Targeting the Neuro-vascular Presynaptic Signalling in STROKE: Evidence and Therapeutic Implications

Abstract

Background

Stroke is one of the leading causes of death and long-term adult disability worldwide. Stroke causes neurodegeneration and impairs synaptic function. Understanding the role of synaptic proteins and associated signalling pathways in stroke pathology could offer insights into therapeutic approaches as well as improving rehabilitation-related treatment regimes.

Purpose

The current study aims to analyse synaptic transcriptome changes in acute and long-term post-stroke (1 day, 7 day timepoints), especially focusing on pre- and postsynaptic genes.

Methods

We performed data mining of the recent mRNA sequence from isolated mouse brain micro-vessels (MBMVs) after transient middle cerebral artery occlusion (tMCAO) stroke model. Using the SynGO (Synaptic Gene Ontologies and annotations) bioinformatics platform we assessed synaptic protein expression and associated pathways, and compared synaptic protein changes at 1 day and 7 day post-stroke.

Results

Enrichment analysis of the MBMVs identified significant alterations in the expression of genes related to synaptic physiology, synaptic transmission, neuronal structure, and organisation. We identified that the synaptic changes observed at the 7 day timepoint were initiated by the regulation of specific presynaptic candidates 1 day (24h) post-stroke, highlighting the significance of presynaptic regulation in mediating organising of synaptic structures and physiology. Analysis of transcriptomic data from human postmortem stroke brains confirmed similar presynaptic signalling patterns.

Conclusion

Our findings identify the changes in presynaptic gene regulation in micro-vessels following ischaemic stroke. Targeting presynaptic active zone protein signalling could represent a promising therapeutic target in mitigating ischaemic stroke.

Introduction

Ischaemic stroke constituted 62.4% of all new strokes in 2019 and is one of the leading causes of death and disability globally. ¹ Extensive fundamental and clinical research has addressed the mechanisms leading to the disruption and recovery of the blood-brain barrier (BBB) after stroke. ² The BBB is a unique structure of micro-vessels, which maintains central nervous system (CNS) homeostasis and comprises of endothelial cells in concert with astrocytes, pericytes, a basement membrane and innervating neurons. ³ The endothelium, neurons, and non-neuronal cells (pericytes, astrocytes and microglia) constitute a functional unit, termed the neuro-vascular unit, which plays a major role in the haemodynamic response under physiological conditions and in injury.^{3, 4} One of the major pathophysiological features of ischaemic stroke includes synaptic failure with impaired transmitter release. ⁵

Neuronal circuitry allows inter-synaptic crosstalk in physiological conditions which is facilitated by transmitter release from the synaptic cleft and its diffusion to neighbouring synapses. ⁶ Synapses function as integrated networks that regulate neuronal activity, which is vital for learning, memory, and behaviour. Synaptic vesicles (SVs) at the axon terminal are believed to dock and fuse at a specialised structure at the presynaptic site, the active zone and release the neurotransmitter. ⁷ According to ultrastructural analysis, the active zone is an electron-dense network of cytoskeletal filaments and scaffolding proteins regulating the SV cycle and mediating the presynaptic functions. Within the presynapse, the multidomain scaffold proteins Bassoon and Piccolo play crucial roles in tethering and priming SVs for release. Bassoon regulates the autophagy in presynaptic terminals, and its loss results in the ubiquitination and removal of vesicle proteins from the presynapse.^8–10 Investigating the role of synaptic dysfunction in neurological disorders remains challenging due to limited and typically late-stage access to human tissue and the inadequate representation of human disease characteristics in existing animal models.

