Even though the housefly Musca domestica shows clear circadian rhythms in its behavioural and physiological processes, a circadian pacemaker system controlling these rhythms has not yet been described morphologically in this species. In M. domestica, neurons immunoreactive to pigment-dispersing factor (PDF), a neurotransmitter/neuromodulator of circadian information arising from a circadian clock and transmitted to target cells, are similar in their number and distribution to the PDF neurons of Drosophila melanogaster. In D. melanogaster these neurons co-localize PER protein and have been identified as clock neurons in that species. Here we report PDF-immunoreactive cells in the housefly's brain during postembryonic development in the larval and pupal stages, as well as in the adult fly soon after eclosion. In the housefly's brain, there are three groups of PDF-immunoreactive neurons: two groups with small (sPDFMe) and large (lPDFMe) cell bodies in the proximal medulla of the optic lobe; and one group in the dorsal protocerebrum (PDFD). Three out of four sPDFMe can be detected during the first hour of larval development, but the fourth sPDFMe is observed in the larva only from 48 hours after hatching, along with five lPDFMe neurons, seen first as two subgroups, and three out of four PDFD neurons. During postembryonic development these neurons show changes in their structure and immunoreactivity. New PDF neurons are observed during pupal development but these neurons mostly do not survive into adulthood. In the adult fly's brain, the PDF neurons have also been examined in double-labelled preparations made with a second antibody directed against the product of one of several clock genes: period (per), timeless (tim), or cryptochrome (cry). Among them, only immunoreactivity to CRY-like protein has been detected in the brain of M. domestica and has shown a daily rhythm in its concentration, as examined immunocytochemically. CRY was co-localized with PDF in the sPDFMe of the housefly's brain fixed during the day. The possibility that the sPDFMe neurons are the housefly's clock neurons is discussed.
