Small molecule neurolysin activators, potential multi-mechanism agents for ischemic stroke therapy

1. Introduction

As we often learn from nature, elucidation and understanding of the brain’s endogenous self-protective and recovery mechanisms have been an emerging approach in the stroke research field to fulfill the unmet need for development of new therapies [1]. For acute ischemic stroke, which constitutes a multifactorial pathological process involving excitotoxicity, oxidative stress, blood–brain barrier (BBB) hyperpermeability, neuroinflammation, edema formation, and cell death, the basic premise of this approach is that mimicking or engaging the brain’s self-protective mechanisms could afford cerebroprotection and effective stroke therapy. One such mechanism identified by our group is peptidase neurolysin (Nln), which is viewed as a compensatory and cerebroprotective enzyme in the post-stroke brain and is being investigated as a potential drug target for acute, ischemic stroke therapy [2]. This focused opinion paper briefly discusses the multi-faceted function of Nln and its peptide substrates in the ischemic brain and details the progress made in development of small molecule Nln activators as a potential new class of multi-mechanism agents for the treatment of acute ischemic stroke.

2. Neurolysin, its peptide substrates, and their relevance to ischemic stroke

Nln (EC 3.4.24.16) is a monomeric zinc endopeptidase belonging to the M3 family of peptidases that was discovered and studied in detail by Frédéric Checler and his colleagues [3]. While abundant throughout the body, Nln is highly expressed in the cytosol and mitochondria of brain cells, and depending on the cell type, it is also found on the plasma membrane. The most characterized function of Nln is linked to the hydrolysis of extracellular neuropeptides among which neurotensin (NT), bradykinin (BK), and substance P (SP) are inactivated, whereas angiotensin I, dynorphin A(1–8), and metorphamide are converted to angiotensin-(1–7) and enkephalins, respectively. In addition, there is some evidence that angiotensin II, hemopressin, and somatostatin may also be inactivated by Nln [3,4]. As shown in numerous experimental and clinical studies, most of these extracellular substrates and their associated peptidergic systems are remarkably associated with the pathogenesis of acute ischemic stroke. More specifically, NT, BK, and SP are involved in enhancing BBB hyperpermeability, neuroinflammation, excitotoxicity, and oxidative stress in the ischemic brain, whereas angiotensin-(1–7) and enkephalins act to reverse these pathogenic mechanisms [2,5], uniquely positioning Nln to simultaneously inactivate several deleterious and generate cerebroprotective peptides. This hypothesized function also appears to be the reason for the observed upregulation of Nln one day after ischemic insult in primary cortical neurons [6,7] and stroke-affected mouse brain as a compensatory reaction [4]. Importantly, the cerebroprotective function of Nln was documented in a mouse ischemic stroke model using a bidirectional approach, in which brain overexpression of Nln via an AAV-driven vector has led to cerebroprotection, whereas its inhibition with a small molecule inhibitor has aggravated the stroke outcomes [5]. Currently, it is unclear whether hydrolysis of cytosolic and mitochondrial substrates of Nln contributes to its proposed cerebroprotective function. However, there is experimental evidence that Nln is also upregulated in mitochondria after ischemic stroke in mice [4] and that its mitochondrial substrates may play a role in pathogenesis of other neurodegenerative disorders [8]. Also, it is important to recognize that Nln is unlikely to have substantial function in neural repair processes in the chronic phase of stroke because its extracellular substrates do not function in a unidirectional way as they do in acute stroke [9]. Whether cytosolic or mitochondrial substrates of Nln play a role in neural repair mechanisms is unknown and should be the focus of future investigation.

3. Discovery of small molecule Nln activators

3.1. Identification of hit structures

Based on the hypothesized cerebroprotective function of Nln in the post-stroke brain, a substantial effort has been made in the last several years to identify and develop small, ‘drug-like’ molecules that can selectively enhance the catalytic activity of the peptidase. For this, the structure of Nln was explored by a computational approach to select a druggable surface pocket, followed by molecular docking and in silico screening of ~140,000 compounds, and identification of two dipeptides, L-histidyl-L-histidine and L-histidyl-L-tyrosine as potential Nln activators [10]. Further detailed in vitro pharmacological studies determined that these dipeptides increase Nln-mediated hydrolysis of a synthetic fluorescence substrate as well as endogenous substrates BK, NT, and angiotensin I in a concentration-dependent manner. Notably, the identified dipeptides enhanced the rate of synthetic substrate hydrolysis by both recombinant (human and rat) and mouse brain-purified Nln (micromolar A₅₀ and ≥ 300% A_max), while minimally affecting activity of closely related enzymes thimet oligopeptidase, neprilysin, angiotensin-converting enzyme (ACE), and ACE2 [10]. This was the first study documenting that the activity of Nln can be enhanced by small molecules and offered a chemical scaffold for development of high-potency, ‘drug-like’ Nln activators. Importantly, the binding site of Nln activators and the mechanism(s) leading to enhanced catalytic activity of this peptidase are poorly understood and are a subject of our ongoing studies.

