Small peptides derived from the autoinhibitory XY linker of phospholipase C-β isoforms inhibit enzyme activity and reduce inflammation

Abstract

Phospholipase C-β (PLCβ) signaling plays a pivotal role in peripheral nociception during inflammation and pain transduction. These enzymes are associated with G-protein coupled receptors of pro-inflammatory and algesic agents that sensitize nociceptors and promote their hyperexcitability. Despite their validation as therapeutic targets, PLCβ isoforms are yet considered undruggable due to the difficulties to identify potent and selective modulators. Here, we address this question and use the autoinhibitory XY linker present in these enzymes as a source of peptide inhibitors of PLCβ activity. We report that peptides patterned after this motif penetrate the cell membrane and interact with PLCβ3 to inhibit PIP₂ hydrolysis and the consequent calcium release from endoplasmic reticulum. These peptides selectively target PLCβ isoforms, without affecting PLCγ-dependent signaling. In primary nociceptor cultures, active peptides attenuate bradykinin-induced electrogenesis and TRPV1 sensitization reducing nociceptor hyperexcitability. Noteworthy, intraplantar administration of a lead peptide prevented inflammation and hypersensitivity in a mouse model of inflammatory pain, highlighting a therapeutic potential. Collectively, our findings indicate that peptides derived from the autoinhibitory XY linker act as selective PLCβ inhibitors with in vivo anti-inflammatory and antinociceptive activity, providing novel pharmacological tools for this enzyme family.

Graphical abstract

Highlights

• •Peptides patterned after PLCβ XY linker selectively inhibit these enzymes.
• •Peptide sequence defines PLCβ binding and activity.
• •These peptides are cell-permeable and show a lack of toxicity.
• •PLCβ peptides attenuate bradykinin-induced electrogenesis and TRPV1 sensitization.
• •The lead peptide shows preventive anti-inflammatory and antinociceptive activities.

1. Introduction

Phospholipases C (PLC) catalyze the hydrolysis of phospholipid phosphatidylinositol 4,5-bisphosphate (PIP₂) into inositol 1,4,5-trisphosphate (IP₃) and diacylglycerol (DAG). IP₃ binding to its receptors promotes calcium release from intracellular stores, while DAG activates protein kinases C (PKC). These second messengers trigger a wide range of intracellular events, ranging from ion channel modulation to exocytosis or cell proliferation and differentiation [1], [2], [3], [4]. Mammalian PLCs include six subtypes of isozymes with different isoforms: β (1−4), γ (1,2), δ (1,3,4), ε, ζ and η (1,2), each one with tissue-specific expression patterns [1].

PLCβ isoforms are relevant in somatosensory perception of exogenous and endogenous stimuli, and in neurotransmission to the central nervous system. Several studies have demonstrated PLCβ expression, especially the PLCβ3 isoform, in different neuronal subpopulations of dorsal root ganglia (DRG) and trigeminal ganglia [5], [6]. In these cells, PLCβ isoforms modulate neuronal excitability through G-protein coupled receptors (GPCRs), as they are differentially activated by Gα_q and Gβγ subunits [7], [8]. Interestingly, the PLCβ pathway modulates Transient Receptor Potential (TRP) channel activity [9]. This type of Ca^2 + -permeable non-selective cation channels work as molecular sensors of thermal, mechanical and chemical stimuli [10]. By far, the best-characterized member is the vanilloid receptor 1 (TRPV1), gated by the vanilloid compound capsaicin, noxious heat (> 43°C), acidic pH or depolarizing voltages [11]. TRPV1 is expressed in subpopulations of Aδ and C-fibers, where it is a key detector and transducer of painful stimuli [12], [13].

Upon injury, peripheral tissues release pro-inflammatory mediators and sensitize nociceptors, leading to hypersensitivity and inflammatory pain [14]. These agents activate multiple sensitizing mechanisms, which are orchestrated by GPCRs-PLCβ pathways. A prominent example is TRPV1 sensitization [15], [16]. Activation of PKCε downstream of PLCβ phosphorylates serine residues in TRPV1, lowering the activation threshold to its activating stimuli [17], [18]. This effect synergizes with the PLCβ-induced reduction in PIP₂ levels to promote TRPV1 sensitization [19], [20], [21]. Furthermore, some pro-algesic agents increase the exocytosis of vesicles carrying TRPV1, rising its expression on the neuronal surface [22], [23]. Given the pivotal role of PLCβ in this inflammatory sensitization, peripheral PLCβ modulation is a valuable approach for the development of anti-inflammatory and antinociceptive drugs.

Currently, the aminosteroid U73122 ((1-(6-((17β-3-methoxyestra-1,3,5(10)-trien-17-yl)amino)hexyl)-1H-pyrrole-2,5-dione)) is widely used as a specific PLC inhibitor [24]. However, many studies have demonstrated the lack of specificity of this compound and off-target inhibitory effects [25], [26], [27], [28], partly due to its ability to alkylate cysteine groups [29]. Additionally, U73122 modulates TRP channels through a PLC-independent mechanism, being a potent agonist of TRPA1 and TRPM4 and a blocker of TRPM3 [30], [31]. Unexpectedly, in cell-free assays with purified PLC proteins, U73122 directly activates PLCβ2, PLCβ3 and PLCγ1 rather than inhibiting them [32]. Due to the off-target effects of U73122 and the involvement of PLCβ proteins in multiple signaling pathways, development of selective PLCβ inhibitors is crucial for understanding PLCβ-mediated cellular processes and for developing potential therapeutic agents.

The structure of PLCβ isoforms includes a N-terminal pleckstrin homology domain, four tandem EF hands, a catalytic triose phosphate isomerase (TIM) barrel domain, a C2 domain and an exclusive C-terminal region with a proximal and a distal domain [7], [33], [34], [35]. Traditionally, the search for enzyme inhibitors was focused on compounds targeting the active site. Unfortunately, this protein region is well preserved among PLC subfamilies which complicates the identification of selective modulators. For instance, high-throughput screening pipelines identified three hit compounds that inhibited the activity of purified PLCβ3, but also of PLCδ1 and PLCγ1 [36]. Intriguingly, nothing else is known about them. Another strategy for PLCβ inhibition was based on the necessary interaction with Gα_q proteins for PLCβ activation. Long peptides based on the helix-turn-helix motif in the proximal domain of PLCβ3, which contacts with Gα_q, disrupted this protein-protein interaction [33], [37]. However, through this site, Gα_q also interacts with other proteins such as p63RhoGEF, and these peptides also inhibited their activation. Altogether, these complications have prevented druggability of PLCβs.

In this scenario, two independent autoinhibitory mechanisms precisely regulate PLCβ isoforms to maintain a low basal activity [38]. The Hα2^’ helix located in the proximal CT domain interacts with a region close to the active site and inhibits its enzymatic activity, until this Hα2^’ helix is displaced by Gα_q binding [34]. Additionally, the XY linker connecting X and Y regions of the TIM barrel catalytic domain blocks the active site when PLCβ is inactive. This autoinhibitory loop is composed of an unconserved N-terminal region, followed by a high number of negatively-charged residues and an ordered region. This includes a short 3₁₀ helix known as lid helix that occludes the active site [39]. Upon PLCβ activation, protein recruitment to the cell membrane promotes removal of the lid helix from the active center, driven by electrostatic repulsions between the acidic part of the linker and the negative charge on the membrane surface [38]. Thus, these autoinhibitory regions are unexploited targets for development of selective PLCβ modulators.

Here, we describe the discovery of small peptide analogues patterned after the XY linker as modulators of PLCβ enzymatic activity. We report that peptides derived from this autoinhibitory region bind to PLCβ isoforms and significantly inhibit their activity, as evidenced from PIP₂ hydrolysis and intracellular calcium transients. Additionally, inhibition of the PLCβ pathway in a neuronal model attenuates bradykinin-induced action-potential firing and TRPV1 sensitization. Finally, the best peptide in terms of activity and solubility shows promising anti-inflammatory and antinociceptive activity in a murine model of inflammatory pain. Collectively, our findings indicate that peptides patterned after the autoinhibitory XY linker act as functionally selective PLCβ inhibitors, that can be tailored to other PLC subfamilies to provide lead compounds for this family of undruggable targets.

2. Materials and methods

2.1. Building of a PLCβ3-XY linker model

To evaluate the interaction between the autoinhibitory XY linker and the isoform PLCβ3, a ligand-receptor complex was built using the structure 3OHM deposited in the Protein Data Bank (https://www.rcsb.org/). This 3D-structure describes the binding of human PLCβ3 to the alpha subunit of the mouse Gq protein, which is required for PLCβ activation. Gα_q protein and non-relevant molecules or ions for the activity of the XY linker were removed. Then, the crystallized region of the XY linker (575-TDEGTASSEVNATEEM-590) was isolated from the rest of the PLCβ3 structure (chain A, residues 12–471 and 593–882) and converted into an independent ligand (chain B, residues 575–590). To enable this separation, residues 591 and 592 were removed. Finally, the complex was repaired with FoldX5 to minimize its free energy by changing residues orientation with bad torsion angles, Van der Waals clashes or high total energy [40] (https://foldxsuite.crg.eu/). Complex was built with YASARA (version 25.1.13, YASARA Biosciences GmbH, https://www.yasara.org/). The reprocessed structure of phospholipase C-β and Gq signaling complex (PDB code: 7SQ2) showed minimal differences in the catalytic domain and the XY linker (RMSD: 0.086 Å).

2.2. Computational design of PLCβ-modulating peptides

The design of PLCβ-modulating peptides started with the fragmentation of the XY linker into smaller overlapping peptides, moving from the N-terminus to the C-terminus with an offset of 1 amino acid. Peptide-receptor complexes containing XY linker fragments ranging from 4 to 10 amino acids were generated with YASARA by removing the remaining portion of the XY loop. The binding free energy of shortened peptides was calculated with FoldX5. These values were normalized to peptide length to compare the binding of peptides of different sizes. More negative values mean better interaction with PLCβ3.

The most promising peptides, according to binding energy or proximity to PLCβ3 active site, were subjected to a sequence space search to increase their affinity. First, shortened peptide-receptor complexes were repaired with FoldX5. Then, the peptide sequence was mutated to a poly-alanine chain to later explore the fitting of the 20 natural amino acids at each position, regardless of residues in consecutive positions. After running the virtual mutagenesis, position-specific scoring matrices were generated with the binding energy of each amino acid normalized to that of the most favorable residue at each position. In these heat maps the binding energy of each residue was represented according to a color scale, with blue colors indicating a good fit, while red ones a bad fit. All computational procedures were carried out with FoldX5 (https://foldxsuite.crg.eu/).

A selection of the most energetically favorable amino acids at each position and residues with different physicochemical properties were combined to obtain designed peptides with increased theoretical affinity. Non-relevant positions for the binding of peptides were kept as in the wild-type XY linker sequence. The binding free energy of designed peptides was estimated and normalized per residue, and peptides were ranked from highest to lowest predicted affinity for PLCβ3. Non-covalent interactions between peptides and PLCβ3 were predicted with the PLIP tool (https://plip-tool.biotec.tu-dresden.de/plip-web/plip/index), using default thresholds for each interaction type [41]. Figures were drawn with open source PyMol (The PyMol Molecular Graphics System 3.0 Schrodinger, LLC, https://www.pymol.org/).

2.3. Sequence conservation among PLCβ isoforms

Canonical protein sequences of human PLCβ1 (Q9NQ66), PLCβ2 (Q00722), PLCβ3 (Q01970) and PLCβ4 (Q15147) were downloaded from UniProt (https://www.uniprot.org/). Sequences were aligned with ClustalOmega from the European Bioinformatics Institute (EMBL-EBI, https://www.ebi.ac.uk/). Based on peptide-receptor complexes, PLCβ3 residues within 3.5 Å of peptides were identified with PyMol. Conservation of these contact residues was analyzed among PLCβ isoforms.

2.4. Cell cultures

Human Embryonic Kidney 293 (HEK293) cells (ECACC, 85120602) were cultured in T25 cm² flasks with DMEM high glucose GlutaMAX (61965–026, Gibco), supplemented with 10% fetal bovine serum (10500–064, Gibco) and 1% penicillin-streptomycin (15140–122, Gibco), at 37 °C and 5% CO₂. Upon 80% confluence, cells were split with 0.25% Trypsin-EDTA (25200–056, Gibco) for 4 min at room temperature. Trypsin reaction was stopped by adding supplemented medium, and cells were plated at the desired density on specific coated surfaces with 0.01% poly-L-lysine (P4707, Sigma-Aldrich). For calcium imaging experiments, HEK293 cells were plated onto 12 mm coverslips at 20,000 cells. For calcium microfluorimetry assays, cells were plated on 96-well black clear bottom plates (3603, Corning) at 30,000 cell/well. Cells were used 48 h after plating.

