Evidence that phosphoinositide metabolism in rat cerebral cortex stimulated by pilocarpine, physostigmine, and pargyline in vivo is not changed by chronic lithium treatment.

The effect of chronic versus acute administration of lithium on receptor-linked phosphoinositide metabolism was assessed by comparing the change in the cerebral cortex levels of myo-inositol 1-phosphate in response to pilocarpine, physostigmine, or pargyline in rats. Rats were exposed to either 29 consecutive days of LiCl injections or 27 and 39 days of dietary Li2CO3, followed by injected LiCl at the end of the diet to insure a constant level of exposure to the drug. In each experiment, an acute group received a single injection of LiCl 20-24 h before they were killed. One hour before being killed, some of the animals acutely exposed to lithium and some of the animals chronically exposed to lithium each received pilocarpine, physostigmine, or pargyline. At the conclusion of the experiment, the rats were killed and brain levels of myo-inositol 1-phosphate and lithium were determined. A differential production of myo-inositol 1-phosphate in groups receiving acute versus chronic lithium would provide evidence of a change in receptor-linked phosphoinositide metabolism due to the chronic administration of lithium. Brain levels of myo-inositol 1-phosphate are dependent on tissue lithium concentrations; consequently, significant differences observed in brain lithium levels between the groups receiving acute versus chronic lithium prevented a meaningful assessment of the effect of the mode of lithium administration on the production of myo-inositol 1-phosphate in those groups. Stepwise multiple regression analysis and the measured brain lithium levels were used to assess the response of myo-inositol 1-phosphate levels to stimulation in animals receiving acute or chronic lithium treatment.(ABSTRACT TRUNCATED AT 250 WORDS)

Chronically administered lithium alters neither myo-inositol monophosphatase activity nor phosphoinositide levels in rat brain.

Rats were exposed to either 29 consecutive days of LiCl injections or 27 and 39 days of dietary Li2CO3, followed by injected LiCl at the end of the diet to insure a constant level of exposure to the drug. At the end of the period of chronic exposure to lithium, the rats were sacrificed and brain myo-inositol-1-phosphate phosphohydrolase (myo-inositol monophosphatase) activity was measured. In none of the experiments was there any difference in the lithium-sensitive activity toward myo-inositol monophosphatase when comparing the control and chronic groups. These brains and those from another group of rats that had been given Li2CO3 in their diet for 41 days, followed by 7 additional days of LiCl injections, were also examined for changes in the levels of the phosphoinositides. No reproducible differences in the absolute tissue levels of those lipids were found when control and chronic lithium groups were compared. These results are contrary to published reports which suggest that myo-inositol monophosphatase activity increases and that the phosphatidylinositol level decreases in rat brain as a result of chronic administration of lithium.

Evidence that inositol 1-phosphate in brain of lithium-treated rats results mainly from phosphatidylinositol metabolism.

In cerebral cortex of rats treated with increasing doses of LiCl, the relative concentrations of Ins(1)P, Ins(4)P and Ins(5)P (when InsP is a myo-inositol phosphate) are approx. 10:1:0.2 at all doses. In rats treated with LiCl followed by increasing doses of pilocarpine a similar relationship occurs. myo-Inositol-1-phosphatase (InsP1ase) from bovine brain hydrolyses Ins(1)P, Ins(4)P and Ins(5)P at comparable rates, and these substrates have similar Km values. The hydrolysis of Ins(4)P is inhibited by Li+ to a greater degree than is hydrolysis of Ins(1)P and Ins(5)P. D-Ins(1,4,5)P3 and D-Ins(1,4)P2 are neither substrates nor inhibitors of InsP1ase. A dialysed high-speed supernatant of rat brain showed a greater rate of hydrolysis of Ins(1)P than of D-Ins(1,4)P2 and a lower sensitivity of the bisphosphate hydrolysis to LiCl, as compared with the monophosphate. That enzyme preparation produced Ins(4)P at a greater rate than Ins(1)P when D-Ins(1,4)P2 was the substrate. The amount of D-Ins(3)P [i.e. L-Ins(1)P, possibly from D-Ins(1,3,4)P3] is only 11% of that of D-Ins(1)P on stimulation with pilocarpine in the presence of Li+. DL-Ins(1,4)P2 was hydrolysed by InsP1ase to the extent of about 50%; both Ins(4)P and Ins(1)P are products, the former being produced more rapidly than the latter; apparently L-Ins(1,4)P2 is a substrate for InsP1ase. Li+, but not Ins(2)P, inhibited the hydrolysis of L-Ins(1,4)P2. The following were neither substrates nor inhibitors of InsP1ase; Ins(1,6)P2, Ins(1,2)P2, Ins(1,2,5,6)P4, Ins(1,2,4,5,6)P5, Ins(1,3,4,5,6)P5 and phytic acid. myo-Inositol 1,2-cyclic phosphate was neither substrate nor inhibitor of InsP1ase. We conclude that the 10-fold greater tissue contents of Ins(1)P relative to Ins(4)P in both stimulated and non-stimulated rat brain in vivo are the consequence of a much larger amount of PtdIns metabolism than polyphosphoinositide metabolism under these conditions.

