PDF file, 51K, Ratio to baseline for plasma VEGF from the days 2 or 28 time points for all animals administered a single dose of MNRP1685A at varying dose levels. Each dose level contained n=4 animals. Plasma VEGF was measured using an MSD assay.

PDF file, 57K, Ratio of plasma PlGF or cNRP1 to baseline for each subject at all time points evaluated (Days 14, 21) from the Phase Ia study.

PDF file, 72K, Binding experiments were carried out by surface plasmon resonance (SPR) measurements on a ProteOn XPR36 (Bio-Rad Laboratories) instrument at 25{degree sign}C. For the binding competition experiment, PlGF (20μg/ml) was immobilized at high surface densities (3000-4000 RU) on an activated ProteOn GLC sensor chip using standard amine coupling procedures as described by the manufacturer. Analytes (Nrp1, Anti-Nrp1B) and 1:1 molar ratio mixture of analytes (Nrp1 + Anti-Nrp1B) were injected in phosphate-buffered saline (PBS), 0.005% v/v Tween-20 (pH7.4) at a flow rate of 100μl/min and sensorgrams for association and disassociation phases were recorded. Analytes were injected for 150 s and allowed to disassociate for 450 s. Data was processed with the ProteOn Manager software (version 2.0, Bio-Rad).

PDF file, 57K, Baseline levels of Plasma PlGF for the Phase I study with MNRP1658A and Avf0776 were measured using MSD assay. Equivalent levels and similar distribution of PlGF was observed between the two clinical trials.

This paper investigates the characteristics of highly viscous non-Newtonian gelled propellant sprays formed by internal impingement of micro air jets on the propellant stream. The newly developed atomizer is an improved alternative to the externally impinging jet designs, in that the high-momentum micro jets of atomizing air located internally and upstream of the atomizer exit orifice continuously impart high shear to the oncoming gelled propellant jet. Breakup and atomization is achieved within a short time and at a much smaller supply pressure of atomizing air than the other atomizers considered for this application. Spray characterization studies reveal that the Sauter mean droplet size of gelled Jet A1 spray under normal ambient conditions is smaller than 50 μm at gas-liquid mass ratios (GLR) that are also smaller than other atomizers. High-speed visualization and droplet size measurements reveal relatively larger droplets by film breakup at low atomizing air mass flow rate. The prompt breakup mode induced at high atomizing air mass flow rate is responsible for generation of smaller sized droplet population clustered in chevron structures and desirable for ignition and flame stabilization. The strong dependence of viscosity of gelled Jet A1 on the gel mass flow rate and shear induced by micro air jets help in achieving similar spray characteristics as that of Newtonian ungelled Jet A1 at low GLR. Dimensional analysis of the breakup process yielded Ohnesorge number, momentum flux ratio, and GLR as possible correlating parameters for SMD and the experimental droplet size data of gelled Jet A1 sprays is reduced in the form of an empirical correlation.

Alzheimer's disease (AD) is characterized by amyloid plaques, neurofibrillary tangles, and synaptic and neuronal loss. The mechanism of neuron death in AD, however, remains unexplored. Recently, a rare autosomal dominant coding mutation, T835M, was discovered in the Un-coordinated 5c (Unc5c) netrin receptor gene that segregated with late-onset AD (LOAD). This mutation leads to cell death in HEK-293T cells and reduces survival in the presence of neurotoxic stimuli in cultured primary neurons[1]. Combining this result with the robust expression of Unc5c in hippocampus, we hypothesize that UNC5C T835M mutation predisposes to LOAD by making neurons more vulnerable to cell death induced by pathogenic Aβ and Tau and Unc5c death domain activation. We employed the mouse knock in (KI) model of Unc5c T85M that were crossed with 5XFAD amyloid mouse model. Brain sections obtained from Unc5cKI/KI;5XFAD mice and littermate controls were imaged by immunofluorescence confocal microscopy for neurons (NeuN), Aβ deposits (Thiazine Red, Aβ42, Aβtotal), apoptosis (TUNEL), astrocytes (GFAP), and microglia (Iba1). We also employed unbiased proteomics. In vitro studies using primary neuronal culture obtained from KI and wildtype (WT) control mice were also employed to study the effects of various stressors including Aβ42, staurosporine and glutamate. Overall, our studies show that homozygous KI mice are very similar to WT littermate controls in terms of the histology, protein and RNA expression or in cell death. However, proteomics analysis of KI and wildtype (WT) mice brains showed upregulation of apoptotic proteins and down-regulation of neuronal proteins. Moreover, when neurons were counted in NeuN-stained brain sections of Unc5cKI/KI;5XFAD mice, we observed that ∼40% of neurons were lost in cortical layer 5 of Unc5cKI/KI;5XFAD compared to UNC5C+/+;5XFAD mice. Neuron loss in Unc5cKI/KI;5XFAD mice correlated strongly with the presence of Aβ deposits in layer 5. Primary KI neurons showed an increased cell death in the presence of cytotoxic stressors. We anticipate that our study will lead to a greater understanding of the underlying molecular pathway involved in AD-related neuron loss and how Unc5c mediates AD risk, and identify novel therapeutic targets for reducing neuron loss in AD and potentially other neurodegenerative diseases.