Ischaemia-induced depletion of blood flow rapidly deprives neurons of oxygen and energy substrates. This abrupt disruption activates autophagic degradation in neurons and loss of synapses, ultimately leading to acute neurological dysfunction and, if prolonged, permanent neurological damage. ¹¹ The neuronal loss and reduced synaptic density in the brain¹², ¹³ aligned with glutamate release is followed by the breakdown of neuronal ionic balance, induction of glutamate-mediated pathways and cell death due to a massive calcium influx. This glutamate excitotoxicity is such an initial event when it occurs in an ischaemic lesion that it results in a subsequent cascade of injury to neurons. ¹⁴ Sustained hypoxic ischaemia depletes ATP and interferes with glutamatergic synaptic transmitters’ ability to transfer information. This might be connected to the inactivation of voltage-gated calcium channels (VGCC) and the docking problem of glutamate vesicles caused by faulty presynaptic protein phosphorylation. ¹⁵ The inability of the dynamic turnover of synaptic proteins is likely to be the cause of the damage to the synaptic superstructure observed in the cerebral artery occlusion (MCAO) animals following reperfusion. ¹⁶ Ischaemic stroke impedes the SVs recirculation through the inhibition of synaptic endocytosis and exocytosis. ¹⁷ Transcriptome analysis thus, allows to identify the molecular targets that trigger the signalling changes, for example, affecting the synaptic functions, post-stroke and aids in determining the therapeutic options.

Here, we analysed the transcriptome data of the mouse brain micro-vessels (MBMVs) from the available database^{18, 19} for potential alterations in synaptic changes specifically emphasising the presynaptic changes. We have performed the SynGO (Synaptic Gene Ontologies and annotations) analysis to elucidate the pre- and postsynaptic candidates’ changes and related signalling alterations. Further, we sought to identify the highly regulated presynaptic active zone candidates to understand the protein-protein interactions through the STRING (Search Tool for the Retrieval of Interacting Genes/Proteins) database. All the significantly regulated synaptic proteins were found in MBMVs at 1 day and 7 days post-stroke injury in mice.

Methods

Data Mining From mRNA Sequences of MBMVs After tMCAO Stroke Induction

To gain a comprehensive understanding of alterations in synaptic proteins following ischaemia, we investigated the online database encompassing mRNA sequencing analysis of mice (transient middle cerebral artery occlusion [tMCAO] model) brain micro-vessels (MBMVs) from 1 day (24h) post-stroke and 7 days post-stroke. ¹⁸ The sequencing data from two other independent studies were analysed for a comparative and broader understanding of synaptic signalling.^{18, 19}

In-silico Analysis

We focused on the analysis of the transcriptome changes of the synaptic genes, upon stroke induction using in-silico bioinformatics tools such as SynGO and STRING.

Identification of Synaptic Candidates

All synaptic candidates with biological processes and localisation are specifically detected using SynGO. ²⁰ In the MBMVs of the stroke model, this database detected all synaptic candidates expressed at two different time points at the mRNA level. Pre- and postsynaptic candidates were identified in the brain at various biological sites and stages. Website: https://www.syngoportal.org/

Regulation of Synaptic Proteins

Based on fold change between the sham and stroke groups of the ipsilateral hemispheres in the post-stroke mouse model at 1 day (24h) and 7 days, significantly regulated synaptic candidates were established. The range of fold change that is commonly accepted for identifying all significantly regulated expressed candidates is set to +1.5 or −1.5 or above. Utilising the SynGO database, we determined the biological processes associated with certain presynaptic and postsynaptic genes from the significant gene list. Significantly regulated candidates are identified by a P value of .05.

Protein-protein Interaction Analysis

The STRING was utilised to anticipate protein-protein interactions with 0.400 medium degree of confidence, demonstrating maximum and potential interaction with partner proteins. This database determines associations that are both functional and physical. Website: https://string-db.org/

Statistical Analysis

The presynaptic active zone candidates’ significant changes were determined using 1-way ANOVA, with post hoc comparison using Bonferroni post-test between three conditions like sham versus stroke 1 day ipsilateral, sham versus stroke 7 days ipsilateral and sham 1 day versus 7 days ipsilateral. Analysis was performed using GraphPad Prism.

Results

Based on the SynGO analysis, numerous regulated synaptic candidates have been detected in both pre- and postsynaptic areas in the sunburst plot. Out of the 1923 background gene list, 322 synaptic genes were expressed under all circumstances both the day after the stroke and 7 days later. At this timepoint, the synaptic compartment expressed both upregulated and downregulated genes equally.

