Ribosomal Synthesis of Backbone-Cyclic Peptides Compatible with in Vitro Display

Article abstract of DOI:10.1021/jacs.8b05327

Backbone-cyclic peptides are an attractive class for therapeutic development. However, in vitro display technologies coupled with ribosomal synthesis are intrinsically inapplicable to such "phenotypes" because of loss of the C-terminal peptide region linking to "genotype". Here, we report a methodology enabling the display of backbone-cyclic peptides. To achieve this, genetic code reprogramming was utilized to implement a rearrangement strategy involving the ribosomal incorporation of a designer initiator containing a thiazolidine-protected cysteine and 2-chloroacetoamide (ClAc) side chain, followed by an α-thio acid and cysteine at downstream positions. Upon expression of the linear peptide, spontaneous thioester rearrangement occurs between the α-thioester and the thiol group of the cysteine, liberating the α-thio group and resulting in cross-linking to the upstream ClAc side-chain group. Then selective deprotection of the thiazolidine-protected cysteine immediately promotes intramolecular native chemical ligation, as demonstrated for various sequences and ring sizes. In this approach, the backbone-cyclic peptides retain their C-terminal peptide regions via the side-chain thioether covalent linkage, making them compatible with in vitro display.

Full text of DOI:10.1021/jacs.8b05327

Subscriber access provided by Iowa State University | Library
Article
Ribosomal synthesis of backbone-cyclic
peptides compatible with in vitro display
Ryo Takatsuji, Koki Shinbara, Takayuki Katoh, Yuki Goto, Toby
Passioura, Ryo Yajima, Yamato Komatsu, and Hiroaki Suga
J. Am. Chem. Soc., Just Accepted Manuscript • Publication Date (Web): 16 Jan 2019
Downloaded from http://pubs.acs.org on January 16, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 34
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Ribosomal synthesis of backbone-cyclic peptides compatible
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
with in vitro display
Ryo Takatsuji, Koki Shinbara, Takayuki Katoh, Yuki Goto, Toby Passioura, Ryo Yajima,
Yamato Komatsu, Hiroaki Suga*
Department of Chemistry, Graduate School of Science, The University of Tokyo, 7-3-1 Hongo,
Bunkyo-ku, Tokyo 113-0033, Japan
*Correspondence to: hsuga@chem.s.u-tokyo.ac.jp
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 34
1
2
3
4
5
6
7
Abstract
8
9
Backbone-cyclic peptides are an attractive class for therapeutic development. However, in
vitro display technologies coupled with ribosomal synthesis are intrinsically inapplicable to such
“phenotypes” due to loss of the C-terminal peptide region linking to “genotype”. Here we report
a methodology enabling the display of backbone-cyclic peptides. To achieve this, genetic code
reprogramming was utilized to implement a rearrangement strategy involving the ribosomal
incorporation of a designer initiator containing a thiazolidine-protected cysteine and 2-
chloroacetoamide (ClAc) sidechain, followed by an -thio acid and cysteine at downstream
positions. Upon expression of the linear peptide, spontaneous thioester rearrangement occurs
between the -thioester and the thiol group of the cysteine, liberating the -thio group and
resulting in crosslinking to the upstream ClAc sidechain group. Then, selective deprotection of
the thiazolidine-protected cysteine immediately promotes intramolecular native chemical ligation,
as demonstrated for various sequences and ring sizes. In this approach, the backbone-cyclic
peptides retain their C-terminal peptide regions via the sidechain thioether covalent linkage,
making them compatible with in vitro display.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Introduction
In vitro display technologies, such as ribosome, DNA, or mRNA display enable the affinity-
based screening of vast numbers of peptides (or proteins) against targets of interest1-4.
Conventional methods are able to display peptides consisting of only the 20 proteinogenic amino
acid due to limitations imposed by the natural genetic code. This limits not only the usable
building blocks but also the scaffold architecture, such that only linear or cyclic forms closed by
2
ACS Paragon Plus Environment

Page 3 of 34
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
a disulfide bond(s) between Cys sidechains are generally accessible. However, recent advances
in genetic code reprogramming methodologies have enabled the expression of peptides with
diverse non-proteinogenic building blocks and macrocyclic scaffolds closed by non-reducible
bonds. For instance, the FIT (Flexible In vitro Translation) system5,6, in which genetic code
reprogramming is facilitated by flexible tRNA-acylation ribozymes (flexizymes)5,7, has made it
possible to express nonstandard peptides macrocyclized (via spontaneous thioether bond
formation between an N-terminal chloroacetamide group and a downstream Cys sidechain) and
containing D- and/or N-methyl-L-amino acids8-12. Integration of the FIT system with mRNA
display, referred to as RaPID (Random nonstandard Peptide Integrated Discovery)(Figure 1a),
allows the in vitro selection of thioether-macrocyclic peptide ligands against therapeutic protein
targets from libraries in excess of 1012 unique compounds, leading to the discovery of novel and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
potent drug candidates13-18
.
In any of the aforementioned display technologies, the “phenotype” imparted by the peptide
must be linked to its “genotype” via the peptide C-terminal region. In the case of mRNA display
(including RaPID), the C-termini of individual peptides are fused to puromycin attached to the
3'-end of cognate mRNAs3,4. This linkage allows for iterative rounds of affinity selection against
a target protein and amplification of active peptide species from the library, and is therefore
indispensable. This means that current methods are able to display macrocyclic peptides
consisting of only N-terminus-to-sidechain (Figure 1a) or sidechain-to-sidechain linkages, but not
N-terminus-to-C-terminus (Figure 1b) linkages. To wit, backbone-cyclized peptide libraries
cannot be screened using current in vitro display methodologies. However, some naturally
occurring macrocyclic peptides, such as cyclosporine A, exhibit N-terminus-to-C-terminus
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 34
1
2
3
4
5
6
7
8
9
linkage19. This motivated us to devise a new methodology overcoming this structural limitation,
and through which backbone-cyclized peptides could be displayed on mRNA. Here we report a
method for ribosomal synthesis of backbone-cyclized peptides without loss of the downstream
peptide region, achieving this end (Figure 1c for example).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Results and Discussion
-Thio acid incorporation into peptide backbones
To establish the method, we envisioned two multi-reaction steps: (S1) a step that establishes
a linkage of the downstream peptide region with a sidechain on the region to be backbone-
cyclized, and (S2) intramolecular native chemical ligation (NCL)20 between the N-terminal amino
group and a downstream thioester that is directly introduced through ribosomal translation,
leading to the formation of a backbone-cyclized scaffold (Figure 1c).
To introduce a thioester bond into the expressed peptide, we planned to reprogram a codon
to assign an α-thio acid. It is well known that some α-hydroxy acids are effectively incorporated
into nascent peptide chains with efficiency comparable to the corresponding α-amino acids when
charged onto appropriate tRNAs21-23. Despite the fact that α -thio acids are analogous to α -
hydroxy acids, it has been reported that ribosomal thioester bond formation using α-thio acids is
very poor24. Therefore, we first investigated the incorporation of α-thio acids into nascent peptide
chains by means of FIT (Figure S1).
We chose three α-thio acids, HSFCl, HSV, and HSG (Figure 2a), which are α-thio analogs of p-
chlorophenylalanine, valine, and glycine, respectively, to investigate ribosome-mediated thioester
bond formation. The respective α-thio acids esterified with cyanomethyl (CME for HSFCl) or 3,5-
4
ACS Paragon Plus Environment

