Engineered L-serine hydroxymethyltransferase from streptococcus thermophilus for the synthesis of α,α-dialkyl-α-amino acids

Article abstract of DOI:10.1002/anie.201411484

α,α-Disubstituted a-amino acids are central to biotechnological and biomedical chemical processes for their own sake and as substructures of biologically active molecules for diverse biomedical applications. Structurally, these compounds contain a quaternary stereocenter, which is particularly challenging for stereoselective synthesis. The pyridoxal-5′-phosphate (PLP)-dependent l -serine hydroxymethyltransferase from Streptococcus thermophilus (SHMT_Sth; EC 2.1.2.1) was engineered to achieve the stereoselective synthesis of a broad structural variety of α,α-dialkyl-α-amino acids. This was accomplished by the formation of quaternary stereocenters through aldol addition of the amino acids D-Ala and D-Ser to a wide acceptor scope catalyzed by the minimalist SHMT_Sth Y55T variant overcoming the limitation of the native enzyme for Gly. The SHMT_Sth Y55T variant tolerates aromatic and aliphatic aldehydes as well as hydroxy- and nitrogen-containing aldehydes as acceptors.

Full text of DOI:10.1002/anie.201411484

Angewandte
Chemie
DOI: 10.1002/anie.201411484
Biocatalysis
Engineered l-Serine Hydroxymethyltransferase from Streptococcus
thermophilus for the Synthesis of a,a-Dialkyl-a-Amino Acids**
Karel Hernandez, Igor Zelen, Giovanna Petrillo, Isabel Usꢀn, Claudia M. Wandtke,
Jordi Bujons, Jesffls Joglar, Teodor Parella, and Pere ClapØs*
Abstract: a,a-Disubstituted a-amino acids are central to
biotechnological and biomedical chemical processes for their
own sake and as substructures of biologically active molecules
for diverse biomedical applications. Structurally, these com-
pounds contain a quaternary stereocenter, which is particularly
challenging for stereoselective synthesis. The pyridoxal-5’-
phosphate (PLP)-dependent l-serine hydroxymethyltransfer-
ase from Streptococcus thermophilus (SHMTSth; EC 2.1.2.1)
was engineered to achieve the stereoselective synthesis of
Figure 1. Structures of myriocin and sphingofungines E and F.
a broad structural variety of a,a-dialkyl-a-amino acids. This
was accomplished by the formation of quaternary stereocenters
through aldol addition of the amino acids d-Ala and d-Ser to
a wide acceptor scope catalyzed by the minimalist SHMTSth
Y55T variant overcoming the limitation of the native enzyme
for Gly. The SHMTSth Y55T variant tolerates aromatic and
aliphatic aldehydes as well as hydroxy- and nitrogen-contain-
ing aldehydes as acceptors.
ters with a strict control of stereoselectivity.[1f,h,2] Biocatalytic
methods have become increasingly significant due to their
potential to confer controlled stereoselectivity and efficiency
under mild aqueous conditions.[3]
The most concise enzymatic route toward the preparation
of optically pure a,a-dialkylated b-hydroxy-a-amino acids
involves the stereoselective addition of an a-substituted a-
amino acid to an aldehyde. Aldol additions of Gly to
aldehydes have been previously achieved using Gly-depen-
dent aldolases,[4] but biocatalytic aldol additions of other a-
amino acids are rare.[3a,d–f] Furthermore, most of them are
marred by low catalytic efficiency and poor diastereomeric
control over the incoming b-hydroxy group.
