Drug Delivery Systems and Lipid Nanoparticles

Drug delivery systems (DDS) and lipid nanoparticles (LNPs) are engineered biomaterial, polymeric, and supramolecular carrier platforms that control the temporal release rate, pharmacokinetic circulation half-life, spatial tissue biodistribution, and intracellular trafficking of therapeutic payloads. While small molecules with balanced physicochemical properties can often be administered as conventional immediate-release tablets (governed by the Biopharmaceutics Classification System), advanced delivery vehicles are essential both for maintaining steady therapeutic plasma concentrations of short-half-life small molecules and peptides over weeks to months, and for enabling macromolecular nucleic acid medicines—including antisense oligonucleotides and siRNAs, messenger RNA (mRNA) vaccines, and CRISPR-Cas9 gene editors—that would otherwise be degraded within minutes by serum nucleases and excluded by anionic cell membranes.

The evolution of modern drug delivery spans three technological eras: (1) macroscopic and microscopic controlled-release polymers (pioneered in the 1960s–1980s by Judah Folkman, Robert Langer, and Takeru Higuchi), which shifted administration from pulsatile peak-and-trough dosing to zero-order and sustained depot release; (2) PEGylated "stealth" liposomes (developed by Alec Bangham, Gregoriadis, and Yechezkel Barenholz), which encapsulated cytotoxic chemotherapeutics in phospholipid bilayers to evade reticuloendothelial clearance; and (3) four-component pH-ionizable lipid nanoparticles (LNPs) (pioneered by Pieter Cullis and colleagues), whose pH-switchable surface charge enabled the clinical translation of the first FDA-approved siRNA therapeutic (patisiran, 2018) and the COVID-19 mRNA vaccines (BNT162b2 and mRNA-1273, 2020).

Zero-Order Osmotic Pumps, Higuchi Diffusion, and PLGA Microspheres

Conventional immediate-release dosage forms produce sawtooth plasma concentration–time profiles that oscillate between subtherapeutic troughs and toxic peaks. Controlled-release systems regulate drug flux through three distinct kinetic regimes:

1. Higuchi Matrix Diffusion (t^1/2 Kinetics): In 1961, Takeru Higuchi derived the release rate of a dispersed drug from an insoluble planar or spherical polymer matrix, demonstrating that the cumulative fraction of drug released (M_t / M_∞) scales with the square root of time: M_t / M_∞ = k_H t^1/2, because the diffusion path length increases as the receding dissolution front moves inward. More generally, polymeric systems combining Fickian diffusion and polymer chain relaxation obey the Korsmeyer–Peppas power law: M_t / M_∞ = k tⁿ (where n = 0.43 for Fickian diffusion from a sphere, 0.43 < n < 0.85 for anomalous transport, and n = 0.85–1.0 for Case II zero-order erosion).
2. Elementary Osmotic Pumps (OROS, Zero-Order Kinetics): To achieve a constant release rate independent of gastrointestinal pH and motility (d*M*/d*t* = constant, n = 1.0), Felix Theeuwes developed the osmotic push–pull tablet: a core containing drug and an osmotic driving agent (such as NaCl or polyethylene oxide hydrogel) is coated with a rigid semi-permeable cellulose acetate membrane pierced by a laser-drilled orifice (~0.2–0.5 mm). Water enters across the membrane down an osmotic pressure gradient (Δπ), pumping dissolved or suspended drug out through the orifice at a constant zero-order rate for 12 to 24 hours (used in extended-release nifedipine, methylphenidate, and paliperidone).
3. Biodegradable PLGA Microsphere Depots: For parenteral sustained release over 1 to 6 months, peptides and small molecules are encapsulated via oil-in-water emulsion solvent evaporation into 20–100 μm microspheres of poly(D,L-lactic-co-glycolic acid) (PLGA). In subcutaneous or intramuscular tissue, water hydrolyzes the ester bonds of the PLGA backbone into endogenous lactic acid and glycolic acid, which enter the citric acid cycle and are exhaled as CO₂ and H₂O. Adjusting the lactide-to-glycolide monomer ratio (e.g., 50:50 vs. 75:25 or 85:15; higher lactide content increases methyl-group hydrophobicity and slows hydrolysis) tunes release duration from 1 month to 6 months in commercial depots of leuprolide (Lupron Depot), goserelin, risperidone, and naltrexone.

Ammonium Sulfate Remote Loading and the EPR Effect (Doxil)

Discovered by Alec Bangham in 1961, liposomes are spherical vesicles (typically 80–150 nm in diameter) composed of one or more concentric phospholipid bilayers surrounding an aqueous core, allowing simultaneous encapsulation of hydrophilic drugs in the aqueous interior and lipophilic drugs within the hydrocarbon acyl chains. However, first-generation unmodified liposomes injected intravenously were opsonized within minutes by plasma complement proteins and rapidly phagocytosed by Kupffer cells and splenic macrophages of the mononuclear phagocyte system (MPS).