Localisation and Biological Process of Synaptic Compartment

Many theories have been proposed to explain how synapse-specific, transcription-dependent forms of synaptic plasticity can be long-lasting. ²¹ Synaptic genes play a key role in a myriad of biological processes necessary for the healthy operation of neural circuits, ranging from the complex regulation of neurotransmitter release and postsynaptic receptor dynamics to the precisely controlled modulation of synaptic plasticity and structural alterations. ²²

Changes in the mRNA Expression of Synaptic Proteins 1 Day (24h) After Stroke

Ischaemic stroke had an extremely rapid effect on the gene expression of presynaptic proteins; effects were visible as early as 1 day post-stroke, especially when compared to the ipsilateral micro-vessels of the sham mice. Only 34 of the 1028 synaptic genes were expressed post-stroke at this early stage, whereas out of 531 just 16 presynaptic genes were expressed 1 day post-stroke condition. Similarly, a few genes of the postsynaptic compartment (18 genes out of 621) were expressed at 1 day post-stroke (Figure 1a). Interestingly, presynaptic genes were significantly downregulated in the immediate aftermath of stroke induction caused by cell death.

Figure 1.

Source: Kestner et al.18

Analysing the cellular localisation of a subset of pre- and postsynaptic genes, some are equally distributed between the pre- (16 genes out of 531) and post- (18 genes out of 621) synaptic structures (Supplementary Table 1). Leveraging cellular annotation analysis, wherein a control gene list serves as a reference from an online database, facilitated a comprehensive understanding of the observed expression patterns.

To highlight the functional implications, such as neurotransmitter release, the precise location of the presynaptic candidates is essential, which further classifies the complex biological mechanisms governing the activity of presynaptic proteins. Post-stroke conditions, especially at the 1 day time point, had a significant effect on several biological functions (11 genes out of 269) (Supplementary Figure 1a). Specifically, there was a notable downregulation of the presynaptic membrane potential (Gabra1, Scn1a, and Hcn1) and presynaptic cytosolic calcium (Calb2) related gene expression (Figure 1c, e).

Furthermore, specific processes within the SVs cycle demonstrated differential regulation. Mctp2 displayed an upregulation, whereas other crucial genes (Cadps, Slc32a1, Sv2c, Syt2, Slc10a2) exhibited significant downregulation (Figure 1g). Processes such as presynaptic dense core vesicle exocytosis and neurotransmitter reuptake were also affected, with downregulation observed in Cadps and Drd2, respectively (Figure 1d, f). These findings underscore the intricate molecular responses within the presynaptic compartment in the acute aftermath of stroke, shedding light on the diverse regulatory landscape governing synaptic function under pathological conditions.

The effects on postsynaptic candidates (7 genes out of 218) at the 1-day post-stroke were less pronounced than those on presynaptic candidates (Supplementary Figure 1a). The regulation of postsynaptic membrane potential (involving genes Gabra1, Hcn1, Gabrg2, Glra2) and postsynaptic membrane neurotransmitter receptor expression (involving genes Rapsn, C1ql2, Erbb4) are the only two biological processes that notably revealed significant downregulation (Figure 1h, i). These data suggest a clear sensitivity of presynaptic compartment to the acute conditions manifested by stroke as highlighted by this differential response. Remarkably, the regulatory effect on postsynaptic candidates was significantly less than that of the more significant changes observed in their presynaptic counterparts (Figure 1a, c–i). The complex and dynamic nature of synaptic regulation in post-stroke is further highlighted by this temporal and compartmental disparity in the molecular response.

Changes in mRNA Expression of Synaptic Genes 7 Days After Stroke

Following 7 days post-stroke, a significant increase in the expression of genes related to synapses was observed in the pre- and postsynaptic compartments in the ipsilateral part of the brain. Specifically, even 7 days after the stroke, the presynaptic compartment continued to exhibit strong effects. Of the several hundreds of genes examined, a significant portion, 228 genes out of 1082, all showed elevated expression following the intermediate-term effects of stroke. The cellular localisation gene expression profile, as shown in (Figure 1b) clearly demonstrates the widespread impact of intermediate-term stroke effects on all synaptic genes. This manifestation highlights the long-lasting and pervasive molecular changes within the synaptic machinery, providing important information about the intermediate-term effects of stroke on synaptic regulation.