Page 5 of 34
Journal of the American Chemical Society
1
2
3
4
5
6
7
dinitrobenzyl (DNB for HSV and HSG) groups were charged onto tRNAAsnE2
(for simplicity,
CUA
8
9
described as tRNACUA below) using appropriate flexizymes (eFx for CME and dFx for DNB)5,6.
Although HSFCl and HSG were charged onto tRNACUA in satisfactory yields, HSV was not charged
well, presumably due to steric hindrance of the sidechain and/or poor solubility in the reaction
buffer (Figure 2b). We thus chose HSFCl-tRNACUA and HSG-tRNACUA for testing ribosome-
mediated thioester bond formation in a FIT reaction lacking release factor 1 (RF1) and in which
the UAG codon was vacant. In the translation of an mRNA (mr1-X5), the UAG codon was
decoded by HSFCl-tRNACUA and HSG-tRNACUA, yielding the model peptides (p1-X5), p1-SFCl5 and
p1-SG5, respectively (Figure 2c). MALDI-TOF-MS analysis of the respective products showed
the expected mass values (Figure 2d and Figure S2b), and MS/MS analysis confirmed their
sequence identity (Supplementary Figure S2a and S2c). To the best of our knowledge, this is the
first demonstration that the ribosome is able to efficiently incorporate an -thio acid into a peptide
backbone. To assess the expression level of each thioester-containing peptide, we performed
tricine-SDS-PAGE analysis relying on the quantitative autoradiographic detection of [14C]-Asp
in the C-terminal peptide region. The concentration of p1-SFCl5 was determined to be 1.78 µM,
which was comparable to that of a control peptide p1-F5 containing phenylalanine (2.03 µM) and
high enough for further applications, whereas the expression of p1-SG5 was a magnitude lower
(Figure 2e). Thus, the following experiments were performed using HSFCl.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Strategy for sidechain-sidechain crosslinking prior to native chemical ligation
Direct thioester formation using HSFCl allowed us to examine NCL between the amino group
of an N-terminal Cys residue and the thioester, (step S2). We previously reported a method to
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 34
1
2
3
4
5
6
7
8
9
generate an N-terminal Cys with a free -amine using two recombinant enzymes, peptide
deformylase and methionine aminopeptidase, allowing backbone cyclization via intramolecular
NCL25. Unfortunately, this enzymatic method would not allow control of the “timing” of the
thioester exchange reaction such that S1 preceded S2. Therefore, we devised an alternative
strategy, which was novel with respect to translated peptides, to chemically generate an N-
terminal amino group on the N-terminal Cys1 residue. We considered that (R)-thiazolidine-4-
carboxylic acid (Thz1), which has been sometimes used in ribosomal translation as a proline
analogue26,27, could be used as a precursor of N-terminal N-methyl-cysteine (MeNCys1). Although
Thz1 could be deprotected by hydroxylamine or methoxyamine28, we were concerned that such
reagents would react with nucleobases to cause mutagenesis29,30, thus making them unsuitable for
application to in vitro display. Therefore, we chose the NaBH3CN reduction method to deprotect
Thz1, which is unlikely to interfere with nucleobases31. By means of MALDI-TOF-MS, we
indeed confirmed that Thz1-Cab2 used as an initiator reassigned by Thz1-Cab2-tRNAfMetCAU was
ribosomally incorporated and then cleanly converted to MeNCys1 in a model peptide by the
treatment with NaBH3CN (vide infra).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
At this point, it was necessary to develop a strategy for achieving the S1 step, and we naively
considered the following scheme: when (S)-2-amino-4-(2-chloroacetamido)butanoic acid (Cab)
was incorporated adjacent to Thz1, and when a Cys12 was incorporated at a downstream position
of HSFCl9, the chloroacetamido group on Cab2 would spontaneously react with the thiol group on
Cys12 to form a thioether linkage (R1 in Figure 3). This large thioether-macrocyclic peptide could
then be treated with NaBH3CN to liberate N-terminal MeNCys1. Subsequently, the sidechain thiol
group would attack the thioester bond to undergo NCL (S-S acyl migration followed by S-N acyl
6
ACS Paragon Plus Environment

