1. The subcellular distribution of 2-oxoglutarate:glyoxylate carboligase was investigated in a normal human liver, a liver from a patient with pyridoxine-resistant primary hyperoxaluria type I and rat livers subjected to various degrees and types of trauma. 2. On continuous sucrose gradients most of the carboligase fractionated with a peak equilibrium density of 1.19–1.20 g/cm 3 and paralleled the distribution of the major peaks of monoamine oxidase, glutamate dehydrogenase and cytochrome oxidase and can be considered to be mitochondrial. Various proportions of the carboligase and mitochondrial marker enzymes were found to be ‘extramitochondrial’ (at or near the top of the sucrose gradients), depending on the liver source and the severity of trauma to which they were subjected. 3. Carboligase, monoamine oxidase (outer membrane marker) and glutamate dehydrogenase (matrix marker) were released from mitochondria by the homogenization and centrifugation procedures, to the extent of 19.9%, 32.4% and 11.5% respectively in hyperoxaluric liver, 12.5%, 17.9% and 8.2% in normal human liver and 3.0%, 4.9% and 3.8% in control rat liver. The proportion of extramitochondrial cytochrome oxidase (inner membrane marker) was virtually undetectable in both human and rat livers. However, sonication of rat liver homogenates or the addition of the detergent Triton X-100 caused a massive release of all four enzymes. 4. The extramitochondrial carboligase was probably in the form of a free protein of very high molecular weight or aggregate, rather than associated with a mitochondrion-derived organelle. 5. Subfractionation of a rat liver mitochondrial preparation indicated that most of the carboligase activity paralleled activities of 2-oxoglutarate decarboxylase, citrate synthase and glutamate dehydrogenase and was probably located in the matrix. The data were compatible with a previous suggestion that most of the carboligase is identical with the first decarboxylating component of the 2-oxoglutarate dehydrogenase complex. However, a significant minority of the carboligase activity showed a distribution different from that of 2-oxoglutarate decarboxylase, possibly due to structural alterations in the multi-enzyme complex. 6. We conclude that 2-oxoglutarate:glyoxylate carboligase is probably wholly mitochondrial and that the apparent cytosolic form of this enzyme is due to mitochondrial damage. Therefore previous suggestions that the lack of this component causes primary hyperoxaluria type I are no longer tenable.
1. Colony-stimulating activity appeared in the perfusate of the isolated rat liver during perfusions with either whole rat blood, rat plasma or an artificial perfusate of Eagle's medium and albumin. 2. Dialysable inhibitors of colony formation were also released during perfusions. 3. Colony-stimulating activity in artificial perfusate could be enhanced by the addition of rat plasma in vitro. Concentrations of cycloheximide that inhibited albumin synthesis by the liver did not inhibit the release of colony-stimulating activity.