The Problem
The scientific community originally became interested in the concept of intracellular delivery due to our need to selectively move specific DNA fragments into the cell nucleus. Since then, micro and nanotechnologies for intracellular drug delivery have grown into a market projected to reach a valuation of over $220 billion by 2026. Despite the collective efforts of researchers, however, drug delivery has remain plagued by inefficient translocation of proteins (Zoltek et al., 2023), nucleic acids (Dowdy, 2023) and small molecules (Yazdi and Murphy, 1994) across biological membranes, limiting their application as therapeutic compounds and research tools.
One major source of inefficiency is the internalization of drug molecules by cells through phagocytosis, which traps them in endosomal vesicles where they have no therapeutic effect. These molecules are likely to be degraded in the lysosome or pumped back out of the cell. Improving the rate at which drugs escape from the endosome into the cytoplasm is referred to as the endosomal escape challenge. For small-molecule drugs, up to a million molecules could be pumped out of a cell for every one molecule which escapes the endosome into the cytoplasm (Yazdi and Murphy, 1994). With reference to genetic therapies, endosomes can trap 99% of therapeutically-administered RNA (Dowdy, 2023). Improving the rate at which a drug enters a cell’s cytosol can greatly improve its efficacy. For example, anticancer chemotherapeutics augmented with escape mechanisms can improve survival times of mouse models by up to 66% (Polli et al., 2022).However, the exact mechanisms by which current endosomal escape platforms work are unclear and inconsistent (Qiu et al., 2023). This limits their effective use in drug development, where the ability to rationally design drugs is critical for efficient research.
While one method to combat this intrinsic cellular resistance to drug uptake could be to increase dosage, this results in challenges regarding off-target toxicity. High drug dosages often result in serious side effects, placing a hard limit on the effectiveness with which we can utilise any given drug. For example, doxorubicin is a widely used chemotherapeutic that is dose-limited by its high cardiac toxicity (Gong et al., 2018). While attempts to design prodrug modifications to improve a drug’s selectivity have been successful in specific cases, this approach remains impractical to accomplish for every single drug in existence. We now face a dilemma – we have a myriad of highly potent chemical species that we are unable to functionally deliver into cells, and that we cannot target to the specific tissues that we wish to drug.
Our Solution
We are developing a modular system which addresses both of these issues. Refractile bodies (R bodies) are enormous, stable polymeric proteins which extend like needles in acidic endosomal environments (Polka & Silver, 2016). Previous work has shown that R-body extension is fast, reversible, extremely robust, and tunable by directed evolution (Polka & Silver, 2016). In the wild, they are able to puncture acidic food vacuole of Paramecia eukaryotes, releasing toxins into them. Using R bodies as scaffolds for transporting drugs and triggering extension after internalisation would mechanically create a route for drug release from inside the endosome into the cytosol. Our team has been working on optimising a variety of established biomolecule-compatible chemical conjugation methods to create designer R bodies that are capable of chaperoning the delivery of specific drugs into specific cell types. This has involved the research and development of both site-specific and bio-orthogonal conjugation methods that allows for the attachment of payloads and targeting molecules to the outer surface of the R body. If successful, this would ultimately result in a novel drug delivery system that is both modular (working for a diverse set of drugs) and target specific (able to specifically deliver therapeutics into target tissues with low, or even no off-target consequences). Thus we aim to create a novel modular system capable of delivering diverse drug loads into diverse cell types.
By tackling two major issues of drug delivery, our system will greatly improve the efficiency and potency of many drugs, improving treatment options for many patients. Some uses for such a system could be enhancing chemotherapeutic delivery into cancer cells, or enhancing gene therapy delivery (using a mRNA/DNA cargo) into dysfunctional cell types (Li et al., 2023) — the modularity and adaptability of our system will be an important feature for ensuring that it remains useful as a drug delivery system.
Project Planning
Our team spent around 2 months thoroughly investigating various applications for R bodies. Our ideas ranged wildly from bioMEMS devices to mechanosensory circuits to binary switches for logic gates! Eventually we decided that repurposing the natural function of R bodies was the soundest option. Keep reading to see the thought process that validated our project selection!
