Factor IX Padua: them that have, give

In this issue of Blood, Finn et al have taken a factor IX variant with increased specific activity associated with thrombophilia and used it to improve gene therapy of hemophilia B in dogs,¹  and Cantore et al have shown similar results in mice.² 

Historically, gene therapy for hemophilia B has been an attractive avenue for clinical research. The F9 gene is cloned, the amount of protein expression required to ameliorate the disease is reasonable, there are many suitable vectors available, and excellent animal models are available. However, there remain obstacles that obstruct the goal of cure for hemophilia B. The adeno-associated virus (AAV) vectors most commonly used now can be difficult to make at pharmacologic scale. Perhaps half of the vector particles employed in clinical trials may be empty capsids devoid of “payload” (DNA containing the transgene of interest; R. J. Samulski, personal communication, August 30, 2012). Pre-existing immunity may hinder efficient gene transfer in some patients, and the highly evolved immune response to any viral-based vector makes repeat gene transfer problematic.

Despite these problems, current vectors with optimized promoter/regulatory elements, and part of an intron, have brought about clinically significant expression of factor IX in hemophilia B patients enrolled in various clinical trials of gene transfer to muscle or liver. The amount of factor IX that has been stably expressed in such patients has been on the order of approximately 1% to 2% for muscle-directed therapy³  and 2% to 11% for liver-directed therapy,⁴  which should convert severe hemophilia B to mild or moderate forms of the disease.

The amount of tissue (muscle or liver) subject to gene transfer with current vectors obviously cannot be increased, and the efficiency of gene transfer is not likely to improve much in the foreseeable future. Modifications of promoters and regulatory sequences to improve transcription have probably gone as far as they can, and optimization of codon usage likewise has been employed to increase the efficiency of translation. So, where to find the next increments of improvement?

The gain-of-function mutation Factor IX Padua may present the opportunity for further much-needed improvements. Factor IX Padua is a spontaneous mutation found in a thrombophilic family where arginine 338 in the factor IX protein is changed to leucine.⁵  Affected individuals have factor IX protein with > 8 times the normal specific activity, presumably with more efficient generation of thrombin. This is most unfortunate for the affected family members, to be sure; however Finn et al used the sequence of the abnormally potent factor IX for construction of improved gene therapy vectors to treat hemophilia B.¹  They created a canine factor IX in which the arginine 338 was converted to leucine to simulate the human factor IX Padua protein, and generated an AAV-6 vector expressing this canine factor IX Padua protein. They tested the construct in hemophilia B dogs by regional delivery to muscle tissue, resulting in canine factor IX activity of 3.5% to 8% after 100 days. Within a week of gene transfer, the whole blood clotting time (which is quite sensitive to factor IX levels) was normalized. More importantly, spontaneous bleeding (typically 5 or more events per year) stopped in the treated dogs. Cantore et al (also in this issue) report on lentivirus vectors with codon-optimized canine factor IX cDNAs, and the Arg338Leu mutation, with a liver-specific promoter.²  When administered to hemophilia B mice, the resultant factor IX antigen levels were 8% of normal for the wild-type canine factor IX construct, and 20% for the codon-optimized construct. The Arg338Leu mutation further increased the specific activity > 5-fold, resulting in total factor IX activity of 125% in mice treated with constructs containing optimized codons and the factor IX Padua mutation. Factor IX activity of > 400% was seen when codon optimized, factor IX Padua mutation human factor IX vectors were tested in mice.² 

Questions remain to be answered. For instance, any change in the amino acid sequence may increase the risk of inhibitor antibody development, as has been demonstrated in the normal factor VIII variants with different propensity for inhibitors, despite normal structure and function.⁶  Finn et al have done some preliminary work to investigate the immunogenicity of the factor IX Padua protein that is encouraging¹ ; however, the number of dogs tested was too small to draw definite conclusions. Inhibitors to factor IX were not seen in the mice treated by Cantore et al,²  but most mouse strains are not especially prone to make antibodies to human factor IX.⁷  It is likely that gene transfer with the factor IX Padua construct early in life would lead to immune tolerance, but in an adult hemophilia B patient extensively treated with normal factor IX, would the alteration at amino acid 338 prove to be immunogenic?

There also is the theoretical potential for thrombogenicity because of enhanced coagulation activity might prove to be thrombogenic; after all, the factor Padua kindred came to attention because of thrombophilia.⁵  However, current gene transfer methods are unlikely to result in uniform transduction of enough cells to make this a serious concern, at this time.

The factor IX Padua construct could be employed in human clinical trials where muscle is targeted, as was done in the first hemophilia B gene therapy trial,³  or it might be used for liver-directed therapy. Both muscle-targeted and liver-target gene therapy approaches have their theoretical and practical advantages.^3,4  A high-activity factor IX construct could bring the effective yield of factor IX from muscle-targeted gene transfer up to the levels achieved by liver-targeted therapy as done by Cantore et al.²  In fact, a phase 1/2 clinical trial is approved to begin testing safety, efficacy, and optimal dose of an AAV serotype 8 vector containing the factor IX Padua construct to be given intravenously, and which will presumably transduce hepatocytes.⁸ 

Interestingly, even before the natural occurrence of the factor IX Padua, in vitro mutagenesis of the arginine at amino acid 338 to alanine was shown to have 3-fold increased specific activity in vitro compared with wild-type factor IX, because of improved factor X binding and thereby increased catalytic activity.⁹  In the case of factor IX Padua, it may be time to take from those who have too much, and give to those with too little.