Lithium-induced changes in electrolyte balance and tissue electrolyte concentration

Dysregulation of renal aquaporins and epithelial sodium channel in lithium-induced nephrogenic diabetes insipidus

Lithium is used commonly to treat bipolar mood disorders. In addition to its primary therapeutic effects in the central nervous system lithium has a number of side effects in the kidney. The side effects include nephrogenic diabetes insipidus with polyuria, mild sodium wasting, and changes in acid/base balance. These functional changes are associated with marked structural changes in collecting duct cell composition and morphology, likely contributing to the functional changes. Over the past few years, investigations of lithium-induced renal changes have provided novel insight into the molecular mechanisms that are responsible for the disturbances in water, sodium, and acid/base metabolism. This includes dysregulation of renal aquaporins, epithelial sodium channel, and acid/base transporters. This review focuses on these issues with the aim to present this in context with clinically relevant features.

Lithium-induced NDI in rats is associated with loss of α-ENaC regulation by aldosterone in CCD

Lithium-induced nephrogenic diabetes insipidus (Li-NDI) is associated with increased urinary sodium excretion and decreased responsiveness to aldosterone and vasopressin. Dysregulation of the epithelial sodium channel (ENaC) is thought to play an important role in renal sodium wasting. The effect of 7-day aldosterone and spironolactone treatment on regulation of ENaC in rat kidney cortex was investigated in rats with 3 wk of Li-NDI. Aldosterone treatment of rats with Li-NDI decreased fractional excretion of sodium (0.83 ± 0.02), whereas spironolactone did not change fractional excretion of sodium (1.10 ± 0.11) compared with rats treated with lithium alone (1.11 ± 0.05). Plasma lithium concentration was decreased by aldosterone (0.31 ± 0.03 mmol/l) but unchanged with spironolactone (0.84 ± 0.18 mmol/l) compared with rats treated with lithium alone (0.54 ± 0.04 mmol/l). Immunoblotting showed increased protein expression of α-ENaC, the 70-kDa form of γ-ENaC, and the Na-Cl cotransporter (NCC) in kidney cortex in aldosterone-treated rats, whereas spironolactone decreased α-ENaC and NCC compared with control rats treated with lithium alone. Immunohistochemistry confirmed increased expression of α-ENaC in the late distal convoluted tubule and connecting tubule and also revealed increased apical targeting of all three ENaC subunits (α, β, and γ) in aldosterone-treated rats compared with rats treated with lithium alone. Aldosterone did not, however, affect α-ENaC expression in the cortical collecting duct (CCD), which showed weak and dispersed labeling similar to that in rats treated with lithium alone. Spironolactone did not affect ENaC targeting compared with rats treated with lithium alone. This study shows a segment specific lack of aldosterone-mediated α-ENaC regulation in the CCD affecting both α-ENaC protein expression and trafficking, which may explain the increased sodium wasting associated with chronic lithium treatment.

Lithium and inositol: effects on brain water homeostasis in the rat

RATIONALE

Since its earliest use in psychiatry, lithium has been known to alter body water homeostasis. Although lithium is also known to decrease the concentration of inositol, an important brain osmolyte, little is known of the effects of lithium on brain water homeostasis.

OBJECTIVE

To determine whether lithium alters brain water homeostasis, and, if so, whether the mechanism involves changes in inositol concentration.

MATERIALS AND METHODS

Rats were fed regular food or regular food plus lithium chloride for either 11 days or 5 weeks. Brains were dissected and assayed for tissue water by the wet-dry method and for inositol by gas chromatography-mass spectrometry.

RESULTS

We found a statistically significant (p=0.05, corrected) 3.1% mean elevation in frontal cortex tissue water in 5-week lithium-fed rats (86.7+/-3.9%), compared to control rats (83.6+/-2.6%). Inositol concentration correlated inversely with percent tissue water (r=-0.50, p=0.003, corrected) in pooled samples of 5-week lithium-fed rats, and was significantly lower in frontal cortex and hippocampus of 5-week lithium-fed rats, compared to controls. Rats fed lithium for 11 days did not differ significantly from controls on either variable.

