Genetic inactivation of Kcnj16 identifies Kir5.1 as an important determinant of neuronal PCO2/pH sensitivity

Abstract

The molecular identity of ion channels which confer PCO(2)/pH sensitivity in the brain is unclear. Heteromeric Kir4.1/Kir5.1 channels are highly sensitive to inhibition by intracellular pH and are widely expressed in several brainstem nuclei involved in cardiorespiratory control, including the locus coeruleus. This has therefore led to a proposed role for these channels in neuronal CO(2) chemosensitivity. To examine this, we generated mutant mice lacking the Kir5.1 (Kcnj16) gene. We show that although locus coeruleus neurons from Kcnj16((+/+)) mice rapidly respond to cytoplasmic alkalinization and acidification, those from Kcnj16((-/-)) mice display a dramatically reduced and delayed response. These results identify Kir5.1 as an important determinant of PCO(2)/pH sensitivity in locus coeruleus neurons and suggest that Kir5.1 may be involved in the response to hypercapnic acidosis.

FIGURE 1.

Deletion of the Kcnj16 gene. A, targeting strategy. Top, restriction map of the WT Kir5.1 (Kcnj16) gene used to construct the targeting vector. The entire 419-amino acid sequence of Kir5.1 is encoded on a single exon. Dotted lines define the limits of the recombinagenic arms. Middle, targeting construct in which a neomycin resistance gene (Neo) was inserted to replace amino acids 1 to 360 of the Kir5.1 open reading frame. Bottom, targeted Kcnj16 allele. Restriction enzyme sites: B, BamHI; X, XbaI; R, EcoRI. B, genotype analysis. DNA from ear or tail biopsies was analyzed by PCR using a three-primer set. A 225-bp fragment from the WT Kir5.1 gene was amplified using a forward primer (5′-CTGCTTGCAGTTTGAAGGAAG-3′). This corresponds to codons 325–331 of the mouse Kir5.1 gene and a reverse primer (5′-CATTCATCTTGTGGGGACAGGACGGTCT-3′) corresponding to anticodons 389–397. A 325-bp PCR product from the successfully targeted gene was amplified using the reverse primer from the Kir5.1 gene and a forward primer (5′-AGGGGGAGGATTGGGAAGACAATAGCA-3′) complementary to sequences in the 3′ region of the integrated neomycin resistance gene. PCR cycle parameters were 94 °C for 30 s then 30 cycles of 94 °C for 15 s, 60 °C for 20 s, and 72 °C for 40 s. Samples were run on a 1.8% agarose gel.

FIGURE 2.

Discharge response of LC neurons from Kcnj16^(+/+) and Kcnj16^(−/−) mice to NH₄Cl withdrawal. A, representative current clamp recordings showing the spontaneous firing activity of WT (top) and Kcnj16^(−/−) (bottom) neurons before (CTRL), during NH₄Cl (10 mm) withdrawal, and after returning to base-line level (BL).

FIGURE 3.

LC neurons from Kcnj16^(−/−) mice show a decreased and delayed response to NH₄Cl. A and B, time course of IFF (Hz) change upon 3-min application of 10 mm NH₄Cl from WT (A) and Kcnj16^(−/−) (B) neurons is shown. C and D, ΔIFF (Hz) calculated as the difference between the peak and the base-line spontaneous firing frequency, for WT and Kcnj16^(−/− neurons during NH₄Cl application (C) and withdrawal (D). Note that the negative values in C indicate a firing rate decrease. E, latency of NH₄Cl effect calculated as the difference between the time at IFF peak and the start of drug superfusion. F, regression line slope values of the rising phase of the response to NH₄Cl pre-pulse expressed as a mean ± S.E. (error bars) of percent variation with respect to control.

FIGURE 4.

Extracellular recordings confirm that the discharge response of LC neurons from Kcnj16^(−/−) mice to NH₄Cl is decreased and delayed. A and B, time course of IFF (Hz) change upon 3-min application of 10 mm NH₄Cl from WT (A) and Kcnj16^(−/−) (B) neurons. C and D, ΔIFF (Hz) calculated as the difference between the peak and the baseline spontaneous firing frequency, for Kcnj16^(+/+) and Kcnj16^(−/−) neurons during NH₄Cl application (C) and withdrawal (D). Note that the negative values in C indicate firing rate decrease. E, latency of NH₄Cl effect calculated as the difference between the time at IFF peak and the start of drug superfusion. F, regression line slope values of the rising phase of the response to NH₄Cl pre-pulse expressed as a mean ± S.E. of % variation with respect to control.