Page 7 of 34
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
migration to the N-methyl-amino group of MeNCys1), forming a macrocyclic scaffold (R2 in
Figure 3). During NCL, the thioester in the main backbone originally produced in translation
would be cleaved, but the thioether linkage formed between the Cab2 and Cys12 residues prior
to NCL would prevent division into two fragments, and the resulting backbone-cyclized peptide
would be compatible with mRNA display (Figure 3b).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To test the above strategy, we designed an mRNA template (mr2-W8) expressing a model
peptide, p2-W8, in a FIT reaction lacking Met and RF1, and in which the AUG and UAG codons
were reassigned to Thz1-Cab2 and HSFCl9, respectively (Figure 3a). Following the 30 min
translation reaction, 33 µg/mL chloramphenicol was added to the translation, aiming at slowing
down the translation to avoid the formation of pre-matured peptide side-products. Under these
conditions, expression of p2-W8 produced a peptide with a mass value consistent with loss of
hydrogen chloride (Figure S3a), suggesting that the expected thioether-macrocyclic peptide, tcp2-
W8-SFCl9 (tcp denotes thioether-cyclic peptide), was produced (Figure 3b). Since macrocyclic
peptides are not generally susceptible to MS/MS cleavage, the thioester bond in tcp2-W8-SFCl9
was opened by addition of an excess amount of free Cys at pH 10, producing a linear peptide with
a Cys-adduct, p2-W8-Cys (Figure 4a), as confirmed by MS (Figure 4b). MS/MS analysis of the
resulting acyclic peptide suggested that the thioether bond did not form between Cab2 and the
downstream Cys12 sidechain as planned (Figure 4c). Rather, the data appeared consistent with
thioether bond formation between Cab2 and the -thio group of the HSFCl9 residue (Figure 4d).
This observation made us revise our original strategy (Figure 3b) to an alternative (Figure 5b).
Based on the observed MS/MS data indicating formation of the Cab2-SFCl9 thioether linkage,
the first event occurring after ribosomal synthesis of p2-W8 was likely to be thioester exchange
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 34
1
2
3
4
5
6
7
8
9
between HSFCl9 and HSCys12 (the subscript HS denotes the sidechain thiol), establishing the Trp8-
SCys12 thioester linkage between the backbone carbonyl group and the thiol sidechain group of
HSCys12 (Rtex in Figure 5, forming p2-W8-SC12). This event might occur rapidly prior to
thioether bond formation, presumably due to fast nucleophilic attack of the primary thiol of
HSCys12 at the secondary thioester in closer proximity than the Cab2 chloroacetamido sidechain
near the N-terminus. This trans-thioesterification would liberate the secondary -thiol of HSFCl9,
inducing subsequent thioether bond formation to generate a Cab2-SFCl9 thioether linkage, yielding
tcp2-W8-SC12 (Figure 5b). Although the MS/MS data clearly supports this event as the major
pathway, it should be noted that even if the minor (originally proposed) pathway occurs in parallel
the final backbone cyclic moiety would be the same in both cases, as discussed below.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We next conducted the NaBH3CN reduction of Thz1 in tcp2-W8-SC12. We observed an
approximately 2 Da increase in mass value, indicating generation of the N-terminal MeNCys1
(deprotected tcp2-W8-SC12) (Figure S3a-b). The thiol group of MeNCys1 could spontaneously
undergo NCL with the Trp8-SCys12 thioester to generate an N-methyl-peptide bond yielding
bcp2-W8 (bcp denotes backbone-cyclic peptide), but the mass value would not change before and
after the NCL in this case (Figure 5b). Therefore, we confirmed the product in two ways. First,
the product after NaBH3CN reduction was treated with 2-iodoacetamide (IAA), where if the
desired NCL did not occur, we expected to see a single AA-adduct on the peptide at MeNCys1;
whereas if it occurred and gave backbone cyclized bcp2-W8, a double AA-adduct to the thiol
sidechains of both MeNCys1 and Cys12 should be observed (Figure 5c). In fact, we observed a sole
peak corresponding to the double AA-adduct (bcp2-W8 + 2AA) (Figure S3c), indicative of
macrocyclization. However, since we did not consider this IAA modification of bcp2-W8
8
ACS Paragon Plus Environment

