Secreted Isoform of Human Lynx1 (SLURP-2): Spatial Structure and Pharmacology of Interactions with Different Types of Acetylcholine Receptors

Abstract

Human-secreted Ly-6/uPAR-related protein-2 (SLURP-2) regulates the growth and differentiation of epithelial cells. Previously, the auto/paracrine activity of SLURP-2 was considered to be mediated via its interaction with the α3β2 subtype of the nicotinic acetylcholine receptors (nAChRs). Here, we describe the structure and pharmacology of a recombinant analogue of SLURP-2. Nuclear magnetic resonance spectroscopy revealed a 'three-finger' fold of SLURP-2 with a conserved β-structural core and three protruding loops. Affinity purification using cortical extracts revealed that SLURP-2 could interact with the α3, α4, α5, α7, β2, and β4 nAChR subunits, revealing its broader pharmacological profile. SLURP-2 inhibits acetylcholine-evoked currents at α4β2 and α3β2-nAChRs (IC50 ~0.17 and >3 μM, respectively) expressed in Xenopus oocytes. In contrast, at α7-nAChRs, SLURP-2 significantly enhances acetylcholine-evoked currents at concentrations <1 μM but induces inhibition at higher concentrations. SLURP-2 allosterically interacts with human M1 and M3 muscarinic acetylcholine receptors (mAChRs) that are overexpressed in CHO cells. SLURP-2 was found to promote the proliferation of human oral keratinocytes via interactions with α3β2-nAChRs, while it inhibited cell growth via α7-nAChRs. SLURP-2/mAChRs interactions are also probably involved in the control of keratinocyte growth. Computer modeling revealed possible SLURP-2 binding to the 'classical' orthosteric agonist/antagonist binding sites at α7 and α3β2-nAChRs.

Figure 1. Amino acid sequence alignment and comparison of the spatial structure of SLURP-2 with other human Ly-6/uPAR proteins and three-finger snake neurotoxins.

(A) The sequence of the water-soluble domain of human Lynx-1 (ws-Lynx1) is shown without the GPI consensus sequence at the C-terminus. The positively charged (Arg/Lys), negatively charged (Asp/Glu), and His and Cys residues are highlighted. The fragments corresponding to β-strands in the spatial structures of the proteins are underlined. The loop regions are highlighted with a yellow background. (B) The 2D ¹H,¹⁵N-HSQC spectrum of rSLURP-2 (0.5 mM, 5% dioxane, pH 5.0, 37 °C). The obtained resonance assignments are shown. The resonances of the side-chain groups are indicated by a superscripted “s”. The signals corresponding to the unfolded/aggregated protein are marked by asterisks. The relative population of this protein form did not exceed 10%. (C) The set of the best 20 rSLURP-2 structures were superimposed over the backbone atoms in regions with a well-defined structure. The three loops and ‘head’ of the protein are labeled. Disulfide bonds are shown in orange. (D) Ribbon representation of the spatial structure of rSLURP-2 with mapped dynamic NMR data. ¹⁵N relaxation rates were measured at 60 MHz (pH 5.0, 37 °C) for 0.08 mM rSLURP-2 in water without the addition of dioxane (Supplementary Fig. S4). The backbone fragments affected by dynamic processes on the ps-ns time scale (heteronuclear NOE < 0.7) or μs-ms time scale (R₁ · R₂ product > 20 s⁻² or HN signals broadened beyond the detection limit) are highlighted. Additional electrostatic and hydrogen bonding interactions that stabilize the protein fold are shown. Backbone amide and carbonyl groups are indicated by blue and red spheres, respectively. (E) Comparison of the spatial structures of rSLURP-2, ws-Lynx1 and the WTX[P33A] mutant. Aromatic/hydrophobic, positively charged (including His), negatively charged, and Cys residues are indicated in green, blue, red, and orange, respectively.

Figure 2. rSLURP-2 binds to different nAChR subunits in human brain extracts.

Affinity purification was performed with rSLURP-2 that was covalently coupled to magnetic beads or non-coupled beads (Ctrl) using human temporal cortical homogenates. The samples were subjected to gel electrophoresis and Western blotting along with the homogenate samples used for affinity purification (Input), followed by the detection of nAChR subunits. Representative blot images obtained in one (of two) independent experiment are shown. The representative blot images from the second experiment are shown in Supplementary Fig. S5. The two bands were observed for the β2 subunit. As shown previously, only the lower band corresponded to the β2 subunit.

