In Situ Cell-Free Protein Synthesis
FULL PAPER
treatment of the corresponding ester with hydrazine hydrate.[46,47] Ester
10c was prepared by the treatment of (S)-leucine with 1-adamantanol, di-
methyl sulfite, and para-toluenesulfonic acid.[48]
19455 Da (Figure 6d) corresponds to substitution of the
chlorinated free amino acid 29 for all ten of the (2S,3S)-iso-
leucine residues found in native PpiB (Figure 6a), while the
next most abundant peak at 19431 Da results from nine re-
placements (each substitution that corresponds to replacing
a methyl group with chlorine increases the mass by an aver-
age of 20.5 Da, although the natural carbon and chlorine
isotope abundance in each of these chlorinated PpiBs results
in five dominant ions 1 Da apart, differing in intensity by
15% or less, so the difference indicated by the spectrometer
is (20Æ5) Da). With chloride 29, the mass spectrum shows
that the average degree of incorporation is above 90%. For
hydrazides 23 and 25, the S30 extract simply provides a way
to avoid having to carry out problematic deprotection as a
separate step; however, for chloride 28 the S30 extract also
continuously replenishes the free amino acid 29 as it is con-
sumed through lactonization. As a result, incorporation is
observed by using ester 28, but not with the same initial con-
centration of the free amino acid 29.
(S)-4-Fluoroleucine hydrazide (23) was prepared by the protection of
(S)-leucine, bromination of phthalimide 19, treatment of bromide 20 with
silver fluoride, and reaction of fluoride 21 with hydrazine hydrate.[20] (S)-
4,5-Dehydroleucine hydrazide (25) was prepared from N-phthaloyl-4,5-
dehydroleucine methyl ester (22)[49] by reaction with hydrazine hy-
drate.[20] M.p. 130–1328C; 1H NMR (300 MHz, D2O): d=5.06 (m, 1H),
4.93 (m, 1H), 4.15 (dd, J=9, 6 Hz, 1H), 2.62 (dd, J=14, 6 Hz, 1H), 2.54
(dd, J=14, 9 Hz, 1H), 1.78 ppm (s, 3H); 13C NMR (100 MHz, D2O): d=
168.9, 138.5, 116.9, 50.8, 39.8, 21.0 ppm; HRMS (ESI, +ve) m/z: calcd for
C6H14N3O: 144.1137; found: 144.1140 [M+H]+.
AHCTUNGERTG(NNUN 2S,3R)-4-Chlorovaline methyl ester (28) was isolated through HPLC
from the mixture obtained by chlorination of ester 3c.[43] 1H NMR
(400 MHz, CD3OD): d=4.27 (d, J=4 Hz, 1H), 3.87 (s, 3H), 3.74 (dd, J=
12, 8 Hz, 1H), 3.66 (dd, J=12, 6 Hz, 1H), 2.35–2.45 (m, 1H), 1.10 ppm
(d, J=8 Hz, 3H); 13C NMR (100 MHz, CD3OD): d=170.2, 55.7, 53.8,
47.0, 39.5, 13.5 ppm; HRMS (ESI, +ve) m/z: calcd for C6H13NO2Cl:
166.0635 and 168.0605; found: 166.0633 and 168.0610 [M+H]+; m/z:
calcd for C6H12NO2ClNa: 188.0454 and 190.0425; found: 188.0452 and
190.0428 [M+Na]+; further details are provided in the Supporting Infor-
mation.
Treatment of amino acid derivatives with S30 extract from E. coli BL21
Star (DE3): S30 extract from E. coli BL21 Star (DE3) was prepared as
previously reported.[22] Stock solutions of the amino acid derivatives 1a–f,
2, 3a–g, 4, 5a–c, 6a,b, 7a,b, 8, 9a–c, 10a–c, 11a–d, 12, 13a–c, 14, 15a,b,
16, 17a–c, and 18 were prepared in water, ethanol, or dimethyl sulfoxide
(DMSO) according to solubility. An aliquot (4 mL) of each stock solution
was diluted to a final concentration of 2 mm with 4-(2-hydroxyethyl)-1-pi-
perazineethanesulfonic acid (HEPES) buffer (116 mL, 50 mm, pH 7.5)
and the S30 extract (80 mL). The mixtures were incubated at 378C for 6 h
and centrifuged at 12000 rpm for 10 min. Each supernatant was passed
through an Amicon Ultra-4 (YM-10) centrifugal filter device, and the fil-
trates were analysed with HPLC by using the Waters AccQ.Tag method,
with reference to amino acid standard solutions and the background
amino acid concentration of the S30 extract. Accordingly, a sample of
each filtrate (20 mL) was treated with AccQ.Fluor borate buffer (80 mL)
and reconstituted AccQ.Fluor reagent (20 mL). The mixtures were ana-
lysed by using an AccQ.Tag column (C18, 4 mm, 150ꢁ3.9 mm), eluting
with a gradient of acetonitrile in AccQ.Tag eluent. Representative HPLC
traces are provided in the Supporting Information.
