Methodology for the preparation of 2-argininylbenzothiazole

Article abstract of DOI:10.1021/jo7019543

(Chemical Equation Presented) An efficient process to produce kilogram quantities of a key argininylbenzo[d]thiazole intermediate was developed for the preparation of the tryptase inhibitor RWJ-56423. A variety of activated arginine esters and benzo[d]thi

Full text of DOI:10.1021/jo7019543

Structurally, RWJ-56423 (1) is a dipeptide containing de-
rivatized arginine and L-proline amino acids. The arginine resi-
due bears both a benzo[d]thiazole heterocycle and an L-proline
residue which has both an N-acetyl and 4-hydroxyl group
attached.
Methodology for the Preparation of
2-Argininylbenzothiazole
Birdella D. Kenney,*,† Michael Breslav,‡ Rosie Chang,‡,
Roland Glaser,§ Bruce D. Harris,‡ Cynthia A. Maryanoff,|
John Mills,‡ Armin Roessler,§ Brigitte Segmuller,† and
Frank J. Villani, Jr.‡
An early approach to synthesize RWJ-56423 (1) was to use
ornithine as a precursor for arginine (Scheme 1). Coupling of
Boc-ornithine ester 3 with the derivatized L-proline amino acid
2 using classical coupling conditions gave 4 (Scheme 1). The
subsequent preparation of acid 5 and the Weinreb4 amide 6 was
successful; however, attempting to react this amide with
2-lithiobenzothiazole gave none of the desired product 7 but
rather the loss of the N-acetyl group on the proline residue.
As a consequence of this result, we made three modifications
to this approach: first we decided to introduce the proline
residue last in the synthesis to avoid interference from the
N-acetylproline function; second, we used the Grignard reagent
(benzothiazol-2-ylmagnesium halide) rather than the lithium
analogue to conduct the reaction at a more practical temperature;
third, we used protected arginine-activated esters rather than
ornithine as an arginine precursor. Thus, we selected three
suitable activated arginine esters (11a, 11b, and 12) to react
with different benzothiazol-2-ylmagnesium halides to prepare
the 2-argininylbenzo[d]thiazole 13.5 Compound 13 can then be
converted to RWJ-56423 (1). The benzothiazol-2-ylmagnesium
halides (9a and 9b) were prepared by reacting benzo[d]thiazole
8 with excess Grignard reagents such as EtMgBr, EtMgCl, or
t-BuMgCl at 0-5 °C. The reaction conditions and the results
obtained from coupling the benzothiazol-2-ylmagnesium halides
with the selected activated esters are summarized in Table 1.
The reaction of the thiolester 11a6 with benzothiazol-2-
ylmagnesium bromide (9a) gave the ketone 13 in moderate yield
(63%, Table 1, entry 1) on a 400 g laboratory scale. However,
upon scale up of the reaction to 2 kg, the yield dropped to 25%
(Table 1, entry 2). We observed the solution of benzothiazol-
2-ylmagnesium bromide change to a thick suspension with
prolonged stirring. This suggested that a lesser amount of
benzothiazol-2-ylmagnesium bromide was in solution on scale
up and thereby caused lower yields for the subsequent coupling
reaction. The outcome of the coupling of imidazolide 11b with
benzothiazol-2-ylmagnesium halides was dependent on the
Grignard reagent being used. Experimental results showed that
treatment of 11b with 9a resulted in isolation of a 71% yield
of product 13 (Table 1, entry 3) on a 50 g laboratory scale,
compared to 90% yield when 9b was used (Table 1, entry 4)
on a 2 kg scale. During the formation of 9a by reacting 8 with
EtMgBr in ether, the product formed as a heavy precipitate,
making it difficult to determine the amount of Grignard reagent
available for the reaction. On the other hand, the benzothiazol-
2-yl MgCl Grignard reagent 9b formed by reacting 8 with
EtMgCl in THF was soluble in the reaction solvent. These
results demonstrated a noticeable advantage for the use of the
chloro Grignard reagent 9b over the bromo analogue 9a.
