Lyoprotectants and bulking agents: what else is in a lyophilized peptide vial

Open two vials of lyophilized peptide and they can look entirely different. One holds a firm white cylinder that keeps its shape when tilted. The other holds a barely visible film on the glass, or a few flakes that scatter with static when the stopper comes out. The difference is rarely the peptide. It is what was dissolved alongside the peptide when the solution was frozen — and in a large fraction of research-grade material, the answer is nothing at all.

Excipients in a lyophilized formulation are not filler. They perform two functions that are often conflated: building a physical cake with acceptable mechanical and reconstitution properties, and chemically protecting the solute through the freezing and drying transitions. These are different jobs, accomplished by different classes of compound, and the material that performs one well is frequently useless at the other. Understanding which is present, and which is absent, explains a good deal of the variation in how nominally identical lyophilized peptides behave in storage.

The two distinct stresses of freeze-drying

It is convenient to speak of “lyophilization stress” as a single thing, but the process imposes two separable insults with different mechanisms and mitigations.

Freezing stress comes first. As ice nucleates and grows, everything that is not water is excluded from the crystal lattice and concentrated into a shrinking interstitial channel. A solute that began at millimolar concentration can reach tens of percent by weight in that freeze-concentrate. Concentration alone drives association reactions with non-linear rate dependence, but the more consequential effect is on the buffer, whose components do not concentrate uniformly: the less soluble species saturates first and crystallizes out, leaving its counterpart behind in solution. Sodium phosphate is the textbook case: the dibasic salt crystallizes as Na₂HPO₄·12H₂O, and the remaining solution becomes acidic. Reported pH shifts run to three units, and in phosphate-buffered saline as much as four, with the drop beginning shortly after ice nucleation at temperatures between roughly −0.5 and −4 °C. A peptide formulated at neutral pH can therefore spend the freezing step at pH 4 or below, at high concentration, against a large and growing ice-water interface — a surface that promotes unfolding and interfacial aggregation much as an air-water interface does.

Drying stress follows. Sublimation removes the ice, then secondary drying removes water that was never frozen: the hydration shell hydrogen-bonded to the solute itself. For a structured peptide, that shell is part of what holds the conformation together, and its removal is a thermodynamic event independent of anything that happened during freezing. A formulation can survive freezing intact and still be damaged during drying, which is why cryoprotection and lyoprotection are sometimes specified as separate requirements even when one excipient provides both.

Interfacial stress is the thread connecting them, and it is the one case where a surfactant rather than a sugar is the appropriate tool. Low concentrations of polysorbate 20 or 80 compete with the peptide for the ice-water and air-water interfaces, reducing surface-driven aggregation. The tradeoff is that polysorbates are themselves chemically unstable — they autoxidize to peroxides and hydrolyze to free fatty acids — so a surfactant added to solve an interfacial problem can introduce an oxidative one, particularly for sequences containing methionine, cysteine, or tryptophan.

How sugars protect: water replacement and vitrification

Two mechanisms are invoked to explain why disaccharides stabilize solutes through drying, and the current understanding is that both operate rather than that one is correct.

The water replacement hypothesis is chemical. As the hydration shell is stripped, sugar hydroxyl groups hydrogen-bond to the polar groups the water vacated, substituting for the lost interactions and preserving local structure. This requires intimate molecular contact between sugar and peptide — the central constraint on excipient selection, since an excipient that separates into its own phase cannot satisfy it.

The vitrification hypothesis is physical. The dried sugar forms an amorphous glass of extremely high viscosity in which the peptide is kinetically immobilized. Motions required for unfolding, association, or bimolecular chemistry slow by many orders of magnitude, so degradation that is thermodynamically favorable does not proceed on a useful timescale. The governing parameter is the glass transition temperature, Tg. Below Tg the matrix behaves as a rigid glass; above it, mobility rises steeply and the protective effect degrades. Two glass transitions matter for different reasons: Tg′, the glass transition of the maximally freeze-concentrated solution, sets the temperature ceiling for primary drying, above which the cake collapses; the Tg of the finished dry solid sets the ceiling for storage and shipping.

Both mechanisms predict the same failure mode. Anything that removes the sugar from molecular contact with the peptide — crystallization, phase separation — abolishes protection, and anything that lowers Tg toward ambient temperature does the same. Residual moisture does both, which is why water content and excipient selection cannot be evaluated independently.

Crystalline bulking agents versus amorphous stabilizers

Mannitol is the most widely used bulking agent in freeze-dried formulations, precisely because it crystallizes. A crystalline matrix produces a mechanically robust cake, and its high eutectic temperature permits primary drying at higher product temperature with faster sublimation — a real manufacturing advantage. What crystalline mannitol does not do is protect the peptide. Having crystallized into its own phase, it is no longer in molecular contact with the solute and satisfies neither the water replacement nor the vitrification requirement. A mannitol-only formulation is a formulation with an excellent cake and no chemical protection.

