Accelerating RNA therapeutic testing with liver microphysiological platforms

visit cn-bio.com A liver microphysiological system to study the delivery and efficacy of oligonucleotide-based therapeutics Oliver Culley, Alexandra Morse, Yassen Abbas, and Tomasz Kostrzewski CN Bio, Cambridge, UK APPLICATION NOTE CN Bio’s Organ-on-a-chip Systems, which include the PhysioMimix® Single- and Multi-Organ lab-benchtop instruments, enable scientists to model human biology in the lab through rapid and predictive human tissue-based studies. The technology bridges the gap between traditional cell culture and human studies, advancing towards the simulation of human biological conditions to support the accelerated development of new therapeutics in application areas including oncology, infectious diseases, metabolism and inflammation. Learn more at cn-bio.com Introduction As new modalities enter clinical trials, the demand for human-relevant in vitro models is increasing. Traditional pre-clinical animal models often fail to accurately predict drug efficacy and toxicity, leading to high late-stage drug attrition rates. New modalities, such as oligonucleotide therapies, are highly specific to humans and require a more human-centric development approach. Non-human species are frequently unsuitable for testing, and the most relevant options, like non-human primates (NHPs), are costly and ethically challenging. Therefore, human-relevant new approach methodologies (NAMs), such as organ-on-a-chip (OOC) technologies, are essential for improving the predictive power of pre-clinical testing. OOC platforms, also known as microphysiological systems (MPS), are designed to replicate the functional biomarkers of cells and tissues in a more physiologically relevant manner by culturing primary human cells on perfused 3D scaffolds (Rubiano et al., 2021). MPS have the potential to offer a more physiological approach to evaluating and designing oligonucleotide therapies before first-in-human trials. Oligonucleotide-based therapeutics, or RNA-based therapeutics, are short sequences that interfere with specific RNA molecules, including ASOs, RNAi, siRNA, miRNA, and aptamers (Roberts et al., 2020). Interest in these therapies grew from understanding the human genome and their faster drug discovery process compared to traditional small compounds. Recent advances, such as lipid nanoparticles (LNPs) and GalNAc-conjugations, enable targeted drug delivery to the liver. GalNAc targets the liver via the asialoglycoprotein receptor (ASGPR) on hepatocytes. GalNAc-conjugated oligonucleotides bind to ASGPRs, are endocytosed, and release the therapeutic oligonucleotide to target specific mRNA sequences (Debacker et al., 2020). This method enhances potency, reduces off-target effects, and allows for longer dosing intervals with lower immunogenicity. For this reason, GalNAc-based therapeutics hold significant potential for treating genetic, metabolic, and infectious diseases. Due to the rapid uptake of oligonucleotides by hepatocytes through GalNAc-associated endocytosis, this strategy has been employed to design therapies targeting liver disease. It has shown promise in treating previously 'undruggable' diseases where traditional drugs have failed. Consequently, we chose to investigate both siRNAs and ASOs in our PhysioMimix® Liver MPS and Oligonucleotide delivery assay, using GalNAc-conjugation to target primary human hepatocytes (PHHs). Aim We introduce the PhysioMimix Liver MPS to study the delivery of oligonucleotides into the liver and their uptake by PHHs. The Liver MPS produces functional and stable PHH microtissues, enabling the study of gene knockdown and dosing strategies over 14 days. To address the growing need for appropriate in vitro testing of such therapeutics, we target PHHs with fluorescently labelled siRNA or ASOs, with and without GalNAc conjugation, in our Oligonucleotide delivery assay. Our results show effective uptake by PHHs and subsequent knockdown of the target genes of interest. Materials & Methods All experiments were performed using the PhysioMimix Single-organ System and bespoke Multi-chip Liver-12 plates, consisting of 12 wells containing individual collagen-coated scaffolds, with PHHs seeded at a density of 0.6x106 cells/scaffold (Figure 1A-B). Pores within the scaffold, in combination with a perfused system, allow for flow of cell culture