Effects of lithium on phosphoinositide metabolism in vivo.

All of the known pathways for metabolizing the phospholipase C (EC 3.1.4.10) products of phosphoinositide metabolism eventually lead to myo-inositol monophosphates and products that are hydrolyzed by myo-inositol 1-phosphatase (EC 3.1.3.25). That enzyme is inhibited by lithium (Ki about 1 mM). In animals treated with LiCl, elevations of myo-inositol 1-phosphate (1-IP) occur in brain that appear to result from endogenous neural activity for they are diminished by the anesthetics halothane and pentobarbital. Lithium is thus a useful tool for assessing endogenous in vivo cerebral phosphoinositide metabolism. The 1-IP elevation is also useful for revealing in vivo central nervous system (CNS) receptor activity that is stimulated by endogenous or exogenous processes such as the effects of centrally acting drugs and of seizures. Stimulation of the CNS in the presence of lithium causes myo-inositol to be sequestered in 1-IP in proportion to the amount of stimulation. Thus if the inositol level falls sufficiently resynthesis of the phosphoinositides may be compromised and receptor response to stimuli may be reduced. Evidence for such an occurrence would support the theory that this is one mechanism by which lithium acts in the therapy of manic illness. We extended our efforts to identify such a lowering of phosphoinositide levels to mice where cerebral metabolism can be halted more rapidly than in rats. However, the only change detected was a small elevation in phosphatidylinositol 4-phosphate. We were successful, however, in causing all of the phosphoinositides to be reduced in rat cerebral cortex by pilocarpine stimulation after lithium treatment, a procedure that causes seizures. The same procedure causes the largest reduction in cortical myo-inositol levels that we have observed, and thus may represent the point where the inositol decrement is sufficient to interfere with resynthesis of the lipids.

Effects of systemically administered lithium on phosphoinositide metabolism in rat brain, kidney, and testis.

A single subcutaneous dose of 10 mEq/kg LiCl gives rise to an increase in the cerebral cortex level of myo-inositol-1-P (I1P) that closely follows cortical lithium levels and, at maximum, is 40-fold above the control value. Kidney and testis show smaller increases in I1P level following LiCl administration. The I1P level is still sixfold greater than that of untreated rat cortex 72 h later. In cortex, parallel increases also occur in myo-inositol-4-P (I4P) and myo-inositol 1,2-cyclic-P (cI1,2P), whereas myo-inositol-5-P (I5P) remains unchanged. The cortical increases in I1P and I4P levels are partially reversed by administering 150 mg/kg of atropine 22 h after the LiCl, treatment that does not affect cI1,2P. When doses of LiCl from 2 to 17 mEq/kg are given, the cerebral cortex levels of I1P and myo-inositol, measured 24 h later, are found to reach a plateau at about 9 mEq/kg of LiCl, whereas cortical lithium levels continued to increase with greater LiCl doses. Levels of all three of the brain phosphoinositides are unchanged by the 10 mEq/kg LiCl dose, as is the uptake of 32Pi into these lipids. Chronic dietary administration of LiCl for 22 days showed that the effects of lithium on I1P and myo-inositol levels persist for that period. Over the course of the chronic administration of the lithium, levels of I1P, myo-inositol, and of lithium in cortex remained significantly correlated. We believe that these increases in inositol phosphates result from endogenous phosphoinositide metabolism in cerebral cortex and that lithium is capable of modulating that metabolism by reducing cellular myo-inositol levels. The size of the effect is a function of both lithium dose and the degree of stimulation of receptor-linked phosphoinositide metabolism. This property of lithium may explain part of its ability to moderate the symptoms of mania. Our chronic study suggests that prolonged administration of LiCl does not result in compensatory changes in myo-inositol-1-P synthase or myo-inositol-1-phosphatase.

