Optimized synthesis of L-m-tyrosine suitable for chemical scale-up

Article abstract of DOI:10.1021/op700093y

This paper demonstrates how L-m-tyrosine 1 can be synthesized on larger-scale via enzyme-catalyzed kinetic resolution of N-acyl m-tyrosine methyl ester 4. N-Acyl m-tyrosme methyl ester 4 was prepared by a modification of Erlenmeyer's azalactone synthesis followed by hydrogenation of the resultant dehydroamino acid 12. The optimized four-step synthesis utilizes cheap and readily available starting materials and circumvents difficult purification protocols.

Full text of DOI:10.1021/op700093y

Organic Process Research & Development 2007, 11, 1069–1075
Optimized Synthesis of L-m-Tyrosine Suitable for Chemical Scale-Up
Cara E. Humphrey,* Markus Furegati, Kurt Laumen, Luigi La Vecchia, Thomas Leutert, J. Constanze D. Müller-Hartwieg, and
Markus Vögtle
Preparations Laboratories, DiscoVery Technologies, NoVartis Institutes for Biomedical Research, Basel, Switzerland
Abstract:
form of stearyl-tyrosine, as a adjuvant added to antigens or
14
vaccines to increase immune response. The unnatural amino
acid was found to be a key motif in three classes of peptidic-
This paper demonstrates how L-m-tyrosine 1 can be synthesized
on larger-scale via enzyme-catalyzed kinetic resolution of N-acyl
m-tyrosine methyl ester 4. N-Acyl m-tyrosine methyl ester 4 was
prepared by a modification of Erlenmeyer’s azalactone synthesis
followed by hydrogenation of the resultant dehydroamino acid 12.
The optimized four-step synthesis utilizes cheap and readily
available starting materials and circumvents difficult purification
protocols.
1
5–17
nucleoside antibiotics; the mureidomycins,
cins and more recently the napsamycins.
the pacidamy-
18
19,20
L-m-Tyrosine 1 is an unnatural amino acid that is both
difficult and expensive to obtain in large quantities required for
chemical scale-up. The price of L-m-tyrosine increases dramati-
cally by a factor of approximately 350 compared to the natural
amino acid L-tyrosine.
Introduction
R-Amino acids play an important role in drug discovery.1
The unnatural amino acid L-m-tyrosine, 1, has found broad
application in medicinal chemistry and in the study of metabolic
2,3
pathways in the central nervous system since its early
synthesis and proof of configuration. m-Tyrosine and ana-
4–6
Surveying the literature revealed that N-R-Cbz-L-m-tyrosine
2 and O-benzyl-N-Boc-L-m-tyrosine 3 have been prepared
21
22
logues thereof have shown application in the treatment of
7
,8
9
10,11
Parkinson’s disease, Alzheimer’s disease, athritis,
as
on bench scale by rhodium-catalyzed asymmetric hydrogena-
12
13
23
agents for pancreatic disorders, in agrichemicals and, in the
tion as shown in Scheme 1. The amino acrylic acid methyl
esters were prepared by Wadsworth–Emmons reaction between
Boc- or Cbz-amino-(dimethoxy-phosphoryl)-acetic acid methyl
ester and the corresponding substituted benzaldehyde.
*
To whom correspondence should be addressed. Tel: +41 61 6962970. Fax:
+
(
41 61 6968757. E-mail: cara.brocklehurst@novartis.com.
1) (a) For books covering amino acid synthesis on a large scale, see:
Blaser, H. U. Asymmetric Catalysis on Industrial Scale; Wiley-VCH:
Weinheim, 2004; ISBN 3-527-30631-5. (b) Ager, D. Handbook of
Chiral Chemicals, 2nd ed.; CRC Press: Boca Raton, 2006; ISBN
24
L-m-Tyrosine 1 has been also prepared by Williams using
an optically active oxazinone as outlined in Scheme 2. This
route, although highly selective and resulting in high ee’s, would
be of considerable expense if scaled up to kilogram scale due
1
-57444-664-9. For reviews dealing with the synthesis of amino acids
and derivatives, see:(c) Duthaler, R. O. Tetrahedron 1994, 50, 1539–
1
8
2
650. (d) Perdih, A.; Sollner Dolenc, M. Curr. Org. Chem. 2007, 11,
01–832. (e) Pascal, R.; Boiteau, L.; Commeyras, A. Top. Curr. Chem.
005, 259, 69–122. (f) Cativiela, C.; Diaz-de-Villegas, M. D. Tetra-
25
to the high cost of the auxiliary.
As the first method of choice, asymmetric hydrogenation
was used to prepare L-m-tyrosine 1 on large scale in our
hedron: Asymmetry 2007, 18, 569–623.
(
2) Hollunge, G.; Persson, S. A. Acta Pharmacol. Toxicol. 1974, 34, 391–
3
98.
(
(
(
(
(
(
(
3) Fell, V.; Greenway, A. M.; Hoskins, J. A. Biochem. Med. 1979, 22,
46–255.
4) Sealock, R. R.; Speeter, M. E.; Schweet, R. S. J. Am. Chem. Soc.
(14) Penney, C. L.; Dionne, G.; Nixongeorge, A.; Bona, C. A. Vaccine
1994, 12, 629–632.
2
(15) Isono, F.; Inukai, M.; Takahashi, S.; Haneishi, T.; Kinoshita, T.;
Kuwano, H. J. Antibiot. 1989, 42, 667–673.
1
951, 73, 5386–5388.
5) Tong, J. H.; Petitcle, C.; Diorio, A.; Benoiton, N. L. Can. J. Bio-
chemistry 1971, 49, 877–881.