CONCLUSIONS

This is the first report of a lithium-induced increase in brain tissue water. Although the mechanism is unclear, it does not appear to result from changes in brain inositol concentration or blood sodium concentration. This finding may have implications for the therapeutic or toxic effects of lithium on brain, because increased tissue water can augment cell excitability.

Effect of chronic lithium chloride on membrane adenosine triphosphatases in certain postural muscles of rats

Lithium has been extensively used as an antidepressant in the treatment of manic depressive disorders requiring chronic administration. Here, we report a study of the effect of long-term lithium treatment on the activities of membrane adenosine triphosphatases (ATPases) in certain postural muscles of rat. Specifically, Ca(2+)-ATPase, Na+,K(+)-ATPase and Mg(2+)-ATPase activities were measured in the soleus, extensor digitorum longus and plantaris muscles following 6 weeks of treatment with LiCl. Increases were observed in the Na+,K(+)-ATPase activity whereas the Mg(2+)-ATPase activity decreased with prolonged LiCl treatment. The most pronounced effect was a highly significant (P < 0.001) increase in the mitochondrial Ca(2+)-ATPase and Na+,K(+)-ATPase activity to almost 50-100% above the control. The increases in the mitochondrial Ca(2+)-ATPase activity of extensor digitorum longus and plantaris were 70% and 100%, respectively. The corresponding increases in the Na+,K(+)-ATPase activity were 127%, 99% and 87% for soleus, extensor digitorum longus and plantaris, respectively. Irrespective of the differences in the fiber pattern and physiological function, all three muscles responded in a similar way to Li+. The changes in the membrane ATPases reflect a deranged ATP turnover, thus affecting the overall energy state of the animal. Based on these results, we hypothesize that Li+ produces its effects by interfering with cation transport processes. Since Li+ affects the neural excitability of the cell it is suggested that the stimulation of the ATPases may be important in the psychotropic properties of the ion.

Chronic progressive renal lesions induced by lithium

New Zealand white rabbits, eight fed lithium (Li) (50 to 250 mmole LiCl/kg food) and seven controls (C) had sequential open renal biopsies at zero, one, three, six, and 12 months. A distinctive histological lesion, consisting of cytoplasmic vacuolation and accumulation of glycogen in cells lining distal convoluted tubules and collecting ducts, was present in Li (one, three, six, and 12 months), but was absent in Li prior to lithium (zero months) and in C (zero, one, three, six, and 12 months). Histological changes of chronic focal interstitial nephropathy namely, interstitial fibrosis, (quantitated by point counting), tubular atrophy and cast formation (quantitated by digitization), and glomerular sclerosis (determined as the percent of sclerosed glomeruli) showed significant differences between Li and C from as early as one month (interstitial fibrosis, P less than 0.02; tubular atrophy, P less than 0.05; casts, P less than 0.05), and up to 12 months (glomerular sclerosis, P less than 0.05). Distal tubular dilatation and microcyst formation, (quantitated by digitization) was also marked in Li compared with C from one month (P less than 0.05). The degree of distal tubular dilatation and other changes of interstitial nephropathy tended to progress with duration of lithium exposure. Macroscopically, Li kidneys (12 months) were pale, granular, and exhibited microcysts. Raised blood urea (P less than 0.02) and serum creatinine levels (P less than 0.05) were also late features (12 months) of lithium-induced nephropathy. The data support the view that lithium induces chronic renal lesions. The precise relationship between the distinctive distal tubular lesion, distal tubular dilatation and focal interstitial nephropathy remains speculative.

