Contact Us

Advances in Delivery of mRNA by Lipid Nanoparticles

Advances in Delivery of mRNA by Lipid Nanoparticles
Loading... 390 view(s)
Advances in Delivery of mRNA by Lipid Nanoparticles


Intracellular delivery of in vitro transcribed (IVT) mRNA as a vaccine or therapeutic is generally achieved by use of lipid nanoparticles (LNPs), which prevent degradation by RNases in blood and are internalized by various mechanisms. LNPs for mRNA delivery, as reviewed elsewhere, are usually formulated as mixtures of four components: ionizable cationic lipids, phospholipids, cholesterol, and polyethylene glycol (PEG) lipids.  

Among the nearly 350 publications on the delivery of mRNA by LNPs in the NIH PubMed database for 2023 to now, the following three reports were selected for this blog because of their potential broad impact: 

  1. Lipid Adjuvants for Increasing mRNA Vaccine Immunogenicity 
  1. LNPs for Muscle-Selective mRNA Delivery with Minimized Off-Tissue Transfection 
  1. In Vivo Editing of Hematopoietic Stem Cells by Antibody-LNP Delivery of mRNA 

In each case, TriLink products were used for either IVT synthesis of CleanCap® nucleoside-modified mRNA or as measurable reporter genes for optimizing the selectivity of LNP formulations. 

  1. Lipid Adjuvants for Increasing mRNA Vaccine Immunogenicity

An adjuvant is an ingredient used in some vaccines to induce a stronger immune response and thereby help vaccines work better. In 2019, Miao et al. discovered that LNP-mRNA vaccines comprised of ionizable lipids with novel cyclic amine head groups can enhance adaptive immune responses in mice. However, due to slow clearance of these lipids from the body, it would be risky to apply them for adjuvanting in humans.  

Follow-up investigations recently reported by Li et al. (2023) solved this problem by introducing an ester linkage into the lipid backbone for esterase-mediated metabolism to facilitate tissue clearance and thereby improve the tolerability and safety of LNPs. They synthesized and screened a combinatorial library of 480 ionizable lipids, each with a different cyclic amine head group and ester-linked tails (Figure 1).  


FIGURE 1. Depiction of an ionizable lipid with a cyclic amine headgroup (red) and ester linker (purple) to lipid tails (blue). Adapted from Li et al. (2023) by Jerry Zon.  

The library of cyclic ionizable lipids was screened based on two criteria: (1) whether a lipid can enable potent mRNA transfection in vivo and (2) whether a lipid can effectively activate immune cells. Two LNPs containing commercially available, FDA-approved ionizable lipids MC3 or ALC-0315 were used as controls.  

To evaluate the transfection efficiency, LNPs were formulated using the library of lipids and loaded with IVT firefly luciferase (Fluc) mRNA prepared with TriLink’s pseudouridine-5’-triphosphate. The resultant Fluc mRNA-LNPs were then screened in mice by quantification of Fluc-mediated bioluminescence 6 h post-injection. Six top-performing cyclic ionizable lipids were thus identified. 

To evaluate the efficiency of immune-cell activation by these six top performers, each was incorporated into an LNP formulation loaded with TriLink’s CleanCap® 5-methoxyuridine-modified ovalbumin (OVA) mRNA, which is commonly used as a model mRNA vaccine. A panel of in vitro measurements and in vivo data obtained with vaccinated mice collectively demonstrated that all six LNP formulations made from these cyclic ionizable lipids were more immunostimulatory compared to the control LNPs made with MC3 or ALC-0315. 

  1. LNPs for Muscle-Selective mRNA Delivery with Minimized Off-Tissue Transfection

According to Chen et al. (2023), LNP delivery of mRNA can result in substantial off-target mRNA expression in the liver, potentially causing safety issues. To mitigate this concern, they used a combinatorial synthesis and screening approach analogous to that mentioned above to find LNP formations for highly muscle-specific mRNA delivery, thus minimizing off-tissue effects in the liver and other organs. Using 324 different LNPs loaded with TriLink’s CleanCap® FLuc mRNA as a reporter, transfected HeLa cells were quantitatively assayed for FLuc expression.  

The best-performing cyclic ionizable lipid, termed iso-A11, was then used to prepare LNP-FLuc mRNA for comparison with the ionizable lipid SM-102, which is a component in Moderna’s COVID-19 vaccine. Each LNP formulation of FLuc mRNA was separately administered to mice by intramuscular (i.m.) injection at the same dose, which led to comparable transfection efficiency at the injection site based on IVIS imaging 6 h post-injection. Remarkably, iso-A11 showed no measurable transfection in off-tissue organs, whereas SM-102 showed high transfection in the liver and spleen, indicating exceptional muscle specificity for iso-A11. Additional studies led to the suggestion that this muscle specificity was due to the difference in pKa values for iso-A11 vs. SM-102. 

Next, they investigated the potential for iso-A11 in gene editing applications, wherein possible off-target editing in non-target cells is a concern. Iso-A11 and SM-102 LNPs were loaded with TriLink’s CleanCap® Cre mRNA (5moU) for i.m. administration to genetically engineered mTmG reporter mice. In these mice, a red fluorescent protein (tdTomato) is ubiquitously expressed on cell membranes. The functional delivery of Cre mRNA facilitates Cre-mediated genetic recombination such that tdTomato is replaced by a membrane-bound enhanced green fluorescent protein (EGFP). 

