Gastrointestinal smooth muscles exhibit the muscarinic cationic current (mI_Cat). Studies in guinea-pig ileum have suggested that both M₂ and M₃ subtypes of muscarinic receptor participate in the activation of mI_Cat through pertussis toxin (PTX)-sensitive G protein G_o and some isozyme of phopholipase C (PLC) other than regulated by G_q/G₁₁ proteins (Zholos & Bolton, 1997; Komori et al. 1998; Yan et al. 2003; Zholos et al. 2004; Okamoto et al. 2004). To provide some insights into the activation mechanism of mI_Cat, we have studied the ileal smooth muscle mI_Cat in M₂ or M₃ subtype knockout (KO) or M₂/M₃-double KO mouse as well as wild type (WT). The mI_Cat was recorded from single ileal longitudinal myocytes using whole-cell patch clamp techniques, where the cells were bathed in a CsCl-based medium and dialysed intracellularly with another CsCl-based one containing a BAPTA/CaCl buffer (calculated Ca²⁺ = 100 nM), unless otherwise stated. All mice were humanely killed. In WT-derived cells held under voltage-clamp at −50 mV, cumulative applications of nonselective muscarinic agonist carbachol (CCh; 1 to 300 μM) produced mI_Cat in a concentration-dependent manner. An M₂-preferring antagonist (methoctramine; 300 nM) caused a rightward parallel shift of the CCh concentration-response curve with an increased EC₅₀ value which was consistent with M₂ mediation of mI_Cat, while an M₃-preferring antagonist (4-DAMP; 30 nM) severely depressed the E_max without a change in the EC₅₀ value. Injection of PTX to WT animals (100 μg/kg, i.p.) 70-74 hrs before experiments caused marked reduction of mI_Cat. A bath application of a PLC inhibitor (U73122; 1 μM), but not an inactive analogue (U73343; 1 μM), almost abolished mI_Cat. Current-voltage (I-V) relation for mI_Cat exhibited a U-shaped curve in a voltage range of 0 and −120 mV. All the above features of mI_Cat resembled those described for mI_Cat in guinea-pig ileal myocytes. In cells from the M₂-KO and M₃-KO types, only a small mI_Cat was evoked even by maximally effective CCh, of which amplitude at 100 μM was estimated to be 19.3 ± 3.7 pA (n=9) for the former type and 11.5 ± 1.5 pA (n=15) for the latter, much smaller than that in WT cells (195.0 ± 28.3 pA, n=13). No current was evoked in the M₂/M₃-double KO. Infusion of GTPγS (200 μM) via patch pipettes induced an mI_Cat-like inward current in all four types, and there was no noticeable difference in the current amplitude among them (150-260 pA). CCh′s effects in producing Ca²⁺-activated K⁺-current (I_K-Ca) via the M₃/G_q/PLC system and in reducing cAMP levels via the M₂/G_i/o/adenylate cyclase system were evaluated under appropriate experimental conditions. CCh (100 μM) evoked I_K-Ca in M₂-KO cells with a similar amplitude to that seen in the WT, but was without effect in the M₃-KO. CCh (1 μM) inhibited the isoprenaline-stimulated cAMP accumulation by 50 % in M₃-KO type and 30 % in WT, but rather enhanced by 30 % in the M₂-KO. Indicated from these results was that even with either M₂ or M₃ subtype lacking, either of the two muscarinic signaling systems is fully operative, in contrast to the mI_Cat-inducing system for whose full activation both M₂ and M₃ are indispensable. By non-stationary noise analysis of whole-cell current it was found that the unitary channel conductances underlying the mI_Cat (at 100 μM CCh) in M₂-KO and M₃-KO cells were 10 pS and 2 pS, respectively. These values clearly differed from the corresponding value (40 pS) estimated for WT mI_Cat. An inward current carried via 10 pS channels was evoked by a diacylglycerol (DAG) analogue (OAG). The I-V curve for mI_Cat in M₂-KO or M₃-KO cells was relatively linear compared with that for the WT mI_Cat. To see if mI_Cat is sensitive to a rise in cytosolic Ca²⁺, some experiments were carried out where the BAPTA/CaCl buffer was absent inside the cell. There was seen a significant potentiation of mI_Cat upon voltage-gated Ca²⁺ entry in WT cells, but not in M₂-KO or M₃-KO cells. Our data suggest that the ileal longitudinal myocytes exhibit three types of mI_Cat. The two of them arise because of the respective openings of 2-pS channels linked to M₂ subtype and of 10-pS channels linked to M₃ subtype via DAG-dependent pathways, while the other arises due to 40-pS channel opening for which both subtypes are absolutely needed and prominently contributes to CCh-evoked mI_Cat (in our experimental conditions). There may be a possibility that functional supermolecular units formed by M₂/M₃ hetero-complexes together with G_o protein, some PLC isozyme and 40-pS cationic channels may exist to control the channel opening. Instead, the functional units might involve 2-pS channels, which are transited to 40-pS channels and opened upon activation of the M₂/M₃ hetero-complexes.