Peptide fragment coupling using a continuous-flow photochemical rearrangement of nitrones

Article abstract of DOI:10.1002/anie.201300504

Go with the flow: A general approach for amide bond formation by way of a continuous-flow photochemical rearrangement of nitrones was described (see scheme). Simple aryl-alkyl amide bonds as well as complex peptide bonds were constructed efficiently with a residence time less than 20minutes. A tetrapeptide was synthesized in this way and the method could be applied to peptide fragment coupling. Copyright

Full text of DOI:10.1002/anie.201300504

Angewandte
Chemie
DOI: 10.1002/anie.201300504
Synthetic Peptides
Peptide Fragment Coupling Using a Continuous-Flow Photochemical
Rearrangement of Nitrones**
Yuan Zhang, Melissa L. Blackman, Andrew B. Leduc, and Timothy F. Jamison*
Amide bonds are prevalent in nature as the key chemical
linkage of peptides and proteins and are also found in
pharmaceutically relevant compounds and important syn-
thetic polymers.[1] Conventional amide bond formation relies
on the condensation of carboxylic acids and amines and
generally uses stoichiometric amounts of activating agents
and other additives.[2] However, the use of activating agents in
peptide synthesis can compromise their utility because of
drawbacks such as epimerization, high cost, and waste
generation.[2b,3] In addition, despite the power of native
chemical ligation (NCL) strategies for fragment assembly of
polypeptides,[4] their reliance on the presence of N-terminal
cysteine residues limits their broader application and suggests
a need for alternative amide bond formation processes.[2a]
Toward this goal, we herein describe a continuous flow
approach of amide bond construction that proceeds by way of
photochemical rearrangement of a nitrone with stereochem-
ical fidelity and, notably, is thus well suited for peptide
fragment coupling (Scheme 1).
peptide bond formation using this approach has been
described.[8] This photochemical amide bond formation
process is attractive for several reasons, including: 1) a lack
of stoichiometric amounts of activating agents or additives for
nitrone formation from hydroxylamines and aldehydes,[9]
2) the ability to effect peptide ligation at a range of amino
acid residues, and 3) bypassing isolation of the often unstable
oxaziridine intermediates.[6] In addition, by using a straight-
forward, easily assembled, continuous-flow reactor system,[10]
the efficiency of this photochemical process is greatly
enhanced.
Continuous-flow photochemical processes have well es-
tablished advantages over conventional batch transforma-
tions in reaction efficiency, yield, reproducibility, and material
throughput.[11,12] In batch reactions, light penetration is
limited to a narrow layer within the reaction mixture.
Whereas in continuous-flow setups, the narrower reaction
channels and increased surface-to-volume ratio greatly
enhance light absorption, even at high substrate concentra-
tions. High material throughput can be realized simply by
flowing the reaction mixture for a longer period of time.
Furthermore, because the irradiation period can be precisely
controlled by the flow rate, over-irradiation-related side
reactions and decomposition pathways are often minimized
or avoided altogether.
The photochemical continuous-flow reactor system that
we constructed is depicted in Figure 1. Quartz tubing was
placed around the water-cooled quartz immersion well of
Scheme 1. Photochemical amide synthesis.
Photochemical rearrangements of nitrone to oxaziri-
dine,[5] and oxaziridine to amide[6] have been individually
documented starting in the 1950s. However, there are only
sporadic reports of a one-pot photochemical rearrangement
of nitrones to amides,[7] and to the best of our knowledge, no
[*] Dr. Y. Zhang, Dr. M. L. Blackman, Dr. A. B. Leduc,
Prof. Dr. T. F. Jamison
Figure 1. Continuous flow photochemical reactor.
