Development of a novel therapeutic suppressor of brain proinflammatory cytokine up-regulation that attenuates synaptic dysfunction and behavioral deficits

Article abstract of DOI:10.1016/j.bmcl.2006.10.028

We report the development of a novel, aqueous-soluble, safe, small molecule, experimental therapeutic that suppresses injury-induced, proinflammatory cytokine increases in the brain, with resultant attenuation of synaptic protein biomarker loss and improvement in hippocampus-dependent behavioral deficits. A GMP production scheme for the active pharmaceutical ingredient, compound 17, is presented. The development and large-scale availability of this novel compound allow exploration of new, potentially disease-modifying, therapeutic approaches to CNS disorders.

Full text of DOI:10.1016/j.bmcl.2006.10.028

Bioorganic & Medicinal Chemistry Letters 17 (2007) 414–418
Development of a novel therapeutic suppressor of brain
proinflammatory cytokine up-regulation that attenuates
synaptic dysfunction and behavioral deficits
Wenhui Hu, Hantamalala Ralay Ranaivo, Saktimayee M. Roy, Heather A. Behanna,
Laura K. Wing, Lenka Munoz, Ling Guo, Linda J. Van Eldik and D. Martin Watterson*
Center for Drug Discovery and Chemical Biology, Northwestern University, 303 E. Chicago Avenue, Mail Code W896,
Chicago, IL 60611, USA
Received 26 September 2006; accepted 12 October 2006
Available online 17 October 2006
Abstract—We report the development of a novel, aqueous-soluble, safe, small molecule, experimental therapeutic that suppresses
injury-induced, proinflammatory cytokine increases in the brain, with resultant attenuation of synaptic protein biomarker loss
and improvement in hippocampus-dependent behavioral deficits. A GMP production scheme for the active pharmaceutical ingre-
dient, compound 17, is presented. The development and large-scale availability of this novel compound allow exploration of new,
potentially disease-modifying, therapeutic approaches to CNS disorders.
Ó 2006 Elsevier Ltd. All rights reserved.
Over-production of proinflammatory cytokines contrib-
utes to the progression of pathology in a diverse array of
diseases,1–5 and the recent approval of protein thera-
pies5–8 that seek to control the level or activity of proin-
flammatory cytokines provides a proof-of-concept for
targeting proinflammatory cytokine levels in drug dis-
covery. However, the macromolecular therapeutic ap-
proach is not as amenable to clinical use in some
diseases, such as central nervous system (CNS) disor-
ders, creating an unmet need for small molecule,
brain-penetrant compounds as new classes of disease-
modifying drugs. An accumulating body of evidence
from animal model studies and clinical observations
supports the concept1,3,4,9,10 that development of such
therapies for CNS disorders could have potential impact
on both acute and chronic disorders as diverse as Alz-
heimer’s disease (AD), stroke, traumatic brain injury,
neuropathic pain, Parkinson’s disease, multiple sclero-
sis, amyotrophic lateral sclerosis, age-related macular
degeneration, and non-AD dementias. The prevailing
hypothesis in this regard is that targeting the up-regula-
tion of proinflammatory cytokine production by activat-
ed glia, the main cellular source of cytokines in the CNS,
will alter disease progression through attenuation of the
subsequent neuronal synaptic dysfunction, the cellular
basis of clinical symptoms and behavioral alterations.
However, there are no such current therapies available
and there are no consensus molecular targets for such
discovery efforts. The current state of the art requires,
therefore, the use of the more classical and unbiased
functional approach to drug discovery.11
Recently, we described12,13 a contemporary version of
the functional approach for CNS drug discovery that
resulted in a novel lead compound with the desired
in vivo functions in animal models of disease. The
discovery platform used de novo lead compound gener-
ation that started with an inactive phenylpyridazine
fragment to generate a focused synthetic chemical
library.12,14,15 The proprietary library was screened for
the ability to suppress, in a concentration-dependent
manner, the increased production of proinflammatory
cytokines by activated glia in cell culture.12,13 Cell-based
screening hits were then tested in hierarchal, in vivo
screens to eliminate highly toxic compounds and to
identify compounds with good in vivo efficacy in animal
models of disease. We identified13 a novel, safe, orally
active, lead compound (MW01-5-188WH; compound 1
in Table 1), that was efficacious in animal models of
AD-relevant pathophysiology.
