Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Mar 1.
Published in final edited form as: J Thromb Haemost. 2023 Nov 26;22(3):633–644. doi: 10.1016/j.jtha.2023.11.010

Humanization and functional characterization of enhanced coagulation factor IX variants identified through ancestral sequence reconstruction

Christopher W Coyle 1, Kristopher A Knight 1, Harrison C Brown 2, Stephan N George 2, Gabriela Denning 2, Gianna M Branella 3, Kenneth C Childers 4, P Clint Spiegel 4, H Trent Spencer 5, Christopher B Doering 5,†
PMCID: PMC10922771  NIHMSID: NIHMS1949528  PMID: 38016519

Abstract

Background:

Laboratory resurrection of ancient coagulation factor IX (FIX) variants generated through ancestral sequence reconstruction (ASR) led to the discovery of a FIX variant, designated An96, that possesses enhanced specific activity independent of, and additive to that provided by human p.Arg384Lys, referred to as FIX-Padua.

Objective:

The goal of the current study was to identify the amino acid substitution(s) responsible for the enhanced activity of An96 and create a humanized An96 FIX transgene for gene therapy application.

Methods:

Reductionist screening approaches, including domain swapping and scanning residue substitution, were employed and guided by one-stage FIX activity (OSA) assays. In vitro characterization of top candidates included recombinant high purity preparation, specific activity determination and enzyme kinetic analysis. Final candidates were packaged into adeno-associated viral (AAV) vectors and delivered to hemophilia B mice.

Results:

Five of 42 total amino acid substitutions in An96 appear sufficient to retain the enhanced activity of An96 in an otherwise human FIX variant. Additional substitution of the Padua variant further increases the specific activity 5-fold. This candidate, designated ET9, demonstrates 51-fold greater specific activity than hFIX. AAV2/8-ET9 treated hemophilia B mice produced plasma FIX activities equivalent to those observed previously for AAV2/8-An96-Padua, which were 10-fold higher than AAV2/8-hFIX-Padua.

Conclusions:

Starting from computationally inferred ancient FIX sequences, novel amino acid substitutions conferring activity enhancement were identified and translated into an AAV-FIX gene therapy cassette demonstrating high potency. This ASR discovery and sequence mapping refinement approach represents a promising platform for broader protein drug and gene therapy candidate optimization.

Keywords: factor ix, protein engineering, gene therapy, Hemophilia B, Adeno-Associated Virus

Introduction

Factor IX (FIX) is a zyomgen of a vitamin K-dependent serine protease critical to the coagulation network. Deficiencies in FIX activity result in hemophilia B, an X-linked recessive genetic disorder with a worldwide annual incidence of 1 in 25,000 male births [1]. Historically, the treatment paradigm for hemophilia B has involved intravenous infusions of plasma-derived or recombinant FIX to prevent and/or treat bleeding episodes. However, in 2023 the first gene therapy for hemophilia B was approved [2]. This product consists of an adeno-associated viral (AAV) vector encoding a codon-optimized variant of human FIX (hFIX), termed hFIX-Padua, driven by a liver-specific promoter [3]. The Padua variant was not discovered by rational drug design, but instead was identified serendipitously in a family in Padua, Italy presenting with inherited thrombophilia linked to a single nucleotide change in the F9 gene (c. 31134G>T) that results in the FIX variant NP_000124.1:p.Arg384Leu, herein referred to as “Padua.” [4]. It was determined that this single amino acid substitution increases the specific activity of activated hFIX (hFIXa) 7 – 8-fold. Currently, all clinical stage AAV gene therapies for hemophilia B utilize this variant as it was shown to elevate FIX activity levels from the moderate to low mild hemophilia B ranges (achieved with wild-type hFIX in 1st generation AAV gene therapies) to the upper mild to supraphysiologic ranges in 2nd generation gene therapies depending on the dose, AAV serotype used, and transgene cassette configuration [5, 6]. However, global utilization may be limited as the first approved AAV-FIX gene therapy comes at a price tag of $3.5 million dollars in the United States [7].

Other than the inclusion of FIX-Padua in hemophilia B gene therapies, all other approved bioengineered FIX products involve addition of protein sequences or chemical conjugates that increase the circulating half-life of hFIX (e.g., immunoglobulin Fc-domain, albumin, or glycopegylation) [8]. None of these enhancements are being pursued in clinical gene therapy trials to increase vector potency and thereby reduce vector requirements. However, progress towards further enhancing FIX activity has been demonstrated in preclinical studies where specific substitutions or combinations of substitutions including p.Val132Ala, p.Glu323Ala, p.Arg364Tyr, p.Arg384Ala/Glu/Leu, and p.Tyr389Arg have been shown to improve the specific activity of recombinant and/or gene therapy produced hFIX [9–11]. Building off our hypothesis that most protein drug and gene therapy transgene product performances can be enhanced above that of native human sequences, our group has taken an evolutionary approach utilizing the ancestral sequence reconstruction (ASR) platform to discover sequence enhancements existing in nature [12]. This approach allows for rapid scanning of the evolutionary landscape in terms of both ortholog diversity and natural selection trajectories. Recently, we reported our initial ASR findings for coagulation factor VIII (FVIII), FIX, and von Willebrand factor (VWF) [13–15]. Subsequently, others have applied the ASR approach to pharmaceutically relevant entities beyond coagulation factors such as Cas9 and base editors [16–19]. In each of our studies, we identified sets of amino acid substitutions that confer enhancement to the biosynthetic efficiency, specific activity and/or half-lives of these critical coagulation factors. Within each inferred ASR variant exists positive, neutral, and potentially negatively acting substitutions. Thus, additional mapping studies are required to identify the specific substitutions that are necessary and sufficient to confer the functional enhancements observed in initial ASR screens. Ideally, next-generation protein drugs and gene therapies will contain the fewest alterations possible from the human protein thereby minimizing the risk of immunogenicity and neutralizing anti-drug antibodies.

FIX-An96 is a representative early mammalian FIX estimated to have been in existence prior to the Cretaceous-Paleogene boundary [15]. Our studies of laboratory resurrected FIX-An96 demonstrate the variant to have a specific activity enhancement comparable to FIX-Padua without a substitution at p.Arg384Leu. Instead, FIX-An96 possesses 42 non-human substitutions spread over each FIX domain [signal peptide(sp) – propeptide – γ-carboxyglutamic acid-rich domain (Gla) – epidermal growth factor-like (EGF) domain 1 (EGF1) – EGF domain 2 (EGF2) – linker – activation peptide (ap) – protease (pro)]. The mechanism of FIX-An96 enhanced specific activity was shown to be independent of, and additive to, p.Arg384Leu as recombinant An96-Padua displayed a specific activity of 6,450 IU/mg, which is 3.2-fold greater than that of hFIX-Padua (2013 IU/mg), and 3.1-fold greater than that of An96 (2090 IU/mg). This enhancement also translated to higher in vivo plasma FIX activities following AAV2/8-FIX-An96 ± Padua delivery [15]. In the current study, we sought to map the critical substitutions necessary and sufficient to confer FIX-An96 comparable activity in a humanized FIX variant predicted to be safer for clinical use.

Materials & Methods

Engineering of human/ancestral hybrid FIX expression plasmids

All cDNA constructs were purchased from Integrated DNA Technologies (Coralville, IA) as gblocks gene fragments. They were subcloned into a liver-directed expression AAV cassette [20, 21] and/or pcDNA3.4 plasmid using XhoI and NotI restriction sites. Single amino acid variants were created utilizing overlap extension PCR [22] with primers according to Supplementary Table S4. Plasmid identity was confirmed with Sanger DNA sequencing (Azenta Life Sciences, Piscataway, New Jersey).