Why protein nano and microcarriers are used in drug delivery
There is a pressing need for controlled drug release for sustained treatment of chronic or persistent medical conditions and diseases. Guided drug delivery and in vivo stability is difficult because therapeutic compounds need to survive numerous transport barriers and binding targets throughout the body. Use of nano and microscale protein-based polymers have become increasingly used for drug and vaccine delivery to cross these biological barriers and — through blood circulation — reach their molecular site of action. Compared to synthetic polymers, protein-based polymers have the advantages of good biocompatibility, biodegradability, environmental sustainability, cost effectiveness and availability (Jao et al., 2017). Their degradation products are generally non-toxic and well-tolerated by the body (Jao et al., 2017). Their ease of modification enables enhanced stability or targeting capability, easily tunable physicochemical properties and versatility for a wide range of therapeutic contexts, including targeted drug delivery for cancer therapy, vaccines, and gene therapy. Regular, repetitive polymeric structures are particularly promising solutions as they enable uniform distribution of binding sites for drugs, allowing for higher loading capacities and more controlled release profiles (Zielińska et al., 2023). They also tend to display enhanced cellular uptake by endocytosis (Zielińska et al., 2023). R bodies possess all of these favourable attributes, and so we have envisioned them as an idea solution for a drug delivery system.
Simplicity is beauty - R bodies are nature’s toxin delivery system
In 1938, the existence of ‘killer’ and ‘sensitive’ Paramecia aurelia was discovered (Smith-Sonnenborn & van Wagtendonk, 1964). When killer and sensitive strains are cultured in the same dish, the sensitive individuals die in hours to days. The death of sensitive strains was proven to result from exposure to (and subsequent phagocytosis of) toxic particles, which were identified as Caedibacter spp., bacterial endosymbionts that killer Paramecia strains release from their cytoplasm (Preer et al., 1953). By the 1960s, the cytotoxic properties of Caedibacter bacteria had been traced to two components: first, inclusion bodies termed ‘refractile bodies’ (R-bodies), named for their appearance under phase-contrast microscopy, and second, a toxin presumably localised to the bacterial cytoplasm whose molecular identity is unidentified (Preer et al., 1966; Preer & Preer, 1964; Schrallhammer, 2010).
Electron microscopy studies show that R-bodies are proteinaceous ribbons that exist in a coiled conformation inside bacteria (Preer et al., 1966). The coiled R-body is structured as a hollow layered cylinder which, in response to certain stimuli, can unroll in a telescopic fashion. Acidic pH change triggers R-bodies to unroll after a Caedibacter bacterium is incorporated into a sensitive Paramecium’s phagosome and the phagosome matures functionally into a lysosome. By unrolling, the R-body expands its length over 20 times (up to 20 μm) and ruptures the bacterial cell wall and phagolysosome. The putative toxin contained within the Caedibacter cell is thus released into the sensitive Paramecium’s cytoplasm, leading to the death of sensitive Paramecia (Pond et al., 1989). However the R-body itself is non-cytotoxic (Pond et al., 1989), which was an important nudge along our design pathway.
Before our team committed to repurposing R bodies, we verified that they could satisfy other requisite features of successful drug delivery systems. Most importantly, stability of a protein-based nanocarrier-drug formulation needs to be carefully considered under different thermal, pH, solvent, oxidising or reducing conditions to ensure the formulation remains intact and effective throughout the storage, transport and administration processes (Jao et al., 2017). Uniquely, R bodies are reported to exhibit extreme physical and chemical stability. Early SDS-PAGE studies reported that treatments including combinations of 8 M urea, 10% SDS, 5% 2-mercaptoethanol or dithiothreitol, boiling at 100°C for up to 1 h or 6 M guanidine thiocyanate could not significantly dissociate R bodies into their monomeric constituents (Pond et al., 1989). Presuming R bodies retain their ability to extend under most pharmaceutically-relevant conditions (which we aimed to verify in our project), their intrinsic stability would be a great asset in in vivo applications, transport and purification, especially as most drug delivery biopolymer scaffolds suffer from stability issues (Zielińska et al., 2023).
Nature had created a simple, presumably extremely stable, brute-force protein to break open endosomes as they acidify. The subsequent mechanism of cell damage ostensibly relies only on the delivery of a separate cargo molecule through the lysed endosome. Our team envisioned combining R bodies with a drug payload to tackle the pernicious problem of endosomal escape.