Strain dependent rate of Li+ elimination associated with toxic effects of lethal doses of lithium chloride in mice

Strain differences in response to the administration of two lethal doses (700 and 900; mg/kg) of lithium chloride were studied in eight week old males from six genetic strains of mice. Two parameters were considered; (a) toxicity (time to death) and (b) hypothermia. Lithium distribution in the body (blood, seven tissues, excreta and urine) were evaluated for each strain following IP injection of 200 mg/kg dose of LiCl. The strain differences were significant for toxicity. The order of susceptibility of the strains was 129/ReJ greater than S.W. greater than C3H/S greater than DBA/2 = Balb/c greater than C57/6J with a 15-fold difference between the most susceptible and the least susceptible strain at the 900 mg/kg dose. Strain differences for hypothermic response at both doses were not significant. Significant strain differences were also observed for lithium distribution in different parts of the body, excreta and urine. The concentration of Li+ found in urine and excreta was positively correlated with resistance (time to death at 900 mg/kg LiCl) to the toxic effect of lithium. The lithium concentration in blood, muscle and lung on the other hand reflected a negative correlation with toxicity. The susceptibility of a strain could be characterized by its inherent lithium excretory ability, particularly through urine. It may suggest an involvement of membrane transport mechanisms in determining toxicity to lithium compounds.

A comparison between the natriuretic response to potassium in lithium- and amiloride-treated rats

The effect of an increase of the potassium content of the food on the natriuretic and diuretic effects of lithium and amiloride was examined in Wistar rats pretreated with the respective drugs for a period of three weeks. Lithium (60 mmol/kg) or amiloride (200 mg/kg) was added to the food and the animals had access to water and a 0.46 M NaCl solution. An increase of the potassium content of the food resulted in a potassium dose-dependent increase of the natriuretic effect in both lithium-treated and amiloride-treated rats. The effect was temporary in lithium-treated rats but persistent in amiloride-treated rats. Lithium treatment resulted in a 10-fold increase of the water intake compared to control rats and amiloride-treated rats. An increase of the potassium intake reduced the water intake in the lithium-treated animals and modestly increased the water intake in amiloride-treated rats. The results show that the exaggerated natriuretic effect produced by administration of potassium to rats treated with lithium is not a specific phenomenon. It is suggested that a lowered ability of the kidneys to secrete potassium may be responsible for the exaggerated natriuretic response to potassium in amiloride- and lithium-treated rats.

Effect of lithium on prolactin responses to thyrotropin releasing hormone in patients with manic state

The plasma thyrotropin (TSH) and prolactin (PRL) responses to thyrotropin releasing hormone (TRH) were studied before and during lithium treatment for 3-4 weeks in 6 patients with manic states and 8 control subjects. The plasma TSH responses to TRH were not different between the two groups before lithium treatment. Lithium administration did not alter non-stimulated secretion of TSH in any groups, but resulted in exaggerated responses of plasma TSH to TRH in both groups. No difference between two groups was observed in plasma TSH responses to TRH. The basal plasma PRL concentration did not differ between the two groups and was not affected by lithium administration to either group. The plasma PRL responses to TRH in female subjects were greater than those in male subjects. In females, the plasma PRL responses to TRH in manic patients were significantly higher than those in control subjects before the treatment. Lithium administration caused enhanced responses to TRH in patients when compared to pretreatment levels, but not in control subjects. Although the small number of male subjects limits conclusions, pretreatment plasma PRL responses to TRH in male manic patients were apparently greater than those in control subjects. However, lithium administration appeared not to affect the responses of plasma PRL to TRH in manic patients when compared to pretreatment levels. The augmented responses of plasma PRL to TRH in patients with manic states suggest the existence of some abnormality in the hypothalamo-pituitary axis. Thus, the effect of the anti-manic agent on PRL secretion in manic patients may suggest the mechanism by which the drug affects manic symptoms.

