Isoform inflation and annotation heterogeneity can confound Kunitz-repertoire comparisons in blood-feeding animals: a gene-level reappraisal
This study demonstrates that isoform inflation and annotation heterogeneity in public genomes confound cross-phylum comparisons of Kunitz repertoires, revealing that apparent expansions in blood-feeding animals are often artifacts of data processing rather than genuine biological convergence.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
For millions of years, the ability to drink blood has evolved independently in many different groups of animals, from leeches and mosquitoes to ticks and vampire bats. To survive a meal, these creatures must overcome the host's natural defense system, which tries to stop bleeding and clot the blood. To do this, they inject a cocktail of proteins that act as anticoagulants, keeping the blood flowing. Scientists have long noticed that many of these animals use a specific type of protein structure, known as the Kunitz domain, to perform this task. Because this same tool appears in such different animals, it was widely assumed that the act of feeding on blood forces these species to evolve and expand their collections of these specific proteins, a pattern known as convergent evolution. The medicinal leech was often cited as a rare exception to this rule, possessing very few of these proteins compared to its blood-feeding cousins.
However, a new analysis suggests that this popular story might be built on a counting error rather than biological reality. Researchers re-examined the genetic blueprints of eight different animal species, including five blood-feeders and three that do not feed on blood. Instead of counting every single version of a protein listed in a database, which can include multiple slightly different copies of the same gene, they focused strictly on counting the unique genes themselves. They found that when they looked at the actual number of genes, the blood-feeding animals did not consistently have larger collections of these anticoagulant proteins than the non-blood-feeding animals. In fact, some animals that do not feed on blood had more of these genes than the average blood-feeder. The apparent abundance seen in earlier studies was largely an illusion created by how the data was organized; some genome projects list many more variations of the same gene than others, making those species look like they have a much larger arsenal than they actually do.
The study also investigated whether the leech was truly an exception with a small collection of these proteins. Even after correcting for the counting method, the leech still had very few of these genes, but the researchers found that the difference between leeches and other animals was not as clear-cut as once thought. More importantly, they discovered that the way scientists annotate genomes—how they label and count genes—varies wildly between different research groups and software programs. One team might count a single gene as dozens of different entries, while another counts it as just one. This inconsistency means that comparing protein lists across different species is like trying to compare the weight of apples to oranges when the scales are calibrated differently. The researchers showed that this variation in counting methods can inflate the apparent number of proteins by nearly five times in some cases, completely distorting the picture of what these animals actually possess.
Beyond simply re-counting the genes, the team asked a harder question: could we ever prove with certainty that blood feeding causes these protein families to expand? To do this, they would need to find pairs of closely related animals, one that feeds on blood and one that does not, with perfectly matched genetic data. They searched through all available public databases but found that such perfect pairs are incredibly rare. In fact, they could only locate one potential pair that met their strict standards for comparison. Furthermore, they ran computer simulations to see if they had enough data to detect a real pattern if one existed. The results showed that even if they gathered every available genome, the number of independent blood-feeding lineages is too small to provide a statistically strong answer. The data simply does not have enough power to confirm or deny the theory with the current tools and available species.
The researchers conclude that while blood-feeding animals certainly do use anticoagulants, the idea that they universally expand their collections of Kunitz proteins is not supported by the evidence when the data is cleaned up. The previous belief that these animals have a massive, convergent expansion of these genes appears to be a side effect of how the data was counted and organized, rather than a biological fact. The study serves as a cautionary tale for the field of evolutionary biology, reminding scientists that before declaring a pattern of convergence, they must ensure they are comparing apples to apples. Until genome data is standardized and more matched pairs of animals are available, the true extent of how blood feeding shapes these specific genetic families remains uncertain. The lesson is not that the animals don't use these tools, but that our current way of measuring them is too noisy to tell the full story.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.