Page 9 of 34
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
conclusive, we applied a second method in which MS/MS analysis was performed on the final
product (Figure 5d). The MS/MS spectrum of bcp2-W8 showed a series of fragmentations in the
C-terminal liner peptide region (b9’, b10’, b11’, b12’, and b13’) indicating thioether formation on
HSFCl9, as well as internal peptide fragments including the Trp8-MeNCys1 amide bond (i[bA-yB],
i[bA-yC], i[bA-yE], and i[bB-yC]), directly indicating the expected backbone cyclization. Further,
we did not observe strong peaks consistent with fragmentation of the macrocyclic peptide derived
from the originally proposed pathway. We thus concluded that the revised pathway was the major
process occurring in backbone cyclization of bcp2-W8 closed by the Trp8-MeNCys1 backbone
amide bond (Figure 5b).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The observed NCL efficiency might depend on the reactivity of the thioester, dictated by the
choice of amino acid residue at the adjacent upstream position32. To verify if other residues could
affect ligation efficiency, we substituted Trp8 in p2-W8 with Gly8, Lys8, or Val8 (p2-G8, p2-K8,
and p2-V8). MALDI-TOF-MS analysis of the respective peptides showed clean expression of the
full length of peptide via spontaneous thioether bond formation (Figure S4a, S4d and S4g), and
the subsequent deprotection of Thz1 to MeNCys1 also yielded products with the expected MS
values (Figure S4b, S4e and 4h). The IAA treatment of bcp2-G8 and bcp2-K8 demonstrated MS
values consistent with their double AA-adduct (Figure S4c and S4f), indicating the completion of
backbone macrocyclization. However, IAA treatment of bcp2-V8 showed two peaks, consistent
with formation of both the double and single AA-adducts, suggesting that backbone
macrocyclization was not complete (Figure S4i). We speculated that because of the steric
hindrance of Val8, the intramolecular NCL in tcp2-V8 was slower than tcp2-W8, tcp2-G8, and
tcp2-K8 under acidic conditions with NaBH3CN. In order to accelerate the cyclization, we added
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 34
1
2
3
4
5
6
7
8
9
an additional step to neutralize the reaction buffer after NaBH3CN reduction. This resulted in
observation of a single peak consistent with formation of the double AA-adduct of bcp2-V8
(Figure S4j), indicating that the complete backbone cyclization for bcp2-V8 was also achieved.
Next, we wondered if the thioester exchange between HSFCl9 and Cys12 followed by
backbone-cyclization was dependent on the linking sequence, Ile10-Gly11. We also wondered if
the length of this linking sequence was of importance. We thus designed two sequences, in one
of which the Ile10-Gly11 sequence was replaced with Ile10-Pro11 (p3-W8) and a second in which
a Ser residue (Ser11) was inserted to give Gly11-Ser12-Cys13 (p4-W8) (Figure 6a). Both peptides
were ribosomally synthesized and cyclized by the procedure outlined above. Mass spectra of the
product gave peaks corresponding to the doubly AA-modified bcp3-W8 and bcp4-W8, indicating
generation of the desired backbone-cyclized peptides closed by the Trp8-MeNCys1 bond (Figure
6b and 6c). This suggests that the linking sequence composition and length between HSFCl and the
downstream Cys residue are not critical.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We further confirmed that this backbone-cyclization methodology could be applied to other
sequences. We designed two linear peptide sequences, p5-L11 and p6-Q14, forming 11- and 14-
residue rings closed by L11-MeNCys1 and Q14-MeNCys1 backbone amide bonds, respectively
(bcp5-L11 and bcp5-Q14, Figure 6d). MS analysis demonstrated that treatment of the expressed
p5-L11 and p6-Q14 peptides with NaBH3CN, neutralization, and then IAA coupling yielded the
respective backbone cyclic peptides with double AA-modification (Figure 6e and 6f). Thus, since
the macrocyclization process proceeded smoothly in all constructs irrespective of sequence (three
distinct sequences were tested) or ring size (8, 11, and 14 residue-rings were tested) even at
10
ACS Paragon Plus Environment

Page 11 of 34
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
distinct ligation residues (Trp, Lys, Val, Leu, and Gln were tested), this strategy of backbone
macrocyclization appears very likely to be applicable to a wide range of sequences and ring sizes.
To examine if this backbone-macrocyclization methodology could be compatible with the
mRNA display, two experiments were performed. First, we designed a mRNA template mr7,
expressing p7-L10 (Figure 7a), with UAG stop codon where ribosome would stall in the absence
of RF1. After the translation of p7-L10, 1.2 µM puromycin (Pu) was added to trap the C-terminus
of p7-L10-tRNA instead of chloramphenicol. After the chemical transformation using the
protocol established in this report, the product was analyzed by LC/ESI-MS. We observed a single
major peak in the LC diagram (Figure 7b), corresponding to an ionized mass (m/z) of 1168.47
(Figure 7c). This mass value was consistent with the calculated mass of [M–2H]2- (m/z = 1168.42),
indicating that bcp7-L10-Pu was formed. In addition, MALDI-TOF analysis of the IAA-treated
bcp-L10-Pu allowed us to detect the MS values consistent with their double AA-adduct (Figure
7d), indicating the completion of backbone macrocyclization. Second, we performed the analysis
of mRNA under the conditions of incubation with 1M NaBH3CN, 1M NaOAc at pH 4.8 for 21
hours, giving no damage of the mRNA template (Supplementary Figure S5). These results firmly
show that the technology developed in this work enables display of backbone-macrocyclic
peptides on their cognate puromycin-mRNA conjugates
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Conclusion
NCL has been applied to the synthesis of various peptides and proteins20,33. This fundamental
chemical concept is also found in a naturally occurring genetically encoded system, so-called
inteins34. Intein systems have been utilized for the expression of backbone-cyclic peptides,
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 34
1
2
3
4
5
6
7
8
9
referred to as split-intein circuit ligation of peptides and proteins (SICLOPPS)35. Although the
SICLOPPS method combined with reverse two-hybrid or other in-cell screening systems allows
the identification of de novo backbone-cyclic peptide ligands against protein targets of interest36-
38, its library size is limited to 109 or less, resulting in relatively rare discovery of bioactive
peptides with very high potencies. By contrast, in vitro display techniques (including RaPID),
which are compatible with libraries of 1012 compounds or greater, have more reliably yielded
potent macrocyclic peptide ligands against a wide range of targets including both intra- or extra-
cellular proteins, and exhibiting KD (or IC50) values in the low nM range or below. Unfortunately,
because such display methods rely on the linkage of peptide C-terminus to cognate mRNA
template, backbone-cyclization has been incompatible to the method; i.e. the displayable scaffold
of macrocyclic peptides on mRNA consists of N-terminus-to-sidechain or sidechain-to-sidechain
linkages.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Here we have devised a new methodology compatible with in vitro display of backbone-
cyclic peptides. Several key technical advances have been made to realize this. First,
incorporation of an α-thio acid, HSFCl, into the peptide main chain was achieved by genetic code
reprogramming. Although the classic study using a puromycin analog where the -amino group
is substituted with -thio group suggests the ribosome capability for the formation of thioester
bond39, this system is not elongatable after thioester bond formation. To the best of knowledge,
there is no report in literature that an -thio acid is ribosomally incorporated and elongated into
peptide chain. Thus, this work represents the first example of such events. Second, the sidechain
thiol of a Cys residue embedded at a position downstream of the HSFCl-thioester bond was
spontaneously exchanged with the ribosomally formed thioester, yielding a new thioester bond
12
ACS Paragon Plus Environment