Human Embryonic Kidney 293 T (HEK293T) cells (ECACC, 12022001) were cultured in T25 cm² flasks with DMEM (21969–035, Gibco), supplemented with 10% fetal bovine serum (10500–064, Gibco) and 1% penicillin-streptomycin (15140–122, Gibco), at 37 °C and 5% CO₂. Upon 80% confluence, cells were split with supplemented medium. After centrifuging the cell suspension at 300 x g for 5 min at 25 °C, pellet was resuspended in the appropriate volume of supplemented medium, and cells were plated at the desired density for transfection protocols.

Immortalized human keratinocyte HaCaT cells (Cytion, 300493) were cultured in T25 cm² flasks with DMEM high glucose GlutaMAX (61965–026, Gibco), supplemented with 10% fetal bovine serum (10500–064, Gibco) and 1% penicillin-streptomycin (15140–122, Gibco). Upon 70% confluence, cells were washed with 0.05% PBS-EDTA for 10 min at 37 ˚C and then detached with 0.05% Trypsin-EDTA (25300–056, Gibco) for 5 min at 37 ˚C. Trypsin reaction was stopped by adding supplemented medium, and cells were plated onto 12 mm coverslips at a density of 20,000 cells. HaCaT cells were used 48 h after plating.

2.5. Drugs for in vitro assays

PLCβ-modulating peptides were synthesized by Biomedal S.L. (Sevilla, Spain) using a solid-phase method, with the N-terminus acetylated and the C-terminus amidated. Trifluoroacetic acid was replaced by the acetate salt and purity was always above 85%. Peptides PL2001 (Ac-NATEEM-NH₂) and PL2002 (Ac-TASSEV-NH₂) were dissolved in PBS-1% NH₄OH at 100 mM stock solution. Peptides PL1621 (Ac-SSEVNA-NH₂), PL2003 (Ac-SSMTNY-NH₂), PL2004 (Ac-NKMEMF-NH₂), PL2204 (Ac-KRIYSSNV-NH₂) and the randomized PL2204 (Ac-IVSNRYSK-NH₂) were dissolved in 100% DMSO at 100 mM stock solution. Peptide PL2005 (Ac-IYSSNV-NH₂) was prepared at 25 mM in 100% DMSO. All peptides were further diluted from the stock up to 100 µM in extracellular solution or cell medium.

Fluorescent peptides PL2003 (5-FAM-Ahx-SSMTNY-NH₂) and PL2204 (5-FAM-Ahx-KRIYSSNV-NH₂) were synthesized by Sb-Peptide (Saint Égrève, France). The fluorophore 5-FAM was conjugated to the N-terminus through the spacer 6-aminohexanoic acid (Ahx). As with non-fluorescent peptides, the C-terminus was amidated and trifluoroacetic acid was replaced by the acetate salt. Fluorescent peptides were prepared at 50 mM in 100% DMSO and diluted from the stock up to 100 µM in extracellular solution.

U73122 hydrate (U6756, Sigma-Aldrich) was dissolved in 100% DMSO at 2 mM stock solution, which was diluted to 2 µM in extracellular solution or cell culture medium for calcium and PIP₂ imaging assays, respectively, or to 5 µM in cell culture medium for MEA recordings. Thapsigargin (T9033, Sigma-Aldrich) was dissolved in 100% DMSO to prepare a 1 mM stock solution and diluted to 5 µM or 1 µM in assay buffer. Atropine sulphate salt monohydrate (A0257, Sigma-Aldrich) was dissolved in distilled water at 200 mM and diluted to 100 µM in assay buffer. For calcium imaging experiments, Fluo-4 AM was dissolved in 100% DMSO at 2 µg/µL. Acetylcholine (A6625, Sigma-Aldrich) was prepared in distilled water at 50 mM and diluted to 1 µM in extracellular solution. For MEA recordings, capsaicin (M2028, Sigma-Aldrich) was dissolved in 100% DMSO at 50 mM and diluted to 500 nM in extracellular solution. Bradykinin acetate salt (B3259, Sigma-Aldrich) was prepared in distilled water at 1 mM and diluted to 1 µM in extracellular solution.

2.6. Calcium microfluorimetry assays

Calcium responses to acetylcholine (ACh) in HEK293 cells were characterized in a microfluorimetry assay with the Fluo-4 NW calcium assay kit (F36206, Invitrogen). 48 h after cell plating on 96-well black plates, growth medium was replaced by 90 µL Fluo-4 NW dye mix prepared in assay buffer (HBSS, 20 mM HEPES), containing 2.5 mM probenecid to prevent dye exclusion. Cells were treated with 100 µM atropine or 5 µM thapsigargin and co-incubated with the dye for 1 h at 37 °C and 5% CO₂. Cells were stabilized for 5 min at 37 °C inside of the microplate reader CLARIOstar Plus (BMG Labtech GmbH, Ortenberg, Germany). Calcium transients were measured along 20 cycles spaced 150 s. After cycle 10, 100 µM ACh was automatically injected, and responses in the presence of each tested compound were measured in triplicate. Intracellular calcium responses to ACh were validated using a calcium-free assay buffer (20 mM HEPES, HBSS (14175–095, Gibco)). The fluorescent dye was excited at 485 nm and emission collected at 520 nm.

2.7. Fluorescence calcium imaging

Non-ratiometric calcium imaging experiments were conducted with the fluorescent indicator Fluo-4 AM (F14201, Thermo Fisher Scientific). HEK293 cells were loaded with 6 µg/mL Fluo-4 AM prepared in extracellular solution (in mM: 140 NaCl, 20 D-mannitol, 10 HEPES, 5 glucose, 4 KCl, 2 MgCl₂ and 1.8 CaCl₂ adjusted to pH 7.4 with NaOH) containing 0.05% (w/v) Pluronic F-127 (P6867, Thermo Fisher Scientific) for 1 h at 37 °C. PLCβ peptides (100 µM), vehicle or U73122 (2 µM) were co-incubated with the calcium dye for 1 h. Afterward, cells were washed with extracellular solution for at least 20 min at 37 °C before starting the imaging. PLCβ was activated by applying two pulses of 1 µM ACh for 30 s, separated by a 180-s wash with extracellular solution to recover basal fluorescence levels. At the end, 1 µM ionomycin (I3909, Sigma-Aldrich) was applied to check cell integrity. Calcium imaging was conducted using an inverted fluorescence microscope (ZEIS Axiovert 200b) coupled to a Hamamatsu FLASH 4.0 LT camera (C11440–42U30, Hamamatsu, Sunayama-cho Japan). The calcium-sensitive dye was excited at 483 nm for 200 ms using the optical beam combiner Lambda 721 (Sutter Instrument, Novato, CA, USA), and fluorescence emission was collected at 512 nm every 3 s. Fluorescence intensity from individual cells was processed with the HC image DIA software (Hamamatsu Photonics).

Reponses to each stimulus were analyzed by measuring the fluorescence peak and subtracting the average basal fluorescence during the 30 s preceding stimulation. Only signals ≥ to 20 arbitrary fluorescence units were analyzed. Individual cell response to ACh was normalized to ionomycin signal. Relative ACh response was calculated by normalizing the ΔF_ACh / ΔF_Ionomycin ratio to the corresponding ratio in vehicle-treated cells. The relative percentage of ACh-responsive cells was determined as the ratio of cells responding to ACh and ionomycin to the total number of ionomycin-responsive cells, and then normalized to the corresponding ratio in vehicle-treated cells.

To assess the selectivity of PLCβ-modulating peptides, HaCaT cells were loaded with 6 µg/mL Fluo-4 AM in 0.05% (w/v) Pluronic F-127, prepared in extracellular solution (20 mM HEPES, 120 mM NaCl, 5 mM KCl, 1 mM MgCl₂, 1 mg/mL sodium pyruvate and 1 mg/mL glucose) containing 2 mM CaCl₂. PL2003 or PL2204 (100 µM), vehicle (0.1% DMSO) or U73122 (2 µM) were co-incubated with the Ca²⁺ dye for 1 h. Before Ca²⁺ measurements, cells were washed with 2 mM Ca²⁺-containing extracellular solution for at least 20 min. Calcium imaging recordings were initiated in Ca²⁺-free extracellular solution. Intracellular Ca²⁺ stores were depleted by 1 µM thapsigargin (TG) in 0.1 mM EGTA for 3 min. Once fluorescence signal returned to baseline, the Ca²⁺-free extracellular solution was replaced by 2 mM Ca²⁺-containing extracellular solution to induce store-operated calcium (SOC) entry. Cells responding to TG and SOC entry with signals ≥ 20 arbitrary fluorescence units were included in the analysis. SOC responses were normalized to the corresponding TG signals. Relative SOC activity was calculated by normalizing the ΔF_SOC / ΔF_TG ratio to the ratio obtained from vehicle-treated cells.

2.8. Cellular uptake of peptides

HEK293 cells were plated at a density of 20,000 cells onto 35 mm µ-Dish, low, ibiTreat (80136, ibidi). After 24 h, nuclei were first labeled with 2 µg/mL Hoechst 33342 (62249, Thermo Scientific) for 10 min at 37 ˚C, followed by washing with extracellular solution (in mM: 140 NaCl, 20 D-mannitol, 10 HEPES, 5 glucose, 4 KCl, 2 MgCl₂ and 1.8 CaCl₂, adjusted to pH 7.4 with NaOH). Cells were then labeled at the plasma membrane with 1:5000 CellMask Deep Red (C10046, Invitrogen) for 10 min at 37 ˚C, followed by washing with extracellular solution. Next, cells were incubated with fluorescent peptides PL2003 or PL2204 for 30, 60 or 120 min at 37 ˚C. Finally, live cells were imaged with a 20x objective in the confocal microscope LSM900 Airyscan 2 (Zeiss, Jena, Germany) after additional washing with extracellular solution.

Fluorescence intensity per cell associated with PLCβ peptides was quantified with ImageJ (Wayne Rasband, NIH) by delineating cells based on the membrane marker and superimposing these regions of interest onto the peptide fluorescence signal. The number of vesicles per cell was calculated by dividing the total number of vesicles per image by the number of cells.

2.9. Subcellular fractionation for membrane isolation

HEK293 cells cultured in T125 cm² flasks to 90% confluence were washed with extracellular solution (in mM: 140 NaCl, 20 D-mannitol, 10 HEPES, 5 glucose, 4 KCl, 2 MgCl₂ and 1.8 CaCl₂, adjusted to pH 7.4 with NaOH) and stimulated with 1 µM ACh for 60 s to activate and concentrate PLCβ at the plasma membrane. Cells were rapidly washed with ice-cold extracellular solution and scraped on ice. The cell suspension was centrifuged at 400 x g for 10 min at 4 ˚C. The resulting pellet was resuspended in lysis buffer (in mM: 250 sucrose, 150 NaCl, 20 Tris-HCl, 1 EDTA, pH 7.4) supplemented with 1% protease inhibitor cocktail (P8340, Sigma-Aldrich) and homogenized on ice with 20 strokes using a Dounce homogenizer. The homogenate was centrifuged at 830 x g for 5 min at 4 ˚C to remove pelleted nuclei. The supernatant was then ultracentrifuged at 100,000 x g for 2 h at 4 ˚C to yield a crude cytosolic fraction and a plasma membrane pellet. Finally, the pellet was solubilized in resuspension buffer (150 mM NaCl, 20 mM Tris-HCl, 1 mM EDTA, 1% Triton X-100 and 1% protease inhibitor cocktail, pH 7.4) and sonicated twice for 5 min on ice.

2.10. Western blotting

PLCβ3 expression in the membrane, cytosolic and total homogenate fractions was assessed by Western blotting following a previously described protocol [42]. Furthermore, total protein was extracted from HEK293 cells passage 30 in lysis buffer (150 mM NaCl, 50 mM Tris-Base, 1 mM EDTA, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, pH 8) containing 1:100 Halt Protease Inhibitor Cocktail EDTA-free (87785, Thermo Scientific). After 40 min of incubation with the lysis buffer, solubilized proteins were collected by centrifugation at 9400 x g for 15 min. Protein concentration was determined with Pierce BCA Protein Assay Kit (23225, Thermo Scientific).