Spatial spread of tau pathology in Alzheimer's disease correlates with cognitive decline. Recent studies provide growing evidence that the spread of tau pathology in Alzheimer's disease is mediated through the extracellular environment. Tau antibodies can potentially intercept extracellular tau and block cell-to-cell spread. Conventional full effector function tau antibodies may promote tau clearance via microglia engulfment, which may enhance microglial activity. We investigated whether effector function is required for blocking the spread of pathological tau through antibody treatment. Effector function is mediated by antibody interactions with immune cells via Fcγ receptors (FcγRs) and complement that can be modulated by mutations in the IgG Fc region. We compared the effects of two antibodies directed against tau phosphorylated on serine 409 (pS409-Tau), one with full effector function and the other effectorless, on tau pathology, neuron health and microglial function. Both antibodies blocked the spread of tau pathology in Tau-P301L transgenic (Tg) mice in vivo as measured by immunohistochemistry and Western blotting. Similarly, both antibodies protected cultured neurons against extracellular tau-induced toxicity. However, in cultured microglia only the full-effector antibody enhanced tau uptake, and promoted release of proinflammatory cytokines. In neuron-microglia co-cultures, only effectorless anti-tau protected neurons, suggesting full-effector tau antibodies can induce indirect toxicity via microglia. We conclude that effector function is not required for efficacy in vivo, and that effectorless tau antibodies may represent a safer approach to targeting the spread of tau pathology by bypassing potential deleterious inflammatory consequences of antibody treatment, while still preserving the ability to limit the spread of toxic proteins in vivo.

Inhibiting the amyloidogenic processing of the amyloid precursor protein (APP) by blocking the activity of the β-site APP cleavage enzyme (BACE) has been extensively pursued as a therapeutic strategy for Alzheimer's disease (AD). Indeed, brief treatment with high doses of anti-BACE1 antibodies effectively reduced peripheral and central Aβ production in wild-type mice and nonhuman primates (Atwal et al., STM 2011), suggesting a potential therapeutic role for this antibody. However, it is still unclear what level of Aβ reduction could translate to a therapeutically meaningful impact on the progression of disease pathology if treatment were continued chronically. Therefore, to test the effectiveness of the anti-BACE1 antibodies for reducing/preventing A β plaque deposition and associated pathology, we treated transgenic mice expressing London mutant (V717I) human APP (hAPPlon) for four months starting at an age prior to the onset of plaque deposition. Female transgenic hAPPlon mice or non-transgenic littermate controls received either short-term treatment (3 IP injections, each separated by 4 days), with control antibody or anti-BACE1 antibodies (30mg/kg, 100mg/kg) starting at 2–3 months-of-age, or chronic treatment (weekly IP injections for 16 weeks) of a control antibody or anti-BACE1 antibody (100mg/kg), starting at 10.5 months-of-age. Serum, plasma and brain tissue was collected for PK/PD and pathology analysis. We found that short-term anti-BACE1 treatment (100mg/kg) modestly reduced guanidine-extracted total brain A β 40 and A β 42 levels by 17% and 31%, respectively, compared with control-treated hAPPlon mice when measured by ELISA. Chronic anti-BACE1 treatment significantly reduced A β plaques by ∼50% throughout the brain, compared with control-treated hAPPlon mice. This was associated with a reduction in total brain A β 40 and A β 42 levels by 73% and 55%, respectively. Plaque-associated neuroinflammation, as measured by microglia and astrocyte activation, was also significantly reduced. Modest but chronic inhibition of BACE1 activity with high affinity anti-BACE1 antibodies can significantly reduce A β plaque deposition and associated pathology in a mouse model of AD. This suggests that partial inhibition of BACE1 could be an effective therapeutic strategy in human AD patients as well.