(16) Haneishi, T.; Inukai, M.; Shimizu, K.; Isono, F.; Sakaida, Y.; Kinoshita,
T. Eur. Pat. Appl. EP 253413, 1988.
6) Byrkjeda, A.; Mostad, A.; Romming, C. C. Acta Chem. Scand. 1974,
B28, 750–756.
(17) Haneishi, T.; Inukai, M.; Shimizu, K.; Isono, F.; Sakaida, Y.; Kinoshita,
T. Eur. Pat. Appl. EP 317292, 1989.
7) Ungerstedt, U.; Fuxe, K.; Goldstei, M.; Battista, A.; Ogawa, M.;
Anagnoste, B. Eur. J. Pharmacol. 1973, 21, 230–237.
8) Ahmad, S.; Phillips, R. S.; Stammer, C. H. J. Med. Chem. 1992, 35,
(18) Chen, R. H.; Buko, A. M.; Whittern, D. N.; McAlpine, J. B. J. Antibiot.
1989, 42, 512–520.
(19) Chatterjee, S.; Nadkarni, S. R.; Vijayakumar, E. K. S.; Patel, M. V.;
Ganguli, B. N.; Fehlhaber, H. W.; Vertesy, L. J. Antibiot. 1994, 47,
595–598.
1
410–1417.
9) Coburn, C. A.; Stachel, S. J.; Vacca, J. P. PCT Int. Appl. WO
004062625, 2004.
10) Head, J. C.; Archibald, S. C.; Warrellow, G. J.; Porter, J. R. PCT Int.
Appl. WO 19961465, 1999.
11) Inaba, T.; Haas, J.; Shiozaki, M.; Littman, N. M.; Yasue, K.; Andrews,
S. W.; Sakai, A.; Fryer, A. M.; Matsuo, T.; Laird, E. R.; Suma, A.;
Shinozaki, Y.; Hori, Y.; Imai, H.; Negoro, T. PCT Int. Appl. WO
2
(20) Nadkarni, S. R.; Chatterjee, S.; Patel, M. V.; Desikan, K. R.; Ganguli,
B. N.; Fehlhaber, H. W.; Rupp, R. H. Eur. Pat. Appl. EP 487756,
1992.
(
(
(21) Nicolaou, K. C.; Murphy, F.; Barluenga, S.; Ohshima, T.; Wei, H.;
Xu, J.; Gray, D. L. F.; Baudoin, O. J. Am. Chem. Soc. 2000, 122,
3830–3838.
2
005058884, 2005.
(22) Wang, W.; Xiong, C. Y.; Zhang, J. Y.; Hruby, V. J. Tetrahedron 2002,
58, 3101–3110.
(
12) Okamoto, S.; Okada, Y.; Okunomiya, A.; Naito, T.; Kimura, Y.;
Yamada, M.; Ohno, N.; Katsuura, Y.; Nojima, H.; Shishikura, T. Eur.
Pat. Appl. EP 284632, 1988.
(23) Knowles, W. S. Ang. Chem., Int. Ed. 2002, 41, 1998–2007.
(24) Bender, D. M.; Williams, R. M. J. Org. Chem. 1997, 62, 6690–6691.
(25) Available in 1 g quantities from Sigma Aldrich at approximately 130
USD/g.
(
13) Weston, L. A.; Bertin, C.; Schroeder, F. PCT Int. Appl. WO
2
006086474, 2006.
1
0.1021/op700093y CCC: $37.00

2007 American Chemical Society
Vol. 11, No. 6, 2007 / Organic Process Research & Development
•
1069
Published on Web 10/17/2007

Scheme 1. Synthesis of L-m-tyrosine via rhodium-catalyzed
asymmetric hydrogenation of amino acrylic acid methyl
esters
Scheme 3. Proposed hydrolase-catalyzed kinetic resolution
of N-acyl m-tyrosine methyl ester 4
Scheme 4. Synthesis of racemic N-acyl m-tyrosine methyl
ester 4 via Knoevenagel condensation
Scheme 2. Synthesis of L-m-tyrosine via condensation of
m-(benzyloxy)benzyl bromide with an optically active
oxazinone
Preparations Laboratories. However, the reaction was not
successful. The DuPHOS catalyst required experienced handling
due to its high sensitivity to moisture and air. In multiple trials
the reaction stalled part of the way through, probably due to
catalyst poisoning. Because of time pressure we decided to seek
an alternative.
Enzymatic reactions are widely used in our unit as this
expertise is available in-house. This led us to explore the
preparation of L-m-tyrosine via enzyme-catalyzed kinetic resolu-
tion of racemic N-acyl m-tyrosine methyl ester 4 using Alcalase
racemic N-acyl m-tyrosine methyl ester 4 utilizing cheap and
readily available starting materials and with the fewest possible
chemical steps.
26,27
based on work by Roper and co-workers
and previous
28
knowledge gained in-house. The advantages of such a route
include the use of cheap and readily available enzymes, in
reactions which are easily scaled up and show very high
selectivity. This route would also provide us with both enan-
tiomers of the resultant m-tyrosine which may be of interest
during the drug discovery and early development phases.