Page 13 of 34
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
between the main chain and Cys sidechain. This rearrangement was initially not planned, but
during the course of our experiments we found that it proceeded rapidly and efficiently prior to
thioether bond formation as described below. Third, the liberated thiol group of HSFCl attacks the
upstream Cab residue bearing a chloroacetamido sidechain, forming a thioether-cyclic peptide.
This Cab residue was introduced at the N-terminus together with Thz, a protected N-methyl-Cys
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
MeNCys), as a Thz-Cab residue by reprogramming of the initiation of translation. Finally,
deprotection of the Thz group under mild reducing conditions converted it to MeNCys, the thiol
sidechain of which induces backbone-cyclization via NCL. Despite the complexity of the above
reactions, each step of the process worked with acceptable efficiency to produce backbone-cyclic
peptides without loss of the original downstream C-terminal peptide region, meaning that the
methodology should be compatible with in vitro display methods (Figure 5c).
This methodology leverages the full power of genetic code reprogramming assisted by
flexizymes. When combined with further nonproteinogenic amino acid incorporation, such as D-
and N-methyl-L-amino acids, which is certainly possible, RaPID display of backbone-cyclized
pseudo-natural peptides will be achievable through in vitro selection against protein targets of
interest. This opens a critical door to the discovery of new therapeutic modalities and diagnostic
molecules based on such pseudo-natural peptides.
Methods
Additional experimental procedures are provided in the Supplementary Information.
In vitro translation using the FIT system
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 34
1
2
3
4
5
6
7
8
9
The reconstituted cell-free translation system used in this study contained all the
necessary components for translation except for RF1 (and methionine, if necessary). The
concentration of cysteine (0.05 mM) was decreased compared to that of the other 19 proteinogenic
amino acids (0.5 mM) to reduce the possibility of reaction of produced thioester-containing
peptide with cysteine. The composition of the FIT reaction was as follows; 50 mM HEPES-KOH
(pH 7.6), 12 mM magnesium acetate, 100 mM potassium acetate, 2 mM spermidine, 20 mM
creatine phosphate, 2 mM DTT, 2 mM ATP, 2 mM GTP, 1 mM CTP, 1 mM UTP, 0.1 mM 10-
formyl-5,6,7,8-tetrahydrofolic acid, 0.5 mM 19 proteinogenic amino acids other than cysteine,
and 0.05 mM cysteine and 1.5 mg/mL E. coli total tRNA along with 0.73 µM AlaRS, 0.03 µM
ArgRS, 0.38 µM AsnRS, 0.13 µM AspRS, 0.02 µM CysRS, 0.06 µM GlnRS, 0.23 µM GluRS,
0.09 µM GlyRS, 0.02 µM HisRS, 0.4 µM IleRS, 0.04 µM LeuRS, 0.11 µM LysRS, 0.03 µM
MetRS, 0.68 µM PheRS, 0.16 µM ProRS, 0.04 µM SerRS, 0.09 µM ThrRS, 0.03 µM TrpRS,
0.02 µM TyrRS, 0.02 µM ValRS, 0.6 µM MTF, 2.7 µM IF1, 0.4 µM IF2, 1.5 µM IF3, 0.26 µM
EF-G, 10 µM EF-Tu, 10 µM EF-Ts, 0.25 µM RF2, 0.17 µM RF3, 0.5 µM RRF, 0.1 µM T7 RNA
polymerase, 4 µg/mL creatine kinase, 3µg/mL myokinase, 0.1 µM pyrophosphatase, 0.1 µM
nucleotide-diphosphatase kinase, 1.2 µM ribosome and 400 nM DNA templates.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In vitro translation was carried out using the FIT system in the presence of 100 µM
HSFCl-tRNACUA (and 100 µM Thz-Cab-tRNAfMetCAU, if necessary). Translation reactions were
started by adding the acyl tRNA(s) to the FIT system at 37˚C, and translation time was set to 30
min. To promote thioether bond formation, additional incubation was performed for 90 min in
the presence of 33 µg/mL chloramphenicol except for in the case of expression of p1-X5 and p7-
L10.
14
ACS Paragon Plus Environment

Page 15 of 34
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
MALDI-TOF-MS and MS/MS analysis
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
After in vitro translation, the solutions were desalted using an SPE C-tip (Nikkyo
Technos) and eluted with 80% acetonitrile, 0.5% acetic acid solution 50% saturated with the
matrix (R)-cyano-4-hydroxycinnamic acid (Bruker). MALDI-TOF-MS and MS/MS
measurements were performed under linear positive mode using an Autoflex II (Bruker) or
ultrafleXtreme (Bruker) with external calibration (Peptide Calibration Standard II, Bruker).
Tricine-SDS-PAGE analysis for translated peptides
For autoradiography analysis, the translation reactions (2.5 µl) were performed in the
presence of 50 µM [14C]-Asp instead of 500 µM cold Asp, and terminated by adding an equal
volume of 2×Tricine-SDS-PAGE loading buffer (900 mM Tris-HCl (pH 8.45), 8% (w/v) SDS,
30% (v/v) glycerol). The translation products were analyzed by 15% tricine-SDS-PAGE (150 V
constant for 40 min.) and autoradiography was performed using an FLA-5100 (Fujifilm Life
Science) or a Typhoon FLA 7000 (GE Healthcare). The amount of peptide product was quantified
based on the band intensities of known amounts of [14C]-Asp as standards. Values reported are
the average of three independent reactions, and error bars represent the standard deviation.
N-terminal MeNCys liberation of expressed peptides possessing an N-terminal Thz residue
2.5 µL of the MeNCys liberation buffer (1 M NaBH3CN, 1 M NaOAc (pH 4.8)) was
added to 2.5 µL of translation solution. After incubation at 25˚C for 21 hours, N-terminal Thz
was cleanly deprotected to N-terminal MeNCys.
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 34
1
2
3
4
5
6
7
8
9
Thioester bond opening in thioether-cyclic peptides by the addition of excess cysteine
To 5µL of peptide solution, 5µL of the thioester-opening buffer (800 mM Na2CO3-
NaHCO3 (pH 10.0), 20 mM cysteine, 40 mM TCEP-HCl) was added. The reaction was completed
in 1 hour at 42 ˚C.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Free thiol detection in peptides with IAA
5 µL of IAA buffer (100 mM IAA, 1 M HEPES-NaOH (pH 8.0)) was added to 5µL of
peptide solution. The reaction was completed in 1 hour at 25 ˚C.
Backbone macrocyclization acceleration in neutral condition
To 5µL of peptide solution, 6 μL of 1 M HEPES-NaOH (pH 8.0) and 1 μL of 400 mM
TCEP-HCl were added and incubated at 25 ̊C for 3 hours to promote backbone macrocyclization.
Supporting Information
The supporting information is available free of charge on the ACS publications website,
URL: http://pubs.acs.org.
Supporting figures, supporting materials and methods, and supporting references.
Acknowledgements
16
ACS Paragon Plus Environment