We have cloned and characterized a novel pyridoxal-5’-
phosphate (PLP)-dependent l-serine hydroxymethyltransfer-
ase from Streptococcus thermophilus (SHMTSth; EC 2.1.2.1).[5]
SHMTSth was found to catalyze aldol additions of Gly to
aldehydes with good stereoselection under kinetic control
(48C).[6]
T
he a,a-disubstituted a-amino acids are central to biotech-
nological and biomedical chemical processes, e.g., the syn-
thesis of biologically active molecules with beneficial bio-
medical application in fields as diverse as immunology (e.g.,
myriocin, Figure 1), infectious diseases (e.g., the fungicides
sphingofungines E and F, Figure 1), cardiovascular diseases
(e.g., DOPA decarboxylase inhibitors), and neurology (e.g.,
lactacystin).[1]
The synthesis of chirally pure a,a-dialkylated b-hydroxy-
a-amino acids remains a challenging task, because it requires
the construction of stereogenic tetrasubstituted carbon cen-
[*] Prof. Dr. P. Clapꢀs
C. M. Wandtke
Biotransformation and Bioactive Molecules Group
Instituto de Quꢁmica Avanzada de CataluÇa IQAC-CSIC
Jordi Girona 18–26, 08034 Barcelona (Spain)
E-mail: pere.clapes@iqac.csic.es
Institut fꢃr Anorganische Chemie, Universitꢄt Gçttingen
Tammannstr. 4, 37077 Gçttingen (Germany)
Dr. T. Parella
Servei de Ressonꢅncia Magnꢆtica Nuclear
Departament de Quꢁmica
K. Hernandez, Dr. J. Bujons, Dr. J. Joglar, Prof. Dr. P. Clapꢀs
Dept Quꢁmica Biolꢂgica y Modelizaciꢂn Molecular
Instituto de Quꢁmica Avanzada de CataluÇa, IQAC-CSIC
Jordi Girona 18–26, 08034 Barcelona (Spain)
Universitat Autꢇnoma de Barcelona, Bellaterra (Spain)
[**] This work was supported by the Spanish MINECO (CTQ2012-31605
and CTQ2012-32436, BFU2012-35367), Generalitat de Catalunya
(2009 SGR 00281), ERA-IB MICINN, PIM2010EEI-00607
(EIB.10.012. MicroTechEnz-EIB., www.fkit.unizg.hr/miten), and
COSTAction: CM1303 Systems Biocatalysis. K.H. acknowledges the
CSIC for a JAE Predoctoral contract program. G.P. acknowledges the
Generalitat de Catalunya and Biochemize S.L. for a DI Predoctoral
contract.
I. Zelen
Instituto de Biologꢁa Molecular de Barcelona, IBMB-CSIC
Baldiri Reixach 13–15, 08028 Barcelona (Spain)
Prof. Dr. I. Usꢂn
Instituciꢂ Catalana de Recerca i Estudis AvanÅats (ICREA)
Instituto de Biologꢁa Molecular de Barcelona, IBMB-CSIC
Baldiri Reixach 13–15, 08028 Barcelona (Spain)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201411484.
G. Petrillo
Dept Quꢁmica Biolꢂgica y Modelizaciꢂn Molecular
Instituto de Quꢁmica Avanzada de CataluÇa, IQAC-CSIC
Instituto de Biologꢁa Molecular de Barcelona, IBMB-CSIC (Spain)
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
The a-methylserine hydroxymethyltransferase from Para-
coccus sp. AJ110402 (MSHMTPc, EC 2.1.2.7) catalyzes the
interconversion between a-methyl-l-serine and d-Ala in the
presence of tetrahydrofolate.[3e] Interestingly, SHMTSth wild
type catalyzes the same reaction using l-Ser and Gly as
substrates (Figure 2) sharing 50% sequence identity with
MSHMTPc (see Figure S1 in the Supporting Information, SI).
reaction conditions, Y55T gave the best results (37%) with
full stereoselectivity; that is, (2S,3R) configured diastereoiso-
mer 3aa (Figure 4, compound numeration: first letter refer-
ring to the substituent of R1 of the amino acid donor and
second one to the R2 of the aldehyde acceptor).