However, further scale-up using 9b gave a lower yield (40%,
Table 1, entry 5) of product 13. The lactam (12) was also tested
Johnson & Johnson Pharmaceutical Research and DeVelopment,
Spring House, PennsylVania 19477, Johnson & Johnson
Pharmaceutical Research and DeVelopment,
Raritan, New Jersey 08869, and Cilag AG,
8205 Schaffhausen, Switzerland
birdellakenney@yahoo.com
ReceiVed September 5, 2007
An efficient process to produce kilogram quantities of a key
argininylbenzo[d]thiazole intermediate was developed for the
preparation of the tryptase inhibitor RWJ-56423. A variety
of activated arginine esters and benzo[d]thiazole nucleophiles
were evaluated as coupling partners. Our work led to the
selection and optimization of an argininyl imidazolide ester
and benzothiazol-2-yl MgCl nucleophile. This paper focuses
on the preparation, use, and stability of the benzothiazol-2-
yl Grignard reagents.
RWJ-564231 (1) is a selective, potent, small-molecule inhibi-
tor of human mast cell tryptase having therapeutic potential for
treating allergic or inflammatory disorders.2 This synthetic
pharmacologic agent is an argininyl-tryptase inhibitor which
shows selectivity over other serine proteases such as kallikrein,
factor Xa, and chymotrypsin.3
* Address correspondence to this author.
† Johnson & Johnson Pharmaceutical Research and Development, Raritan.
‡ Johnson & Johnson Pharmaceutical Research and Development, Spring
House.
Current address: Suffolk County Community College, 530 College Rd,
Selden, N.Y. 11784.
§ Cilag AG.
| Currently at Cordis Corp.
(1) Costanzo, M.; Maryanoff, B.; Hecker, L.; Schott, L.; Yabut, S.; Zhang,
H.; Andrade-Gordon, P.; Kauffman, J.; Lewis, J.; Krishnan, R.; Tulinsky,
A. J. Med. Chem. 1996, 39, 3039-3043.
(2) Clark, J. M.; Moore, W. R.; Tanaka, R. D. Drugs Future 1996, 21,
811-816.
(3) Costanzo, M. J; Yabut, S. C.; Almond, H. R., Jr.; Andrade-Gordon,
P.; Corcoran, T. W.; de Garavilla, L.; Kauffman, J. A.; Abraham, W. M.;
Recacha, R.; Chattopadhyay, D.; Maryanoff, B. E. J. Med. Chem. 2003,
46(18), 3865-3876.
(4) Deng, J.; Hamada, Y.; Shioiri, T.; Matsunaga, S.; Fusetani, N. Angew.
Chem., Int. Ed. 1994, 33, 1729-1731. Kahn, M. PCT Int. Appl. WO9630396
A1, 1996 (Example 2).
(5) Yuan, G.; Xue, K. Acta Chim. Sin. 1990, 48, 931-935.
(6) Horiki, K. Synth. Commun. 1977, 7, 251-259.
10.1021/jo7019543 CCC: $37.00 © 2007 American Chemical Society
Published on Web 11/13/2007
9798
J. Org. Chem. 2007, 72, 9798-9801

SCHEME 1. Original Synthetic Strategy To Prepare RWJ-56423
TABLE 1. Preparation of 2-Arginylbenzothiazole 13a
(50-100 g), we were unable to identify the cause of our
problematic scale-up issues because yields and purities were
greater than 90%. Attention was then turned to a careful exam-
ination of the stability of the Grignard reagents 9a and 9b.
Stability studies of benzothiazol-2-yl MgBr (9a) and ben-
zothiazol-2-yl MgCl (9b) were carried out at 0 °C under nitrogen
(Figure 1). Reacting 8 with either EtMgBr or EtMgCl prepared
the desired Grignard reagent (9a or 9b). Aliquots were taken
in either 15 min or 1 h intervals of the initial reaction time and
immediately quenched into dilute acid. Based upon the weight
percent analysis of the aliquots taken, the amount of benzo[d]-
thiazole was determined over a 6 h period. The mostly insoluble
9a showed no change in the amount of benzo[d]thiazole over
time. On the other hand, the soluble benzothiazol-2-yl MgCl
(9b) showed decreased amounts of benzo[d]thiazole over time.