Page 17 of 34
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
This work was partially supported by CREST for Molecular Technologies, JST to H.S., and
KAKENHI (16H06444 and 26220204 to H.S.; 26560429 to T.K.; 16H01131 and 17H04762 to
Y.G.) from the Japan Society for the Promotion of Science.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
References
(1)
(2)
(3)
Mattheakis, L. C.; Bhatt, R. R.; Dower, W. J., An in-vitro polysome display system for
identifying ligands from very large peptide libraries. Proc. Natl. Acad. Sci. USA 1994,
91, 9022-9026.
Odegrip, R.; Coomber, D.; Eldridge, B.; Hederer, R.; Kuhlman, P. A.; Ullman, C.;
FitzGerald, K.; McGregor, D., CIS display: In vitro selection of peptides from libraries
of protein-DNA complexes. Proc. Natl. Acad. Sci. USA 2004, 101, 2806-2810.
Nemoto, N.; Miyamoto-Sato, E.; Husimi, Y.; Yanagawa, H., In vitro virus: Bonding of
mRNA bearing puromycin at the 3'-terminal end to the C-terminal end of its encoded
protein on the ribosome in vitro. FEBS Lett. 1997, 414, 405-408.
(4)
(5)
(6)
(7)
Roberts, R. W.; Szostak, J. W., RNA-peptide fusions for the in vitro selection of
peptides and proteins. Proc. Natl. Acad. Sci. USA 1997, 94, 12297-12302.
Murakami, H.; Ohta, A.; Ashigai, H.; Suga, H., A highly flexible tRNA acylation
method for non-natural polypeptide synthesis. Nat. Methods 2006, 3, 357-359.
Goto, Y.; Katoh, T.; Suga, H., Flexizymes for genetic code reprogramming. Nat. Protoc.
2011, 6, 779-790.
Xiao, H.; Murakami, H.; Suga, H.; Ferré-D’Amaré, A. R., Structural basis of specific
tRNA aminoacylation by a small in vitro selected ribozyme. Nature 2008, 454, 358-
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 34
1
2
3
4
5
6
7
361.
8
9
(8)
Goto, Y.; Ohta, A.; Sako, Y.; Yamagishi, Y.; Murakami, H.; Suga, H., Reprogramming
the translation initiation for the synthesis of physiologically stable cyclic peptides. ACS
Chem. Biol. 2008, 3, 120-129.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(9)
Goto, Y.; Murakami, H.; Suga, H., Initiating translation with D-amino acids. RNA 2008,
14, 1390-1398.
(10)
Kawakami, T.; Murakami, H.; Suga, H., Messenger RNA-programmed incorporation of
multiple N-methyl-amino acids into linear and cyclic peptides. Chem. Biol. 2008, 15,
32-42.
(11)
(12)
(13)
Katoh, T.; Tajima, K.; Suga, H., Consecutive elongation of D-amino acids in
translation. Cell Chem. Biol. 2017, 24, 46-54.
Katoh, T.; Iwane, Y.; Suga, H., Logical engineering of D-arm and T-stem of tRNA that
enhances D-amino acid incorporation. Nucleic Acids Res. 2017, 45, 12601-12610.
Yamagishi, Y.; Shoji, I.; Miyagawa, S.; Kawakami, T.; Katoh, T.; Goto, Y.; Suga, H.,
Natural product-like macrocyclic N-methyl-peptide inhibitors against a ubiquitin ligase
uncovered from a ribosome-expressed de novo library. Chem. Biol. 2011, 18, 1562-
1570.
(14)
(15)
Hayashi, Y.; Morimoto, J.; Suga, H., In vitro selection of anti-Akt2 thioether-
macrocyclic peptides leading to isoform-selective inhibitors. ACS Chem. Biol. 2012, 7,
607-613.
Morimoto, J.; Hayashi, Y.; Suga, H., Discovery of macrocyclic peptides armed with a
mechanism-based warhead: isoform-selective inhibition of human deacetylase SIRT2.
18
ACS Paragon Plus Environment

