The delivery problem

The most common question a newcomer asks about this whole field is the simplest one: why the needle? The answer is not preference or convention — it is physics and biochemistry, and it starts with what a peptide actually is. A peptide is a short chain of amino acids joined by peptide bonds. That bond is the exact chemical linkage your digestive system exists to take apart. Swallow a peptide and you are feeding it to a machine purpose-built to destroy it.

The teardown is sequential and redundant. Stomach acid (around pH 1–2) unfolds the molecule and begins hydrolysis; pepsin starts cleaving it. In the small intestine the pancreatic proteases — trypsin, chymotrypsin, the carboxypeptidases — cut it into fragments, and brush-border peptidases on the intestinal wall finish the job. What began as a signaling molecule is absorbed as its constituent amino acids: nutrition, not medicine. The body cannot tell your research peptide apart from a mouthful of egg white.

Suppose a fragment survives the gauntlet of the gut lumen. It still has to cross the intestinal wall to reach the bloodstream, and here peptides fail on a second, independent count: they are the wrong shape to pass a cell membrane. Most peptides are large, water-loving, often electrically charged, and studded with hydrogen-bond donors and acceptors. Medicinal chemists call this territory “beyond the rule of five” — peptides violate the classic drug-likeness guidelines on nearly every axis.

The wall offers two ways through and blocks both. The gaps between cells (the paracellular route) are sealed by tight junctions with angstrom-scale pores — far too small. The route through a cell (transcellular) means crossing a lipid bilayer that actively repels polar, charged molecules. The result is brutal: unmodified peptides typically show oral bioavailability under 1–2%, and for larger ones it rounds to zero.

If the gut is out, why not rub it on? Because skin is a barrier engineered by evolution to keep the outside world out. Its outer layer, the stratum corneum, is a “brick and mortar” lattice of dead cells in lipid — and it passively admits only small, fat-soluble molecules. The rough cutoff is famous enough to have a name: the 500-dalton rule. Molecules much heavier than that do not meaningfully permeate intact skin.

Even placed directly into the blood, a native peptide is often gone in minutes. Native GLP-1 has a circulating half-life of roughly one to two minutes — the enzyme DPP-4 clips it and the kidney clears the rest. This is why the drugs are not the raw hormones but engineered analogs: DPP-4-resistant substitutions, and fatty-acid chains (acylation) that let the molecule hitch a ride on albumin and evade clearance, stretching a two-minute signal into a multi-day one. That single trick is what turned a fragile gut hormone into a once-weekly therapy — the story told in how GLP-1 actually works.

“Oral” peptides are the most oversold idea in this space, and the barriers above are exactly why. Making a peptide survive the gut and cross the wall and do so reproducibly is one of the hardest problems in drug delivery — not a formulation you can achieve by putting powder in a capsule, dissolving it under the tongue, or calling it “liposomal.” When a product makes an oral or sublingual claim without rigorous, molecule-specific pharmacokinetic data behind it, the safe assumption is that most of the dose never reaches your bloodstream at all.

Consider the one true success story, precisely because it shows how much machinery is required. Oral semaglutide reaches the market only by being co-formulated with a permeation enhancer, SNAC, that locally buffers stomach acid and briefly loosens the gastric lining to let a sliver of drug slip through. Even then, bioavailability is around 1% — so the tablet must carry many times more drug than the injection to compensate. And it is fragile: it works only taken on an empty stomach under strict water and timing constraints, and absorption still varies widely from person to person and day to day. That is the state of the art for oral peptides — an enormous engineering effort to claw back a single percent, under conditions most real-world users won’t hold to.

None of this means oral peptides are impossible — enteric coatings, protease inhibitors, self-emulsifying systems (SEDDS), and ingestible microinjectors are all real research. But proof-of-concept in a lab is not the same as a product that works in your hand, and the gap between those two things is where most of the overselling lives.

The nose is the one needle-free route with a genuine, established track record — but only within strict limits, and understanding those limits is the whole point. The nasal mucosa is thin, richly vascularized, has real surface area across the turbinates, and drains straight into the bloodstream, skipping first-pass metabolism by the liver. That makes it attractive. It becomes efficacious only when three conditions line up at once:

• Small. Roughly under ~1 kDa crosses the mucosa reasonably; larger peptides need absorption enhancers and pay for size with lower, more variable uptake.
• Potent. Nasal bioavailability is often only single-digit-to-low-double-digit percent, so the route only works when the small fraction that does get in is still a therapeutic amount.
• Fast, or brain-bound. The nose suits molecules where rapid onset matters — and the olfactory and trigeminal pathways offer a partial 'nose-to-brain' shortcut that can sidestep the blood-brain barrier.

The proof is in what already ships nasally. Desmopressin (a small vasopressin analog) and the GnRH analogs nafarelin and buserelin are routine nasal sprays. Salmon calcitonin is delivered nasally despite being larger than the ideal cutoff — it gets away with it because it is extraordinarily potent. Oxytocin has a long nasal history, and the melanocortin agonist bremelanotide (PT-141) began life as an intranasal formulation before its delivery route changed.

The limits teach as much as the successes. Mucociliary clearance sweeps the spray toward the throat in roughly fifteen to twenty minutes, so the absorption window is short; the mucosa has its own enzymes; a head cold can tank uptake; and you can only instill a few microliters per nostril. The honest summary: the nose works for small, potent molecules where a low and somewhat variable absorbed fraction is still enough. It is a real route — but a narrow one, not a general escape from the needle.

This is a problem the field is actively, and cleverly, chipping away at. The most promising fronts:

• Designed-in permeability. Cyclic and bicyclic peptides, N-methylation, and 'stapled' peptides aim to build molecules that are protease-resistant and membrane-permeable from the start, rather than bolting on an enhancer.
• Ingestible devices. Capsule-scale robotic applicators (the self-orienting SOMA family) that inject a payload into the gut wall from inside the GI tract — sidestepping the absorption problem mechanically.
• Microneedle patches. Dissolving arrays that place a peptide just past the stratum corneum, painlessly, without a syringe.
• Smarter depots. Microsphere and in-situ-gel long-actings — already used for octreotide and leuprolide — that turn one injection into weeks or months of steady release.

The through-line is worth holding onto: the needle dominates not because the field lacks imagination, but because every alternative route is a fight against barriers that evolution spent a very long time perfecting. The winners will be the molecules and devices engineered specifically to beat one barrier at a time.

Established: the gut digests and blocks unmodified peptides (oral bioavailability typically under 1–2%); intact skin excludes molecules much above ~500 Da; native peptide half-lives are on the order of minutes; and injection remains the reliable default. A small set of small, potent peptides work nasally.

Open: whether permeability can be reliably designed into larger peptides; how far ingestible-device and microneedle delivery scale beyond early products; and — the reader’s practical question — which, if any, of the oral and sublingual products now sold actually deliver drug, a question only molecule-specific pharmacokinetic data can answer.