A focused library of mutants (94 clones) was generated by
site-saturation mutagenesis at residue Y55 (i.e., NNK degen-
eracy, Figures S8 and S9). d-Ser and d-Thr were screened as
putative donor substrates. We hypothesized that if d-Ser was
tolerated, then d-Ala could also be a good donor. Using 2a as
acceptor, two new variants, Y55S and Y55C, consistently
exhibited aldol activity on d-Ser (Figures S10 and S11,
Table S2), whereas no active clones were found for d-Thr.
The analogue of d-Ser, d-Cys, undergoes an irreversible side
reaction with the PLP bound to SHMTSth furnishing a thiazo-
lidine and deactivating the enzyme (Figures S12A and S13).[7]
SHMTSth Y55T, Y55S, and Y55C did not show retroaldol
activity toward l-Thr (Figure S14), which is consistent with
the results reported for the SHMT from the B. stearothermo-
philus variant Y51F.[8]
The best donor for SHMTSth variants and the wild type
was d-Ala (Figure 3). SHMTSth Y55Twas the best catalyst for
d-Ala and d-Ser donors, whereas the wild-type enzyme was
the best for Gly but was inactive with d-Ser.
Figure 2. MSHMTPc and SHMTSth catalyze the interconversion of a-
methyl-l-Ser to d-Ala and l-Ser to Gly, respectively, in the presence of
tetrahydrofolate (THF).
The alignment of the MSHMTPc and SHMTSth protein
sequences revealed three residues that are presumably
involved in the donor selectivity: T60, H70, and, to a lesser
extent, T236 of MSHMTPc, which correspond to Y55, Y65,
and H229 in SHMTSth (Figure S1).[3e] Hence, the SHMTSth
variants Y55T, Y65H, H229T, Y55T/Y65H, Y55T/H229T,
Y65H/H229T, and Y55T/Y65H/H229T were constructed,
overexpressed, purified to homogeneity, and tested as cata-
lysts for the aldol addition of d-Ser (1a) to 2a (Table 1); such
activity is absent in the wild-type enzyme. Under identical
Table 1: SHMTSth wild type and variants as catalyst for the aldol addition
of d-Ser (1a) to 2a.[a]
Figure 3. Relative initial reaction rates in % (vo/voGly_SHMT wt) of the
addition of d-Ser (1a) (black), d-Ala (1b) (grey), and Gly (1c) (white)
to 2a catalyzed by SHMTSth variants and the wild type; vo =initial
reaction rate; voGly_SHMT wt =initial reaction rate of the aldol addition of
1c to 2a for wild-type SHMTSth
.
SHMTSth catalyst
Aldol adduct 3aa [%][b]
[c]
wild type
Y55T
Y65H
H229T
–
37 (50[d])
Surprisingly, the SHMTSth wild type could also catalyze
the aldol addition of d-Ala to 2a. This could not be
anticipated because when d-Ala is incubated with SHMTSth
wild-type a transamination took place yielding pyridoxamine
phosphate, pyruvate, and apo-SHMTSth (Figure S15).[9] The
aldolase activity was regained when the acceptor aldehyde
was present (Figures S16 and S17). The transamination
reaction was reduced by 10-fold for SHMTSth Y55T, Y55S,
and Y55C variants suggesting the key role of Y55 in
controlling the catalytic promiscuity of SHMTSth (Figure
S15).
[c]
–
[c]
–
Y55T/Y65H
Y55T/H229T
Y65H/H229T
Y55T/Y65H/H229T
10
8
–
[c]
21
[a] Reaction (1 mL), d-Ser (0.1 mmol), 2a (0.1 mmol), PLP (0.3 mmol),
pH 6.5 (the amino acid is the system buffer), DMF or DMSO (20% v/v),
SHMTSth catalyst (2 mg protein). [b] Aldol product formed after 24 h.
[c] Not detected. [d] Reaction optimized (see Figures S2–S7): water/
DMSO 1:1, [aldehyde]/[d-Ser] 1:4, pH 6.5 at 258C.
