Lipid Nanoparticle Formulation Variables and Peptide Encapsulation Efficiency

Lipid nanoparticle formulation is often treated as a problem the mRNA vaccine era solved. Peptide cargoes like GLP-1 break that assumption. Their charge profile, conformational flexibility, and susceptibility to hydrolysis make them a fundamentally different problem, one where variables that help nucleic acid retention can actively work against peptide stability.

Why encapsulating peptides like GLP-1 in lipid nanoparticles is harder than encapsulating nucleic acids

Peptide encapsulation in lipid nanoparticles is not a solved or default problem. The physicochemical properties of peptides set them apart from mRNA and siRNA from the outset. Their moderate molecular weight, amphoteric charge, conformational flexibility, and susceptibility to hydrolysis mean that a formulation optimized for a nucleic acid cannot simply be borrowed and expected to hold a peptide. The variables that drive efficient nucleic acid encapsulation, particularly strong electrostatic interactions with a heavily charged polyanion, can destabilize a peptide that depends on a narrower range of conditions to remain folded and intact.

GLP-1 illustrates this immediately. It is a moderate-length peptide of 30 or 31 amino acids with an amphipathic helical structure, which makes it prone to aggregation when its hydrophobic face is exposed to an incompatible environment and to enzymatic degradation when that structure is disturbed. There is no single property to optimize. The problem is relational, a matter of matching the assembly conditions to the peptide's own instability. The central tension is easy to state and difficult to resolve. Every formulation decision that follows is an attempt to shift the balance between retaining the drug inside the particle and losing it to the surrounding solvent, to the membrane, or to aggregation.

How lipid composition determines whether a peptide is captured or excluded at the moment of particle formation

The identity of the lipid components is the first and most foundational decision. Ionizable lipid, helper phospholipid, cholesterol, and PEGylated lipid together set the electrostatic environment, membrane fluidity, and aqueous compartment geometry that either attract or repel a given peptide at the moment of self-assembly. Composition is not a background condition. It is the active filter that decides whether a peptide enters the particle or is left outside.

Ionizable lipids are the most consequential single component. At low pH during formulation they carry a positive charge that can electrostatically complex with negatively charged regions of a peptide. Upon re-protonation in the acidic endosome, they regain their positive charge and enable endosomal escape. This pH-switchable character is what allows encapsulation and release to be decoupled, a property that matters as much for peptide delivery as it does for nucleic acids. But the switch only functions if the peptide's own charge profile is compatible with the lipid's protonation window.

Helper phospholipids, of which DSPC and DOPE are representative examples, modulate bilayer rigidity and the size of the aqueous core. More fusogenic phospholipids, which tend to be unsaturated and cone-shaped, create a more disordered membrane that can accommodate peptide molecules at the interface. The choice between a more rigid and a more fluid membrane is therefore not a neutral one for peptide cargo. It changes where a peptide can reside and how tightly the particle holds it.

Cholesterol and PEGylated lipids complete the compositional picture. But too high a cholesterol fraction reduces the aqueous volume fraction and compresses the space available for hydrophilic cargo. A compositionally mismatched lipid system cannot be rescued by ratio or pH optimization alone.

How molar ratios among lipid components shift the encapsulation equilibrium

Once lipid identity is fixed, the molar ratio among components becomes the primary lever for pushing encapsulation efficiency higher or lower. The relative proportion of ionizable lipid to phospholipid to cholesterol to PEG-lipid controls the charge density on the forming particle, the internal aqueous volume, and the membrane permeability that determines whether peptide molecules stay inside or partition back out. Ratio is where compositional intent meets quantitative control.

Higher ionizable lipid fraction increases the electrostatic driving force for peptide complexation at low pH. Above a threshold, the ionizable lipid fraction can disrupt bilayer integrity and cause peptide aggregation or leakage. The result is a compositional optimum, not a monotonic increase with more cationic lipid. Formulators who assume that more charge means more loading will find themselves on the wrong side of that curve.

Cholesterol ratio behaves similarly in a non-linear way. At moderate fractions it stabilizes the membrane and reduces leakage, which is exactly what a fragile peptide needs. The ratio thus sets a geometric constraint as much as an energetic one.

