Chapter 1: Current state-of-the-art
The facilitator of entry: equilibrative nucleoside transporter 2 (ENT2)
The equilibrative nucleoside transporter 2 (ENT2), also known as SLC29A2, is highly expressed in human skeletal muscle tissues1. In particular, ENT2 is nitrobenzylmercaptopurine riboside (NBMPR)-insensitive (concentrations up to 1 μM) transporter important in the first step of exogenous nucleosides and nucleobases2 salvage pathway, i.e., their transport across plasma membrane1. Thus, ENT2 is important for obtaining of exogenous nucleic acid precursors by the cell. ENT2 has 4-fold higher affinity for exogenous inosine compared to another, NBMPR-sensitive (0.1–1 nM concentration range), ENT1 nucleoside transporter, also known as SLC29A11. The uptake of uridine is mediated by both ENT1 and ENT2 with roughly the same efficiency. However, only ENT2-mediated uridine influx is inhibited by hypoxanthine nucleobase3. These findings suggested that human ENT2 might be important in adenosine and its metabolites transport in skeletal muscle tissues1,3,4.
Anti-DNA antibody fragment-driven ENT2-dependent protein transduction
MRL-lpr /lpr mice harboring spontaneous autosomal recessive lymphoproliferation (lpr ) mutation, resulting in massive T-cells proliferation and high levels of anti-DNA antibodies, develop early onset systemic lupus erythematosus (SLE)-like syndrome5. Many hybridomas secreting monoclonal anti-DNA antibodies were derived by fusing the MRL-lpr /lpr mice spleen cells and various myeloma cells (e.g. , Sp2/0-Ag14, MPC-11, FOX-NY, etc.). 3E10 represents one of these anti-DNA monoclonal antibodies (mAbs)6.
It was demonstrated that antigen-binding (Fab) and single-chain variable (scFv) fragments of 3E10 mAb, when covalently fused to a protein or an enzyme, facilitate their penetration into cytosol and nuclei of multiple cell types7. Importantly, it was demonstrated that 3E10 scFv penetration into cell nuclei can only be blocked when high (100 μM) NBMPR concentrations are used, prompting that NBMPR-insensitive ENT2 transporter is involved8. Furthermore, ENT2 absence on the cell surface blocked penetration, whereas its reconstitution restored 3E10 scFv transport8. Thus, it was convincingly demonstrated that 3E10 Fab and scFv penetration is ENT2-dependent process. It was hypothesized8 that 3E10 mAb fragments bind nucleosides or nucleobases, which are transported by ENT2 directly into the cytosol.
Anti-DNA antibody fragments target skeletal and heart muscles in vivo
Two facts strongly suggest that in vivo 3E10 will target and deliver its fusion proteins/enzymes specifically into skeletal muscle tissues. First, ENT2 is highly expressed in human skeletal muscles. Second, 3E10 scFv intracellular penetration is ENT2-dependent. This was addressed in 2003, when scFv of 3E10 mAb were systemically injected into tail veins of mice. It was found that 4 hours after injection 3E10 scFv was present in ~20% of the skeletal muscle cells (gastrocnemius) of mice9. The authors stated that infrequently scFv was present in renal tubular cells and heart tissue could not be analyzed, whereas 3E10 scFv was not detected in brain, lung, intestine, spleen, liver, pancreas, ovary, and skin tissues9. The corresponding images, however, were not demonstrated.
In a very recent publication, after systemic intravenous injection, 3E10 Fab fragment fused to human pancreatic α-amylase (VAL-0417 construct, Valerion Therapeutics) substantially decreased the amount of cytosolic insoluble polysaccharides (also known as Lafora bodies) in heart tissue of a mouse model of Lafora neurodegenerative disease10. In the same study, intramuscular (i.m.) injection of VAL-0417 into gastrocnemius muscle resulted in high construct concentration at 2 and 24 h post-injection in the injected muscle, strongly suggesting muscle uptake. Similar approach (i.m. injection) was used by Valerion Therapeutics (previously known as 4s3 Bioscience) to successfully deliver myotubularin in a mouse model of X-linked myotubular myopathy and as a result to improve the skeletal muscle function11. In addition, the same company develops yet another 3E10 Fab and human acid α-glucosidase (GAA) fusion (FabGAA), also known as VAL-1221, for Pompe disease characterized by the accumulation of both lysosomal and cytoplasmic glycogen in both skeletal and heart muscles. Recently, VAL-1221 or its similar construct was used to demonstrate that FabGAA fusion accesses the cytosol directly bypassing the endo-lysosomal pathway thus proving the hypothesis set forth in the previous section12.
Valerion Therapeutics – the origin of intellectual portfolio
The 3E10 mAb was discovered by Dr. Richard H. Weisbart at the University of California, Los Angeles (UCLA). Dr. Weisbart and UCLA filed approximately 30 patent applications (and about half of these applications resulted in awarded patents) related to 3E10 mAb, Fab, Fv, or scFv and their usage for protein transduction. Valerion Therapeutics licensed the UCLA intellectual portfolio most likely before 2012. Subsequently, Valerion Therapeutics filed at least 30 patent applications (including under the name 4s3 Bioscience) covering the usage of 3E10 mAb and its fragments further.
The “Internalizing moieties” (US10221250B2) patent by Valerion Therapeutics covers in detail the final amino acid sequence of the 3E10 mAb and its variants (single amino acid substitutions). In particular, Weisbart and colleagues found that a single amino acid mutation (D31N, Asp→Asn) in the complementarity-determining region 1 (CDR1) of variable heavy (VH) chain of 3E10 mAb increased its binding to DNA and cellular penetration7. Virtually all other modifications described in US10221250B2 patent are in the framework (FR) regions of 3E10 variable regions, hence addressing the humanization of mAb.
In this Solution, NextGenRnD reveals the connection between DNA recognition, cell penetration, and skeletal muscle targeting/binding processes. This enabled the rational design of at least 32 novel potential skeletal muscle-targeting moieties, which were generated using targeting moiety minimization approach. The absolute majority of these original peptide targeting moieties has a molar mass of approximately 5,000 g/mol (5 kDa). These targeting moieties should be applicable for skeletal muscle delivery of antibodies, proteins, enzymes, peptides, antisense oligonucleotides, mRNA, and DNA. Neither of these targeting moieties is patent protected. Neither of these targeting moieties is based on 3E10 mAb. Finally, we provide strategies for further optimization of the targeting moieties specified.