2
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
The presence of a cosolvent is of paramount importance
for obtaining good conversions, even with water-soluble
aldehydes. Indeed, product formation at practical rates was
only observed in water/DMSO 1:1 as demonstrated for the
addition of d-Ala to 2m (Figure S7(1)–S7(2)). Cosolvent
molecules may cluster in the active site of enzymes affecting
the enzyme–substrate interactions in different ways, for
example, by altering the rate of conversion.[10]
active with aldehydes 2l–n. Both Y55S and Y55C variants
were generally less efficient than Y55T and the wild type,
confirming the initial rational redesign of the biocatalyst.
Aromatic aldehydes with electron-withdrawing groups, 2l and
2n, led to high-yielding reactions, whereas for those with
electron-donating groups, 2o and 2p, no aldol product was
detected. Benzaldehyde (2j) was a weak substrate, which is
consistent with previously reported studies.[11]
Next, the four SHMTSth biocatalysts were screened for the
aldol additions of 1a, 1b, and 1c to aldehydes 2a–p (Figure 4).
SHMTSth Y55T and Y55C were the best catalysts for aldol
additions of d-Ser, whereas the wild type was the best for Gly
additions (Figure 4). For d-Ala additions, SHMTSth wild type
or Y55T were the best, with the former being remarkably
Aldol additions of d-Ser (1a) to aldehydes catalyzed by
SHMTSth Y55T were highly (95:5) to fully diastereoselective,
within the limits of NMR detection (Table 2). Mechanisti-
cally, the approach of any electrophile to the enzyme–donor
complex invariably takes place at the si-face (i.e., always S-
configured C2 stereocenter) with no exceptions known so
Table 2: Reaction conversion, yields of isolated product, and diastereomeric ratios for the aldol addition of 1a–c to aldehydes 2 catalyzed by SHMTSth
Y55T and SHMTSth wild type.
R1
R2
CH2OH (a)
SHMTSth Y55T
CH3 (b)
SHMTSth wild type
H (c)
SHMTSth wild type
SHMTSth Y55T
Conv.[a] (yield)[b]
[%]
d.r.[c]
3/4
Conv.[a] (yield)[b]
[%]
d.r.[c]
3/4
Conv.[a] (yield)[b]
[%]
d.r.[c]
3/4
Conv.[a] (yield)[b]
[%]
d.r.[c]
3/4
47 (28)
32 (20)
>95:5
>95:5
24 (21)
27 (11)
95:5
61 (36)
45 (22)
>95:5
–
–
72:28
95:5
15 (5)
67:33
44 (21)
31 (16)
22 (14)
>95:5
>95:5
87:13
44 (18)
25 (9)
–
93:7
91:9
–
39 (23)
38 (6)
>95:5
>95:5
86:14
–
–
–
–
–
–
34 (22)
31 (23)
40 (22)
95:5
44 (31)
40 (21)
65:35
86:14
52 (35)
71 (20)
85:15
95:5
18 (10)
33 (13)
49:51
50:50
>95:5
–
–
25 (16)
35 (20)
42:58
40:60
–
–
29 (10)
29 (15)
37:63
50:50
12 (5)
95:5
24 (12)
71:29
14 (9)
95:5
55 (29)
29:71
48 (26)
76:24
59 (48)
33:67
–
–
94 (31)
83 (51)
8:92
43[d]
8:92
92:8
86 (65)
60 (56)
87:13
50:50
21 (15)
95:5
44:56
61 (43)
[a] Conversion: percentage of aldol adduct (3+4) produced. [b] Yield of isolated product (3+4). [c] d.r.: diastereomeric ratio assessed by NMR
spectroscopy (see SI). Products were enantiomerically pure (see text). [d] Reaction was performed at analytical scale.
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
additions of Gly catalyzed by wild-type SHMTSth showed
moderate to low diastereoselectivity and conversions
(Table 2).