In vitro expression and testing of FIX constructs

FIX expression plasmids were transfected into low passage Huh7 cells (Creative Biolabs, Shirley, NY) with TransIT-X2 (Mirus Bio, Madison, Wisconsin) in antibiotic-free medium supplemented with 15μg/mL vitamin K1 (Sigma Aldrich, St. Louis, MO), according to the Mirus Bio instructions. Twenty-four hours post transfection, wells were washed with Dulbecco’s phosphate buffered saline (DPBS; Gibco – Thermo Fisher Scientific, Waltham, MA) and switched to 0.5mL of Freestyle 293 medium (Gibco) supplemented with 15μg/mL vitamin K1 . After an additional 24 hrs, aliquots of conditioned medium were collected and analyzed via one stage coagulation assay (OSA) as described below. Data were normalized to activity present in the conditioned media from hFIX expressing cells and referred to as ‘relative FIX expression rate’.

Purification of FIX variants

Expi293F Cells (Gibco) were transiently transfected with linearized FIX transgene cassettes in the pcDNA3.4 backbone. Two days following transfection, cells were moved to a T75 flask and geneticin G417 sulfate was added at a concentration of 400μg/mL until antibiotic selection was complete (~2 weeks). Polyclonal selected cell lines were then expanded into Erlenmeyer Flasks in Freestyle F-17 medium (Gibco) with 100μg/mL G418 and 15μg/mL vitamin K. Conditioned medium was collected every 48–72 hrs, subjected to centrifugation at 1,250 x g for 15 min, filtered through a 0.22μM polyethersulfone (PES) membrane and stored at −20°C with 0.02% NaN3. FIX proteins were purified as previously described [15], but briefly involved an initial capture step using a Capto MMC resin (Cytiva, Marlborough, Massachusetts) and final polishing step using a RESOURCE Q column with a CaCl2 elution gradient [23, 24]. FIX concentrations were determined using absorbance at 280nm and an estimated molar extinction coefficient. For biochemical studies of hFIX, the commercial product, BeneFIX (Pfizer, New York, NY), was utilized. FIX activity measurement and thrombin generation assays of purified proteins were performed as previously described [15].

FIX activation

FIX zymogen was mixed with 1:100 to 1:1000 ratios of hFXIa (Prolyix, Essex Junction, Vermont) for 2–3 hrs at 37°C. FXIa was removed via immunoprecipitation with an anti-FXIa antibody (GMA-070, Green Mountain Antibodies, Burlington, Vermont) and protein G-agarose beads (Pierce Protein Biology, ThermoFisher Scientific, Waltham, Massachusetts).

FIXa enzyme kinetic studies with peptide substrate

Spectrozyme FIXa substrate was purchased from BioMedica Diagnostics (Windsor, NS, Canada). For experiments where the substrate concentration was kept constant, 1 mM of substrate was mixed with 0–800nM in the recommended buffer (50 mM TRIS, 100 mM NaCl, 5 mM CaCl2, pH 7.4, with 33% (v/v) ethylene glycol) and immediately put on a plate reader spectrometer to measure the change in absorbance at 405 nm (Δ405 nm). Initial velocities were obtained using Softmax Pro 4.3.1 Software (Graphpad Software, San Diego, CA).

Tenase enzyme kinetic studies

Recombinant proteins were purchased from Prolytix (Essex Junction, Vermont). Protein purity was assessed by SDS-PAGE and Coomassie-blue visualization. For FVIII/FVIIIa, Advate (Takeda, Lexington, MA), a recombinant full-length FVIII was utilized. PCPS lipids were made as previously described [25]. FVIII (5 nM) was activated with 50 nM of thrombin for 30 sec, followed by thrombin inhibition with recombinant hirudin (75 nM) for 15 sec. FVIIIa (1 nM final concentration) was added to a mixture of FIXa (0.1 nM final concentration) and PCPS lipids (20 μM) in HBS buffer with 10mM CaCl2. After allowing the FXase complex to form for 15 sec, FX was added (various concentrations) and aliquots of the reaction were quenched. The slope of each FXa generation curve was plotted against the FX concentration used and MichaelisMenten kinetic parameters were fit using GraphPad Prism. Experiments where the FIXa-FVIIIa Kd was measured were performed as described above, except 400nM of FX was used, FVIIIa concentration was varied from 0–40nM, and the hirudin concentration was a 1.5x molar excess of the FVIII concentration.

FIXa modeling

Molecular modeling was initially performed with AlphaFold [26]. The model was subsequently aligned with a previously described model of the intrinsic FXase complex based on the homologous prothrombinase complex (PDB code: 7TPP) [27, 28]. Structural figures of the lead candidate, designated ET9, were generated with PyMOL Molecular Graphics System, Version 2.0 (Schrödinger).

Adeno-associated viral vector production and testing

FIX transgenes were liver codon optimized (LCO) as previously described [20] and subcloned into an AAV transfer plasmid utilizing AAV2 ITRs. LCO FIX transgenes genes were under the direction of an HHS4-TTR enhancer-promoter and included a minute virus of mouse (MVM) intron, and a minimal synthetic β-globin polyadenylation sequence. Plasmids were transformed into Stbl3 competent cells and following monoclonal colony selection, grown to a volume of 500mL in terrific broth with 200 µg/mL ampicillin. Plasmids were prepared using a Qiagen plasmid Mega kit (Qiagen, Germantown, MD) and sequence was confirmed with Sanger sequencing. Additionally, intact ITR sequences were confirmed by SmaI/AhdI restriction enzyme (New England Biolabs, Ipswich, MA) digestion and ITR Sanger sequencing (Azenta Life Sciences, Piscataway, New Jersey). Plasmids were then packaged into single-stranded AAV8-capsid viral vectors by Charles River Laboratories (Rockville, MD). Viral vectors were titered using qPCR with QuantStudio3 Real-Time PCR system and PowerUP SYBER Green (Applied Biosystems, Waltham, MA) and interpolated from a plasmid standard.

Male hemophilia B mice (exon 1–3 disrupted, Jackson Laboratory strain #004303), aged 7 to 14 weeks, were injected intravenously into the lateral tail vein with varying doses of AAV (previously diluted into a volume of 100 µL of PBS with 0.001% pluronic F-68). Blood was collected from the retro-orbital plexus into 0.38% (m/v) sodium citrate (final concentration). Plasma was isolated and FIX activity determined by OSA as described above.

Statistical analysis

All data analysis, statistical testing and graph creation were performed using GraphPad Prism 9.5.1 software (GraphPad Software, San Diego, CA).

Results

Mapping functional residues necessary for FIX-An96 enhanced activity

FIX-An96 possesses 42 amino acid substitutions with respect to hFIX [15]. We hypothesized that not all 42 substitutions are required to maintain the enhanced specific activity of FIX-An96. As an initial screen for the domain locations of the functional substitutions, a panel of hybrid FIX-An96/hFIX molecules were generated and cloned into a mammalian expression plasmid under the control of a synthetic liver-specific promoter (Figure 1A). The resulting molecules were tested in vitro for expression and activity by transient transfection into low passage Huh-7 cells and assay of the conditioned media by FIX OSA (Figure 1B). This assay allows for the detection of hybrid proteins with more efficient FIX biosynthesis (i.e., protein translation and/or secretion) or enhanced specific activity, but it does not distinguish between the two possibilities. Of the initial 14 hybrids generated, only 3 (FIX-108, FIX-109, and FIX-112) displayed FIX activity in the conditioned medium at levels non-inferior to FIX-An96. The common FIX-An96 domains specific to these 3 variants are the EGF2 and protease domains, suggesting that each domain contains amino acid substitutions necessary for the activity enhancement observed for FIX-An96.

FIGURE 1.

