A five-minute sublingual window for a GLP‑1 peptide

Peptides are not allergic to innovation. They are allergic to shortcuts.

The big reason injectable glucagon-like peptide‑1 (GLP‑1) receptor agonists became a category is obvious: they work. The less glamorous reason delivery remains the moat is also obvious too: most peptides do not cross biological barriers easily, and the routes that do work tend to be the ones we already know how to control.

A new International Journal of Pharmaceutics paper claims a step toward that moat: a long‑acting, dual fatty‑acid–conjugated GLP‑1 receptor agonist called TE‑8105 (previously tested in a Phase 1/2a setting) can be dosed sublingually in diabetic mice with durable metabolic effects, if you get the formulation and the “under‑the‑tongue” timing right (PubMed).

Why sublingual is tempting, and why it usually disappoints

If you could make a peptide work through the mouth, you get a set of practical advantages all at once: you bypass stomach enzymes, dodge some of the gut’s protease gauntlet, and (in principle) avoid some of the food-effect headaches that come with oral peptide approaches.

But the mouth is not a loophole. The mucosa is still a barrier, contact time is finite, and the total amount of drug that can plausibly cross is often small.

That is why a lot of “needle‑free peptide” work ends up being a game of margins: small increases in permeability, small increases in residence time, small improvements in stability, and then hoping the pharmacology is forgiving enough that those margins add up.

A recent formulation-focused octreotide study makes the point in a different disease area. Octreotide can be given as an injection or as an oral capsule, but oral exposure is low and dosing can be constrained by food effects. A buccal film approach tries to trade those constraints for a controlled mucosal interface, where permeation enhancers and film microstructure become the knobs you can actually turn (PubMed).

What TE‑8105 is, in the only way that matters for this story

This is not primarily a “new receptor biology” story. It is a pharmacokinetics plus formulation story.

The authors describe TE‑8105 as a long‑acting GLP‑1 receptor agonist with two fatty acid chains. Their core hypothesis is simple: fatty‑acid conjugation might do more than extend half‑life after injection. It might also help with mucosal permeability, or at least make the absorption math less punishing.

If you only absorb a small fraction of a dose, a short‑half‑life peptide can still fail because the absorbed fraction vanishes too quickly. A longer‑half‑life molecule can sometimes make low absorption feel less catastrophic.

The “sweet spot” they say mattered in mice

In db/db diabetic mice, they systematically tuned several parameters and used metabolic outcomes (blood glucose, body weight, intake measures) as the readouts.

The combination they report as most effective was:

3 mg/mL TE‑8105 in a liquid formulation
pH 7.6
~5 minutes of mucosal contact time

They then adapted those parameters into prototype soft gel and tablet formats.

Two formulation details stand out because they are the kind of things that often become the practical bottleneck later:

Excipient synergy: phenol (0.3–0.5%) and mannitol (4%) reportedly improved efficacy beyond either alone.
A “real product” tell: peppermint oil further improved the tablet’s performance, which reads like an early attempt to solve the human-factor problem of “will anyone actually hold this under their tongue for long enough.”

The comparison that explains why the team cares about half-life

The paper’s most market-relevant claim is not that sublingual dosing “worked.” It is the durability angle.

They report that sublingual TE‑8105 and semaglutide achieved similar therapeutic potency on the order of ~7% relative to the subcutaneous route, but TE‑8105’s longer plasma half‑life supported once-every-2-days dosing for sustained glycemic control, whereas semaglutide required daily administration under their experimental setup (PubMed).

There is a quiet, important point hidden in that comparison: for needle‑free peptides, you do not always need high bioavailability. You need predictable bioavailability and a pharmacokinetic profile that makes the absorbed fraction matter.

The grown-up constraints (and why they matter more than the mouse curves)

It is hard to read any sublingual peptide paper without immediately asking the unromantic questions:

Variability: human mucosa, saliva volume, and technique are messy. If exposure varies widely, the whole format may become clinically awkward.
Behavioral compliance: five minutes under the tongue is plausible. But it is also long enough to be annoying, especially at scale.
Local tolerability: excipients that help absorption can also irritate mucosa. That trade-off decides whether a delivery idea becomes a product.
Manufacturing realism: “prototype tablet” is not “stable commercial tablet.” Long‑acting fatty‑acid conjugates can be finicky.

None of those uncertainties negate the signal. They define what the next signal has to look like.

What would actually move this from interesting to real

For this story to graduate from “promising mouse formulation” to “format shift,” the next persuasive data is boring on purpose:

human pharmacokinetics and pharmacodynamics for sublingual TE‑8105 (or a close analog)
tolerability and usability in a realistic dosing routine
a clean read on variability across people

If those pieces line up, the takeaway is not just that TE‑8105 found a clever trick. It is that fatty‑acid conjugation plus an oromucosal interface might be a repeatable pattern for certain peptide classes, not a one-off.