When Marcelo Melo, a genetics researcher at the University of Iowa, started studying keratins, they seemed innocuous. This family of more than 50 fibrous structural proteins makes up hair, skin, and nails.
But soon, he realized he was entering uncharted territory.
“People sometimes find interesting variants in these families of proteins, but they tend to ignore it or move on… because of the limitations in available assays,” he said. From colleagues to core facilities, everyone was reluctant to take on these tricky proteins.
“There are very few groups really looking at keratins because of how difficult it is to work with them,” Melo said. Melo needed better assays, and he found the answer in an article dating back more than 30 years: a largely forgotten method called far-western blotting. Overcoming skepticism from his colleagues, Melo showed that using the right tools can unlock novel insights into complex proteins that underlie human disease.
A hairy challenge
Melo first encountered keratins as a graduate researcher in the lab of Hatem El-Shanti, a geneticist at the University of Iowa who focuses on childhood diseases. At the time, El-Shanti was especially interested in a condition called loose anagen syndrome, where hair easily comes out of the scalp. “The disorder improves with the onset of puberty, so it doesn’t get a lot of attention,” El-Shanti said. “But it has a substantial impact on the child’s psyche and emotional status.”
The name of the condition refers to the anagen, or growth, phase of hair follicles. Loose anagen syndrome is thought to occur because the roots of growing hair follicles are malformed. Disrupted keratin fibers may prevent the hair from anchoring properly in the follicle, making it easy to uproot the hair with just a gentle pull.
In some families, loose anagen syndrome is inherited from generation to generation, suggesting the condition may have a genetic origin. Researchers in El-Shanti’s lab were working with one such family where, across three generations, there were eight individuals with loose anagen syndrome. Five of them agreed to provide samples for Melo and El-Shanti to study their condition.
Using genome sequencing, the researchers found that multiple affected members of the family had a mutation in the keratin 32 gene (KRT32), which is known to be involved in hair structure.1 The mutation changed one amino acid in the keratin protein, which Melo suspected might be enough to change the protein’s ability to bind another important keratin molecule and destabilize each follicle’s mechanism for holding hair in place.
However, proving this experimentally proved to be a challenge. The very same properties that make keratins strong building blocks for the body make them especially hard to study in the laboratory. They can’t be easily dissolved, digested, or purified, and their 3D assemblies are fortified by strong covalent bonds.
One traditional way to show that two proteins interact would be with a proximity ligation assay, which uses antibodies and oligonucleotide probes to assess whether two proteins are near each other. If a mutation interferes with binding, the proteins should no longer be close together. However, Melo knew this wouldn’t work well for keratins because they are structural proteins with fixed positions in the cell. Even if two keratin proteins failed to bind, they might still be stuck near each other.
He considered other protein-protein interaction assays that use antibodies to detect proteins, but they all had the same problem: the antibodies couldn’t distinguish between keratins because their amino acid sequences were so similar. Melo’s colleagues were similarly stumped. Even scientists at a university core facility, who had worked with hundreds of researchers on a wide range of protein targets, were hesitant to get involved in a keratin project. “It was hard to encounter that early on in the project,” Melo said. “It wasn’t very encouraging for us to continue pursuing this.”
Using an old method for a modern study
The solution to Melo’s problem appeared in an unexpected place: a little-discussed method first proposed in the early 1990s called far-western blotting.2,3,4 “This is an older method that has been somewhat abandoned,” said El-Shanti.
The name is a reference to the more widely used method, western blotting, that uses gel electrophoresis to separate proteins by size and antibodies to identify a specific protein. Far-western blotting extends this approach to identify protein-protein interactions. To determine whether two proteins bind, one of the proteins is considered the “prey” while the other is the “bait.” The prey is loaded onto a gel, either purified or as part of cell lysate. Through gel electrophoresis, the prey protein is isolated based on weight and transferred to a membrane, similar to western blotting. There, the protein is denatured and then renatured to allow it to adopt its native conformation so that, once the bait protein is added to the membrane, they can bind to each other. Binding is detected using a reporter system, which is typically an antibody tagged with horseradish peroxidase that can produce luminescent signal.