FIGURE 5.

NH₄Cl-evoked outward and inward current in LC neurons from Kcnj16^(+/+) and Kcnj16^(−/−) mice. A and B, representative voltage clamp recordings from Kcnj16^(+/+) (A) and Kcnj16^(−/−) (B) neurons before and during the superfusion and withdrawal of NH₄Cl. The holding potential was −60 mV. C and D, amplitudes of the NH₄Cl-induced outward (C) and inward (D) current for Kcnj16^(+/+) and Kcnj16^(−/−) calculated as the difference between base-line and peak current (ΔI, pA). E and F, latency (E) and regression line slope (%) (F) of the NH₄Cl-induced currents. The latency was calculated as the difference between the time at peak current and the start of NH₄Cl superfusion. Values shown are means ± S.E. (error bars).

FIGURE 6.

NH₄Cl-evoked outward current reverses polarity at K⁺ equilibrium potential. A and B, whole-cell I-V relationships in Kcnj16^(+/+) (A) and Kcnj16^(−/−) (B) neurons before (control; filled squares) and during the perfusion of NH₄Cl (open squares). C, I-V plots of the NH₄Cl-induced outward current (I_NH4Cl) in Kcnj16^(+/+) (filled circles) and Kcnj16^(−/−) (open circles) neurons. I_NH4Cl was calculated by subtracting the current obtained before NH₄Cl application from the current obtained in the presence of NH₄Cl. The reversal potential for I_NH4Cl was −106 ± 1.9 mV (n = 6). This potential for Kcnj16^(−/−) neurons could not be determined. D and E, I/V plots calculated before (control; filled squares) and during NH₄Cl withdrawal (open squares) for Kcnj16^(+/+) (D) and Kcnj16^(−/−) (E) neurons. F, I/V plots of the NH₄Cl-induced inward current (I_NH4Cl) in Kcnj16^(+/+) (filled circles) and Kcnj16^(−/−) (open circles) neurons. Reversal of NH₄Cl-induced inward current was not observed in either Kcnj16^(+/+) or Kcnj16^(−/−) neurons. The holding potential was −60 mV; n = 6.

FIGURE 7.

LC neurons from Kcnj16^(−/−) mice show abnormal response to hypercapnic acidosis. A, representative current clamp recordings showing the response of WT (upper) and Kcnj16^(−/−) (lower) neurons to 15% CO₂ and after returning to control solution (5% CO₂). B and C, time course of IFF (Hz) change upon 5-min application of 15% CO₂ for Kcnj16^(+/+) (B) and Kcnj16^(−/−) (C) neurons. D, ΔIFF (Hz) values calculated as the difference between the peak (15% CO₂) and the base-line (5% CO₂) spontaneous firing frequency for Kcnj16^(+/+) and Kcnj16^(−/−) neurons. E, regression line slope (%) of the effect induced by 15% CO₂ superfusion and calculated as detailed in Fig. 3, for Kcnj16^(+/+) and Kcnj16^(−/−) neurons. Data are means ± S.E. (error bars) of 5–8 independent experiments for each group. Statistical significance was calculated by using unpaired Student's t test and ANOVA.

FIGURE 8.

Model for the role of Kir5.1 in LC neuronal excitability and modulation by NH₄⁺ and CO₂. NH₄⁺ ions enter the neuron by means of transporters and permeable ion channels. NH₄⁺ (pK_a 9.2 at pH 7.4) in the bath solution is in equilibrium with a small amount of NH₃ that is highly lipophilic and which diffuses freely across cell membrane. NH₃ then accumulates in the neuron, becomes protonated, and transiently alkalinizes the cytoplasm (NH₃ + H₂O ↔ NH₄⁺ + OH⁻). During the washout of NH₄Cl from the recording chamber, NH₃ diffuses out of the neuron, and the NH₄⁺, which was either generated inside or entered the neuron, rapidly acidifies the cytoplasm (NH₄⁺ + H₂O ↔ NH₃+ H₃O⁺). This increased [H⁺] inhibits Kir4.1/Kir5.1 activity, depolarizing the cell and increasing the neuronal firing rate. By contrast, CO₂ diffuses freely across cell membrane during hypercapnia and, by means of carbonic anhydrases (CA), is rapidly converted to bicarbonate and H⁺_, which inhibits Kir4.1/Kir5.1 channels. In Kir5.1 knock-out animals this chemosensitive signaling mechanism is absent resulting in a reduced response to both NH₄Cl and hypercapnic acidosis.