3.2. Development of advanced hit peptidomimetic molecules

Recognizing the suboptimal plasma and brain stability of the identified dipeptides, and their low potency, a detailed structure–activity relationship study was conducted to develop higher potency peptidomimetic molecules with optimal drug-like properties, including optimal protein binding, BBB permeability, and metabolic stability [11,12]. By iterative excision of amino acid functionality, ~80 compounds were synthesized and tested leading to identification of three peptidomimetic compounds 9d, 10c, and 11a with more than 20-fold increase in Nln activation potency over the hit dipeptide molecules. Similar to the hit dipeptides, these compounds could substantially increase hydrolysis of both synthetic and endogenous substrates of Nln and had negligible effect on activity of closely related peptidases. Importantly, a set of in vitro experiments focusing on the ‘drug-like’ properties of these peptidomimetic compounds was also carried out revealing substantial improvement in plasma and brain stability and BBB permeability compared to the hit dipeptides [11]. Based on these findings, characterization of the peptidomimetic activators was extended to study their pharmacokinetic profile and brain penetration in mice after bolus intravenous administration, with additional experiments to determine their tissue binding and P-gp-mediated transport [13]. While in this study all three compounds were detectible in the mouse brain, 11a had the highest brain concentration and brain uptake clearance. As expected, the brain uptake of 11a was further increased in the ischemic hemispheres of mice in an experimental model of stroke, compared to the non-ischemic hemisphere [13]. Collectively, these detailed in vitro and in vivo pharmacokinetic studies confirmed the improved ‘drug-like’ properties and BBB permeability of 11a indicating that it can serve as an excellent chemical scaffold for further hit-to-lead optimization of these peptidomimetic Nln activators.

4. Expert opinion

Ischemic stroke remains a major therapeutic challenge with the only approved cerebroprotective treatment being restoration of blood flow via tissue plasminogen activator or endovascular thrombectomy. One reason for the ineffective drug discovery outcomes in this field is that almost all neuroprotective approaches developed to date have been based on one disease-one target-one drug paradigm. It is now largely recognized that the underlying mechanisms of stroke pathogenesis are complex networks of multiple pathways suggesting that multi-target approaches for stroke therapy may have a better chance to reach patients [14]. Among such approaches, one drug therapy engaging several therapeutic targets will likely be more advantageous over a multiple medication therapy, since it could preclude issues related to different pharmacokinetic properties and potential drug interactions of multiple agents, each of which modulates a specific therapeutic target. This would especially be the case when several overlapping or parallel pathways are being targeted. Therefore, as it appears to be by nature’s design, Nln is uniquely positioned to inactivate several potent and independent peptidergic systems which contribute to progression of stroke injury and to simultaneously facilitate activation of other peptidergic systems which function to protect the brain from ischemia (Figure 1). In other words, Nln can serve as a single pharmacological target to modulate function of multiple, independent mechanisms, and pathways that are critically involved in pathogenesis of stroke. Increasing activity of endogenous Nln by small molecules or delivery of recombinant Nln constructs are the main avenues to exploit this mechanism therapeutically during the acute phase of stroke and compensate for the long period (~24 h) required for cytosolic Nln to translocate to the plasma membrane and mitochondria. Considering issues related to brain delivery of a 75 kDa protein with no known transport mechanism to the brain, small molecule activators appear to be more favorable in this quest. In this regard, substantial progress has been made to identify small molecule Nln activators and characterize the phenomenon of Nln activation, elucidate the structure–activity relationship of His-dipeptides, develop advanced hit peptidomimetic Nln activators, and study their ‘drug-like’ properties, including in vivo pharmacokinetics and accessibility to the brain. Furthermore, our ongoing studies focus on development of bioisosteres from the current peptidomimetic molecules and detailed structure–activity relationship of newly identified non-peptide Nln activators, which have considerable potential to offset metabolic stability and BBB permeability challenges. The ultimate goal of these studies is to obtain high-potency, ‘drug-like’ Nln activators with optimal pharmacokinetic properties, which can reach the brain at pharmacological concentrations upon peripheral administration and enhance activity of the peptidase in the brain. These activator molecules will be used for detailed proof-of-concept studies to evaluate the effect of Nln activation on stroke outcome. If successful and further confirmed in larger, independent studies, small molecule Nln activators could indeed become a new class of multi-target drugs for treatment of ischemic stroke patients. In this scenario, it is conceivable that these drugs could also be investigated for the treatment of other neurological disorders (e.g. traumatic brain injury or neuropathic pain) in which the above-described Nln peptide substrates play critical roles in pathogenic mechanisms. Notably, key aspects of preclinical and clinical studies should be carefully considered in these efforts to successfully navigate the translational pipeline and minimize the chance of failures [15]. One question that is little explored but critical for these studies is related to the potential adverse effects of Nln activators. The closest relevant information is perhaps from studies with recombinant Nln, documenting negligible effects on basic physiological parameters in mice, and lack or minimal toxicity in cultured cells [16,17]. However, more detailed and expanded studies are warranted to understand the potential side effects of these agents before planning clinical trials.

Figure 1.

Peptidase neurolysin is a multi-mechanism target for acute ischemic stroke therapy. Neurolysin inactivates a number of cerebrotoxic peptides in the ischemic brain (e.g. SP, NT, BK) leading to attenuation of microvascular hyperpermeability, edema formation, neurogenic inflammation, oxidative stress, and cell death after stroke. In addition, neurolysin promotes formation of angiotensin (1–7) (Ang 1–7, from Ang I /angiotensin I/) and enkephalins (ENK, from dynorphin A1–8 /Dyn 1–8/ and metorphamide /Met/) in the post-stroke brain, leading to enhanced cerebroprotection. Small, ‘drug-like’ molecules that access the brain and enhance activity of neurolysin are currently being investigated as potential cerebroprotective agents for stroke therapy.