To characterize the structural frame of the catalytic
system, we determined the X-ray structure of apo-SHMTSth
wild type (PDBID-4WXB), in complex with Gly-PLP
(PDBID-4WXF), and with a mixture of l-Thr-PLP and Gly-
PLP (PDBID-4WXG) (Figures S91–S94). Models of external
aldimines (Figure S95) were generated for the SHMTSth
(Figures 6 and S96) based on the structure of the wild-type
l-Thr-PLP complex (Figures 6A and S93).
Figure 4. Screening of the SHMTSth wild type and variants Y55T, Y55S,
and Y55C as catalysts for the aldol additions of 1a–c to aldehydes 2a–
p. For the numerical values, see Tables S4–S6.
far.[3d,6,12] This will ensure the full enantioselectivity of the
reactions catalyzed by SHMT with any donor/acceptor
combination. Moreover, the aldol adducts obtained were
always anti (3) with respect to the a-amino and the b-hydroxy
functions (i.e., always si!re attack) (Figure 5 and SI). Only
Figure 6. Modeled structures of the active site of SHMTSth wild type
with the bound external aldimines of l-Thr (A), d-Ala (B), and d-Ser
(D), and of the SHMT Y55T variant with the bound external aldimine
of d-Ser (C). Aldimines (yellow), protein residues (grey or green,
meaning that they belong to different protein monomers), and the
mutated T55 residue (C) (orange). Models were built from the
corresponding X-ray structure (Figure S93).
These models suggest that d-Ala could form the aldimine
intermediate (Figure S95 I) with wild-type SHMTSth (Fig-
ure 6B) as that observed for the l-Thr-PLP complex. In this
ꢀ
conformation, the hydrogen on the a-carbon (H Ca), which
Figure 5. X-ray diffraction analysis for the determination of the abso-
lute configuration of 3ah as example. ORTEP-type plot displaying one
molecule with 50% probability ellipsoids.[13]
is removed to form the quinonoid intermediate (Figure S95
II), is located at 4.4 ꢀ of the E57 and at 3.9 ꢀ of a water
molecule, which appears to be activated by H bonds to the
PLP-phosphate and E57 and H126. The d-Ser aldimine model
for the Y55T variant (Figure 6C) shows that the space
generated is partially occupied by the hydroxymethyl group
of d-Ser, which establishes H bond interactions with the
hydroxy group of T55 and the carboxylate group of d-Ser. In
aldehyde 2g gave moderate diastereoselectivity (87:13 3ag/
4ag). Additions of d-Ala (1b) to aldehydes proceeded for
both the SHMTSth wild type and SHMTSth Y55T, thus allowing
to compare the stereochemical outcome for identical sub-
strates (Table 2). The SHMTSth Y55T, unlike the wild type,
catalyzed the additions with high stereoselectivity (i.e., anti
(3), except for the acceptor 2m for which the syn product (4)
was the major product). SHMTSth wild type provided syn/anti
mixtures in most cases, the anti being the major adduct,
whereas the syn was predominantly formed with aromatic
aldehydes. The high selectivity of the SHMTSth Y55T variant
for d-Ala and d-Ser can be explained by mechanistic
considerations which are consistent with the computational
models described below and in Figure S96. On the other hand,
ꢀ
this conformation, the H Ca would be even closer to residue
E57 (3.0 ꢀ) and to the water molecule (3.3 ꢀ). Hence, one of
these could operate as the catalytic base that triggers the
quinonoid formation. Preliminary results with E57 and H126
variants (Table S14) showed that none of these two residues is
essential for activity. Thus, the crystallographic water seems
ꢀ
the most plausible candidate to mediate the H Ca proton
abstraction.[3d] The structure of wild-type SHMTSth suggests
that d-Ser could also form the aldimine I complex (Fig-
ure 6D). However, as compared with the Y55T variant, here
4
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
K. K. Baldridge, D. Hilvert, Angew. Chem. Int. Ed. 2006, 45,
the hydroxy group of d-Ser would adopt an opposed
6824 – 6826; Angew. Chem. 2006, 118, 6978 – 6980; c) F. P.