Relative effects of hyperbaric oxygen on cations and catecholamine metabolism in rats: protection by lithium against seizures

Analysis of lithium (Li+) in the brain and blood after intraperitoneal injection (i.p.) shows that initially its concentration is high in blood and negligible in the brain. Subsequently its concentration increases in the brain and disappears from the blood. Lithium itself affects neurological actions but the mechanisms remain obscure. It also modifies the toxic action of oxygen at high pressure (OHP), which causes convulsions, either suppressing or exacerbating it. These clearly separate effects correspond with the presence of Li+ in the blood (suppression) or in the brain (exacerbating). Determination of the effect of Li+ and OHP upon cations, catecholamines, ammonia, tyrosine hydroxylase, and monamine oxidase on brain and blood tissue showed that there was very little correspondence between changes in the cations either with Li+ or the toxic effects of OHP. On the other hand, OHP developed a sustained blood and brain hyperammonemia in rats which could be negatively modified by Li+ in the blood. The latter effect also corresponded with a prolongation of convulsive latency. Changes in brain catecholamines, tyrosine hydroxylase, monoamine oxidase and tyrosine were effected by Li+ and potentiated by OHP. These data suggest that Li+ and OHP mediate their effects relatively more through developing hyperammoneic states in both blood and brain than by altering cation concentrations in these tissues.

Combined effects of alcohol and lithium on body composition in the rat model

Alcohol markedly enhances Li+ retention in blood and muscle in rats taking Li and alcohol. The marked enlargement of the exchangeable Na+ space which occurs with alcohol alone, Li+ alone, and to an even greater extent with combined therapy, results from extensive mobilization of "bound" bone sodium, which is transferred to intracellular sites, demonstrably in skeletal muscle, and presumably in other cells. This 87% increase in measured intracellular Na+, and concomitant decrease in K+, would be predicted to produce profound effects on the intracellular ionic milieu, and on membrane function. Lithium is associated with a decrease in food and fluid intake, and this effect is enhanced by alcohol. However the decrease in intake did not result in nutritionally significant deficits in either group, and therefore the observed abnormalities in body composition did not result from caloric, protein, K, Na, or Ca deficiency.

The effect of lithium on electrolyte transport by the in situ choroid plexus of the cat

1. The effects of lithium on electrolyte transport were studied by using the cat choroid plexus isolated in a chamber in situ. 2. Lithium infused intravenously to produce plasma lithium concentrations up to 5 m-equiv/l. caused an increase in plasma magnesium with no effect on the concentration of magnesium in the chamber fluid. 3. When 22NaCl was infused intravenously the chamber fluid/plasma ratio of 22Na was nearly 1 in the first 30 min sample and at the steady state it was significantly greater than 1. 4. When lithium chloride (1.5 m-equiv/l.) or potassium chloride (6.6 m-equiv/l.) was added to the chamber at the start of a collection period with plasma 22Na in the steady state, the 22Na content of the chamber fluid promptly increased 118 and 68%, respectively, above the control value with no increase in secretory rate. 5. The addition of ouabain to the chamber fluid, in addition to the lithium chloride or potassium chloride, tended to stimulate or have no significant effect on 22Na uptake at a concentration of 10(-5) M and to reduce it as well as the secretory rate at 10(-3) M. 6. The date are compatible with there being two functionally separate sodium transport systems in the choroid plexus. One transports sodium accompanied by an anion and water to provide the fluid secreted into the chamber (c.s.f.) and the other operates primarily to regulate the potassium concentration of the c.s.f. by pumping potassium out in exchange for sodium. 7. Lithium can be transported by both systems to a limited extent and the presence of lithium in the c.s.f. stimulates the sodium-potassium regulating pump.

Potentiation of lithium toxicity by ethanol in rats and mice

The effects of ethanol treatments, both acute (single, intraperitoneal) and chronic (forced drinking of ethanol for 10 mo), on the distribution, excretion of lithium and urine output were studied in rats. Retention of lithium induced by ethanol appeared to be responsible for the potentiation of lithium toxicity. The potential hazard in the interaction between lithium and ethanol is discussed.

