JLK Isocoumarin Inhibitors: Selective v- Secretase Inhibitors That Do Not Interfere With Notch Pathway In Vitro or In Vivo

Abstract

The enzymatic activity of γ-secretase is essential for the generation of amyloid beta (Aβ) peptides, a process closely associated with the progression of Alzheimer’s disease. Initially, inhibiting γ-secretase appeared to be a promising therapeutic strategy to reduce Aβ levels and potentially slow or halt disease development. However, the discovery that γ-secretase also processes other critical transmembrane proteins, including Notch, which is crucial for various biological functions in both embryonic development and adult life, raised significant concerns regarding the viability of this approach. Previously, a new class of non-peptidic compounds was identified that selectively inhibits Aβ production in vitro without disrupting the Notch signaling pathway. The current study extends these findings by demonstrating that, in vivo, these inhibitors do not interfere with the Notch pathway responsible for somite formation in zebrafish embryos. Further, the selectivity of these compounds, referred to as JLK inhibitors, is confirmed by showing that they do not affect the processing of E-cadherin, a substrate impacted by other γ-secretase inhibitors. Additionally, JLK inhibitors were found not to interfere with β-site APP cleaving enzymes (BACE1 and BACE2), α-secretase activity, the proteasome system, or GSK3β kinase. These findings position JLK inhibitors as the only known agents capable of selectively reducing Aβ production without undesired effects on the processing of other proteins.

Keywords: γ-secretase, inhibitors, amyloid beta peptide, Notch, NICD, presenilins, Alzheimer’s disease, protein processing, cadherins, proteasome, GSK3β, BACE

Introduction

Amyloid beta (Aβ) peptides, which progressively accumulate during Alzheimer’s disease, are produced through proteolytic processing of the β-amyloid precursor protein (βAPP). This process involves sequential cleavage by two enzymes, β-secretase and γ-secretase. The γ-secretase enzyme determines the C-terminal end of Aβ peptides, a factor that significantly influences their biophysical characteristics. Among these peptides, Aβ42 is particularly aggregation-prone and appears early in disease progression, contributing significantly to neurodegeneration.

Due to its pivotal role in Aβ peptide production, γ-secretase emerged as a potential therapeutic target. Presenilin 1 and 2 (PS1 and PS2), two proteins genetically linked to familial forms of Alzheimer’s disease, are believed to be involved in the γ-secretase complex. Although there is substantial evidence supporting their involvement, the exact enzymatic nature of presenilins and whether they constitute the sole γ-secretase remain subjects of scientific debate. Notably, Aβ production has been observed in cells lacking presenilins, suggesting the existence of both presenilin-dependent and presenilin-independent γ-secretase pathways. The complete characterization of γ-secretase is still pending definitive clarification.

Several γ-secretase inhibitors have been developed to suppress Aβ production. Many of these compounds act by targeting presenilins and have successfully reduced Aβ levels in vitro. However, these agents often lack substrate selectivity and also inhibit the processing of several other important transmembrane proteins, such as Notch, ErbB-4, nectin 1α, and E-cadherin. Inhibition of these proteins can disrupt essential physiological processes because their biological activity depends on regulated proteolytic cleavage.

A notable example is the Notch protein, which is cleaved to produce the Notch intracellular domain (NICD). This domain translocates to the nucleus, where it functions as a transcription factor regulating genes critical for embryonic development and adult tissue maintenance. Therefore, non-selective γ-secretase inhibition can have harmful effects.

Previously, a group of non-peptidic γ-secretase inhibitors was identified that did not interfere with presenilins but effectively blocked Aβ production. These agents were also shown to preserve NICD production in vitro, indicating that the Notch pathway remains intact. The present study confirms this finding in vivo using a zebrafish model of somite segmentation. Additionally, these inhibitors were further analyzed for specificity and shown not to affect the activity of α-secretase, β-secretase (BACE1 and BACE2), the proteasome, or GSK3β kinase. This distinct profile supports their therapeutic potential as selective γ-secretase inhibitors that reduce Aβ production without disturbing other essential cellular processes.

Materials and Methods

Aβ and sAPPα Recovery and Detection

Human embryonic kidney (HEK293) cells, including both wild-type and Swedish-mutated APP-expressing lines, were utilized to evaluate the effect of JLK inhibitors on amyloid beta production. Cells were preincubated for seven hours either in the absence or presence of JLK inhibitors. Aβ42 and Aβ40 peptides were subsequently immunoprecipitated using specific antibodies and analyzed through established immunodetection methods. Soluble APPα (sAPPα) was collected from the same conditioned media and partially immunoprecipitated using designated antibodies for additional analysis.