Endocytic pathways for R body uptake
As Paramecia rely on bacteria to translocate R bodies into endosomes, we needed to verify whether isolated, purified R bodies might reach endosomes when present in extracellular fluid. The route of internalisation taken by a micro- or nano-size particle depends primarily on its size, shape, surface charge and surface functionalisation (Polka & Silver 2016; Polli et al., 2022). We assessed whether R bodies have properties amenable to cellular internalisation.
Large particles measuring approximately 500 nm or greater in diameter tend to internalise from the fluid phase into parenchymal cells, such as tumour cells, via macropinocytosis (Cai, 2023). R bodies measure 450 nm by 250 nm when coiled into cylinders, which makes it likely that macropinocytosis will serve as a major endocytic pathway. Macropinocytosis occurs in a constitutive, non-selective manner (Sousa de Almeida et al., 2021) and tends to be upregulated in neoplastic cells (particularly KRAS mutant cells) (Liu & Qian, 2022) which makes this process ideal for R body delivery. However, as macropinocytosis is conserved among all parenchymal cells (Liu & Qian, 2022), our team also had to consider how to limit R body delivery to target sites and minimise off-target effects.
Drug conjugation strategies
We aimed to create a customisable cargo-delivery chassis by exploiting the regular, polymeric arrangement of monomers of our protein.
Structural studies of R bodies have demonstrated that the two monomer peptides, RebA and RebB, are arranged side-by-side in a thin, flat ribbon (Cai, 2023). One face is composed of a lattice of RebA and RebB C-termini, and the other face is composed of a lattice of RebA and RebB N-termini. One must imagine a waffle, where each square of the waffle is the cross-section of a monomer's C-terminus on one side and their N-terminus on the other side.
When the R body ribbon is in its coiled conformation, all monomer C-termini face the lumen and all N’ face the external environment (Cai, 2023). The majority of the peptide is unavailable for chemical reaction as it is sequestered within the thickness of the R body ribbon (Cai, 2023). Thus four classes of conjugation sites exist: RebA C-terminus, RebB C-terminus, RebA N-terminus, and RebB N-terminus.
Our team scoured the literature to select four orthogonal bioconjugation strategies that could be used at these sites. Orthogonality was a key design consideration as it ensures the modularity of our product. Four distinct attachment sites and four versatile attachment strategies allows for customizable configurations where different cargo can be selectively attached to specific sites using any combination of the available strategies. This design offers flexibility in cargo loading and site utilization, enabling various combinations such as using different strategies at different sites, attaching diverse cargo types, or leaving certain sites unmodified as needed.
The table below represents the strategies we selected, and rejected, for this purpose. To streamline future research, entrepreneurship and ethical decisions we aimed to use bioconjugation methods validated in existing FDA-approved drug conjugate products. However, we would like to highlight that we opted to reject several strategies due to time and funding constraints. Those we believe to be feasible future additions to the R body conjugation toolbox are presented below. Navigate to our Engineering page to read about our successes with conjugation!
| Selected Strategies | ||||
|---|---|---|---|---|
| Conjugation Strategy | Sites of Conjugation | R Body Modification | Reason for Selection | |
| Sortase A Enzymatic Conjugation | LPXTG motifs at C’ only of RebA or RebB | Engineering an LPXTG motif into the target monomer peptide sequence | Ease, literature | |
| Non-canonical Amino Acid (ncAA) Incorporation by Amber Codon Suppression | ncAA at N’ or C’ of RebA or RebB | Engineering an UAG amber codon into the target monomer mRNA | ARX517 is in Phase I clinical trials and uses acetyl-phenylalanine, whereas we plan to use azide-phenylalanine | |
| Thiol-Maleimide Conjugation | Cysteine residues at N’ or C’ of RebA or RebB | Engineering a cysteine linker into the target monomer peptide sequence | Trastuzumab duocarmazine (SYD985) is an FDA-approved precedent. | |
| 2-PCA N-terminus Bioconjugation | N’ amines only of RebA or RebB | None | Simultaneous modification of RebA and RebB on one side. This is the least orthogonal strategy as it does not differentiate RebA and RebB on N’ | |
| Tyrosine Click Chemistry | Tyrosine residues at N’ or C’ of RebA or RebB | Engineering a tyrosine linker into the target monomer peptide sequence | ||