N-Acyl m-Tyrosine Ethyl Ester via Knoevenagel Con-
densation. Scheme 4 details the initial synthesis route used to
prepare 3 kg of racemic N-acyl m-tyrosine methyl ester 4 via
Knoevenagel condensation of 3-hydroxybenzaldehyde with
dimethylmalonate. This route to unnatural amino acids was
chosen due to considerable previous experience with this robust
28
Results and Discussion
and reliable chemistry. After subsequent hydrogenation of the
double bond and benzyl protection diester 8 was converted to
oxime 9 by reaction with isoamyl nitrite. Isoamyl nitrite may
constitute a safety risk as it decomposes exothermically at
elevated temperatures. Differential scanning calorimetry (DSC)
measurements showed the onset of a large exotherm at 160 °C
We envisaged the production of L-m-tyrosine 1 by acetyl-
deprotection of the product of kinetic resolution, as detailed in
Scheme 3. Previous experience told us that the N-acetyl
protected amino acid methyl ester would be a good substrate
for hydrolase-catalyzed kinetic resolution. In order to make this
process valuable, we needed to find a large-scale synthesis of
-1
of 1857 J g . Oxime hydrolysis and removal of the benzyl
protecting group gave N-acyl m-tyrosine methyl ester 4 in 70%
overall yield. Careful purification of benzyl-protected amino acid
10 was required before hydrogenation. Large amounts of water
(
(
(
26) Roper, J. M.; Bauer, D. P. Synthesis 1983, 1041–1043.
27) Roper, J. M. U.S. Patent 4,414,404, 1983.
28) Laumen, K.; Ghisalba, O. Eng. Life Sci. 2006, 6, 193–194.
1
070
•
Vol. 11, No. 6, 2007 / Organic Process Research & Development

Scheme 5. Synthesis of racemic N-acyl m-tyrosine methyl
ester 4 via Erlenmeyer amino acid synthesis
Scheme 6. Alcalase-catalyzed kinetic resolution of N-acyl
m-tyrosine methyl ester 4
12 occurred in a single reaction to give N-acyl m-tyrosine methyl
ester 4 in quantitative yield (overall yield 56%).
When this reaction was scaled to 100 g we encountered a
number of problems with the isolation of azalactone 11. When
the reaction mixture was poured into a mixture of ethanol/water
(1:2) the major product isolated was acid 13 formed by
hydrolysis of the required product. In order to circumvent this
problem, the reaction mixture from the first step was poured
directly into methanol containing sodium acetate, thus reducing
the two-step procedure to a one-pot procedure (see Scheme 5,
left hand side- branch). This approach showed a yield of 68%,
more than 12% higher in yield than the two steps performed
individually.
were required for reaction quenching, otherwise any residual
zinc poisoned the palladium catalyst used in hydrogenation and
stalled the reaction.
The synthetic route was high yielding but constituted six
chemical steps. It became obvious that a shorter route would
not only save considerable time but might prove to be more
economically favorable.
Enzymatic Kinetic Resolution. Finally, racemic N-acyl
m-tyrosine methyl ester 4 was converted to (S)-N-acyl m-
tyrosine 5 via Alcalase (protease from Bacillus licheniformis)-
catalyzed kinetic resolution, followed by acetyl-deprotection to
give L-m-tyrosine 1 in 42% yield over two steps (Scheme 6).
The enzyme was highly selective for the (S)-enantiomer, thus
giving the required L-m-tyrosine 1 in an excellent ee of >99.5%.
The Alcalase shows a remarkably broad substrate tolerance
and high enantioselectivity against a range of nonproteinogenic
racemic amino acid derivatives. It has been previously shown
to hydrolyze N-acetyl-protected amino acid esters of mono-,
N-Acyl m-Tyrosine Methyl Ester via Erlenmeyer Azlac-
tone Synthesis. A route to N-acyl m-tyrosine methyl ester 4,
based upon the Erlenmeyer synthesis of amino acids via
29–31
oxazolone formation, is outlined in Scheme 5 (main branch).
In a trial scale reaction, m-benzyloxybenzaldehyde was
reacted with N-acetylglycine to yield 2-methyl-oxazol-5-one 11
in high yield as a bright yellow solid. The presence of a single
(
Z)-geometric isomer for this oxazolone was confirmed by X-ray
analysis.
The ring was opened using sodium acetate in methanol to
28
32
di- or trisubstituted phenylanilines, and even tert-leucine,
yield the acrylic acid methyl ester in moderate yield. Purification
of the solid formed by trituration with 50% EtOAc in hexanes
gave the product as a pale yellow crystalline solid. Simultaneous
benzyl deprotection and hydrogenation of dehydroamino acid
with high enantioselectivity.
Conclusions
In summary, we have developed a new enzymatic method
to prepare L-m-tyrosine, which is complementary to asymmetric
(
(
(
29) Erlenmeyer, E. Justus Liebigs Ann. Chem. 1893, 275, 1–20.
30) Dakin, H. D. J. Biol. Chem. 1929, 439, 446.
31) Konkel, J. T.; Fan, J. F.; Jayachandran, B.; Kirk, K. L. J. Fluorine
Chem. 2002, 115, 27–32.
(32) Laumen, K.; Ghisalba, O.; Auer, K. Biosci. Biotechnol. Biochem. 2001,
65, 1977–1980.
Vol. 11, No. 6, 2007 / Organic Process Research & Development
•
1071

hydrogenation. Such a method might be considered whenever
both enantiomers of the unnatural amino acid are required.
Additionally, the synthesis of key intermediate N-acyl m-
tyrosine methyl ester 4 has been successfully reduced from an
six-step to a two-step procedure, starting from cheap and readily
available starting materials, requiring minimal purification, and
unlike the longer synthesis in which isoamyl nitrite was utilized,
it does not include problematic reagents. The overall yield for
the optimized two-step synthesis is comparable to the six-step
synthesis at 68%; however, the shortened procedure developed
herein has the following four advantages: (a) the time needed
for the two-step synthesis is far less than performing a six-step
synthesis; (b) the space/time yield is higher compared to the
longer six-step synthesis; (c) the two-step synthesis developed
is not only ecological but economical more sound, and finally,
was monitored by HPLC using the conditions shown below.