A significant symmetry was revealed by the cellular localisation analysis of the pre- (122 genes out of 531) and post- (134 genes out of 621) synaptic compartments, where a considerable number of genes were found to be equally distributed between the two compartments. Interestingly, there was a notable downregulation of these genes’ expression levels, suggesting a consistent and strong influence on both synaptic locations. The collective downregulation of gene expression was especially noticeable in the intermediate-term consequences after stroke. The detailed picture of the localised and impacted proteins highlights the long-lasting effects of stroke on the complex spatial dynamics of synaptic regulation and clarifies the persistent changes in the mRNA expression in pre- and postsynaptic domains (Supplementary Tables 2, 3).

Intermediate-term post-stroke conditions impacted scores of genes. The expression of the pre- (84 genes out 269) and post- (65 genes out of 218) synaptic compartment is influenced by numerous biological processes (Supplementary Figure 1b). Presynaptic candidates remain severely affected in the 7d post-stroke group, similar to the post-1-day condition (Figure 1c–g, 2a–e). The presynaptic genes significantly affected the SVs cycle (Syn2, Ncs1, Amph, Sncb, Dnajc6, Napb, Dnm1, Cspg5, Ppfia3, Syt2, Syt9, Syp, Nrxn1, Atp6v1g2, Cacna1b, Stx1b, Sv2a, Rab3b, Rph3a, Kch1, Prkcg, Prrt2, Cplx3, Sh3gl2, Cdk5r1, Slc32a1, Tmem163, Slc17a6, Pacsin1, Unc13a, Sv2c, Erc2, Snap91, Syngr3, Cacnb4, Cplx1, Cadps, Sv2b, Bsn, Camk2a, Npy, Slc17a7, Slc30a3, Rims1, Syn1, Pclo, Snap25, Rimbp2, Syt1), the neurotransmitter uptake (Slc6a3), the presynaptic cytosolic calcium levels regulation (Ncs1, Calb2, Atp2b3, Cacna1b, Tspoap1, Kcnh1, Calb1, Erc2, Cacnb4, Sv2b, Npy, Cnr1, Rimbp2), the presynaptic membrane potential regulation (Gria1, Gabra2, Kcna2, Gria4, Gabrb1, Kcnc4, Grin1, Kcnj11, Grin2b, Grik2, Kcnh1, Gria3, Kcnh1, Gria3, Gabra1, Kcnmb4, Gria2, Gabra5, Gabra3, Kcnc1, Grik3, Scn1a, Hcn1, Kcnc2, Kcnq5, Kcnj9, Kcnj3, Grin2a, Kcna4, Rimbp2), the exocytosis of presynaptic dense core vesicles (Dnm1, Cadps, Rims1, Snap25), and the neurotransmitter reuptake (Slc1a2, Slc6a1, Syngr3, Drd2) (Figure 2a–e).

Figure 2.

Source: Kestner et al.18

Our results further demonstrate that the effects on postsynaptic candidates are significantly stronger at the 7-day post-stroke period than they are at the 1-day post-stroke conditions (Figure 1h–I, 3a–c). An in-depth analysis of the impacted postsynaptic compartment demonstrated the participation of altered candidates in numerous biological mechanisms. Indeed, the processes including the exocytosis of postsynaptic dense core vesicles (Syt6), regulation of postsynaptic neurotransmitter receptor activity (Nptxr, Dlgap2, Shisa9, Neto1, Neto2, Nptx1, Shisa7, Cacng5, Cnih2, Nptx2), regulation of postsynaptic cytosolic calcium levels (Atp2b2, Calb1), regulation of postsynaptic membrane neurotransmitter receptor levels (Kif2c, Nptxr, Ctnnd2, Nsg2, Syt3, Dlg4, Olfm1, Prkcz, Pak3, Igsf11, Cplx1, Cacng8, Grid2, Cntnap2, Slc12a5, Adam22, Tmem108, Olfm2, Neto1, Lgi1, Neto2, Nrxn3, Cnih3, Hpca, Pacsin1, Nptx1, Cacng2, Caly, Frrs1l, Camk2a, Akap5, Cacng3, Dlg2, Sst, Nptx2, Arc) where only Kif2c is shown upregulation, and regulation of postsynaptic membrane potential (Gria1, Gabra2, Chrna4, Grin3a, Kcna2, Gria4, Gabrb1, Kcnc4, Glrb, Grin1, Gabrd, Kcnd2, Grin2b, Gabrg2, Grik2, Gria3, Gabra1, Gabrg3, Gria2, Gabra5, Chrm1, Gabra4, Gabra3, Hcn1, Rgs7, Gabrb3, Grin2a). The increased regulatory influence on these various biological processes highlights how postsynaptic responses to the long-lasting effects of stroke are dynamic and ever-changing over a prolonged period post-ischaemia injury (Figure 3a–c).