To overcome rapid clearance and passive drug leakage, Yechezkel Barenholz, Alberto Gabizon, and colleagues engineered PEGylated liposomal doxorubicin (Doxil, approved 1995)—the first FDA-approved nanomedicine—using three biophysical innovations:

• Stealth PEGylation: Incorporating 5 mol% of a phospholipid conjugated to 2,000 Da poly(ethylene glycol) (DSPE–PEG₂₀₀₀) creates a hydrated, sterically repulsive polymer brush layer on the vesicle exterior that prevents protein opsonization and extends human circulation half-life to ~55–70 hours.
• High-T_m Rigid Bilayer: Constructing the bilayer from fully saturated hydrogenated soy phosphatidylcholine (HSPC, T_m ≈ 52 °C) and cholesterol (38 mol%) forms a liquid-ordered phase that prevents premature doxorubicin leakage in plasma at 37 °C.
• Transmembrane Ammonium Sulfate ((NH₄)₂SO₄) Gradient Remote Loading: Liposomes are prepared with an internal 250 mM (NH₄)₂SO₄ buffer. Trace neutral NH₃ diffuses out of the liposome, generating an inward-directed pH gradient (interior pH ≈ 5.2). When amphiphilic weakly basic doxorubicin (p*K*_a ≈ 8.3) is added externally, neutral doxorubicin diffuses across the bilayer, becomes protonated inside the acidic core, and precipitates with sulfate anions (SO₄²⁻) into crystalline fibrous bundles (>95% loading efficiency, ~10,000–15,000 drug molecules per 85 nm vesicle). These long-circulating nanoparticles preferentially extravasate through leaky fenestrated tumor vasculature (pore sizes 100–600 nm) via Yasuhiro Matsumura and Hiroshi Maeda's enhanced permeability and retention (EPR) effect, while sparing the tight capillaries of the heart to drastically reduce anthracycline cardiotoxicity.

The Four-Component LNP Formulation and Ionizable Lipid p*K*_a (~6.2–6.5)

Delivering polyanionic nucleic acids—such as a 21-base-pair siRNA (~14 kDa, ~42 negative phosphate charges) or a 4,000-nucleotide mRNA (~1.3 MDa, ~4,000 negative charges)—into the cytoplasm requires condensing the nucleic acid into a protective nanoparticle that crosses the cell membrane and escapes the endolysosomal compartment. Early cationic liposomes made with permanently positively charged quaternary ammonium lipids (such as DOTMA or DOTAP) complexed RNA efficiently in vitro, but proved toxic and ineffective in vivo because their permanent positive charge agglutinated negatively charged red blood cells, activated complement cascades, and caused severe hepatotoxicity.

Pieter Cullis and colleagues solved this paradox by replacing permanently charged lipids with ionizable cationic lipids—lipids bearing a tertiary amine headgroup attached to branched hydrophobic tails—paired in a precise four-component molar architecture (typically 50 : 38.5 : 10 : 1.5 mol%) assembled via rapid microfluidic ethanol-dilution mixing:

1. Ionizable Cationic Lipid (~50 mol%): Examples include DLin-MC3-DMA (MC3) in the hepatic siRNA drug patisiran (Onpattro), ALC-0315 in the Pfizer/BioNTech COVID-19 mRNA vaccine (BNT162b2), and SM-102 in the Moderna mRNA vaccine (mRNA-1273). Crucially, the apparent surface p*K*_a of the ionizable lipid inside the LNP is tuned precisely to 6.2–6.5.
2. Cholesterol (~38.5 mol%): Fills packing defects between lipid acyl chains, stabilizes the nanoparticle core, and facilitates membrane fusion during endosomal escape.
3. Helper Phospholipid (~10 mol%): Typically 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC) or fusogenic DOPE, which stabilizes the outer monolayer and promotes non-bilayer phase transitions.
4. Diffusible PEG-Lipid (~1.5 mol%): Typically DMG-PEG₂₀₀₀ or ALC-0159 bearing short C₁₄ myristoyl chains; prevents particle aggregation during microfluidic manufacture and storage, then sheds rapidly in blood.

