T-type voltage-gated channels, Na⁺/Ca²⁺-exchanger, and calpain-2 promote photoreceptor cell death in inherited retinal degeneration

Abstract

Inherited retinal degenerations (IRDs) are a group of untreatable and commonly blinding diseases characterized by progressive photoreceptor loss. IRD pathology has been linked to an excessive activation of cyclic nucleotide-gated channels (CNGC) leading to Na⁺- and Ca²⁺-influx, subsequent activation of voltage-gated Ca²⁺-channels (VGCC), and further Ca²⁺ influx. However, a connection between excessive Ca²⁺ influx and photoreceptor loss has yet to be proven.

Here, we used whole-retina and single-cell RNA-sequencing to compare gene expression between the rd1 mouse model for IRD and wild-type (wt) mice. Differentially expressed genes indicated links to several Ca²⁺-signalling related pathways. To explore these, rd1 and wt organotypic retinal explant cultures were treated with the intracellular Ca²⁺-chelator BAPTA-AM or inhibitors of different Ca²⁺-permeable channels, including CNGC, L-type VGCC, T-type VGCC, Ca²⁺-release-activated channel (CRAC), and Na⁺/Ca²⁺ exchanger (NCX). Moreover, we employed the novel compound NA-184 to selectively inhibit the Ca²⁺-dependent protease calpain-2. Effects on the retinal activity of poly(ADP-ribose) polymerase (PARP), sirtuin-type histone-deacetylase, calpains, as well as on activation of calpain-1, and − 2 were monitored, cell death was assessed via the TUNEL assay.

While rd1 photoreceptor cell death was reduced by BAPTA-AM, Ca²⁺-channel blockers had divergent effects: While inhibition of T-type VGCC and NCX promoted survival, blocking CNGCs and CRACs did not. The treatment-related activity patterns of calpains and PARPs corresponded to the extent of cell death. Remarkably, sirtuin activity and calpain-1 activation were linked to photoreceptor protection, while calpain-2 activity was related to degeneration. In support of this finding, the calpain-2 inhibitor NA-184 protected rd1 photoreceptors.

These results suggest that Ca²⁺ overload in rd1 photoreceptors may be triggered by T-type VGCCs and NCX. High Ca²⁺-levels likely suppress protective activity of calpain-1 and promote retinal degeneration via activation of calpain-2. Overall, our study details the complexity of Ca²⁺-signalling in photoreceptors and emphasizes the importance of targeting degenerative processes specifically to achieve a therapeutic benefit for IRDs.

Video Abstract

Introduction

Inherited retinal degenerations (IRDs) are a large group of genetically heterogeneous, potentially blinding diseases of the retina [1]. The most common form of IRD is retinitis pigmentosa (RP), with a prevalence of approximately 1:4000, affecting more than two million patients worldwide [2]. RP patients commonly experience reduced night vision due to primary degeneration of rod photoreceptors, and subsequently suffer from a progressive visual field constriction known as ‘tunnel vision’ due to a secondary degeneration of cone photoreceptors linked to the preceding rod loss [2]. The second messenger cyclic-guanosine-monophosphate (cGMP) has been identified to play a central role in the pathobiology of many genetically distinct types of IRD [3], and may be directly or indirectly associated with the activity of histone deacetylases (HDACs), poly(ADP-ribose) polymerases (PARPs), cyclic nucleotide-gated channels (CNGCs), and calpain-type proteases [3,4,5].

cGMP- and Ca²⁺-signalling play a crucial role for the phototransduction, the first step of vision, in photoreceptors. cGMP is produced by guanylyl cyclase, the activity of which is inhibited by Ca²⁺ [5]. In darkness, cGMP opens CNGC, causing influx of Na⁺ and Ca²⁺. Together, Ca²⁺ and cGMP form a feedback loop which controls the levels of both second messengers. CNGC-mediated ion influx is countered by the Na⁺/Ca²⁺/K⁺ exchanger (NCKX) and by the ATP-driven Na⁺/K⁺ exchanger (NKX) [5]. As a result, in the dark, a photoreceptor cell is depolarised at approximately − 40 mV [6, 7]. The consequent activation of voltage-gated Ca²⁺-channels (VGCCs) mediates further Ca²⁺ influx and synaptic glutamate release [6, 8]. In light, the enzyme phosphodiesterase-6 (PDE6) rapidly hydrolyses cGMP, leading to CNGC closure, Ca²⁺ decrease, and photoreceptor hyperpolarization. Subsequently, VGCCs close, ending synaptic neurotransmitter release [5]. In IRDs, a mutation-induced cGMP accumulation is likely to over-activate CNGCs and to produce an abnormal influx of Na⁺ and Ca²⁺ into photoreceptors [9].

The rd1 mouse is a naturally occurring IRD model first described by Keeler in the 1920s [10] and has been employed extensively as an animal model of retinitis pigmentosa [11]. It is characterized by early onset, rapid retinal degeneration due to a mutation in the rod-photoreceptor-specific Pde6b gene [12]. This causes dysfunction of cGMP-phosphodiesterase-6 (PDE6) and subsequent accumulation of cGMP in rods. This in turn results in primary rod degeneration, starting at postnatal day 9 (P9), with a peak of cell death at about P13, and with essentially all rods lost by P21. This is followed by secondary cone photoreceptor cell loss occurring between P18 to P60 [13, 14]. The primary rd1 degeneration has often been connected to an excessive Ca²⁺ influx, notably through cGMP-dependent CNGCs [15]. Excessive Ca²⁺ is thought to promote activity of Ca²⁺-dependent calpain-type proteases and rd1 photoreceptor death [16]. Accordingly, many studies over the past decades have focused on the role of Ca²⁺ in IRDs, albeit without conclusive results until today [17]. While some studies indicated that Ca²⁺-channel blockers or genetic ablation of Ca²⁺-permeable channels may preserve photoreceptors [18, 19], other work suggested that inhibition of Ca²⁺-permeable channels either does not reduce retinal cell death or may even promote degeneration [20, 21]. Overall, these conflicting results raise the question whether rd1 rod cell death is linked to Ca²⁺ overload, and, if so, which Ca²⁺-permeable channels might be responsible.

A candidate for a Ca²⁺-permeable channel causing rd1 pathology is VGCC. Functional VGCCs are composed of pore-forming α₁ subunit proteins, encoded by CACNA1x genes, of which there are 10 isoforms in the mammalian genome [22]. In the case of Ca_V1.1–1.4 channels (known as L-type channels), these are encoded by CACNA1S, -C, -D and -F, respectively. The Ca_V2.1–2.3 channels (termed P/Q-, N- and R-type) are encoded by CACNA1A, -B and -E, accordingly. The T-type Ca_V3.1–3.3 channels are encoded by CACNA1G, -H and -I [22, 23]. The accessory α₂δ and β subunits are important for channel folding, subsequent transport to the cell surface, and their integration into specific domains in polarized cells such as neurons. Both Ca_V1 and Ca_V2 classes of channels form a heteromeric complex, co-assembling with one of four β subunits (encoded by CACNB1-4), and one of four α₂δ subunits (encoded by CACNA2D1-4) [22].

Changes in membrane potential triggered by either CNGC or VGCC activity may affect the activity of the Na⁺/Ca²⁺ exchanger (NCX). This is a bi-directional regulator of cytosolic Ca²⁺, which in forward mode transports Ca²⁺ out of cells; however, in reverse mode NCX may import extracellular Ca²⁺ [24]. As NCX does not require ATP for ion transport, the direction of Ca²⁺ movement through the channel depends on the net electrochemical gradients for Na⁺ and Ca²⁺, such that membrane depolarization can augment Ca²⁺ influx [24]. Ion transport by the NCX is electrogenic, with a stoichiometry of three Na⁺ ions exchanged for each Ca²⁺ ion [25]. In mammals, three different NCX genes have been identified: SLC8A1 encoding NCX1, SLC8A2 encoding NCX2, and SLC8A3 encoding NCX3 [26].

Store-operated Ca²⁺ entry (SOCE) is a ubiquitous Ca²⁺ signalling pathway, which is triggered physiologically when the endoplasmic reticulum (ER) Ca²⁺ stores are depleted. In this situation Ca²⁺-levels are replenished through Ca²⁺-release-activated Ca²⁺-channels (CRAC) [27]. SOCE involves a complex choreography between the plasma membrane (PM) protein “Orai” (encoded by ORAI1 and ORAI2) and the ER-resident Ca²⁺-sensing stromal interaction molecules (STIMs, encoded by STIM1 and STIM2). The depletion of ER Ca²⁺ is sensed by STIM1 and its homolog STIM2, causing the opening of Orai channels to drive Ca²⁺ entry into the cell [27].

Excessive Ca²⁺ may activate Ca²⁺-dependent calpain-type proteases [16]. In the calpain family the isoforms calpain-1, calpain-2, and calpain-5 have been connected to neurodegenerative diseases [28,29,30]. The best-characterized calpain isoforms are calpain-1 and calpain-2 [30, 31], where the former is often linked to processes like synaptic plasticity and long-term potentiation, important for learning and memory [32]. Calpain-2 may have a more prominent role in cytoskeletal remodelling and cell motility [33]. Recent studies have suggested that calpain-1 and calpain-2 may play opposing roles, where calpain-1 may be neuroprotective while calpain-2 may be causing neurodegeneration [34].