25 µg of total protein and 15 µg of membrane, cytosolic or total homogenate fractions were prepared in loading buffer (60 mM Tris-HCl, 40 mM DTT, 0.01% bromophenol blue, 2% SDS, 5% glycerol, pH 6.8) and heated up to 95 °C for 5 min before gel loading. Proteins were separated by SDS-PAGE in 7.5% polyacrylamide gels. Electrophoresis ran in running buffer (0.3% Tris-Base, 1.44% glycine, 0.1% SDS, pH 8.4) at 100 V during the first 10 min, and then at 150 V until the end. Separated proteins were transferred onto a 0.45 µm nitrocellulose membrane (Bio-Rad) using a wet blotting system for 2 h at 100 V and 4 °C in transfer buffer (25 mM Tris-Base, 190 mM glycine, 0.1% SDS, 20% methanol, pH 8.4). Thereafter, the membrane was blocked with 5% BSA-TBST for 1 h. Isoform PLCβ3, β-tubulin and α-actin were probed with specific primary antibodies anti-PLCβ3 (1:100 in 1% BSA-TBST; sc-133231, Santa Cruz Biotechnology), anti-β-tubulin (1:1000 in 2.5% BSA-TBST; 10094–1-AP, Proteintech) and anti-α-actin (1:1000 in 2.5% BSA-TBST; A2066, Sigma-Aldrich). Additionally, integrin α6/β1 was probed as a membrane-fraction control with the anti-integrin antibody (1:2000 in 1%BSA-TBST; ab181551, Abcam) [43]. Membranes were incubated overnight (16 h) at 4 °C. After six 5-min washes with TBST (20 mM Tris-Base, 150 mM NaCl, 0.1% Tween-20, pH 7.5), membranes were incubated for 1 h at room temperature with secondary antibodies anti-mouse IgG-HRP (1:20,000 in 1% BSA-TBST; A4416, Sigma-Aldrich) to detect PLCβ3 or anti-rabbit IgG-HRP (1:25,000 in 2.5% BSA-TBST; A0545, Sigma-Aldrich) to detect β-tubulin, α-actin and integrin α6/β1. After six 5-min washes with TBST, proteins were detected with the substrate SuperSignal West Pico Plus (34577, Thermo Scientific). Chemiluminescence was visualized with the ChemiDoc imaging system (Bio-Rad).

2.11. Surface plasmon resonance

PLCβ3 from the HEK293 membrane protein extract was captured using a specific PLCβ3 antibody (sc-133231, Santa Cruz Biotechnology) in a Biacore X100 instrument (Cytiva, Marlborough, USA) with two flow cells. The antibody was first immobilized on both flow cells of a Sensor Chip Protein G (29179316, Cytiva) to ∼ 8500 response units (RU) by injecting 20 µg/mL for 900 s at 5 µL/min. To obtain the pull-down of PLCβ3, the membrane protein extract was then passed only over the “active” flow cell at 5 µL/min, until reaching immobilization levels of 250 RU (Supplementary Figures 5B and 5C), whereas the reference flow cell was left without the membrane protein extract. Immobilization levels were estimated from resonance unit values after the initial exponential phase of dissociation. Afterwards, the binding of increasing concentrations of PL2003, PL2204 or the randomized PL2204 (12.5, 25, 50, 100, 125, 150 and 200 µM), prepared in HBS-EP+ Buffer (BE100826, Cytiva), was assessed in multi-cycle kinetic assays by injecting the peptides over both flow cells for 250 s at 5 µL/min, allowing the signal to return to baseline between cycles. Measurements were performed at 25 ˚C. Peptide-PLCβ3 interaction was determined by subtracting the signal of reference flow cell from the signal of active flow cell. Binding levels were calculated at the end of peptide injection after baseline subtraction.

2.12. PLCδ-PH-GFP transfections and PIP₂ imaging

HEK293T cells were plated on 24-well plates at different densities. After 24 h, those wells with approximately 60% confluence were co-transfected with a mix of 400 ng of PIP₂ reporter PLCδ-PH-GFP encoding plasmid (Addgene plasmid #21179, gift from Tobias Meyer) [44] and 400 ng of human bradykinin receptor B₂R plasmid (cDNA Resource Center #BDKB200000, Rolla, MO, USA), using the transfection reagent FuGENE HD (E2311, Promega). 24 h post-transfections, cells were split and plated onto 10 mm coverslips coated with 20 µg/mL poly-D-lysine (P6407, Sigma-Aldrich). Transfected cells were cultured for 24 h before imaging.

PL2003 (100 µM), PL2204 (100 µM, 10 µM, 1 µM), vehicle or the unspecific PLC inhibitor U73122 (2 µM) were pre-incubated for 1 h before starting recordings. Basal fluorescence levels of cells were monitored for the first 30 s in an extracellular solution with calcium (in mM: 160 NaCl, 10 HEPES, 10 glucose, 2.5 KCl, 2 CaCl₂ and 1 MgCl₂ adjusted to pH 7.4 with NaOH). Translocation of PLCδ-PH-GFP from the plasma membrane to the cytosol was induced by a single pulse of 250 nM bradykinin (05–23–0500, Calbiochem) for 2 min. Live-cell imaging was carried out with a Nikon Eclipse TE2000-E inverted fluorescence microscope (40x objective) coupled to a IXON ultra 897 EMCCD camera (Oxford Instruments Andor, Belfast, UK). GFP was excited at 470 nm using the pE-800^fura Illumination System (CoolLED Ltd, Andover, UK), and fluorescence emission was collected at 535 nm every 2 s. Image analysis was done with the NIS-Elements software (Nikon Instruments Inc, NY, USA).

PLCδ-PH-GFP translocation was quantified by measuring the fluorescence increase induced by bradykinin, normalized to basal fluorescence levels (ΔF/F₀). The time to reach maximum fluorescence intensity was evaluated from the onset of PLCδ-PH-GFP translocation. To compare the kinetics of PIP₂ hydrolysis, fluorescence traces were normalized to their maximum and minimum values. The dose-response curve of peptide PL2204 was fitted to non-linear regression Eq. (1), where y is the % of inhibition; x is the peptide concentration; a and b are the minimum and maximum response, respectively; IC50 is the peptide concentration that elicits a halfway response between the maximum and the minimum, and the hill slope describes the steepness of the curve.

y = a + (b - a)/(1 + (IC50/x)^{hill slope}) (1)

2.13. Cell viability assay

HEK293 cells were plated on 96-well plates at a density of 40,000 cells per well. After 24 h, cells were treated with PL2204 at different micromolar concentrations (1, 10, 100 µM), its vehicle (0.1% DMSO) or 0.1% SDS as a cytotoxic control. 24 h post-treatment, MTT reagent (M2128, Sigma-Aldrich) was added to a final concentration of 0.45 mg/mL, and cells were incubated for 1 h at 37 ˚C. The medium was then removed and formazan crystals were solubilized in 100% DMSO with shaking for 1 min. Absorbance was measured at 570 nm, with 620 nm as a reference wavelength, using the microplate reader CLARIOstar Plus (BMG Labtech GmbH, Ortenberg, Germany). Relative cell viability was calculated by subtracting the absorbance at 620 nm from that at 570 nm and normalizing to vehicle. The cytotoxicity threshold was established at 70% cell viability according to ISO 10993–5, with values below this cutoff considered cytotoxic.

2.14. Animals

Procedures were conducted in accordance with approval from the UMH Ethical Committee and the regional government (code: 2022 VSC PEA 0078–2), adhering to European Community guidelines (2010/63/EU) and the ethical standards of the International Association for the Study of Pain [45]. For behavioral experiments, the Institutional Animal and Ethical Committee at Universidad Miguel Hernández de Elche (UMH, Elche, Spain) approved the use of a cohort of 18 C57BL/6JRccHSd male mice (14–16 weeks old, 28–36 g; Harlan, The Netherlands), bred at the animal facility (Servicio de Experimentación Animal, UMH, Elche, Spain). For the primary culture of dorsal root ganglion neurons, neonatal Wistar rats (3–5 days-old, Harlan, The Netherlands) were obtained from the breeding stock at UMH. All animals were housed under controlled conditions (22 ± 1 °C, 55 ± 20% relative humidity) with a 12-hour light/dark cycle (lights on from 8:00 a.m.–8:00 p.m.). Efforts were made to habituate the animals to handling to minimize pain and stress.

2.15. Primary culture of DRG neurons

Dorsal root ganglion neurons (DRGs) were isolated from neonatal Wistar rats (3–5 days-old), humanely euthanized by decapitation. DRGs were harvested from all the spinal levels, with nerve fibers removed, and enzymatically digested for 1 h at 37 °C and 5% CO₂ with a 2.5 mg/mL collagenase type IA solution (C9891, Sigma-Aldrich) in DMEM high glucose GlutaMAX (10566–016, Gibco) medium supplemented with 1% penicillin-streptomycin (15140–122, Gibco). Thereafter, ganglia were mechanically dissociated using a 1 mL micropipette. Single cell suspension was filtered through a 100 µm cell strainer and sensory neurons washed after three successive centrifugations at 300 x g for 5 min in DMEM high glucose GlutaMAX medium supplemented with 1% penicillin-streptomycin and 10% fetal bovine serum (10500–064, Gibco). Finally, DRG neurons were plated in drops on multielectrode array chambers coated with 8.3 µg/mL poly-L-lysine (P9155, Sigma-Aldrich) and 5 µg/mL Cultrex 3-D Culture Matrix Laminin I (3446–005–01, Bio-Techne R&D Systems). After 30 min allowing cell adhesion, cultures were supplemented with 50 ng/mL 2.5S nerve growth factor (N6009, Sigma-Aldrich) and 0.6 µg/mL cytosine arabinoside (C1768, Sigma-Aldrich) in the complete medium. Neuronal cultures were maintained at 37 °C and 5% CO₂ for 48 h before starting recordings.

2.16. Multielectrode array recordings

Extracellular action potentials were recorded in thin multielectrode planar arrays containing 60 electrodes (59 recording electrodes plus one internal reference electrode) with a diameter of 30 µm and an interelectrode distance of 200 µm. The electrical activity was recorded with the MEA1060-INV System at a sampling frequency of 25 kHz, controlled by MC_Rack software version 4.6.2 (Multi Channel Systems MCS GmbH, Reutlingen, Germany). Recordings were performed at 34.5 °C with the temperature controller TCO2 (Multi Channel Systems MCS GmbH). PLCβ peptides (100 µM) or vehicle (0.1% DMSO) were pre-incubated for 1 h in DMEM high glucose GlutaMAX medium supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin, in absence of nerve growth factor. The effect of the unspecific PLC inhibitor U73122 was also tested at 5 µM. To evaluate TRPV1 sensitization, three pulses of 500 nM capsaicin (P1, P2 and P3) were applied for 30 s. Between P1 and P2, cells were washed for 150 s with extracellular solution (in mM: 140 NaCl, 20 D-mannitol, 10 HEPES, 5 glucose, 4 KCl, 2 MgCl₂ and 1.8 CaCl₂, adjusted to pH 7.4 with NaOH). Between P2 and P3, after a 120-s wash, the neuronal culture was stimulated with 1 µM bradykinin for 8 min to activate the PLCβ pathway and potentiate TRPV1. At the end, neuronal excitability was assessed by applying a short pulse of 40 mM KCl. Stimuli were applied using a peristaltic perfusion system PPS2 (Multi Channel Systems MCS GmbH) at a flow rate of 5 mL/min.

Recordings were analyzed using MC_Rack software (Multi Channel Systems MCS GmbH). Raw data were filtered with a second order high pass Butterworth filter, using a cutoff frequency of 50 Hz to remove background noise. Spikes were detected in the filtered data when their amplitude exceeded a threshold automatically set for each electrode at ± 5 µV the standard deviation of the recording. Mean spike frequency for each stimulus was calculated in the 60-s time interval upon stimulation. Only KCl-responsive electrodes were included in the analysis. TRPV1 sensitization was quantified by comparing responses to capsaicin before (P2) and after applying bradykinin (P3) through the relation P3 - P2. The percentage of BK response or TRPV1 sensitization under each treatment was determined by normalizing the corresponding values to those of vehicle.

2.17. Experimental design of behavioral experiment

To maximize the information obtained from the authorized number of animals, the cohort of 18 mice was divided into three groups of six (Supplementary Figure 8). Number of animals and statistical analyses described in the Statistical Analyses subsection (2.18) were similar to previous experimental studies on diverse models of pain [46], [47], [48]. Two groups of animals were exposed to inflammation (i.pl. CFA, n = 6 mice each), while the third served as a control (i.pl. saline, n = 6 mice). To evaluate the preventive effect of PL2204 on inflammation and hypernociception, one of the CFA-exposed groups received vehicle (5% DMSO in saline), and the other was pre-treated with PL2204. Treatments were randomly allocated to homogeneous groups of animals in terms of baseline von Frey sensitivity, using the Random function of Excel. Researchers were blinded to the treatment in the CFA-treated groups (vehicle-CFA vs. PL2204-CFA) but unblinded for the control group. The peptide treatments for the CFA-injected animals were labelled by a person not involved in the experiments and the key was revealed at the end of the procedure. To assess possible therapeutic effects of the peptide, when both CFA-treated groups displayed similar behavior (8 days post-CFA injection), the vehicle-treated group was switched to PL2204, while the PL2204-treated group received vehicle (Supplementary Figure 8). The control group, which did not receive CFA, was injected with vehicle in the same manner.

