The Alzheimer's disease-linked protease BACE2 cleaves VEGFR3 and modulates its signaling.

The ß-secretase BACE1 is a central drug target for Alzheimer's disease. Clinically tested, BACE1-directed inhibitors also block the homologous protease BACE2. Yet, little is known about physiological BACE2 substrates and functions in vivo. Here, we identify BACE2 as the protease shedding the lymphangiogenic vascular endothelial growth factor receptor 3 (VEGFR3). Inactivation of BACE2, but not BACE1, inhibited shedding of VEGFR3 from primary human lymphatic endothelial cells (LECs) and reduced release of the shed, soluble VEGFR3 (sVEGFR3) ectodomain into the blood of mice, non-human primates and humans. Functionally, BACE2 inactivation increased full-length VEGFR3 and enhanced VEGFR3 signaling in LECs and also in vivo in zebrafish, where enhanced migration of LECs was observed. Thus, this study identifies BACE2 as a modulator of lymphangiogenic VEGFR3 signaling and demonstrates the utility of sVEGFR3 as a pharmacodynamic plasma marker for BACE2 activity in vivo, a prerequisite for developing BACE1-selective inhibitors for a safer prevention of Alzheimer's disease.

JNJ-67569762, A 2-Aminotetrahydropyridine-Based Selective BACE1 Inhibitor Targeting the S3 Pocket: From Discovery to Clinical Candidate.

The discovery of a novel 2-aminotetrahydropyridine class of BACE1 inhibitors is described. Their pKa and lipophilicity were modulated by a pending sulfonyl group, while good permeability and brain penetration were achieved via intramolecular hydrogen bonding. BACE1 selectivity over BACE2 was achieved in the S3 pocket by a novel bicyclic ring system. An optimization addressing reactive metabolite formation, cardiovascular safety, and CNS toxicity is described, leading to the clinical candidate JNJ-67569762 (12), which gave robust dose-dependent BACE1-mediated amyloid ß lowering without showing BACE2-dependent hair depigmentation in preclinical models. We show that 12 has a favorable projected human dose and PK and hence presented us with an opportunity to test a highly selective BACE1 inhibitor in humans. However, 12 was found to have a QT effect upon repeat dosing in dogs and its development was halted in favor of other selective leads, which will be reported in the future.

Seizure protein 6 and its homolog seizure 6-like protein are physiological substrates of BACE1 in neurons.

BACKGROUND: The protease BACE1 (beta-site APP cleaving enzyme) is a major drug target in Alzheimer's disease. However, BACE1 therapeutic inhibition may cause unwanted adverse effects due to its additional functions in the nervous system, such as in myelination and neuronal connectivity. Additionally, recent proteomic studies investigating BACE1 inhibition in cell lines and cultured murine neurons identified a wider range of neuronal membrane proteins as potential BACE1 substrates, including seizure protein 6 (SEZ6) and its homolog SEZ6L. METHODS AND RESULTS: We generated antibodies against SEZ6 and SEZ6L and validated these proteins as BACE1 substrates in vitro and in vivo. Levels of the soluble, BACE1-cleaved ectodomain of both proteins (sSEZ6, sSEZ6L) were strongly reduced upon BACE1 inhibition in primary neurons and also in vivo in brains of BACE1-deficient mice. BACE1 inhibition increased neuronal surface levels of SEZ6 and SEZ6L as shown by cell surface biotinylation, demonstrating that BACE1 controls surface expression of both proteins. Moreover, mass spectrometric analysis revealed that the BACE1 cleavage site in SEZ6 is located in close proximity to the membrane, similar to the corresponding cleavage site in SEZ6L. Finally, an improved method was developed for the proteomic analysis of murine cerebrospinal fluid (CSF) and was applied to CSF from BACE-deficient mice. Hereby, SEZ6 and SEZ6L were validated as BACE1 substrates in vivo by strongly reduced levels in the CSF of BACE1-deficient mice. CONCLUSIONS: This study demonstrates that SEZ6 and SEZ6L are physiological BACE1 substrates in the murine brain and suggests that sSEZ6 and sSEZ6L levels in CSF are suitable markers to monitor BACE1 inhibition in mice.