Systemic cholinergic agents induce seizures and brain damage in lithium-treated rats.

Administration of pilocarpine or physostigmine to rats treated with lithium chloride produced sustained limbic seizures, widespread brain damage, and increased concentrations of D-myo-inositol-1-phosphate (a metabolite of the phosphoinositides, lipids involved in membrane receptor function) in the brain. The syndrome was preventable with atropine. The physostigmine doses and concentrations of blood lithium that caused the syndrome are similar to those considered appropriate for psychiatric chemotherapy.

In vivo phosphorylation of neurofilament proteins in the central nervous system of immature rat and rabbit.

Neurofilament proteins from brain and spinal cord of immature rat (20-35 days of age) and rabbit (15-17) days of age) were prepared by an axonal flotation technique. Examination of rat filament preparations by electron microscopy revealed a preponderance of 10 nm diameter filaments that were usually loosely aggregated although some bundles of more tightly packed filaments were present as well. The neurofilament triplet proteins of the rat and rabbit central nervous system were found to be phosphorylated 24 hr after the intracerebral injection of [32P]orthophosphate when examined by sodium dodecyl sulfate polyacrylamide gel electrophoresis followed by fluorography. Examination of each eluted neurofilament protein from both species showed that [32P]phosphate was retained after reelectrophoresis and fluorography. Evidence is presented that the [32P]phosphate is covalently linked to the purified neurofilament proteins by phosphoester bonds.

Evidence that lithium alters phosphoinositide metabolism: chronic administration elevates primarily D-myo-inositol-1-phosphate in cerebral cortex of the rat.

The administration of LiCl (3.6 mequiv./kg/day) to adult male rats for 9 days results in an increase in the cerebral cortex level of myo-inositol-1-phosphate (M1P) to 4.43 +/- 0.52 mmol/kg (dry weight) compared with a control level of 0.24 +/- 0.02 mmol/kg. This establishes that the previously observed acute effect of lithium on M1P (Allison et al., 1976) is both prolonged and augmented by repeated doses of lithium. Larger doses of LiCl over a 3-5 day period result in even larger increases in M1P and a 35% decrease in myo-inositol. In each case, 90% of the increase is due to the D-enantiomer, evidence that lithium is largely producing this effect via phospholipase C-mediated phosphoinositide metabolism. Data are presented showing that lithium is an uncompetitive inhibitor of the hydrolysis of both D- and L-M1P by M1P'ase.

Evaluation of in vivo brain site perfusion with the push-pull cannula.

The effect of tissue site perfusion with push-pull cannulas on the integrity of brain tissue was evaluated during in vivo studies in conscious New Zealand rabbits. The mean rate of perfusion-induced tissue disruption was 1.09 microgram tissue wet wt/min as estimated from the perfusate levels of endogenous nervous system specific proteins (S-100 and 14-3-2). The level of intravenously injected 125I-labeled bovine serum albumin recovered in perfusate samples indicated that approximately 2.0 nl serum/min penetrated the vascular barrier into the perfused tissue site. The appearance of intraventricularly injected solutes in perfusate samples was subjected to regression analysis. It was demonstrated that nonspecific variations in the appearance of one substance in the perfusate could be controlled for by a regression on the levels of another concurrently recovered inert substance. The experiments demonstrate that tissue site perfusion can provide useful access to brain extracellular fluid when suitable controls are incorporated for recovery variation and blood-brain barrier seepage.