FIGURE 1

Open in a new tab

Anccstral/human factor IX (hFIX) hybrid schematics and expression. (A) Schematics of the 17 initial An96-hFIX hybrids generated are shown. Blue regions indicate the hFIX sequence, and gold regions represent the factor (F)IX-An96 sequence. The asterisks represent addition of the p.Val132Ala variant. The numbers in parentheses are the numbers of amino acid substitutions with respect to the hFIX sequence. (B) Relative FIX activities were obtained by first measuring FIX activity production rates (international units per 106 cells per 24 hours) using timed media collection, measurement of FIX activity by 1-stage coagulation assay, and cell counting within each well. The rates were then normalized to the values obtained for hFIX and presented as “Relative FIX Activity.” The dashed horizontal line denotes the relative expression rate for FIX-An96 (positive control). (C) FIX expression rates were similarly assessed for the C-terminal protease domain hybrids again relative to hFIX. Error bars indicate sample SD. Statistical analysis was performed using a 1-way analysis of variance with post hoc Tukey multiple comparisons test. Asterisks denote hybrids with significant expression below FIX-An96, where *P < .033, **P < .002, ***P < .0002, and ****P < .0001. Three or more independent experiments with 2 technical replicates each were performed in B and C. EGF, epidermal growth factor; ns, not significant.

Within the EGF2 domain, there are 5 amino acid substitutions, one of which was previously described (p.Val132Ala) to confer a specific activity enhancement [10]. Of note, many extant mammalian species outside of the human/primate lineage and including rodents and ungulates (hoofed mammals) naturally share alanine at this position. Constructs containing only the FIX-An96 N-terminal (FIX-115) or C-terminal (FIX-114) halves of the protease domain were included to interrogate this largest domain at higher resolution (Figure 1A). While neither FIX-114 nor FIX-115 matched FIX-An96 performance in the transient transfection assay, FIX-114 displayed partial enhancement, possibly limited only by the lack of p.Val132Ala. Therefore, three additional constructs were generated to reintroduce p.Val132Ala and interrogate protease subdomain substitutions within the FIX-An96 C-terminal half: FIX-114 + p.Val132Ala (FIX-201), N-terminal FIX-An96 subdomain of FIX-114 + p.Val132Ala (FIX-202), and C-terminal An96 subdomain of FIX-114 + p.Val132Ala (FIX-203) (Figure 1A). FIX-201 displayed enhancement indistinguishable from the parent FIX-An96 variant. However, both FIX-202 and FIX-203 demonstrated < 50% of the parent FIX-An96 enhancement suggesting that gain-of-function substitutions are present within each subdomain of the C-terminal half of the FIX-An96 protease domain (Figure 1C).

Collectively, there are 11 amino acid substitutions in these two subdomains relative to hFIX. Of note, none of these substitutions occur at Arg384. To identify the functional FIX-An96 protease domain substitutions, a negative screen was performed whereby single amino acids in FIX-201 were substituted for their hFIX counterpart anticipating that substitution away from gain of function residues would result in a decrease in FIX enhancement compared to the parent molecule (Supplementary Table S1 and Figure 2). Six of the 11 FIX-An96 C-terminal protease domain substitutions reduced the FIX enhancement: p.Lys323Glu, p.Trp326Val, p.Asn338Asp, p.Arg339Lys, p.Arg362Lys, and p.Ser367Leu. The other five substitutions (p.Ser313Asn, p.Asn361His, p.Ile368Val, p.Tyr399Phe, and p.Lys404Arg) conferred no significant effect.

FIGURE 2.

FIGURE 2

Open in a new tab

Negative single amino acid substitution screen of factor (F) IX-201. FIX expression rates of single amino acid variants listed in Supplementary Table S1 were determined as described in Figure 1. Brackets indicate the parental hybrid from which each subvariant was derived. Error bars indicate sample SD. Statistical analysis was performed using a 1-way analysis of variance with post hoc Tukey multiple comparisons test. Asterisks denote hybrids with significant expression below An96, where *P < .033, **P < .002, ***P < .0002, and ****P < .0001. Three independent experiments with 2 technical replicates each were performed. hFIX, human factor IX; ns, not significant.

Having observed the substitutions within FIX-201 to be necessary for the FIX-An96 enhancement, a final panel of constructs was generated. Each construct contained p.Val132Ala plus at least one An96 C-terminal protease domain substitution (Supplementary Table S2). Although each of these humanized FIX-An96 constructs possessed enhancement significantly above hFIX, the two variants that contained only 2 or 3 substitutions in the C-terminal protease domain (designated FIX-401 and FIX-402) displayed reduced enhancement compared to FIXAn96 (Figure 3A). However, the two constructs that contained 4 and 5 substitutions within the C-terminal protease domain (designated FIX-403 and FIX-404, respectively) demonstrated enhancement indistinguishable from FIX-An96 revealing that p.Val132Ala, p.Glu323Lys, p.Asp328Asn, p.Lys362Arg, and p.Leu367Ser (FIX-403) are necessary and sufficient to recreate the enhancement of FIX-An96 in a molecule with 99% identity to hFIX (Figure 3A).

FIGURE 3.

FIGURE 3

Open in a new tab

Human factor IX (hFIX)-An96 constructs ± Padua. Factor (F)IX expression rates were determined as described in Figure 1. (A) The data for constructs possessing the native p.Arg384, and (B) the data from related constructs containing the p.Arg384Leu (Padua) substitutioa Error bars indicate sample SD. Statistical analysis was performed using a 1-way analysis of variance with post hoc Tukey multiple comparisons test. Asterisks denote hybrids with significant expression below (A) An96and (B) An96-Padua. where *P < .033, **P < .002, ***P < .0002, and ****P < .0001. Three independent experiments with 2 technical replicates each were performed. ns, not significant.

Addition of Padua Variant (Arg384Leu) to FIX-403

Previously, we demonstrated the specific activity enhancement conferred by the Padua variant is additive to the enhancement present in FIX-An96. To confirm that this observation remained valid for FIX-403, p.Arg384Leu was introduced into FIX-403 to create a construct designated ET9. Similar to our previous findings for FIX-An96 ± Padua, ET9 demonstrated 4 – 5-fold greater enhancement over hFIX-Padua or FIX-An96 and equivalent enhancement to that described previously for FIX-An96-Padua (Figure 3B) [15].

In order to address the possibility of increased FIX biosynthesis (transcription, translation, and secretion) as the mechanism driving enhanced FIX activity measurements, an ELISA was performed on conditioned media samples from transiently transfected Huh7 (Supplementary Figure S1A) and HEK293T/17 (Supplementary Figure S1B) cells. While FIX-403 showed moderately (< 2-fold) higher FIX antigen concentrations compared to hFIX, hFIX-Padua, and ET9 when expressed from Huh7 cells, the antigen levels of all four constructs were equivalent when expressed from HEK293T/17 cells. These data do not support the hypothesis that the 5- and 25-fold enhanced relative activities of FIX-403 and ET9 observed in the transient transfection assay (Figure 3B) result from more efficient biosynthesis.

Procoagulant activities of recombinant FIX variants

To further interrogate the mechanism(s) driving FIX-An96, FIX-403, and ET9 activity enhancement, stable polyclonal cell populations expressing each FIX variant were generated, conditioned media was collected, and high purity recombinant FIX preparations were generated as described previously [15]. Subsequently, the specific activities of each preparation were determined from OSA measurements of serial dilutions of each recombinant FIX preparation (Supplementary Table S3 and Figure 4A). Nearly parallel activity curves were observed for each variant supporting the validity of the activity comparisons. As expected, overlapping curves were generated for the parental:humanized progeny pairs, An96:FIX-403 and An96-Padua:ET9, again supporting the conclusion that the full activity enhancement was conferred by identified residues. Both FIX-An96 and FIX-403 display specific activities (2,182 and 1,681 IU/mL, respectively) equivalent to hFIX-Padua (2,108 IU/mg). Substitution of p.Arg384Leu into FIX-403 (i.e., ET9) further increased the specific activity to 10,644 IU/mg, which is indistinguishable to that of An96-Padua (10,553 U/mg). Notably, ET9 specific activity is 5-fold greater than that of the hFIX-Padua, and 51-fold higher than commercial recombinant hFIX.

FIGURE 4.

