New studies suggest that the accepted picture of everything the genome can encode, the full set of genes and the proteins they make, may be incomplete. The real picture may be far more complex.

In every organism sequenced so far, genes sit in a line along the chromosome, and each one is copied into RNA from its own location. That RNA message is then read to build a protein. A new study shows that proteins do not always come from a single gene. Pieces of the message from one gene can be spliced onto pieces from a gene on a different chromosome, producing a hybrid protein that no single gene encodes.

Hybrid Proteins Inside Immune Cells

The study focused on macrophages, the patrol cells of the immune system that swallow bacteria and raise the alarm. It also examined one hybrid built from the gasdermin D gene, Gsdmd, and a second gene called Tmem106a. To confirm that the hybrid protein exists, the study attached a small molecular tag to it and made an antibody that recognizes only that tail.

The study found more than 30,000 candidate hybrid messages in mouse macrophages. Some were present under normal conditions and others switched on by inflammation or by signals that promote tissue repair. The hybrid message rose in lung macrophages during influenza, in brain immune cells during bacterial meningitis and in abdominal macrophages after exposure to bacterial toxin. Human macrophages carried more than 900 hybrid messages. Thirty-three of them draw on the same parent genes as mouse versions. The question is: how exactly do these hybrids happen?

How Two Genes Become One Message

Most genes are small coding sections called exons. These are scattered through long stretches of DNA filled with non-coding sections called introns. When a gene is copied, the whole stretch becomes RNA. Splicing then cuts out the introns and joins the exons into a finished messenger RNA.

The same splicing machinery can join the messages of two different genes. In this process, called trans-splicing, exons from one gene’s RNA are joined to exons from another gene’s RNA. This produces a composite message that is read into a “hybrid” protein.

When a macrophage detects a bacterial signal, chromosomes that ordinarily sit apart inside the nucleus loop toward one another, bringing distant genes into contact. The loops depend on a protein that organizes how chromosomes fold. Blocking the splicing machinery erases the hybrid message. Removing the folding protein prevents the loops and the hybrid while the parent genes carry on normally.

Many Proteins From One Gene

One segment of DNA can encode many proteins. Alternative splicing lets the same gene include or skip different exons. Nearly all mammalian genes undergo it. Different start and stop sites add more variety. More than half of mammalian genes have alternative start sites, and at least 70% have more than one site where the message can end.

Proteins can change again after they are made. Cells attach phosphate groups, small proteins, fatty acids or sugar chains or cut a protein to activate it. Each change can alter what the protein does, where it goes in the cell and how long it lasts.

Telling Real Hybrids From Look-Alikes

Trans-splicing is well documented elsewhere. Trypanosomes and nematodes splice a short leader sequence onto the front end of many messages, and trypanosomes onto all of them, because their genes are arranged in long runs. Flowering plants need it to assemble several mitochondrial genes. The new study suggests mammals may do it too.

Three look-alikes can mimic it. A DNA deletion or rearrangement can fuse two genes so they are copied as one. Copying can run past the end of one gene into its neighbor, producing a double-length message that is then spliced normally. Sequencing enzymes can also hop between RNA molecules, creating false hybrids.

The study read RNA directly, and DNA tests found no rearrangement behind the Gsdmd hybrid, whose parent genes sit on separate chromosomes. Many candidates confirmed by a second method joined close neighbors, where run-on copying remains possible. Proof demands the right cells and exacting methods.

Trans-splicing adds a new type of protein variation to that list. It doesn’t appear to be random. It appears to be controlled. Specific signals switch it on. It also depends on how chromosomes fold, and it joins the same pieces at the same place each time. It may also depend on context and differ between cell types. If so, trans-splicing greatly expands the range of working proteins that a set of genes in DNA can encode.