In Vitro and In Vivo Notch Assays

To assess the impact of JLK inhibitors on Notch signaling, HEK293 cells expressing a modified Notch construct were used for in vitro analysis. For in vivo experiments, zebrafish embryos were obtained from natural spawning, manually dechorionated, and maintained at 28 degrees Celsius in 30% Danieau’s solution. Treatments with either a known γ-secretase inhibitor (DAPT), JLK compounds, or a vehicle control (DMSO) were initiated at the sphere stage and continued through the somite development stages. After treatment, embryos were fixed in paraformaldehyde and subjected to whole-mount in situ hybridization using a digoxygenin-labeled RNA probe specific for the deltaC gene, a marker of active Notch signaling. This method followed standard procedures for examining developmental gene expression in zebrafish embryos.

E-Cadherin Cleavage

HEK293 cells were washed with phosphate-buffered saline and resuspended in an ice-cold hypotonic buffer containing MOPS and potassium chloride. The cells were homogenized on ice using a tight-fitting homogenizer. Post-nuclear supernatants were obtained through centrifugation, followed by isolation of crude membrane fractions by higher-speed centrifugation. Membranes were resuspended in an assay buffer with protease inhibitors and incubated at 37 degrees Celsius for two hours in the presence or absence of JLK compounds or a known γ-secretase inhibitor. Samples were analyzed by Western blot to assess E-cadherin processing.

BACE1 and BACE2-Expressing Cells and β-Secretase Assay

HEK293 cells were stably transfected with constructs encoding 1D4-tagged BACE1 or BACE2 proteins. Transfection was performed according to the manufacturer’s guidelines, and expression of BACE proteins was confirmed through gel electrophoresis, Western blot analysis, and immunodetection using antibodies specific to the 1D4 epitope. Positive clones showed expression of 75 kDa and 50 kDa proteins corresponding to BACE1 and BACE2, respectively. The functional β-secretase activity in these cells was assessed using a fluorometric assay as described in prior studies.

Proteasome Assay

Purified 20S proteasome from bovine pituitary, with a specific activity based on p-nitroaniline release, was used to evaluate the effect of JLK inhibitors on proteasome function. A fluorogenic substrate, Z-Gly-Gly-Leu-7-amino-4-methylcoumarin, was incubated with the enzyme in a Tris-HCl buffer containing magnesium chloride, dithiothreitol, and EDTA. Reactions were conducted at 37 degrees Celsius with or without JLK inhibitors. Proteasome activity was measured by monitoring fluorescence at defined excitation and emission wavelengths.

GSK3β Assay

The activity of glycogen synthase kinase 3 beta (GSK3β) was assessed using a luciferase-based reporter assay that responds to β-catenin signaling. The 293T cells used in this assay were stably transfected with the LEF-OT reporter construct. Lithium chloride served as a positive control and elicited strong reporter activation. This method follows previously established protocols.

Results and Discussion

JLK Inhibitors Prevent Aβ40 and Aβ42 Recovery From Cells Expressing Wild-Type and Swedish-Mutated APP

The effects of JLK2 and JLK6 inhibitors on Aβ40 and Aβ42 production were tested in HEK293 cells overexpressing either wild-type or Swedish-mutated APP. Both inhibitors demonstrated a dose-dependent reduction in the levels of secreted Aβ40 and Aβ42 in these cellular models. The half-maximal inhibitory concentration (IC50) values for JLK2 and JLK6 were approximately 30 micromolar for both Aβ species, suggesting similar inhibitory potency. The consistent IC50 values across the different APP variants imply that both peptides may be generated by the same proteolytic enzyme or by enzymes with similar affinities for the inhibitors. Previous findings support the notion that JLK compounds inhibit γ-secretase activity, as their action protects intermediate APP cleavage products, C83 and C99, which are typically processed by γ-secretase.

JLK Inhibitors Do Not Affect Notch Pathways In Vitro and In Vivo

The selectivity of JLK compounds was examined by evaluating the Notch signaling pathway in both in vitro and in vivo settings. In cultured HEK293 cells expressing a truncated Notch construct, the production of the Notch intracellular domain (NICD) was observed as a result of γ-secretase-mediated cleavage. Unlike the γ-secretase inhibitor DFK167, which effectively blocked NICD generation, JLK inhibitors did not impair NICD production. This indicates that JLK compounds do not interfere with Notch processing in vitro.

In vivo studies using zebrafish embryos focused on the expression patterns of deltaC in the presomitic mesoderm, which serve as markers of active Notch signaling. Treatment with the γ-secretase inhibitor DAPT disrupted deltaC expression, resulting in a disorganized “salt-and-pepper” pattern typical of impaired Notch signaling. Conversely, embryos treated with the JLK2 compound maintained normal stripe-like deltaC expression in 90 percent of cases, demonstrating preserved Notch pathway activity. These observations support the in vitro data and confirm that JLK inhibitors selectively block amyloid-beta (Aβ) production without compromising Notch signaling during embryonic development.