FIGURE 4

Open in a new tab

One-stage coagulation assay activity curves and thrombin generation assays for recombinant factor (F)IX variants ± Padua. (A) Dilutions (0.001–6.8 nM) of FIX variants were added to FIX-dcficient plasma, and time to fibrin clot formation was measured by one-stage coagulation assay. Slopes were calculated by linear regression of data obtained from 3 independent experiments (R2 > 0.98). Error bars indicate sample SD. Slopes (95% Cl) for human FIX (hFIX), hFIX-Padua, FIX-403, An96, An96-Padua, and ET9 were −17.54 (−18.52 to −16.57), −17.34 (−17.99 to −16.69), −18.53 (−19.04 to −18.03), 16.15 (−16.80 to −15.49), −17.12 (−17.78 to −16.46), and −16.93 (−17.94 to −15.93), respectively. (B) Endogenous thrombin potential (ETP) and (C) peak thrombin as determined by thrombin generation assay. Data points indicate the average of 3 independent experiments. Error bars indicate SD. WT, wild-type.

The thrombin generation assay (TGA) represents a complementary and more comprehensive assessment of procoagulant activity as it takes place over a significantly longer timescale than the OSA and produces a multiparameter readout including a time-dependent thrombin generation curve (thrombogram), peak thrombin concentration, and endogenous thrombin potential (ETP). TGA parameter comparisons of recombinant FIX variants added to hemophilia B plasma revealed similar relative procoagulant activity trends as those observed by OSA with one striking difference, which was consistently higher procoagulant activity of ET9 compared to An96Padua which was even more pronounced at the lowest FIX concentrations. While in the OSA, ET9 and An96Padua were identical, in the TGA, the following FIX activity ranking was observed: ET9 > An96Padua > FIX-403 ≈ An96 ≈ hFIX-Padua > hFIX (Figure 4B-C and Supplementary Figure S2).

FIX-403 and ET9 enzyme kinetics

Others have shown that the specific activity enhancement conferred by hFIX-Padua (p.Arg384Leu) as well as p.Arg384Ala, is not apparent in FIXa chromogenic peptide substrate assays [11, 29]. Similarly, activated hFIX, hFIX-Padua, FIX-403, and ET9 all displayed similar reaction velocities towards a synthetic FIXa substrate (Supplementary Figure S3). Next, the activities of activated FIX-403 and ET9 were compared to hFIXa and hFIXa-Padua using a reconstituted FXase assay with purified components (Figure 5A). In this assay, proteolytic cleavage of a chromogenic FXa substrate is utilized as an indirect measurement of FXase product (FXa) accumulation, which can be interpolated from a standard curve generated using known concentrations of highly purified FXa and saturating concentrations of Xa substrate. Applying the Michaelis-Menten kinetic model, specific parameters, Km and kcat are estimated and presented in Table 1. This analysis reveals that the ASR identified amino acid substitutions appear to operate through enhancement of substrate turnover with FIX-403a displaying a kcat 3.1 times higher than hFIXa (7.67 s−1 versus 2.44 s−1, respectively) and 1.4 times higher than hFIXa-Padua (5.49 s−1). ET9a displays the greatest kcat enhancement at 12.17 s−1, which is 5-fold greater than hFIXa and 2.2 and 1.6-fold greater than hFIXa-Padua and FIX-403a, respectively. Similar to previous observations, substitution of the Padua mutation variant results in Km elevations [29] of 2.2 and 1.2-fold, respectively, for hFIXa-Padua and ET9a. Catalytic efficiency estimates (kcat/Km) of hFIXa and hFIXa-Padua were similar with FIX-403a and ET9a being 1.5 and 2.1-fold higher, respectively.

FIGURE 5.

FIGURE 5

Open in a new tab

Tenaso activity and factor (R)VIIIa binding among FIXa variants. Using purified components, FXase complexes were assembled and analyzed for (A) FIXa kinetic parameters as well as (B) FVIIIa binding. Data points represent values obtained from 2 independent experiments. (A) Lines represent the Michaelis- Menten fit of the data. R2 values of 0.79. 0.91,0.94, and 0.92 were obtained for human FIX (hFIX), hFIX-Padua. FIX-403, and ET9, respectively. (B) Lines represent one-site total binding fit of the data. R2 values of 0.79, 0.96, 0.98, and 0.98 were obtained for hFIXa, hFIXa-Padua FIXa-403, and ET9a, respectively.

Table 1:

Michaelis-Menten kinetic parameters of hyperactive variants in the purified tenase assay

FIXa variant Km (nM) 95% CI kcat (s−1) 95% CI Vmax (nM/min) kcat/Km (M−1s−1) KDapp (nM) 95% CI
hFIX 23.7 13.1 to 40.6 2.44 2.17 to 2.73 14.61 1.03×108 7.80 1.76 to 125.3
hFIX-Padua 52.9 35.6 to 77.7 5.49 5.02 to 6.04 32.94 1.04×108 1.08 0.59 to 2.00
FIX-403 47.6 34.6 to 65.0 7.67 7.14 to 8.27 46.03 1.61×108 2.08 1.37 to 3.21
ET9 56.1 38.4 to 80.9 12.17 11.1 to 13.4 73.01 2.18×108 0.26 0.20 to 0.34
Open in a new tab

Similar to the p.Arg384Leu and p.Arg384Ala variants previously described, the activity enhancements observed for FIX-403 and ET9 also appear to require FVIIIa [11, 29]. The initial step in FVIIIa cofactor activity involves FVIIIa·FIXa complex formation driven by direct affinity of the two macromolecules. Since FXase activity is directly proportional to the concentration of FVIIIa·FIXa complex and addition of FVIIIa results in a >1000-fold increase in FXase reaction rate, the binding affinity of FIXa for FVIIIa can be estimated by monitoring FXa generation over a range of FVIIIa (or FIXa) concentrations as previously described [11, 29]. To determine the binding affinity of ASR-based variants enhancement for FVIIIa, a similar modified FXase assay design was utilized (Figure 5B). Using this approach, apparent Kd (Kdapp) estimates obtained for hFIXa, hFIXa-Padua, FIXa-403, and ET9a were: 7.80, 1.08, 2.08, and 0.26 nM, respectively. Consistent with the relative procoagulant activities of the FIX variants, the relative order of FVIIIa binding affinities as well as the magnitude of Kdapp differences follow a similar trend suggesting a role for improved FVIIIa binding contributing to the activity enhancement.

Molecular modeling of FIX-403 non-human substitutions

To structurally rationalize the FIX-403 non-human substitutions, we queried the five amino acid residue changes, along with the Padua variantusing AlphaFold [26]. The resultant model was subsequently superimposed with a previously modeled FXase complex (Figure 6A), which is a model that was initially generated through sequence homology to the recently determined prothombinase complex and further refined for energetic minimization and conformational docking, along with the addition of the PPACK FIXa inhibitor (Figure 6B) [27]. The sole light chain substitution is p.Val132Ala, which removes hydrophobicity from the EGF1 and EGF2 interface. Within the catalytic domain, the remaining substitutions are spread throughout the FIXa structure and do not make direct contact with the protease active site. The substitutions, p.Glu323Lys, p.Lys362Arg, and p.Leu367Ser are all solvent-exposed; both p.Lys362Arg and p,Leu367Ser are adjacent to a putative allosteric network of amino residues connecting the Ca2+ loop to the substrate binding pocket [30]. In contrast to the solvent-exposed substitutions, p.Asp338Asn is partially buried in the structure of the catalytic domain and resides on the N-terminus of an alpha helix that packs against the EGF2 domain.

FIGURE 6.

FIGURE 6

Open in a new tab

Model of the FXaso complex with ET9. (A) Model of factor VIII (Al, slate; A2, cyan; A3, dark blue; Cl, dark teal; C2, light teal) bound to ET9 (heavy chain [HC], tan; light chain [LCJ. orange). (B) Model of ET9 HC highlighting An96 substitutions (green spheres), the Padua variant (magenta sphere), Ca2+ ion (yellow sphere), putative allosteric network (blue), and substrate inhibitor PPACK (white sticks).