After 24 h, the reaction mixture was cooled to room temper-
ature, and the toluene removed by distillation. The residue was
dissolved in dichloromethane (10 L) and washed with brine (5
L) and twice with saturated sodium bicarbonate (2 × 5 L). The
combined aqueous phases were extracted with dichloromethane
(2 × 5 L), and the combined organic phases were dried over
anhydrous sodium sulfate and concentrated in vacuo to give a
clear red colored oil. The oil was stirred under hexane (5 L)
until pink crystals formed, which were removed by filtration
1
(5106 g, 97%). Mp 73.6–73.7 °C. H NMR (400 MHz, DMSO-
6
d ) δ 3.76 (3H, s), 3.79 (3H, s), 6.85 (1H, s), 6.86 (1H, d, J )
7
7
.8 Hz), 6.91 (1H, d, J ) 7.8 Hz), 7.24 (1H, t, J ) 7.8 Hz),
.64 (1H, s), 9.78 (1H, s). C NMR (101 MHz, DMSO-d ,
6
13
DEPT) δ 53.4 (t), 115.8 (s), 119.0 (s), 121.6 (s), 125.4 (q),
30.8 (s), 134.0 (q), 142.7 (s), 158.3 (q), 164.6 (q), 167.1 (q).
(d) this synthesis could be easily adapted to amino acids
1
containing different aryl-substitution patterns.
HRMS (FAB) calcd for C12
H O
12 5
+ Na 259.0577, found
12 5
259.0576. Anal. Calcd for C12H O : C, 61.01; H, 5.12; N, 0.
Found: C, 61.04; H, 5.08; N, < 0.3. HPLC (Macherey-Nagel
CC 70/4 Nucleosil 100-3 C18 HD, 20–100% MeCN (6′) 100%
MeCN (1.5′) 100–20% MeCN (0.5′)) Rf (aldehyde) ) 1.28 min;
Experimental Section
All reagents were obtained commercially and used as
received unless otherwise noted. The Alcalase used was Novo
Nordisk, Alcalase 2.5 L, type DX, PMN04666, 2.67 AU/g
proteolytic activity. TLC was performed on Merck silica gel
plates 60 F-254, Art. no. 5729, with detection by UV (254 nm)
Rf (product) ) 2.72 min; purity 78%.
2-(3-Hydroxy-benzyl)-malonic Acid Dimethyl Ester 7. Ma-
lonic acid dimethyl ester 6 (5106 g, 21.6 mol) was dissolved
in methanol (20 L) containing Pd/C (5%) (120 g) and shaken
under 1.2 bar H for 4 h. After 100% H consumption the
2 2
or molybdate dip (3 g KMnO
00 mL H O).
Reverse-phase HPLC analyses were performed on an Agi-
lent-1100 machine using a Macherey-Nagel CC 70/4 Nucleosil
4 2 2
; 20 g K CO ; 5 mL 5% NaOH;
3
2
reaction was stopped, and the reaction mixture was filtered
through a plug of Hyflo to remove the carbon and any palladium
impurities. The methanolic solution obtained was concentrated
100-3 C18 HD column, acetonitrile and water both containing
1
in vacuo to yield a clear yellow oil (5164 g, > 99%). H NMR
0.05% TFA, a column temperature of 35 °C, with a flow rate
(
400 MHz, DMSO-d
6
) δ 2.98 (2H, d, J ) 8.2 Hz), 3.59 (6H,
of 1.0 mL/min and measuring at 216 nm. The standard gradient
used was 5–100% MeCN over 6 min, 100% MeCN for 1.5
min followed by 100–5% MeCN over 0.5 min.
Normal-phase chiral separations were performed using either
a Chiralpak AD-H 250 × 4.0 mm column in the case of 4, a
hexane/ethanol mixture at 80:20, a column temperature of 20
s), 3.77 (1H, t, J ) 7.8 Hz), 6.57 (1H, s), 6.59 (1H, s), 7.03
1
3
(
d
1
(
1H, t, J ) 7.0 Hz), 9.31 (1H, s). C NMR (101 MHz, DMSO-
, DEPT) δ 34.8 (d), 53.0 (t), 53.4 (s), 114.3 (s), 116.2 (s),
19.8 (s), 130.0 (s), 139.6 (q), 158.0 (q), 169.5 (q). HRMS
FAB) calcd for C12 + Na 261.0733, found 261.0733;
calcd for C12 + K 277.0473, found 277.0472. HPLC
Macherey-Nagel CC 70/4 Nucleosil 100-3 C18 HD, 20–100%
MeCN (6′) 100% MeCN (1.5′) 100–20% MeCN (0.5′)) R
.49 min; purity 91%.
-(3-Benzyloxy-benzyl)-malonic Acid Dimethyl Ester 8. Di-
6
14 5
H O
14 5
H O
°C, with a flow rate of 1.0 mL/min and measuring at 210 nm;
(
or a Chirobiotic TAG 250 × 4.6 mm column in the case of 1,
)
a water/methanol mixture at 40:60, a column temperature of
f
2
2
0 °C, with a flow rate of 1.5 mL/min and measuring at 215
nm.
NMR was performed using a 400 MHz Varian machine,
2
methyl ester 7 (5164 g, 21.7 mol) was dissolved in acetone (15
L) in a 25-L reactor containing potassium carbonate (3600 g,
1
AS 400 Oxford. H shifts were referenced to DMSO-d
ppm and CDCl at 7.25 ppm with tetramethylsilane as internal
standard for H NMR. MS was measured using VG Platform
Fisons Instruments), Spectraflow 783 Detector, HP 1100 Series
6
at 2.49
26.1 mol) and dimethylformamide (90 mL). Benzylbromide (3
3
1
L, 25.3 mol) was added slowly via a dropping funnel to give a
yellow suspension. The reaction mixture was heated to 68 °C
(
(internal temperature) and heated at reflux for 20 h. After this
HPLC. Melting points were measured using a Büchi, B-545
machine.
time a white solid had formed, which was removed by filtration.
