Gonzalo Hortelano

Cell encapsulation: Promise and progress

In cell encapsulation, transplanted cells are protected from immune rejection by an artificial, semipermeable membrane, potentially allowing transplantation (allo- or xenotransplantation) without the need for immunosuppression. Yet, despite some promising results in animal studies, the field has not lived up to expectations, and clinical products based on encapsulated cell technology continue to elude the scientific community. This commentary discusses the reasons for this, summarizes recent progress in the field and outlines what is needed to bring this technology closer to clinical application.

Induction of angiogenesis by implantation of encapsulated primary myoblasts expressing vascular endothelial growth factor

We previously demonstrated that intramuscular implantation of primary myoblasts engineered to express vascular endothelial growth factor (VEGF) constitutively resulted in hemangioma formation and the appearance of VEGF in the circulation. To investigate the potential for using allogeneic myoblasts and the effects of delivery of VEGF‐expressing myoblasts to non‐muscle sites, we have enclosed them in microcapsules that protect allogeneic cells from rejection, yet allow the secretion of proteins produced by the cells.

Effects of genetic fusion of factor IX to albumin on in vivo clearance in mice and rabbits.

Individuals with haemophilia B require replacement therapy with recombinant or plasma‐derived coagulation factor IX (fIX). More benefit per injected dose might be obtained if fIX clearance could be slowed. The contribution of overall size to fIX clearance was explored, using genetic fusion to albumin. Recombinant murine fIX (MIX), and three proteins with C‐terminal epitope tags were expressed in HEK 293 cells: tagged MIX (MIXT), tagged mouse serum albumin (MSAT) and MFUST, in which MIX and MSAT were fused in a single polypeptide chain. Proteins MFUST and MIXT were two‐ to threefold less active in clotting assays than MIX. In mice, the area under the clearance curve (AUC) was reduced for MFUST compared with MSAT or plasma‐derived MSA (pd‐MSA); the terminal catabolic half‐life (t0·5) did not differ amongst the three proteins. Two minutes after injection, >40% of the injected MFUST was found in the liver, compared with <10% of either MSAT or pd‐MSA. In rabbits, the AUC for MFUST was reduced compared to MIXT, MSAT, or pd‐MSA, while the t0·5 of the fusion protein fell between that of MIXT and MSAT or pd‐MSA. Similar results were obtained with non‐radioactive fused or non‐fused recombinant human fIX in fIX knockout mice. The clearance behaviour of the fusion protein thus more closely resembled that of fIX than that of albumin despite a modest increase in terminal half‐life, suggesting that fIX‐specific interactions that are important in determining clearance were maintained in spite of the increased size of the fusion protein.

Persistent Delivery of Factor IX in Mice: Gene Therapy for Hemophilia Using Implantable Microcapsules

Severe hemophilia B is a life-threatening, life long condition caused by absence of or defective coagulation factor IX. Gene therapy could provide an alternative treatment to repeated injection of plasma-derived concentrate or recombinant factor IX. We have previously described the use of implantable microcapsules containing recombinant myoblasts to deliver human factor IX in mice. This study reports the generation of improved myoblast-specific expression vectors. Mouse myoblast clones transfected with the various vectors secreted factor IX in vitro, at rates between 70 and 1000 ng/10(6) cells/day. The recombinant myoblast clones were then encapsulated and implanted into mice. Immunocompetent mice implanted with encapsulated myoblasts had up to 65 ng of factor IX per milliliter in their plasma for up to 14 days, after which antibodies to human factor IX became detectable, and this coincided with decreased factor IX in mouse plasma. In immunodeficient mice, however, factor IX delivery was maintained at a constant level for at least 6 weeks (end of experiment). Interestingly, the highest-secreting myoblast clone in vitro did not deliver the highest level of hFIX in vivo. This discrepancy observed between performance in vitro and in vivo may have important implications for the development of gene therapy protocols based on recombinant cells.

Therapeutic levels of human Factor VIII in mice implanted with encapsulated cells: potential for gene therapy of haemophilia A

A gene therapy delivery system based on microcapsules enclosing recombinant cells engineered to secrete a therapeutic protein has been evaluated. The microcapsules are implanted intraperitoneally. In order to prevent cell immune rejection, cells are enclosed in non‐antigenic biocompatible alginate microcapsules prior to their implantation into mice. It has been shown that encapsulated myoblasts can deliver therapeutic levels of Factor IX (FIX) in mice. The delivery of human Factor VIII (hFVIII) in mice using microcapsules was evaluated in this study.

Reduction of GAG storage in MPS II mouse model following implantation of encapsulated recombinant myoblasts.

Hunter syndrome, mucopolysaccharidosis type II (MPS II), is a X‐linked inherited disorder caused by the deficiency of the enzyme iduronate‐2‐sulfatase (IDS), involved in the lysosomal catabolism of the glycosaminoglycans (GAG) dermatan and heparan sulfate. Such a deficiency leads to the intracellular accumulation of undegraded GAG and eventually to a progressive severe clinical pattern. Many attempts have been made in the last two to three decades to identify possible therapeutic strategies for the disorder, including gene therapy and somatic cell therapy.