To try and settle the long-standing controversy on the role of Ca²⁺ in IRD pathogenesis, we performed an initial bio-informatic analysis of differentially expressed genes (DEGs) in rd1 mouse retina based on both whole-retina RNA sequencing (RNA-seq) and single-cell RNA sequencing (scRNA-seq) datasets. We found numerous DEGs to be enriched in Ca²⁺ associated pathways, arguing for an important function of Ca²⁺-signalling. We then used Ca²⁺ chelation and Ca²⁺-permeable channel inhibition in organotypic rd1 retinal explant cultures to identify the possible sources of high photoreceptor Ca²⁺. To assess the role of Ca²⁺-dependent proteolysis, we also studied the overall activity of calpains, and then focussed on calpain-1 and − 2. We found that calpain-2 contributed to photoreceptor cell death, a finding confirmed by the protective effect of the selective calpain-2 inhibitor NA-184. Altogether, we show that Ca²⁺ contributes to rd1 photoreceptor cell death, regulating, among other things, the activity of PARPs and sirtuins, and that Ca²⁺ overload and photoreceptor degeneration is likely caused by T-type VGCC and NCX. Surprisingly, inhibition of either CNGC or CRAC did accelerate retinal degeneration.

Materials and methods

Animals

For retinal explant cultures C3H/HeA Pde6b^rd1/rd1animals (rd1), their congenic wild-type C3H/HeA Pde6b^+/+ counterparts (wt) [35], and B6.129SvJ;C3H/HeA-CNGB1^tm double-mutant mice (rd1*Cngb1^−/−) were used [19]. The rd1*Cngb1^−/− double mutants were generated by an intercross of rd1 and Cngb1^−/− [19]. Animals were used regardless of gender. The stock has been maintained by repeated backcrossing over ten generations to make a congenic inbred strain, homozygous for both gene mutations.

In rd1 mouse retina, rod photoreceptors start to degenerate at post-natal day (P) 9 with the peak of cell death occurring at P13 and the rod degeneration essentially complete by P21 [36]. The last time-point at which the overall retinal morphology and cellular composition of the retina is still comparable between rd1 and wt retina is at P11. In rd1*Cngb1^−/− double mutants, the degeneration progresses significantly slower, with the onset of rod death at around P12, the peak of cell death at P18 and the end of photoreceptor degeneration at around two months post-natal [19].

Animals were housed under standard white cyclic lighting and had free access to food and water. Animal protocols compliant with § 4 of the German law of animal protection were reviewed and approved by the competent authority (Einrichtung für Tierschutz, Tierärztlicher Dienst und Labortierkunde, Registration No. AK02/19M, AK01/20M AK05/22M).

Analysis of RNA-Seq, scRNA-seq data and differential expression

The mRNA expression comparison between rd1 and wt mouse employed datasets downloaded from the GSE62020 database [37], while single-cell RNA expression analysis used data from the GSE212183 database [38]. To characterize rd1, we performed a differential analysis (fold change > 1.2, p < 0.05) [39, 40] comparing rd1 to wt using the “limma” package of R language. Heatmaps and volcano plots used to visualize Ca²⁺-related genes were created using the packages “pheatmap” and “ggplot2”. To infer functional annotations of rd1 genes, gene ontology (GO) [m5.go.v2022.1.Mm.symbols.gmt] of differentially expressed genes (DEGs) was supplemented by gene set enrichment analysis (GSEA; version 4.2.2). The statistical significance was defined as false discovery rate (FDR) < 0.05, and the overrepresentation of indicated GO gene sets in the ranked gene lists were presented by the normalized enrichment score (NES). GO enrichment analyses were conducted for the selected common DEGs using “GOplot” and “enrichplot” packages, p < 0.05. Additionally, the software packages “ggpubr”, “corrplot”, “fmsb” and “ggalluvial” were used to generate box plot, balloon plot, deviation plot, correlation plot, radar plot, and alluvial diagram, respectively.

Retinal explant culture

To study the effects of various drugs on photoreceptor enzyme activities and cell death, we used organotypic retinal explant cultures derived from wt, rd1, and rd1*Cngb1^−/− animals. Mouse retinal explants can be maintained under defined conditions for culturing periods of at least four weeks, with only minor cell loss induced by the culture situation [41, 42]. Retinas were explanted at postnatal day 5 (P5). The explants were cultured on a polycarbonate membrane (Order No.: 83.3930.040; 0.4 μm TC-inserts, SARSTEDT, Hildesheim, Germany) with complete medium (Gibco R16 medium with supplements) [43]. The two retinas obtained from a single animal were split across different experimental groups to maximize the number of independent observations acquired per animal. After explantation, the complete R16 medium was changed every two days along with the pharmaceutical compounds including 10 µM BAPTA-AM [44] (ab120503; Abcam, Cambridge, UK), 50µM L-cis-diltiazem [45] (ab120532; Abcam), 20 µM CM4620 [46] (HY-101942; MedChemExpress, Sollentuna, Sweden), 40 µM SN-6 [47] (HY-107658; MedChemExpress), 100 µM D-cis-diltiazem [45] (ab120260; Abcam), 10 µM TTA-A2 [48] (HY-111828; MedChemExpress), 15 µM DS5565 [49] (HY-108006; MedChemExpress), and 1 µM NA-184 [50] (kindly provided by Michel Baudry, Western University, Pomona, CA, USA), respectively. The dosing of compounds was chosen based on the published IC₅₀ values (see references given behind each compound) and the retinal dose-responses curves collected in Figure S2. All compounds were dissolved in DMSO at a final medium concentration of no more than 0.1% DMSO. Cultures were ended at P11 (rd1 short-term cultures), P23 (rd1 long-term treatment) and P17 (rd1*Cngb1^−/−) by either fixation with 4% paraformaldehyde (PFA) or without fixation and direct freezing in liquid N₂. Explants were embedded in Tissue-Tek (Sakura Finetek Europe B.V., Alphen aan den Rijn, The Netherlands) and sectioned (14 μm) in a cryostat (Thermo Fisher Scientific, CryoStar NX50 OVP, Runcorn, UK).

TUNEL staining

TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labelling) assay kit (Roche Diagnostics, Mannheim, Germany) was used to label dying cells. Histological sections from retinal explants were dried and stored at − 20 °C. The sections were rehydrated with phosphate-buffered saline (PBS; 0.1 M) and incubated with proteinase K (1.5 µg/µL) diluted in 50 mM TRIS-buffered saline (TBS; 1 µL enzyme in 1 mL TBS) for 15 min. This was followed by 3 times 5 min TBS washing and incubation with blocking solution (10% normal goat serum, 1% bovine serum albumin, and 1% fish gelatine in phosphate-buffered saline with 0.03% Tween-20). TUNEL staining solution was prepared using 21 parts of blocking solution, 18 parts of TUNEL labelling solution, and 1 part of TUNEL enzyme. After blocking, the sections were incubated with TUNEL staining solution overnight at 4 °C. Finally, sections were washed 2 times with PBS, mounted using mounting medium with DAPI (ab104139; Abcam), and imaged by microscopy.

Calpain activity assay

This assay allows resolving the overall calpain activity in situ on unfixed tissue sections. Retinal tissue sections were incubated and rehydrated for 15 min in a calpain reaction buffer (CRB) (25 mM HEPES, 65 mM KCl, 2 mM MgCl₂, and 1.5 mM CaCl₂ in ddH₂O; pH 7.2) with 2 mM dithiothreitol (DTT). Tissue sections were incubated for 3 h at 37 °C in CRB with tBOC-Leu-Met-CMAC (25 µM; A6520; Thermo Fisher Scientific, OR, USA). Then, sections were washed with PBS and incubated with ToPro (1:1000 in PBS, Thermo Fisher Scientific) for 15 min. Afterwards, tissue sections were washed twice in PBS (5 min) and mounted using Vectashield without DAPI (Vector Laboratories Inc., Burlingame, CA, USA) for immediate visualization by microscopy.

PARP activity and PAR staining

The PARP in situ activity assay is based on the incorporation of a fluorescent NAD⁺ analogue and allows resolving the overall PARP enzyme activity on unfixed tissue sections [51]. Such sections were incubated and rehydrated for 10 min in PBS. The reaction mixture (10 mM MgCl₂, 1mM dithiothreitol, and 50 µM 6-Fluo-10-NAD⁺ (Cat. Nr.: N 023; Biolog, Bremen Germany) in 100 mM Tris buffer with 0.2% Triton X100, pH 8.0) was applied to the sections for 3 h at 37 °C. After three 5 min washes in PBS, sections were mounted in Vectashield with DAPI (Vector Laboratories) for subsequent microscopy.