N-Acyl m-Tyrosine Ethyl Ester via Knoevenagel Con-
densation (according to Scheme 4). 2-(3-Hydroxy-benzylidene)-
malonic Acid Dimethyl Ester 6. 3-Hydroxybenzaldehyde (2705
g, 22.2 mol) was added to a 25-L reactor containing dimeth-
ylmalonate (2538 mL, 22.2 mol) dissolved in toluene (10 L),
acetic acid (250 mL, 4.37 mol) and pyridine (25 mL, 0.43
mmol) were added to give a light yellow cloudy solution. The
mixture was heated to 110 °C (internal temperature) and stirred
The filtrate was concentrated in vacuo to give the product as a
1
yellow oil (7471 g, > 99%). H NMR (400 MHz, DMSO-d
6
)
δ 3.04 (1H, d, J ) 8.2 Hz), 3.07 (1H, d, J ) 7.0 Hz), 3.58
(6H, s), 3.87 (1H, t, J ) 7.8 Hz), 5.04 (1H, s), 6.75 (1H, m),
6.83–6.87 (2H, m), 7.17 (3H, t, J ) 7.8 Hz), 7.29–7.43 (5H,
13
m). C NMR (101 MHz, DMSO-d , DEPT) δ 34.8 (d), 53.0
6
(t), 53.2 (s), 69.7 (d), 113.6 (s), 115.9 (s), 121.8 (s), 128.4 (s),
128.5 (s), 129.1 (s), 130.1 (s), 137.7 (q), 139.8 (q), 159.0 (q),
(using an overhead stirrer) at reflux for 24 h under nitrogen
169.5 (q). HRMS (FAB) calcd for C19
H O
20 5
+ Na 351.1203,
with a Dean–Stark trap to collect the water formed. The reaction
found 351.1202; calcd for C19
H O + K 367.0942, found
20 5
1
072
•
Vol. 11, No. 6, 2007 / Organic Process Research & Development

3
67.0942. HPLC (Macherey-Nagel CC 70/4 Nucleosil 100-3
C18 HD, 20–100% MeCN (6′) 100% MeCN (1.5′) 100–20%
MeCN (0.5′)) R ) 4.55 min; material taken on crude to next
step.
-(3-Benzyloxy-phenyl)-2-[hydroxyimino]-propionic Acid
catalysis in the next step. The organic phase was dried
over anhydrous sodium sulfate and concentrated in vacuo
to give a brown oil (2888 g, 88%), which partially
crystallized on standing. Mp 78.3–78.9 °C. H NMR (400
6
MHz, DMSO-d ) δ 1.78 (3H, s), 2.82 (1H, dd, J ) 9.4
f
1
3
Methyl Ester 9. Dimethyl ester 8 (3328 g, 10.1 mol) was added
to a 25-L reactor containing methanol (15 L) to give a clear
yellow solution. Sodium methoxide (2160 mL, ∼5.4 M in
MeOH, 11.8 mol) was added dropwise, resulting in a small
exotherm. The reaction was not allowed to heat over 40 °C,
and once addition was finished, the reaction was cooled to 10
and 13.7 Hz), 2.96 (1H, dd, J ) 5.5 and 13.7 Hz), 3.57
(3H, s), 4.43 (1H, m), 5.05 (2H, s), 6.78 (1H, d, J ) 7.8
Hz), 6.84–6.88 (2H, m), 7.18 (1H, t, J ) 7.8 Hz),
1
3
7.31–7.43 (5H, m), 8.33 (1H, d, J ) 7.8 Hz). C NMR
(101 MHz, DMSO-d , DEPT) δ 22.9 (t), 37.4 (d), 52.5
6
(t), 54.2 (s), 69.7 (d), 113.5 (s), 116.3 (s), 122.1 (s), 128.4
(s), 128.5 (s), 129.1 (s), 129.9 (s), 137.8 (q), 139.5 (q),
158.9 (q), 170.0 (q), 172.9 (q). HRMS (FAB) calcd for
°C (internal temperature). Isoamyl nitrite (1500 mL, 11.1 mol)
was added to the cooled mixture over 30 min with care not to
increase the temperature over 15 °C. Note: the DSC of isoamyl
nitrite showed a large exotherm starting at 160 °C. The reaction
mixture was stirred at 10 °C for 2 h, and after this time the
solution was allowed to warm to room temperature. 2 N HCl
C
C
19
H
H
21NO
21NO
4
+ H 328.1543, found 328.1543; calcd for
+ Na 350.1363, found 350.1362. Anal. Calcd
19
4
4
for C19H21NO : C, 69.71; H, 6.47; N, 4.28. Found: C,
69.26; H, 6.36; N, 3.95. HPLC (Macherey-Nagel CC 70/4
(5 L) was added to pH 7, and the mixture was allowed to stir
Nucleosil 100-3 C18 HD, 20–100% MeCN (6′) 100%
for 30 min before concentrating to remove the methanol. The
thick beige suspension was diluted with methanol (1 L) and
filtered to remove the solid. The solid was placed in a large
rotation flask and azeotroped with toluene (2 × 2.5 L) to remove
residual water. The beige solid formed was resuspended in
hexane (5 L) and stirred at 0 °C for 1–2 h. The solid was
MeCN (1.5′) 100–20% MeCN (0.5′)) R
>99%.