media across the formed liver microtissues (Figure 1C). PHHs were treated with fluorescently labelled GalNAc-siRNAs (Creative Biogene) targeting the gene GAPDH. These included: GalNAc-siGAPDHCy5; Naked-siGAPDH-Cy5; or GalNAc-siScrambled-FAM. In addition to GalNAc-siRNAs, we investigated GalNAc-ASOs (Creative Biogene) targeting the gene for the production of albumin protein. The following ASOs were used: GalNAc-ALB ASO-Cy5; Naked-ALB ASO-Cy5; or GalNAc-Scrambled ASO-FAM. Scaffolds were imaged for Cy5 or FAM label using a confocal microscope (Nikon A1-R); and RNA was isolated to test for knockdown of the GAPDH gene using the qPCR technique. We investigated uptake and gene knockdown over a range of concentrations (10 nM – 100 nM) of ASOs as well as the effect of single dose vs repeated dose administered in our Liver MPS-based Oligonucleotide delivery assay. Media exchanges were performed every 2-3 days, and media samples were used to assess cell health and damage. Hepatocyte cell health and functionality were evaluated by quantifying the production of lactate dehydrogenase (LDH, Promega’s colormetric assay) and albumin protein by ELISA (AssayPro), respectively. Uptake was evaluated by immunocytochemistry following fixation with 4% PFA alongside staining for albumin (ab207327, abcam) as well as Hoechst 33342 (Invitrogen) using half the scaffold at the experiment end. Zymo Research’s RNA Shield and Quick-RNA Microprep Kit were used to extract RNA from the remaining half scaffold. Invitrogen’s High-Capacity RNA-to-cDNA Kit, followed by qPCR using TaqMan Fast Advanced Master Mix, and TaqMan Gene Expression Assays were used to assess target gene expression. Genes of interest were expressed as quantitation cycle (Cq) values normalised using reference gene Cq values (18S, HPRT1 or TBP), displayed as ΔΔCq expression, having first normalised to that of the vehicle control (Livak and Schmittgen, 2001). Results & Discussion We used labelled siRNAs and ASOs, with and without GalNAc conjugation, to explore the use of new modalities in our PhysioMimix Liver MPS-based Oligonucleotide delivery assay. PHHs were seeded into individual wells on Day 0 with media changes and sampling occurring every 2-3 days oligonucleotides were added once (Day 4), or three times (Day 4, 6, and 8), or treated with a vehicle control (ddH2 O), (Figure 1D). Stable and functional PHHs can be generated on collagen-coated scaffolds to evaluate the uptake of labelled oligonucleotides (Figure 1E). Readouts of cell damage (LDH) and cell health (albumin) showed no difference among the siRNAs used. LDH levels typically started around 1.5- 2.0 RAU/106 cells, consistent across all experiments. Albumin production remained between 20-40 μg/day/106 cells for this PHH donor, with peak production typically occurring on Day 6 (Figure 2A). Greatest knockdown of the gene of interest (GAPDH) was observed with the treatment of PHHs with GalNAc-conjugated targeting siRNAs (GalNAc-siGAPDH, Figure 2B). Liver microtissues remained functional at experiment take-down, shown by positive albumin staining, which also acts to co-label the PHHs (Figure 2C). The uptake of Cy5-labeled siRNA by PHHs in the scaffold pores was confirmed, as shown in Figure 2C. Although the signal intensity was higher in the Nkd-siGAPDH images compared to the GalNAc-siGAPDH images (magenta, Figure 2C), indicating more uptake of this modality, the GalNAc-conjugated modality had a greater effect at the gene expression level (Figure 2B). This difference may be due to varying means of uptake/ endocytosis or processing of the siRNAs. We next investigated whether there was a dose-dependent knockdown of GAPDH with GalNAc-conjugated targeting siRNAs. PHHs were treated three times with siRNA (on Days 4, 6, and 8) at concentrations of 10, 50, or 100 nM. There were no adverse effects on cell health, as indicated by low LDH levels and stable albumin production over the 11-day experiment (Figure 3A). We observed a dose-dependent knockdown of GAPDH gene expression with GalNAc-siGAPDH, while no concentration effect was seen in the Nkd-siGAPDH condition (Figure 3B). An increase in siRNA uptake with increasing concentration was also observed (magenta, Figure 3C). Using antisense oligonucleotides (ASOs) targeting albumin gene (ALB) expression we first treated PHHs once at Day 4, over a 14-day period. Comparing uptake of labelled-ASOs at