Ferreting out the dynamics of ferritin expression

Ferritin is a remarkable molecule. A shell-like megapolymer of similar 20-kd light (L) and 21-kd heavy (H) subunits, it mediates a phase transition of soluble iron to solid ferric oxide, incorporating up to 2000 metal atoms into its mineral core. All living organisms exploit some form of this versatile molecule, using ferritin as an intracellular iron-storage depot. In mammals, small amounts of iron-poor ferritin escape from cells to circulate in the blood. Although the origin of serum ferritin is uncertain, it is widely considered to be a good clinical indicator of body iron stores.

Last year, Ferreira and colleagues made the important observation that H and L subunits must have nonredundant functions, because mice lacking H ferritin die early in gestation (Ferreira et al, J Biol Chem. 2000;275:3021-3024). Presumably, this is because H ferritin uniquely possesses a ferroxidase activity that is necessary for iron assimilation. In this issue (page 525), they extend their findings, showing that animals with only one functional H ferritin gene have half the normal levels of H ferritin but make up the difference by increased expression of L ferritin.

These results are intriguing for 3 reasons. First, they suggest that the L-to-H stoichiometry of isoferritins is less important than previously suspected, because the mutant mice have no evidence of abnormal iron homeostasis. Second, they indicate that regulation of ferritin gene expression is complex. Mechanisms of transcriptional and translational regulation have been studied in detail, but no existing information readily explains the dynamic relationship between H and L ferritin production in mice lacking one H ferritin gene. Finally, they suggest a new etiology for isolated hyperferritinemia in otherwise normal human patients.