After 6 h at 0 °C, only 70% of active 9b remained, and the
disappearance of 30% of 9b was associated with the formation
of a yellow viscous oily residue. On the other hand, additional
studies showed that the reverse addition of up to 80% of the
benzo[d]thiazole /THF solution into EtMgCl/THF afforded a
stable solution of 9b for at least 1 h at 10 °C without the
formation of the yellow viscous oily residue. However, the
addition of excess benzo[d]thiazole/THF solution into the
solution of EtMgCl resulted in the immediate formation of the
yellow viscous oily residue observed previously. This result
suggested that the presence of excess benzo[d]thiazole catalyzes
the decomposition of 9b to the yellow oily residue, while excess
EtMgCl impedes its formation.7 In addition to that, we noticed
that prolonged stirring of the 9b solution in the presence of air
introduced by opening of the reaction flask caused its faster
decomposition. The isolation, purification, and characterization
of the yellow oily residue was later determined to be the benzo-
[d]thiazole trimer 15. The decomposition to the benzo[d]thiazole
Grignard
entry reagent
method of
prepn
activated method of
yield
scale (%)
ester
prepn
1
2
3
4
5
6
7
8
9a
9a
9a
9b
9b
9b
9b
9b
EtMgBr (a)
EtMgBr (a)
EtMgBr (a)
EtMgCl (b)
EtMgCl (b)
EtMgCl (b)
EtMgCl (b)
t-BuMgCl (c)
11a
11a
11b
11b
11b
12
d
d
e
e
e
f
400 g
2 kg
50 g
2 kg
4 kg
5 g
63
25
71
90
40
95
75
80
12
f
50 g
10 kg
11b
e
a Reagents and conditions: (a) THF, 10 °C, EtMgBr in Et2O, 10-15
min; (b) THF, 10 °C, EtMgCl in THF, 10-15 min; (c) THF, 3-5 °C,
t-BuMgCl in THF, 10-15 min; (d) Ph2POCl, Et3N, PhSH, 0 °C, 15 min;
(e) CDI, THF, rt, 30 min; (f) CDI, DMF, TEA, rt, 30 min.
in the reaction with benzothiazol-2-yl MgCl 9b. Surprisingly,
this reaction produced the desired ketone 13 in excellent yield
on a 5 g scale (95%, Table 1, entry 6). However, scaling up the
reaction to 50 g (Table 1, entry 7) resulted in a lower (75%)
conversion to product 13. Increasing the equivalents of EtMgCl
compared to benzo[d]thiazole resulted in the consumption of
12; however, this also resulted in the formation of ethyl ketone
14 as a byproduct.
The results obtained from the reactions listed in Table 1 point
to one major observation; regardless of which activated ester is
used, the large-scale reactions resulted in compromised yields.
We continued process development to evaluate the following
critical parameters: addition rate, exclusion of moisture,
elimination of light, and order of addition. On laboratory scale
FIGURE 1. Comparative stability of 2-benzothiazole Grignards.
J. Org. Chem, Vol. 72, No. 25, 2007 9799

SCHEME 2. Mechanism for the Formation of Benzo[d]thiazole Trimer (15) in the Presence of Oxygen and Magnesium
trimer explained the lowered yields on large-scale compared to
the laboratory scale reactions.
experimental conditions were explored in an open reaction
vessel, a significant amount of trimer was observed. These
results showed that the exclusion of air was an uncontrolled
critical parameter on large scale. The Grignard reaction was
being compromised by air entering the large-scale reactors.
Air dissolved in the THF solvent or EtMgCl reagent solution
prior to use are potential sources for oxygen getting in the
reactor.
