We have localized a fraction of the enzyme, purine nucleoside phosphorylase (PNP), to the centrioles and basal bodies of mammalian, avian, and protozoan cells. Two completely independent methods were used, one based on the ultrastructural cytochemistry of the enzyme activity and one based on immunofluorescence microscopy using an antibody raised in rabbit against purified human PNP. PNP catalyzes the reversible conversion of purine nucleosides and inorganic phosphate to the corresponding purine bases and ribose-1-phosphate. Its partial localization to centrioles and basal bodies raises the possibility that purine compounds are involved in centriole replication and/or in the regulation of microtubule assembly in vivo. No centriolar PNP could be detected in primary skin fibroblast from two infants with severe immunodeficiency disease associated with the absence of soluble PNP. This raises the possibility that defects in centriole function may contribute to the impaired division and maturation of T lymphoid precursor in this inherited disorder. Initially, the immunofluorescence analyses were complicated by a residual centriole-binding antibody that persisted in immunoglobulins from immune animals after complete removal of anti-PNP by affinity chromatography. Binding was abolished by exposure of cells to sodium periodate, indicating that this (and possibly other) "spontaneous" anticentriole antibodies in rabbit serum may be directed against carbohydrates.
The flagella of populations of three protozoan species (Ochromonas, Euglena, and Astasia) were amputated and allowed to regenerate. The kinetics of regeneration in all species were characterized by a lag phase during which there was no apparent flagellar elongation; this phase was followed by elongation at a rate which constantly decelerated as the original length was regained. Inhibition by cycloheximide applied at the time of flagellar amputation showed that flagellar regeneration was dependent upon de novo protein synthesis. This was supported by evidence showing that a greater amount of leucine was incorporated into the proteins of regenerating than nonregenerating flagella. The degree of inhibition of flagellar elongation observed with cycloheximide depended on how soon after flagellar amputation it was applied: when applied to cells immediately following amputation, elongation was almost completely inhibited, but its application at various times thereafter permitted considerable elongation to occur prior to complete inhibition of flagellar elongation. Hence, a sufficient number of precursors were synthesized and accumulated prior to addition of cycloheximide so that their assembly (elongation) could occur for a time under conditions in which protein synthesis had been inhibited. Evidence that the site of this assembly may be at the tip of the elongating flagellum was obtained from radioautographic studies in which the flagella of Ochromonas were permitted to regenerate part way in the absence of labeled leucine and to complete their regeneration in the presence of the isotope. Possible mechanisms which may be operating to control flagellar regeneration are discussed in light of these and other observations.