| Rejected Strategies | ||||
| Strategy | Reason for Rejection | |||
| Native chemical ligation | Native chemical ligation (NCL) relies on an N terminus catalytic cysteine residue.2 This would interfere with thiol-maleimide orthogonality. NCL could replace thiol-maleimide conjugation if the latter proves unsuccessful. | |||
| Typical split intein conjugation | Split inteins are trans-splicing peptide sequences that spontaneously self-ligate two fragments (IntN and IntC). Despite their utility and ease of use, typical split intein conjugation would require the fusion of an approximately 100 amino acid long IntN fragment to RebA or RebB. As the monomers are only 114 and 105 amino acids long themselves, we were concerned whether the IntN fragment would interfere with assembly and decided not to test with strategy. Typical split inteins also often use a cysteine-based catalytic mechanism which would interfere with thiol-maleimide orthogonality (Elleuche & Pöggeler, 2010; Burton et al., 2020; Wang & Zhang, 2019). Further investigation is needed to confirm whether split inteins might find application R body conjugation. | |||
| Atypical split intein conjugation | Atypical split inteins based on the recently discovered VidaL system have remarkably short IntN fusion fragments of 16 amino acids. They use a serine catalyst, have reaction half-lives below 1 min, show robust performance under various temperatures and salinities, and do not suffer from premature IntC cleavage unlike typical split inteins. However the IntC fragment is 124 amino acids, too long for solid-phase synthesis, therefore necessitating an additional recombinant expression step as well as a large fragment that potentially might interfere with R body extension (Burton et al., 2020; Wang & Zhang, 2019). Further investigation is needed to confirm whether VidaL split inteins might find application R body conjugation. | |||
| SpyTag-SpyCatcher ligation | A versatile, robust and efficient ligation method that spontaneously forms an isopeptide bond between two defined peptides under a range of conditions. The main disadvantage is the method’s directionality, which requires SpyCatcher to be fused to the R body. At 12kDa SpyCatcher is bigger than the RebB or RebA monomers and could interfere with assembly. A drug-SpyTag adduct must also be synthesised for this strategy, which would require a novel in vitro synthesis strategy (Neugebauer et al., 2017). Further investigation is needed to confirm whether SpyTag-SpyCatcher might find application R body conjugation. | |||
| Enzyme-mediated covalent bond formation | We selected Sortase A for enzymatic conjugation due to its widespread usage and extensive literature. However, protein farnesyltransferase (PFTase) and N-myristoyltransferase (NMT) have found increasing application in bioconjugation as alternatives to sortase A. Their key disadvantage is the use of commercially unavailable substrates, albeit relatively simple to synthesize in-house (Stephanopoulos & Francis, 2011). Greater funding and chemical synthesis capacity would enable further investigation into the utility of these enzymes. | |||
Drug release strategies
Next our team contemplated a cleavable linker for drug release within the endosome. We referred to the standard strategies used in antibody-drug conjugates (ADCs) for inspiration (Khongorzul et al., 2020). The kinetics and mechanism of release were the key considerations. As wild-type R bodies abruptly extend and rupture endosomes at pH 5.7 or below (Polka and Silver, 2016), any cleavable linker must decompose within 20 minutes (Mollaev et al., 2019; Wang et al., 2017) before R body extension ruptures the endosome and intra-vesicular acidity or enzyme concentration is diluted. It should be noted, however, that R bodies have been successfully engineered to extend over a range of pH values (Polka and Silver, 2016), which expands the options of cleavable linkers with different kinetic properties that can be utilised.
Our team was able to test one cleavable linker — a hydrazone group found in aldoxorubicin, a maleimide-containing prodrug of doxorubicin that has completed phase 2b clinical trials for treating soft tissue sarcomas (Chawla et al., 2015), and excitingly, found some evidence for endosomal escape, though hydrazone instability may have been the culprit. We also believe cathepsin B-cleavable sequences — a widely-utilised strategy with rapid cleavage at pH 6.0 (Khramtsov et al., 2023) — would be highly compatible with sortase A bioconjugation, and would also be a promising step for future investigation.
