SESSION 3A TRANSLATIONAL STRATEGIES

C16 HOW CAN ACADEMIC LABS CONTRIBUTE TO THERAPY DEVELOPMENT IN ALS? A MEDICINAL CHEMISTRY PERSPECTIVE

LIPINSKI C

Melior Discovery, Connecticut, United States

E-mail address for correspondence: [email protected]

Keywords: medicinal chemistry, rules and filters, chemistry garbage

An academic labs contribution to therapy development can be ruined by something totally outside of biology, namely miserable chemistry.

Here are some of the ugly facts about miserable chemistry. Seven out of eight commercially available screening compounds are chemistry trash; not worth screening for therapy development and quite likely not even worth screening as tools or probes in a chemical biology target validation exercise. It is not easy to discern true screening positives from false positives. The worst false positives are chemistry related. There is something in the chemical structure that causes the compound to appear active in a screen but in a biological sense the activity is bogus. This is really troublesome because chemistry related false positive activity holds up in replicate assays. The big pharmaceutical companies independently recognized this problem and junked a third to a half of their legacy screening collections. The bottom line; remove the chemistry garbage before you screen not after you screen.

Other dirty laundry facts you should know. Always confirm chemical structure. The label on the vial and the actual chemical structure can be very different. Always verify the chemistry on an active by re-synthesis of the compound before spending significant effort. Somewhere between 10–50% of actives do not hold up when the compound is remade and retested, even if the original sample identity and purity were OK.

There is a wealth of knowledge about miserable chemical structures but it is buried in the medicinal chemistry literature and not easily accessible to biologists. There are all kinds of rules and filters to get rid of the chemical garbage. Dealing with bad chemistry is not rocket science but it can be very difficult for a biologist. Think of a veteran medicinal chemist as a master of pattern recognition who can mentally connect a chemical structure to biological information. Seasoned medicinal chemists are like board certified pathologists in pattern recognition skills.

There is a big people-disconnect between academic biology and medicinal chemistry. The vast majority of medicinal chemists live in industry. Learning medicinal chemistry pattern recognition takes time, maybe 10–15 years to have a chance to really get good. The academic medicinal programs are few and because of the time factor the bulk of the knowledge lies in only a small number of senior faculty as opposed to graduate students or post docs.

A collaboration or connection to a medicinal chemist is a priceless asset if, as a biologist, you really want to make a contribution to ALS therapy development. You can expect a certain degree of conflict with the medicinal chemist. They may tell you the chemical compounds you are testing for your biology community peer reviewed publication are worthless. They will make “snap” judgements about chemical quality in just a few seconds by looking at chemical structures. My advice; if you want to positively connect to medicinal chemists put a beautiful looking chemical structure in the first few slides of your biology presentation.

C17 SMALL MOLECULE SCREENING FOR MOTOR NEURON PROTECTION: FROM BENCH TO CLINIC

BORDET T, ABITBOL JL, BERNA P, PRUSS R

Trophos, Marseilles, France

E-mail address for correspondence: [email protected]

Keywords: TRO19622, clinical trial, drug screening

Rationalizing drug screening for ALS remains a challenge. The vast majority of cases are sporadic, and it is not known to what extent the disease pathophysiology in these patients reflects that in the familial forms (e.g. SOD1-related). In addition, different cell types (e.g. lower motor neurons, upper motor neurons, astrocytes and microglia) are involved and probably need to be targeted to obtain efficacy. Finally, and in common with many human neurodegenerative diseases, ALS affects post mitotic and highly differentiated cells that have specific properties that cannot be modeled by any experimental system. As a result, there is no perfectly validated molecular target and, to date, no animal model predictive of clinical efficacy in ALS. In such a context, the use of a phenotypic screening process to discover drugs that intervene in the fundamental pathways controlling survival and death of motor neurons appears a reasonable approach.

Starting in 2000, Trophos adopted that unconventional route and screened a library of 45,000 “drug-like” compounds on purified rat embryonic motor neurons deprived of trophic factors and selected TRO19622 (chemical name: cholest-4-en-3-one, oxime) as the most promising drug candidate [1]. This cholesterol-like small molecule is orally bioavailable and crosses the blood brain barrier. It rescued motor neurons from facial nerve axotomy in neonatal rats and promoted axonal regeneration after peripheral nerve crush in mice. It also significantly delayed disease onset and improved the survival of SOD1G93A mice. Extensive systematic pharmacological profiling showed that TRO19622 interacts physically with two constituents of the mitochondrial permeability transition pore suggesting that its neuroprotective action is mediated, at least in part, through the modulation of the mitochondrial membrane permeabilization. Interestingly TRO19622 demonstrated efficacy in several models of pathological conditions where mitochondrial dysfunction has been evidenced such as painful peripheral neuropathies [2].

TRO19622 has completed regulatory preclinical assessment of pharmacology, safety, toxicity and pharmacokinetics showing no major signs of toxicity either in acute or long term regulatory safety toxicity studies. TRO19622 has successfully completed Phase 1/1b studies in both healthy volunteers and ALS patients demonstrating the product is well tolerated, has an excellent safety profile and that once-a-day dosing achieves the predicted exposure level required for efficacy, based on preclinical models. A pivotal European multi-centric Phase 2/3 trial of TRO19622 as add-on to riluzole is planned to begin in 2009 including 470 ALS patients. Although promoting motor neuron survival and nerve regeneration may not be sufficient to arrest all the events underlying this complex disease, the preclinical data obtained to date suggest such a compound will provide a step forward in the understanding of the disease process and will be of benefit to patients.