2-Acetylamino-3-(3-hydroxy-phenyl)-propionic Acid Methyl
Ester 4. Benzyl-protected m-tyrosine 10 (1800 g, 5.5 mol)
was dissolved in methanol (18 L) containing Pd/C (10%)
f
) 3.74 min; purity
(180 g) and shaken under 1.2 bar H
2
for 4 days. After
removed by filtration and dried under vacuum at 40 °C
100% H consumption the reaction was stopped, and the
2
1
overnight (2527 g, 83%). H NMR (400 MHz, DMSO-d
6
) δ
reaction mixture was filtered through a plug of Hyflo to
remove the carbon and any palladium impurities. The
methanolic solution obtained was concentrated in vacuo
to yield a viscous brown oil (1323 g, >99%). The product
could be crystallized according to the procedure used in
3.69 (3H, s), 3.80 (2H, s), 5.02 (2H, s), 6.75 (1H, d, J ) 7.8
Hz), 6.82–6.84 (2H, m), 7.17 (1H, t, J ) 7.8 Hz), 7.30–7.43
13
(
5H, m), 12.50 (1H, s). C NMR (101 MHz, DMSO-d
6
, DEPT)
δ 30.7 (d), 52.9 (t), 69.7 (d), 112.9 (s), 116.0 (s), 121.7 (s),
1
1
1
28.4 (s), 128.5 (s), 129.1 (s), 130.1 (s), 137.7 (q), 138.6 (q),
50.0 (q), 159.1 (q), 164.8 (q). HRMS (FAB) calcd for
the next synthesis. Mp 112.6–112.7 °C. H NMR (400
MHz, DMSO-d ) δ 1.78 (3H, s), 2.76 (1H, dd, J ) 9.4
6
C
C
17
H
17NO
17NO
4
+ Na 322.1050, found 322.1049. Anal. Calcd for
: C, 68.22; H, 5.72; N, 4.68. Found: C, 68.01; H,
and 13.7 Hz), 2.89 (1H, dd, J ) 5.5 and 13.7 Hz), 3.57
(3H, s), 4.38 (1H, m), 6.58–6.60 (3H, m), 7.03 (1H, t, J
17
H
4
1
3
5
.68; N, 4.56. HPLC (Macherey-Nagel CC 70/4 Nucleosil 100-3
C18 HD, 20–100% MeCN (6′) 100% MeCN (1.5′) 100–20%
MeCN (0.5′)) R ) 4.03 min; purity 86%.
-Acetylamino-3-(3-benzyloxy-phenyl)-propionic Acid Methyl
) 8.0 Hz), 8.31 (1H, d, J ) 7.4 Hz), 9.29 (1H, s).
C
NMR (101 MHz, DMSO-d , DEPT) δ 22.9 (t), 37.4 (d),
6
f
52.5 (t), 54.3 (s), 114.2 (s), 116.5 (s), 120.2 (s), 129.8
(s), 139.2 (q), 157.9 (q), 170.0 (q), 172.9 (q). HRMS
2
Ester 10. Oxime 9 (3000 g, 10.0 mol) was added to a 30-L
reactor containing acetic acid (14 L) and acetic anhydride
(FAB) calcd for C12
238.10733; calcd for C12
260.08922. Anal. Calcd for C12
N, 5.90. Found: C, 60.33; H, 6.23; N, 5.84. HPLC
(Macherey-Nagel CC 70/4 Nucleosil 100-3 C18 HD,
20–100% MeCN (6′) 100% MeCN (1.5′) 100–20% MeCN
H
15NO
4
+ H 238.10738, found
+ Na 260.08933, found
: C, 60.75; H, 6.37;
H
15NO
4
(
5.6 L, 59.3 mol) to give a cloudy yellow solution. The
reaction mixture was warmed in a warm water bath to 40
C (internal temperature) and zinc powder (1700 g, 26
H
15NO
4
°
mol) was added in portions. The internal temperature was
kept between 40 and 45 °C throughout the zinc addition.
After complete addition, the grey mixture was stirred at
(0.5′)) R ) 2.80 min; purity >99%.
f
N-Acyl m-Tyrosine Methyl Ester via Erlenmeyer Azlac-
tone Synthesis (according to Scheme 5). 4-[1-(3-Benzyloxy-
phenyl)-meth-(Z)-ylidene]-2-methyl-4H-oxazol-5-one 11. m-
Benzyloxybenzaldehyde (47.5 g, 224 mmol) was added to a
350 -mL flask containing N-acetylglycine (28.8 g, 246 mmol)
and sodium acetate (20.6 g, 251 mmol), and acetic anhydride
(112 mL, 5.27 mmol) was added. The mixture was stirred (using
an overhead stirrer) at 115 °C for 3.5 h under nitrogen to give
a dark brown solution. The starting materials took 30 min to
completely dissolve. The reaction was monitored by TLC in
10% EtOAc in hexanes (Rf (SM) ) 0.25, Rf (P) ) 0.18). The
reaction mixture was cooled in an ice bath to 20 °C and then
poured into water (2 L) diluted with ethanol (1 L). The yellow/
4
0 °C for a further 20 h. The mixture was quenched (in
two portions due to limited reactor size) with water (12.5
L, Note: a large amount of water was used to remove any
zinc that could poison the Pd-catalyst in the next step),
then the mixture was filtered and the solid was washed
with ethyl acetate and water. The filtrate was concentrated
to remove the acetic acid, and the residue was dissolved
in water (10 L) and ethyl acetate (5 L). The aqueous phase
was extracted three more times with ethyl acetate (3 × 5
L), and the combined organic phases were washed three
times more with water (5 L). Thorough washing is
important to remove all zinc salts that may disturb
Vol. 11, No. 6, 2007 / Organic Process Research & Development
•
1073

brown solid formed upon stirring for 30 min at room temper-
ature was collected by filtration and dried at room temperature
under a filter paper over the weekend to give a yellow/brown
lumpy solid. The solid was crushed to give a bright yellow solid
(1H, d, J ) 7.8 Hz), 7.30–7.44 (7H, m), 9.67 (1H, s). 13C
NMR (101 MHz, DMSO-d , DEPT) δ 23.1 (t), 52.9 (t),
69.9 (d), 116.5 (s), 116.7 (s), 123.2 (s), 127.6 (q), 128.4
(s), 128.6 (s), 129.2 (s), 130.4 (s), 131.4 (s), 135.4 (q),
137.6 (q), 159.0 (q), 166.3 (q), 170.1 (q). HRMS (FAB)
6
(
59.7 g, 91%) and taken on crude for the next synthesis step.