For the detection of PAR, we used an immunostaining enhanced with 3,3′-diaminobenzidine (DAB) staining. The procedure is initiated by quenching of endogenous peroxidase activity using 40% MeOH and 10% H₂O₂ in PBS with 0.3% Triton X-100 (PBST) in retinal tissue sections for 20 min. Sections were further incubated with 10% normal goat serum (NGS) in PBST for 30 min, followed by anti-PAR antibody (1:200; ALX-804-220-R100; Enzo Life Sciences, Farmingdale, NY, USA) incubation overnight at 4 °C. Incubation with the biotinylated secondary antibody (1:150, Vector in 5% NGS in PBST) for 1 h was followed by the Vector ABC Kit (Vector Laboratories, solution A and solution B in PBS, 1:150 each) for 1 h. DAB staining solution (0.05 mg/mL NH₄Cl, 200 mg/mL glucose, 0.8 mg/mL nickel ammonium sulphate, 1 mg/mL DAB, and 0.1 vol. % glucose oxidase in phosphate buffer) was applied evenly, incubated for precisely 3 min, and immediately rinsed with phosphate buffer to stop the reaction. Sections were mounted in Aquatex (Merck, Darmstadt, Germany).

Sirtuin/HDAC activity assay

This assay allows detecting overall HDAC activity in situ on fixed tissue sections and is based on an adaptation of the FLUOR DE LYS®-Green System (Biomol, Hamburg, Germany). Retinal sections were exposed to 50 µM FLUOR DE LYS®-SIRT1 deacetylase substrate (BML-Kl177-0005; ENZO, New York, USA) with 2 mM NAD⁺ (BML-KI282-0500; ENZO) in assay buffer (50 mM Tris/HCl, 137 mM NaCl; 2.7 mM KCl; 1mM MgCl2; pH 8.0) for 3 h at 37 °C. Sections were then washed in PBS and fixed in methanol at -20 °C for 20 min. Slides were mounted with FLUOR DE LYS® developer II concentrate (BML-KI176-1250; Enzo, New York, USA) diluted 1:5 in assay buffer overnight for subsequent microscopy.

Fig. 1

Whole tissue RNA-seq analysis highlights Ca²⁺-signalling-related changes in rd1 retina. A Heatmap for post-natal day 13 RNA-Seq data comparing the expression of Ca²⁺-related differentially expressed genes (DEGs) between rd1 and wild-type (wt) retina. In this group of genes, 32 were down- and 57 were up-regulated in rd1 retina. Color-coding indicates the logFC values (positive values in red, negative values in blue). B Gene set enrichment analysis (GSEA) showing upregulation of four Ca²⁺-related gene ontology (GO) pathways in rd1 retina (false discovery rate, FDR < 0.05). C Circle plot showing 10 different Ca²⁺-related GO terms (including biological processes (BPs) and molecular functions (MFs)) enriched in DEGs in the rd1 situation (p < 0.05). Color-coding of circles indicates the logFC values (positive values in red, negative values in blue); numbers in trapezoids indicate differentially expressed genes (red: upregulated; blue: downregulated); inner circle shows z-scores (increasing scores in red, decreasing scores in blue)

Immunohistochemistry for calpain-1/2 and NCXs

Sections were rehydrated with PBS for 15 min and then incubated with a blocking solution (10% NGS, 1% BSA, and 0.3% PBST) for 1 h. The primary antibodies, rabbit-anti-calpain-2 (1:200; ab39165; Abcam), rabbit-anti-calpain-1 (1:100; ab39170; Abcam), rabbit-anti-NCX1 (1:200; LS-B15461; LifeSpan BioScience, Washington, USA), rabbit-anti-NCX2 (1:200; BS-1997R; BIOSS; Massachusetts, USA) and mouse-anti-NCX3 (1:100; NB120-2869; Novus Biologicals, Colorado, USA) were diluted in blocking solution and incubated overnight at 4 °C rinsing with PBS for 3 times 10 min each; this was followed by incubation with the secondary antibodies, goat-anti-rabbit AlexaFluor488 (1:400; A11034; Molecular Probes; Oregon, USA), goat-anti-rabbit AlexaFluor568 (1:300; A11036; Molecular Probes), and goat-anti-mouse AlexaFluor568 (1:500; A11031; Molecular Probes), for 1 h. The sections were further rinsed with PBS for 3 times 10 min each and mounted with mounting medium with DAPI (Abcam).

Microscopy and image analysis in retinal cultures

Images of organotypic explant cultures were captured using a Zeiss Imager Z.2 fluorescence microscope, equipped with ApoTome 2, an Axiocam 506 mono camera, and HXP-120V fluorescent lamp (Carl Zeiss Microscopy, Oberkochen, Germany). Excitation (λExc.)/emission (λEm.) characteristics of filter sets used for different fluorophores were as follows (in nm): DAPI (λExc. = 369 nm, λEm = 465 nm), AF488 (λExc. = 490 nm, λEm = 525 nm), AF568 (λExc. = 578 nm, λEm = 602 nm), and ToPro (λExc. = 642 nm, λEm = 661 nm). The Zen 2.3 blue edition software (Zeiss) captured images (tiled and z-stack, 20× magnification). Sections of 14 μm thickness were analysed using 15 Apotome Z-planes.

For quantification of positive cells in ONL, we proceeded as follows: The number of cells in six different rectangular ONL areas was counted manually based on the number of DAPI-stained nuclei and used to calculate an average ONL cell size. This average ONL cell size was used to calculate the total number of cells in a given ONL area. The percentage of positive cells was then calculated by dividing the absolute number of positive cells by the total number of ONL cells. ONL thickness was determined by manual counts on DAPI stained retinal sections. For each count, a vertical column was placed on nine different positions of the section and the end-to-end distances of the photoreceptor layer in each of these positions were recorded manually. The nine individual counts were averaged to give the mean value of photoreceptor thickness for one explant.

Statistical analysis and figure preparation

Two-way comparisons were analysed using Student’s t-test. Multiple comparisons were made using a one-way analysis of variance (ANOVA) test with Tukey multiple comparison post-hoc test. Calculations were performed with GraphPad Prism 8 (GraphPad Software, La Jolla, CA, USA). Levels of significance were as follows: *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. Data in Fig. 8A was normalized by linear scaling according to the formula: χ_scaled = χ-χ_min/χ_max-χ_min, using SPSS Statistics 26 (IBM, Armonk, New York, USA). The figures were prepared using Photoshop 2022 and Illustrator 2022 (Adobe, San Jose, CA, USA). Bioinformatic analyses were performed by R software (Version 4.0.1). Figure 9 was created using BioRender.com.

Results

Ca²⁺-related genes and pathways during the photoreceptor degeneration of rd1

To investigate whether and to what extent Ca²⁺ was involved in the progression of rd1 retinal cell death, we initially performed RNA-seq to assess gene expression differences between rd1 and wt mice at post-natal day (P) 13. The GSE62020 database [37] (accessed June 2023) was analysed to identify differentially expressed genes (DEGs). A fold change > 1.2 and a p-value < 0.05 were considered to indicate significant changes [39, 40]. Compared to wt, rd1 retina expressed 666 up-regulated genes and – coincidentally – 667 down-regulated genes at P13 (Table S1). Among these, 89 DEGs were found to be associated with Ca²⁺-signalling, of which 32 were down-regulated and 57 were up-regulated in rd1 retina (Fig. 1A). Gene set enrichment analysis (GSEA) revealed four gene ontology (GO) terms involved in Ca²⁺-signalling (Fig. 1B, FDR < 0.05). Ten Ca²⁺-related biology processes (BPs) and molecular functions (MFs) were identified by DEGs GO enrichment analysis (Fig. 1C, p < 0.05).

Ca²⁺ plays an important role during photoreceptor degeneration

To further investigate Ca²⁺-related pathways specifically in degenerating rd1 photoreceptors, we performed single-cell (sc) RNA-seq analysis to assess transcriptional differences between the rd1 and wt situation. A visualization in the form of a volcano plot showed the genes up- or down-regulated in rod photoreceptors from GSE212183 [38]. In total, 1233 DEGs were found to be differentially expressed between rd1 and wt rod photoreceptors at P13, where 655 DEGs were up-regulated and 578 were down-regulated (Figure S1A, Table S2). 74 Ca²⁺-related genes in rod photoreceptors are presented, of which 19 were down-regulated and 55 were up-regulated (Fig. 2A). These rod DEGs were linked to 19 Ca²⁺-related GO terms (Table S4), among which the top-5 GO terms for rod photoreceptors were selected (Fig. 2B). There were 779 DEGs found in rd1 cone photoreceptors at P13 (Figure S1B, Table S3). 38 Ca²⁺-related genes were differentially expressed in cone photoreceptors, of which 11 were down-regulated and 27 were up-regulated (Fig. 2C). These cone DEGs were enriched in 6 Ca²⁺-related biological processes (BP) (Table S5), the top-5 BP are shown (Fig. 2D).