JLK Inhibitors Do Not Affect E-Cadherin Processing

E-cadherin is a transmembrane protein essential for cell adhesion, undergoing a cleavage process similar to that mediated by γ-secretase. This processing is characterized as γ-secretase-dependent because it is prevented by γ-secretase inhibitors and abolished in the absence of presenilin. However, unlike typical γ-secretase-mediated proteolysis occurring within the transmembrane domain, E-cadherin cleavage appears to take place at the edge of the inner membrane leaflet. This cleavage plays a critical role in regulating the assembly of adherens junctions and controlling intracellular β-catenin levels, thereby influencing the Wnt signaling pathway.

Unlike other γ-secretase inhibitors such as L-685,458, JLK7 does not inhibit the γ-secretase-mediated cleavage of E-cadherin C-terminal fragments (CTF2). This suggests that JLK inhibitors do not interfere with E-cadherin processing, preserving its functional role in cellular adhesion and signaling.

JLK Inhibitors Do Not Affect BACE1, BACE2, and the α-Secretase Pathway

An important aspect of secretase inhibitor development is ensuring selectivity, especially given that amyloid precursor protein (βAPP) undergoes cleavage by α- and β-secretases that directly regulate Aβ production. Inhibition of the α-secretase pathway generally leads to increased Aβ production, so nonselective inhibition of α-secretases could have adverse effects.

The effects of JLK inhibitors on the endogenous α-secretase pathway were investigated, revealing that compounds JLK2 and JLK6 do not alter the levels of secreted APPα. Additionally, fluorimetric assays measuring the activity of ADAM10 and ADAM17, the main α-secretase candidates, confirmed that JLK inhibitors do not affect α-secretase activity in vitro.

BACE1 is considered the primary β-secretase involved in Aβ production, with BACE1 deficiency abolishing Aβ production in embryonic cortical neurons and mouse brains. BACE2, a related enzyme, does not significantly contribute to typical Aβ production but may be involved in generating alternative truncated toxic Aβ species. Two stable cell lines overexpressing either BACE1 or BACE2 were established, and fluorimetric assays demonstrated that JLK2, JLK6, and JLK7 did not affect the in vitro activity of these enzymes. These findings align with data from cells expressing the C99 βAPP fragment, which produces Aβ solely through γ-secretase cleavage without requiring prior β-secretase activity. JLK compounds inhibit Aβ production from C99-expressing cells, confirming that γ-secretase is the only secretase targeted by JLK inhibitors among the secretase family.

JLK Inhibitors Do Not Affect the Proteasome

The proteasome is a multicatalytic complex involved in various cellular functions, including the regulation of cell death. Inhibitors of the proteasome, such as lactacystin, can induce proapoptotic effects. Proteasome inhibition has significant implications for neurodegenerative diseases, as protein aggregates that accumulate in these conditions can impair proteasome function, disrupting normal protein degradation. This phenomenon has been specifically associated with Alzheimer’s disease. For instance, the calpain/proteasome inhibitor MG132 increases Aβ levels, and the proteasome is known to degrade presenilin proteins, influencing βAPP processing.

Studies showed that proteasome activity is completely unaffected by JLK inhibitors in vitro. This finding is consistent with previous data indicating that JLK compounds do not alter proteasomal degradation of endogenous or overexpressed presenilins, suggesting no interference with proteasome function.

JLK Inhibitors Do Not Affect GSK3β Activity

Glycogen synthase kinase 3 beta (GSK3β) activity was assessed using a luciferase reporter assay that responds to β-catenin signaling. β-Catenin, a substrate of GSK3β, is phosphorylated and subsequently degraded; when GSK3β is inhibited, phosphorylation and degradation are prevented, allowing β-catenin to accumulate and influence transcription. Lithium, known to inhibit GSK3β, was shown to reduce Aβ production without affecting the Notch pathway, mimicking some effects observed with JLK compounds.

In this assay, lithium served as a positive control and robustly activated the reporter. However, none of the JLK compounds induced reporter activity, indicating that JLK inhibitors do not affect GSK3β activity in vivo.

Conclusion

JLK compounds represent a unique class of γ-secretase inhibitors that selectively target the γ-secretase pathway involved in βAPP processing without affecting other γ-secretase-mediated pathways. This high degree of selectivity is a critical attribute, arguably more important than absolute potency. JLK inhibitors are valuable tools for both in vitro experiments and certain in vivo models, such as zebrafish embryos. Ongoing studies in animal models of Alzheimer’s disease aim to further assess the ability of JLK compounds to reduce Aβ accumulation in vivo.