Department of Chemistry, Massachusetts Institute of Technology
Cambridge, MA 02139 (USA)
E-mail: tfj@mit.edu
[**] We are grateful to the Novartis-MIT Center for Continuous
Manufacturing for financial support. We thank several colleagues at
MIT (Dr. Andrew S. Kleinke, Dr. James J. Mousseau, Dr. Ping Zhang,
and Kurt Armbrust) and at Novartis (Dr. Guido Koch, Dr. Berthold
Schenkel, Dr. Gerhard Penn, Dr. Benjamin Martin, and Dr. Jçrg
Sedelmeier) for insightful discussions. We also thank Eric Standley
(MIT) and Li Li (MIT) for obtaining mass spectrometric data.
a 450 W medium-pressure mercury lamp. The reaction
solution was introduced into the tubing using a syringe
pump and collected into a flask after passing through a 20 psi
back-pressure regulator. The complete UV setup was oper-
ated safely within an aluminium foil-lined photobox that was
easily accommodated by a standard laboratory fume hood.[13]
We first tested the photochemical rearrangement of
simple alkyl-aryl nitrone 1a in the continuous-flow system.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201300504.
Angew. Chem. Int. Ed. 2013, 52, 4251 –4255
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4251

.
Angewandte
Communications
Table 2: Substrate scope for the rearrangement of alkyl-aryl nitrones.
Nitrone 1a was readily prepared from methyl hydroxylamine
hydrochloride and p-tolualdehyde.[14] MeCN was found to be
optimal among the solvents evaluated (tetrahydrofuran
(THF), MeCN, dimethylformamide (DMF), isopropanol
(iPrOH), benzene) and was used as the reaction solvent for
further study.[15]
Optimization of the reaction conditions is summarized in
Table 1. We initially chose 0.02m as the substrate concen-
tration and used a cooling water bath to maintain the reaction
temperature at 408C. With a residence time (tR) of 30 minutes,
Entry[a]
Nitrone
R1
R2
tR
[min]
TFA
[equiv]
Yield
[%]
1
2
3
4
1b
1c
1d
1e
1 f
1g
1h
1i
Me
Me
Me
Me
Me
Me
Me
Me
Bn
p-OMe
m-OMe
3,4-(OCH2O)
o-Me
2,4,6-Me
p-Br
p-F
p-CF3
p-OMe
p-Me
10
15
5
15
10
20
15
10
10
10
–
–
83
70
74
60
65
90
84
81
87
89
0.1
0.25
0.25
0.25
0.25
0.1
–
5
Table 1: Photochemical rearrangement of nitrone 1a to amide 3a.
6[b]
7
8
9
10
1j
1k
Bn
0.25
Entry[a]
Concentration
[m]
tR
[min]
TFA
[equiv]
T
[8C]
Ratio
[a] All experiments were conducted using quartz tubing (0.460 mL or
0.625 mL, i.d.=0.762 mm). See the Supporting Information for details.
[b] A Pyrex sleeve (280 nm cutoff) was placed around the UV lamp to
prevent debromination.
(1a:2a:3a)[b]
1
2
3
4
0.02
0.02
0.05
0.05
0.05
0.05
0.1
30
30
30
10
10
10
30
–
40
40
40
40
92
92
92
nd:2:3
0.1
0.1
0.1
0.1
0.25
0.25
nd:nd:1
nd:1:26
1:11:21
nd:1:19
nd:nd:1
1:4:21
5
6[c]
7
[a] All experiments were conducted using 0.625 mL quartz tubing with an
inner diameter (i.d.) of 0.762 mm. [b] All ratios were determined by
1H NMR studies. nd=not detected. [c] the yield of isolated 3a was 93%.
Scheme 2. Synthesis of nitrone 6. Cbz=benzyloxycarbonyl.
all the starting nitrone 1a was consumed, but the oxaziridine
intermediate 2a was not fully converted into the amide
product 3a (Table 1, entry 1). The addition of a catalytic
amount of trifluoroacetic acid (TFA, 0.1 equiv) facilitated
conversion of oxaziridine into amide,[16] and afforded nearly
exclusive amide formation with a 30 minute tR, even at 0.05m
substrate concentration (Table 1, entry 3). Removal of the
cooling water bath resulted in a rise of the reaction temper-
ature to 928C and further expedited amide formation
(entry 5). Finally, increasing the amount of TFA to 0.25 equiv-
alents drove the reaction to completion, with a 10 minute tR,
and gave 3a in 93% yield of the isolated product (entry 6).