Conclusion
This study has demonstrated the ability of the S30 extract to
remove a range of protecting groups for direct incorporation
of the resulting amino acids into a protein. This approach is
more efficient because it not only decreases the number of
synthetic steps required in the preparation of unnatural
amino acids, but also provides a versatile method to circum-
vent problems associated with chemical instability of amino
acids during both their deprotection and protein synthesis.
The method has been demonstrated to be suitable for the
incorporation of the fluoro- and dehydroleucines 26 and 27
and chlorovaline 29 as substitutes for leucine and isoleucine,
respectively, at levels of 90% or above. These high levels
are more than adequate for applications such as isotopic la-
belling or the fluorination of proteins, for use in spectro-
scopic studies for example, in which the unmodified protein
is not detectable. These levels are also suitable to investigate
general rather than specific effects of amino acid modifica-
tions on protein structure and function. Further, the cell-
free system with the S30 extract allows for a complete site-
specific incorporation of an unnatural amino acid through
the addition of a mutant aminoacyl tRNA synthetase and
cognate suppressor tRNA, under which conditions the incor-
poration levels would be expected to be quantitative.
Cell-free protein synthesis: Plasmid DNA encoding for His6-PpiB with
expression under control of the phage T7 promoter (pND1098) was car-
ried out according to a previous report.[10] Plasmid DNA was prepared
from E. coli DH5a/pND1098 with the Qiagen Plasmid Maxi kit. T7
RNA polymerase (50000 U mLÀ1) was obtained from New England Bi-
oLabs Inc. (MA, USA). Cell-free protein synthesis was carried out by
using a reported procedure[8,10,21,22] with the following few modifications.
(S)-Alanine and RNasin were not added to the inner mixture (500 mL).
T7 RNA polymerase (2 mL) was added to each reaction mixture instead
of the plasmid encoding for this enzyme. An aliquot of tRNA solution of
5 mL was added instead of 10 mL. The final concentration of the solvent
(ethanol or DMSO) used to dissolve some of the amino acid derivatives
1a–f, 2, 3a–g, 4, 5a–c, 6a,b, 7a,b, 8, 9a–c, 10a–c, 11a–d, 12, 13a–c, 14,
15a,b, 16, 17a–c, and 18 was ꢀ2%, which was established through con-
trol experiments to have no effect on protein synthesis. The His6-PpiB se-
quence (with an additional C-terminal asparagine residue;[10] mass=
Experimental Section
19221 Da,
N-formyl-His6-PpiB=19250 Da)
is MHHHHHHMVT
FHTNHGDIVI KTFDDKAPET VKNFLDYCRE GFYNNTIFHR
VINGFMIQGG GFEPGMKQKA TKEPIKNEAN NGLKNTRGTL
AMARTQAPHS ATAQFFINVV DNDFLNFSGE SLQGWGYCVF
AEVVDGMDVV DKIKGVATGR SGMHQDVPKE DVIIESVTVS
EN. The in vitro cell-free reaction mixture containing expressed His6-
PpiB was centrifuged at 12000 rpm for 2 min and the supernatant was ap-
plied to a Ni-ion affinity column equilibrated with 20 mm sodium phos-
phate, 0.5m NaCl, and 20 mm imidazole at pH 7.5 and 48C. Bound pro-
teins were eluted by application of 20 mm sodium phosphate, 0.5m NaCl,
Amino acid derivatives: With the exception of hydrazides 2 and 9a and
ester 10c, all the amino acids illustrated in Figures 1–3 are available from
Sigma–Aldrich, Merck Pty. Ltd., Auspep, TCI Chemicals, or Aurora Fine
Chemicals LLC, although for the purposes of this investigation many
were prepared from the corresponding free amino acids (see the Support-
ing Information). Phenylhydrazide 2 was synthesized from methyl ester
3a by using phenylhydrazine.[45] Hydrazide 9a was prepared from (S)-
lysine by esterification with thionyl chloride in methanol, followed by
Chem. Eur. J. 2013, 19, 6824 – 6830
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6829