2.17.1. Model of Inflammation

Inflammation and the associated nociceptive sensitization were induced by subcutaneous injection of 10 µL of 0.5 mg/mL CFA (Complete Freund’s Adjuvant, F5881, Sigma Aldrich, USA) in saline into the plantar side of the left hind paw using a Hamilton syringe with a 30-gauge needle. Control animals received 10 µL of saline.

2.17.2. Drugs tested in behavioral experiments

PL2204 was dissolved in 5% DMSO in saline at 10 mg/mL. A single dose of PL2204 (25 µL, 250 µg) was administered via a 0.3 mL syringe (324826, BD Micro-Fine Demi, BD Medical, France) into the plantar side of the left hind paw, either 2 h before (preventive effect) or 8 days after (curative effect) the induction of inflammation. Control mice received 5% DMSO in saline.

2.17.3. Paw inflammation assessment

Paw thickness was measured before and at 2 h, 24 h, and 4, 7, 8, and 9 days after CFA or saline injection. A manual caliper with a precision of 0.02 mm was used, always at the same location between the insertions of the 1st and 5th digits.

2.17.4. Antinociceptive evaluation

Mechanical sensitivity was assessed by evaluating the hind paw withdrawal thresholds on days −3, −2, −1, and at 3 h, 24 h and 4, 8, and 9 days after CFA injection. Mice were placed in Plexiglas® chambers (10 × 10 × 14 cm) with a wire grid bottom and habituated for 1 h before testing. Von Frey filaments equivalent to 0.04, 0.07, 0.16, 0.4, 0.6, 1 and 2 g were used, applying first the 0.4 g filament and increasing or decreasing the strength according to the response [49]. Four additional filaments were sequentially applied since the first change of response (from negative to positive or from positive to negative). The sequence of the last six responses was used to calculate the withdrawal threshold following the method described by Dixon [50].

Heat Sensitivity was assessed using a plantar test apparatus [51] (Ugo Basile, Italy). Animals were habituated for 1 h before testing. Radiant heat was applied to the plantar surface of the hind paw, with the beam intensity adjusted to achieve baseline latencies of 8–12 s in control mice. A cutoff time of 15 s was applied. The mean hind paw withdrawal latencies were obtained from the average values of three separate trials, taken at 5- to 10-min intervals, to reduce the possible influence of thermal sensitization on the response.

2.18. Statistical analyses

For cellular studies, data normality was first assessed with Shapiro-Wilk test. Depending on the distribution, statistical significance between two groups was determined with an unpaired Student’s t-test or the Mann-Whitney U test. Statistical significance between more than two groups was assessed with a one-way ANOVA followed by Bonferroni’s post hoc test or a Kruskal-Wallis followed by Dunn’s test. Data from fluorescence calcium imaging assays were analyzed with a two-way ANOVA or a mixed model (ACh pulses, treatments and their interaction), followed by post hoc Tukey or Sidak tests. Data points were considered outliers and removed if they exceeded 1.5 times the interquartile range (Q3 – Q1) above the third quartile (Q3) or below the first quartile (Q1). The number of data points per group is shown in brackets, while the number of independent experiments (N) is indicated in the figure legends. For the in vivo experiment, a two-way ANOVA or a mixed model (within-factor: "time"; between-factor: "treatment", and their interaction) were used, followed by post hoc Tukey tests to compare groups and by Dunnett’s tests to compare vs. baseline. In the manuscript, data are expressed as mean ± Standard Error of the Mean or as median with the 95% confidence interval. Statistical significance was set at p < 0.05 and analyses were done with GraphPad Prism 9.4.1 (GraphPad Software Inc., San Diego, CA, USA).

3. Results

3.1. PLCβ peptides patterned after the autoinhibitory XY linker

Using computational techniques of molecular modelling, we designed peptide analogues of the XY linker to inhibit the activity of PLCβ isoforms. As illustrated in Fig. 1A, only a small region at the C-terminal end of the XY linker is crystallized in the PLCβ3 structural model (PDB code: 3OHM). This flexible region covers catalytic domain surface where the active center of PLCβ isoforms is located. In a first step, the XY linker was fragmented into smaller overlapping peptides sequentially moving from the N-terminus to the C-terminus with an offset of 1 amino acid (Fig. 1B). Consequently, a set of shortened peptides containing 4–10 amino acids were selected from the XY linker using the structural context of PLCβ. This strategy allowed selected peptides to be bound to PLCβ3, while the remaining XY linker was removed. A total of 70 peptide-PLCβ3 complexes were modelled, and the interaction was graded by estimating the theoretical binding free energy (Table 1 and Supplementary Tables 1). Normalization by peptide length suggests that small peptides (6-mers) exhibited values similar to those of longer peptides. Therefore, we focused on 6-mer peptides as they may exhibit more favorable physicochemical properties such as solubility, stability, and membrane permeability than longer counterparts.

Fig. 1.

PLCβ peptides patterned after the autoinhibitory XY linker. (A) Catalytic domain of PLCβ3 isoform (blue surface) interacting with the crystallized region of autoinhibitory XY linker (pink cartoon). Approximate location of active site (dashed box). PDB code: 3OHM. (B) XY linker sequence cut into smaller overlapping peptides moving from the N-terminus to the C-terminus with an offset of 1. (C) Left panel, anchor points of wild-type peptide 579-TASSEV-584 (pink, PL2002) bound to PLCβ3 isoform with predicted ΔG per residue. Right panel, PLCβ3 residues that directly interact with this peptide (blue). Solid blue lines: hydrogen bonds. Dashed yellow lines: salt bridges. (D) Sequence space search to optimize interaction of wild-type peptide with PLCβ3. Heat map shows normalized binding energies of the 20 natural amino acids (x-axis) at each position (y-axis) according to a color scale from blue (best fit) to red (worst). (E) The most favorable residues (Designed column) at each position (Position column) of the wild-type peptide (Wild type column) were combined to obtain new peptides. (F) Left panel, designed peptide 579-IYSSNV-584 (yellow, PL2005) shows increased theoretical affinity compared to wild-type (ΔG per residue). Right panel, PLCβ3 residues that directly interact with this peptide (blue). Solid blue lines: hydrogen bonds. Dashed green lines: π-stacking. Dashed grey lines: hydrophobic interactions. ΔG: binding free energy.

Table 1.

Predicted ΔG per residue of the resulting 6-mer peptides obtained from the fragmentation of the XY linker.

Peptides (6 Aa)	ΔG per residue (kcal/mol)
575-TDEGTA-580	-0.192
576-DEGTAS-581	-0.207
577-EGTASS-582	-0.099
578-GTASSE-583	0.080
579-TASSEV-584	-0.039
580-ASSEVN-585	0.217
581-SSEVNA-586	-0.077
582-SEVNAT-587	-0.071
583-EVNATE-588	-0.147
584-VNATEE-589	-0.046
585-NATEEM-590	-0.416

More negative value means better interaction. ΔG: binding free energy. Aa: amino acid.

The 6-mer peptide with the most negative binding energy, and thereby the best interaction, was 585-NATEEM-590 (Table 1). This fragment comes from the C-terminus of the XY linker, distant from the active center. In this peptide, M590 is buried in the structure of PLCβ3, being a potential anchor point (Supplementary Figure 1A). The other 6-mer peptides with the best binding energies are at the N-terminus of the XY linker (575-TDEGTA-580 and 576-DEGTAS-581) (Table 1). These fragments are derived from the XY region known as lid helix that physically blocks the active site of the protein [38]. In this region stands out residue E583, which is inserted into the binding interface, and it could be critical for the interaction with PLCβ3 (Fig. 1C). Thus, we focused on peptides having strong binding energies containing E583 (581-SSEVNA-586 and 579-TASSEV-584), together with peptide 585-NATEEM-590. Using this overlapping selection, the whole crystallized sequence of the XY linker was considered in the design strategy.

The binding of 6-mer peptides to the PLCβ3 catalytic center was evaluated in a sequence space search to obtain peptides with an improved theoretical binding free energy. Using a virtual mutagenesis approach, position-specific scoring matrices for each peptide were obtained with a prediction of the binding free energy of the 20 natural amino acids at each position (Fig. 1D). Heat maps depict restrictive positions in red, where only specific residues fit, and permissive positions in blue, where the 20 amino acids show an optimal interaction. An example of a restrictive position in peptide 585-NATEEM-590 is M590 (Supplementary Figure 1B). This residue is buried in a binding pocket, that only can accommodate the methionine or a hydrophobic amino acid. In contrast, S581 and S582 in peptide 579-TASSEV-584 are clear examples of permissive positions (Fig. 1D). They are oriented towards the opposite direction of the PLCβ3 binding surface allowing almost any amino acid to be fitted. A selection of the best amino acids per position in terms of binding energy with different physicochemical properties were combined to obtain new sequences with an increased theoretical binding affinity, always keeping wild-type residues in permissive positions to save computational time (Fig. 1E). Designed peptides obtained with their predicted binding energies can be found in the source data folder in Mendeley Data (doi: 10.17632/tjdg3svyd4.1). Most of them show an improved binding energy compared to their corresponding wild-type sequence. For each wild-type peptide selected, a designed peptide ranked into the top 10 of binding energies, with a balance of polar and hydrophobic amino acids or positive and negative charges, was evaluated (Table 2).

Table 2.

Summary of PLCβ-modulating peptides selected for their functional validation.

Peptide	Sequence
PL1621 (wild type)	Ac-SSEVNA-NH2
PL2003 (designed)	Ac-SSMTNY-NH2
PL2001 (wild type)	Ac-NATEEM-NH2
PL2004 (designed)	Ac-NKMEMF-NH2
PL2002 (wild type)	Ac-TASSEV-NH2
PL2005 (designed)	Ac-IYSSNV-NH2

Ac: acetylated.

Potential non-covalent interactions of wild-type and designed peptides to the PLCβ3 surface were predicted with the PLIP software. Peptide 579-TASSEV-584 (PL2002) interacts with PLCβ3 through residue E583, whose side chain forms a salt bridge with the amino group of the K468 side chain (Fig. 1C), and creates a hydrogen bond with the N-group of the peptide bond between N414 and E413. Notably, the designed sequence 579-IYSSNV-584 (PL2005) improves its binding free energy (Fig. 1F). The substitution of the alanine at position 580 by a tyrosine introduces a π-π interaction with the aromatic Y648 side chain. Besides, Y580 forms a hydrogen bond through its hydroxyl group with the imidazole of H332. The interchange of the critical E583 by an asparagine does not seem to affect the binding, creating a new hydrogen bond with the E413 side chain, whose γ-carbon also interacts with the Y580 aromatic ring.

The main anchor site for peptide 581-SSEVNA-586 (PL1621) is the residue E583, similarly to 579-TASSEV-584 (PL2002) (Supplementary Figure 2A). In the designed peptide 581-SSMTNY-586 (PL2003), a new hydrogen bond is created through the interaction of the hydroxyl group of Y586 with the amino group of the R470 side chain (Supplementary Figure 2D). Moreover, the aromatic ring of Y586 contacts with the β-carbon of K468. The mutant threonine introduced at position 584 establishes another hydrogen bond with D417 side chain and with the carbonyl group of the peptide bond between D417 and V416.

Finally, peptide 585-NATEEM-590 (PL2001) binds to PLCβ3 through the buried M590 and other contact points (Supplementary Figure 1A). A586 establishes hydrophobic contacts through its side chains with the γ-carbon of R470 and the β-carbon of K468. The hydroxyl group of T587 forms a hydrogen bond with the guanidinium group of R470, while the β-carbon of its side chain contacts with the γ-carbon of Q422. Although E588 is not oriented to the binding surface, the deprotonated carboxyl group of its side chain at physiological pH could interact with the protonated guanidinium group of R470 through a salt bridge. Beyond, β and γ carbons of E589 interact with the side chain of A426 and A423, respectively. Interestingly, in the designed peptide 585-NKMEMF-590 (PL2004), phenylalanine introduction in the buried position 590 allows a T-shaped π-stacking interaction with the F412 aromatic side chain, and hydrophobic contacts with V465 and L593 (Supplementary Figure 1D). Substitution of A586 by a lysine introduces a hydrogen bond with the carbonyl group of the peptide bond between K469 and R470.