Further increasing the substrate concentration led to dimin-
ished conversion (entry 7). Thus, 0.05m was used as the
reaction concentration for further studies.
With these conditions in hand, a series of alkyl–aryl
nitrone substrates were investigated (Table 2). In general,
formation of amides from nitrones with either electron-rich or
electron-poor aryl group proceeded smoothly with good to
excellent yields of the isolated product. Substrates with
electron-rich aryl groups tended to require a shorter tR and
in some cases did not require TFA to accelerate the
subsequent rearrangement (Table 2, entries 1–3, 9).
When a 0.05m solution of nitrone 6 in MeCN was
subjected to the continuous flow photochemical conditions,
the rearrangement was facile and neither TFA nor heating
was required to promote the reaction (Table 3). Entries 1–4
demonstrate the reaction outcome with different residence
times. The oxaziridine intermediate from nitrone 6 existed as
a 3:2 mixture of two diastereomers 7a and 7b. The
stereochemistry of the oxaziridine rings in 7a and 7b was
tentatively assigned as shown. A ten minute tR was sufficient
for full conversion of both the starting nitrone and the
oxaziridine intermediates, and dipeptide 8 was isolated in
59% yield (Table 3, entry 4). In comparison, the correspond-
ing batch reaction with a ten minute reaction time only gave
a mixture of oxaziridines and amide in 7:1 ratio (entry 5).
Concentration is paramount in photochemical transfor-
mations, according to the Beer–Lambert law, and a clean
reaction is rarely obtained at a preparatively useful concen-
tration such as 0.1m in batch processes.[18] We were thus
pleased to find that increasing the initial substrate concen-
tration to 0.1m required only a marginally longer residence
time (12.5 minutes) for complete conversion, and more
importantly, no detrimental effect on the yield of isolated 8
was observed (entry 7). It is important to note that no
epimerization on either stereogenic center in dipeptide 8 was
Encouraged by the feasibility of the photochemical
rearrangement of alkyl-aryl nitrones to form amides, we
next investigated peptide bond formation employing this
transformation. Nitrone 6 was readily synthesized from Ala-
1
observed by H NMR spectroscopy or HPLC analysis.[13]
Interestingly, in the rearrangement of nitrone 6, another
amide (11) was isolated in 20% yield along with the desired
dipeptide 8. The mechanism of formation of 11 is proposed in
Scheme 3.[8b,19] Upon irradiation, homolytic cleavage of the
derived hydroxylamine
4 and Val-derived aldehyde 5
(Scheme 2).[14,17]
4252
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 4251 –4255

Angewandte
Chemie
Table 3: Photochemical rearrangement of nitrone 6.
Scheme 4. Individual photochemical rearrangement of 7a and 7b.
Entry[a]
tR
[min]
Concentration
[m]
Ratio
(6:7:8)[b]
1
2
3
4
1
2.5
5
10
–
0.05
0.05
0.05
0.05
0.05
0.1
1:9[c]:1
nd:4:3
a mixture of 8 and 11, however in different ratios (Scheme 4).
This information, together with a previous theoretical study
that showed the substituent anti to the nitrogen lone pair
migrates with no significant activation barrier,[20] prompts us
to propose that nitrogen inversion in the oxaziridine ring
nd:1:5
nd:nd:1[d]
nd:7:1
5[e]
6
10
12.5
nd:1:6
nd:nd:1[d]
À
competes with N O bond cleavage in this process. The
7
0.1
difference between the selectivity of 7a and 7b could be
explained by disparate rates of nitrogen inversion owing to
different spatial environments. Further investigation of the
rearrangement process and efforts to increase the selectivity
are currently ongoing.