Fig. 2

Photoreceptor scRNA-Seq and Ca²⁺ chelation reveal critical role for Ca²⁺-signalling in rd1 cell death. A Volcano plot for scRNA-Seq data showing Ca²⁺-related differentially expressed genes (DEGs) in rd1 vs. wild-type (wt) rod photoreceptors. B Circle plot showing top-5 Ca²⁺-related GO terms enriched for Ca²⁺-related DEGs in rd1 rod photoreceptors (p < 0.05). C Volcano plot for scRNA-Seq data showing Ca²⁺-related differentially expressed genes (DEGs) in rd1 vs. wild-type (wt) cone photoreceptors. D Circle plot showing top-5 Ca²⁺-related GO terms enriched for Ca²⁺-related DEGs in rd1 cone photoreceptors (p < 0.05). E The TUNEL assay labelled dying cells (magenta) in wt and rd1 retinal explant cultures. DAPI (grey) was used as nuclear counterstain. Untreated (Untr.) wt and rd1 retina were compared to retina treated with 10 µM BAPTA-AM (BAPTA). Scatter plot displaying percentage of TUNEL-positive cells in the outer nuclear layer (ONL). Statistical testing: Student’s t-test performed between rd1 Untr. and 10 µM BAPTA-AM (BAPTA). Untr. wt: n = 4; Untr. rd1: 10; BAPTA rd1: 19; error bars represent SD; *** = p < 0.001. INL = inner nuclear layer, GCL = ganglion cell layer; scale bar = 50 µm

In the following, we used the Ca²⁺ chelator BAPTA-AM in rd1 organotypic retinal explants, with the aim of confirming the link between Ca²⁺ and photoreceptor degeneration. The TUNEL assay was used to quantify the numbers of dying cells in the outer nuclear layer (ONL). A dose-response for BAPTA-AM treatment on rd1 explants was established and revealed 10 µM as a suitable concentration for further experiments (Figure S2A). In wt retinal explants, a relatively low number of ONL cells (1.08% ± 0.21, n = 4) were positive for the TUNEL assay, when compared with their rd1 counterparts (4.84% ± 0.73, n = 10). BAPTA-AM treatment significantly reduced rd1 ONL cell death to 3.34% (± 0.92, n = 19, p < 0.001; Fig. 2E). While BAPTA-AM treatment appeared to be well tolerated in wt retina (Figure S1D), it did not preserve photoreceptor viability in long-term treatment of rd1 retina lasting until P23 (Figure S3A). In contrast, BAPTA-AM accelerated photoreceptor degeneration in explant cultures derived from rd1*Cngb1^−/− double-mutant mice (Figure S1C), indicating that depletion of intracellular Ca²⁺ in photoreceptors lacking functional CNGC was detrimental.

Expression of Ca²⁺ permeable channels in retina and rod photoreceptors

A whole range of Ca²⁺-permeable channels may potentially contribute to increased intracellular Ca²⁺-levels in photoreceptors. Notably, the second messenger cGMP activates photoreceptor CNGC, leading to Na⁺ and Ca²⁺ influx and, indirectly, via ensuing changes in membrane polarization, to additional Ca²⁺-influx through VGCC [5]. High intracellular Na⁺ levels produced by CNGC may reverse the directionality of NCX [52], possibly resulting in Ca²⁺-influx instead of efflux. Moreover, high Ca²⁺, via the activation of calmodulin-dependent protein kinase 2 (CaMK2), may potentiate Ca²⁺-influx mediated by CRAC [53].

To assess the expression of these four groups of Ca²⁺-permeable channels, we used RNA-seq and scRNA-seq to screen for any possible expression changes connected to rd1 photoreceptor degeneration. While in whole retina RNA-seq numerous changes were found in genes encoding for CNGC, VGCC, CRAC, and NCX isoforms, the most prominent expression changes appeared after P18, i.e. at a time when most rd1 rod photoreceptors have already been lost (Fig. 3A). At P13, RNA-seq data indicated that genes coding for VGCC were up-regulated while genes coding for CNGC were down-regulated (Fig. 3B). To pinpoint rd1 photoreceptor specific gene expression changes in Ca²⁺-permeable channels, we used scRNA-seq to investigate the critical P11 to P17 time-frame (Fig. 3C; cf. Figure S3B for a corresponding scRNA-seq analysis for cone photoreceptors). At the peak of photoreceptor degeneration, at P13, we found an upregulation of genes coding for CNGC, VGCC, and NCX (Fig. 3D), while CRAC genes appeared to be down-regulated. However, we also found an upregulation in the expression of the Camk2g gene, where CaMK2 may increase CRAC activity independent of CRAC gene expression changes (Figure S3C).

Fig. 3

Inhibition of differentially expressed Ca²⁺-channels in rd1 retina either promote or reduce photoreceptor cell death. A Balloon plot showing time-dependent whole-retina expression changes (post-natal day (P) 7 to P21) of cyclic nucleotide-gated channel (CNGC), Ca²⁺-release activated channel (CRAC), Na⁺/Ca²⁺ exchanger (NCX), and voltage-gated Ca²⁺ channel (VGCC). B Deviation plot highlighting expression changes for CNGC, CRAC, NCX, and VGCC in P13 rd1 whole retina. C Balloon plot showing scRNA-seq data and time-dependent expression changes (post-natal day (P) 11 to P17) of CNGC, CRAC, NCX, and VGCC in rd1 rod photoreceptors. D Deviation plot highlighting expression changes for CNGC, CRAC, NCX, and VGCC in P13 rd1 rod photoreceptors. E TUNEL assay labelling dying cells (magenta) in rd1 and wild-type (wt) retinal explant cultures. DAPI (grey) was used as a nuclear counterstain. Untreated (Untr.) wt and rd1 retina were compared to retina treated with 50 µM CNGC inhibitor (L-cis diltiazem), 20 µM CRAC inhibitor (CM4620), 40 µM NCX inhibitor (SN-6), 100 µM L-type VGCC inhibitor (D-cis diltiazem), 10 µM T-type VGCC inhibitor (TTA-A2), and 15 µM α₂δ subunit VGCC ligand (DS5565). Scatter plot showing percent TUNEL-positive cells in outer nuclear layer (ONL). Dashed line indicates rd1 untr. situation, data points below this threshold indicate protective effects, data points above suggest destructive effects. Statistical testing: one-way ANOVA and Tukey’s multiple comparison post hoc test. Untr. wt: n = 5; Untr. rd1: 18; L-cis rd1: 6; CM4620 rd1: 15; SN-6 rd1: 9; D-cis rd1: 12; TTA-A2 rd1: 6; DS5565 rd1: 6; error bars represent SD; significance levels: *** = p < 0.001; **** = p < 0.0001. INL = inner nuclear layer, GCL = ganglion cell layer; scale bar = 50 µm

Overall, the RNA-seq and scRNA-seq data suggested important changes in the expression of Ca²⁺-permeable channel-related genes in rd1 photoreceptor degeneration. Still, from the RNA expression data alone it was not possible to unambiguously deduce which type of Ca²⁺-permeable channel was causally involved in rd1 cell death.

Inhibition of Ca²⁺ permeable channels impact rod photoreceptor viability

To characterize the role of different Ca²⁺-permeable channels functionally, we employed an array of different channel inhibitors, including the CNGC inhibitor L-cis-diltiazem (L-cis), the CRAC inhibitor CM4620, the NCX inhibitor SN-6, the L-type VGCC inhibitor D-cis-diltiazem (D-cis), the T-type VGCC inhibitor TTA-A2, and the α₂δ subunit VGCC ligand DS5565. Retinal explant cultures derived from rd1 and wt animals were treated with these inhibitors and the effects analysed using the TUNEL assay to detect dying cells in the ONL. To assess a possible cross-reactivity of the inhibitors used, we also employed explant cultures derived from rd1*Cngb1^−/− double-mutant mice, i.e. retinas in which rods lack functional CNGC. For L-cis and D-cis suitable drug concentrations for retinal treatments had been established previously [45]; for inhibitors where such data were missing (i.e. SN-6, TTA-A2, DS5565, CM4620), dose-response curves were generated to select appropriate treatment concentrations (Figures S2).

As shown previously [16, 54], in wt retina the number of TUNEL positive cells in the ONL was significantly lower than in its rd1 counterpart (Fig. 3E, Table S6). Remarkably, treatment with both L-cis and CM4620 significantly increased the numbers of TUNEL positive cells in both rd1 and wt ONL (Fig. 3E, Figure S1D, Table S6). In other words, L-cis and CM4620 exhibited clear photoreceptor toxicity that was independent of the rd1 mutation. In contrast, SN-6 and TTA-A2 significantly decreased photoreceptor cell death in rd1 retina (Fig. 3E, Table S6). Also, in the retina of rd1*Cngb1^−/− double-mutant mice SN-6 significantly decreased TUNEL positive cells (Figure S1C), yet, in wt retina ONL cell death was significantly increased by SN-6 (Figure S1D). Like BAPTA-AM, SN-6 did not show long-term protection in rd1 ONL (Figure S3A). To assess the expression of NCX at the protein level, immunostaining was performed using antibodies directed against NCX1, NCX2, and NCX3. When compared to negative control, only NCX1 was found to be expressed in rd1 and wt retina, notably in photoreceptor segments and Müller glial cells (Figure S3D).