Page 19 of 34
Journal of the American Chemical Society
1
2
3
4
5
6
7
Angew. Chem. Int. Ed. Engl. 2012, 124, 3479-3483.
8
9
(16)
(17)
(18)
Ito, K.; Sakai, K.; Suzuki, Y.; Ozawa, N.; Hatta, T.; Natsume, T.; Matsumoto, K.; Suga,
H., Artificial human Met agonists based on macrocycle scaffolds. Nat. Commun. 2015,
6, 6373.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Song, X.; Lu, L.Y.; Passioura, T.; Suga, H., Macrocyclic peptide inhibitors for the
protein-protein interaction of Zaire Ebola virus protein 24 and karyopherin alpha 5.
Org. Biomol. Chem. 2017, 15, 5155-5160.
Kawamura, A.; Munzel, M.; Kojima, T.; Yapp, C.; Bhushan, B.; Goto, Y.; Tumber, A.;
Katoh, T.; King, O. N.; Passioura, T.; Walport, L. J.; Hatch, S. B.; Madden, S.; Muller,
S.; Brennan, P. E.; Chowdhury, R.; Hopkinson, R. J.; Suga, H.; Schofield, C. J., Highly
selective inhibition of histone demethylases by de novo macrocyclic peptides. Nat.
Commun. 2017, 8, 14773.
(19)
(20)
(21)
(22)
(23)
Joo, S. H., Cyclic peptides as therapeutic agents and biochemical tools. Biomol. Ther.
2012, 20, 19.
Dawson, P. E.; Muir, T. W.; Clark-Lewis, I.; Kent, S., Synthesis of proteins by native
chemical ligation. Science 1994, 266, 776-779.
Fahnestock, S.; Rich, A., Ribosome-catalyzed polyester formation. Science 1971, 173,
340-343.
Ohta, A.; Murakami, H.; Higashimura, E.; Suga, H., Synthesis of polyester by means of
genetic code reprogramming. Chem. Biol. 2007, 14, 1315-1322.
Fahnestock, S.; Rich, A., Synthesis by ribosomes of viral coat protein containing ester
linkages. Nat. New Biol. 1971, 229, 8.
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 34
1
2
3
4
5
6
7
8
9
(24)
(25)
(26)
Ellman, J. A.; Mendel, D.; Schultz, P. G., Site-specific incorporation of novel backbone
structures into proteins. Science 1992, 255, 197-200.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Kawakami, T.; Ohta, A.; Ohuchi, M.; Ashigai, H.; Murakami, H.; Suga, H., Diverse
backbone-cyclized peptides via codon reprogramming. Nat. Chem. Biol. 2009, 5, 888.
Bekhor, I. J.; Mohseni, Z.; Bavetta, L. A., Inhibition of proline-C14 incorporation into
rat liver ribosomes by thiazolidine-4-carboxylic acid in a cell-free system. Proc. Soc.
Exp. Biol. Med. 1965, 119, 765-769.
(27)
Hartman, M. C.; Josephson, K.; Lin, C. W.; Szostak, J. W., An expanded set of amino
acid analogs for the ribosomal translation of unnatural peptides. PLoS One 2007, 2,
e972.
(28)
(29)
Villain, M.; Vizzavona, J.; Rose, K., Covalent capture: a new tool for the purification of
synthetic and recombinant polypeptides. Chem. Biol. 2001, 8, 673-679.
Stolarski, R.; Kierdaszuk, B.; Hagberg, C. E.; Shugar, D., Hydroxylamine and
methoxyamine mutagenesis: displacement of the tautomeric equilibrium of the
promutagen N6-methoxyadenosine by complementary base pairing. Biochemistry 1984,
23, 2906-2913.
(30)
Stolarski, R.; Kierdaszuk, B.; Hagberg, C. E.; Shugar, D., Mechanism of
hydroxylamine mutagenesis - tautomeric shifts and proton-exchange between the
promutagen N6-methoxyadenosine and cytidine. Biochemistry 1987, 26, 4332-4337.
Tsukiji, S.; Pattnaik, S. B.; Suga, H., An alcohol dehydrogenase ribozyme. Nat. Struct.
Mol. Biol. 2003, 10, 713.
(31)
(32)
Hackeng, T. M.; Griffin, J. H.; Dawson, P. E., Protein synthesis by native chemical
20
ACS Paragon Plus Environment

Page 21 of 34
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
ligation: Expanded scope by using straightforward methodology. Proc. Natl. Acad. Sci.
USA 1999, 96, 10068-10073.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(33)
(34)
(35)
(36)
Thapa, P.; Zhang, R. Y.; Menon, V.; Bingham, J. P., Native chemical ligation: a boon to
peptide chemistry. Molecules 2014, 19, 14461-14483.
Shah, N. H.; Muir, T. W., Inteins: Nature's gift to protein chemists. Chem. Sci. 2014, 5,
446-461.
Scott, C. P.; Abel-Santos, E.; Wall, M.; Wahnon, D. C.; Benkovic, S. J., Production of
cyclic peptides and proteins in vivo. Proc. Natl. Acad. Sci. USA 1999, 96, 13638-13643.
Horswill, A. R.; Savinov, S. N.; Benkovic, S. J., A systematic method for identifying
small-molecule modulators of protein-protein interactions. Proc. Natl. Acad. Sci. USA
2004, 101, 15591-15596.
(37)
(38)
Tavassoli, A.; Benkovic, S. J., Genetically selected cyclic-peptide inhibitors of AICAR
transformylase homodimerization. Angew. Chem. Int. Ed. Engl. 2005, 44, 2760-2763.
Young, T. S.; Young, D. D.; Ahmad, I.; Louis, J. M.; Benkovic, S. J.; Schultz, P. G.,
Evolution of cyclic peptide protease inhibitors. Proc. Natl. Acad. Sci. USA 2011, 108,
11052-11056.
(39)
Gooch, J.; Hawtrey, A. O., Synthesis of thiol-containing analogues of puromycin and a
study of their interaction with N-acetylphenylalanyl-transfer ribonucleic acid on
ribosomes to form thioesters. Biochem. J. 1975, 149, 209-220.
Figure Legends
21
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 34
1
2
3
4
5
6
7
8
9
Figure 1| Compatibility of macrocyclization with in vitro display. (a) The N-terminal ClAc
group reacts with the sidechain of a downstream Cys, forming a thioether macrocyclic peptide,
which retains the fusion of downstream peptide region (shown in cyan) to Pu-mRNA. Pu indicates
puromycin. This system is compatible with in vitro display techniques. (b) Backbone
macrocyclization via native chemical ligation (NCL), producing detachment of the downstream
peptide region from Pu-mRNA. This system is incompatible with in vitro display techniques. (c)
A new strategy to display backbone-cyclized peptides via an N-methyl-peptide bond, which is
compatible with in vitro display technologies. Peptides are linked to mRNA by two steps (S1 and
S2). Details are discussed in this work.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2| Ribosome-mediated thioester bond formation. (a) Chemical structure of the α-thio
acids used in this study. (b) Quantification of microhelix RNA thioacylation by acid-PAGE
analysis. (c) The mRNA sequence (mr1) encoding the p1-X5 peptide. X represents an α-thio acid
of choice as shown in (a). (d) MALDI-TOF-MS of the thioester-containing peptide P1-sFp-Cl
expressed from mr1 in the FIT system. (e) Quantification of peptides expressed from mr1 in the
FIT system. The expression levels (µM in the reaction mixture) of peptides were quantified based
on the band intensities of known amounts of [14C]-Asp as standards. Lane 1, p1-F5 suppressed
with phenylalanine F-tRNACUA; p1-SFCl5 suppressed with HSFCl-tRNACUA; p1-SG5 suppressed
with HSG-tRNACUA
.
Figure 3| The initial strategy for linking upstream and downstream sequences to form
backbone-cyclized peptides compatible with in vitro display. (a) The mRNA sequence (mr2-
22
ACS Paragon Plus Environment

