The calmodulin-binding, short linear motif, NSCaTE is conserved in L-type channel ancestors of vertebrate Cav1.2 and Cav1.3 channels

Abstract

NSCaTE is a short linear motif of (xWxxx(I or L)xxxx), composed of residues with a high helix-forming propensity within a mostly disordered N-terminus that is conserved in L-type calcium channels from protostome invertebrates to humans. NSCaTE is an optional, lower affinity and calcium-sensitive binding site for calmodulin (CaM) which competes for CaM binding with a more ancient, C-terminal IQ domain on L-type channels. CaM bound to N- and C- terminal tails serve as dual detectors to changing intracellular Ca(2+) concentrations, promoting calcium-dependent inactivation of L-type calcium channels. NSCaTE is absent in some arthropod species, and is also lacking in vertebrate L-type isoforms, Cav1.1 and Cav1.4 channels. The pervasiveness of a methionine just downstream from NSCaTE suggests that L-type channels could generate alternative N-termini lacking NSCaTE through the choice of translational start sites. Long N-terminus with an NSCaTE motif in L-type calcium channel homolog LCav1 from pond snail Lymnaea stagnalis has a faster calcium-dependent inactivation than a shortened N-termini lacking NSCaTE. NSCaTE effects are present in low concentrations of internal buffer (0.5 mM EGTA), but disappears in high buffer conditions (10 mM EGTA). Snail and mammalian NSCaTE have an alpha-helical propensity upon binding Ca(2+)-CaM and can saturate both CaM N-terminal and C-terminal domains in the absence of a competing IQ motif. NSCaTE evolved in ancestors of the first animals with internal organs for promoting a more rapid, calcium-sensitive inactivation of L-type channels.

Figure 1. Amino acid sequence conservation of calmodulin binding, amino terminal (NSCaTE) and carboxyl terminal (Pre-IQ/IQ) motifs in L-type Cav1 channels.

(A) Running window of amino acid similarity of aligned Cav1.2, Cav1.3 and four representative invertebrate L-type channels (red asterisk in B). DI, DII, DIII, DIV are the location of the four major domains, each domain consisting of six transmembrane helices. (B) Multiple alignment of N-terminal sequences illustrating the conservation of NSCaTE and a downstream methionine (Met₂) in L-type channels of coelomate animals. (C) C-terminal sequence alignments illustrating the nearly invariant Pre-IQ/IQ motifs in coelomate animals, with a recognizable “IQ” even in single-celled, Paramecium.

Figure 2. Gene tree of L-type calcium channels illustrating the evolution of calmodulin binding, amino terminal (NSCaTE) and carboxyl terminal (Pre-IQ/IQ) motifs in L-type Cav1 channels.

All animal phyla including single cell protozoans (such as Paramecium) and simple multicellular organisms (placozoans, sponge) have an IQ motif. A N-terminal NSCaTE (xWxxx(IorL)xxxx) motif evolved in common invertebrate ancestors (coelomates) and retained in all major phylogenetic groups. While the Pre-IQ and IQ motifs are featured in all L-type channels of metazoans, the NSCaTE motif is missing in some Arthropod species, including many insects, suggesting that NSCaTE is not an essential feature of L-type channels. NSCaTE was lost in Ca_v1.1 and Ca_v1.4 after the speciation of L-type channels to four gene isoforms in vertebrates. Almost every NSCaTE containing L-type channel has a downstream methionine (Met₂) from the start codon (Met₁) which could serve as an alternative translational start site for inclusion (Met₁) or exclusion (Met₂) of NSCaTE in L-type channels.

Figure 3. No difference in levels between snail LCa_v1 (L-type) channel mRNA transcripts containing NSCaTE (LCa_v1-Met₁) and those lacking NSCaTE (LCa_v1-Met₂).

A) Illustration of the location of primer sets for qPCR amplification of snail mRNA transcripts containing or lacking NSCaTE in LCa_v1 channels. Numbers refer to the position of the DNA sequence from LCav1-Met₁ translational start site. B) Graph illustrates the lack of significant difference in mRNA expression levels (relative to HPRT control values, scaled such that the sum of all signal values across tissues are equal) in whole animals or in specific tissues of snails containing or lacking the LCav1_NSCaTE sequence. Degree of correlation with adjusted R² value = 0.95.

Figure 4. No major differences in biophysical properties between snail LCa_v1 channel containing a full-length N-terminus with NSCaTE translated from upstream methionine Met₁ or truncated N-terminus missing NSCaTE generated from the downstream methionine Met₂.

LCa_v1-Met₁ and LCa_v2-Met₂ were transfected in HEK-293T cells alongside mammalian α₂δ₁ and β1b accessory subunits and recorded in 10 mM Barium (Ba) or Calcium (Ca) containing extracellular solution using patch clamp electrophysiology. Intracellular solution contained 9 mM EGTA. (A) Representative current traces generated from voltage steps (−40 mV to 60 mV in 10 mV steps) from a holding potential of −60 mV) illustrating the typical buffer resistant (9 mM EGTA) calcium-dependent inactivation when calcium is the charge carrier, leaving residual voltage-dependent inactivation when barium replaces calcium in the extracellular solution. (B) Normalized current-voltage relationships (n = 10), transformed and Boltzmann-fitted as activation curves in (C). (D) Steady-state availability curves (Ba_ex: n = 10, Ca_ex: n = 6), generated by measuring the fraction of maximal current generated after a 10 s sustained prepulse voltage from −100 to +50 mV in 10 mV steps). (E) Time of recovery from inactivation (n = 4) measured as the fraction of maximal current recovery after time delays, plotted on a log scale.