[a] All entries were conducted in MeCN at 408C using a quartz tubing
(0.625 mL, i.d.=0.762 mm). [b] All ratios were determined by 1H NMR
studies. nd: not detected. [c] 7a:7b=3:2. [d] Yield of isolated 8 is 59%.
[e] Batch reaction was conducted in a quartz NMR tube (i.d.=5 mm)
placed next to the quartz tubing for ten minutes.
Next, a broader range of dipeptide formation was inves-
tigated (Table 4). All nitrone starting materials were readily
synthesized from the corresponding hydroxylamines and
aldehydes.[14,17] Although we had demonstrated that the initial
substrate concentration could be as high as 0.1m (Table 3,
Table 4: Photochemical formation of dipeptides from nitrones.
Entry[a] Substrate
Product[b]
1
12
13 (54%)
15 (60%)
2
Scheme 3. Proposed mechanism for the formation of 11.
14
3
À
weak N O bond in the oxaziridine intermediate generates
16
17 (56%)
N,O-diradical 9. H-migration leads to the desired amide (8),
while migration of the alkyl group gives aminal 10, which
further degrades to amide 11. The ratio of 8 and 11 remains
constant throughout the course of the reaction. It is also worth
noting that similar by-products were not observed in the
photochemical rearrangement of alkyl-aryl nitrones 1a–k.
The formation of 11 is somewhat surprising, because both
previously reported theoretical studies[20] and experimental
data[21] suggest that only the substituent anti to the nitrogen
lone pair migrates during this process. Because nitrone 6
exists exclusively as its Z-stereoisomer,[22] we expect both
oxaziridines 7a and 7b to be trans-substituted. In both cases,
therefore, an H atom should be at the anti position and
migrate selectively. To understand the stereochemical course
of the rearrangement better, 7a and 7b were prepared from
epoxidation of the corresponding imine and separated
chromatographically.[8b,23] When 7a and 7b were subjected
individually to the photochemical conditions, each provided
4
18
19 (50%)
21 (50%)
5[c]
20
6[d]
22
23 (37%)
[a] All experiments were conducted at 0.05m in MeCN at 408C with
tR =10 min using quartz tubing (0.460 mL or 0.625 mL, i.d. =0.762 mm)
unless otherwise noted. See the Supporting Information for details.
[b] Yields of the isolated product are shown in parenthesis. [c] tR =5 min.
[d] Experiment was conducted in benzene with tR =20 min.
Angew. Chem. Int. Ed. 2013, 52, 4251 –4255
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4253

.
Angewandte
Communications
entry 7), the solubility of some of the substrates in MeCN was
limited to 0.05m. All reactions proceeded smoothly under
continuous flow photochemical conditions to give the corre-
sponding dipeptide in reasonable to good yields. The rear-
rangement reaction tolerated a variety of functional groups.
Nitrones derived from protected amino acids with hydro-
phobic side chains (Val, 6), polar uncharged side chains (Ser,
12), aromatic rings (Tyr, 18), charged side chains (Lys, 16;
Glu, 20; and Arg, 22), and proline (14) were all amenable
substrates for this transformation. Amides of the type shown
for 11 were also observed in all entries, with yields of the
isolated product ranging from 10–19%.[13]
Scheme 7. Photochemical bi-directional amide bond formation.
S-protected cysteine-derived nitrones (S-Bn, S-tBu, S-
trityl) are not suitable substrates for this transformation
because of the decomposition of the oxaziridine intermedi-
ates.[24] Despite this limitation, the substrate scope of this
photochemical process represents a high degree of comple-
mentarity to NCL and allows peptide fragment coupling of
a much broader range of amino acid residues.
To expand the scope of this photochemical peptide bond
formation process further, sterically hindered nitrones 24 and
26 were prepared.[14,25] Upon irradiation, dipeptides 25 and 27
were obtained without compromised yields, albeit longer
residence times were required (Scheme 5).