Figure 3.

Source: Kestner et al.18

The long-term impact of stroke on overall synaptic functions is indicated by the significant disruption and strong downregulation of all genes present in pre- and postsynaptic compartments.

Regulation of Active Zone Presynaptic Proteins in Post-stroke Conditions

The regulatory impact of presynaptic active zone proteins on the SVs cycle is critical for coordinating the release of neurotransmitters into the synaptic cleft. ²³ However, the CNS experiences severe disruptions following the induction of a stroke in the ipsilateral region of the mouse brain. To understand their effects after stroke, we sought to analyse how the presynaptic candidates, more specifically the active zone candidates, are regulated at different time points. Notably, from the analysis data, active zone candidates like Bassoon, Piccolo, Rim1 and Rimbp2 were significantly downregulated in response to stroke-induced damage (Figure 4). Interestingly, Rim2 showed no discernible changes, pointing to the potential for isoform-specific differences that make it less vulnerable to the effects of ischaemia. These all-presynaptic active zone proteins are cardinal to implicate the neurotransmitter release process—if one of them especially Bassoon or Piccolo (large scaffold protein) are dearth, the system results in failure of the presynaptic functions. In the MBMVs, surprisingly presynaptic proteins are also affected post-stroke. Bassoon exhibits a significant downregulation under both circumstances, with fold changes of −1.39 and −2.88 for the day-1 and 7 days ipsilateral between sham vs stroke model, respectively. On the other hand, at same conditions, the Piccolo fold change is −1.43 and −3.43, respectively. Likewise, fold change of Rim1 and Rimbp2 is −1.10, −3.10 and −0.73, −3.90, respectively (Figure 4a–d). The intricate molecular responses within the presynaptic compartment are illuminated by the regulation of the active zone, offering important insights into the effects of stroke on neurotransmitter release dynamics in the nervous system.

Figure 4.

Source: Kestner et al.18

Further to analyse the interacting partners and thus delineate the complex signalling mediated by the altered active zone proteins, we employed STRING database to comprehend the protein-protein interaction (Supplementary Figure 2a–d). Indeed, the interacting partners of these active zone proteins are mostly the other active zone proteins, thus highlighting the coordinated functions of these scaffolding proteins at the presynaptic compartment.

Discussion

Induction of stroke via the tMCAO in mice serves as a valuable tool for comprehending gene expression dynamics in various biological processes within the pre- and postsynaptic neurons of the brain. ²⁴ Transcriptome analysis helps to identify the regulated gene expression from the entire set of mRNAs which is produced from the genome.^{25, 26} While investigating the post-stroke effects, specific attention was devoted to two distinct time points, 1 day (24h) and 7 day post-stroke, shedding light on the short-term and intermediate-term impacts after transient middle cerebral artery occlusion in mice.

