Differences in the large extracellular loop between the K(+)-Cl(-) cotransporters KCC2 and KCC4

Abstract

K(+)Cl(-) cotransporters (KCCs) play fundamental physiological roles in processes such as inhibitory neurotransmission and cell volume regulation. Mammalian genomes encode four distinct KCC paralogs, which share basic transport characteristics but differ significantly in ion affinity, pharmacology, and relative sensitivity to cell volume. Studies to identify divergence in functional characteristics have thus far focused on the cytoplasmic termini. Here, we investigated sequence requirements of the large extracellular loop (LEL) for function in KCC2 and KCC4. Mutation of all four evolutionarily conserved cysteines abolished KCC2 transport activity. This behavior differs from that of its closest relative, KCC4, which is insensitive to this mutation. Chimeras supported the differences in the LEL of the two cotransporters, because swapping wild-type LEL resulted in functional KCC2 but rendered KCC4 inactive. Insertion of the quadruple cysteine substitution mutant of the KCC4 loop, which was functional in the parental isoform, abolished transport activity in KCC2. Dose-response curves of wild-type and chimeric KCCs revealed that the LEL contributes to the different sensitivity to loop diuretics; a KCC2 chimera containing the KCC4 LEL displayed an IC(50) of 396.5 mum for furosemide, which was closer to KCC4 (548.8 mum) than to KCC2 (184.4 mum). Cell surface labeling and immunocytochemistry indicated that mutations do not affect trafficking to the plasma membrane. Taken together, our results show a dramatic and unexpected difference in the sequence requirements of the LEL between the closely related KCC2 and KCC4. Furthermore, they demonstrate that evolutionarily highly conserved amino acids can have different functions within KCC members.

FIGURE 1.

Cysteines of the large extracellular loop are highly conserved. Partial alignment of KCCs reveals high conservation of the four cysteines (marked with an asterisk) in the LEL between TMDs 5 and 6. The prediction of the TMDs is according to the hydropathy model of KCC2 (14). mm, Mus musculus; rn, Rattus norvegicus; dr, D. rerio; dm, D. melanogaster; ce, C. elegans.

FIGURE 2.

Single cysteine mutations in the LEL abolish transport activity of KCC2. HEK-293 cells were assayed for ⁸⁶Rb⁺ uptake 48 h after transfection with the constructs indicated below the columns. An empty vector was used for mock transfection. KCC2 wild type (KCC2_wt)-transfected cells displayed a 5 times higher activity compared with mock-transfected cells. All single cysteine mutants in the extracellular loop of KCC2 were transport-inactive because their transport activity showed no significant difference from the empty vector. The plot depicts the mean of three experiments ± S.D. (error bars). One-way ANOVA of all determinants demonstrated a significant multivariate effect (F = 198.67; p < 0.001). ANOVA without KCC2_wt revealed no significant difference between mutants and mock-transfected cells (F = 0.3; p = 0.94).

FIGURE 3.

Double, triple, and quadruple cysteine KCC2 mutants are transport-inactive. The ⁸⁶Rb⁺ uptake of KCC2_wt and compound mutants were measured in transfected HEK-293 cells. All double cysteine (A) and triple and quadruple cysteine (B) KCC2 mutants were transport-inactive because their transport activity showed no significant difference from the empty vector. Plots depict mean of three experiments ± S.D. (error bars). One-way ANOVA of all determinants demonstrated a significant multivariate effect for the double mutants (F = 31.1; p < 0.001) and for the triple and quadruple mutants (F = 3316.99; p < 0.001). ANOVA without KCC2_wt revealed no significant difference between the double mutants (F = 0.16; p = 0.98) or the triple and quadruple mutants (F = 1.64; p = 0.17) compared with mock-transfected cells.

FIGURE 4.

Cysteine mutations in the LEL differentially affect transport activity of KCC2 and KCC4. HEK-293 cells were assayed for ⁸⁶Rb⁺ uptake 48 h after transfection with the constructs indicated below the columns. An empty vector was used for mock transfection. A, ⁸⁶Rb⁺ uptake of the quadruple cysteine mutant KCC2_q (C287S,C302L,C322S,C331L) resulted in a decrease of the transport activity compared with KCC2_wt (p < 0.001). Treatment with 1 mm NEM did not increase the transport activity of the quadruple mutant in contrast to its effect on KCC2_wt. B, KCC4-transfected cells revealed increased transport activity in relation to mock-transfected cells. ⁸⁶Rb⁺ uptake of KCC4_wt and the quadruple mutant KCC4_q (C308S,C323L,C343S,C352L) is similar. The plot depicts the mean of three experiments ± S.D. (error bars). ***, p < 0.001; n.s., not significant.

FIGURE 5.

Chimera of KCC2 and KCC4 are expressed in HEK-293 cells. A, topology model of the chimeras KCC2_2-4-2 and KCC4_4-2-4. White circles and black circles represent amino acid residues of KCC2 and KCC4, respectively. The extracellular cysteines are marked in red. B, immuncytochemical labeling of KCC2_wt, the quadruple mutant KCC2_q, the chimera KCC2_2-4-2, and the quadruple cysteine mutant chimera KCC2_2-4q-2 upon transient transfection in HEK-293 cells. In the wild type and the mutants, KCC2 immunoreactivity was detected at the plasma membrane and the perinuclear region. C, immunoblot analysis of the chimera KCC2_4-2-4 and the quadruple cysteine mutant chimera KCC4_4-2q-4. Both chimeras are expressed in HEK-293 cells.

FIGURE 6.

The large extracellular loop of KCC2 but not of KCC4 requires the parental backbone. HEK-293 cells were assayed for ⁸⁶Rb⁺ uptake 48 h after transfection with the constructs indicated below the columns. An empty vector was used for mock transfection. A, KCC2_wt (white bars), KCC4_wt (black bars), and the chimera KCC2_2-4-2 (striped bars) showed transport activity. The transport activity of KCC2_wt and the chimera KCC2_2-4-2 is comparable. KCC2_2-4q-2 displayed strongly reduced transport activity compared with KCC2_wt (p < 0.001). B, KCC4_4-2q-4 and KCC4_4-2-4 are transport-inactive. The plot depicts mean of three experiments ± S.D. ***, p < 0.001; *, p < 0.05; n.s., not significant.

FIGURE 7.

Concentration-response profile for inhibition of KCC2, KCC4, and KCC2_2-4-2 by furosemide. Transfected HEK-293 cells with KCC2, KCC4, and KCC2_2-4-2 were exposed to increased furosemide concentration from 2 to 2000 μm in the preincubation and uptake buffer. The activity of untreated cells of the respective KCC was normalized to 100%. IC₅₀ values were determined by using logarithmic regression analysis. The plot depicts the mean of four experiments ± S.D. ***, p < 0.001.

FIGURE 8.

Cell surface expression of transport-inactive cysteine mutants. Transfected cells were labeled with sulfosuccinimidyl 2-(biotinamido)-ethyl-1,3-dithiopropionate. The cell surface amount of the indicated proteins was measured by quantifying the eluate by immunoblot analysis. A, KCC2_wt and KCC2_q are at similar levels at the surface, whereas KCC2_2-4q-2 was decreased at the surface compared with KCC2_wt (p = 0.021). B, KCC4_wt and the two transport-inactive chimeras KCC4_4-2-4 and KCC4_4-2q-4 are equally expressed at the cell surface. The plot depicts the mean of four experiments ± S.D. (error bars). *, p < 0.05; n.s., not significant.