34 in 49% yield (Scheme 7).[26] This result suggests that this
transformation might find applications in protein engineering
to create multifunctional chimeric proteins.[27]
In conclusion, we have demonstrated a general and
versatile approach for amide bond formation using a contin-
uous-flow photochemical rearrangement of nitrones. Not only
can this method be used to synthesize simple amides, but also
to generate more complex peptide bonds. Importantly, this
transformation represents a novel approach for peptide
fragment coupling and expands the current scope of protein
synthesis. In addition, by adopting a continuous-flow setup,
the efficiency of the photochemical process can be signifi-
cantly enhanced. Further investigations of its applications in
cyclic peptide synthesis and protein ligation are currently
ongoing.
Received: January 20, 2013
Published online: March 4, 2013
Keywords: amides · continuous flow · nitrones · peptides ·
.
photochemistry
Scheme 5. Sterically hindered peptide bond formation.
We next investigated the formation of a longer peptide
(Scheme 6). Nitrone 30 was prepared from Val/Gly-derived
hydroxylamine 28 and Phe/Ala-derived aldehyde 29.[14,17]
When a solution of 30 (0.05m in MeCN) was subjected to
the described photochemical conditions (tR = 10 min), tetra-
peptide 31 was isolated in 52% yield.
[1] The Amide Linkage: Structural Significance in Chemistry,
Biochemistry and Materials Science (Eds.: A. Greenberg, C. M.
Breneman, J. F. Liebman), Wiley, Hoboken, NJ, 2002.
[2] For recent reviews on amide bond synthesis, see: a) V. R.
Pattabiraman, J. W. Bode, Nature 2011, 480, 471 – 479; b) E.
Valeur, M. Bradley, Chem. Soc. Rev. 2009, 38, 606 – 631.
[3] R. B. Merrifield, J. Am. Chem. Soc. 1963, 85, 2149 – 2154.
[4] P. E. Dawson, T. W. Muir, I. Clarklewis, S. B. H. Kent, Science
1994, 266, 776 – 779.
The double photochemical rearrangement of bis-nitrone
33 also proceeded smoothly, giving the symmetrical diamide
[5] J. S. Splitter, M. Calvin, J. Org. Chem. 1958, 23, 651 – 651.
[6] J. S. Splitter, M. Calvin, J. Org. Chem. 1965, 30, 3427 – 3436.
[7] Representative examples of nitrone to amide rearrangement:
a) H. Kondo, K. Tanaka, K. Kubo, T. Igarashi, T. Sakurai,
Heterocycles 2004, 63, 241 – 247; b) Y. Zeng, B. T. Smith, J.
Hershberger, J. Aubꢀ, J. Org. Chem. 2003, 68, 8065 – 8067; c) E.
Bourguet, J.-L. Baneres, J.-P. Girard, J. Parello, J.-P. Vidal, X.
Lusinchi, J.-P. Declercq, Org. Lett. 2001, 3, 3067 – 3070.
[8] To the best of our knowledge, there are only two examples of
related peptide formation from oxaziridines. See: a) P. Duhamel,
B. Goument, J.-C. Plaquevent, Tetrahedron Lett. 1987, 28, 2595 –
2596; b) S. Wenglowsky, L. S. Hegedus, J. Am. Chem. Soc. 1998,
120, 12468 – 12473.
[9] J. W. Bode, Top. Organomet. Chem. 2013, 44, 13 – 34.
[10] For recent reviews on continuous-flow experiments, see: a) D.
Webb, T. F. Jamison, Chem. Sci. 2010, 1, 675 – 680; b) S. V. Ley,
Scheme 6. Formation of the tetrapeptide 31 from nitrone 30.
4254
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 4251 –4255

Angewandte
Chemie
I. R. Baxendale, CHIMIA 2008, 62, 162 – 168; c) K. Geyer, T.
Gustafsson, P. H. Seeberger, Synlett 2009, 2382 – 2391; d) R. L.
