Large conductance Ca²⁺ and voltage-activated potassium (BK) channels regulate cell excitability and their gating can be modulated by the auxiliary b, γ and LINGO subunits^1,2. BK:gamma1 channels activate at resting potentials in 0 Ca²⁺ and this shift in activation V_1/2 (V_1/2ACT) was reduced by mutation of F273 in the g1 transmembrane domain³. Interestingly, the novel regulatory BK subunit LINGO2 also caused a negative shift in V_1/2ACT when co-expressed with BK channels and induced rapid inactivation⁴, rather like the effects of LINGO1². Given that gamma and LINGO proteins appear similar in structure⁵, we performed an alanine scan of phenylalanine residues in the transmembrane region of LINGO1 and LINGO2 and co-transfected these with BK cDNA, to ascertain their effect on BK channel activity.

         HEK cells were transfected with BK:LINGO:eGFP cDNA (100:500:150ng ratio) and currents were studied under voltage clamp. Experiments were carried out using the inside-out configuration of the patch clamp technique with equimolar 140 mM K⁺ solutions, at 37℃. The cytosolic surface of the patch was exposed to the same solutions and the [Ca²⁺]_i  ranged from 100nM-10µM Ca²⁺. Deletion of the last four residues of LINGO2 (BK:LINGO2_DMKMI)  abolished inactivation and significantly shifted V_1/2ACT more negatively to 104±1 mV (n=6) in 100nM Ca²⁺ compared to 130±2 mV (n=8) in full length BK:LINGO2 (p<0.05, paired t-test). Two TM mutants, LINGO2_F558A and LINGO2_F560A shifted V_1/2ACT positively to 162±2 and 160±2 mV respectively_, in 100 nM Ca²⁺, compared to BK:LINGO2 (n=5, p<0.05, ANOVA) and these values were practically identical to the V_1/2ACT recorded in BK alone. In five experiments with BK:LINGO2_F550A the mean V_1/2ACT was 152±2 (n=5) but this shift failed to reach statistical significance compared to BK:LINGO2. Similarly, the V_1/2ACT  obtained with the BK:LINGO2_F564A mutant 134±4 mV, (n=7, p>0.05) was indistinguishable from BK:LINGO2. Although the V_1/2ACT of BK:LINGO2_F552A mutant channels (139±1 mV, n=6) was not significantly different to BK:LINGO2, currents recorded with this mutant failed to show inactivation at any voltage in 100nM, 1µM or 10µM Ca²⁺. Co-expression of BK:LINGO1 produced rapidly inactivating currents under voltage clamp and these had a V_1/2ACT  of 113±3mV (n=6),  in agreement with previous studies². Mutation of the equivalent residue to LINGO2_F552 in LINGO1 (F568A) was next performed, to establish if it also failed to inactivate when co-transfected with BK cDNA and recorded under the same conditions. In 5 experiments the V_1/2ACT of the BK:LINGO1_F568A mutant was shifted significantly positive to 140±2 mV compared to BK:LINGO1 (p<0.05)  and inactivation was completely abolished.

In conclusion, the data suggest that LINGO2_F550, LINGO2_F552, LINGO2_F564 do not contribute significantly to the shift in BK V_1/2ACT induced by LINGO2, whereas LINGO2_F558 and LINGO2_F560 contributed equally to the negative shift. However, mutations of the residue equivalent to gamma1_F273 in either LINGO1 (F568) or LINGO2 (F552) abolished inactivation of BK channels at all voltages and [Ca²⁺]_i tested. These data suggest that this conserved residue in LINGO plays a critical role in allowing the LINGO proteins to associate with BK channels or that its mutation modified the structure or alignment of the LINGO tail to prevent inactivation.