Keywords: Alzheimer’s disease; Pyridazine; Process chemistry;
Proinflammatory cytokine; Neurodegeneration.
*
Corresponding author. Tel.: +1 312 503 0656; fax: +1 312 503
0007; e-mail: m-watterson@northwestern.edu
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.10.028

W. Hu et al. / Bioorg. Med. Chem. Lett. 17 (2007) 414–418
415
Table 1. Identification of lead compound
R4
N
N R3
N N
R3
IL-1bb
a
Compound
R4
MW
N
1
394.47
408.50
395.46
374.48
332.40
352.82
332.40
331.41
330.43
2.5 0.7
N
N
2
3
4
5
6
7
8
5.3 0.3
25.8 3.1
6.1 1.2
8.3 2.1
9.5 2.3
N
N
N
N
N
N
N
CH3
Cl
N
N
N
N
CH3
CH3
CH3
46.1 11.6
17.7 3.6
15.5 5.2
N
N
9
10
CH3
CH3
336.47
268.36
31.4 3.5
11
CH3
48.9 12.8
a MW = molecular weight.
b Concentration (lM) for 50% inhibition; means SEM; IL-1b = interleukin-1b.
Compound 1 has attractive efficacy and safety screen-
ing profiles,13 but its aqueous solubility is not ideal
for clinical development. Extensive analyses indicate
the significant influence of a compound’s molecular
properties on pharmacokinetics, toxicity, and efficacy,
with success generally associated with low MW and
good aqueous solubility, especially drugs that fall into
the top class of the biopharmaceutics classification
scheme.16–20 However, increasing the aqueous solubil-
ity of a compound while decreasing MW and retaining
in vivo biological function is a non-trivial goal often
considered paradoxical, especially for CNS lead com-
pounds. We summarize in this report the discovery of
a novel lead compound, compound 5 (Table 1), that
has improved molecular properties but retains in vivo
efficacy. In addition we report the development of a
synthetic scheme for good manufacturing protocol
(GMP) production of compound 17, a highly soluble
and stable dihydrochloride hydrate of compound 5.
Compound 17 is the active pharmaceutical ingredient
(API) currently in clinical development.21
The rationale for efforts to improve molecular proper-
ties while retaining activity was to replace the R4 phen-
yl ring of compound 1 (Table 1). Compound 5 fulfilled
these criteria. Compounds 2–6 (Table 1) all have a MW
similar to, or less than, compound 1, but have better
computed logS values (respectively, À5.4, À5.3, À4.4,
À4.9, À4.1, and À4.6 for compounds 1–6). Compounds
2, 4, 5, and 6 retain activity in the glia cell-based screen
for concentration-dependent suppression of IL-1b pro-
duction, but there was some loss of activity with com-
pound 3 (Table 1). Previous screening results showed15
that there is no activity when R4 = H. Therefore, com-
pound 5 was the most promising due to its potential
improved aqueous solubility, retention of activity, and
a decreased MW (332.40 compared to 394.47). Experi-
mental determination of aqueous solubility confirmed
that compound 5 (0.016 mg/ml) is much more soluble
in water than compound 1 (<0.002 mg/ml).
Interestingly, an analog of compound 5 has been previ-
ously described in the literature.22 The CNS drug min-

416
W. Hu et al. / Bioorg. Med. Chem. Lett. 17 (2007) 414–418
aprine is also a 3-amino-4-methyl-6-phenylpyridazine.
However, minaprine is not active as a suppressor of pro-
inflammatory cytokine production by activated glia,
suggesting that the details of R3 may be critical for this
difference in function. In order to gain insight into key
features of compound 5 that distinguish it from minap-
rine, we examined the effects of diversification at R3 in
the context of compound 5 (synthesis of compounds
7–11) on retention of activity. Compounds 7–11 (Table
1) are analogs of compound 5 that differ at R3. Diversi-
fication included alternatives to the pyrimidinyl ring,
including compounds with a pyrazinyl (compound 7),
pyridinyl (compound 8), phenyl (compound 9), or cyclo-
hexyl (compound 10) ring. Preference for an aromatic
group at R3 for full activity in the context of a methyl
at R4 is suggested by the loss of activity with a cyclohex-
yl (compound 10) or methyl (compound 11) at R3, and
the partial retention of activity with a pyridinyl or
phenyl (compounds 8 and 9). However, this hypothesis
is contradicted by compound 7, which has a pyrazinyl
at R3 yet has a lower activity comparable to the cyclo-
hexyl (compound 10) or methyl (compound 11) analogs.