Hartman, J. P. McMullen, K. F. Jensen, Angew. Chem. 2011, 123,
7642 – 7661; Angew. Chem. Int. Ed. 2011, 50, 7502 – 7519.
[11] For recent reviews on photochemistry in continuous flow, see:
a) M. Oelgemçller, O. Shvydkiv, Molecules 2011, 16, 7522 – 7550;
b) J. P. Knowles, L. D. Elliott, K. I. Booker-Milburn, Beilstein J.
Org. Chem. 2012, 8, 2025 – 2052.
[12] For contributions to this field from our laboratories, see: a) A. C.
Gutierrez, T. F. Jamison, J. Flow Chem. 2011, 1, 24 – 27; b) A. C.
Gutierrez, T. F. Jamison, Org. Lett. 2011, 13, 6414 – 6417; c) B.
Shen, M. W. Bedore, A. Sniady, T. F. Jamison, Chem. Commun.
2012, 48, 7444 – 7446; d) J. W. Tucker, Y. Zhang, T. F. Jamison,
C. R. J. Stephenson, Angew. Chem. 2012, 124, 4220 – 4223;
Angew. Chem. Int. Ed. 2012, 51, 4144 – 4147.
[19] W. D. Emmons, J. Am. Chem. Soc. 1957, 79, 5739 – 5754.
[20] E. Oliveros, M. Riviꢁre, J. P. Malrieu, C. Teichteil, J. Am. Chem.
Soc. 1979, 101, 318 – 322.
[21] A. Lattes, E. Oliveros, M. Riviꢁre, C. Belzecki, D. Mostowicz, W.
Abramskj, C. Piccinni-Leopardi, G. Germain, M. Van Meers-
sche, J. Am. Chem. Soc. 1982, 104, 3929 – 3934.
[22] A. G. Dias, C. E. V. Santos, F. Z. G. A. Cyrino, E. Bouskela,
P. R. R. Costa, Bioorg. Med. Chem. 2009, 17, 3995 – 3998.
[23] Please see the Supporting Information for the synthesis of 7a
and 7b. The stereochemistry was assigned in accordance with
Table 3.
[24] It is known that thioethers can react with oxaziridines, leading to
either N- or O-transferred products. See: A. Armstrong, I. D.
Edmonds, M. E. Swarbrick, Tetrahedron Lett. 2003, 44, 5335 –
5338.
[13] Please see the Supporting Information for details.
[14] S. I. Medina, J. Wu, J. W. Bode, Org. Biomol. Chem. 2010, 8,
3405 – 3417.
[15] Benzene is also an amenable solvent, although the reaction rate
is slower than with MeCN.
[16] D. Xing, X. Xu, L. Yang, Synthesis 2009, 3399 – 3404.
[17] A. Moulin, J. Martinez, J.-A. Fehrentz, J. Pept. Sci. 2007, 13, 1 –
15.
[18] “Photochemical Methods”: A. Albini, L. Germani in Handbook
of Synthetic Photochemistry (Eds.: A. Albini, M. Fagnoni),
Wiley-VCH, Weinheim, 2010.
[25] J. Calveras, M. Egido-Gabꢂs, L. Gꢃmez, J. Casas, T. Parella, J.
Joglar, J. Bujons, P. Clapꢀs, Chem. Eur. J. 2009, 15, 7310 – 7328.
[26] For recent examples of a bi-directional approach in organic
synthesis, see: a) Y. Zhang, J. D. Rainier, Org. Lett. 2009, 11,
237 – 239; b) V. B. Naidoo, M. Rautenbach, J. Pept. Sci. 2012, 18,
317 – 325.
[27] For recent reviews on protein linkers, see: a) P. E. Wright, H. J.
Dyson, Curr. Opin. Struct. Biol. 2009, 19, 31 – 38; b) W. Wriggers,
S. Chakravarty, P. A. Jennings, Biopolymers 2005, 80, 736 – 746.
Angew. Chem. Int. Ed. 2013, 52, 4251 –4255
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4255