JNJ-40255293, a novel adenosine A2A/A1 antagonist with efficacy in preclinical models of Parkinson's disease.

Adenosine A2A antagonists are believed to have therapeutic potential in the treatment of Parkinson's disease (PD). We have characterized the dual adenosine A2A/A1 receptor antagonist JNJ-40255293 (2-amino-8-[2-(4-morpholinyl)ethoxy]-4-phenyl-5H-indeno[1,2-d]pyrimidin-5-one). JNJ-40255293 was a high-affinity (7.5 nM) antagonist at the human A2A receptor with 7-fold in vitro selectivity versus the human A1 receptor. A similar A2A:A1 selectivity was seen in vivo (ED50's of 0.21 and 2.1 mg/kg p.o. for occupancy of rat brain A2A and A1 receptors, respectively). The plasma EC50 for occupancy of rat brain A2A receptors was 13 ng/mL. In sleep-wake encephalographic (EEG) studies, JNJ-40255293 dose-dependently enhanced a consolidated waking associated with a subsequent delayed compensatory sleep (minimum effective dose: 0.63 mg/kg p.o.). As measured by microdialysis, JNJ-40255293 did not affect dopamine and noradrenaline release in the prefrontal cortex and the striatum. However, it was able to reverse effects (catalepsy, hypolocomotion, and conditioned avoidance impairment in rats; hypolocomotion in mice) produced by the dopamine D2 antagonist haloperidol. The compound also potentiated the agitation induced by the dopamine agonist apomorphine. JNJ-40255293 also reversed hypolocomotion produced by the dopamine-depleting agent reserpine and potentiated the effects of l-dihydroxyphenylalanine (L-DOPA) in rats with unilateral 6-hydroxydopamine-induced lesions of the nigro-striatal pathway, an animal model of Parkinson's disease. Extrapolating from the rat receptor occupancy dose-response curve, the occupancy required to produce these various effects in rats was generally in the range of 60-90%. The findings support the continued research and development of A2A antagonists as potential treatments for PD.

Synaptic transmission changes in fear memory circuits underlie key features of an animal model of schizophrenia.

Non-competitive antagonists of the N-methyl-d-aspartate receptor (NMDA) such as phencyclidine (PCP) elicit schizophrenia-like symptoms in healthy individuals. Similarly, PCP dosing in rats produces typical behavioral phenotypes that mimic human schizophrenia symptoms. Although schizophrenic behavioral phenotypes of the PCP model have been extensively studied, the underlying alterations of intrinsic neuronal properties and synaptic transmission in relevant limbic brain microcircuits remain elusive. Acute brain slice electrophysiology and immunostaining of inhibitory neurons were used to identify neuronal circuit alterations of the amygdala and hippocampus associated with changes in extinction of fear learning in rats following PCP treatment. Subchronic PCP application led to impaired long-term potentiation (LTP) and marked increases in the ratio of NMDA to 2-amino-3(5-methyl-3-oxo-1,2-oxazol-4-yl)propionic acid (AMPA) receptor-mediated currents at lateral amygdala (LA) principal neurons without alterations in parvalbumin (PV) as well as non-PV, glutamic acid decarboxylase 67 (GAD 67) immunopositive neurons. In addition, LTP was impaired at the Schaffer collateral to CA1 hippocampal pathway coincident with a reduction in colocalized PV and GAD67 immunopositive neurons in the CA3 hippocampal area. These effects occurred without changes in spontaneous events or intrinsic membrane properties of principal cells in the LA. The impairment of LTP at both amygdalar and hippocampal microcircuits, which play a key role in processing relevant survival information such as fear and extinction memory concurred with a disruption of extinction learning of fear conditioned responses. Our results show that subchronic PCP administration in rats impairs synaptic functioning in the amygdala and hippocampus as well as processing of fear-related memories.