When Melo found the method, he was intrigued. “I had never heard of this,” Melo said. However, past studies only used it to find a protein’s potential binding targets. He couldn’t find any examples of researchers using far-western blotting to measure how a mutation changed protein-protein interactions. “But I was confident in my western blotting abilities and decided to give it a try,” he said.
Melo loaded purified keratin 32 with and without the mutation onto adjacent lanes of a gel as his prey. Then, he added its potential binding partner, keratin 82, as the bait. Two bands appeared: a faint band in the lane with the mutant protein and a much stronger band in the lane with the wild-type protein, indicating that the mutation had attenuated the interaction between the two keratin proteins.
Based on these data, the researchers proposed a model of how the attenuated binding between keratin 32 and 82 may interfere with the formation of hook-like structures on follicle cells that are required to anchor the hair in the shaft. The model is still hypothetical; conclusively proving it requires electron microscopy studies of actual hair follicles. Moreover, El-Shanti and Melo emphasized that loose anagen syndrome is a heterogeneous disease with distinct causes in different families. However, they still believe that this finding provides a useful basis for understanding the role of keratin in some forms of the disease. “We think a lot of the structural [forms of loose anagen syndrome] work in a similar way, by preventing anchoring,” Melo said.
Defying skepticism
For El-Shanti and Melo, their discovery of the mutant keratin’s functional consequence was more than just insight into the disease. It showed that this resurrected method wasn’t just a relic of the past: in combination with modern genome sequencing methods, far-western blotting could unlock critical functional insights about mutations.
Colleagues in the field had been skeptical. Reviewers critiqued their grant proposals for using outdated methods, and when Melo presented the work at conferences, he could see skepticism in his audience’s faces. However, the data always won them over. “Now, a lot of people remark that this is extremely cool,” El-Shanti said.
Melo believes that far-western blotting can make protein-protein interaction studies accessible to more labs. “It uses equipment and tools that essentially every lab has access to, rather than expensive kits or microscopes,” he said. For his study of loose anagen syndrome, he used previously published protocols, but found that they needed some modifications to work effectively in his experiment. For example, because his goal was to quantify change in binding based on band intensity, unlike previous studies that were just looking for binding, he needed better control over the amount of protein loaded onto the gel. To do this, he first quantified the prey protein to ensure equal amounts in all wells, and used a traditional western blot as a loading control to compare relative amounts of binding signal.
He also extended the protocol to use either a stain or an antibody to detect the prey so that the assay could work with even small amounts of protein. After all, to study keratin, he had to purify it from in vitro systems, which often yielded little protein. “We tried a variety of different assays,” he said. “The far western was the only one that we could get to work.”
He realized this refined protocol could be useful to other researchers. “Even people who are not studying keratins are probably encountering some of the same issues, and there were really no protein-protein interaction assays that we could use for this,” Melo said. This inspired him to publish his optimized protocol earlier this year in Frontiers in Molecular Biosciences.5
Melo hopes that having access to experimental methods like far-western blotting can help motivate researchers to study structural proteins involved in genetic diseases.
“My hope is that, when people encounter variants in genes such as keratins, they don’t immediately just move on and decide not to investigate it,” he said. “If the community realizes it’s possible with equipment and resources they have available at their lab, this will advance the field quite a bit.”
References
1. Melo M, et al. Heterozygous KRT32 variant is responsible for autosomal dominant loose anagen hair syndrome. HGG Advances. 2025;6(4):100495.
2. Kaelin WG, et al. Expression cloning of a cDNA encoding a retinoblastoma-binding protein with E2F-like properties. Cell. 1992;70(2):351-64.
3. Blackwood EM, et al. Max: a helix-loop-helix zipper protein that forms a sequence-specific DNA-binding complex with Myc. Science. 1991;251(4998):1211-7.
4. Wu Y, et al. Detecting protein–protein interactions by far western blotting. Nature Protocols. 2007;2(12):3278-84.
5. Melo M, et al. Quantifying cytoskeletal protein interactions with far Western blotting. Frontiers in Molecular Biosciences. 2026;13:1761723.