Figure 3. Effect of rSLURP-2 on nAChRs expressed in Xenopus oocytes.

(A,C,E) Electrophysiological recordings of currents evoked by 100, 10, and 100 μM of ACh at α3β2, α4β2, and α7 nAChRs, respectively. The concentrations of ACh were similar to the half activation concentrations (EC₅₀) of the corresponding receptor subtypes. The 5-second pulses of ACh are indicated by the arrows above the traces. The concentration of rSLURP-2 is indicated above the traces. (B,D,F) Dose-response curves of ACh-evoked currents by rSLURP-2 at the α3β2, α4β2, and α7 nAChRs. Each point represents the average ± S.E. of seven, six and seven independent experiments, respectively. The Hill equation (y = A0/(1+([rSLURP-2]/IC₅₀)^nH)) was fitted to the normalized data (% of control) obtained at the α4β2 and α7 receptors. For the α4β2 receptor, the value of the scaling parameter (A0) was fixed at 100%. The calculated A0 value for the α7 receptor was 122 ± 3%. The calculated IC₅₀ and nH parameters were 1.7 ± 0.4 μM and 0.54 ± 0.03, and 3.0 ± 0.2 μM and 1.90 ± 0.14, for the α4β2 and α7 receptors, respectively. (G) Electrophysiological recordings of the ACh-evoked current at α7 nAChR in the absence and presence of 30 nM rSLURP-2. Currents were elicited by 5-second pulses of 40 μM ACh. ACh pulses were interrupted by periods of silence. A representative trace from nine independent experiments is shown. Green traces represent the responses evoked by ACh in the absence of the compound, orange traces are the responses evoked by the same ACh test pulse in the presence of rSLURP-2, and blue traces are the responses evoked by ACh after terminating the rSLURP-2 application. (H) Cartoon demonstrating the priming activity of SLURP-2 at α7 nAChRs. ACh and SLURP-2 molecules are indicated as yellow and red spheres, respectively. The open α7 nAChR channels are indicated by green and orange cores. Closed channels have a white core.

Figure 4. SLURP-2 reduces nicotine-mediated phosphorylation of ERK1/2 in PC12 cells.

(A) An increase in ERK1/2 phosphorylation in PC12 cells was elicited by 25 μM nicotine, and this effect was blocked by pre-incubation with 10 μM recombinant SLURP-2. Values are presented as the ratio of phosphorylated (pERK1/2) to total ERK1/2 (totERK1/2) protein and are normalized to the 25 μM nicotine group (n = 8, mean ± S.E). **indicates a significant difference (p < 0.01, ANOVA followed by Dunnett’s multiple comparisons test) compared with the 25 μM nicotine group. (B) Representative images of Western blots summarized in (A).

Figure 5. Effects of rSLURP-2 on the growth of Het-1A cells.

(A) Influence of rSLURP-2, rSLURP-1, and ws-Lynx1 on Het-1A cell growth (% of control, n = 6–12, mean ± S.E.). The Hill equation (y = A1(100%-A1)/(1+([protein]/IC₅₀)^nH)) was fitted to the data measured using WST-1 reagent. The data for rSLURP-1 were obtained from ref. . The calculated EC₅₀, nH and A1 parameters were 7.6 ± 1.0 nM, 1.5 ± 0.4 and 116 ± 1% (rSLURP-2/Het-1A), and 4.3 ± 0.6 nM, 1.4 ± 0.2 and 60 ± 1% (rSLURP-1/Het-1A), respectively. (B) Effects of rSLURP-2 (1 μM), Atr (1 μM), Mec (10 μM), α-Bgtx (1 μM), MII (1 μM) and their co-application on the growth of Het-1A cells after 48 hours. Each bar is the mean ± S.E. of four independent experiments performed in triplicate. The pairwise statistical analysis of the data groups measured with and without rSLURP-2 was done using t-test. Data indicated as **(p < 0.01) and ***(p < 0.001) are significantly different from each other. Multiple comparisons of all the data groups with the ‘control’ group and with the ‘rSLURP-2’ group were done using ANOVA followed by special Dunnett´s multiple comparisons test. Data marked with ^¤ and ^#(p < 0.05), ^¤¤ and ^##(p < 0.01), ^¤¤¤ and ^###(p < 0.001) are significantly different from ‘rSLURP-2’ and the ‘control’, respectively. (C) Effects of rSLURP-2 (1 μM) and its co-application with atropine (Atr, 1 μM), Mec (10 μM), α-Bgtx (1 μM), or α-conotoxin MII (MII, 1 μM) on the morphology of Het-1A cell nuclei after 24 and 48 hours. The cells nuclei were colored with Hoechst 33342 and propidium iodide.