1
H NMR (400 MHz, DMSO-d
.14 (1H, dd, J ) 2.6 and 8.0 Hz), 7.17 (1H, s), 7.32–7.47
6H, m), 7.73 (1H, d, J ) 7.8 Hz), 7.89 (1H, s). The sample
6
) δ 2.38 (3H, s), 5.13 (2H, s),
calcd for C19
calcd for C19
4
Anal. Calcd for C19H19NO : C, 70.14; H, 5.89; N, 4.30.
H
19NO
4
+ H 326.13854, found 326.13869;
7
(
H
19NO
4
+ Na 348.12042, found 348.12063.
1
13
contained 18% acidic by-product 13 by H NMR. C NMR
101 MHz, DMSO-d , DEPT) δ 16.0 (t), 70.0 (d), 118.2 (s),
18.4 (s), 125.5 (s), 128.6 (s), 128.6 (s), 129.1 (s), 130.3 (s),
30.6 (s), 133.5 (q), 135.0 (q), 137.4 (q), 159.0 (q), 167.5 (q),
68.0 (q). HRMS (FAB) calcd for C18 + H 294.11247,
+ Na 316.09441, found
16.09436. HPLC (Macherey-Nagel CC 70/4 Nucleosil 100-3
Found: C, 69.97; H, 5.78; N, 4.45. HPLC (Macherey-
Nagel CC 70/4 Nucleosil 100-3 C18 HD, 5–100% MeCN
(
6
1
1
1
(6′) 100% MeCN (1.5′) 100–5% MeCN (0.5′)) R
min; > 92% purity.
f
) 4.97
H
15NO
3
2-Acetylamino-3-(3-hydroxy-phenyl)-propionic Acid Methyl
Ester 4. Acrylic acid methyl ester 11 (40.8 g, 124 mmol) was
dissolved in methanol (627 mL) containing Pd/C (10%) (4.23
found 294.11238; calcd for C18H15NO
3
3
C18 HD, 5–100% MeCN (6′) 100% MeCN (1.5′) 100–5%
2 2
g) and shaken under 1.1 bar H for 4 h. After 100% H
MeCN (0.5′)) Rf (aldehyde) ) 5.57 min; Rf (product) ) 6.11 min.
consumption the reaction was stopped, and the reaction mixture
was filtered through a 3 cm plug of Hyflo to remove the carbon
and any palladium impurities. The methanolic solution obtained
was concentrated in vacuo to yield a clear pale brown gummy/
foamy solid (30.4 g, > 99%). If necessary the solid was
crystallized from 10% dichloromethane in hexanes to give a
(Z)-2-Acetylamino-3-(3-benzyloxy-phenyl)-acrylic Acid 13 as
Isolated By-Product. Isolated as a the major product from 100 g
1
scale-up of the previous reaction. H NMR (400 MHz, DMSO-
d
6
) δ 1.95 (3H, s), 5.11 (2H, s), 7.34–7.54 (9H, m), 9.48 (1H,
+
s), 12.68 (1H, br s). MS (ES ) calcd for C18
found 312; (ES ) calcd for C18
H
17NO
4
+ H 312,
-
1
H
17NO
4
– H 310, found 310.
cream colored solid. Mp 105.1–105.7 °C. H NMR (400 MHz,
HPLC (Macherey-Nagel CC 70/4 Nucleosil 100-3 C18 HD,
–100% MeCN (6′) 100% MeCN (1.5′) 100–5% MeCN (0.5′))
) 4.55 min.
Z)-2-Acetylamino-3-(3-benzyloxy-phenyl)-acrylic Acid Meth-
yl Ester 12 (one-pot procedure). m-Benzyloxybenzaldehyde
100 g, 471 mmol) was added to a 1-L flask containing
N-acetylglycine (100 g, 466 mmol) and sodium acetate
42.9 g, 522 mmol), and acetic anhydride (232 mL, 2450
6
DMSO-d ) δ 1.78 (3H, s), 2.75 (1H, dd, J ) 9.0 and 13.7 Hz),
5
R
2.88 (1H, dd, J ) 5.9 and 13.7 Hz), 3.57 (3H, s), 4.37 (1H, dd,
J ) 5.7 and 8.0 Hz), 6.57–6.60 (3H, m), 7.03 (1H, t, J ) 8.0
f
13
(
Hz), 8.31 (1H, d, J ) 7.8 Hz), 9.30 (1H, br s). C NMR (101
MHz, DMSO-d , DEPT) δ 22.9 (t), 37.4 (d), 52.5 (t), 54.3 (s),
114.2 (s), 116.5 (s), 120.2 (s), 129.8 (s), 139.2 (q), 157.9 (q),
170.0 (q), 172.9 (q). HRMS (FAB) calcd for C12 + H
6
(
H
15NO
15NO
15NO
4
(
238.10738, found 238.10739; calcd for C12
H
4
+ Na
) + Na
mmol) was added. The mixture was stirred (using an
overhead stirrer) at 115 °C for 18 h under nitrogen to give
a dark brown solution. The reaction was monitored by
260.08933, found 260.08923; calcd for 2(C12
H
4
4
497.18944, found 497.18949. Anal. Calcd for C12H15NO : C,
60.75; H, 6.37; N, 5.90. Found: C, 59.94; H, 6.09; N, 5.97.