The compounds D-cis and DS5565 targeting L-type and α₂δ VGCC did not lead to significant reduction of TUNEL positive cells in rd1 ONL (Fig. 3E, Table S6).

Taken together, the experiments with Ca²⁺-permeable channel blockers confirmed that Ca²⁺-signalling was indeed important for photoreceptor viability. Contrary to expectations, activity of CNGC and CRAC appeared to have a pro-survival role, while activity of T-type VGCC and NCX was detrimental for rd1 rod photoreceptors. The role of NCX appeared ambiguous, as its inhibition improved survival of rd1 photoreceptors but promoted death in wt ones, suggesting that NCX directionality might have been reversed in the rd1 situation. Finally, L-type VGCC activity appeared to be unrelated to rd1 degeneration.

Calpain-2 contributes to rd1 photoreceptor degeneration

The reduction of TUNEL-positive cells after Ca²⁺ chelation and Ca²⁺ permeable channel inhibition strongly pointed to a link between high intracellular Ca²⁺-levels and rd1 photoreceptor cell death. To further investigate Ca²⁺-induced cell death, we performed GSEA analysis on rd1 P13 RNA-seq data, in which DEGs were enriched in the GO biological process (BP) pathways, including “Positive regulation of proteolysis” (GO: 0045862; normalized enrichment score (NES) = 1.78, p < 0.0001; Fig. 4A), “Membrane protein proteolysis” (GO: 0033619; NES = 1.62, p = 0.003; Fig. 4B), “Membrane protein ectodomain proteolysis” (GO: 0006509; NES = 1.6, p = 0.027; Fig. 4C), and “Positive regulation of proteolysis involved in protein catabolic process” (GO: 1903052; NES = 1.6, p < 0.0001; Fig. 4D). Previous studies had connected the activity of Ca²⁺-dependent calpain-type proteases to rd1 retinal degeneration [55, 56] and we hypothesized that the neurodegenerative calpain-2 isoform [34] might be responsible for retinal cell death. Thus, we used the recently developed and highly specific calpain-2 inhibitor NA-184 [57] to treat rd1 explant cultures. A dose-response curve for the effects of NA-184 was compiled (Figure S2F) and a concentration of 1 µM was chosen for further experiments. As before, in wt retina the numbers of TUNEL positive cells in the ONL were low (1.08% ± 0.44, n = 4) when compared with untreated rd1 retina (4.8% ± 0.47, n = 7; Fig. 4E). NA-184 treatment significantly reduced rd1 ONL TUNEL positive cells (3.55% ± 0.53, n = 7, p < 0.01; Fig. 4E), implying a causative involvement of calpain-2 in the degenerative process.

Fig. 4

Ca²⁺-dependent proteolysis and calpain-2 are associated with rd1 photoreceptor cell death. A Gene set enrichment analysis (GSEA) showing enrichment of differentially expressed genes (DEGs) in the biological process (BP) “Positive regulation of proteolysis” (GO: 0045862). B DEGs enriched in the BP “Membrane protein proteolysis” (GO: 0033619). C DEGs enriched in the BP “Membrane protein ectodomain proteolysis” (GO: 0006509). D DEGs enriched in the BP "Positive regulation of proteolysis involved in protein catabolic process” (GO:1903052). E TUNEL assay labelled dying cells (magenta) in wild-type (wt) and rd1 retinal explant cultures. DAPI (grey) was used as nuclear counterstain. Untreated (Untr.) wt and rd1 retina compared to retina treated with the calpain-2 selective inhibitor NA-184. Scatter plot shows percentage of TUNEL-positive cells in outer nuclear layer (ONL). Statistical comparison: Student’s t-test performed between rd1 Untr. and rd1 NA-184. INL = inner nuclear layer, GCL = ganglion cell layer; scale bar = 50 µm

Calpain activity changes after interventions targeting intracellular Ca²⁺

To further study how Ca²⁺ chelation, activity of Ca²⁺ permeable channels, and calpain-2 inhibition regulated overall calpain, we investigated calpain activity using a general in situ activity assay and immunolabelling for activated calpain-1 and calpain-2. Calpain activity and calpain-2 activation were rather low in wt retina when compared with rd1 retina (Fig. 5A, C; Table S7A, C). BAPTA-AM significantly reduced both overall calpain activity and calpain-2 activation specifically (Fig. 5A, C; Table S7A, C; dose-response curve shown in Figure S2A). However, L-cis and CM4620 significantly increased both calpain activity and calpain-2 activation (Fig. 5A, C; Table S7A, C; dose response curve for CM4620 shown in Figure S2B). General calpain activity and calpain-2 activation were also significantly reduced after treatment with SN-6, D-cis, TTA-A2, and NA-184 (Fig. 5A, C; Table S7A, C). Dose-response curves for SN-6, TTA-A2, and NA-184 on rd1 calpain activity are shown in Figure S2C, D, and F, respectively. Interestingly, DS5565 significantly reduced overall calpain activity, while it did not decrease calpain-2 activation, when compared to rd1 control (Fig. 5A, C; Table S7A, C; dose response curve for DS5565 shown in Figure S2E).

Fig. 5

Ca²⁺-channel inhibitors differentially regulate general calpain activity, activation of calpain-1 and -2. A Calpain activity assay (blue) in wild-type (wt) and rd1 retinal explant cultures, nuclear counterstain with ToPro (red). Untreated (Untr.) wt and rd1 retina were compared to retina treated with BAPTA-AM, L-cis diltiazem, CM4620, SN-6, D-cis diltiazem, TTA-AS, DS5565, and NA-184. Scatter plot showing percent calpain activity positive cells in outer nuclear layer (ONL). Untr. wt: n = 11; Untr. rd1: 23; BAPTA rd1: 10; L-cis rd1: 6; CM4620 rd1: 5; SN-6 rd1: 9; D-cis rd1: 6; TTA-A2 rd1: 7; DS5565 rd1: 6; NA-184 rd1: 8. B Activated calpain-1 (cyan) immunostaining in wt and rd1 retinal explant cultures with DAPI (grey) as nuclear counterstain. Untreated specimens were compared to retina treated with compounds as in A. Scatter plot showing percent ONL cells displaying calpain-1 activation. Untr. wt: n = 7; Untr. rd1: 17; BAPTA rd1: 9; L-cis rd1: 9; CM4620 rd1: 10; SN-6 rd1: 8; D-cis rd1: 10; TTA-A2 rd1: 8; DS5565 rd1: 8; NA-184 rd1: 9. C Activated calpain-2 (yellow) immunostaining in rd1 and wt retinal explant cultures with DAPI (grey) as nuclear counterstain. Untreated specimens were compared to retina treated with compounds as in A. Scatter plot showing percent ONL cells displaying calpain-2 activation. Untr. wt: 10; Untr. rd1: 21; BAPTA rd1: 9; L-cis rd1: 9; CM4620 rd1: 10; SN-6 rd1: 7; D-cis rd1: 10; TTA-A2 rd1: 8; DS5565 rd1: 6; NA-184 rd1: 9. Note the significant elevation of calpain activity/calpain-2 activation caused by CNGC and CRAC inhibition. The opposite is observed for calpain-1 activation. Statistical testing: one-way ANOVA and Tukey’s multiple comparison post hoc test; significance levels: * = p < 0.05; ** = p < 0.01; *** = p < 0.001; **** = p < 0.0001; error bars represent SD; INL = inner nuclear layer, GCL = ganglion cell layer; scale bar = 50 µm

In wt retina the numbers of calpain-1 activation in ONL were relatively similar when compared with that of rd1 (p > 0.05; Fig. 5B and Table S7B). BAPTA-AM significantly increased the percentage of activated calpain-1 in rd1 ONL, while L-cis did not (Fig. 5B; Table S7B). The number of activated calpain-1 positive cells in rd1 ONL was significantly raised by CM4620 and SN-6, but not by D-cis, TTA-A2, and DS5565 (Fig. 5B; Table S7B). Interestingly, the calpain-2 inhibitor NA-184 significantly increased calpain-1 activation in rd1 retina (Fig. 5B; Table S7B).

Relationships between activation of calpain-1, calpain-2, and cell death

To capture the presumed interactions between calpain-1, calpain-2, and cell death numerically, we applied Spearman’s rank correlation coefficient analysis (Spearman analysis) to the corresponding datasets to assess the strength and direction of the monotonic relationship between variables [58]. Figure 6A illustrates that calpain-1 activation in rd1 ONL correlated negatively with calpain-2 activation with R = -0.62 (p < 0.01), and similarly with cell death indicated by TUNEL positive cells (R = -0.65, p < 0.0001, Fig. 6B). Further, our hypothesis that calpain-1 activation may not be connected to cell death, while calpain-2 activation is, was supported by the fact that the ratio of calpain-2/calpain-1 was positively correlated with TUNEL (R = 0.61, p < 0.01, Fig. 6C), a similar association as would be expected for the effect of calpain-2 alone.

