What Is a Molten Globule and Why Does It Matter?

A molten globule is a partially folded form of a protein that keeps its overall compact shape and basic structural scaffolding but loses the precise, locked-in arrangement of atoms that makes a fully folded protein work. Think of it as a protein caught between two states: not yet unraveled into a floppy chain, but not snapped into its final rigid shape either. The concept emerged in the early 1980s, and it has turned out to be far more than a laboratory curiosity. Molten globules show up in normal cell processes, in the progression of diseases like Alzheimer’s and Parkinson’s, and in the shelf-life problems that plague biologic drugs.

A Protein Halfway Between Order and Chaos

To picture a molten globule, it helps to know that a finished protein has two layers of organization that matter here. The first layer consists of local patterns along the chain: stretches that coil into spirals or lie flat next to each other like the pleats of a fan. The second layer is the way those local patterns pack together in three-dimensional space, locking side chains into specific contacts that give the protein its unique shape and function. A molten globule keeps most of the first layer intact while losing the second. The protein stays roughly the same size as its finished form but its interior is loose and mobile, with side chains shifting around rather than sitting in fixed positions.

Early work using disulfide bonds to probe the structure confirmed exactly this picture: a compact conformation with native-like secondary structure content but poorly defined tertiary structure.¹ Multiple biophysical techniques, including circular dichroism and fluorescence measurements, can detect the hallmarks of a molten globule: the presence of helices and sheets alongside the absence of specific fixed contacts between distant parts of the chain.²

That looseness in the interior has a practical consequence. Hydrophobic side chains that are normally buried and hidden in a native protein become partially exposed in a molten globule, making the surface stickier. This increased stickiness is both useful and dangerous, as we will see throughout the rest of this article.

How Scientists Spot One

A single technique rarely clinches the identification. Researchers typically combine several spectroscopic methods, each of which reveals a different aspect of the protein’s state. Circular dichroism in the far-ultraviolet range shows whether helices and sheets are present; near-ultraviolet circular dichroism and fluorescence anisotropy reveal whether side chains are locked in place. Fluorescent dyes that bind to exposed hydrophobic patches light up in a molten globule but stay dark in a fully folded protein. Nuclear magnetic resonance can show whether hydrogen atoms throughout the chain are in fixed or mobile environments.

Infrared spectroscopy has proved especially useful for small proteins whose structure is dominated by flat sheets rather than spirals. Because it can simultaneously probe both the secondary structure content and the compactness of the protein, FT-IR is well suited to catching temperature-induced molten globule states that might be missed by other methods.³ In practice, the combination of multiple readouts is what gives researchers confidence: if a protein is compact, retains secondary structure, and has a mobile interior with exposed hydrophobic surface, the case for a molten globule is strong.

What Pushes a Protein Into This State

Molten globules can be triggered in the lab by changing the environment around a protein. Lowering the pH is one of the most reliable methods. When the solution becomes acidic, charges on the protein shift, weakening the precise contacts that hold the interior together. Studies on interleukin-2 showed that between pH 4 and pH 2, the protein’s tertiary structure gradually loosened while its overall size and secondary structure stayed essentially the same.⁴ Similar effects appear at low concentrations of chemical denaturants, and mild heat can do the job as well.

The same acid-induced loosening has been documented in the prion protein, the molecule implicated in mad cow disease and related conditions. At pH 2.0, the prion protein adopts a molten globule-like state with native-like secondary structure and increased hydrophobic surface area.⁵ This observation matters because conditions inside certain cellular compartments can be quite acidic, meaning proteins may transiently visit molten globule states during their normal life cycle inside the cell.

A Key Pit Stop on the Folding Road

Evidence accumulated since the 1960s suggested that proteins might pass through states intermediate between rigid and fully unfolded. The Russian biophysicist Oleg Ptitsyn proposed in the 1970s that such an intermediate played a role in how proteins fold, and a compact, partly folded species was first observed experimentally in the early 1980s. The term “molten globule” was coined in 1983.²

Since then, the molten globule has been recognized as a state that lies on both the kinetic and sometimes the thermodynamic pathway between a protein’s unfolded chain and its finished form.⁶ Understanding the properties of this intermediate helps researchers work out how a protein navigates from a floppy strand of amino acids to a precisely folded machine, and why the process sometimes goes wrong.