The results indicate that only certain types of aromatic
rings can substitute for the pyrimidine ring in the con-
text of 3-amino-6-phenylpyridazines represented by
compounds 1 and 5. Although compound 5 is distinct
from minaprine in its structural requirements at R3, it
is similar to minaprine in its retention of partial activity
with the presence of a benzyl at R6 instead of a phenyl
(compound 12) or loss of activity with the presence of a
methyl (compound 13).21 However, further pursuit of
structure–activity relationships was not done as part of
this investigation because the discovery of compound 5
provided attainment of our discovery goal, and the
diversification at R3 revealed a strict preference for an
arylpiperazine, a feature distinguishing compound 5
from minaprine.
tion at a potential therapeutic dose, or after acute, esca-
lating-dose oral administration at doses up to 40 times a
potential therapeutic dose (Fig. 1). It also shows cardio-
vascular safety, as measured by lack of prolongation of
cardiac QTc interval in guinea pigs following FDA rec-
ommended protocols. Compound 5 demonstrates brain
penetration (LogBB=0.45; Fig.S1) 21 comparable to
other CNS drugs and compound 1.13 Compound 5 is
efficacious in an AD-relevant mouse model of human
amyloid-beta (Ab)-induced injury (Fig. 2). A single, dai-
ly oral administration of compound 5 at a low dose
(2.5 mg/kg) for 2 weeks, beginning 21 days after initia-
tion of the human Ab injury, effectively suppressed the
up-regulation of the proinflammatory cytokines inter-
leukin-1b (IL-1b), tumor necrosis factor alpha (TNFa),
and S100B measured at day 60 after the start of injury.
The number of activated GFAP-positive astrocytes and
F4/80-positive microglia was also reduced. The loss of
the synaptic marker proteins synaptophysin and PSD-
95 was attenuated. Deficits in the Y-maze, a hippocam-
pus-dependent behavioral task, were ameliorated. These
data document that compound 5 retains the desired
in vivo functions of the original lead compound 1.
The discovery of compound 5 as an analog of com-
pound 1 that retains full in vivo functions of compound
1, but has a lower MW and better aqueous solubility,
provided the starting point for development and charac-
terization of an API to allow clinical investigations.
Standard diversification chemistries used for the synthe-
sis of compounds such as those in Table 1 are not attrac-
tive for chemical processes amenable to large-scale
GMP production. For example, a Pd-catalyzed reaction
from a common 4-chloropyridazine precursor (com-
pound 6) was used initially to synthesize compound 5
and other analogs, summarized in Scheme S1,21 as part
of the in-parallel synthetic approach to diversification.12
However, Pd-catalyzed reaction protocols raise specific
concerns for GMP production and are not readily
scalable. Therefore, we developed an efficient produc-
tion protocol (Scheme 1) to facilitate manufacture for
clinical studies. In addition, the protocol offers the
potential for use of water as an alternative solvent in
the penultimate amination step. We also added as a last
A seminal discovery for the studies summarized here is
that compound 5 retains the in vivo functions of com-
pound 1 (Figs. 1, 2, and S121). Specifically, compound
5 mimics compound 1 in oral bioavailability, safety,
and CNS disease model efficacy. Compound 5 produces
no histological liver injury after chronic oral administra-
Figure 1. Oral administration of compound 5 does not reveal in vivo liver or cardiac toxicity. (A) Chronic therapeutic dose test of liver injury: there is
no histological evidence of liver injury after oral administration of compound 5 (2.5 mg/kg) to mice once daily for 2 weeks. (B) High dose test of liver
injury: no histological evidence of liver injury is seen after oral administration of increasing doses up to 100 mg/kg to mice once daily for 3 days. (C)
Physiological test for cardiotoxicity at high dose: no evidence of prolongation of QTc interval after oral administration (15 mg/kg) to guinea pigs;
sotalol is a positive control compound.