In the presence of oxygen, two competitive pathways are in
effect: (1) the nucleophilic acylation of activated Boc-Arg-Mtr
with benzothiazol-2-yl MgCl which leads to product and (2)
an oxygen quench of benzothiazol-2-yl MgCl which leads to
its degradation (formation of the trimer). LC/MS analysis of
the reaction mixtures showed evidence that benzothiazolone
(17a, Scheme 2) and benzo[d]thiazole dimer (19, Scheme 2)
are along the reaction pathway to form benzo[d]thiazole trimer
(15). To confirm this finding, we subjected commercially
available benzothiazolone (17b) to benzothiazol-2-yl MgCl (9b)
under the same reaction conditions. The in situ product matched
the benzo[d]thiazole trimer (15) by HPLC retention times and
LC/MS exact mass. Floating particulates in the Grignard reagent
were observed, which were thought to be residual magnesium
metal leftover from the Grignard reagent formation. In preparing
benzothiazol-2-yl MgCl (9b), magnesium metal was added. As
a result, the formation of the benzo[d]thiazole trimer (15) was
facilitated. The Grignard reagent was not at any time filtered
before use.
The solution for this decomposition problem was to maintain
excess EtMgCl during the formation of 9b. However, excess
EtMgCl reacts with 11a, 11b, and 12 and gives rise to undesired
side products. It is at this point that we replaced EtMgCl with
the more sterically hindered t-BuMgCl to prepare 9b. The
t-BuMgCl is unreactive toward the imidazolide (11b). An excess
of t-BuMgCl could be used without concern of a side reaction
with imidazolide 11b. With this new procedure, Boc-Arg-Mtr
ketone (13) could be successfully prepared on a 10 kg scale in
80% yield using 4 equiv of benzo[d]thiazole and 12 equiv of
t-BuMgCl in a predictable reproducible manner (Table 1, entry
8). As a safety precaution, the reaction mixture was heavily
blanketed with inert argon due to the pyrophoric nature of
t-BuMgCl. The argon gas also minimizes any oxygen absorption
and subsequent benzo[d]thiazole degradation.
This modification provided a reliable and scaleable process
in preparing the key intermediate argininyl-benzo[d]thiazole 13
to meet the RWJ-56423 drug substance demand. However, one
issue remained: we still needed to determine the mechanism
for the formation of the benzo[d]thiazole trimer.
It was proposed that oxygen played a key role in the formation
of the benzo[d]thiazole trimer (15), regardless to the order of
addition. The benzothiazol-2-yl MgCl formation in which
oxygen was rigorously excluded via an inert atmosphere of
nitrogen and a 1:1.1 ratio of EtMgCl/benzo[d]thiazole resulted
in 100% recovered benzo[d]thiazole with no formation of the
trimer (15) following an acid quench. However when the same
The proposed mechanism (Scheme 2) reflects the essential
role of oxygen and magnesium in the formation of the benzo-
[d]thiazole trimer (15). Following deprotonation of benzo[d]-
thiazole (8) with EtMgCl, benzothiazol-2-yl MgCl (9b) is oxi-
dized to benzo[d]thiazole peroxide (16) by oxygen. Magnesium
then reduces the peroxide to benzothiazolone (17a). Nucleophilic
addition of benzothiazol-2-yl MgCl (9b) to benzothiazolone
(17a) results in the metalated dimer 18.
The metalated dimer (18) aromatizes to dimer (19) followed
by the nucleophilic addition of a second equivalent of ben-
zothiazol-2-yl MgCl (9b) to form the metalated trimer (20).
Quenching the reaction mixture with acidic water then leads to
benzo[d]thiazole trimer (15).
(7) Kenney, B. D. et al. Benzothiazole Grignard Chemistry: Methodology
for Preparation of a Benzothiazole Ketone. Abstract of Papers, 37th National
Organic Chemistry Symposium, Bozeman, MT, June 2001, Abstract 98.
9800 J. Org. Chem., Vol. 72, No. 25, 2007

mixture was transferred into a cold (5-10 °C) mixture of 2 M HCl
(300 mL, 600 mmol) and EtOAc (150 mL) while stirring continued.