Much of the heritability of late-onset Alzheimer's disease (LOAD) remains unaccounted for, despite progress from genome-wide association studies. The overall contribution of rare variants to the risk of LOAD remains to be determined and, to date, few rare variants in LOAD have been identified and functionally characterized. We performed exome and whole-genome sequencing in multiplex AD family, in which we previously discarded mutation in APP, PSEN1 and 2. We identified a rare variant that segregates in families with LOAD, is associated with disease in large case/control studies, and leads to increased neuronal cell death. A missense variant in UNC5C segregated with disease in a large pedigree with autosomal dominant inheritance of LOAD. To confirm the association, we screened a collection of families enriched for LOAD and identified additional families where the variant segregates with disease. The variant was also enriched in large collections of LOAD cases compared to controls (metaP = 0.006, OR = 2.15). The variant alters an evolutionarily conserved residue in the hinge region of UNC5C and leads to increased cell death in vitro. These studies identified a novel pathway in AD pathogenesis and suggest that a strategy that combines pedigree analysis with case/control studies will be valuable in the identification of rare variants in complex disease.

Several rare genetic variants cause early-onset Alzheimer's disease(EOAD), including mutations in amyloid-β precursor protein (APP), presenilin-1 (PSN1), and presenilin-2 (PSN2). However, these variants account for less than 5% of all AD cases. The overall contribution of rare variants to the risk of late-onset Alzheimer's disease (LOAD) is unknown and, to date, few rare variants in LOAD have been identified and functionally characterized. To elucidate novel risk variants, we performed linkage analysis and whole genome sequencing in a large pedigree with apparent dominant inheritance of LOAD and identified a candidate rare variant in the Netrin receptor UNC5C (rs137875858). This variant segregates with disease in two families enriched for LOAD and shows association with LOAD in four independent case/control cohorts. The consequence of this variant is a single amino acid change from a highly conserved threonine residue, T835, to a methionine. T835M is located in the hinge region of UNC5C just upstream of the death domain, which may affect protein structure and death domain access. UNC5C plays a role in axonal guidance during development and is highly expressed in the adult hippocampus and cerebellum. To characterize the functional role of T835M in LOAD, we expressed T835M in 293T cells and found that overexpression of this variant leads to growth defects caused by enhanced apoptosis. Moreover, T835M-expressing hippocampal neurons showed increased vulnerability when treated with Aβ, suggesting that neurons in T835M-carriers are more susceptible to cell death in the presence of amyloid. These studies implicate an important role for UNC5C in regulating neuronal cell death, and identify a novel pathway that may contribute to Alzheimer's disease pathogenesis.

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease that causes progressive paralysis due to motor neuron death. Several lines of published evidence suggested that inhibition of epidermal growth factor receptor (EGFR) signaling might protect neurons from degeneration. To test this hypothesis in vivo, we treated the SOD1 transgenic mouse model of ALS with erlotinib, an EGFR inhibitor clinically approved for oncology indications. Although erlotinib failed to extend ALS mouse survival it did provide a modest but significant delay in the onset of multiple behavioral measures of disease progression. However, given the lack of protection of motor neuron synapses and the lack of survival extension, the small benefits observed after erlotinib treatment appear purely symptomatic, with no modification of disease course.

Boosting antibody uptake in brain would substantially improve our ability to modulate various CNS drug targets. To better understand antibody properties relative to the blood-brain barrier (BBB) and as a potential Alzheimer's therapeutic intervention, we developed an antibody that targets BACE1. By utilizing a receptor-mediated transcytosis approach targeting the transferrin receptor (TfR), we explored how we could enhance uptake of our therapeutic antibody across the BBB. A function-blocking antibody targeting BACE1 (anti-BACE1) was developed and characterized in vitro. A bispecific antibody with low affinity to TfR and with high affinity to BACE1 (anti-TfR/BACE1) was generated to enhance brain uptake. Brain antibody levels and Abeta reduction were evaluated following systemic delivery of anti-BACE1 or anti-TfR/BACE1 in mice. Anti-BACE1 is a highly selective, potent antagonist of BACE1 that inhibits BACE1 activity by binding to an exosite. Anti-BACE1 reduces Abeta production in cultured human cell lines expressing APP and in primary mouse neurons. Systemic dosing of mice with anti-BACE1 resulted in sustained reductions in peripheral Abeta concentrations. High doses of anti-BACE1 also transiently reduced CNS Abeta concentrations. Enhancing brain uptake of anti-BACE1 by utilizing a bispecific anti-TfR/BACE1 resulted in substantially improved brain exposure leading to a greater and more sustained reduction in brain Abeta with lower dose levels than anti-BACE1 alone. BACE1 can be targeted in a highly selective manner through passive immunization with BBB penetrating anti-TfR/BACE1 antibodies. These data also provide proof-of-concept for brain penetrating antibodies, expanding our ability to develop CNS antibody therapies.