Time course of lithium-induced alterations in renal and endocrine function in normal and Brattleboro rats with hypothalamic diabetes insipidus

1. A lithium chloride (1.1 g/kg) supplemented diet was given to Long Evans (LE) and Brattleboro (DI) rats to investigate its actions in the presence (LE) and absence (DI) of vasopressin. 2. During the first 24 h, Li-supplemented LE rats displayed an initial water deficit (drinking less than renal output), increased plasma antidiuretic (ADH) titres and slightly increased plasma renin activities (PRA) and plasma osmolarities. Such changes were qualitatively similar to those seen in rats fed a normal diet, but deprived of water for 24 hours. After 12 days, the Li-supplemented rats had elevated plasma ADH titres, but reduced pituitary oxytocic and antidiuretic activities. 3. The urinary losses of Na, K and Cl exceeded dietary intakes in LE rats on the introduction of the Li-supplement, and the urinary osmolarity fell by 50%. Electrolyte balances were gradually re-established, although drinking and urine production increased in parallel to reach twice the control values by day 12 of the supplement. 4. Aldosterone and corticosterone secretory rates and their peripheral plasma concentrations were unchanged both after 24 h and 28 days of the Li-supplement. 5. Li elicited no water deficit or saluresis in DI rats, and although the polyuria and polydipsia were exacerbated, urinary osmolarity did not change over the 12 day observation period. 6. Li increased Ca excretion in both rat types; after 12 days the PRA of DI but not LE animals were increased. 7. It is concluded that the overall renal actions of Li are tempered by vasopressin rather than adrenocorticosteroids.

Lithium transport by the colon of normal and sodium-depleted rats

1. The transport of Li by colonic epithelium has been examined in normal and Na-depleted rats. 2. Substitution of Li for Na with lumen of the conon causes the transepithelial electrical potential difference (p.d.) and short-circuit current to fall to low levels and the electrical resistance of fall moderately. Recovery occurs by fairly slowly after removal of Li. 3. Li absorption increases linearly with increasing concentration in the lumen and is significantly faster in Na-depleted rats. Increasing the luminal Na concentration reduces Li absorption from solutions of low Li concentration. 4. Comparison of absorption rates with secretion rates in rats given Li systemically, together with measurements of Li distribution across the epithelium in relationship to the transepithelial p.d. indicate that Li transport is predominatly or entirely passive. Interference with Li absorption by Na suggests, however a mucosal membrane carrier which, since Li absorption rises after Na depletion, may be increased in the Na-depleted state.

Plasma renin activity in depressed patients treated with increasing doses of lithium carbonate

Plasma renin activity (PRA) was measured in the supine position and after active upright stance in patients with endogenous depression and in a group of healthy volunteers serving as controls. In the depressed patients, PRA was further investigated in the same conditions during treatment with increasing doses of lithium carbonate. Basal PRA values were lower in depressed patients than in normal controls, particularly in the upright stance, and tended to rise gradually during lithium therapy. These findings suggest that lithium may work as a stimulant of the renin-angiotensin system, and possibly as an antidepressant, by way of producing functional activation of the norepinephrine system independent of its action on the water and electrolyte balance.

The distribution of lithium, sodium and magnesium in rat brain and plasma after various periods of administration of lithium in the diet

1 The tissue solubilizer Soluene-100 provides an efficient and easy means of preparing small amounts of rat tissue for cation analysis. 2 Administration of lithium ions to rats for two days to 42 days by the addition of lithium chloride to the diet at a concentration of 30 mmol/kg dry weight results in (a) the uniform distribution of lithium throughout the brain at a concentration comparable to that found in plasma; (b) decrease in the brain sodium concentration: (c) a decrease in brain magnesium concentration and an increase in plasma magnesium concentration; (d)no change in brain water content. 3 The inclusion of LiCl in the diet at a concentration of 30 mmol/kg dry food gives consistent and predictable plasma and brain levels of lithium in the rat without the occurrence of serious side effects over periods of up to 42 days.

The distribution of lithium and its effects on the distribution and excretion of other ions in the rat

1. In rats, lithium (ca 1 mEquiv/kg body weight) decreased brain sodium and magnesium, bone sodium and calcium and increased muscle calcium, plasma magnesium, urinary calcium and urine volume.2. Lithium was particularly concentrated in bone.