The temporal dimension allowed for a nuanced understanding of the evolving biological consequences in the injured brain region. Upon closer examination, it was observed that in the 1 day post-stroke condition, the presynaptic compartment exhibited a pronounced impact, particularly in the SVs processes. In contrast, the postsynaptic biological processes did not experience a commensurate perturbation level at this early stage. This is consistent with two recent studies that compared RNA sequencing data from micro-vessels in a similar model (tMCAO) with publicly available RNA sequencing data from lesion-site samples of stroke patients and transcriptomic analyses of human stroke brain tissue. ²⁷ We have performed similar SynGO analysis on both the data sets from these studies showing a strong perturbation of presynaptic processes, supporting our result (Supplementary Figure 3).

This discovery emphasises how the biological reactions to stroke are dynamic and ever-changing, with long-term consequences for both synaptic compartments. The research asserts how complex the temporal dynamics of gene expression are in pre- and postsynaptic neurons after stroke induction. A more complex understanding of the biological effects of stroke on the mouse brain is made possible by the simultaneous influence on presynaptic processes and the long-term, broad-spectrum alterations in both compartments. Here, the main focus of the study is to understand the regulation of the mRNA of the active zone presynaptic proteins, which are significantly altered in both conditions, 1 day and 7 days post-stroke in the micro-vessels of ipsilateral tMCAO mouse brain. Where protein-protein interactions between different active zone presynaptic proteins are maximum, involvement in the regulation of various biological systems such as synaptic vehicle function, active zone structure, regulation of synaptic vehicle, etc, are evident in our analysis.

Our analysis identified Bassoon, Piccolo, Rim1, and Rimbp2 to be significantly regulated after stroke and that all of them are specifically affected by SV function, which is indirectly responsible for neurotransmitter release in the synaptic cleft. Mechanistically, Bassoon, Piccolo, Rim1 and Rimbp2 have been shown to play crucial roles in various aspects of SV cycle and regulate the neurotransmitter release.^{23, 28} Bassoon constitutively limits autophagy in presynaptic terminals, and its loss results in the ubiquitination and removal of vesicle proteins from the presynapse. Further, Bassoon loss triggers changes in neurogenesis and maturation of synapses and circuits in hippocampus through altered neurotrophic signalling.^{29, 30} Neurogenesis involves two processes, the proliferation and maturation of neural progenitor cells. Regulating these processes is crucial for identifying possible targets for therapeutic interventions promoting functional recovery after stroke. ³¹ Indeed, previous studies have shown reduced synaptic density post-stroke in humans across different time points.^{13, 32} Others using rodent models for stroke, observed decreased expression levels of synaptosome-associated protein 29 (SNAP-29) and postsynaptic density protein 95 (PSD-95),^{33, 34} suggesting extensive studies to be conducted detailing the synaptic signalling in stroke to develop novel therapeutic avenues.

In the current study, we used mRNA sequence data from isolated micro-vessels due to the lack of literature on mRNA sequencing performed on isolated neurons from either mice or humans. It was indeed surprising to see the signatures of several synaptic proteins attached to the micro-vessels and therefore considered the mRNA sequence of MBMVs to analyse the changes in synaptic proteins. In addition, we are limited to only two different time points to compare the changes in synaptic proteins as they were the only time points used by Kestner et al. In future, it would be valuable to conduct a comprehensive analysis using mRNA sequences from neurons isolated at various time points post-stroke, should they become available. Nevertheless, our current study in MBMVs could identify the impacted genes and their regulation patterns during these time spans was made easier by our findings. The development of targeted therapies for stroke patients could be facilitated by this thorough knowledge.

Conclusion

The pre-and postsynaptic candidates that are substantially regulated at transcriptome levels following stroke induction in the tMCAO mouse model were the focus of this study. Our analysis leads us to the conclusion that genes governing several biological processes at both the pre- and postsynaptic compartments are severely regulated (largely downregulated) at an intermediate-term timepoint (7 days) post-stroke. This severe regulation begins at the presynaptic compartment as early as 1 day after the stroke. Neuronal structure and functions are both impacted by stroke, and in recent decades, there has been a major advancement in our understanding of the underlying causes of this condition. There are only a few effective treatments available despite these advancements, especially for patients who are nearing the end of their recuperation. Consequently, there is an urgent need for targeted therapies aimed at enhancing brain plasticity and improving long-term clinical outcomes in stroke patients.