In vivo testing of ET9

Previously, we showed that the enhanced activity of An96 measured in vitro using OSA and TGA translated to 60- and 6-fold in vivo potency enhancement over hFIX and hFIX-Padua, respectively [15]. In the current study, we utilized identical methodology, up-and-down staircase method combined with a saphenous vein bleed challenge, to determine the 50% effective dose (ED50) of ET9. Consistent with our previous findings for An95-Padua and hFIX-Padua, in the current study, ET9 displayed an ED50 10-times lower than hFIX-Padua (0.256 µg/kg and 2.34 µg/kg, respectively) (Supplementary Figure S4). These results further validate the translational potential of ET9 in clinical protein replacement and gene therapy applications.

To investigate the in vivo performance of FIX-403 and ET9 in a gene therapy setting, liver-directed and codon-optimized transgene cassettes were packaged into AAV8 capsid vector particles and injected into hemophilia B mice at varying doses. Bi-weekly post-AAV-FIX injection, blood was collected, and FIX activity determined by OSA and compared to our previously published data for AAV-hFIX-Padua , AAV-An96, and AAV-An96-Padua (Figure 7). Consistent with our in vitro observations, in vivo performance of the AAV2/8-ET9 vector across dose range between 109 – 1012 gc/kg was overlapping with that observed previously for AAV2/8-An96-Padua. At the highest dose tested (4.1 × 1011 gc/kg), mean AAV2/8-ET9 activity was 8.7 IU/mL, which is ~7-higher than observed previously for AAV2/8-hFIX-Padua at a dose of 5 × 1011 gc/kg (1.2 IU/mL). Furthermore, AAV2/8-ET9 achieved plasma FIX activity levels similar to the latter vector at a dose level 25-times lower (0.79 IU/mL at 2 × 1010 gc/kg).

FIGURE 7.

FIGURE 7

Open in a new tab

Comparison of adeno-associated virus (AAV) 2/8 vectors encoding factor (F)IX transgenes in vivo. Comparison of FIX variant AAV2 (inverted terminal repeat) and 8 (capsid) 12 weeks after AAV administration at varying doses. Data points represent the group mean, and error bars represent sample SD (n = 3–6 mice/ group). hFlX. human factor IX.

Discussion

In gene therapy candidates, the importance of potency as a critical attribute has grown in appreciation as clinical data continue to accumulate. Observations of vector related toxicities and the relatively slow progress made towards overcoming manufacturing challenges such as scale limitations and cost of goods remain major commercialization challenges [31–33]. Currently, there are three primary approaches to improving vector potency, 1) increasing gene transfer efficiency, 2) optimizing transgene mRNA transcription and/or stability, and 3) enhancing the performance of the transgene product through protein bioengineering. AAV-FIX gene therapies represent the state of the art in terms of optimizations at all three levels, i.e., capsid serotype transition from AAV2 to AAV3, 5, 8 and variants thereof, promoter/transgene (synthetic liver-directed promoters and codon optimized FIX transgenes with reduced CpG content), and transgene product engineering (inclusion of the Padua variant) [10, 15, 20, 34–37]. Together, these modifications appear to be critical to the clinical successes observed.

The first approved AAV-FIX gene therapy for hemophilia B is the most expensive single drug developed to date [38]. One strategy for reducing cost is the development of more potent transgene cassettes that allow for lower dosing and therefore more doses produced per manufacturing campaign. Our team has focused on the development of bespoke optimization strategies for gene therapy transgene cassettes at the nucleic acid and amino acid levels [13, 15, 20, 21, 39]. For the latter, we have explored extant and ancient (inferred) interspecies diversity as a discovery method followed by identification of functional sequences and engineering them into the relevant human transgene products. Although this approach comes with a theoretical risk of increased immunogenicity, we and others have not observed this outcome in preclinical and clinical studies to date. In contrast, existing data frequently support the notion that leading gene therapy approaches, such as liver-directed AAV gene therapy and hematopoietic stem cell lentiviral gene therapy, promote immune tolerance to the transgene products (for review see Patel et al.) [40, 41]. Therefore, a priori, and consistent with the outcomes observed using the Padua variant, protein engineering need not be considered off-limits during the development of future gene therapy candidates.

Previously, we utilized the ASR approach to identify FIX variants with enhanced properties and discovered a top variant, FIX-An96 that displayed 11-fold higher specific activity than hFIX. Addition of the Padua variant to FIX-An96 (An96-Padua) further boosted the specific activity to a level nearly 60-fold higher than hFIX, likely representing the most active FIX variant described to date [8] . In the current study, we systematically identified the amino acids necessary and sufficient for the enhancement observed in FIX-An96 by removing them from FIX-An96 and subsequently engineering them into hFIX. The resulting construct, FIX-403, possessed 5 non-human amino acid substitution located in two domains; EGF2 (p.Val132Ala) and protease (p.Glu323Lys, p.Asp338Asn, p.Lys362Arg, and p.Leu367Ser). Variant p.Val132Ala has been described previously by Chang et al [42] in a comprehensive EGF1/2 domain alanine scanning experiment and shown to increase kcat in the presence of FVIIIa. Interestingly, p.Val132Ala is naturally present in most non-human mammals and the majority of ancestral FIX variants inferred during our previous study [15]. Despite being categorized as nonhuman substitutions, p.Val132Ala and p.Glu323Lys have a minor allele frequencies of 1.09 ×10−5 and 8.53 ×10−4, respectively [43, 44]. Additionally, Lin et al described the combinations of p.Val132Ala with p.Glu323Ala and p.Arg384Ala/Leu having the additive benefits of each individual residue. It is intriguing that using orthologous approaches (i.e., alanine scanning and ASR) two common residues, p.Val132 and p.Glu323, were identified as positions where substitutions result in gain of function. These data combined with the identification and characterization of the Padua gain of function substitution by Stafford, Simioni, Arruda and Samelson-Jones further support the hypothesis that FIX activity has been suppressed by negative selection during human evolution [4, 11, 45]. Furthermore, ET9 builds upon the TripleL (p.Val132Ala/Glu323Ala/Arg384Leu) variant described by Kao et al to have a 22-fold specific activity difference (4308.8 ± 182.0 versus 197.9 ± 8.7 IU/mg) by OSA and antigen ELISA [46]. Through modification of p.Glu323Lys, as opposed to alanine in TripleL, and addition of the p.Asp338Asn, p.Lys362Arg, and p.Leu367Ser substitutions, ET9 specific activity, as determined by clotting assay and spectrophotometric protein determination, increased another 2.5-fold. Kao et al also showed that p.Val132Ala/Glu323Ala/Arg384Ala (or Leu) enhances FIXa performance by both increasing the apparent affinity for FVIIIa (10-fold decrease in Kdapp) and increasing kcat (<3-fold) [8, 46]. Kinetic studies of hFIXa-403 and ET9a uncovered a further increase in kcat (~5-fold increase) with little alteration of Km other than the consistent increase with addition of p.Arg384Leu as has been observed previously [11, 29, 46]. Molecular modeling was performed to provide speculative mechanistic information as follows. p.Val132Ala possibly disrupts the Van der Waals packing at the hydrophobic interface between the EGF1 and EGF2 domains. Valine is considerably more hydrophobic than alanine [47] and has substantially more surface area, which likely accounts for the relatively large effect observed by us and others. p.Asp338Asn may destabilize the alpha helix that packs against the EGF2, as the negative charge of the carboxyl group of aspartic acid serves as a helix capping residue that stabilizes the N-terminal positive charge of a helix macrodipole. Both of these substitutions may function by lowering the energetic barrier to putative conformational changes required for catalytic activity upon FVIIIa binding. p.Glu323Lys is completely exposed on the opposing face of the FVIIIa/FIXa interface and thus may be involved in electrostatically steering FIXa towards FVIIIa binding. Tighter binding between FIXa and FVIIIa results in higher FVIIIa-FIXa complex concentrations, thus potentially improving FX turnover and is the proposed mechanism by which p.Arg384Leu exerts its effects [29]. As stated above, p.Lys362Arg and p.Leu367Ser substitutions reside in the region of the FIXa catalytic domain proposed to possess a signal transduction network of interactions connecting the Ca2+ loop to the substrate binding pocket [30]. p.Lys362Arg may form a salt bridge with p.Glu434 stabilizing the sodium binding loops and the active site, possibly rectifying a known shortcoming of FIXa: a poorly formed active site [48–50]. p.Lys362Arg is also located on the autolysis loop predicted to regulate binding to FX binding and activation. Previous studies have shown p.Lys362Ala to reduce FX activation [51]. Lastly, p.Leu367Ser may stabilize the Ca2+ active conformation and signal relay to the active site. Here, the p.Leu367Ser substitution forms a putative hydrogen bond with p.Glu375 located on the Ca2+ loop.