In summary, six PLCβ peptides based on the autoinhibitory XY linker were computationally designed. These include three peptides with the wild-type XY sequence and their respective designed peptides with improved theoretical affinity by increasing the number of non-covalent interactions with the binding surface.

3.2. PLCβ-derived peptides inhibit PLCβ-mediated intracellular calcium transients

The inhibitory effect of PLCβ-modulating peptides was assessed recording the intracellular Ca²⁺ transients triggered after PLCβ activation by the muscarinic acetylcholine receptor 3 (m3AChR), endogenously expressed in HEK293 cells [52]. Stimulation with acetylcholine (ACh) elicited a robust calcium response (Fig. 2A). Given the potential effect of ACh on mAChR and nicotinic receptors (nAChRs) [53], [54], ACh responses in HEK293 cells were characterized in Ca²⁺ microfluorimetry assays (Supplementary Figure 3). Blockade of m3AChR with atropine abolished Ca²⁺ responses. Inhibition of the endoplasmic reticulum Ca²⁺-ATPase with thapsigargin (TG) also blocked ACh responses after depleting intracellular Ca²⁺ stores (Supplementary Figure 3A). Furthermore, removal of Ca²⁺ from the extracellular solution did not affect ACh responses (Supplementary Figure 3B). These results reveal that ACh responses in HEK293 cells are mediated by m3AChR coupled to PLCβ, discarding the potential involvement of endogenous nAChRs. This cell line also expresses the protein PLCβ3, used as a reference for the design of PLCβ-modulating peptides (Supplementary Figure 3C).

Fig. 2.

PLCβ-derived peptides inhibit PLCβ-mediated intracellular calcium transients. (A) Representative traces of intracellular Ca²⁺ transients to two pulses of 1 µM ACh (P1, P2) and a final pulse of 1 µM ionomycin in HEK293 cells treated with vehicle (0.1% DMSO), one of the most potent PLCβ peptides (PL2003) or U73122. (B-D) Left panels, peptides patterned after the XY linker significantly attenuate ACh-induced intracellular Ca²⁺. The activity of wild-type XY sequence peptides (purple) is compared to that of designed peptides (orange) and U73122 (yellow). Percentage of ACh response normalized to vehicle. Right panels, PLCβ peptides do not significantly alter the percentage of ACh-responsive cells, unlike U73122. Percentage of responsive cells normalized to vehicle. Bar graphs indicate mean + Standard Error of Mean. Dots are the mean of each recording (N = 3 independent experiments). Number of cells per condition in brackets (n) in left panels. Two-way ANOVA or mixed model followed by Tukey’s tests (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001). ACh: acetylcholine.

PLCβ activity was evaluated after applying two pulses of 1 µM ACh (Fig. 2A). Peptides were tested in pairs to compare activity of wild type to mutated sequences. Peptides (100 µM) were pre-incubated for 1 h to facilitate cell permeability followed by target interaction. As illustrated in Fig. 2B-2D, left panels, 6-mer peptides attenuated ACh-induced intracellular Ca²⁺ raise, showing a slight loss of inhibitory activity after the second stimulation (p < 0.05, ACh P1 vs. ACh P2). These peptides produced an inhibitory effect, similar to that of the unspecific PLC inhibitor, U73122. However, U73122 showed a high inter-assay variability in its inhibition efficacy (40.5 ± 9.4%, p < 0.01 vs. vehicle, Fig. 2B; 65.4 ± 4.3%, p < 0.0001 vs. vehicle, Fig. 2C; 51.9 ± 7.5%, p < 0.0001 vs. vehicle, Fig. 2D, left panels).

As shown in Fig. 2B-2D, left panels, there were not significant differences in the inhibitory potency of wild-type and designed peptides. PL1621 and PL2003 were the most potent peptides, inhibiting PLCβ activity in response to the first ACh pulse by around 50% (52.0 ± 4.9% and 44.6 ± 7.2%, p < 0.01 vehicle vs. PL1621 and PL2003, respectively, Fig. 2B, left panel). The pair PL2001-PL2004 showed the best theoretical binding energies, but they caused a lower reduction of intracellular Ca²⁺ levels compared to PL1621 and PL2003 (32.0 ± 10.7% and 31.3 ± 8.9%, p < 0.05 vehicle vs. PL2001 and PL2004, respectively, Fig. 2C, left panel). PL2002 also induced a significant reduction of ACh responses (39.5 ± 5.9%, p < 0.001 vs. vehicle, Fig. 2D, left panel). Unfortunately, its designed counterpart PL2005 showed solubility problems in different solvents. Nevertheless, a modest inhibitory activity was observed (31.3 ± 7.1%, p < 0.05 vs. vehicle, Fig. 2D, left panel). In all assays, U73122 significantly reduced the number of cells responding to ACh without compromising ionomycin responses (81.5 ± 10.0%, p < 0.05 vs vehicle, Figs. 2B, 66.7 ± 9.3%, p < 0.0001 vs. vehicle, Figs. 2C and 62.0 ± 12.0%, p < 0.0001 vs. vehicle, Fig. 2D, right panels). Unlike U73122, PLCβ-modulating peptides did not significantly alter the percentage of ACh-responsive cells. Together, these data suggest that peptides patterned after the XY linker induce a partial catalytic inhibition of PLCβ enzymes.

3.3. PL2003 penetrates cells

Given the inhibitory effect of PLCβ peptides on PLCβ-mediated intracellular calcium transients, we selected PL2003 as a representative to evaluate its cell-penetrating ability. Peptide internalization was monitored in live cells by using a fluorescent PL2003 conjugated to the fluorescein derivative 5-FAM (HPLC data provided by the synthesis company in the Supplementary File 5-FAM-PL2003 HPLC). After 30, 60 and 120 min of incubation, PL2003-treated cells displayed FAM fluorescence unlike vehicle-treated cells (Figs. 3A and 3B) (Supplementary Figures 4A and 4B). PL2003 localized intracellularly and attached to the plasma membrane in the same optical Z-plane as the nucleus (Fig. 3C and Supplementary Figure 4C) (PL2003 Z-stack video in Supplementary Materials), both locations compatible with PLCβ distribution [55]. The longer the incubation time, the more intracellular PL2003 was accumulated, with intracellular fluorescence significantly increased after longer exposures (Fig. 3D).

Fig. 3.

PL2003 penetrates cells and binds to pulled-down PLCβ3. (A,B) Representative pictures of live HEK293 cells treated with (A) vehicle (0.2% DMSO) or (B) fluorescent PL2003 100 µM (green) for 60 min. Cell membrane labeled with CellMask (red) and nuclei with Hoechst (blue). Scale bars: 20 µm. (C) Zoomed-in view of the highlighted square region in B with a brightfield background, showing intracellular PL2003 localization. Scale bars: 10 µm. (D) PL2003 fluorescence is detectable after 30 min of incubation and increases significantly with longer incubation times. Violin plot data expressed as median (dashed line) with Q1 and Q3 quartiles (dot lines). Number of cells per condition in brackets (n) from one experiment (N = 1). Kruskal-Wallis followed by Dunn’s test (*p < 0.05, ****p < 0.0001). (E) Representative SPR sensorgrams of PL2003 binding to pulled-down PLCβ3 (reference flow cell subtracted from active flow cell). PL2003 (12.5–200 µM) was injected between 100 and 350 s. Traces normalized to baseline. (F) PL2003 binding to pulled-down PLCβ3 increases with concentration. Data expressed as mean + Standard Error of Mean (N = 3 independent experiments). Binding values fitted to a linear regression (y = 0.022x + 0.171). SPR: surface plasmon resonance.

3.4. PL2003 binds to pulled-down PLCβ3

Once we found cellular uptake of PL2003, we investigated its potential binding to PLCβ3 by surface plasmon resonance. For this purpose, we first stimulated HEK293 cells with ACh and used a subcellular fractionation protocol to obtain a membrane protein extract enriched in PLCβ3 (Supplementary Figure 5A) (full images of membrane blots are available in the source data folder in Mendeley Data (doi: 10.17632/tjdg3svyd4.1)). This isoform was captured on a sensor chip using a specific antibody (Supplementary Figures 5B and 5C). As illustrated in Fig. 3E, PL2003 interacted with immobilized PLCβ3 over a range of micromolar concentrations and association steadily increased until the end of the injection. PL2003 binding was concentration-dependent and showed low response values, consistent with the limited molecular mass of the hexapeptide (742.8 Da) compared to PLCβ3 (152 kDa) (Fig. 3F).

3.5. PL2003 inhibits PIP₂ hydrolysis in cells

To further investigate the mode of action of PL2003, we evaluated its effect on the hydrolysis of membrane PIP₂. The level of PIP₂ was optically monitored with the fluorescent construct PLCδ-PH-GFP, encoding a selective PIP₂ binding domain (PH) conjugated to the green fluorescent protein (GFP) [44]. For this purpose, HEK293T cells were co-transfected with the PIP₂ reporter and the bradykinin receptor 2 (B₂R) that also signals through PLCβ. Under basal conditions, the fluorescent reporter is located to the inner leaflet of the plasma membrane bilayer, interacting with PIP₂, with minimal expression in the cytosol (Fig. 4A, left panels). In contrast, bradykinin (BK) stimulation of B₂R activates PLCβ enzymatic activity, which hydrolyses PIP₂, promoting the translocation of the reporter PLCδ-PH-GFP to the cytosol (Fig. 4A, right panels) (PLCδ-PH-GFP translocation videos in Supplementary Materials). Translocation of PLCδ-PH-GFP caused a remarkable increment of cytosolic fluorescence in HEK293 cells pre-incubated with vehicle (Fig. 4B). Noteworthy, 100 µM peptide PL2003 pre-incubated for 1 h, attenuated by 40% the cytosolic fluorescence increase induced by BK (Figs. 4B and 4C), akin to 2 µM U73122 (ΔF/F₀ PL2003: 0.118 [95% CI: 0.101–0.140], PL2003 vehicle: 0.201 [95% CI: 0.174–0.220], U73122: 0.116 [95% CI: 0.097–0.123], U73122 vehicle: 0.201 [95% CI: 0.174–0.220], p < 0.0001 vs. vehicle). In contrast, neither PL2003 nor U73122 modified the time to reach the maximum PLCβ activity, measured from the onset of PLCδ-PH-GFP translocation (Fig. 4D). Normalization of representative PLCδ-PH-GFP traces to their maximum and minimum values yielded traces with similar slopes (Fig. 4E). These findings indicate that PL2003 does not alter the kinetics of PIP₂ hydrolysis, exhibiting a reaction velocity almost identical to that reported in the presence of vehicle.

Fig. 4.

PL2003 inhibits PIP₂ hydrolysis. (A) Representative pictures of HEK293T cells transfected with the PIP₂ probe PLCδ-PH-GFP and pre-treated with vehicle (0.1% DMSO), PL2003 or U73122, before (basal) and after applying bradykinin (BK). Scale bars: 20 µm. (B) Representative fluorescence traces illustrating BK-induced translocation of PLCδ-PH-GFP from plasma membrane to cytosol. A slight fluorescence rundown is observed during basal fluorescence assessment. (C) PL2003 attenuates PLCδ-PH-GFP translocation similarly to PLC inhibitor U73122. (D) PL2003 or U73122 does not modify the time to reach maximum translocation. (C,D) Violin plot data expressed as median (dashed line) with Q1 and Q3 quartiles (dot lines). Number of cells analyzed in brackets (n), from N = 3 or 4 independent experiments in presence of PL2003 or U73122, respectively. Mann-Whitney U (****p < 0.0001). (E) The slope of PIP₂ hydrolysis is similar after vehicle, PL2003 or U73122. BK: bradykinin.

These data suggest that PL2003 directly inhibited the enzymatic activity of PLCβ isoforms reducing the hydrolysis of the membrane PIP₂ and the release of IP₃-mediated Ca²⁺ from the endoplasmic reticulum.