Abstract Purpose: MNRP1685A is a monoclonal antibody to neuropilin-1 (NRP1). We evaluated blood-based pharmacodynamic biomarkers of MNRP1685A in two phase I studies to assess exposure/response relationships to inform target dose and regimen selection. Experimental Design: The phase I studies evaluated escalating doses of MNRP1685A as a single agent or in combination with bevacizumab. Plasma placental growth factor (PlGF), VEGF, and circulating NRP1 (cNRP1) were evaluated at multiple time points using meso-scale discovery (MSD) assays and ELISA, respectively. Plasma PlGF was also measured in a phase I/II trial of bevacizumab in metastatic breast cancer (AVF0776). The association between PlGF and MNRP1685A dose was described by a sigmoid Emax model. cNRP1 and MNRP1685A PK profiles were described using a two-target quasi-steady state (QSS) model. Results: A dose- and time-dependent increase in plasma PlGF and cNRP1 was observed in all patients treated with MNRP1685A. PK/PD analysis showed that bevacizumab and MNRP1685A had an additive effect in elevating PlGF. Predictions based on the two-target QSS model showed that the free drug concentration to maintain greater than 90% saturation of membrane NRP1 (mNRP1) and cNRP1 is about 8 μg/mL. Conclusion: These data show that MNRP1685A inhibits the VEGF pathway in humans as assessed by an increase in plasma PlGF. MNRP1685A seems to enhance bevacizumab-mediated VEGF pathway blockade, as showed by an increase in the magnitude of PlGF elevation when combined with bevacizumab. PK/PD analysis of biomarkers in the phase I population allowed identification of doses at which apparent maximal pathway modulation was observed. Clin Cancer Res; 18(21); 6040–8. ©2012 AACR.

The imbalance between Aβ production and/or clearance is hypothesized to be the major contributor to pathogenesis in Alzheimer's disease (AD). Inhibiting the sequential proteolytic cleavage of amyloid precursor protein (APP) by β- or γ-secretase blockade has been extensively pursued as a therapeutic strategy for AD treatment. The β-site APP cleavage enzyme (BACE) remains a popular target for small molecule inhibitors and antibodies because of its rate-limiting role in Aβ production. Targeting BACE1, versus its close homolog BACE2, is favored because BACE1 is highly expressed in neurons of the brain, whereas BACE2 is not. Although much is known about the role of BACE1 and BACE2 in processing human APP in in vitro and in vivo over-expression systems, less is known about their role in processing endogenous mouse APP in the brain and periphery. To determine the contribution of BACE1 and BACE2 to endogenous Aβ production in the brain and periphery, we measured brain and plasma Aβ40 levels by ELISA from BACE1ko, BACE2ko, and BACE1/2 double ko mice. Endogenous Aβ40 levels were measured in wild-type or BACE1ko mice to determine the selectivity of anti-BACE1 antibodies and small molecules against BACE1. In this study, we demonstrated that BACE1 contributes to ∼90% of endogenous brain and 50-60% of plasma Aβ40 production, whereas BACE2 does not contribute to either brain or plasma Aβ40 production. APPko mice had no detectable Aβ40 levels, demonstrating the specificity of the assay. Transient handling-induced increases in peripheral Aβ40 are mediated by BACE1, not BACE2. Anti-BACE1 antibodies are selective in inhibiting BACE1 activity, since no further reductions in Aβ40 levels were observed in BACE1 ko mice. BACE1 is the major β-secretase involved in the production of endogenous mouse Aβ in the brain and periphery, whereas BACE2 has little to no role. The remaining peripheral Aβ production in BACE1/2 ko mice suggests that there may be additional β-secretases yet to be identified.

