Bacteria have a habit of sometimes reversing half their genome. In the picture above, this has evidently happened a few times (whether in M. smegmatis, M. tuberculosis, or both, is not clear from the picture, but we can be sure it has happened somewhere). If the two sites where the inversion happens (bacterial genomes are loops) lie within genes, those genes get crossed over, just like chromosomes in multicellular sex.
That's my insight for tonight. It's probably not original.