3.6. Positively-charged amino acids at the N-terminus of PL2005 improve solubility while preserving activity

PL2003 was formulated in saline solution for later in vivo testing, but a strong aggregation was observed which limited its potential therapeutic applications. Thus, we designed an aqueous-soluble 6-mer peptide for in vivo pharmacological studies. The designed peptide PL2005 (579-IYSSNV-584), exhibited one of the best binding free energies (Fig. 1F). However, this peptide was initially discarded due to its moderate potency blocking ACh responses (31.3 ± 7.1%, p < 0.05, Fig. 2D, left panel) and its poor aqueous solubility when prepared in 0.4% DMSO in extracellular solution. Nonetheless, because its apparent high binding energy we reasoned that this peptide could be a good therapeutic candidate and it was selected to improve its solubility. For this task, two positive residues (KR) were introduced at the N-terminus to improve water solubility and cell penetration, particularly of neuronal membranes that exhibit a negative resting membrane potential. As depicted in Table 3, the addition of K577 and R578 notably improved the binding energy when compared to the original sequence (577-EGIYSSNV-584). Peptide 577-KRIYSSNV-584 (hereinafter named PL2204) showed a normalized binding free energy similar to PL2005 (ΔG per residue: −1.108 kcal/mol). The guanidinium group of R578 interacts with the Q340 side chain through hydrogen bonds, creating a new anchor point (Fig. 5A). Moreover, the bulky and polar side chain of R578 rotates the side chain of the adjacent I579, which contacts with the β-carbon of the D364 side chain and the indole W366.

Table 3.

ΔG per residue of 8-mer peptides obtained by addition of two positive charges at the N-terminus of PL2005 (579-IYSSNV-584).

Peptides (8 Aa)	ΔG per residue (kcal/mol)
577-KRIYSSNV-584	-1.196
577-RKIYSSNV-584	-1.166
577-RRIYSSNV-584	-0.753
577-EGIYSSNV-584	-0.615
577-KKIYSSNV-584	-0.588

Peptide 577-EGIYSSNV-584 containing wild-type residues at positions 577 and 578 included. More negative value means better interaction. ΔG: binding free energy. Aa: amino acid.

Fig. 5.

Positively-charged amino acids at the N-terminus of PL2005 improve solubility while preserving activity. (A) Left panel, designed 8-mer peptide 577-KRIYSSNV-584 (yellow, PL2204) shows similar ΔG to the original 6-mer PL2005. Right panel, interactions between the 8-mer peptide and PLCβ3. PLCβ3 residues directly interacting in blue. Solid blue lines: hydrogen bonds. Dashed green lines: π-stacking. Dashed grey lines: hydrophobic interactions. (B) Left panel, PL2204 attenuates ACh-induced intracellular Ca²⁺ transients, while a randomized PL2204 peptide not. Percentage of ACh response normalized to vehicle. Right panel, PL2204 slightly reduces the percentage of ACh-responsive cells. Percentage of responsive cells normalized to vehicle. PL2204 and randomized PL2204 with their corresponding vehicles (0.1% DMSO) from independent experiments. Data of bar charts expressed as mean + Standard Error of Mean. Dots are the mean of each recording from N = 3 independent experiments. Number of cells per condition in brackets (n) in the left panel. Mixed model followed by Tukey (*p < 0.05, **p < 0.01, ***p < 0.001). (C) Representative fluorescence traces illustrating BK-induced translocation of PLCδ-PH-GFP from plasma membrane to cytosol in vehicle (0.1% DMSO) or PL2204 (100 µM). A slight fluorescence rundown is observed during basal fluorescence assessment. (D) 10–100 µM PL2204 attenuates PLCδ-PH-GFP translocation-associated fluorescence. (E) Concentration-dependent inhibitory effect of PL2204. IC₅₀ value of 2.5 ± 1.2 µM (mean ± SEM). (F) Similar time to reach maximum translocation after 1–100 µM PL2204 or vehicle. (G) The slope of PIP₂ hydrolysis is similar in PL2204 or vehicle. (D,F) Violin plot data expressed as median (dashed line) with Q1 and Q3 quartiles (dot lines). Number of cells analyzed in brackets (n), from N ≥ 3 independent experiments. Kruskal-Wallis followed by Dunn’s test (***p < 0.001, ****p < 0.0001). ΔG: binding free energy. ACh: acetylcholine. BK: bradykinin. Veh: vehicle. Rand: randomized.

PL2204 was soluble in a physiological saline solution and DMSO. The activity of PL2204 on the intracellular Ca²⁺ transients triggered by ACh was tested following the previously described protocol. PL2204 at 100 µM induced a reduction of ACh responses of approximately 30% (31.8 ± 4.9%, p < 0.01 vs. vehicle), revealing that modifications in the sequence of PL2005 did not change its PLCβ inhibitory activity (Fig. 5B, left panel). The sequence dependence of its inhibitory activity was validated using a randomized peptide containing the same amino acids composition in a random sequence. The randomized PL2204 completely lost its inhibitory activity, indicating that PLCβ inhibition depends on the peptide amino acid sequence. In contrast to previous PLCβ peptides, the percentage of ACh-responsive cells was slightly reduced after PL2204 (90.6 ± 3.3%, p < 0.05 vs. vehicle, Fig. 5B, right panel). In order to evaluate possible toxic effects of the peptide, we conducted a MTT cell viability assay which yielded absence of toxicity after exposure to increasing concentrations during 24 h (1–100 µM, Supplementary Figure 6).

PL2204 also inhibited the hydrolysis of PIP₂ and attenuated the BK-triggered translocation of the PIP₂ probe PLCδ-PH-GFP from the cell membrane to the cytosol (Figs. 5C and 5D) (PLCδ-PH-GFP translocation videos in supplementary materials). As shown in Fig. 5E, PL2204 followed a sigmoidal dose-response relationship in the micromolar range with a maximum inhibition at 100 µM (41.8 ± 2.8%, p < 0.0001 vs. vehicle). The peptide induced similar inhibition at 10 µM (37.2 ± 3.2%, p < 0.0001 vs. vehicle), whereas it lost effect at 1 µM. The IC₅₀ was 2.5 ± 1.2 µM. Like PL2003 and U73122, PL2204 did not alter the time required to reach the maximum PLCδ-PH-GFP translocation (Fig. 5F), displaying identical kinetics to vehicle (Fig. 5G).

Taken together, these results support that rational addition of two positive charges to PL2005 resulted in a more soluble peptide in saline solutions. Although PL2003 appears to be more potent than PL2204 in calcium imaging screenings, both peptides similarly inhibit the PLCβ activity in the PIP₂ assay, suggesting a similar in vitro potency.

3.7. PL2204 penetrates cells and binds to pulled-down PLCβ3 unlike its randomized control peptide

Cell permeability of the positively-charged PL2204 was evaluated in live cells following the previously described strategy for PL2003. Cells treated with fluorescent PL2204 exhibited intracellular fluorescence, indicating cellular uptake (Fig. 6A) (Supplementary Figure 7) (HPLC data provided by the synthesis company in the Supplementary File 5-FAM-PL2204 HPLC). PL2204 significantly accumulated in cells after 30 min of incubation, and accumulation was higher after longer incubation times, reaching similar levels with 60 and 120 min exposures (Fig. 6B). Notably, PL2204 incubations significantly increased the number of rounded intracellular fluorescent vesicles per cell when compared to PL2003, suggesting a greater involvement of the endocytic pathway (Fig. 6C).

Fig. 6.

PL2204 penetrates cells and binds to pulled-down PLCβ3 unlike its randomized control. (A) Representative pictures of live HEK293 cells treated with fluorescent PL2204 100 µM (green) for 60 min. Cell membrane labeled with CellMask (red) and nuclei with Hoechst (blue). Scale bars: 20 µm. (B) PL2204 fluorescence is detectable after 30 min of incubation and higher intracellular concentrations are obtained after 60–120 min exposures. Violin plot data expressed as median (dashed line) with Q1 and Q3 quartiles (dot lines). Number of cells per condition in brackets (n) from one experiment (N = 1). Kruskal-Wallis followed by Dunn’s test (****p < 0.0001). (C) PL2204 internalizes into cells, showing a significant increase in vesicle number compared to PL2003. Violin plot data expressed as median (dashed line) with Q1 and Q3 quartiles (dot lines). Mann-Whitney U (**p < 0.01). (D) Representative SPR sensorgrams of PL2204 (left panel) and randomized PL2204 (right panel) binding to pulled-down PLCβ3 (reference flow cell subtracted from active flow cell). Peptides (12.5–200 µM) were injected between 100 and 350 s. Traces normalized to baseline. (E) PL2204 binding to pulled-down PLCβ3 increases with concentration, displaying compelling difference with a randomized PL2204 peptide. Data expressed as mean + Standard Error of Mean (N = 3 independent experiments). Binding values fitted to a linear regression (PL2204: y = 0.104x + 0.884) (Randomized PL2204: y = 0.048x + 0.626). SPR: surface plasmon resonance.

Surface plasmon resonance revealed also PL2204 interaction with pulled-down PLCβ3 in a concentration-dependent manner (Figs. 6D, left panel, and 6E). Remarkably, a peptide containing the same residues as PL2204 in a randomized order (randomized PL2204) showed strongly reduced binding levels when compared to functional PL2204, with lower values at all tested concentrations (Figs. 6D, right panel, and 6E). These results, together with functional data, suggest that the effective interaction and PLCβ inhibition induced by PL2204 requires sequence-dependent interactions, rather than resulting from nonspecific binding.

3.8. PLCγ activity is preserved in the presence of PL2003 and PL2204

The selectivity of PLCβ-modulating peptides for PLCβ isoforms was assessed recording Ca²⁺ transients evoked by store-operated calcium channels (SOC). Previous works demonstrated that SOC activation in human keratinocytes requires PLCγ activity [56]. To evaluate this mechanism, HaCaT cells were treated with thapsigargin in Ca²⁺-free extracellular solution. TG depleted intracellular Ca²⁺ stores, thereby triggering SOC activation through a PLCγ-dependent mechanism. Subsequent addition of extracellular Ca²⁺ elicited a robust response as a result of Ca²⁺ influx through SOC (Fig. 7A). This process was significantly inhibited in cells pre-treated with the unspecific PLC inhibitor U73122 (55.4 ± 9.4, p < 0.05 vs. vehicle, Figs. 7A and 7B, left panels). In contrast, SOC activity remained unaltered in the presence of PL2003 or PL2204 compared to vehicle-treated cells (Figs. 7A, right panel, and 7B, middle and right panels). These findings suggest that these PLCβ-modulating peptides did not alter PLCγ activity, thus substantiating their selectivity for PLCβ isoforms.

Fig. 7.

PLCγ activity is preserved in the presence of PL2003 and PL2204. (A) Representative fluorescence traces of TG-induced SOC activity in HaCaT cells treated with vehicle (0.1% DMSO), U73122 (left panel) or PL2003 and PL2204 (right panel). Depletion of intracellular calcium stores by 1 µM TG in Ca²⁺-free extracellular solution activates SOC through a mechanism partly dependent on PLCγ. Re-addition of extracellular calcium triggers SOC-mediated calcium influx. (B) U73122 attenuates SOC response, whereas PL2003 and PL2204 have no effect. SOC activity expressed as the ratio ΔF_SOC / ΔF_TG, normalized to vehicle-treated cells. Data expressed as mean + Standard Error of Mean. Dots are the mean of each recording (N = 3 or 2 independent experiments for PL2003-PL2204 or U73122, respectively. Number of cells per condition in brackets (n). Unpaired t-test (*p < 0.05). TG: thapsigargin. SOC: store-operated calcium channels.

3.9. PL2003 and PL2204 inhibit BK-induced neuronal electrogenic activity and TRPV1 sensitization

BK induces PIP₂ hydrolysis through PLCβ, leading to PKC activation by DAG and Ca²⁺ that, in turn, phosphorylates TRPV1 potentiating its activity [9], [21]. In addition, BK promotes neuronal excitability by modulating the activity of other ion channels [57]. Thus, we next investigated whether PLCβ inhibition could attenuate BK-mediated neuronal electrogenesis and TRPV1 sensitization in sensory neurons. For this purpose, primary cultures of neonatal rat dorsal root ganglion (DRG) neurons were seeded on multielectrode arrays (MEA) to monitor neuronal networks activity. Repeated neuronal stimulation with three 500 nM capsaicin pulses (P1, P2 and P3), interspersed with washing periods, induced a progressive TRPV1 desensitization, evidenced by a significant decrease in the neuronal firing evoked by P2 and P3 (Figs. 8A and 8C). However, application of 1 µM BK between P2 and P3 for 8 min promoted an increase in the neuronal electrogenic activity and prevented desensitization of P3 (Figs. 8B, first panel, and 8D). These data indicate that BK activation of PLCβ signaling hyperexcites sensory neurons incrementing their electrical firing.

Fig. 8.