Dry Versus Wet Interiors

Not all molten globules are the same. Research on the protein barstar revealed at least two distinct flavors. In a “dry” molten globule, the hydrophobic core is still sealed off from water even though the side chains within it have started to rearrange. The protein’s interior is looser than in the native state, but solvent has not yet penetrated. In a “wet” molten globule, water molecules have broken into the core, making it even more mobile and solvent-accessible while the overall secondary structure still holds.⁷

The distinction matters because the dry form appears earlier on the unfolding pathway and is closer to the native state, while the wet form sits further along toward full denaturation. Different experiments may capture different points on this spectrum, which partly explains why early literature sometimes gave conflicting descriptions of what a molten globule “really” looks like. The answer is that there is a range, and proteins can stall at different points along it depending on conditions.

Doing Real Work Inside Cells

The molten globule is not just a laboratory artifact or an accidental halfway house during folding. It has been predicted and confirmed experimentally that molten globule states exist in living cells and play roles in normal physiology.⁸ Two of the best-understood functional roles involve getting proteins across membranes and helping molecular chaperones do their jobs.

Many proteins need to cross lipid membranes to reach the cellular compartment where they function. A fully folded, rigid globule is too bulky and inflexible to thread through a membrane pore efficiently. Early proposals suggested that molten globule states, being compact yet flexible, could serve as suitable intermediates for translocation across biological membranes.⁹ Work on bacterial toxins later confirmed this idea: the insertion intermediates that embed themselves in target-cell membranes are not fully unfolded but have the features of a molten globule state, the same state observed at early stages of folding.¹⁰

Chaperones, the cellular helpers that prevent proteins from clumping together, also interact with molten globule forms. The chaperone alpha-crystallin, which is abundant in the eye lens, appears to recognize proteins in an aggregated molten globule state by sensing their large increase in hydrophobicity compared with the soluble molten globule forms that appear during normal folding.¹¹ In bacterial secretion systems, the translocator protein PopB exists as a molten globule that is stabilized by binding its dedicated chaperone PcrH, forming a complex with partial rigid structure but few tertiary contacts.¹² In both cases, the looseness of the molten globule is the feature, not a defect: it allows the protein to be recognized, stabilized, or reshaped by its partner.

Native Molten Globules and Intrinsically Disordered Proteins

Some proteins never fold into a rigid shape at all. These intrinsically disordered proteins function while remaining flexible, and researchers have recognized that “native molten globules” represent the most folded end of the disorder spectrum. A study of the nuclear coactivator binding domain of CBP showed that it exists as a native molten globule, adopting a compact structure with fluctuating secondary and tertiary contacts that allows it to bind multiple different partners.¹³ The flexibility lets these proteins sample many shapes, increasing the odds that they will find one that fits whichever partner happens to be nearby. This “conformational selection” mechanism has implications for signaling and gene regulation, where proteins often need to interact with many different targets.

The Dark Side: Aggregation and Disease

The same exposed hydrophobic surface that makes molten globules useful for membrane crossing and chaperone binding also makes them prone to sticking to one another. When molten globule-like intermediates accumulate, they can clump into insoluble aggregates. In more than fifteen different human diseases, misfolding intermediates with molten globule-like features play a key role in driving aggregation and amyloid formation.¹⁴

The process typically involves a partially folded conformation, sometimes called a pre-molten globule state, that shares many properties with the molten globule but is even more flexible. This flexibility is what allows the necessary rearrangements to form the cross-beta structure that defines amyloid fibrils.¹⁵ Because the pre-molten globule form is more flexible than the mature molten globule, it more easily rearranges into the repeating flat-sheet architecture that makes amyloid fibrils so stable and hard for the body to clear.¹⁶