PEG-lipid fraction is usually kept low, in the low single-digit mole percent range. The PEG layer is essential for colloidal stability, but it competes with the particle's ability to form cleanly and capture cargo during the brief window of assembly. Keeping it low is a matter of protecting the encapsulation event itself.

What emerges from these individual effects is an argument for ratio windows rather than ratio targets. Because each component's effect on encapsulation depends on the levels of the others, a formulator who optimizes one ratio in isolation while holding the rest fixed will reach a local optimum, not a global one. For a therapeutic GLP-1 formulation destined for intranasal delivery, the ratio must be validated not only for encapsulation efficiency in isolation but for the specific pharmacokinetic and tolerability profile required by the nose-to-brain route, a constraint that shapes how nanoparticle biotechnology companies approach formulation iteration in a clinically grounded way.

The role of pH in driving peptide loading and controlling electrostatic interactions during assembly

The aqueous pH of the formulation buffer during LNP assembly is not a passive environmental condition. It is an active driver of encapsulation efficiency because it sets the protonation state of both the ionizable lipid and the peptide cargo. Their electrostatic attraction or repulsion at the moment of mixing determines how much peptide enters the forming particle. pH is therefore a kinetic and thermodynamic tool, not merely a preservation condition to be set and forgotten.

At acidic pH, typically around the pKa of the ionizable lipid, which often falls in the range of 6 to 7, the ionizable lipid is protonated and positively charged. Peptides with negatively charged residues, or with a net negative charge at that pH, are electrostatically drawn into the assembling particle. The question is not simply how acidic the buffer is, but whether the peptide and the lipid carry complementary charges at that specific point on the pH scale.

GLP-1 and related peptides have their own isoelectric point and charge profile across pH. Formulating below the peptide's isoelectric point gives it a net positive charge, which may reduce electrostatic attraction to the ionizable lipid. The pH must therefore be chosen with the specific peptide's charge curve in mind, not just the lipid's pKa. Two peptides with different isoelectric points can require different assembly pH values even in the same lipid system.

After assembly, the external buffer pH is typically raised to physiological range. This neutralizes the ionizable lipid and collapses the electrostatic interaction, which traps the peptide inside rather than releasing it prematurely. The same variable that drives loading at one stage prevents leakage at the next.

The window is bounded on both sides. If the assembly pH is too low, the ionizable lipid can become so highly charged that it disrupts the bilayer architecture entirely or causes the peptide to precipitate rather than encapsulate. The effective range is narrow, and it is defined by the intersection of two charge curves: the lipid's and the peptide's.

How solvent system choice and the aqueous-to-organic mixing process govern nanoparticle architecture and peptide entrapment

LNP formation is a kinetically controlled self-assembly process. When a lipid solution in an organic solvent, typically ethanol, is rapidly mixed with an aqueous peptide solution, the rate of mixing, the solvent polarity gradient, and the temperature together determine whether the particle forms a well-defined structure that entraps the peptide or a poorly organized aggregate that excludes it. The how of assembly matters as much as the what. The ethanol-to-aqueous volume ratio, which maps to the flow rate ratio in microfluidic or bulk mixing formats, directly sets the polarity environment at the moment of nucleation. Speed and order trade off, and the optimal mixing condition depends on how quickly the peptide loses solubility in the changing solvent environment.

Microfluidic mixing, using staggered herringbone micromixers or similar chip formats, achieves reproducible, rapid, and controlled mixing compared to bulk methods such as ethanol injection or thin-film hydration. The value of this approach is not that one platform is categorically superior, but that the reproducibility of the mixing event allows encapsulation efficiency to be a controlled outcome rather than a batch-to-batch variable. A kinetic process can only be optimized if it runs the same way every time.

Ethanol residual concentration after assembly matters for peptide integrity. Residual organic solvent can disrupt peptide secondary structure, and the alpha-helical conformations in GLP-1 are sensitive to solvent polarity. A peptide that survives assembly can still unfold during downstream processing if the solvent is not fully removed.

Temperature during mixing sets lipid fluidity. Below that transition, the membrane becomes rigid and may expel the peptide. A membrane that is fluid enough to accept a peptide during assembly and rigid enough to hold it afterward is the working target.