Fig. 6

Differential correlation of calpain-1 and calpain-2 to photoreceptor cell death. A Spearman analysis comparing the numbers of activated calpain-1 and calpain-2 in the outer nuclear layer (ONL) of untreated rd1 retina and rd1 retina treated with the calpain-2 inhibitor NA-184. B Spearman analysis between the numbers of activated calpain-1 and dying, TUNEL positive cells in untreated or NA-184 treated rd1 ONL. C Spearman analysis between the ratio of calpain-2/calpain-1 and TUNEL positive cells in untreated or NA-184 treated rd1 ONL. D Alluvial diagram showing the relationship between activated calpain-2 and calpain-1, as well as TUNEL positive cells in rd1 ONL after treatment with BAPTA-AM, L-cis, CM4620, SN-6, D-cis, TTA-AS, DS5565, and NA-184

General trends in the relationships between calpain-1, calpain-2, and cell death were visualized in an alluvial diagram where the numbers of activated calpain-1, -2, and TUNEL positive cells in the ONL of the rd1 control group were used as baseline (Fig. 6D). This representation suggests that in rd1 retinal explant cultures treatment with BAPTA-AM, SN-6, D-cis, TTA-A2, DS5565, and NA-184 led to low calpain-2 activation in the ONL when compared to untreated (Fig. 6D). In contrast, L-cis and CM4620 treatments were related to high calpain-2 activation. Most of the low calpain-2 activation was linked to high calpain-1 activation and to a reduction of TUNEL positive cells in ONL (Fig. 6D). Taken together, the above data are in support of our hypothesis that calpain-1 activation was unrelated to cell death while calpain-2 activation was closely connected to it.

Effect of Ca²⁺ and calpain-2 on enzymatic activities of PARP and sirtuin

PARP was previously reported to be connected to photoreceptor cell death [4]. To dissect the relationship of Ca²⁺, calpain, and PARP, we performed an in situ PARP activity assay and immunostaining for PAR, i.e. the product of PARP activity. While PARP activity and PAR-positive cells were infrequent in wt retina when compared with rd1 (Fig. 7A, B; Table S8A, B), treatment with BAPTA-AM significantly reduced both PARP activity and PAR generation in rd1 ONL (Fig. 7A, B; Table S8A, B; BAPTA-AM dose-response in Figure S2A), indicating a Ca²⁺-dependent mechanism for PARP activation. However, the Ca²⁺-channel blockers L-cis and CM4620 significantly increased both PARP activity and PAR in rd1 ONL (Fig. 7A, B; Table S8A, B; CM4620 dose-response in Figure S2B). In contrast, PARP activity and PAR positive cells in rd1 ONL were significantly reduced after treatment with SN-6, D-cis, TTA-A2, and DS5565 (Fig. 7A, B; Table S8A, B; dose response curves for SN-6, TTA-A2, DS5565 shown in Figures S2C, D, and E, respectively). The calpain-2 inhibitor NA-184 neither decreased PARP activity nor PAR generation in rd1 ONL (Fig. 7A, B; Table S8A, B; NA-184 dose-response in Figure S2F).

Fig. 7

Ca²⁺-signalling affects PARP activity, PAR generation, and sirtuin activity. A PARP activity assay (green) was performed in rd1 and wild-type (wt) retinal explant cultures, with DAPI (grey) as nuclear counterstain. Untreated (Untr.) rd1 and wt retina were compared to rd1 retina treated with BAPTA-AM, L-cis, CM4620, SN-6, D-cis, TTA-AS, DS5565, and NA-184. The scatter plot shows percentage of PARP activity positive cells in the outer nuclear layer (ONL). Untr. wt: n = 11; Untr. rd1: 24; BAPTA rd1: 10; L-cis rd1: 5; CM4620 rd1: 5; SN-6 rd1: 7; D-cis rd1: 10; TTA-A2 rd1: 6; DS5565 rd1: 6; NA-184 rd1: 8. B PAR staining (black) was performed in rd1 and wt retinal explant cultures. Untreated wt and rd1 retina were compared to drug-treated retina as in A. Untr. wt: n = 6; Untr. rd1: 17; BAPTA rd1: 7; L-cis rd1: 7; CM4620 rd1: 10; SN-6 rd1: 8; D-cis rd1: 8; TTA-A2 rd1: 8; DS5565 rd1: 8; NA-184 rd1: 9. C Sirtuin activity assay (white) was performed in rd1 and wt retinal explant cultures. Untreated rd1 and wt retina were compared to drug-treated retina as in A. Untr. wt: n = 7; Untr. rd1: 21; BAPTA rd1: 9; L-cis rd1: 9; CM4620 rd1: 10; SN-6 rd1: 8; D-cis rd1: 9; TTA-A2 rd1: 8; DS5565 rd1: 8; NA-184 rd1: 9. Statistical testing: one-way ANOVA with Tukey’s multiple comparison post hoc test performed between rd1 explant cultures. Error bars represent SD; ns = p > 0.05; * = p < 0.05; ** = p < 0.01; *** = p < 0.001; **** = p < 0.0001. INL = inner nuclear layer, GCL = ganglion cell layer; scale bar = 50 µm

The histone deacetylase (HDAC) sirtuin-1 (SIRT1) was suggested to be indirectly regulated by PARP-dependent consumption of NAD⁺ [59]. Thus, we investigated sirtuin activity using an HDAC in situ activity assay based on deacetylation of a SIRT1 specific substrate. In rd1 retina, the number of HDAC or sirtuin activity positive cells in the ONL was higher compared to wt control (Fig. 7C; Table S8C). BAPTA-AM further significantly increased sirtuin activity in rd1 ONL (Fig. 7C; Table S8C). Treatment with L-cis and CM4620 did not change the numbers of sirtuin activity positive cells, as compared to untreated rd1, while sirtuin activity was increased by SN-6 (Fig. 7C; Table S8C). Sirtuin activity did not rise after treatment with D-cis, TTA-A2, and DS5565 (Fig. 7C; Table S8C), but unexpectedly, NA-184 significantly increased the percentage of sirtuin positive cells in rd1 ONL (Fig. 7C; Table S8C).

Enzyme activity patterns are altered by changes in Ca²⁺ or by inhibition of calpain-2

To compare the various processes studied here with each other, we normalized the experimental data by linear scaling, such that the lowest values were set to zero while the highest values were set to one. In wt retina, all TUNEL staining and all enzyme activity markers were generally low when compared with the untreated rd1 group (Fig. 8A). In untreated rd1, the number of ONL cells showing calpain-1 activation was lower than calpain activity, calpain-2 activation, PARP activity, PAR positive cells, and sirtuin activity (Fig. 8A). Photoreceptor degeneration was significantly reduced by treatments with BAPTA-AM, SN-6, TTA-A2, and NA-184, while calpain-1 activation and sirtuin activity were relatively high, compared to other markers (Fig. 8A). In contrast, the treatments with L-cis and CM4620, which increased photoreceptor death, increased calpain activity, calpain-2 activation, PARP activity, and PAR-positive cells more than calpain-1 activation and sirtuin activity (Fig. 8A). In the D-cis and DS5565 treated groups, which did not affect photoreceptor viability, calpain-1 activation in ONL was generally lower than calpain-2 activation (Fig. 8A). Taken together, the activity patterns observed may constitute enzymatic signatures characteristic for either retinal degeneration or protection.

Fig. 8

Enzymatic signatures for rd1 photoreceptor degeneration and Spearman analysis. A Comparison of enzymatic markers across different experimental treatments. Normalized cell numbers positive for TUNEL (magenta), calpain activity (calpain, blue), activated calpain-2 (yellow), PARP activity (PARP, green), PAR (black), activated calpain-1 (cyan), and sirtuin (Sirt) activity (white). B Spearman analysis of enzymatic markers in photoreceptors (calpain activity, calpain-1, calpain-2, PARP activity, PAR, Sirtuin). The asterisks in circles show statistical significance, numbers in squares present the R. C Radar plot for Spearman analysis revealing the correlation between different enzymatic markers and the TUNEL assay. Note that while cell death (TUNEL) was strongly associated with general calpain activity, calpain-2 activation, PARP activity, and PAR, it was negatively correlated with calpain-1 activation and Sirtuin activity

To numerically capture these presumed activity patterns triggered by alterations in Ca²⁺-influx, we performed a Spearman analysis (Fig. 8B). This analysis excluded data from the NA-184 treatment since this calpain inhibitor would not per se change Ca²⁺-influx. General calpain activity and calpain-2 activation was positively correlated with PARP activity and PAR (R > 0.7, p < 0.0001). Calpain-1 activation and sirtuin activity were also positively correlated with each other (R > 0.6, p < 0.0001). A radar plot was used to show the relationship of cell death and enzymatic signatures after Spearman analysis (Fig. 8C). This illustrated that TUNEL positive cells in rd1 ONL were positively correlated with calpain activity, calpain-2 activation, PARP activity and PAR (R > 0.7, p < 0.0001, Fig. 8C). However, TUNEL positive cells were anti-correlated with sirtuin activity and calpain-1 activation. Taken together, these analyses suggest that calpain-1 activation and sirtuin activity were associated with photoreceptor survival, as opposed to general calpain activity, calpain-2, PARP, and PAR, which were strongly connected to cell death.