Seebeck, D. Hilvert, J. Am. Chem. Soc. 2003, 125, 10158 –
10159; d) K. Fesko, M. Uhl, J. Steinreiber, K. Gruber, H.
Griengl, Angew. Chem. Int. Ed. 2010, 49, 121 – 124; Angew.
Chem. 2010, 122, 125 – 128; e) H. Nozaki, S. Kuroda, K.
Watanabe, K. Yokozeki, J. Mol. Catal. B 2009, 56, 221 – 226;
f) H. Nozaki, S. Kuroda, K. Watanabe, K. Yokozeki, J. Mol.
Catal. B 2009, 59, 237 – 242; g) M. Mꢃller, Adv. Synth. Catal.
2012, 354, 3161 – 3174; h) K. Fesko, M. Gruber-Khadjawi,
ChemCatChem 2013, 5, 1248 – 1272; i) C. L. Windle, M. Mꢃller,
A. Nelson, A. Berry, Curr. Opin. Chem. Biol. 2014, 19, 25 – 33.
[4] P. Clapꢄs, X. Garrabou, Adv. Synth. Catal. 2011, 353, 2263 – 2283.
[5] L. Vidal, J. Calveras, P. Clapꢄs, P. Ferrer, G. Caminal, Appl.
Microbiol. Biotechnol. 2005, 68, 489 – 497.
[6] M. L. Gutierrez, X. Garrabou, E. Agosta, S. Servi, T. Parella, J.
Joglar, P. Clapꢄs, Chem. Eur. J. 2008, 14, 4647 – 4656.
[7] L. V. Schirch, M. Mason, J. Biol. Chem. 1962, 237, 2578 – 2581.
[8] B. S. Bhavani, V. Rajaram, S. Bisht, P. Kaul, V. Prakash, M. R. N.
Murthy, N. Appaji Rao, H. S. Savithri, FEBS J. 2008, 275, 4606 –
4619.
[9] a) K. Shostak, V. Schirch, Biochemistry 1988, 27, 8007 – 8014;
b) R. Contestabile, A. Paiardini, S. Pascarella, M. L. di Salvo, S.
DꢅAguanno, F. Bossa, Eur. J. Biochem. 2001, 268, 6508 – 6525;
c) A. C. Eliot, J. F. Kirsch, Annu. Rev. Biochem. 2004, 73, 383 –
415; d) L. Schirch, W. T. Jenkins, J. Biol. Chem. 1964, 239, 3797 –
3800; e) L. Schirch, A. Diller, J. Biol. Chem. 1971, 246, 3961 –
3966.
[10] a) C. Mattos, C. R. Bellamacina, E. Peisach, A. Pereira, D.
Vitkup, G. A. Petsko, D. Ringe, J. Mol. Biol. 2006, 357, 1471 –
1482; b) C. Mattos, D. Ringe, Curr. Opin. Struct. Biol. 2001, 11,
761 – 764.
orientation, stabilized by H bonds with residues E57 and
ꢀ
Y65 and hindering the access to the H Ca atom, which would
be at > 5 ꢀ from the E57 and water molecule. The lack of d-
Ser reactivity with wild-type SHMTSth could thus be due to the
steric hamper caused by the hydroxy group of d-Ser, sitting
ꢀ
between the proposed catalytic water and the H Ca atom.
In summary, we have discovered the SHMTSth Y55T
variant for the efficient, highly enantio- and diastereoselec-
tive synthesis of a,a-disubstituted b-hydroxy-a-amino acids,
which was so far unprecedented for glycine utilizing aldolases.
Received: November 27, 2014
Revised: December 18, 2014
Published online: && &&, &&&&
ꢀ
Keywords: aldol reaction · amino acids · C C bond formation ·
enzyme catalysis · quaternary stereocenters
.