Why particle size and surface charge (zeta potential) are both outputs of formulation choices and independent predictors of encapsulation efficiency and payload stability

Particle size and surface charge are the integrated result of every upstream formulation decision. They must be understood both as diagnostic indicators of formulation quality and as variables in their own right that researchers can target. Measuring them is not enough. Designing for them is the point.

Smaller particles, typically in the sub-200 nm range, have a higher surface-area-to-volume ratio, which increases the proportion of peptide at or near the surface rather than in the interior. This can mean faster payload leakage unless the surface is stabilized by the PEG layer and the cholesterol-reinforced bilayer. A particle that looks uniform under dynamic light scattering can still be losing peptide at an unacceptable rate.

Zeta potential reflects the electrostatic stability of the particle in suspension. Near-neutral zeta potential in physiological conditions is desired for in vivo stealth, but it must be achieved by the ionizable lipid becoming uncharged after assembly, not by eliminating charge during assembly. These are different phenomena at different stages, and conflating them is a common error.

Polydispersity index is the third measurable that connects to encapsulation. Because different size fractions have different encapsulation capacities, high PDI translates directly to variable and unreliable encapsulation efficiency across a batch. The average hides the problem. Only the distribution reveals it.

For intranasal delivery specifically, particle size gains an additional dimension. Particles must be small enough to traverse the olfactory epithelium and mucus layer but large enough to protect the peptide from enzymatic degradation in the nasal cavity. Its GLP-1 therapeutic is built on formulation principles designed from the ground up for peptide encapsulation, where efficiency directly determines whether sufficient drug reaches the brain to achieve a therapeutic effect.

How these variables interact as a system and why single-variable optimization systematically underperforms

The variables described through this piece do not work independently. The mixing speed that produces an optimal size under one lipid composition may produce polydisperse particles under a different one. Maximum peptide encapsulation efficiency is not achieved by sequentially perfecting each variable in isolation. It is achieved by designing for the interactions among them.

The coupling between pH and ionizable lipid fraction is the most consequential. The charge environment set by their combination is what actually drives peptide capture. Size and charge are outputs, but they are outputs of a system, and optimizing an output directly without touching the inputs is aimless.

The strongest objection to this framing is that formulation complexity makes systematic optimization expensive and slow. The cost of under-optimized encapsulation, which shows up as low bioavailability, degraded peptide, and batch-to-batch variability in a clinical or commercial program, far exceeds the cost of upfront systematic characterization. Spending less on formulation upfront is a way of spending more on failure later.

What high encapsulation efficiency actually enables for peptide therapeutics delivered beyond the injection route

High encapsulation efficiency is not an end in itself. Intranasal, oral, and pulmonary routes all impose losses during transit, and any efficiency loss at the encapsulation step is compounded by those downstream losses. The formulation must overperform at the start because the route will erode the margin.

For intranasal peptide delivery, the olfactory and respiratory epithelium, mucociliary clearance, and the enzymatic environment all act to reduce the fraction of drug that reaches its receptor. That requires not just high initial encapsulation but colloidal and chemical stability of the encapsulated peptide throughout the delivery window. Encapsulation efficiency and payload stability are two faces of the same problem.

GLP-1 is a particularly instructive case because its short half-life and susceptibility to DPP-4 enzymatic cleavage mean that an LNP with marginal encapsulation efficiency will deliver a peptide that has already partially degraded inside the particle before it even reaches the receptor region. Formulation quality and therapeutic window are directly coupled. The particle is not a passive container. It is the environment in which the peptide lives from the moment of assembly to the moment of receptor engagement.

The broader platform implication is that a nanoparticle delivery platform built on rigorous formulation science is not a single-drug solution. The same mechanistic understanding of how lipid composition, ratio, pH, solvent, and size jointly control encapsulation can be applied across other peptides and biologics. Lionbio's research program illustrates this point directly: the platform is a set of formulation principles, not a bespoke recipe for one molecule. That distinction is what separates a platform approach from a molecule-by-molecule empirical scramble.

The next frontier is that formulation optimization for non-injectable routes changes which therapeutic indications become reachable at all. The blood-brain barrier shields the brain from systemic circulation, but the olfactory epithelium offers a route that bypasses that shield. The formulation science described here is the gate through which that route becomes viable.