W. Hu et al. / Bioorg. Med. Chem. Lett. 17 (2007) 414–418
417
Figure 2. Compound 5 is efficacious in an Alzheimer’s disease mouse model. Compound 5 suppresses up-regulation of the proinflammatory
cytokines IL-1b, TNFa, and S100B; decreases astrocyte (GFAP) and microglia (F4/80) activation; prevents loss of the synaptic proteins
synaptophysin and PSD-95; and attenuates hippocampal-dependent behavioral deficits in the Y-maze. C, control; I, Ab-injured; I + compd, Ab-
injured + oral administration (2.5 mg/kg once daily for two weeks beginning three weeks after start of injury). Data are means SEM of n = 5–10
mice per group. Significantly different from Ab-injured: **p < 0.01, ***p < 0.001.
step the production of the dihydrochloride hydrate,
compound 17, which is the API,21 as the final product.
Hydrochloride hydrates are a common salt form used
in API production for diverse drugs,23 including the
drug minaprine. As summarized in Scheme 1, the com-
mercially available 2-methyl-4-oxo-4-phenyl butanoic
acid was cyclized with hydrazine to obtain compound
14, which was then dehydrogenated to 4-methyl-6-phen-
ylpyridazinone compound 15. Chlorination with phos-
phorus oxychloride provided the 3-chloropyridazine
compound 16. Compound 5 was then obtained by ami-
nation. The corresponding dihydrochloride hydrate,
compound 17, was synthesized by treatment with HCl
in isopropanol. As detailed in supporting data,21 this
scheme and its variations have been used for generation
of the API at the multi-gram level in the laboratory and
at the kg level under GMP conditions.
O
O
O
OH
O
NH
N
NH
N
c
a
b
Compound 17 is water-soluble and stable. Compound 17
has improved solubility at 37 °C compared to the lead
base compound 5, and compared to compound 18,21 the
analogous dihydrochloride hydrate of compound 1. The
experimental values for compound 17 and compound 18
are, respectively, 322 mg/ml and 0.04 mg/ml. In contrast
to differences in experimentally determined aqueous solu-
bility, these compounds possess similar experimentally
determined logP values (2.3 and 2.7), as measured by oct-
anol/water partition coefficient. Compound 17 has a melt-
ing point >215 °C, a temperature at which it decomposes,
compared to a melting point of 115.7–116.4 °C for the
base compound 5. When compound 17 was tested for
chemical stability in water, dilute acid, or dilute base at
37 °C over 5 days, there was <5% loss of material in water
and acid. Compound 17 was less stable in basic condi-
tions, with only 53% remaining after 5 days. Compound
17 also showed <10% loss of material when treated with
3% (v/v) H2O2 at ambient temperature (24 °C) for up to
48 h. These data provide experimental evidence that the
14
15
N
N
N
N
N
N
N
Cl
N
N
N
e
N
N
d
N
N
. 2HCl.H2O
16
5
17
Scheme 1. Reagents and conditions: (a) N2H4, EtOH, reflux; (b)
CuCl2, CH3CN, reflux; (c) POCl3, CH3CN, reflux; (d) 1-(2-pyrim-
idyl)piperazine, water, reflux; (e) HCl, isopropanol.

418
W. Hu et al. / Bioorg. Med. Chem. Lett. 17 (2007) 414–418
8. Mpofu, S.; Fatima, F.; Moots, R. J. Rheumatology 2005,
44, 271.
molecular properties were improved as a result of refine-
ment, and demonstrate that the API is a water-soluble,
stable product.
9. Guo, Z.; Cupples, L. A.; Kurz, A.; Auerbach, S. H.;
Volicer, L.; Chui, H.; Green, R. C.; Sadovnick, A. D.;
Duara, R.; DeCarli, C.; Johnson, K.; Go, R. C.; Grow-
don, J. H.; Haines, J. L.; Kukull, W. A.; Farrer, L. A.
Neurology 2000, 54, 1316.
10. D’Ambrosio, R.; Perucca, E. Curr. Opin. Neurol. 2004, 17,
731.