[1] a) A. Avenoza, C. Cativiela, F. Corzana, J. M. Peregrina, M. M.
Zurbano, Tetrahedron: Asymmetry 2000, 11, 2195 – 2204;
b) A. V. R. Rao, M. K. Gurjar, T. R. Devi, K. R. Kumar,
Tetrahedron Lett. 1993, 34, 1653 – 1656; c) C. Papa, C. Tomasini,
Eur. J. Org. Chem. 2000, 1569 – 1576; d) S. H. Pines, S. Karady,
M. Sletzinger (Merck and Co., Inc.), US3517057A, 1970, p. 7;
e) C. Cativiela, M. D. Dꢁaz-de-Villegas, Tetrahedron: Asymmetry
1998, 9, 3517 – 3599; f) A. Avenoza, J. H. Busto, F. Corzana, J. M.
Peregrina, D. Sucunza, M. M. Zurbano, Tetrahedron: Asymme-
try 2004, 15, 719 – 724; g) C. R. Strader, C. J. Pearce, N. H.
Oberlies, J. Nat. Prod. 2011, 74, 900 – 907; h) C. Cativiela, M. D.
Dꢁaz-de-Villegas, Tetrahedron: Asymmetry 2000, 11, 645 – 732;
i) H. Vogt, S. Brase, Org. Biomol. Chem. 2007, 5, 406 – 430; j) M.
Tanaka, Chem. Pharm. Bull. 2007, 55, 349 – 358; k) Y. Ohfune, T.
Shinada, Eur. J. Org. Chem. 2005, 5127 – 5143; l) S. H. Kang,
S. Y. Kang, H.-S. Lee, A. J. Buglass, Chem. Rev. 2005, 105, 4537 –
4558; m) T. Degenkolb, A. Berg, W. Gams, B. Schlegel, U. Grꢂfe,
J. Pept. Sci. 2003, 9, 666 – 678.
[11] N. Dꢃckers, K. Baer, S. Simon, H. Grçger, W. Hummel, Appl.
Microbiol. Biotechnol. 2010, 88, 409 – 424.
[12] A. Soler, X. Garrabou, K. Hernꢆndez, M. L. Gutiꢄrrez, E. Busto,
J. Bujons, T. Parella, J. Joglar, P. Clapꢄs, Adv. Synth. Catal. 2014,
356, 3007 – 3024.
[13] CCDC 1032540 contains the supplementary crystallographic
data for compound 3ah. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif. The Flack parameter is
x = ꢀ0.037(24), indicating the high enantiomeric discrimination
power of the analysis for S configuration.
[2] R. Grandel, U. Kazmaier, Eur. J. Org. Chem. 1998, 409 – 417.
[3] a) K. Fesko, L. Giger, D. Hilvert, Bioorg. Med. Chem. Lett. 2008,
18, 5987 – 5990; b) F. P. Seebeck, A. Guainazzi, C. Amoreira,
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
Communications
Biocatalysis
Quaternary stereocenters are obtained
with a stereoselective biocatalyst. Chirally
pure a,a-dialkyl-a-amino acid derivatives
were synthetized through aldol addition
of d-Ala and d-Ser to a variety of acceptor
aldehydes. This reaction is catalyzed by
the engineered pyridoxal-5’-phosphate
dependent biocatalyst SHMTSth Y55T.
(SHMTSth is the l-serine hydroxymethyl-
transferase from Streptococcus thermo-
philus).
K. Hernandez, I. Zelen, G. Petrillo,
I. Usꢁn, C. M. Wandtke, J. Bujons,
J. Joglar, T. Parella,
P. Clapꢂs*
&&& — &&&
Engineered l-Serine
Hydroxymethyltransferase from
Streptococcus thermophilus for the
Synthesis of a,a-Dialkyl-a-Amino Acids
6
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
These are not the final page numbers!