Discussion

Ca²⁺ overload mediated by CNGC and VGCC has repeatedly been connected to cGMP-induced photoreceptor death in IRD models [15, 60, 61]. However, a number of studies using inhibitors of Ca²⁺-permeable channels have reported contradictory effects [62, 63], such that the precise role of Ca²⁺ in photoreceptor degeneration is still unclear. Our initial bioinformatics analysis provided insights into the regulation of Ca²⁺-related genes connected to retinal degeneration. However, since gene expression along does not necessarily reflect the actual protein expression and protein functioning, we then used pharmacological experiments to confirm perturbations of intracellular Ca²⁺ in the diseased retina. homeostasis may cause photoreceptor degeneration, and that blocking Ca²⁺-influx may be neuroprotective. More specifically, our data suggest that Ca²⁺-permeable channels, T-type VGCC, and possibly also NCX, may serve as therapeutic targets for the treatment of IRDs (Fig. 9). Additionally, the activity of the Ca²⁺-activated proteolytic enzyme calpain-1, in contrast to calpain-2, was linked to lower cell death rates. Consequently, an inhibition of calpain-2 turned out to be neuroprotective. Overall, our study indicates that both substantial positive or negative deviations of intracellular Ca²⁺ levels may endanger photoreceptor viability.

Fig. 9

Experimental interventions and their relation to Ca²⁺ signalling in cGMP-dependent rd1 degeneration. In rd1 photoreceptors, the Pde6b mutation induces cGMP accumulation, which activates cyclic-nucleotide-gated channels (CNGC), leading to Na⁺ and Ca²⁺ influx. CNGC-dependent depolarization activates voltage-gated Ca²⁺ channel (VGCC) and may reverse directionality of Na⁺/Ca²⁺ exchanger (NCX), both leading to more Ca²⁺-influx. Additional Ca²⁺-influx may be mediated by Ca²⁺ release activated channel (CRAC). All these Ca²⁺ permeable channels may contribute to intracellular Ca²⁺ overload, which, in turn, may activate calpain-2 directly and poly(ADP-ribose)-polymerase (PARP) indirectly. On the one hand, over-activated calpain-2 increases proteolysis of neuroprotective calpain-1. On the other hand, Ca²⁺-dependent activation of PARP may positively feedback on Ca²⁺-influx via NAD⁺ depletion, which may reduce the protective activity of NAD⁺-dependent sirtuins. Eventually, the activities of calpain-2 and PARP, triggered by high intracellular Ca²⁺-levels promote photoreceptor cell death. The drugs used in this study and their targets are indicated. Red colour indicates destructive processes and drugs, while green labelled proteins and compounds promote photoreceptor survival, grey indicates absence of a clear beneficial or destructive effect

Ca²⁺ homeostasis is critical for the survival of retinal photoreceptors

In different cell types, the second messenger Ca²⁺ is involved in the regulation of a diverse range of cellular processes, including fertilization, metabolism, transcription, and cell death [64]. This diversity of Ca²⁺-signalling related processes requires a precise control of intracellular Ca²⁺ levels, which in a healthy cell are maintained at below 100 nM vs. extracellular levels exceeding 2 mM [64]. In photoreceptors, the over-activation of CNGC caused by high intracellular cGMP levels leads to an influx of Na⁺ and Ca²⁺, depolarizing the cell and activating VGCC, causing further Ca²⁺ influx [5, 61]. However, whether Ca²⁺ is responsible for photoreceptor degeneration is debated, as some reports indicate that blocking Ca²⁺-permeable channels may be neuroprotective [18, 19, 65], while other studies propose the opposite [20, 21]. Our bioinformatic analysis of RNA-seq data obtained at the peak of rd1 photoreceptor cell death (P13) revealed more than 80 DEGs associated with Ca²⁺. GSEA analysis and GO enrichment identified Ca²⁺-related DEGs associated with four pathways linked to proteolysis. Since at P13 the rd1 retina loses mainly rod photoreceptors [14], we then used scRNA-seq to focus on Ca²⁺-related pathways in photoreceptors. DEGs were enriched in Ca²⁺-related GO terms in both rods and cones, however, remarkably, rods displayed gene regulation in more Ca²⁺-related GO terms than cones. The well-studied membrane permeable Ca²⁺ chelator, BAPTA-AM, preserved photoreceptor viability, providing strong evidence for a contribution of excessive Ca²⁺-levels to rd1 rod degeneration. However, exceedingly low Ca²⁺ levels are also known to trigger cell death [66]. In line with this, BAPTA-AM increased the numbers of dying, TUNEL positive cells in wt retina and strongly increased cell death in rd1*Cngb1^−/− double-mutant retina. Together, these findings demonstrate that both exceedingly high Ca²⁺ levels and a depletion of intracellular Ca²⁺ can lead to retinal degeneration [67]. The key question that remains is which Ca²⁺-permeable channel(s) may be responsible for excessive Ca²⁺-influx.

CNGC and CRAC activity does not promote rd1 degeneration

In the present study, we investigated the roles of CNGC, CRAC, NCX, and VGCC in photoreceptor cell death (Fig. 9). CNGC is activated by high cGMP, linking it to the rd1 mutation in PDE6. The CNGC gene family comprises six members (CNGA1-4, CNGB1 and CNGB3), of which CNGA1 and CNGB1 are expressed by rods, while CNGA3 and CNGB3 are expressed in cones [68]. In the whole retina RNA-seq data Cnga1 appeared to be down-regulated, while the scRNA-seq showed it as up-regulated in rod photoreceptors. This discrepancy may have been caused by the ongoing photoreceptor degeneration since in P13 rd1 retina a number of rod photoreceptors have been lost already compared to their P13 wt counterparts [14]. However, the inhibition of CNGC by L-cis-diltiazem accelerated photoreceptor cell death, in line with previous research [20], but in apparent contradiction with data obtained in rd1*Cngb1^−/− double-mutant mice [19]. However, the rd1*Cngb1^−/− double-mutant mice still retain residual CNGC activity and Ca²⁺-influx through homomeric CNGA1 channels [69]. L-cis-diltiazem treatment will abolish even this residual Ca²⁺-influx, probably leading to cell death triggered by too low intracellular Ca²⁺.

CRAC activity has previously been connected to cell death and inhibition of CRAC has been shown to increase cell viability [70]. CaMK2 may potentiate SOCE and CRAC activity via enhancing Stim1 aggregation and interaction with Orai [53]. In Pde6b-mutant mice, the Camk2g gene was found to be up-regulated in the present and in previous studies [71]. However, the Camk2d gene displayed a down-regulation in to the bulk RNA-seq dataset, which warranted an independent confirmation as to whether CRAC contributed to Ca²⁺ overload during rd1 retinal cell death. Selective CRAC inhibition with CM4620 triggered photoreceptor degeneration in wt retina and increased photoreceptor death even further in rd1 retina, indicating that contrary to our expectations CRAC-mediated Ca²⁺-influx was essential for photoreceptor survival.

Taken together, our data indicates that the activity of CNGC and CRAC Ca²⁺-permeable channels appears to be protective rather than destructive, and thus neither of these is linked to rd1 photoreceptor degeneration. We note that these results may have been influenced by our use of early post-natal retina, in which photoreceptors and their outer segments are not yet fully developed. However, given the marked detrimental effects of CNGC and CRAC inhibitors, it seems unlikely that this outcome would change in more mature retina.

T-type VGCC and NCX contribute to Ca²⁺ overload and rd1 photoreceptor cell death

Apart from CNGC and CRAC, we also investigated the importance of different VGCC-types and NCX for rd1 degeneration. NCX is a bi-directional regulator of cytosolic Ca²⁺, capable of mediating both Ca²⁺ influx and Ca²⁺ efflux. NCX inhibition bestows resistance to retinal damage induced by N-methyl-D-aspartate (NMDA) and high intraocular pressure [72]. In our hands, the NCX inhibitor SN-6 [73] reduced calpain activity and prolonged rd1 photoreceptor viability, yet it increased cell death in the wt situation. This suggests a reversal of NCX directionality in the two different genotypes, possibly caused by high intracellular Na⁺ levels [52] and membrane depolarization [24]. Thus, cGMP-dependent over-activation of CNGC may have reversed NCX in rd1 rod photoreceptors. Surprisingly, while BAPTA-AM led to photoreceptor cell death in rd1*Cngb1^−/− retina, perhaps due to Ca²⁺ depletion, NCX inhibition in the rd1*Cngb1^−/− situation attenuated retinal degeneration, indicating that NCX is in forward mode when CNGC activity is low. However, neither BAPTA-AM nor SN-6 preserved the outer retina in the long-term, implying that other mechanisms, such as those triggered by cGMP-dependent protein kinase G (PKG) may promote retinal degeneration independent of elevated Ca²⁺ [74, 75].