Page 23 of 34
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
W8) encoding the p2-W8 peptide. R1 represents the first reaction that spontaneously forms a
thioether bond between Cab2 and Cys12. R2 represents the second reaction (NCL forming an
amide backbone) after deprotection of Thz1. (b) Schematic view of the entire process via R1 and
R2. In this process, the downstream sequence of the peptide would not detach from the backbone-
cyclized peptide, and would thus be compatible with in vitro display. “tcp” and “bcp” stand for
thioether-cyclic peptide and backbone-cyclic peptide, respectively. The C-terminal Ser residue is
shown as “S”. Pu-mRNA shows the potential fusion site at the C-terminal of the peptide.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4| Analysis of tcp2-W8 by MALDI-TOF MS/MS. (a) Schematic view of cleavage of
the thioester in tcp2-W8 by the addition of Cys via thioester exchange followed by intramolecular
amidation. (b) MALDI-TOF-MS of p2-W8-Cys. (c) MS/MS analysis of p2-W8-Cys.
Fragmentation data was inconsistent with the predicted structure of p2-W8-Cys as shown in the
right fragmentation profile. (d) The revised structure of p2-W8-Cys, the fragmentation profile of
which was consistent with the observed data.
Figure. 5| Revised pathway of backbone-cyclized peptide synthesis. (a) The mRNA sequence
(mr2-W8) encoding the p2-W8 peptide. Rtex represents spontaneous thioester exchange between
HSFCl9 and Cys12 prior to R1 and R2. R1 forms a thioether bond between Cab2 and HSFCl9, and
R2 forms a backbone-cyclized peptide, bcp2-W8, via NCL after deprotection of Thz1. (b)
Schematic view of the entire process via Rtex, R1, and R2. In this process, the downstream
peptide region would not detach from the backbone-cyclized peptide, and would thus be
compatible with in vitro display. (c) Treatment of bcp2-W8-SC12 and bcp2-W8 with 2-
23
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 34
1
2
3
4
5
6
7
8
9
iodoacetoamide (IAA). The initial peptide intermediate should be modified with a single AA,
whereas the latter backbone cyclized peptide should be modified with double AA (see Figure S3
for the MS data). (d) MALDI-TOF MS/MS analysis of bcp2-W8. “i[bX-yX]” labels indicate the
peaks corresponding to internal peptide fragments resulting from two fragmentations inside the
macrocyclic moiety. The internal peptide fragments containing the newly generated Trp8-
MeNCys1 amide bond are shown in bold.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6| Backbone-cyclization of various peptides compatible with in vitro display methods.
(a) Two peptides, p3-W8 and p4-W8, with two or three amino acids between the HSFCl9 and the
downstream Cys. (b) MS data of tcp3-W8 (top), bcp3-W8 (middle) after NaBH3CN treatment,
and bcp3-W8 2IAA adduct (bottom). (c) MS data of tcp4-W8 (top), bcp4-W8 (middle) after
NaBH3CN treatment, and bcp4-W8 2IAA adduct (bottom). (d) Two peptides, p5-L11 and p6-Q14,
with different amino acid sequences between the Cab2 and the HSFCl9. (e) MS data of tcp5-L11
(top), bcp5-L11 (middle) after NaBH3CN treatment, and bcp5-L11 2IAA adduct (bottom). (f) MS
data of tcp6-Q14 (top), bcp6-Q14 (middle) after NaBH3CN treatment, and bcp6-Q14 2IAA
adduct (bottom).
Figure 7| Conjugation of a model backbone-cyclic peptide with a puromycin.
(a) The mRNA sequence (mr7) encoding p7-L10 conjugated with a puromycin at the C-
terminus followed by backbone cyclization to yield bcp7-L10. (b) Mass chromatogram at m/z =
1168.42 obtained from LC/ESI-MS analysis of p7-L10. (c) Mass spectrum of p7-L10 isolated at
24
ACS Paragon Plus Environment

Page 25 of 34
Journal of the American Chemical Society
1
2
3
4
5
6
7
the retention of 11.51 min. (d) MALDI-TOF MS of IAA-treated bcp-L10-Pu. In the right panel,
8
9
the region of 2450–2500 m/z was expanded to show their isotope peaks.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
25
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 34
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
26
ACS Paragon Plus Environment

Page 27 of 34
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
27
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 34
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
28
ACS Paragon Plus Environment

Page 29 of 34
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
29
ACS Paragon Plus Environment