Conclusion
R bodies are clearly a promising starting point for researchers seeking easily understandable, easily produced, and easily used tools for circumventing lysosomal degradation of payloads. Despite the clear importance of ensuring that drugs are efficiently taken up by cells to maximise their effect, there remains a gap in the pharmaceutical market for a drug that can do just that. With our experimentation and discussion with stakeholders, it is clear that our low-cost platform for improving endosomal escape will have a place in the pharmaceutical market, and may allow many drugs and types of drugs to more effectively treat patients. Its potential to improve drug potency and efficiency may allow more drugs to be commercially viable, hopefully leading to a wider range of treatments being available to treat a wider range of diseases, which might otherwise not have been developed.
References
Burton, A. J., Haugbro, M., Parisi, E., & Muir, T. W. (2020). Live-cell protein engineering with an ultra-short split intein. Proceedings of the National Academy of Sciences - PNAS, 117(22), 12041–12049. https://doi.org/10.1073/pnas.2003613117
Cai, G. (2023). An Interdisciplinary Investigation of Conformational Changes in Quasi-Crystalline Protein Array R-Bodies in Response to pH [Ph.D., University of Washington]. In ProQuest Dissertations and Theses (2837944634). ProQuest One Academic. https://www.proquest.com/dissertations-theses/interdisciplinary-investigation-conformational/docview/2837944634/se-2?accountid=14757
Chawla SP, Papai Z, Mukhametshina G, et al. First-Line Aldoxorubicin vs Doxorubicin in Metastatic or Locally Advanced Unresectable Soft-Tissue Sarcoma: A Phase 2b Randomized Clinical Trial. JAMA Oncol. 2015;1(9):1272–1280. doi:10.1001/jamaoncol.2015.3101
Elleuche, S., & Pöggeler, S. (2010). Inteins, valuable genetic elements in molecular biology and biotechnology. Applied Microbiology and Biotechnology, 87(2), 479–489. https://doi.org/10.1007/s00253-010-2628-x
Gong, J., Yan, J., Forscher, C., & Hendifar, A. (2018). Aldoxorubicin: a tumor-targeted doxorubicin conjugate for relapsed or refractory soft tissue sarcomas. Drug design, development and therapy, 12, 777–786. https://doi.org/10.2147/DDDT.S140638
Jao, D., Xue, Y., Medina, J., & Hu, X. (2017). Protein-Based Drug-Delivery Materials. Materials, 10(5). https://doi.org/10.3390/ma10050517
Khongorzul, P., Ling, C. J., Khan, F. U., Ihsan, A. U., & Zhang, J. (2020). Antibody-drug conjugates: A comprehensive review. Molecular Cancer Research, 18(1), 3–19. https://doi.org/10.1158/1541-7786.MCR-19-0582
Khramtsov, Y. V., Georgiev, G. P., & Sobolev, A. S. (2023). Selection of an Amino Acid Site with One of the Fastest Cleavage Kinetics by the Endosomal Protease Cathepsin B for Potential Use in Drug Delivery Systems. Doklady. Biochemistry and Biophysics, 509(1), 78–80. https://doi.org/10.1134/S1607672922600221
Li, X., Mahboubeh Kheirabadi, Dougherty, P. G., Kamer, K. J., Shen, X., Estrella, N. L., Suresh Peddigari, Pathak, A., Blake, S. L., Emmanuelle Sizensky, Carmen Del Genio, Gaur, A. B., Mohanraj Dhanabal, Mahasweta Girgenrath, Sethuraman, N., & Qian, Z. (2023). The endosomal escape vehicle platform enhances delivery of oligonucleotides in preclinical models of neuromuscular disorders. Molecular Therapy — Nucleic Acids, 33, 273–285. https://doi.org/10.1016/j.omtn.2023.06.022
Mollaev, M., Gorokhovets, N., Nikolskaya, E., Faustova, M., Zabolotsky, A., Zhunina, O., Sokol, M., Zamulaeva, I., Severin, E., & Yabbarov, N. (2019). Type of pH sensitive linker reveals different time-dependent intracellular localization, in vitro and in vivo efficiency in alpha-fetoprotein receptor targeted doxorubicin conjugate. International Journal of Pharmaceutics, 559, 138–146. https://doi.org/10.1016/j.ijpharm.2018.12.073