Mannitol also carries a specific and underappreciated liability. It can crystallize as a hemihydrate, and the physical form that appears depends on process conditions: hemihydrate formation is favored when solute crystallization occurs at temperatures at or below about −20 °C, while anhydrous forms are favored at or above about −10 °C, with the transition reported near −15 °C. Annealing above that boundary is the standard mitigation. This matters beyond cake appearance: hemihydrate retained in the finished product is a reservoir of water held in the crystal lattice. If it dehydrates during storage, that water is released into the amorphous phase, plasticizing the glass and raising mobility — a lyophile that reads dry by total moisture assay yet behaves as though it were wet. Composition matters too: sugars inhibit mannitol crystallization more strongly than proteins do, and in their presence mannitol has been reported to crystallize preferentially as the hemihydrate, so a mixed sugar-mannitol formulation is not the sum of its parts.

Commercial practice therefore uses both classes deliberately: a crystallizing bulking agent for structure, and an amorphous disaccharide above a minimum mass ratio to the peptide for protection. Glycine serves as a crystallizing bulking agent much as mannitol does, with its own polymorph considerations.

Sucrose versus trehalose

Among amorphous stabilizers the practical choice is usually between two non-reducing disaccharides, and they are not interchangeable.

The clearest difference is thermal. Trehalose forms a dry glass with a Tg measured by differential scanning calorimetry at or above 120 °C, among the highest of the disaccharides; sucrose sits far lower, generally reported in the range of 60–70 °C. Both are comfortably above ambient in principle, but the margin is what matters once residual moisture depresses Tg and a shipment sits on a loading dock in summer. Trehalose formulations also tolerate a higher product temperature during primary drying by virtue of a higher Tg′, which shortens cycles. Absolute Tg′ values reported in the literature vary substantially with measurement method and concentration, but the ordering — trehalose above sucrose — is consistent.

The chemical difference concerns reactivity toward the peptide. Both sugars are non-reducing as supplied, meaning neither presents a free anomeric aldehyde available to condense with a primary amine. Sucrose, however, is a glycoside vulnerable to acid-catalyzed hydrolysis, and inversion yields glucose and fructose, which are reducing. Those products react with lysine side chains and the N-terminal amine via the Maillard pathway, producing glycated species that are heterogeneous, chromatographically messy, and mass-shifted. In freeze-dried model systems the Maillard reaction has been reported to proceed roughly three orders of magnitude more slowly in trehalose than in sucrose. The route to this failure is exactly the acidic freeze-concentrate described above: a formulation buffered at neutral pH that acidifies during freezing creates the conditions for sucrose inversion, and a lysine-rich sequence supplies the amine.

Trehalose has a countervailing weakness. It readily forms a crystalline dihydrate, and crystallization removes it from the amorphous phase with the same loss of protection that afflicts mannitol. Sucrose, by remaining amorphous itself, has been shown to suppress trehalose dihydrate crystallization — one of the motivations for mixed-sugar systems.

What this means for a research vial

Most of the above describes formulated products. A large share of research-grade peptide is not formulated at all — it is lyophilized directly from a purification eluate, typically water and acetonitrile with trifluoroacetic acid, with no excipient added. The result is a peptide-only solid: no bulking agent, hence the thin film or the scatter of static-prone flakes rather than a cake; and no amorphous stabilizer, hence protection that depends entirely on the peptide’s own glass-forming behavior and its residual moisture. For a short peptide with no oxidizable or deamidation-prone residues this is often adequate. For a longer, structured, or chemically labile sequence it is a meaningfully less protected state than a formulated equivalent, and worth knowing which is in hand.

Excipients are also largely invisible to the standard analytical panel. Mannitol, sucrose, and trehalose have no appreciable absorbance at 214 or 280 nm and are essentially unretained under reversed-phase conditions, so they elute near the void and contribute nothing to an area-percent purity figure. A vial that is half excipient by mass can report the same chromatographic purity as one that is pure peptide. What excipients do affect is every calculation anchored to vial mass: a labeled quantity may refer to peptide content or to total solids, and reconstituting to a target concentration on the assumption that the entire contents are peptide will be wrong by whatever fraction is sugar. Amino acid analysis or another peptide content method, rather than the mass on the label, is the reliable anchor — the same gap between chromatographic purity and peptide content that counterion and moisture content produce.

The cake itself is the one piece of formulation information available without instrumentation. A cake that has shrunk from the vial wall, slumped, or fused into a glassy plug has passed above its collapse temperature at some point, during drying or afterward. Material that dissolves noticeably more slowly than it once did, or that produces transient haze on reconstitution, is reporting something about its physical state. Neither observation identifies a degradation product, and neither substitutes for analysis. But both are free, and in a category where excipient composition is frequently undisclosed, the appearance of the solid is often the only formulation data a recipient actually has.