailed Student’s t tests with 95% confidence level (a–d, f–h, n = 4 except n = 5 for Padua_KR-HSA_QMP; e n = 5; precise p-values are given in Supplementary Tables 9–11). FIX-HSA_WT, black; Padua-HSA_QMP, gray; Padua_KA-HSA_QMP, pink; Padua_KR-HSA_QMP, purple. Source data are provided as a Source Data file.