Figure 5. Full-length LCa_v1-Met₁ channels containing NSCaTE have a buffer-sensitive form of calcium dependent inactivation not found in truncated LCa_v1-Met₂ channels lacking NSCaTE.

LCa_v1-Met₁ and LCa_v2-Met₂ were transfected in HEK-293T cells alongside mammalian α₂δ₁ and β2a accessory subunits and recorded using patch clamp electrophysiology. (A) Overlapping sample traces illustrating the ultra-fast calcium-dependent inactivation (red trace) of LCav₁-Met₁ channel currents in low (0.5 mM) EGTA buffering conditions. (B) Graph of fraction of peak current size at 300 ms time point of inactivation decay (R300). * represents a statistically significantly smaller R300 value (p<0.001, ANOVA) for LCav₁-Met₁ calcium currents (red trace), reflecting the buffer-sensitive calcium dependent inactivation uniquely in LCav₁ channels with a full N-terminus, containing NSCaTE.

Figure 6. Snail NSCaTE_LCav1 and mammalian NSCaTE_Cav1.2 peptides assume an alpha-helix upon binding calmodulin (CaM).

(A) 17 mer NSCaTE and 21 mer IQ peptide sequences are used in circular dichroism (Figures 6) and Gel Shift Mobility Assays (Figures 7). (B,C) Differential spectra of NSCaTE peptide indicate a helical transformation with snail and mammalian NSCaTE upon addition of the helix stabilizing agent, trifluoroethanol (TFE: dashed line; no TFE: solid line, Figure 6B) or upon addition to CaM (solid line = 1∶1 CaM:peptide, dashed line = 1∶2 CaM:peptide, Figure 6C). Each trace is obtained by subtracting the 10 µM CaM-alone trace from the corresponding NSCaTE+CaM spectrum. Y axis units are mean residue ellipticity or (θ), (difference in spectra corrected for total protein/peptide concentration).

Figure 7. Gel Mobility Shift Assays illustrate that snail NSCaTE of LCav1 and mammalian NSCaTE of Cav1.2 can displace calcium-calmodulin (Ca²⁺-CaM) prebound to either snail IQ or mammalian IQ motifs.

Gel shift mobility assays of CaM and individual peptides (A) or CaM pre-bound to an IQ peptide competing with increasing amounts of NSCaTE (B). Each lane contains 300pmol wild-type Ca²⁺-CaM; first lane is a CaM-only control for reference (Positions #2,#3,#5). (A) Each subsequent lane has increasing ratios of C-terminal (IQ) and/or N-terminal NSCaTE peptides at the ratios indicated. The changing conformation of CaM by IQ peptide causes a mobility shift of the CaM band (from Positions #2 → #1, top panels). Neither snail NSCaTE_LCav1 or mammalian NSCaTE_Cav1.2 changed CaM mobility alone, at any ratio (Position #3, bottom panels). (B) First lane is again Ca²⁺-CaM-only control (Position #5). Second lane is Ca²⁺-CaM with 1.5×IQ peptide alone control (for the maximum shift reference, Position #4). In subsequent lanes, increasing the amount of added NSCaTE peptide eventually reversed the slow mobility of CaM-IQ peptide back to the faster mobility of CaM without IQ peptide (Position #4 → #5), but not at a 100% effectiveness.

Figure 8. Isothermal Calorimetry (ITC) analysis indicates a 2∶1 stoichiometry of NSCaTE:CaM in the absence of IQ, and a 1∶1 NSCaTE:CaM stoichiometry when CaM is first pre-bound to IQ.

(A) Representative raw sample data for several CaM-peptide titrations. (B) Summary Table of ITC data. NSCaTE or IQ peptides were titrated into Ca²⁺-CaM alone or Ca²⁺-CaM pre-bound to a competing peptide at a 1∶1 ratio. All binding reactions exhibited negative enthalpy under the experimental conditions used (exothermic, ΔH <0). Despite the lack of mobility shift of Ca²⁺-CaM (Figure 7), NSCaTE does bind Ca²⁺-CaM alone. LCa_v1 and Ca_v1.2 IQ peptides have a higher affinity (Kd = 80 nM and 130 nM, respectively) for Ca²⁺-CaM than do either of the NSCaTE peptides (Kd = 0.83 µM and 3.24 µM for Ca_v1.2 and LCa_v1 NSCaTEs). Both mammalian NSCaTE_Cav1.2 and snail NSCaTE_LCav1 is able to bind to Ca²⁺-CaM pre-bound to IQ motifs, and both IQ peptides are able to bind to CaM when it is first bound to NSCaTE, which is consistent with the possibility of a NSCaTE-CaM-IQ complex in vivo. N values indicate that NSCaTE_LCav1 and NSCaTE_Cav1.2 can bind to Ca²⁺-CaM in a 2∶1 stoichiometry, consistent with NSCaTE possibly binding simultaneously to both the N- and C-lobes of CaM, as previously reported .