Day 14, there was greater signal in the Nkd-ALB ASO compared with the GalNAc-ALB ASO (Figure 4A). Despite this we observed more knockdown of ALB gene expression in the GalNAc-ALB ASO conditions (Figure 4B), which was consistent with observations in the GalNAc-siRNAs (Figure 2). Further work could investigate the end cellular location of conjugated vs non-conjugated oligonucleotides in PHHs to better understand these observed differences. Next, we aimed to investigate the effect of two dosing strategies with either a single dose (Day 4) or repeat dose (Day 4, 6 and 8), over a 14-day period. We found comparable results for LDH and albumin between treatment regimens; and similar results when comparing the three different ASOs used (Figure 5A-B). GalNAc-ALB ASO achieved greater knockdown of ALB gene expression compared with Nkd-ALB ASO when administered in a single dose, as well as in a repeat dose (Figure 3C-D). Consistent with the greater knockdown of gene with repeat dosing, we observed greater uptake of ASOs both targeting the albumin gene, and the scrambled control, with the repeat dose compared to the single dose (Figure 5E-F). This highlights the ability of the Liver MPS to facilitate repeat dosing, as would often be the case in the clinical application of new modalities. In this study, we demonstrate the effectiveness of the Liver MPS in examining the uptake, and gene knockdown, of oligonucleotides targeting the liver. We utilized siRNA targeting the GAPDH gene and ASOs targeting the ALB gene. For each modality, we included a non-coding scramblecontrol to assess cellular uptake without gene targeting. We show effective uptake in our system of both siRNA and ASO with a single dose, and increased uptake with a repeated dose. There is greater knockdown of target genes GAPDH and ALB for GalNAc-conjugated siRNA and ASO, respectively, when compared with non-conjugated modalities. Recent studies by other groups have explored the cellular uptake of ASOs using PHHs in our Liver MPS (Majer et al., 2024). Using a multiphoton imaging platform, Majer et al., demonstrated increased uptake of GalNAcASOs at 1 hour compared to Nkd-ASOs. However, at 4 hours, the reverse was observed, with greater signal intensity (labelled ASOs) in PHHs treated with Nkd-ASOs, compared to GalNAc-conjugated ASOs. The authors also distinguished between cuboidal and non-cuboidal shaped hepatocytes and examined the trafficking of endocytosed ASOs. They found no colocalization of intracellular lipids with endocytosed ASOs in cuboidal hepatocytes (i.e., those seen in vivo). Future research could investigate a dose-response assay with fewer treatments but higher concentrations of the new modalities. Furthermore, to investigate the specificity of GalNAc for its receptor on hepatocytes, incorporating another cell type, such as Kupffer cells, could demonstrate a preference for PHHs when using GalNAcconjugated modalities in our system. Including Kupffer cells can also help detect any immune-driven toxicity of the oligonucleotides, as these cells are part of the liver's resident macrophage population. Figures Figure 1. Generation of 3D human liver microtissues using the PhysioMimix® OOC microphysiological system. (A) Overview of the PhysioMimix platform, including the Controller, Docking Station, and three Drivers with Multi-chip Liver-12 plates. (B) Schematic of a Multi-chip Liver-12 plate, showing a cross-section of an open cell culture well with a 3D scaffold for hepatocyte growth. (C) Micropumps drive the circulation of cell culture media, facilitating nutrient and gas exchange. Each plate can culture up to 12 liver microtissues. (D) Timeline for dosing the Liver MPS with oligonucleotides over 14 days. (E) 3D human liver microtissues on a scaffold imaged by brightfield microscopy. Scale bar: 200 μm. Figure 2. Introduction of GalNAc-conjugated, fluorescently labelled siRNAs in the Liver MPS. PHHs treated three times with siRNA (Day 4, 6, and 8) at a concentration of 0 nM and 100 nM. (A) Cells are healthy indicated by low LDH, and stable albumin production observed over the 11-day experiment. (B) Knockdown is greater in target (GAPDH) gene expression with GalNAc-conjugated-siGAPDH compared to nonconjugated (Nkd) siRNA. (C) Immunofluorescent imaging following fixation shows uptake of Cy5-labelled siRNAs (magenta); nuclei (blue), and albumin protein (yellow). Scale