HPLC (Macherey-Nagel CC 70/4 Nucleosil 100-3 C18 HD,
5–100% MeCN (6′) 100% MeCN (1.5′) 100–5% MeCN (0.5′))
TLC in 10% EtOAc in hexanes (Rf (SM) ) 0.25, Rf (P)
)
0
.18) and by HPLC until the aldehyde concentration
reached a minimum. The reaction mixture was cooled in
an ice bath to 20 °C then poured into methanol (2.5 L)
containing sodium acetate (76 g, 926 mmol), and the clear
brown solution was stirred at room temperature for 48 h,
after which time TLC in 50% EtOAc in hexanes indicated
near complete conversion (Rf (SM) ) 0.57, Rf (P) ) 0.13).
The reaction was concentrated in vacuo to remove the
methanol, then diluted with EtOAc (1 L) and washed with
water (2 L). The aqueous layer was washed twice more
with EtOAc (2 × 500 mL), and the combined EtOAc
layers were washed once more with brine (1 L). The
organic layer was dried over anhydrous sodium sulfate
containing charcoal (50 g) before being concentrated in
vacuo to yield a brown crystalline solid. The crude
material was triturated with 40% methylene chloride in
hexane (300 mL) followed by the addition of 100 mL of
hexane (now 30% methylene chloride in hexane), and the
pale yellow solid obtained was washed with heptane (100
R
f
) 3.17 min.
(S)-2-Acetylamino-3-(3-hydroxy-phenyl)-propionic Acid 5. In
a typical experiment, racemic N-acyl m-tyrosine methyl ester
4 (34.1 g, 144 mmol) was suspended in phosphate buffer at
pH 7.5 (300 mL) and 100 µg of Alcalase was added. The
mixture was stirred at room temperature, and the pH was kept
constant using a pH-stat by adding a 1 N NaOH solution. After
about 9 h, the conversion reached 50% (consumption of 72 mL
of the NaOH solution). The aqueous phase was extracted twice
with dichloromethane (2 × 250 mL), and then the combined
organic phases were washed with water, dried over anhydrous
sodium sulfate and concentrated in vacuo to give (R)-N-acyl
m-tyrosine methyl ester (R)-4 (16.6 g, 70 mmol, 49%) as a
colorless resin. The spectroscopic data were identical to those
of racemic N-acyl m-tyrosine methyl ester (()-4. [R] 589
-76.7 (c 1.16, CHCl
mm hexanes/EtOH 80:20 flow 1 mL/min) ee 99.2%. The crude
aqueous phase containing (S)-N-acyl m-tyrosine (S)-5 was used
for the next step without purification.
2
0
)
3
); HPLC (Chiralpak AD-H 250 × 4.0
mL) before being dried in the vacuum oven at 40 °C for
1
2
h (104 g, 68%). Mp 113.8–115.4 °C. H NMR (400
L-m-Tyrosine 1. Concentrated HCl solution (82 mL) was
added to the crude aqueous phase containing (S)-N-acyl
m-tyrosine (S)-5 (409 mL), and the mixture was heated for 2 h
MHz, DMSO-d
6
) δ 1.97 (3H, s), 3.69 (3H, s), 5.11 (2H,
s), 7.02 (1H, dd, J ) 2.1 and 8.0 Hz), 7.11 (1H, s), 7.21
1
074
•
Vol. 11, No. 6, 2007 / Organic Process Research & Development

at 100 °C followed by complete evaporation of the liquid. The
residue was redissolved in water (100 mL), and the pH was
adjusted to 6 by adding 4 N NaOH solution. The precipitate
formed was filtered and dried in vacuo to yield L-m-tyrosine 1
erey-Nagel CC 70/4 Nucleosil 100-3 C18 HD, 5–100% MeCN
(6′) 100% MeCN (1.5′) 100–5% MeCN (0.5′)) R ) 1.527 min;
> 99% purity.
f
20
(
11.1 g, 42%) as colorless crystals. Mp 278.0–278.4 °C. [R] 589
Acknowledgment
4,5
)
-7.9 (c 1, 1 N HCl) [lit.: [R]25
HPLC (Chirobiotic TAG 250 × 4.6 mm H
flow 1.5 mL/min) ee >99.5%. H NMR (400 MHz, DMSO-
) δ 2.88 (1H, dd, J ) 8.2 and 14.5 Hz), 3.05 (1H, dd, J )
.1 and 14.4 Hz), 3.79 (1H, dd, J ) 5.1 and 8.2 Hz), 6.63–6.70
D
) -7.9 (c 2, 1 N HCl)];
O/MeOH 40:60
The authors would like to thank Stephanie Rothe-Pöllet,
Dominique Ducret, Michaela Ketterer, Annina Riepp and
Heiner Schütz for technical assistance; Claudia Betschart and
Siem Veenstra for fruitful discussions; Eric Francotte for
determining the optical purities; and Trixi Wagner for perform-
ing the X-ray analysis.
2
1
d
5
6
1
(
3H, m), 7.12 (1H, t, J ) 8.5 Hz). H NMR (126 MHz, D
DEPT) δ 36.6 (d), 56.0 (s), 114.7 (s), 116.2 (s), 121.5 (s), 130.6
s), 137.0 (q), 156.0 (q), 174.1 (q). HRMS (FAB) calcd for
2
O,
(
Received for review May 2, 2007.
OP700093Y
C
C
9
9
H
H
11NO
11NO
3
+ H 182.08117, found 182.08121; calcd for
+ Na 204.06311, found 204.06308. HPLC (Mach-
3
Vol. 11, No. 6, 2007 / Organic Process Research & Development
•
1075