PL2003 and PL2204 inhibit BK-induced neuronal electrogenic activity and TRPV1 sensitization. Representative recordings of TRPV1 activity in DRG neuronal cultures (A) non-sensitized and (B) sensitized with BK in the presence of PL2003 or PL2204, its vehicle (0.1% DMSO) or PLC inhibitor U73122. Three pulses of 500 nM capsaicin (P1, P2, P3) were applied. In the desensitization protocol (A), extracellular solution was applied between P2 and P3. In sensitization protocols (B), 1 µM BK was applied to activate the PLCβ pathway. Magnified view of BK-induced action potentials within dashed squares. (C, D) Mean spike frequency after each pulse of capsaicin (P1, P2, P3) in non-sensitized (C) and BK-sensitized (D) cultures. (E) PL2003 and PL2204 significantly reduce BK-induced action potentials. Mean spike frequency normalized to vehicle. (F) PL2003 and PL2204 attenuate BK-induced TRPV1 sensitization, expressed as P3 – P2 response normalized to vehicle. (C-F) Violin plot data expressed as median (dashed line) with Q1 and Q3 quartiles (dot lines). Number of electrodes recorded in brackets (n), from N = 3 independent experiments. Kruskal-Wallis followed by Dunn’s test (*p < 0.05, **p < 0.01, ****p < 0.0001) or Mann-Whitney U (***p < 0.001, ****p < 0.0001). BK: bradykinin.

We hypothesized that our best PLCβ-modulating peptides may attenuate both the BK-induced TRPV1 sensitization as well as its direct electrogenic activity. Neuronal cultures were incubated for 1 h with PL2003 or PL2204 (100 µM), its vehicle (0.1% DMSO) or the U73122. As depicted in Fig. 8E left panel, PL2003 significantly reduced the firing of BK-evoked action potentials by 50% compared to vehicle (53.1% [95% CI: 41.2–70.6] inhibition, p < 0.0001), reproducing U73122 inhibitory effect (64.7% [95% CI: 41.4–70.6] inhibition, p < 0.0001). Noteworthy, PL2003-mediated PLCβ inhibition also attenuated BK-induced TRPV1 potentiation compared to vehicle (Fig. 8F, left panel). The presence of PL2003 resulted in a significant reduction of capsaicin evoked action potentials (P3) as compared to vehicle (70.7% [95% CI: 47.0–88.2] inhibition, p < 0.05, Fig. 8F, left panel). Intriguingly, although U73122 effectively reduced BK-mediated action-potential firing, it did not significantly attenuate BK-induced TRPV1 sensitization (Fig. 8F, left panel).

PL2204 also caused a significant reduction in BK-evoked action-potential firing compared to vehicle (Fig. 8E, right panel, 33.4% [95% CI: 11.0–48.1] inhibition, p < 0.0001). Furthermore, as shown in Fig. 8F right panel, PL2204 reduced TRPV1 sensitization by nearly 45% compared to vehicle (43.8% [95% CI: 11.9–68.0] inhibition, p < 0.001). The improved PL2204 peptide inhibited BK-induced electrogenicity and TRPV1 sensitization to a similar extent as PL2003. Therefore, due to its enhanced solubility in saline, PL2204 was selected for in vivo evaluation.

3.10. PL2204 displays in vivo anti-inflammatory and antinociceptive activity

Given the prominent in vitro effects of PL2204, we conducted an in vivo behavioral experiment in mice to investigate a possible effect of the peptide in preventing inflammation and nociceptive hypersensitivity. Mice received an intraplantar dose of PL2204 (250 µg) or vehicle, and 2 h later an injection of CFA or saline at the same site. This yielded 3 groups: (1) CFA-injected mice treated with PL2204, (2) CFA-injected mice treated with vehicle, and (3) a saline-injected group treated with vehicle (Supplementary Figure 8). Inflammation was measured 2 h after CFA injection, whereas mechanical and heat hypersensitivity were assessed 3 and 4 h post-injection, respectively.

Mice treated with vehicle and injected with CFA developed strong and persistent inflammation for 7 days (Fig. 9A, p < 0.001 vs. baseline and control group Vehicle+Saline). In contrast, PL2204-treated mice revealed a reduced inflammation 2 h and 1 d after CFA (Fig. 9A, p < 0.001 vs. Vehicle+CFA). This anti-inflammatory effect gradually disappeared at 96 h, when swelling of ipsilateral paw became similar between vehicle and PL2204 treated mice (Figs. 9A, 4–7d).

Fig. 9.

PL2204 displays in vivo anti-inflammatory and antinociceptive activity. (A) Left panel, mice treated with PL2204 showed less CFA-induced inflammation than those treated with vehicle and injected with CFA. Right panel, representative pictures showing reduced CFA-induced inflammation after pre-treatment with PL2204 or vehicle (B) Mice treated with PL2204 showed alleviation of mechanical hypersensitivity 3 h after CFA (N.S. vs. Vehicle+Saline). On the contrary, CFA-injected mice treated with vehicle developed robust mechanical hypersensitivity at all time points. (C) Mice treated with PL2204 showed slightly reduced heat sensitization 4 h after CFA (N.S. vs. Vehicle+Saline, N.S. vs. baseline) whereas CFA-injected mice treated with vehicle showed significant heat hypersensitivity at 4 h and 1 d. (A-C) Mean + Standard Error of Mean (SEM) values are shown. Two-way Anova and mixed models followed by Tukey and Dunnett’s tests. *p < 0.05, **p < 0.01, ***p < 0.001 vs. control (Vehicle+Saline). ^$p < 0.05, ^$$p < 0.01, ^$$$p < 0.001 vs. baseline. ^###p < 0.001 vs. Vehicle+CFA. n = 6 animals per group. CFA: complete Freund’s adjuvant.

CFA also induced pronounced mechanical hypersensitivity in vehicle-treated mice, significant from 3 h to 4 d after injection (Fig. 9B, p < 0.001 vs. baseline and p < 0.01–0.001 vs. control group). On the contrary, PL2204-treated mice showed significant alleviation of mechanosensitivity 3 h after CFA (Fig. 9B, N.S. vs. control group). Although, hypersensitivity emerged 24 h later in this group (Fig. 9B, p < 0.05 vs. control group and p < 0.01 vs. baseline). In terms of heat nociception, vehicle-treated mice developed significant hyperalgesia 4 h after CFA (Fig. 9C, p < 0.01 vs. baseline) which was slightly reduced in PL2204 pre-treated mice.

After investigating the preventive effects of PL2204, its therapeutic activity was assessed. Eight days after CFA, when all animals with inflammation reached similar paw thickness and mechanical hypersensitivity, mice that were initially treated with vehicle were exposed for the first time to PL2204, whereas mice previously exposed to PL2204 were now treated with vehicle. The control group was kept without inflammation and received also vehicle. Two hours after the treatments, mice newly exposed to PL2204 showed a slight reduction in paw thickness when compared to previous measurements (Supplementary Figure 9A, p < 0.05 vs. day 7) and showed a partial antinociceptive effect (Supplementary Figure 9B, N.S. vs. Vehicle+Saline+Vehicle or PL2204 +CFA+Vehicle). On the contrary, vehicle-treated mice showed unchanged paw swelling and mechanical hypersensitivity (Supplementary Figure 9). Hence, therapeutic activity of PL2204 could still be observed at this later stage of the inflammatory process, although such activity was less pronounced than when administered as a preventive treatment.

4. Discussion

The primary contribution of this study is the development of cell-permeable peptide inhibitors targeting PLCβ enzymes through the combination of in silico, in vitro and in vivo complementary approaches. We report that small peptides derived from the autoinhibitory XY linker of PLCβ3 interact with this isoform and reduce PLCβ enzymatic activity, resulting in an attenuation of inflammation-induced neuronal electrogenesis and TRPV1 sensitization. These peptides are sequence- and PLCβ subfamily-specific, since a random sequence peptide was inactive and peptides after the PLCβ XY linker did not affect PLCγ activity. PL2204, an aqueous soluble 8-mer peptide incorporating two positively-charged residues at the N-terminus, exhibited in vivo anti-inflammatory and antinociceptive activity in the CFA mouse model of inflammatory pain. These peptides represent a significant progress towards the development of novel pharmacological leads and enable druggability of these enzymes that have notably resisted translational pharmacological targeting [36], [37], [58].

We hypothesized that the autoinhibitory XY linker, a region that physically occludes the catalytic site of PLC enzymes, could be a source of peptide-based inhibitors. Hence, we used a rational design approach to identify candidates within this motif. XY linkers with different amino acid sequence can be found in PLCβ, PLCδ and PLCε isoforms, while PLCγ isoforms have a more complex XY region with multiple domains [39], [59]. The sequence diversity of this protein region within different PLC subfamilies may facilitate the development of subfamily-selective inhibitors, a yet unmet goal for this enzyme family. Here, we focused on PLCβ taking advantage of the availability of the PLCβ3 structure [33]. Analysis of the XY linker sequence reveals that most of the residues are conserved (62.5%) among PLCβ1-β2-β3 isoforms, and more divergent when compared to PLCβ4 (Supplementary Figure 10A). In contrast, PLCβ3 residues harboring the XY binding site are mostly conserved among the four PLCβ isoforms (Supplementary Figure 10B), suggesting potential cross-reactivity of peptides patterned after the XY linker. Hence, it is plausible that our small peptides primarily target PLCβ3 but may cross-interact to some extent with the other PLCβ isoforms. Nonetheless, sequence differences in the XY linker provide the basis to rationally evolve peptides toward specific PLCβ isoforms.

Previous peptide-based PLCβ inhibitors were patterned after the helix-turn-helix (HTH) motif that interacts with Gα_q proteins [37]. Such Gα_q-PLCβ interaction has also been inhibited optogenetically using HTH peptides [58]. Although these long peptides strongly inhibit PLCβ in vitro, their use as selective PLCβ pharmacological tools is limited because their mechanism of action also disrupts the activation of Gα_q-interacting proteins such as RhoGEFs [37], [60], [61]. The peptides presented here exploit an endogenous autoinhibitory motif to selectively target PLCβ isoforms. The specificity of this molecular approach relies on the heterogeneity of XY linkers among the different PLC subfamilies, and avoids classical targeting through highly conserved PLC catalytic sites or through modulation of interacting proteins [37], [58]. One caveat for the use of peptides could be their ability to reach the intracellular compartment. To circumvent this relevant issue, previously designed HTH peptides required intracellular genetic expression, likely due to poor cell permeability. In our case, PL2003 and PL2204 showed pronounced cell permeability after a short incubation period. Both peptides were detected intracellularly with greater number of vesicles or involvement of the endocytic pathway for PL2204. While this pathway may represent an additional mechanism facilitating PL2204 uptake, it could also reflect sequestration within vesicles destined to lysosomal degradation, as previously described for positively-charged peptides [62]. This may reduce its effective intracellular availability and thereby contribute to its slightly reduced inhibitory effect when compared to PL2003. Thus, reducing peptide size, while maintaining crucial contact points with the target protein seems to be a critical factor for the effective modulation of intracellular targets. Given the limited availability of peptide inhibitors targeting intracellular proteins, the current findings represent a relevant milestone in the development of novel peptidic pharmacological tools with clinical applications.

Molecular modeling guided the design of peptides patterned after the XY linker based on predicted binding energies. These predictions must be interpreted with caution, as peptide-PLCβ interactions were assessed under favorable conditions, without considering the hindrance of the protein’s XY linker. Furthermore, additional factors such as peptide solubility and stability can influence the observed experimental activity. Thus, while direct correlations of predicted interactions with experimentally-observed activity cannot be unequivocally established, binding predictions provide a valuable framework to select peptides for further experimental characterization.

The activity of PLCβ-modulating peptides was assessed in vitro using m3AChR and B₂R receptors coupled to PLCβ isoforms. PLCβ peptides patterned after the XY linker attenuated Ca²⁺ release from endoplasmic reticulum triggered after m3AChR activation, while maintaining the number of responsive cells. Interestingly, the amplitude of Ca²⁺ transients elicited after a second ACh pulse was still partially reduced after the washout period. These findings are compatible with the mode of action of a moderate-affinity PLCβ inhibitor that hinders PIP₂ access to the catalytic site, partially inhibiting enzymatic activity. Unlike glue-type drugs, these molecules with moderate potency may have the therapeutic advantage of efficiently targeting pathological states while preserving the activity of physiologically working enzyme pools. Such a partial inhibition of PLCβ was corroborated with the most potent peptides PL2003 and PL2204 in a PIP₂ hydrolyzing assay, revealing direct effect on the hydrolysis of the membrane phospholipid PIP₂. In addition, the fact that PL2003 or PL2204 did not slow the kinetics of PIP₂ hydrolysis, reveals competitive inhibition with PIP₂ for the active site of PLCβ. In agreement, both the cell permeability of the peptides and their interaction with pulled-down PLCβ3 in surface plasmon resonance assays provide compelling evidence supporting cellular inhibition of PLCβ. SPR provided valuable information about peptide-PLCβ3 interactions. PLCβ3 was captured from membrane extracts by using a Protein G chip coupled to a monoclonal antibody. Since the epitope of PLCβ is located far from the active site, it does not interfere with peptide binding. A residual fraction of nonspecific proteins may be captured or co-captured with PLCβ3, potentially affecting interaction with the peptides. However, blunted binding of the randomized peptide ensures sequence specificity of PL2204 for the active site of PLCβ3.