Three-Stage pH-Switchable Mechanism and Endosomal Escape

The golden rule of LNP biophysics (apparent p*K*_a ≈ 6.2–6.5) enables the nanoparticle to switch its electrostatic state across three distinct physiological environments:

• Stage 1 — Acidic Microfluidic Formulation (pH 4.0): Because pH 4.0 is >2 units below the lipid p*K*_a, the tertiary amine headgroups are >99% protonated (positively charged). Upon rapid mixing of the ethanol lipid stream with the aqueous pH 4.0 acetate buffer containing polyanionic mRNA or siRNA, electrostatic complexation encapsulates >90%–95% of the RNA cargo inside electron-dense inverse micellar structures.
• Stage 2 — Neutral Physiological Circulation (pH 7.4): Following dialysis into pH 7.4 buffer and intravenous or intramuscular injection, pH 7.4 is ~1 unit above the lipid p*K*_a, rendering the LNP surface predominantly uncharged (neutral). Neutrality prevents nonspecific erythrocyte lysis and immune toxicity. Simultaneously, the short-chain C₁₄ PEG-lipid desorbs from the LNP surface in plasma (t_1/2 ≈ minutes), allowing endogenous Apolipoprotein E (ApoE) to adsorb onto the neutral lipid surface. For intravenously administered LNPs (like patisiran or in vivo CRISPR editors like NTLA-2001), bound ApoE acts as a natural targeting ligand that engages low-density lipoprotein receptors (LDLR) on fenestrated hepatocytes, driving >80% liver uptake via receptor-mediated endocytosis.
• Stage 3 — Endosomal Acidification and Inverted Hexagonal (H_II) Escape (pH 5.5–6.0): Once internalized into an endosome, V-type H⁺-ATPase proton pumps acidify the endosomal lumen to pH 5.5–6.0 (below the lipid p*K*_a). The ionizable lipid re-protonates (becomes positively charged) and forms electrostatic ion pairs with anionic phospholipids (phosphatidylserine and bis(monoacylglycero)phosphate) in the inner leaflet of the endosomal membrane. Because the ionizable lipid has a small headgroup and splayed, branched polyunsaturated or esterified tails (a cone-shaped molecular geometry), its ion pair with anionic endosomal lipids transitions from a planar lamellar bilayer (L_α) into a membrane-disrupting non-bilayer inverted hexagonal (H_II) phase, puncturing the endosome to release intact mRNA or siRNA into the cytoplasm before lysosomal degradation. Furthermore, adding a fifth Selective Organ Targeting (SORT) lipid (Daniel Siegwart, 2020) alters the plasma protein corona to redirect LNP delivery beyond the liver to the lungs (via permanently cationic SORT lipids) or spleen (via anionic SORT lipids).

Why must the ionizable lipid in an mRNA or siRNA lipid nanoparticle have a p*K*_a between 6.2 and 6.5?

A p*K*_a of 6.2 to 6.5 allows the lipid to act as a three-stage electrostatic switch: at pH 4.0 during manufacturing, it is positively charged to bind and encapsulate negatively charged RNA; in the bloodstream at pH 7.4, it becomes neutral to avoid red blood cell lysis and toxicity; and inside the acidifying endosome (pH 5.5–6.0), it becomes positively charged again to fuse with anionic endosomal membranes and release the RNA into the cytoplasm.

How do intravenously injected lipid nanoparticles naturally target hepatocytes in the liver?

Lipid nanoparticles are formulated with ~1.5 mol% of a PEG-lipid anchored by short C₁₄ acyl chains (such as DMG-PEG₂₀₀₀). Once injected into the bloodstream, this short-chain PEG coat rapidly sheds off the nanoparticle surface, exposing the neutral lipid core. Circulating plasma Apolipoprotein E (ApoE) adsorbs onto the exposed LNP surface and binds to low-density lipoprotein receptors (LDLR) abundantly expressed on hepatocytes, mediating rapid endocytosis into the liver.

How does remote ammonium sulfate loading pack 10,000+ doxorubicin molecules into a single Doxil liposome?

Doxil liposomes are prepared with an internal 250 mM ammonium sulfate ((NH₄)₂SO₄) solution. Neutral ammonia (NH₃) gas diffuses out across the lipid bilayer, leaving protons behind and creating an acidic interior (pH ~5.2). When weakly basic doxorubicin is added outside at neutral pH, uncharged doxorubicin diffuses into the liposome, becomes protonated in the acidic interior, and precipitates with sulfate ions into insoluble crystalline bundles trapped inside the core.

1. Cullis, P. R., & Hope, M. J. (2017). "Lipid nanoparticle systems for enabling gene therapies." Molecular Therapy, 25(7), 1467–1475. https://doi.org/10.1016/j.ymthe.2017.03.013
2. Hou, X., Zaks, T., Langer, R., & Dong, Y. (2021). "Lipid nanoparticles for mRNA delivery." Nature Reviews Materials, 6(12), 1078–1094. https://doi.org/10.1038/s41578-021-00358-0
3. Barenholz, Y. (2012). "Doxil®—The first FDA-approved nano-drug: lessons learned." Journal of Controlled Release, 160(2), 117–134. https://doi.org/10.1016/j.jconrel.2012.03.020
4. Cheng, Q., Wei, T., Farbiak, L., Johnson, L. T., Dilliard, S. A., & Siegwart, D. J. (2020). "Selective organ targeting (SORT) nanoparticles for tissue-specific mRNA delivery and CRISPR-Cas gene editing." Nature Nanotechnology, 15(4), 313–320. https://doi.org/10.1038/s41565-020-0669-6