VGCC is activated by membrane depolarization, allowing Ca²⁺ entry into the cells [76]. VGCC consists of four non-covalently associated subunits: α₁ (L-type and T-type VGCC), β, α₂δ, and γ [77]. In rd1 photoreceptors, VGCC may indirectly be activated by cGMP-induced over-activation of CNGC [61]. Hence, D-cis-diltiazem [78], DS5565 [79], and TTA-A2 [80] were used to block different VGCC types. D-cis-diltiazem did not prevent retinal degeneration, consistent with previous literature [21, 81, 82]. Interestingly, L-type VGCC genes (Cacna1s and Cacna1f) were shown up- and down-regulated respectively, suggesting there could be variable effects within different L-type VGCC isoforms. Furthermore, while we observed an up-regulation of genes coding for the α₂δ subunits of VGCC in rd1 rods, treatment with the α₂δ subunit VGCC ligand DS5565 did not protect rd1 photoreceptors either. However, DS5565 reduced calpain activity, indicating that the α₂δ subunit was effectively regulating intracellular Ca²⁺ levels. Our data RNA-seq and scRNA-seq also indicate a differential regulation of the β subunit of VGCC (Cacnb2), which, however, does not seem to be related to an increased therapeutic response to VGCC inhibition.

Unexpectedly, it was the inhibition of T-type VGCC that significantly attenuated photoreceptor cell death. This is surprising since T-type channels are generally thought to open at relatively negative membrane potentials of around − 40 to -80 mV, as opposed to L-type channels which may open already at -20 mV [83]. In photoreceptors the resting potential is approx. -40 mV and the membrane potential would reach − 70 mV only during light-induced photoreceptor hyperpolarisation [84]. Thus, in rd1 photoreceptors, which cannot be hyperpolarized by light due to PDE6 dysfunction, one would assume that L-type VGCC rather than T-type should carry most of the Ca²⁺-currents. However, we cannot exclude the possibility that T-type channels change their properties to become permeable at less negative membrane potential, or that the rd1 photoreceptor membrane potential shifts to more negative values.

Overall, our data suggest T-type VGCC and NCX contribute to rd1 retinal degeneration. One may speculate that in the initial phases of an individual photoreceptor´s demise, the membrane potential is still maintained at sufficiently negative values, allowing for an opening of T-type channels and Ca²⁺-influx. Eventually, however, the cell may no longer be able to keep its membrane potential and becomes depolarized, so much so that NCX reverses its direction and turns into a net importer of Ca²⁺. This idea of two different stages in the dysregulation of photoreceptor Ca²⁺-levels may also help to understand why NCX inhibition did not afford long-term protection. While the exact mechanism remains to be clarified, our data importantly highlights T-type VGCC as a potential target for therapeutic [84] intervention.

Calpain-2 causes photoreceptor cell death and negatively regulates calpain-1

High levels of intracellular Ca²⁺ will activate calpain-type proteases and calpain activation may be connected to CNGC activity [19, 85]. Calpain consist of a family of Ca²⁺-activated neutral cysteine proteinase involved in a large number of cellular processes [34, 86]. Originally, the calpain proteolytic system was reported to consist of three different proteins: calpain-1 (or µ-calpain, activated at µM Ca²⁺ concentrations) and calpain-2 (or m-calpain, activated at mM Ca²⁺ concentrations), as well as their endogenous inhibitor, calpastatin [86]. Recent evidence indicates that calpain-1 and calpain-2 play opposite roles, with calpain-1 being neuroprotective while calpain-2 promotes neurodegeneration [34]. We previously found that in rd1 mice, activated calpain-2 was increased rather than calpain-1 [56], and the current GSEA analysis indicated a positive regulation of proteolytic pathways. Thus, we assumed that calpain-2, as part of the proteolytic processes induced by Ca²⁺ [87], led to retinal degeneration. Therefore, we treated rd1 retina with the calpain-2 specific inhibitor NA-184 [57]. This caused a significant decrease in the numbers of TUNEL positive cells in the ONL, highlighting the role of calpain-2 in photoreceptor cell death. On the other hand, calpain-1 activation was decreased in situations where calpain-2 activation was high, possibly due to calpain-2-dependent cleavage of calpain-1´s protease core domain 1 [88]. Accordingly, NA-184 treatment also increased calpain-1 activation. These findings extend one of our previous studies where we found the general calpain inhibitor calpastatin peptide to protect photoreceptors in vitro and in vivo [89]. Yet, since calpastatin peptide inhibits both calpain-1 and − 2, the beneficial effects of calpain-2 inhibition may have been partly offset by the additional calpain-1 inhibition. Taken together, we found calpain-2 activation to be related to photoreceptor cell death, and this was associated with decreased calpain-1 activation.

It is important to note that our general calpain activity assay does not discriminate between specific calpain isoforms and that also isoforms other than calpain-2 could be involved in photoreceptor degeneration. A candidate isoform may be calpain-5 which was shown to be expressed in the retina [90] and found to be involved in inherited forms of uveitis [91]. Calpain overactivation may be relevant also for a variety of other retinal neurodegenerative diseases, including neovascular inflammatory vitreoretinopathy (NIV), diabetic retinopathy (DR), uveitis, glaucoma, and cataractogenesis [92]. Specifically, calpain-2 has been connected to NIV, cataractogenesis, and glaucoma, while calpain-5 was associated with NIV and uveitis [92, 93]. Whatever the case, inhibition of calpain-2 and/or -5 could provide therapeutical benefit in these ocular diseases.

Ca²⁺ differentially regulates PARP and sirtuin activity

PARP and HDAC activity were previously connected to rd1 degeneration and to Ca²⁺-signalling [85, 94, 95]. PARP is a DNA repair enzyme, which catalyses the polymerization of ADP-ribose units – derived from the ADP donor NAD⁺ – resulting in the attachment of either linear or branched PAR polymers to itself or other target proteins [96]. However, excessive PARP activity may drive a specific form of cell death, termed PARthanatos [97], which has also been connected to in retinal degeneration [4]. Accordingly, in certain IRD animal models, PARP inhibition was found to be neuroprotective [16, 98]. In line with previous literature [99, 100], the inhibition of Ca²⁺ -permeable channels reduced PARP activity and PAR generation in rd1 explants, suggesting that Ca²⁺ regulated PARP activity. However, it is yet unclear how Ca²⁺ produces PARP over-activation. The specific calpain-2 inhibitor, NA-184, did not reduce PARP activity in rd1 explants, suggesting that Ca²⁺ controls PARP activity independently of calpain-2 activity [16]. Instead the PARP activity observed may be caused by high levels of Ca²⁺ stimulating respiratory chain activity and leading to higher amounts of reactive oxygen species [101]. This in turn could produce oxidative DNA damage and trigger PARP hyper-activation [96, 102].

HDACs catalyse the removal of acetyl groups from lysine residues of both histone and nonhistone proteins [103]. HDACs are divided into zinc-dependent HDACs, and sirtuins, a family of NAD⁺-dependent HDACs [103]. As PARP is a main consumer of NAD⁺ [104], sirtuin activity can be regulated by PARP according to the NAD⁺ levels [105]. The Sirtuin-1 protein is suggested to be protective in neurons [106]. While we observed decreased PARP activity after most of our experimental interventions, sirtuin activity was increased after treatment with BAPTA-AM and the NCX inhibitor SN-6, suggesting that the resultant decrease in PARP activity may relieve the NAD⁺ shortage induced by PARP. Interestingly, calpain-2 inhibition also increased sirtuin activity without affecting PARP activity. However, currently there is no evidence indicating that calpain-2 cleaves HDACs. Still, HDACs may regulate calpain-2 activation indirectly via epigenetically increasing calpastatin expression [107]. Calpain may also influence mitochondrial biogenesis, as calpain plays a detrimental role upstream of the peroxisome proliferator-activated receptor γ (PPARγ) coactivator-1α (PGC-1α) pathway [108], which regulates the transcription of numerous nuclear-encoded mitochondrial genes [109] and is a key driver of mitochondrial biogenesis [110]. Thus, it appears possible that calpain-2 inhibition may improve rd1 photoreceptor energy metabolism.

Conclusion

In IRD research, the role of Ca²⁺-signalling in disease pathogenesis has remained controversial for a long period of time. The novel results presented here indicate that Ca²⁺-signalling may have both beneficial and detrimental effects in rd1 photoreceptors, depending on the source of Ca²⁺ and probably its intracellular localization. More specifically, inhibition of CNGC and CRAC accelerated cell death, while Ca²⁺ chelation, as well as inhibition of NCX and T-type VGCC increased photoreceptor viability. A selective inhibition of calpain-2 improved photoreceptor viability as well, while, remarkably, activation of calpain-1 and sirtuin-type HDAC was linked to photoreceptor survival. In contrast, general calpain activity and activity of PARP were found to be destructive. While our results propose T-type VGCC and calpain-2 as therapeutic targets for IRD treatment (Fig. 9), a careful context- and genotype-specific evaluation appears indicated as suggested, for instance, by the opposing results obtained in rd1 single-mutant and rd1*Cngb1^−/− double-mutant retina. In perspective, our study illustrates the complexity of Ca²⁺-signalling during photoreceptor degeneration and highlights the need for additional work to further delineate destructive and protective pathways to promote the rational development of new therapeutic approaches for IRD and related retinal diseases.