These ASR substitutions, along with our previously described enhanced ancestral FVIII (termed An53) [13] are predicted to have existed along a common ancestral lineage ~90 million years ago. Therefore, it is reasonable to speculate that the ancient mammalian hemostatic system may have required more active secondary hemostasis components. This corresponds to a pre-platelet, thrombocyte era where primary hemostasis was presumably less efficient and more efficient secondary hemostasis perhaps advantageous. The development of invasive placentation, unique to mammals, and evolution towards live birth has been proposed as the driving force for platelet evolution [52]. By providing a highly effective primary hemostatic mechanism, platelets combined with higher activity coagulation factors may have tipped the hemostatic balance towards thrombosis overtaking as the dominant selective force. If accurate, one would predict to observe a gradual dampening of secondary hemostatic efficiency through the stepwise introduction of amino acid substitutions that confer reduced activity in critical coagulation factors, such as FVIII and FIX. Although this type of evolutionary mechanistic speculation is intriguing, no such understanding is required to utilize ASR to explore the diversity and capture functionality evolution has achieved for the purpose of designing improved protein drugs and gene therapies. The current study provides additional proof of concept for this promising drug development approach.

Supplementary Material

1

Acknowledgements

This work was supported by funding from the National Institutes of Health: National Heart, Lung and Blood Institute (HL137128 and U54 HL141981 to H.T.S. and C.B.D, R15HL135658 and U54HL141981 to P.C.S.), the National Hemophilia Foundation Judith Graham Pool Postdoctoral Research Fellowship to K.C.C., and Hemophilia of Georgia, Gene Therapy Program grant to H.T.S. and C.B.D. Graphic in Figure 1A was created with Biorender.com.

