Where do new genes come from?
When a few members of a family have a gene not found in most other members, one explanation is that the gene is newly evolved, rather than inherited from the common ancestor of that family. (The other possibility is that their ancestor had it, but most descendants lost it.)
New genes often turn out to be copies of old genes, sometimes with modifications that give them very different functions. But a paper just published in Current Biology reports "Emergence of a new gene from an intergenic region", rather than duplication of an existing gene....
One definition of a gene is "a section of DNA which, if altered, has some effect on phenotype (observable traits)." Many genes are transcribed into messenger RNA and then translated into protein (perhaps an enzyme). But there is an increasing list of examples of genes whose phenotypic effects depend on activity by the RNA transcript, which may never be translated into protein. Such genes often have a regulatory function, controlling the expression of other genes.
The gene reported in this paper was detected as an RNA transcript in house mice and some of their close relatives, but not in more distant relatives like rats. Their analysis of the family tree for the gene suggests that it evolved about 3 million years ago, after the last common ancestor of mice and rats lived.
The new gene is apparently not translated into protein. Antibodies designed to detect the protein, if it is made, didn't find any. Also, evolutionary changes in the DNA sequence appear to be indifferent as to protein structure: DNA-base changes that would have no effect on protein amino acid sequence were about as common as those that would effect the protein.
There was a phenotypic effect, however, meeting our criterion for a gene. When the researchers knocked out the DNA corresponding to the transcript, sperm motility decreased. Apparently the RNA transcript had some beneficial effect on expression of other genes.
Although the DNA was in every cell, the RNA transcript was produced almost exclusively in the testis. Apparently conditions for gene expression are looser there than elsewhere, so that a variety of mutations can get a stretch of DNA transcribed. Most such mutations are presumably harmful, but the mice with beneficial mutations (like the one described in this paper) have increasing representation in successive generations.