Permeability of alginate microcapsules to secretory recombinant gene products

Non‐autologous somatic gene therapy is an alternate approach to delivering recombinant gene products through implantation of a “universal” donor cell line engineered to produce a therapeutic gene product. The cells are immunologically isolated by enclosure in immunoprotective microcapsules fabricated from alginate‐poly‐L‐lysine‐alginate. The molecular weight cutoff of these microcapsules was thought to be <100 kd, thus, excluding the immunoglobulins. However, when such microcapsules are fabricated to enclose cells, they show a higher permeability threshold than expected. The secretion rates of recombinant gene products ranging from 21 through 150 to 300 kd (human growth hormone, rat serum albumin, human arylsulfatase A, human immunoglobulin, mouse β‐hexosaminidase, mouse β‐glucuronidase) were similar between the nonencapsulated and encapsulated recombinant cells with the exception of the largest molecular species, the 300‐kd β‐glucuronidase. Its secretion was reduced about eightfold after encapsulation. Increasing the thickness of the membrane by prolonging the coating time with poly‐L‐lysine did not provide a lower molecular weight cutoff. An additional coating with alginate, however, reduced the leakage of the larger molecular species, but the effect was short lived: After 2 weeks in culture, the double‐ and single‐coated microcapsules were equally permeable. Both the increased poly‐L‐lysine and alginate coating were detrimental to the long‐term viability and proliferation of the encapsulated cells. Hence, immunoisolation of encapsulated cells with alginate‐poly‐L‐lysine‐alginate microcapsules cannot provide a molecular weight cutoff below 300 kd.

Sustained and therapeutic delivery of factor IX in nude haemophilia B mice by encapsulated C2C12 myoblasts: concurrent tumourigenesis

This study reports the generation of an immunodeficient murine model for haemophilia B, obtained by breeding factor IX‐deficient mice with an immunodeficient mouse strain, and use of this mouse model to evaluate the long‐term efficacy and safety of a gene therapy strategy for treating haemophilia B. Nude haemophilic mice were implanted with biocompatible microcapsules enclosing recombinant myoblasts secreting human factor IX. The activated partial thromboplastin time (APTT) of plasma of mice thus treated was invariably shortened 3 weeks after microcapsule implantation, and remained shortened for at least 77 days. Shortening of the APTT of the haemophilia mice coincided with the appearance of human factor IX in mice plasmas (up to 600 ng mL−1 on day 77), and normalization of the tail‐bleeding time. Thus, the microencapsulated myoblasts reversed the clinical phenotype of haemophilia B. In contrast, plasmas of immunocompetent haemophilic mice similarly implanted with microcapsules only showed a transient shortening of APTT, and coincident transient delivery of human factor IX antigen. Rapid disappearance of human factor IX from plasmas of immunocompetent mice also coincided with production of antibodies to the human transgene. Significantly, 86% of the nude haemophilia mice developed tumours of myoblast origin. Thus, while this study revealed the feasibility of this gene therapy approach to treat severe haemophilia B, it also highlights the importance of using safer cell lines to prevent tumour development.

An important role for the activation peptide domain in controlling factor IX levels in the blood of haemophilia B mice

The factors responsible for the removal of injected factor IX (fIX) from the blood of individuals with haemophilia B are only partly understood, and may include binding to endothelial or subendothelial sites, passive extravasation related to size or charge, or interactions requiring fIX activation. To investigate these issues, we have produced and characterised recombinant fIX proteins with amino acid changes: delta155-177, an internal deletion which removes most of the activation peptide while retaining the activation cleavage sites; S365A, which inactivates the serine protease activity of fIXa; and K5A, previously shown to eliminate fIX binding of endothelial/subendothelial collagen IV. All proteins were expressed in stably transfected HEK 293 cells, purified by immunoaffinity chromatography, and compared to the wild type HEK 293-derived protein (fIX (WT)). Mutant fIX proteins K5A and delta155-177 exhibited 72 and 202% of the specific activity of fIX (WT), respectively; S365A was without activity. Following intravenous injection in haemophilia B (fIX knockout) mice, recoveries did not differ for fIX (WT) and delta155-177, but were higher for K5A and S365A. The terminal catabolic half-life of delta155-177, alone among the mutants, was increased, by 45% versus fIX (WT). Nine hours post-injection, the observed areas under the clearance curve (AUCs) of delta155-177 and K5, but not S365A, were elevated 2-fold. delta155-177 was equally effective as fIX (WT) in reducing blood loss following tail vein transection in haemophilia B mice. Our results suggest that deletion of the multiple sites of fIX post-translational modification found within the activation peptide eliminated important fIX clearance motifs.

Factors that contribute to the immmunogenicity of therapeutic recombinant human proteins

Use of recombinant human proteins has revolutionized medicine by providing over 200 highly purified hormones and proteins that effectively treat many inherited and acquired peptide hormone and protein deficiencies. With the exception of therapeutic monoclonal antibodies, these biological medicines are synthesized by cultured cells using DNA sequences that would yield proteins with identical amino acid sequences as endogenous human proteins. Therefore, there was the broad expectation that recombinant human biological medicines would be non-immunogenic in patients capable of synthesizing even sub-optimal levels of these therapeutic proteins to which they are innately tolerant. However, the widespread clinical use of recombinant human proteins has demonstrated that nearly all of them are immunogenic. This observation suggests that factors additional to differences in amino acid sequences of endogenous and biotherapeutic proteins contribute to the immunogenicity of therapeutic proteins. The main aim of this review is to summarize some of the factors that are known to contribute to the immunogenicity of recombinant therapeutic proteins.