MABT5102A, a monoclonal antibody (MAb) that binds to beta-amyloid (Abeta), is being evaluated as a disease modifying therapeutic in patients with Alzheimer's disease (AD). In order to gain an understanding of its mechanism of action, the pharmacokinetics (PK), pharmacodynamics (PD) and distribution of MABT5102A were studied in a mouse model of AD and in cynomolgus monkey. MABT5102A PK was studied in human amyloid precursor protein-transgenic (Tg) mice, non-Tg mice and in cynomolgus monkeys following single or multiple dose IV administration. MAb serum concentrations were assayed with an Abeta-coated ELISA. Distribution studies were conducted in mice with I123- or I125-labeled MAbs. The PK of MABT5102A showed dose proportionality over the dose range tested in both species. Mean clearance (CL) of MABT5102A was 28 mL/day/kg in hAPP-Tg mice, nearly two-fold of that observed in non-Tg mice. In monkeys, the mean CL was 4 mL/day/kg. In both species, dose dependent increases in plasma total Abeta levels were observed. A mechanism-based PK/PD model linking MABT5102A disposition with plasma total Abeta levels was able to adequately describe the PKPD profiles in both species. No major differences in the tissue distribution of MABT5102A were observed between Tg and non-Tg mice. The uptake of MABT5102A into brain and CSF represented <0.2% wt/wt of the administered dose. The PK parameters observed for MABT5102A in normal mice and cynomolgus monkeys were as expected in these species. A PKPD model utilizing plasma Abeta as a PD marker was developed to determine the effects of different dosing regimens on the Abeta profile. The low uptake of antibody into the brain is consistent with other reported studies.

Targeted inactivation of genes involved in murine cardiovascular development frequently leads to abnormalities in blood flow. As blood fluid dynamics play a crucial role in shaping vessel morphology, the presence of flow defects generally prohibits the precise assignment of the role of the mutated gene product in the vasculature. In this study, we show how to distinguish between genetic defects caused by targeted inactivation of the neuropilin 1 (Nrp1) receptor and hemodynamic defects occurring in homozygous knockout embryos. Our analysis of a Nrp1 null allele bred onto a C57BL/6 background shows that vessel remodeling defects occur concomitantly with the onset of blood flow and cause death of homozygous mutants at E10.5. Using mouse embryo culture, we establish that hemodynamic defects are already present at E8.5 and continuous circulation is never established in homozygous mutants. The geometry of yolk sac blood vessels is altered and remodeling into yolk sac arteries and veins does not occur. To separate flow-induced deficiencies from those caused by the Nrp1 mutation, we arrested blood flow in cultured wild-type and mutant embryos and followed their vascular development. We find that loss of Nrp1 function rather than flow induces the altered geometry of the capillary plexus. Endothelial cell migration, but not replication, is altered in Nrp1 mutants. Gene expression analysis of endothelial cells isolated from freshly dissected wild-type and mutants and after culture in no-flow conditions showed down-regulation of the arterial marker genes connexin 40 and ephrin B2 related to the loss of Nrp1 function. This method allows genetic defects caused by loss-of-function of a gene important for cardiovascular development to be isolated even in the presence of hemodynamic defects.

Targeted inactivation of genes involved in murine cardiovascular development frequently leads to abnormalities in blood flow. Since blood fluid dynamics play a crucial role in shaping vessel morphology, the presence of flow defects generally prohibits the precise assignment of the role of the mutated gene product in the vasculature. In this study, we show how to distinguish between genetic defects caused by targeted inactivation of the Neuropilin‐1 (Nrp‐1) receptor, and hemodynamic defects occurring in homozygous knockout embryos. Using mouse embryo culture, we establish that hemodynamic defects are already present at E8.5 and continuous circulation is never established in homozygous mutants. The geometry of yolk sac blood vessels is altered and remodeling into yolk sac arteries and veins does not occur. To separate flow‐induced deficiencies from those caused by the nrp‐1 mutation, we arrested blood flow in cultured wild‐type and mutant embryos and followed their vascular development. We find that loss of nrp‐1 function rather than flow induces the altered geometry of the capillary plexus. This method allows genetic defects caused by loss‐of‐function of a gene important for cardiovascular development to be isolated even in the presence of hemodynamic defects.