The resulting brown mixture was stirred for 5 min in a cool water
bath (7-10 °C). The layers were separated, and the aqueous layer
was extracted with EtOAc (150 mL). The combined brownish
organic layers were sequentially washed with saturated aq NaHCO3
(150 mL), water (150 mL), and brine (150 mL). The organics were
subsequently dried over MgSO4 (20 g) for 2 h. The crude mixture
was filtered, concentrated to ca. 140 mL in vacuo, and transferred
to an addition funnel. The crude mixture was added in a steady
stream to a vigorously stirred mixture of heptane (600 mL) and
EtOAc (60 mL). After the mixture was stirred for 15-30 min, the
solids were collected by vacuum filtration, washed with EtOAc/
heptane (1/4) (100 mL × 3), and air-dried overnight (ca. 16 h) to
yield the title compound 13 as a tan solid in 80% yield: mp 227.22
°C; rotamers observed 1H NMR (400 MHz, CDCl3) δ ppm 1.38-
1.46 (m, 9 H) 1.64-1.76 (m, 2 H) 2.05-2.13 (m, 4 H) 2.57-2.61
(m, 3 H) 2.65-2.70 (m, 3 H) 3.20-3.32 (m, 1 H) 3.51 (s, 1 H)
3.80-3.84 (m, 4 H) 5.55-5.66 (m, 2 H) 6.19 (s, 2 H) 6.50 (s, 1
H) 7.53-7.60 (m, 2 H) 7.96-7.99 (m, 1 H) 8.17-8.21 (m, 1 H);
13C NMR (100 MHz, CDCl3) δ11.9 (s,1C), 18.26, 18.34 (s,
1C), 24.1 (s, 1C), 25.38, 25.47 (s, 1C), 28.3 (s, 3C), 31.4 (s, 1C),
40.68, 40.75 (s, 1C), 55.2, 55.4 (s, 1C), 80.6 (s, 1C), 111.6 (s,
2C), 122.4 (s, 1C), 124.7 (s, 1C), 125.9 (s, 1C), 127.3 (s, 1C), 128.2
(s, 1C), 129.2 (s, 1C), 133.7 (s, 1C), 134.6 (s, 1C), 136.6 (s, 1C),
137.2 (s, 1C), 138.6 (s, 1C), 153.4 (s, 1C), 156.2 (s, 1C), 158.3 (s,
1C), 193.20, 193.26 (s, 1C); MH+ 604.3. Anal. Calcd for
C28H37N5O6S2: C, 55.70; H, 6.18; N, 11.60; O, 15.90; S, 10.62.
Found: C, 55.52; H, 6.39; N, 11.33; S, 10.35. In cases where
EtMgCl Grignard was used instead of t-BuMgCl, benzo[d]thiazole
trimer (15) was isolated as an orange solid (33% yield): mp 194-
FIGURE 2. Photomicrograph of crystalline benzo[d]thiazole trimer
(15).
The benzo[d]thiazole trimer8 was isolated and slowly crystal-
lized from dichloromethane and pentane. The trimer was isolated
as orange crystalline clusters (Figure 2) in the absence of any
apparent extended conjugation present in its molecular structure.
Evidence of a possible tautomer was not apparent in our
spectroscopic data analysis as a plausible explanation for the
orange-colored benzo[d]thiazole trimer.
When the initial optimization of the coupling of various
activated acids with benzothiazol-2-yl magnesium halide 9a and
9b failed to yield ketone 13 reproducibly on large scale, an
investigation into the parameters leading to the troublesome
benzo[d]thiazole trimer process impurity 15 was undertaken.
Atmospheric oxygen, metallic magnesium, and excess ben-
zothiazol-2-yl MgCl were shown as interdependent critical
parameters in enabling the unexpected degradation pathway to
benzo[d]thiazole trimer (15). Generating benzothiazol-2-yl MgCl
(9b) with excess t-BuMgCl and carefully excluding oxygen from
the reaction vessel minimized the trimer (15) byproduct.
Utilizing the optimized procedure to ketone 13, the demand for
drug substance (1) was readily met in additional synthetic
campaigns.