11. Sams-Dodd, F. Drug Discovery Today 2006, 11, 465.
12. Wing, L. K.; Behanna, H. A.; Van Eldik, L. J.; Watterson,
D. M.; Ralay Ranaivo, H. Curr. Alzheimer. Res. 2006, 3,
205.
In summary, we have described the discovery and devel-
opment of a novel, water-soluble, safe, experimental
therapeutic with potential to be a new type of disease-
modifying drug. Currently, IND-enabling preclinical
toxicology and pharmacokinetic analyses are validating
and extending the findings of the initial screens reported
here, and GMP manufacture of the API for clinical
studies has been accomplished.21
13. Ralay Ranaivo, H.; Craft, J. M.; Hu, W.; Guo, L.; Wing,
L. K.; Van Eldik, L. J.; Watterson, D. M. J. Neurosci.
2006, 26, 662.
14. Velentza, A. V.; Wainwright, M. S.; Zasadzki, M.;
Mirzoeva, S.; Schumacher, A. M.; Haiech, J.; Focia, P.;
Egli, M.; Watterson, D. M. Bioorg. Med. Chem. Lett.
2003, 13, 3465.
Acknowledgment
This work was supported in part by NIH Grants
AG028561, NS047586, and AG000260.
15. Hu, W.; Ralay Ranaivo, H.; Craft, J. M.; Van Eldik, L. J.;
Watterson, D. M. Curr. Alz. Res. 2005, 2, 197.
16. Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P.
J. Adv. Drug Delivery Rev. 1997, 23, 3.
Supplementary data
17. van de Waterbeemd, H.; Smith, D. A.; Beaumont, K.;
Walker, D. K. J. Med. Chem. 2001, 44, 1313.
18. Wenlock, M. C.; Austin, R. P.; Barton, P.; Davis, A. M.;
Leeson, P. D. J. Med. Chem. 2003, 46, 1250.
19. Vieth, M.; Siegel, M. G.; Higgs, R. E.; Watson, I. A.;
Robertson, D. H.; Savin, K. A.; Durst, G. L.; Hipskind, P.
A. J. Med. Chem. 2004, 47, 224.
20. Veber, D. F.; Johnson, S. R.; Cheng, H.-Y.; Smith, B. R.;
Ward, K. W.; Kopple, K. D. J. Med. Chem. 2002, 45,
2615.
The contents of supporting information include the fol-
lowing: details of synthetic procedures, analytical chem-
istry methods, physical property determinations, NMR
spectra, and biological procedures, including screening
data for oral bioavailability. Supplementary data associ-
ated with this article can be found, in the online version,
at doi:10.1016/j.bmcl.2006.10.028.
References and notes
21. Details of synthetic procedures, analytical chemistry
methods, physical property determinations, NMR spectra,
and biological procedures, including screening data for
oral bioavailability and brain uptake, can be found in
Supporting Information. Information about the clinical
development of compound 17 under the name of Minozac
can be accessed on the website of NeuroMedix (NMX)
Inc., http://www.neuromedix.com.
22. Wermuth, C. G.; Schlewer, G.; Bourguignon, J.-J.; Mag-
hioros, G.; Bouchet, M.-J.; Moire, C.; Kan, J.-P.; Worms,
P.; Biziere, K. J. Med. Chem. 1989, 32, 528.
23. Wermuth, C. G.; Stahl, P. H. In Handbook of Pharma-
ceutical Salts; Stahl, P. H., Wermuth, C. G., Eds.; Wiley-
VCH, 2002; p 249.
1. Craft, J. M.; Watterson, D. M.; Van Eldik, L. J. Expert
Opin. Ther. Targets 2005, 9, 887.
2. Mrak, R. E.; Griffin, W. S. T. Neurobiol. Aging 2005, 26,
349.
3. Wyss-Coray, T. Nat. Med. 2006, 12, 1005.
4. Watkins, L. R.; Maier, S. F. Nat. Rev. Drug Disc. 2003, 2,
973.
5. Maini, R. N.; Taylor, P. C. Annu. Rev. Med. 2000, 51, 207.
6. Scott, D. L.; Kingsley, M. B. N. Engl. J. Med. 2006, 335,
704.
7. Tobinick, E.; Gross, H.; Weinberger, A.; Cohen, H. Med
Gen. Med. 2006, 8, 25.