Connections to Parkinson’s and Alzheimer’s

Alpha-synuclein, the protein that accumulates in the brains of people with Parkinson’s disease, is natively unstructured but adopts partially folded conformations that serve as key intermediates in aggregation and fibrillation.¹⁷ Laboratory experiments have shown that oligomers of alpha-synuclein enriched with loosely held, molten globule-like structure exhibit greater toxicity toward neuronal cells than more tightly organized aggregates. Strikingly, these toxic oligomers did not transform into compact amyloid fibers even after extended incubation, suggesting that the molten globule-like form itself can be the damaging species, not merely a stepping stone on the path to fibrils.¹⁸ Chemical modification of alpha-synuclein by ribose, a process called glycation, also induces soluble molten globular dimers and polymers that are highly cytotoxic.¹⁹

Alzheimer’s disease has its own molten globule connection through apolipoprotein E4, the strongest genetic risk factor for the common form of the disease. ApoE4 undergoes domain interaction, a structural rearrangement that results in a compact form and molten globule formation, producing stable, reactive intermediates with potentially pathological activities.²⁰ Researchers have proposed using small-molecule “structure correctors” to prevent this domain interaction and convert apoE4 into a form resembling the less harmful apoE3 variant.²¹ The idea is conceptually elegant: if you can keep the protein from forming its pathological molten globule intermediate, you may be able to short-circuit the downstream damage.

Trouble for Biologic Drugs

Biologic drugs, such as monoclonal antibodies used in cancer treatment and autoimmune disease, are proteins. That means they are susceptible to the same molten globule pitfalls that affect any protein. During manufacturing, antibodies often pass through low-pH steps meant to inactivate viruses. Even brief exposure to acidic conditions can push some antibody molecules into a transient molten globule state. These partially unfolded monomers then face a fork in the road: they can refold back to their native shape, or they can collide with other molten globule monomers and form aggregates.²²

Aggregated antibodies are a problem for two reasons. They lose their therapeutic activity, and they can trigger unwanted immune responses in patients. Understanding that molten globule intermediates drive this aggregation pathway through successive addition of partially unfolded monomers has helped manufacturers optimize pH-hold times, choose stabilizing excipients, and design formulation conditions that minimize the window of opportunity for these sticky intermediates to find each other.

Watching Molten Globules on a Computer

Because molten globules are transient and structurally heterogeneous, pinning them down experimentally is hard. Molecular dynamics simulations offer a complementary view. Early simulations of the unfolding of bovine pancreatic trypsin inhibitor produced a partially denatured state that was compact but expanded by roughly 11 to 25 percent relative to the native state, with overall secondary structure preserved and local unfolding concentrated in loop regions. A hydrophobic core was maintained, though side-chain packing was compromised. These properties were consistent with a molten globule, providing structural details that experiments alone could not resolve at the time.²³

Simulations of human alpha-lactalbumin, one of the most studied molten globule systems, generated ensembles of compact structures that retained substantial secondary structure and an overall native fold but had very different side-chain contacts from one snapshot to another.²⁴ More recently, constant-pH molecular dynamics, which directly couples the protonation state of amino acids to the protein’s shape, has been able to capture molten globule formation at acidic pH in silico, matching previous experimental observations.²⁵ These computational tools are increasingly valuable for predicting which conditions will push a given protein into a molten globule state, something that matters both for basic science and for industrial protein formulation.

Liquid-Liquid Phase Separation and Viral Factories

One of the newer connections to emerge involves liquid-liquid phase separation, the process by which proteins and other molecules condense into droplet-like compartments inside cells without a surrounding membrane. These condensates are involved in everything from gene regulation to stress responses. Work on the phosphoprotein of respiratory syncytial virus showed that its ability to undergo phase separation and form liquid-like viral factory structures is governed by a molten globule domain within the protein.²⁶ The flexible, partially folded nature of the molten globule domain enables the weak, multivalent interactions that drive droplet formation, while adjacent sequences act as a brake to keep the process in check. This finding opens a new frontier: molten globule domains may turn out to be common engines of biological condensation, linking the decades-old concept of partial folding to one of the hottest areas in cell biology.