Evidence that PGE2 in the dorsal and median raphe nuclei is involved in LPS-induced anorexia in rats.

Anorexia is an element of the acute-phase immune response. Its mechanisms remain poorly understood. Activation of inducible cyclooxygenase-2 (COX-2) in blood-brain-barrier endothelial cells and subsequent release of prostaglandins (e.g., prostaglandin E2, PGE2) may be involved. Therefore, we sought to relate the effects of prostaglandins on the anorexia following gram-negative bacterial lipopolysaccharide treatment (LPS) to neural activity in the dorsal and median raphe nuclei (DRN and MnR) in rats. COX-2 antagonist (NS-398, 10mg/kg; IP) administration prior to LPS (100µg/kg; IP) prevented anorexia and reduced c-Fos expression the DRN, MnR, nucleus tractus solitarii and several related forebrain areas. These data indicate that COX-2-mediated prostaglandin synthesis is necessary for LPS anorexia and much of the initial LPS-induced neural activation. Injection of NS-398 into the DRN and MnR (1ng/site) attenuated LPS-induced anorexia to nearly the same extent as IP NS-398, suggesting that prostaglandin signaling in these areas is necessary for LPS anorexia. Because the DRN and MnR are sources of major serotonergic projections to the forebrain, these data suggest that serotonergic neurons originating in the midbrain raphe play an important role in acute-phase response anorexia.

Subdiaphragmatic vagal deafferentation fails to block the anorectic effect of hydroxycitrate.

We investigated the neural mediation of the feeding suppression through orally administered hydroxycitrate (HCA) in male rats that were fed a high-glucose diet (about 48% glucose). Ten-day ad libitum food intake and body weight regain after previous body weight loss (13% of initial body weight) due to restrictive feeding were measured in rats with sham deafferentation (SHAM; n = 6), subdiaphragmatic vagal deafferentation (SDA; n = 7), and SDA plus celiac-superior mesenteric ganglionectomy (SDA/CGX; n = 9). HCA suppressed the 10-day cumulative food intake in all surgical groups and body weight regain in SDA and SDA/CGX groups. Independent of HCA, SDA and SDA/CGX rats consumed less food and gained less weight compared to SHAM rats. These results demonstrate that all vagal afferents from below the diaphragm and vagal efferents of the dorsal trunk as well as splanchnic nerves (afferents and efferents) are not necessary for the feeding-suppressive effect of HCA in this animal model. Vagal afferents, however, appear to play a role in the control of intake when a high-glucose diet is consumed after a period of restrictive feeding.

Evidence for a role of the 5-HT2C receptor in central lipopolysaccharide-, interleukin-1 beta-, and leptin-induced anorexia.

We examined the role of serotonin (5-HT) and the 5-HT(1A) and 5-HT(2C) receptors in the anorectic effects of centrally administered lipopolysaccharide (LPS), interleukin-1 beta (IL-1 beta), and leptin. Food intake was measured in rats after intracerebroventricular (ICV) injections of LPS (20 ng), IL-1 beta (10 ng), or leptin (1 microg) at lights out, followed by intraperitoneal (IP) injections of either the 5-HT(1A) autoreceptor agonist 8-hydroxy-2-(di-n-propylamino)tetraline (8-OH-DPAT) (125 microg/kg) or the 5-HT(2C) receptor antagonist SB 242084 (0.3 mg/kg) at the onset of anorexia. SB 242084 significantly attenuated the food intake reduction caused by all compounds (all P<.01). IP 8-OH-DPAT attenuated ICV IL-1 beta-induced anorexia (P<.01). We also tested the involvement of the median raphe 5-HT(1A) receptors in peripheral LPS- and IL-1 beta-induced anorexia. Rats were injected intraperitoneally with either LPS (100 microg/kg) or IL-1 beta (2 microg/kg) at lights out, and 8-OH-DPAT (4 nmol) was administered directly into the median raphe nucleus at the onset of anorexia. Median raphe injections of 8-OH-DPAT significantly attenuated both IL-1 beta- and LPS-induced anorexia (both P<.01). These results implicate the 5-HT(2C) receptors in the mediation of central LPS-, IL-1 beta-, and leptin-induced anorexia. Our results also suggest that the midbrain raphe nuclei play a role in mediating the anorectic response to peripheral LPS and IL-1 beta.

