Cockroaches have a reputation for surviving anything. A new study suggests their genomes have been quietly doing the same thing for over 150 million years: soaking up bacterial DNA and keeping it, generation after generation, whether or not it serves any useful function.
The study, published June 15 in the Proceedings of the National Academy of Sciences, identified 40,485 individual pieces of DNA from the bacterium Blattabacterium cuenoti embedded in the genomes of 18 cockroach species. Individual cockroach genomes carried between 93 and 4,900 of these bacterial inserts — an order of magnitude more than any previously documented case of horizontal gene transfer in a multicellular animal.
“This is not a rare event,” said Kyle M. Ewart of the University of Sydney, the study’s lead author. “It has happened thousands of times, and in every cockroach species we looked at.”
Blattabacterium cuenoti is not an infection — it is an obligate intracellular endosymbiont, a bacterium that lives inside specialized fat body cells called bacteriocytes and is passed directly from mother to offspring through the egg. The relationship between cockroaches and Blattabacterium dates back to the common ancestor of cockroaches and termites, more than 150 million years.
The bacterium performs a critical metabolic function: it recycles nitrogen from uric acid waste into amino acids. This service is essential for termites, which live on wood, and has been retained in cockroaches even though their diet is more varied. Termite genomes, notably, contain zero Blattabacterium inserts — termites lost the bacterium early in their evolutionary history.
The mechanism by which bacterial DNA enters the cockroach genome is a form of horizontal gene transfer. Because Blattabacterium lives inside cockroach cells and is present in the germ line (eggs), its DNA occasionally leaks out of the bacterial cells. The cockroach’s own DNA repair machinery sometimes uses the available bacterial DNA as a patch for broken chromosomes. If this happens in a germ cell, the bacterial insert is inherited by future generations.
Why previous studies missed it
Earlier genome sequencing projects saw bacterial sequences in cockroach DNA and dismissed them as contamination. The assumption was that the bacterial DNA came from the insect’s gut microbiome or the environment, not from genuine genomic integration.
Two technical advances allowed Ewart’s team to prove otherwise. First, long-read sequencing technology (PacBio HiFi) produces DNA reads long enough to span the boundary between cockroach and bacterial sequences, showing that the two are physically joined in the genome. Second, the study set a minimum insert length of just 50 base pairs and did not require the inserted DNA to code for a protein — unlike previous surveys that only looked for complete, functional genes.
The result was 40,485 confirmed inserts across 18 cockroach genomes, with Australian Panesthiinae and Geoscapheinae cockroaches carrying the most — consistently over 3,000 per genome.
Mostly junk, but not useless
The vast majority of the bacterial inserts — 74 to 85 percent — are intergenic, meaning they sit between cockroach genes rather than inside them. Only 15 to 26 percent fall within gene regions. Transcriptome analysis on three species showed that 91 to 95 percent of the inserts are not transcribed at all. Only 0.40 to 0.65 percent mapped to exon sequences in RNA contigs.
In other words, the bacterial DNA is overwhelmingly non-functional junk. It persists not because it helps the cockroach, but because it does not hurt enough to be removed by natural selection.
Some of the inserts are surprisingly old. The researchers dated one set of shared inserts to at least 28.7 million years ago, placing them in the common ancestor of the Panesthiinae and Geoscapheinae cockroach subfamilies.
The researchers also found 32 chimeric inserts — sequences composed of up to nine different Blattabacterium segments from distant genomic locations, apparently fused together by the cockroach’s non-homologous end-joining repair pathway before being inserted into the genome.
What this means for horizontal gene transfer
Horizontal gene transfer — the movement of genetic material between distantly related species — has been documented in virtually every branch of life. Bacteria do it constantly, swapping antibiotic resistance genes on plasmids. Eukaryotes also do it, but the scale has generally been thought to be much smaller. The previous record for nuclear horizontal gene transfer in a multicellular eukaryote (outside of bdelloid rotifers) was roughly 282 inserts.
The cockroach results raise the question of whether other animals with long-term intracellular endosymbionts — aphids and their Buchnera, for instance, or cicadas and their Hodgkinia — might carry similar mountains of bacterial DNA that have simply been filtered out by assembly software trained to flag contamination.
“A lot of what was thrown away as contamination in previous genome projects might actually be real horizontal transfer events,” Ewart said.
Source:
- K.M. Ewart et al., “Uncovering thousands of endosymbiont DNA transfer events within single cockroach genomes,” Proceedings of the National Academy of Sciences 123(25), e2604240123 (2026). DOI: 10.1073/pnas.2604240123