Footnotes

Conflict of Interest Disclosures

H.C.B., G.D., S.N.G., C.B.D., H.T.S., K.A.K. and C.W.C. are inventors on patents and patent applications describing the ancestral FIX technology filed by Expression Therapeutics, Emory University, Children’s Healthcare of Atlanta, and Georgia Institute of Technology. C.B.D., H.T.S. and H.C.B. are inventors on liver-directed codon-optimization and promoter technology filed by Emory University and Children’s Healthcare of Atlanta. H.T.S. and C.B.D. are cofounders of Expression Therapeutics, Inc. and own equity in the company. H.C.B., G.D. and S.N.G. are employees of Expression Therapeutics, Inc. and own equity in the company. Expression Therapeutics, Inc. has obtained licenses for FIX-An96, liver codon optimized FIX, and synthetic liver-directed promoter intellectual property. K.C.C., P.C.S., and G.M.B., declare no conflicts of interest. The terms of these arrangements have been reviewed and approved by Emory University in accordance with its conflict of interest policies.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Iorio A, Stonebraker JS, Chambost H, Makris M, Coffin D, Herr C, Germini F. Establishing the Prevalence and Prevalence at Birth of Hemophilia in Males: A Meta-analytic Approach Using National Registries. Ann Intern Med. 2019; 171: 540–6. 10.7326/m19-1208. [DOI] [PubMed] [Google Scholar]
  • 2.Heo YA. Etranacogene Dezaparvovec: First Approval. Drugs. 2023; 83: 347–52. 10.1007/s40265-023-01845-0. [DOI] [PubMed] [Google Scholar]
  • 3.Pipe SW, Leebeek FWG, Recht M, Key NS, Castaman G, Miesbach W, Lattimore S, Peerlinck K, Van der Valk P, Coppens M, Kampmann P, Meijer K, O’Connell N, Pasi KJ, Hart DP, Kazmi R, Astermark J, Hermans C, Klamroth R, Lemons R, Visweshwar N, von Drygalski A, Young G, Crary SE, Escobar M, Gomez E, Kruse-Jarres R, Quon DV, Symington E, Wang M, Wheeler AP, Gut R, Liu YP, Dolmetsch RE, Cooper DL, Li Y, Goldstein B, Monahan PE. Gene Therapy with Etranacogene Dezaparvovec for Hemophilia B. N Engl J Med. 2023; 388: 706–18. 10.1056/NEJMoa2211644. [DOI] [PubMed] [Google Scholar]
  • 4.Simioni P, Tormene D, Tognin G, Gavasso S, Bulato C, Iacobelli NP, Finn JD, Spiezia L, Radu C, Arruda VR. X-linked thrombophilia with a mutant factor IX (factor IX Padua). N Engl J Med. 2009; 361: 1671–5. 10.1056/NEJMoa0904377. [DOI] [PubMed] [Google Scholar]
  • 5.George LA. Factor IX Padua for haemophilia B gene addition: universal adaptation and repeated success. Lancet Haematol. 2022; 9: e465–e6. 10.1016/s2352-3026(22)00178-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Samelson-Jones BJ. Worldwide use of factor IX Padua for hemophilia B gene therapy. Mol Ther. 2022; 30: 2394–6. 10.1016/j.ymthe.2022.06.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Naddaf M. Researchers welcome $3.5-million haemophilia gene therapy - but questions remain. Nature. 2022; 612: 388–9. 10.1038/d41586-022-04327-7. [DOI] [PubMed] [Google Scholar]
  • 8.Samelson-Jones BJ, Arruda VR. Protein-Engineered Coagulation Factors for Hemophilia Gene Therapy. Mol Ther Methods Clin Dev. 2019; 12: 184–201. 10.1016/j.omtm.2018.12.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Nichols TC, Levy H, Merricks EP, Raymer RA, Lee ML. Preclinical evaluation of a next-generation, subcutaneously administered, coagulation factor IX variant, dalcinonacog alfa. PLoS One. 2020; 15: e0240896. 10.1371/journal.pone.0240896. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Kao CY, Lin CN, Yu IS, Tao MH, Wu HL, Shi GY, Yang YL, Kao JT, Lin SW. FIX-Triple, a gain-of-function factor IX variant, improves haemostasis in mouse models without increased risk of thrombosis. Thromb Haemost. 2010; 104: 355–65. 10.1160/th09-11-0792. [DOI] [PubMed] [Google Scholar]
  • 11.Chang J, Jin J, Lollar P, Bode W, Brandstetter H, Hamaguchi N, Straight DL, Stafford DW. Changing residue 338 in human factor IX from arginine to alanine causes an increase in catalytic activity. J Biol Chem. 1998; 273: 12089–94. 10.1074/jbc.273.20.12089. [DOI] [PubMed] [Google Scholar]
  • 12.Lazarus RA, Scheiflinger F. Mining ancient proteins for next-generation drugs. Nat Biotechnol. 2017; 35: 28–9. 10.1038/nbt.3762. [DOI] [PubMed] [Google Scholar]
  • 13.Zakas PM, Brown HC, Knight K, Meeks SL, Spencer HT, Gaucher EA, Doering CB. Enhancing the pharmaceutical properties of protein drugs by ancestral sequence reconstruction. Nat Biotechnol. 2017; 35: 35–7. 10.1038/nbt.3677. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Zakas PM, Coyle CW, Brehm A, Bayer M, Solecka-Witulska B, Radford CE, Brown C, Nesbitt K, Dwyer C, Kannicht C, Spencer HT, Gaucher EA, Doering CB, Lillicrap D. Molecular coevolution of coagulation factor VIII and von Willebrand factor. Blood Adv. 2021; 5: 812–22. 10.1182/bloodadvances.2020002971. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Knight KA, Coyle CW, Radford CE, Parker ET, Fedanov A, Shields JM, Szlam F, Purchel A, Chen M, Denning G, Sniecinski RM, Lollar P, Spencer HT, Gaucher EA, Doering CB. Identification of coagulation factor IX variants with enhanced activity through ancestral sequence reconstruction. Blood Adv. 2021; 5: 3333–43. 10.1182/bloodadvances.2021004742. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Blazeck J, Karamitros CS, Ford K, Somody C, Qerqez A, Murray K, Burkholder NT, Marshall N, Sivakumar A, Lu WC, Tan B, Lamb C, Tanno Y, Siddiqui MY, Ashoura N, Coma S, Zhang XM, McGovern K, Kumada Y, Zhang YJ, Manfredi M, Johnson KA, D’Arcy S, Stone E, Georgiou G. Bypassing evolutionary dead ends and switching the rate-limiting step of a human immunotherapeutic enzyme. Nat Catal. 2022; 5: 952–67. 10.1038/s41929-022-00856-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Hendrikse NM, Sandegren A, Andersson T, Blomqvist J, Makower Å, Possner D, Su C, Thalén N, Tjernberg A, Westermark U, Rockberg J, Svensson Gelius S, Syrén PO, Nordling E. Ancestral lysosomal enzymes with increased activity harbor therapeutic potential for treatment of Hunter syndrome. iScience. 2021; 24: 102154. 10.1016/j.isci.2021.102154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Koblan LW, Doman JL, Wilson C, Levy JM, Tay T, Newby GA, Maianti JP, Raguram A, Liu DR. Improving cytidine and adenine base editors by expression optimization and ancestral reconstruction. Nat Biotechnol. 2018; 36: 843–6. 10.1038/nbt.4172. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Alonso-Lerma B, Jabalera Y, Samperio S, Morin M, Fernandez A, Hille LT, Silverstein RA, Quesada-Ganuza A, Reifs A, Fernández-Peñalver S, Benitez Y, Soletto L, Gavira JA, Diaz A, Vranken W, Sanchez-Mejias A, Güell M, Mojica FJM, Kleinstiver BP, Moreno-Pelayo MA, Montoliu L, Perez-Jimenez R. Evolution of CRISPR-associated endonucleases as inferred from resurrected proteins. Nat Microbiol. 2023; 8: 77–90. 10.1038/s41564-022-01265-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Brown HC, Zakas PM, George SN, Parker ET, Spencer HT, Doering CB. Target-Cell-Directed Bioengineering Approaches for Gene Therapy of Hemophilia A. Mol Ther Methods Clin Dev. 2018; 9: 57–69. 10.1016/j.omtm.2018.01.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Brown HC, Doering CB, Herzog RW, Ling C, Markusic DM, Spencer HT, Srivastava A, Srivastava A. Development of a Clinical Candidate AAV3 Vector for Gene Therapy of Hemophilia B. Hum Gene Ther. 2020; 31: 1114–23. 10.1089/hum.2020.099. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Heckman KL, Pease LR. Gene splicing and mutagenesis by PCR-driven overlap extension. Nat Protoc. 2007; 2: 924–32. 10.1038/nprot.2007.132. [DOI] [PubMed] [Google Scholar]
  • 23.Kovnir SV, Orlova NA, Shakhparonov MI, Skryabin KG, Gabibov AG, Vorobiev, II. A Highly Productive CHO Cell Line Secreting Human Blood Clotting Factor IX. Acta Naturae. 2018; 10: 51–65. [PMC free article] [PubMed] [Google Scholar]
  • 24.Ribeiro DA, Passos DF, Ferraz HC, Castilho LR. Anion-exchange purification of recombinant factor IX from cell culture supernatant using different chromatography supports. J Chromatogr B Analyt Technol Biomed Life Sci. 2013; 938: 111–8. 10.1016/j.jchromb.2013.09.002. [DOI] [PubMed] [Google Scholar]
  • 25.Parker ET, Doering CB, Lollar P. A1 subunit-mediated regulation of thrombin-activated factor VIII A2 subunit dissociation. J Biol Chem. 2006; 281: 13922–30. 10.1074/jbc.M513124200. [DOI] [PubMed] [Google Scholar]
  • 26.Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, Tunyasuvunakool K, Bates R, Žídek A, Potapenko A, Bridgland A, Meyer C, Kohl SAA, Ballard AJ, Cowie A, Romera-Paredes B, Nikolov S, Jain R, Adler J, Back T, Petersen S, Reiman D, Clancy E, Zielinski M, Steinegger M, Pacholska M, Berghammer T, Bodenstein S, Silver D, Vinyals O, Senior AW, Kavukcuoglu K, Kohli P, Hassabis D. Highly accurate protein structure prediction with AlphaFold. Nature. 2021; 596: 583–9. 10.1038/s41586-021-03819-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Childers KC, Peters SC, Spiegel PC Jr. Structural insights into blood coagulation factor VIII: Procoagulant complexes, membrane binding, and antibody inhibition. J Thromb Haemost. 2022; 20: 1957–70. 10.1111/jth.15793. [DOI] [PubMed] [Google Scholar]