Evidence that the anorexia induced by lipopolysaccharide is mediated by the 5-HT2C receptor.

Rats consistently reduce their food intake following injections of bacterial lipopolysaccharides (LPS). Because inhibition of serotonergic (5-HT) activity by 8-OH-DPAT (5-HT(1A) activation) attenuates LPS-induced anorexia, we conducted a series of studies to examine whether other 5-HT-receptors are involved in the mediation of peripheral LPS-induced anorexia. In all experiments, rats were injected with LPS (100 microg/kg body weight [BW] ip) at lights out (hour 0). Antagonists were administered peripherally at hour 4, shortly after the onset of anorexia, which presumably follows the enhanced cytokine production after LPS. Food intake was then recorded during the subsequent 2 h or longer. 5-HT receptor antagonists cyanopindolol and SB 224289 (5-HT(1B)), ketanserin (5-HT(2A)), RS-102221 (5-HT(2C)), and metoclopramide (5-HT(3)) failed to attenuate LPS-induced anorexia. In contrast, both ritanserin (5-HT(2A/C)-receptor antagonist) (0.5 mg/kg BW) and SB 242084 (5-HT(2C)) (0.3 mg/kg BW) attenuated LPS-induced anorexia at doses that did not alter food intake in non-LPS-treated rats (all P<.01). Our results suggest that at least part of the anorexia following peripheral LPS administration is mediated through an enhanced 5-HT-ergic activity and the 5-HT(2C) receptor.

Hypothalamic implants of dilute estradiol fail to reduce feeding in ovariectomized rats.

To investigate further the site where estradiol (E(2)) inhibits food intake, we tested the effects on feeding of subcutaneous and intrahypothalamic implants of 10% E(2) benzoate in cholesterol (CHOL) or CHOL alone. E(2) was implanted subcutaneously in Silastic tubes, and intrahypothalamically via bilateral 29-gauge cannulas into the paraventricular nucleus (PVN) or the medial preoptic area (MPA) of ovariectomized (OVX) Sprague-Dawley and Long-Evans rats. Three-day implant periods followed 3-day baseline periods. Rats were allowed ad libitum access to chow and tap water, and food intake and body weight were measured each day. Subcutaneous 10% E(2) implants in Sprague-Dawley rats reduced food intake 21% on Day 2 and 34% on Day 3 (P's<.01) and decreased 3-day body weight gain 11 g (P<.05). In contrast, 10% E(2) implants in the PVN of Sprague-Dawley rats did not change food intake or body weight. Implants of 10% or 20% E(2) in the MPA also failed to decrease food intake. MPA implants of 10% E(2) decreased body weight gain 8 g (P<.05), but MPA implants of 20% E(2) decreased weight gain only 4 g (P>.05). To determine whether the strain of rat affected our negative results on food intake, we implanted 10% E(2) into the PVN of Long-Evans rats. Again, PVN E(2) did not decrease food intake significantly in comparison to the pretest baseline. PVN E(2) did, however, decrease body weight gain 5 g and decreased food intake 6% compared to rats with implants of CHOL (both P<.05), but these effects appeared to be due to an increase in feeding in the CHOL group in comparison to their baseline. Finally, CHOL and E(2) implants did not impair the responsivity of the PVN because acute implants of norepinephrine (NE) into the PVN of E(2)- or CHOL-treated Long-Evans rats significantly increased food intake. Our results do not support the hypothesis that E(2)'s actions in either the PVN or the MPA are sufficient to account for its inhibitory effects on feeding.