1
212 °C; H NMR (400 MHz, DMSO-d6) δ 6.76 (td, J ) 7.5, 1.1
Hz, 1H), 6.84 (dd, J ) 7.5, 1.1 Hz, 1H), 7.01 (td, J ) 7.5, 1.1 Hz,
1H), 7.19 (dd, J ) 7.5, 1.1 Hz, 1H), 7.44-7.56 (m,4H), 7.95 (dd,
J ) 8.2, 1.1 Hz, 2H), 8.15 (dd, J ) 8.2, 1.1 Hz, 2H), 8.70 s,1H,
NH); 13C (100 MHz, DMSO-d6) δ 78.9 (s, 1C), 110.3 (d, 1C), 120.2
(d, 1C), 121.2 (d, 1C), 122.4 (d, 2C), 122.8 (s, 1C), 123.0 (d, 2C),
125.6 (d, 2C), 126.3 (d, 1C), 126.5 (d, 2C), 135.2 (s, 2C), 145.0
(s, 1C), 153.1 (s, 2C), 174.7 (s, 2C); MH+ 404.1. Anal. Calcd for
C21H13N3S3: C, 62.50; H, 3.25; N, 10.41; S, 23.84. Found: C,
62.25; H, 3.25; N, 10.41; S, 23.84.
2-(Benzo[d]thiazol-2-yl)-2,3-dihydro-2,2′-bibenzo[d]thiazole
(15). To a solution of 8 (2.6 g, 19.2 mmol) in 2 mL of THF was
added 2 M EtMgCl (9.72 mL, 19.4 mmol) while the reaction
temperature was maintained between 0 and 5 °C. Solution 8 was
allowed to stir for 10 min at 0-5 °C. In a separate flask, EtMgCl
(5.1 mL, 10.2 mmol) was added to a solution of benzothiazol-2-
one (17b) (1.40 g, 9.25 mmol) in 3 mL of THF while the reaction
temperature was maintained between 0 and 5 °C. Solution 17b was
allowed to stir for 10 min. Solution 8 was transferred via cannula
to solution 17b while the reaction temperature was maintained
between 0 and 5 °C. The reaction temperature was allowed to warm
to rt. The in situ product was identified by HPLC and LC/MS as
the trimer 15.
Experimental Section
[(1S)-1-(2-Benzothiazolylcarbonyl)-4-[[imino[[(4-methoxy-
2,3,6-trimethylphenyl)sulfonyl]amino]methyl]amino]butyl]car-
bamic Acid 1,1-Dimethylethyl Ester (13). To a slurry of 1,1′-
carbonyldiimidazole (6.48 g, 40 mmol) in THF (23 mL) at rt was
added a solution of Boc-Arg (Mtr)-OH (10, 18.00 g, 33.3 mmol)
in THF (54 mL) over 2 min via cannula. The resulting clear light
yellow imidazolide solution (11b) was stirred at rt for 5 min and
then concentrated to ca. 50 mL in vacuo for 30 min. In a separate
reaction vessel was added a solution of benzo[d]thiazole (55.70 g,
400 mmol) in THF (100 mL) to a solution of t-BuMgCl (200 mL,
400 mmol) in THF (100 mL) at 4 °C over 2 h. After the addition,
the internal temperature of the resulting benzothiazol-2-yl MgCl
(9b) solution was maintained between 6 and 8 °C while stirring
continued for 10 min. The imidazolide solution (11b) was trans-
ferred into an addition funnel and subsequently added into the
benzothiazol-2-yl MgCl (9b) solution over 25 min. The resulting
dark reddish solution was stirred at 10 °C for 15 min. The reaction
Acknowledgment. We are grateful to Dr. Lanny Liebeskind
and Dr. Ahmed Abdel-Magid for invaluable discussions and
feedback. We are also appreciative to Diane Gauthier for NMR
spectroscopic contributions and Dr. George Bu for mass spectra.
Supporting Information Available: 1H NMR, 13C NMR,
DEPT, MS, and elemental analysis and X-ray crystal structure. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(8) Vela-Becerra, J.; Sharma, P.; Cabrera, A.; Alvarez, C.; Toscano, A.;
Penieres, G. Heterocycl. Commun. 2000, 6, 553-556.
JO7019543
J. Org. Chem, Vol. 72, No. 25, 2007 9801