High EGFP signal from editing at the administration muscle sites was observed for both iso-A11 and SM-102 LNPs. However, based on confocal microscopy of tissue sections, the SM-102 treated group also showed high off-tissue editing in the liver and spleen, while editing in these tissues by iso-A11 was undetectable. 

To assess the potential efficacy of iso-A11 LNP for therapeutic vaccines, Chen et al. conducted experiments using the conventional B16F10 melanoma model. They chose to encapsulate tyrosinase-related protein 2 (Trp2) mRNA in the LNPs for an mRNA cancer vaccine based on previous research (Miao et al. 2019) that had proved Trp2 as a validated tumor-associated antigen for melanoma.  

Mice were subcutaneously injected with B16F10-FLuc melanoma cells that had been engineered to express FLuc for monitoring tumor volume by IVIS. Mice were administered Trp2 mRNA encapsulated in iso-A11 LNPs as two doses on day 7 and day 12 after tumor inoculation. A similar administration of phosphate-buffered saline (PBS) served as a negative control. During the 28-day monitoring period, iso-A11 significantly (p <0.01) reduced tumor volume compared to PBS, thereby supporting this approach for cancer immunotherapy. 

  1. In Vivo Editing of Hematopoietic Stem Cells by Antibody-LNP Delivery of mRNA

Hematopoietic stem cells (HSCs) reside in the bone marrow (BM), where they divide throughout life to produce all cells of the blood and immune system through their self-renewal ability. HSC transplantation (HSCT), which replaces diseased HSCs with healthy ones, can be a curative treatment for nonmalignant hematopoietic disorders such as hemoglobin gene (HBB) abnormalities.  

Current autologous HSCT requires ex vivo treatment with reagents for genome editing. A “conditioning” regimen, such as aggressive chemotherapy or whole-body radiation, is then used to eliminate the patient’s own diseased HSCs prior to transplanting the edited HSCs. This allows the engraftment of the genetically modified autologous HSCs in the BM. However, the required conditioning procedures can lead to acute and chronic systemic toxicities, and secondary malignancies due to accumulated DNA damage.  

In a radical departure from this current approach, Breda et al. (2023) developed new methodology to directly modify HSCs in vivo. This research, which used IVT TriLink CleanCap® mRNAs modified with N1-methylpseudouridine, features a targeting antibody (Ab) covalently attached to the surface of LNPs (LNP-Ab) for HSC-specific delivery of mRNAs. The Ab binds to the stem cell factor (SCF) receptor termed KIT, which was chosen because the intracellular uptake process for SCF-KIT is essential for HSC viability, and could facilitate mRNA/LNP-KIT internalization. 

As a first step, it was demonstrated that intravenous administration of FLuc mRNA/LNP-KIT to mice led to detection of FLuc-mediated luminescence in the femur at 24 hours, whereas control FLuc mRNA/LNP attached to an irrelevant IgG Ab did not.  

To assess the feasibility for therapeutic human genome editing, LNP-KIT was used to deliver mRNA encoding a previously reported (Newby et al. 2021) Cas9 adenine base editor (ABE) and, separately, single-guide RNA (sgRNA) targeted to the β-globin sickle cell mutation. Adenine base editing of an A to a G converts the pathogenic HBBS gene mutation to the nonpathogenic HBBG variant.  

Sickle cell specimens from four donors were studied. It was found that a molecular excess of sgRNA/LNP-KIT, relative to ABE mRNA/LNP-KIT, led to an 88% rate of editing based on DNA sequencing. Using reversed-phase HPLC, it was shown that there was a corresponding increase in HBBG protein (up to 92%) and decrease in HBBS protein after in vitro erythroid differentiation. Finally, microscopic inspection showed a nearly complete absence of sickle-shaped cells upon exposure of the erythroblasts to hypoxic conditions. 

These findings collectively support future clinical studies of in vivo delivery of LNP-KIT loaded with ABE mRNA and sgRNA to treat sickle cell disease. Simple intravenous infusion could provide a much simpler and cost-effective alternative to current ex vivo treatment, conditioning, and engraftment procedures that are reported  to exceed $2-3 million. Intravenous infusion to cure other monogenic nonhematopoietic diseases, such as cystic fibrosis and metabolic disorders, may likewise be feasible.  

Concluding Comments 

The above examples indicate that customization of LNP composition can enable increased immunogenicity, tissue targeting, and Ab-directed cell-specific delivery. However, synthesis and screening of large libraries of candidate lipids are daunting tasks. To address this bottleneck, Lewis et al. (2023) recently reported a machine learning algorithm to reduce the burden of  screening ionizable lipids by learning from the breadth and diversity of lipids that have already been tested. Stay tuned for more on this in the next Zone blog, which is focused on artificial intelligence (AI) applied to mRNA vaccines. 


Your comments are welcomed, as usual. 


Please feel free to share this blog with your colleagues or on social media.