Figure 6. Influence of rSLURP-2 on [³H]-NMS binding to mAChRs.

(A,B) Interaction of rSLURP-2 with ³H-NMS binding at M1 and M3 receptors, respectively. Membranes (20 μg of protein) were incubated in the presence of the indicated concentrations of rSLURP-2 and 100 pM ³H-NMS. ³H-NMS binding is expressed as the percent of control binding in the absence of rSLURP-2. Data points are means ± S.E. of four independent experiments performed in quadruplicate. The equation y = 100 * ([NMS]+K_d)/{[NMS]+K_d·(K_a+[rSLURP-2])/(K_a+[rSLURP-2]/α)} was fitted to normalized data. The K_d of ³H-NMS binding (197 and 187 pM for M1 and M3 mAChRs, respectively, see Supplementary Table S2) was determined in parallel saturation experiments. The calculated K_a (equilibrium dissociation constant of rSLURP-2) and α (factor of cooperativity) parameters were 231 ± 81 nM and 1.296 ± 0.051, and 144 ± 56 nM and 1.128 ± 0.013, for the M1 and M3 receptors, respectively. *and **indicate significant differences compared with the control (p < 0.05, p < 0.01, respectively, t-test). (C) Interaction of rSLURP-2 with ³H-NMS binding at M2, M4, and M5 receptors. Membranes expressing 5–20 μg of protein were incubated in the presence of 1.4 μM rSLURP-2 and 150 pM ³H-NMS. Bars represent the mean ± S.E. of specific ³H-NMS binding in two independent experiments performed in quadruplicate. (D) Influence of rSLURP-2 on the dissociation rate of [³H]-NMS at M1 and M3 mAChRs. Time course of [³H]-NMS dissociation in the presence and absence of rSLURP-2 was determined as described in Methods. Specific binding of [³H]-NMS was calculated as the difference between the total and non-specific binding measured in the same experiment and is expressed as the percent of initial binding (ordinate). The exponential decay equation was fitted to the normalized data (% of control binding). The (insert) 4.2 μM rSLURP-2 changed the [³H]-NMS dissociation rate constants (K_off, min⁻¹) from 0.082 ± 0.002 to 0.093 ± 0.002 (n = 4, p < 0.05, t-test) and from 0.066 ± 0.001 to 0.059 ± 0.002 (n = 4, p < 0.05, t-test) at the M1 and M3 receptors, respectively. The dissociation rate constants are presented as the means ± S.E. of the values obtained in four independent experiments performed in triplicate.

Figure 7. Modeled complexes of rSLURP-2 with α3β2 and α7 nAСhRs.

(A,B) rSLURP-2 complexes with ‘closed’ binding sites in the α3β2- and α7-nAСhRs. (C,D) Two solutions of the rSLURP-2 complex with an ‘open’ binding site in α7-nAСhR. Top and side views of the obtained models are shown. The rSLURP-2 molecule is shown in blue, and its disulfide bonds are indicated in orange. The three loops of rSLURP-2 are labeled. For the α3β2 receptor, the α3 subunits are shown in light green, and the β2 subunits are indicated in wheat. For the ‘closed’ binding site of the homopentameric α7 receptor, the principal and complementary subunits are shown in light green and wheat, respectively. For the ‘open’ binding site of the homopentameric α7 receptor, the principal and complementary subunits are indicated in magenta and light green, respectively. Loops C of the α3 and α7 subunits are indicated in red.