bar: 50 μm. A Cell health metrics B Target expression C Figure 3. Dose-dependent knockdown of GAPDH with GalNAcconjugated targeting siRNAs. PHHs treated three times with siRNA (Day 4, 6, and 8) at a concentration of 0, 10, 50 and 100 nM. (A) Cells are healthy indicated by low LDH, and stable albumin production observed over the 11-day experiment. (B) Dose dependent knockdown in target (GAPDH) gene expression with GalNAc-siGAPDH; with no effect of increasing concentration in Nkd-siGAPDH condition. (C) Immunofluorescent imaging following fixation shows uptake of Cy5-labelled siRNAs (magenta); nuclei (blue), and albumin protein (yellow). Scale bar: 100 μm. A B Cell health metrics Target expression C Figure 4. Sustained uptake of ASOs, and effective gene targeting, with single dose. (A) Immunofluorescent imaging following fixation shows uptake of Cy5-labelled ASOs (magenta), nuclei (blue), albumin protein (yellow) and FAM-labelled non-targeting ASO (green). (B) Greatest knockdown of ALB gene observed with GalNAc-ALB ASO compared with other conditions after a single dose at a concentration of 50 nM. Scale bar: 100 μm. A B Figure 5. Comparison using a single and repeat dose of GalNAcconjugated, fluorescently labelled ASOs in the Liver MPS. PHHs were treated with ASOs targeting albumin gene (ALB) expression either once (single; Day 4), or three times (repeat; Day 4, 6 and 8), over a 14-day period. Cells are healthy indicated by low LDH, and stable albumin over the 14- day experiment for the (A) single and (B) repeated dose. Gene expression of GalNAc-conjugated ASO targeting ALB (GalNAc-ALB); non-conjugated ASO targeting ALB (Nkd-ALB), and GalNAc-conjugated, non-targeting ASO (GalNAc-Scramble) in the (C) single and (D) repeated dose conditions. Immunofluorescent imaging shows nuclei (blue), albumin protein (yellow), Cy5-labelled targeting ASOs (magenta), and FAM-labelled non-targeting ASO (green) in the (E) single and (F) repeated dose conditions. Scale bar: 100 μm. A B C D E F Conclusion Our study demonstrates the utility of the primary human Liver MPS and Oligonucleotide delivery assay to overcome the limitations of nonhuman models for evaluating the uptake and gene knockdown efficacy of oligonucleotides targeting the liver. Using the PhysioMimix Single-organ System and Multi-chip Liver-12 plates, we were able to generate stable and functional human liver tissues on 3D collagen coated scaffolds for testing siRNAs and ASOs. Our results showed effective uptake of both siRNA and ASO with a single dose, and increased uptake with repeated dosing. Notably, GalNAc-conjugated siRNA and ASO exhibited greater knockdown of target genes GAPDH and ALB, respectively, compared to non-conjugated modalities. These findings highlight the potential of the PhysioMimix Liver MPS for in vitro testing of new therapeutic modalities and underscore the importance of dosing strategies in achieving functional gene knockdown. We show no effect of knockdown at the protein level. Dosing would likely find a KD that gave a downstream response in terms of cell function, however, here we are dosing too lightly. Liver tissue cells remain targeted but viable. The liver MPS-based Oligonucleotide assay provides a robust physiological approach to enable testing of oligonucleotide therapies before first-in-human trials, as well as an alternative to animal in vivo models. Key messages Oligonucleotide therapies are highly specific to humans and require a more human-centric development approach. MPS replicates the structural and functional biomarkers of cells and tissues in a more physiologically relevant manner. Our study demonstrates the effectiveness of a Liver MPS and Oligonucleotide delivery assay in examining the uptake and gene knockdown, of oligonucleotides targeting the liver. MPS insights facilitate the development of new modality drugs where animal models are less suited, enabling better informed first-inhuman studies. References A. J. Debacker et al., (2020). Mol. Ther. 28 (8):1759-1771 K. J. Livak & T. D. Schmittgen (2001), Methods. 25 (4):402-408 J. Majer et al., (2024). Lab. Chip. 24 (19):4594-4608 T. C. Roberts et al., (2020). Nat. Rev. Drug Discov. 19 (10):673-694 A. Rubiano et al., (2021). Clin. Transl. Sci. 14 (3):1049-1061 +44 (0) 1223 737941 | enquiries@cn-bio.com visit cn-bio.com 332 Cambridge Science Park Milton Road, Cambridge, CB4 0WN CNB1362. v1