The moderate potency and efficacy of the peptides may be influenced by several factors. Since PLCs are membrane proteins that hydrolyze lipids, their catalytic site may be hindered in the internal membrane leaflet, limiting peptide access. In addition, the accessibility of the catalytic site is further restricted under basal conditions, as it is occupied by the XY linker and only accessible when the enzyme is activated. Thus, designed peptides must compete with the enzyme’s own XY linker for binding. These constraints likely remain the major factor limiting the extent of inhibition in physiological conditions. Nonetheless, this limitation may represent an advantage for selective inhibition of active PLCβ under pro-inflammatory conditions, when the XY linker is displaced from the active center. Another relevant factor that could affect the potency and efficacy of the peptides is their chemical and metabolic stability. To further refine our understanding of peptide potency, subsequent studies will focus on stability assessments. It is plausible that peptide degradation could account for the relatively high concentration of peptide (10–100 µM) required for PLCβ inhibition. Relevant sources of peptide breakdown could come from intrinsic degradation after dissolution or from protease activity. N- and C-termini of our peptides are protected by acetylation and amidation, respectively. However, small peptides made of natural L-amino acids remain susceptible to degradation by cytosolic proteases, reducing the amount of inhibitor available for PLCβ binding, and potentially limiting extent and duration of inhibition. While stability has not been assessed, the concentrations of PL2003 and PL2204 used in the study maintain their efficacy across assays, arguing against this as a cause of variable efficacy. Considering the above-mentioned limitations, it is remarkable that these 6/8-mer peptides still exhibit micromolar inhibitory activity. Such concentrations are in line with similar peptides, such as DD04107, whose activity as a neuronal exocytosis blocker was reported without observable toxicity [23] and passed all the security and toxicity screenings of a Phase I Clinical Trial (https:www.clinicaltrialsregister.eu/, EudraCT Number: 2016–002846–21).

PL2003 and PL2204 did not affect store-operated calcium entry in HaCaT cells, a process dependent on PLCγ activity [56], which underscored their selectivity for PLCβ. While the active center of PLC is preserved also in other subfamilies such as PLCδ, the activity of the peptides toward these isoforms could not be assessed in cellular assays due to the current lack of selective activators [1]. Nevertheless, the selectivity versus PLCγ distinguishes our peptides from the unspecific PLC inhibitor U73122, thus allowing specific modulation of PLCβ-dependent mechanisms.

Strategies such as the use of cell penetrating peptides have been used to increase membrane permeability [62]. The most frequently used cell-penetrating peptide is the Tat sequence, which is rich in positively charged amino acids and takes advantage of the negative membrane potential to enhance membrane permeability. This strategy is very efficient in translocating peptides, although due to its high positive charge tends to drive peptides to the cell nucleus which exhibits strong negative charge. We used this approach in a conservative manner to increase the solubility of PL2005 by incorporating two positively charged amino acids at the N-terminus, mimicking naturally occurring cell-penetrating peptides, and to minimize nuclear localization [62]. With this modification, our 8-mer peptide PL2204, with a net positive charge (+2), also inhibited PLCβ enzymatic activity. Notably, a randomized version of PL2204 lost the inhibitory activity and displayed a strongly reduced binding, underscoring the relevance of the specific order of amino acids for the effective interaction with the binding surface of PLCβ.

Our peptides significantly attenuated BK-induced electrogenicity in sensory neurons, consistent with a PLCβ-dependent mechanism of BK-induced excitability. Our findings agree with growing evidence suggesting that depolarizing effects of bradykinin in nociceptive neurons are mediated by inhibition of M-type K⁺ channels and opening of Ca²⁺-activated chloride channels such as ANO1 [57]. These events depend on the rise of intracellular calcium triggered by the PLCβ-IP₃ pathway [57], [63]. PLCβ signaling is also involved in inflammatory sensitization of TRPV1 channels expressed in sensory neurons [64], [65], [66]. Notably, our peptides reduced BK-elicited TRPV1 sensitization. Inhibition of PLCβ signaling with our peptides was less potent in preventing TRPV1 inflammatory potentiation than direct B₂R blockade with the specific antagonist HOE140 [22]. This pharmacological approach likely causes stronger inhibition because B₂R also activate Gα_s proteins coupled to adenylate cyclase [67]. Specifically, adenylate cyclase generates cAMP that activates protein kinase A, which also phosphorylates and sensitizes TRPV1 [68]. Since PLCβ isoforms are not involved in this pathway, a degree of TRPV1 potentiation is expected after the PLCβ-directed peptides. Remarkably, U73122 did not attenuate TRPV1 sensitization despite reducing BK-induced neuronal electrogenic activity. This observation suggests either incomplete PLCβ inhibition by U73122 or engagement of alternative mechanisms, particularly given its already known off-target effects on members of the TRP family [30], [31].

While our main focus was the study of BK response, the inflammatory soup contains additional mediators that trigger neuronal PLCβ activation, including histamine, protease-activated receptor 2 or ATP [14], [64], [65], [66]. The effect of these mediators should then be also sensitive to the PLCβ-modulating peptides, and a stronger overall effect on inflammation could be expected. Furthermore, the PLCβ pathway modulates other members of the TRP family under inflammatory conditions [9]. For instance, TRPA1, a sensor of environmental irritants and endogenous proalgesic agents involved in inflammatory pain [69], undergoes sensitization through PLCβ-dependent mechanisms [70], [71]. Thus, inhibition of PLCβ should attenuate the sensitization triggered by multiple pro-nociceptive mediators under inflammatory pain conditions. Future experiments will address this question.

Local administration of PL2204 before induction of inflammation had a pronounced anti-inflammatory effect detected as early as 2 h after the injury, which lasted up to 24 h. No signs of toxicity were appreciated during the 9 days following administration. This preventive effect is compatible with the observed inhibition of BK signaling mediated by the PLCβ-modulating peptide. BK is a potent inductor of plasma extravasation and hyperalgesia, and it is involved in neurogenic inflammation [72], [73]. In terms of inflammation, PL2204 had a subtle effect when administered 8 days after the induction of inflammation, suggesting less participation of PLCβ at these later stages. Although certain reduction in paw swelling was found, this finding highlights the dynamism of vascular events involved in the inflammatory process. Previous anti-inflammatory effects were described after administration of the PLC inhibitor U73122 [74]. However, this compound has multiple off-target effects [25], [26], [27], [28], [30], [31]. Hence, our findings underscore the specific participation of PLCβ in the initial phases of inflammation.

PL2204 anti-inflammatory effect was accompanied by decrease of mechanical and thermal hypersensitivity in the injured paw during early inflammation stages. This antinociceptive activity could be due to reduced swelling, and it is in line with the decrease in TRPV1 sensitization observed in cellular studies. Our data reveal the involvement of PLCβ signaling in this mouse model of inflammatory sensitization. Although only males were tested, the promising results warrant further experiments in females. Previous studies found similar results in neuropathic pain models, although the PLC inhibitor U73122 was systemically administered [6]. Given the number of processes modulated by PLCβ, it is expected that unwanted effects could occur after such systemic administration. While generalized effects could be interesting for assessing the role of PLCβ in mechanistic studies, the most plausible route for potential clinical applications would be local or topical, always in the presence of pathophysiological alterations involving PLCβ recruitment.

Despite the observed inhibitory activity, further optimization will be necessary to enhance the translational potential of our peptides for therapeutic applications. In this context, peptide lipidation facilitates their anchoring into the lipid bilayer, thereby increasing their membrane concentration [75]. This modification was used by others to modulate protein-protein interactions at the membrane surface, such as those involved in the SNARE complex [76], interaction of GPCRs with heterotrimeric proteins [77] and interaction of intracellular domains involved in channel gating [78]. However, lipid anchoring restricts conformational flexibility and it may be less suitable to promote binding of small peptides to a catalytic site that requires structural adaptability. Flexible linkers could mitigate this limitation by providing greater mobility. Although we do not fully discard lipidation as strategy to improve peptide potency or cutaneous permeability, we do not prioritize this because it could also increase toxicity, most likely by altering membrane physicochemical properties and downstream signaling processes [75]. Peptide stapling can also improve affinity and proteolytic resistance by promoting α-helical conformation [79]. However, this enforced constraint may reduce flexibility and binding to the catalytic site in our peptides. Given their potential degradation by endogenous proteases, substitution of natural L-amino acids with D-amino acids confers significant protection against proteolytic degradation [80] and it seems an appropriate way to increase effective peptide concentration and extend the duration of inhibition.

In conclusion, peptides patterned after the autoinhibitory XY linker are cell-permeable, effective and selective PLCβ inhibitors with in vivo anti-inflammatory and antinociceptive activity. PL2003 partially inhibited PLCβ enzymatic activity and BK-induced TRPV1 sensitization, being an optimal pharmacological tool for in vitro study of PLCβ-dependent processes. PL2204 also showed PLCβ inhibitory activity, and its enhanced solubility in saline enables its potential use for therapeutic application. Indeed, local administration of PL2204 prevented inflammation and hypersensitivity in a mouse model of inflammatory pain. Our peptides provide a framework for understanding the role of PLCβ in cellular and pathophysiological processes through the modulation of PLCβ isoforms, considered until now undruggable targets.

Author contributions

J.A.L. executed and analyzed in silico experiments; designed, performed and analyzed functional validation cell-based assays, cell-penetration assessments, and surface plasmon resonance assays; conducted and analyzed behavioral assays, and wrote the manuscript. D.C. designed, conducted and analyzed behavioral assays, cell-penetration assessments and surface plasmon resonance assays, and wrote the manuscript. I.D. supervised the experiments, conceptualized the project and revised the manuscript. G.F.B. designed and supervised in silico experiments, and revised the manuscript. N.G. and S.S. designed and supervised PIP₂ imaging assays, and revised the manuscript, M.A.B. and R.G.M. synthesized peptides and revised the manuscript. A.F.C. conceptualized the project, provided funding and revised the manuscript. A.F.M. supervised and designed experiments, conceptualized the project, provided funding and wrote the manuscript.

Funding

Financial support of the Spanish Ministry of Science, Innovation and Universities [PID2021–126423OB-C21 to A.F.C. and A.F.M. and C22 to R.G.M.] and from Generalitat Valenciana [GVA-PROMETEO/2021/031] was granted to A.F.M. J.A.L. was a recipient of an FPI fellowship from the Spanish Ministry of Science, Innovation and Universities [PRE2019–091317]. Work in N.G. laboratory was supported by the Wellcome Trust Investigator Award 212302/Z/18/Z and BBSRC Project Grant BB/V010344/1.

CRediT authorship contribution statement

Jorge de Andrés-López: Writing – original draft, Visualization, Methodology, Investigation, Formal analysis. David Cabañero: Writing – original draft, Visualization, Methodology, Investigation, Formal analysis. Isabel Devesa: Writing – review & editing, Supervision, Conceptualization. Shihab Shah: Writing – review & editing, Supervision, Methodology. Gregorio Fernández-Ballester: Writing – review & editing, Supervision, Software, Methodology. Nikita Gamper: Writing – review & editing, Supervision, Resources, Funding acquisition. Mª Ángeles Bonache: Writing – review & editing, Resources. Rosario González-Muñiz: Writing – review & editing, Resources, Funding acquisition. Asia Fernández-Carvajal: Writing – review & editing, Supervision, Funding acquisition, Conceptualization. Antonio Ferrer-Montiel: Writing – original draft, Supervision, Funding acquisition, Conceptualization.

Declaration of Competing Interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Jorge de Andrés-López has patent #WO2025132812A1 issued to ANTALGENICS, S.L. Isabel Devesa has patent #WO2025132812A1 issued to ANTALGENICS, S.L. Gregorio Fernández-Ballester has patent #WO2025132812A1 issued to ANTALGENICS, S.L. Antonio Ferrer-Montiel has patent #WO2025132812A1 issued to ANTALGENICS, S.L. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary material

Supplementary Tables and Figures

5-FAM-PL2003_HPLC

5-FAM-PL2204_HPLC

PL2003_Z-stack_video

PLCdelta-PH-GFP_translocation_PL2003

PLCdelta-PH-GFP_translocation_PL2204

PLCdelta-PH-GFP_translocation_U73122

PLCdelta-PH-GFP_translocation_vehicle

Data availability

Raw data, source data and results of statistical analysis are deposited in Mendeley Data (doi: 10.17632/tjdg3svyd4.1).