C18 SMALL MOLECULE ACTIVATORS OF THE NRF2-ARE PATHWAY FOR TREATMENT OF AMYOTROPHIC LATERAL SCLEROSIS.

HIGGINBOTTOM A, MEAD R, BARBER S, DAFTARY S, SHAW P

University of Sheffield, United Kingdom

E-mail address for correspondence: [email protected]

Keywords: NRF2, drug, astrocyte

Background: Nrf2 drives expression of a battery of Phase II detoxification and anti-oxidant enzymes via its interaction with the antioxidant response element (ARE). When activated, this ‘programmed cell life’ (pro-cell life) response is neuroprotective and may be of clinical value in patients with Amyotrophic Lateral Sclerosis (ALS) in which oxidative stress is heavily implicated. There are a number of molecules which activate this pathway but they tend to be pro-oxidant in themselves and have little capacity for crossing the blood-brain barrier (BBB) and so may not be useful in in vivo models of human neurodegenerative disease.

Objectives: The primary aim of the present study was to conduct a search for small molecule activators of the Nrf2-ARE ‘pro-cell life’ pathway, with increased potency, minimal toxic effects and enhanced opportunity for CNS penetrance compared to the currently available tool molecules.

Methods: A small molecule library of 2,000 compounds (the Spectrum collection), containing natural products as well as 640 FDA approved drugs and molecules likely to modulate CNS function, was screened in an NRF2-ARE reporter assay designed and optimised for high-throughput screening. Hit molecules were assessed in in vitro models of oxidative stress and CNS specific reporter assays and profiled using chemi-informatics to select drug-like molecules with the potential for CNS penetrance. A small number of molecules were selected for further in vivo studies.

Results: In general, the hit molecules identified were able to robustly activate the NRF2-ARE pathway in an astrocytic cell line (C6), but less well in motor neuronal-like cells (NSC34). In addition C6 cells pre-treated with the NRF2 inducer which was then removed were significantly protected against an oxidative insult.

Two molecules with a history of human use were identified which satisfied our stringent criteria for activity in models of oxidative stress and potential for in vivo use and CNS penetrance. The first (Compound 1) is a natural product used in herbal medicine which had sub-micromolar activity in cellular reporter assays (EC₅₀=0.36µM in a C6 NRF2-ARE reporter cell line). The second molecule (Compound 2) has been used in human patients with another neurodegenerative condition. Interestingly the S(+) enantiomer of this compound which has lost primary agonist activity, retained NRF2-ARE inducing activity (EC₅₀=7.43 µM in a C6 NRF2-ARE reporter cell line).

Discussion and Conclusions: These molecules show promise for further development as neuroprotective agents for ALS. Compound 2 may be particularly useful as it has a history of safe use in man, it is known to preferentially partition to the CNS and its use would not be limited by primary agonist activity, allowing higher doses to be investigated.

C19 PROTEIN BIOMARKERS FOR ALS DISEASE PROGRESSION

BOWSER R, RYBERG H, WILSON M, DARKO S, LACOMIS D

University of Pittsburgh School of Medicine, PA, United States

E-mail address for correspondence: [email protected]

Keywords: biomarkers, cystatin c, surrogate markers

Background: Amyotrophic lateral sclerosis (ALS) is a rapidly progressive disease, with the typical time from diagnosis till death of 2–5 years. There are no current diagnostic tests for ALS or surrogate markers of disease progression. Surrogate markers of ALS disease progression would be valuable measurements for drug efficacy in clinical trials and also to identify new targets for drug therapy. ELISA and mass spectrometry were used to identify protein alterations that correlate to disease progression. We identified specific proteins that correlate to clinical parameters of disease progression within ALS patients, and also proteins that identified subpopulations of ALS patients based on site of disease onset.

Objectives: To identify protein biomarkers that correlate to clinical measurements of ALS disease progression, we prospectively collected cerebrospinal fluid (CSF) and blood plasma every 4 months from 24 ALS patients. We also collected CSF and plasma from 14 control subjects over a 2-year time frame.

Methods: We collected blood plasma and CSF samples from 24 ALS (20 SALS and 4 FALS) and 14 control (11 healthy controls, 2 MS, 1 spinal cerebellar ataxia) subjects at the University of Pittsburgh School of Medicine. Samples were collected from ALS patients every 4–6 months over a 3-year time course. Samples were collected from control subjects over a 2-year period. Subjects were age/gender matched. ALS disease progression was monitored with clinical parameters including ALS-FRS, forced vital capacity (FVC), and manual muscle strength. We performed ELISA for cystatin C, transthyretin, neurofilament H, and anti-thrombin III on each CSF and plasma sample. Samples were run in duplicate and experiments repeated three times. Protein levels were correlated to clinical parameters of disease progression.

Results: Cystatin C protein levels in both the CSF and plasma decrease over time in ALS patients that exhibit rapid clinical disease progression. Transthyretin protein levels did not change over time in ALS patients, though there were modest increased levels in the plasma of limb onset ALS and decreased levels in the plasma of bulbar onset ALS. We observed increased levels of phospho-neurofilament H in the CSF of ALS patients and increased levels of anti-thrombin III in both the CSF and plasma of ALS patients during disease progression. Both neurofilament H and anti-thrombin III protein levels were increased in ALS patients when compared to control subjects.

Conclusions: Our data indicates that cystatin C may have utility as a surrogate biomarker for ALS disease progression, and could differentiate fast versus slow progression. In addition, phosphor-neurofilament H and anti-thrombin III protein levels were also altered during ALS disease progression, suggesting additional biochemical pathways altered during the course of disease. Overall levels of transthyretin did not significantly differ during ALS disease progression, though we noted differences in transthyretin levels between limb versus bulbar disease onset.