Neugebauer, M., Böcker, J., Matern, J., Pietrokovski, S. & Mootz, H. (2017). Development of a screening system for inteins active in protein splicing based on intein insertion into the LacZα-peptide. Biological Chemistry, 398(1), 57-67. https://doi.org/10.1515/hsz-2016-0229
Polka, J. K., & Silver, P. A. (2016). A Tunable Protein Piston That Breaks Membranes to Release Encapsulated Cargo. ACS Synthetic Biology, 5(4), 303–311. https://doi.org/10.1021/acssynbio.5b00237
Polli, J. R., Chen, P., Bordeau, B. M., & Balthasar, J. P. (2022). Targeted Delivery of Endosomal Escape Peptides to Enhance Immunotoxin Potency and Anti-cancer Efficacy. The AAPS Journal, 24(3). https://doi.org/10.1208/s12248-022-00698-x
Preer, J. R., Hufnagel, L. A., & Preer, L. B. (1966). Structure and behavior of R bodies from killer paramecia. Journal of Ultrastructure Research, 15(1-2), 131–143. https://doi.org/10.1016/s0022-5320(66)80100-4
Preer, J. R., Siegel, R. W., & Stark, P. S. (1953). The Relationship between Kappa and Paramecin in Paramecium Aurelia. Proceedings of the National Academy of Sciences, 39(12), 1228–1233. https://doi.org/10.1073/pnas.39.12.1228
Qiu, Chong, Fei Xia, Junzhe Zhang, Qiaoli Shi, Yuqing Meng, Chen Wang, Huanhuan Pang, et al. “Advanced Strategies for Overcoming Endosomal/Lysosomal Barrier in Nanodrug Delivery.” Research 6 (January 2023): 0148. https://doi.org/10.34133/research.0148.
Smith-Sonneborn, J. E., & Van Wagtendonk, W. J. (1964). Purification and chemical characterization of kappa of stock 51, Paramecium aurelia. Experimental Cell Research, 33(1–2), 50–59. https://doi.org/10.1016/S0014-4827(64)81011-9
Sousa de Almeida, M., Susnik, E., Drasler, B., Taladriz-Blanco, P., Petri-Fink, A., & Rothen-Rutishauser, B. (2021). Understanding nanoparticle endocytosis to improve targeting strategies in nanomedicine. Chemical Society Reviews, 50(9), 5397–5434. https://doi.org/10.1039/d0cs01127d
Stephanopoulos, N., & Francis, M. B. (2011). Choosing an effective protein bioconjugation strategy. Nature Chemical Biology, 7(12), 876–884. https://doi.org/10.1038/nchembio.720
Wang, X.-W., & Zhang, W.-B. (2019). SpyTag-SpyCatcher Chemistry for Protein Bioconjugation In Vitro and Protein Topology Engineering In Vivo. Bioconjugation, 2033, 287–300. https://doi.org/10.1007/978-1-4939-9654-4_19
Wang, C., Zhao, T., Li, Y., Huang, G., White, M. A., & Gao, J. (2017). Investigation of endosome and lysosome biology by ultra pH-sensitive nanoprobes. Advanced drug delivery reviews, 113, 87–96. https://doi.org/10.1016/j.addr.2016.08.014
Wei, T., Liu, H., Wu, H., Pu, F., & Li, X. (2021). Serine/threonine ligation-assisted chemical synthesis of HMGA1a protein with site-specific post-translational modifications. STAR Protocols, 2(3), 100777–100777. https://doi.org/10.1016/j.xpro.2021.100777
Yazdi, P. T., & Murphy, R. M. (1994). Quantitative Analysis of Protein Synthesis Inhibition by Transferrin-Toxin Conjugates. Cancer Research, 54(24), 6387–6394.
Zhang, Y., Xu, C., Lam, H. Y., Lee, C. L., & Li, X. (2013). Protein chemical synthesis by serine and threonine ligation. Proceedings of the National Academy of Sciences - PNAS, 110(17), 6657–6662. https://doi.org/10.1073/pnas.1221012110
Zielińska, A., Karczewski, J., Eder, P., Kolanowski, T., Szalata, M., Wielgus, K., Szalata, M., Kim, D., Shin, S. R., Słomski, R., & Souto, E. B. (2023). Scaffolds for drug delivery and tissue engineering: The role of genetics. Journal of Controlled Release, 359, 207–223. https://doi.org/10.1016/j.jconrel.2023.05.042