A polyuric syndrome with nephrogenic diabetes insipidus (NDI) is a frequent consequence of prolonged administration of lithium (Li) salts. Studies in the past, mainly the acute and in vitro experiments, indicated that Li ions can inhibit hydroosmotic effect of [8-arginine]vasopressin (AVP) at the step of cAMP generation in vitro. However, the pathogenesis of the NDI due to chronic oral administration of low therapeutic doses of Li salts is not yet clarified. We conducted a comprehensive study to clarify the mechanism by which Li administered orally for several weeks induces polyuria and NDI in rats. Albino rats consuming a diet which contained Li (60 mmol/kg) for 4 wk developed marked polyuria and polydipsia; at the end of 4 wk the plasma Li was 0.7 +/- 0.09 mM (mean +/- SEM; n = 36). Li-treated rats had a significantly decreased (-33%) tissue osmolality in papilla and greatly reduced cortico-papillary gradient of urea (cortex--43%; medulla--64%; papilla--74%). Plasma urea was significantly (P less than 0.001) lower in Li-treated rats (5.4 +/- 0.2 mM) compared with controls (6.8 +/- 0.3 mM). Medullary collecting tubules (MCT) and papillary collecting ducts (PCD) microdissected from Li-treated animals had higher content of protein than MCT and PCD from the control rats. The cAMP accumulation in response to AVP added in vitro was significantly (delta = -60%) reduced. Also, the cAMP accumulation in MCT and PCD after incubation with forskolin was markedly lower in Li-treated rats. Addition of 0.5 mM 1-methyl,3-isobutyl-xanthine did not restore the cAMP accumulation in response to AVP and forskolin in MCT from Li-treated animals. In collecting tubule segments from polyuric rats with hypothalamic diabetes insipidus (Brattleboro homozygotes) the AVP-dependent cAMP accumulation was not diminished. The activity of adenylate cyclase (AdC) in MCT of Li-treated rats, both the basal and the activity stimulated by AVP, forskolin, or fluoride, was significantly (delta approximately equal to -30%) reduced, while the activity of cAMP phosphodiesterase (cAMP-PDIE) in the same segment showed no significant difference from the controls. Also, the content of ATP in MCT microdissected from Li-treated rats and incubated in vitro did not differ from controls. The rate of [14C]succinate oxidation to 14CO2 in MAL was inhibited (-77%) by 1 mM furosemide, which indicates that this metabolic process is coupled with NaCl cotransport in MAL. The rate of (14)CO(2) production from [14C]succinate in MAL was not significantly different between control and Li-treated rats. In MCT of control rats, the rate of [14C]succinate oxidation was approximately 3 times lower than in MAL. The rate of (14)CO(2) production from [(14)C]succinate in MCT of Li-treated rats was significantly (delta +33%) higher than in MCT dissected from control rats. Based on these results, we conclude that at least two factors play an important role in the pathogenesis of NDI consequent to chronic oral administration of Li: (a) decreased ability of MCT and PCD to generate and accumulate cAMP in response to stimulation by AVP; this defect is primarily due to diminished activity of AdC in these tubular segments caused by prolonged exposure to Li; and (b) lower osmolality of renal papillary tissue, due to primarily to depletion of urea, which decreases osmotic driving force for water reabsorption in collecting tubules. On the other hand, NaCI reabsorption in MAL is apparently not affected by chronic Li treatment.