  • 28.Ruben EA, Summers B, Rau MJ, Fitzpatrick JAJ, Di Cera E. Cryo-EM structure of the prothrombinprothrombinase complex. Blood. 2022; 139: 3463–73. 10.1182/blood.2022015807. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Samelson-Jones BJ, Finn JD, George LA, Camire RM, Arruda VR. Hyperactivity of factor IX Padua (R338L) depends on factor VIIIa cofactor activity. JCI Insight. 2019; 5. 10.1172/jci.insight.128683. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Zögg T, Brandstetter H. Structural basis of the cofactor- and substrate-assisted activation of human coagulation factor IXa. Structure. 2009; 17: 1669–78. 10.1016/j.str.2009.10.011. [DOI] [PubMed] [Google Scholar]
  • 31.Ertl HCJ. Immunogenicity and toxicity of AAV gene therapy. Front Immunol. 2022; 13: 975803. 10.3389/fimmu.2022.975803. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Shupe J, Zhang A, Odenwelder DC, Dobrowsky T. Gene therapy: challenges in cell culture scaleup. Curr Opin Biotechnol. 2022; 75: 102721. 10.1016/j.copbio.2022.102721. [DOI] [PubMed] [Google Scholar]
  • 33.Michael Meagher P, Krishnan Mani, Davies Claire, PhD Uncharted Territory: Top Challenges Facing Gene Therapy Development. Genetic Engineering & Biotechnology News: Mary Ann Liebert, Inc. Publishers, 2021. [Google Scholar]
  • 34.Nair N, De Wolf D, Nguyen PA, Pham QH, Samara-Kuko E, Landau J, Blouse GE, Chuah MK, VandenDriessche T. Gene therapy for hemophilia B using CB 2679d-GT: a novel factor IX variant with higher potency than factor IX Padua. Blood. 2021; 137: 2902–6. 10.1182/blood.2020006005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Ran G, Chen X, Xie Y, Zheng Q, Xie J, Yu C, Pittman N, Qi S, Yu FX, Agbandje-McKenna M, Srivastava A, Ling C. Site-Directed Mutagenesis Improves the Transduction Efficiency of Capsid Library-Derived Recombinant AAV Vectors. Mol Ther Methods Clin Dev. 2020; 17: 545–55. 10.1016/j.omtm.2020.03.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Kumar SRP, Xie J, Hu S, Ko J, Huang Q, Brown HC, Srivastava A, Markusic DM, Doering CB, Spencer HT, Srivastava A, Gao G, Herzog RW. Coagulation factor IX gene transfer to non-human primates using engineered AAV3 capsid and hepatic optimized expression cassette. Mol Ther Methods Clin Dev. 2021; 23: 98–107. 10.1016/j.omtm.2021.08.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Quade-Lyssy P, Abriss D, Milanov P, Ungerer C, Königs C, Seifried E, Schüttrumpf J. Next generation FIX muteins with FVIII-independent activity for alternative treatment of hemophilia A. J Thromb Haemost. 2014; 12: 1861–73. 10.1111/jth.12731. [DOI] [PubMed] [Google Scholar]
  • 38.Naddaf M. $3.5-Million Hemophilia Gene Therapy Is World’s Most Expensive Drug. Scientific American, 2022. [Google Scholar]
  • 39.Doering CB, Denning G, Dooriss K, Gangadharan B, Johnston JM, Kerstann KW, McCarty DA, Spencer HT. Directed engineering of a high-expression chimeric transgene as a strategy for gene therapy of hemophilia A. Mol Ther. 2009; 17: 1145–54. 10.1038/mt.2009.35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Patel SR, Lundgren TS, Spencer HT, Doering CB. The Immune Response to the fVIII Gene Therapy in Preclinical Models. Front Immunol. 2020; 11: 494. 10.3389/fimmu.2020.00494. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Samelson-Jones BJ, Arruda VR. Translational Potential of Immune Tolerance Induction by AAV Liver-Directed Factor VIII Gene Therapy for Hemophilia A. Front Immunol. 2020; 11: 618. 10.3389/fimmu.2020.00618. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Chang YJ, Wu HL, Hamaguchi N, Hsu YC, Lin SW. Identification of functionally important residues of the epidermal growth factor-2 domain of factor IX by alanine-scanning mutagenesis. Residues Asn(89)-Gly(93) are critical for binding factor VIIIa. J Biol Chem. 2002; 277: 25393–9. 10.1074/jbc.M105432200. [DOI] [PubMed] [Google Scholar]
  • 43.Karczewski KJ, Francioli LC, Tiao G, Cummings BB, Alföldi J, Wang Q, Collins RL, Laricchia KM, Ganna A, Birnbaum DP, Gauthier LD, Brand H, Solomonson M, Watts NA, Rhodes D, Singer-Berk M, England EM, Seaby EG, Kosmicki JA, Walters RK, Tashman K, Farjoun Y, Banks E, Poterba T, Wang A, Seed C, Whiffin N, Chong JX, Samocha KE, Pierce-Hoffman E, Zappala Z, O’Donnell-Luria AH, Minikel EV, Weisburd B, Lek M, Ware JS, Vittal C, Armean IM, Bergelson L, Cibulskis K, Connolly KM, Covarrubias M, Donnelly S, Ferriera S, Gabriel S, Gentry J, Gupta N, Jeandet T, Kaplan D, Llanwarne C, Munshi R, Novod S, Petrillo N, Roazen D, Ruano-Rubio V, Saltzman A, Schleicher M, Soto J, Tibbetts K, Tolonen C, Wade G, Talkowski ME, Aguilar Salinas CA, Ahmad T, Albert CM, Ardissino D, Atzmon G, Barnard J, Beaugerie L, Benjamin EJ, Boehnke M, Bonnycastle LL, Bottinger EP, Bowden DW, Bown MJ, Chambers JC, Chan JC, Chasman D, Cho J, Chung MK, Cohen B, Correa A, Dabelea D, Daly MJ, Darbar D, Duggirala R, Dupuis J, Ellinor PT, Elosua R, Erdmann J, Esko T, Färkkilä M, Florez J, Franke A, Getz G, Glaser B, Glatt SJ, Goldstein D, Gonzalez C, Groop L, Haiman C, Hanis C, Harms M, Hiltunen M, Holi MM, Hultman CM, Kallela M, Kaprio J, Kathiresan S, Kim B-J, Kim YJ, Kirov G, Kooner J, Koskinen S, Krumholz HM, Kugathasan S, Kwak SH, Laakso M, Lehtimäki T, Loos RJF, Lubitz SA, Ma RCW, MacArthur DG, Marrugat J, Mattila KM, McCarroll S, McCarthy MI, McGovern D, McPherson R, Meigs JB, Melander O, Metspalu A, Neale BM, Nilsson PM, O’Donovan MC, Ongur D, Orozco L, Owen MJ, Palmer CNA, Palotie A, Park KS, Pato C, Pulver AE, Rahman N, Remes AM, Rioux JD, Ripatti S, Roden DM, Saleheen D, Salomaa V, Samani NJ, Scharf J, Schunkert H, Shoemaker MB, Sklar P, Soininen H, Sokol H, Spector T, Sullivan PF, Suvisaari J, Tai ES, Teo YY, Tiinamaija T, Tsuang M, Turner D, Tusie-Luna T, Vartiainen E, Vawter MP, Ware JS, Watkins H, Weersma RK, Wessman M, Wilson JG, Xavier RJ, Neale BM, Daly MJ, MacArthur DG, Genome Aggregation Database C. The mutational constraint spectrum quantified from variation in 141,456 humans. Nature. 2020; 581: 434–43. 10.1038/s41586-020-2308-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.McVey JH, Rallapalli PM, Kemball-Cook G, Hampshire DJ, Giansily-Blaizot M, Gomez K, Perkins SJ, Ludlam CA. The European Association for Haemophilia and Allied Disorders (EAHAD) Coagulation Factor Variant Databases: Important resources for haemostasis clinicians and researchers. Haemophilia. 2020; 26: 306–13. 10.1111/hae.13947. [DOI] [PubMed] [Google Scholar]
  • 45.Samelson-Jones BJ, Finn JD, Raffini LJ, Merricks EP, Camire RM, Nichols TC, Arruda VR. Evolutionary insights into coagulation factor IX Padua and other high-specific-activity variants. Blood Adv. 2021; 5: 1324–32. 10.1182/bloodadvances.2019000405. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Kao CY, Yang SJ, Tao MH, Jeng YM, Yu IS, Lin SW. Incorporation of the factor IX Padua mutation into FIX-Triple improves clotting activity in vitro and in vivo. Thromb Haemost. 2013; 110: 244–56. 10.1160/th13-02-0154. [DOI] [PubMed] [Google Scholar]
  • 47.Kyte J, Doolittle RF. A simple method for displaying the hydropathic character of a protein. J Mol Biol. 1982; 157: 105–32. 10.1016/0022-2836(82)90515-0. [DOI] [PubMed] [Google Scholar]
  • 48.Neuenschwander PF, Deadmond KJ, Zepeda K, Rutland J. Correlation of factor IXa subsite modulations with effects on substrate discrimination. J Thromb Haemost. 2012; 10: 382–9. 10.1111/j.1538-7836.2011.04605.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Hopfner KP, Brandstetter H, Karcher A, Kopetzki E, Huber R, Engh RA, Bode W. Converting blood coagulation factor IXa into factor Xa: dramatic increase in amidolytic activity identifies important active site determinants. Embo j. 1997; 16: 6626–35. 10.1093/emboj/16.22.6626. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Hopfner KP, Lang A, Karcher A, Sichler K, Kopetzki E, Brandstetter H, Huber R, Bode W, Engh RA. Coagulation factor IXa: the relaxed conformation of Tyr99 blocks substrate binding. Structure. 1999; 7: 989–96. 10.1016/s0969-2126(99)80125-7. [DOI] [PubMed] [Google Scholar]
  • 51.Yang L, Manithody C, Olson ST, Rezaie AR. Contribution of basic residues of the autolysis loop to the substrate and inhibitor specificity of factor IXa. J Biol Chem. 2003; 278: 25032–8. 10.1074/jbc.M302174200. [DOI] [PubMed] [Google Scholar]
  • 52.Martin JF, Wagner GP. The origin of platelets enabled the evolution of eutherian placentation. Biol Lett. 2019; 15: 20190374. 10.1098/rsbl.2019.0374. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1949528-supplement-1.doc (2.8MB, doc)

RESOURCES