Tandem Tetrahydroisoquinoline-4-carboxylic Acid/β-Alanine as a New Construct Able To Induce a Flexible Turn

Article abstract of DOI:10.1002/chem.201701045

Tetrahydroisoquinoline-4-carboxylic acid, a constrained β²-amino acid named β-TIC, was synthesised for the first time in enantiopure form. The biocatalytic route applied herein represents one of the few successful examples of enzymatic resolution of β²-amino acids. Model tetrapeptides, namely, Fmoc-l-Ala-β-TIC-β-Ala-l-Val-OBn (Fmoc=fluorenylmethyloxycarbonyl, Bn=benzyl), containing both isomers of β-TIC, were prepared. Both computational and NMR spectroscopy studies were performed. A reverse-turn conformation was observed in the case of (R)-β-TIC enantiomer that was obtained in 99 % enantiomeric excess by enzymatic resolution. The β-TIC/β-Ala construct represents the first example of a flexible turn mimetic containing a cyclic and an acyclic β-amino acid. Furthermore, the presence of an aromatic ring of β-TIC could facilitate non-covalent interactions to increase the potential of this scaffold for the preparation of protein–protein interaction modulators.

Full text of DOI:10.1002/chem.201701045

A Journal of
Accepted Article
Title: Tandem Tetrahydroisoquinoline-4-carboxylic Acid/β-Alanine as a
New Construct Able to Induce a Flexible Turn
Authors: Maria Luisa Gelmi, Alessandro Contini, Andrea Bonetti,
Raffaella Bucci, Davide Tessaro, Francesca Clerici, Sara
Pellegrino, and Donatella Nava
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201701045
Link to VoR: http://dx.doi.org/10.1002/chem.201701045
Supported by

Chemistry - A European Journal
10.1002/chem.201701045
Tandem Tetrahydroisoquinoline-4-carboxylic Acid/-Alanine as a New Construct Able to
Induce a Flexible Turn
Raffaella Bucci, Dr. Andrea Bonetti, Prof. Francesca Clerici,[a] Donatella Nava,[a] Dr. Sara
[a]
†
Pellegrino, [a] Dr. Davide Tessaro, [b] Prof. Maria Luisa Gelmi[a]*
[a]DISFARM, Sezione di Chimica Generale e Organica “A. Marchesini” Università degli Studi
Milano, Via Venezian 21, I-20133 Milano (Italy)
[b]Department of Chemistry, Materials and Chemical Engineering “G. Natta”, Politecnico di Milano,
p.za L. da Vinci 32, I-20133 Milano (Italy)
*
Maria Luisa Gelmi, full professor, marialuisa.gelmi@unimi.it
2
Abstract. Tetrahydroisoquinoline-4-carboxylic acid, a constrained β -amino acid named -TIC,
was for the first time synthesized in enantiopure form. The biocatalytic route here applied
2
represents one of the few successful examples of enzymatic resolution of β -amino acids. Model
tetrapeptides, i.e. Fmoc-L-Ala--TIC--Ala-L-Val-OBn, containing both isomers of -TIC, were
prepared. Both computational and NMR studies were performed. A reverse-turn conformation was
observed in the case of R--TIC enantiomer that was obtained in 99% e.e. by enzymatic resolution.

-TIC/-Ala construct represents the first example of a flexible turn mimetic containing a cyclic
and an acyclic -amino acid. Furthermore, the presence of an aromatic ring of -TIC could
facilitate non-covalent interactions increasing the potential of this scaffold for the preparation of
protein-protein interaction modulators. 
Introduction
It is of general knowledge that peptide shape and conformation influence proteins structure and
function.[1] Loops and turns, in particular, acting as linking motifs between different secondary
structures, are the molecular tools used to reverse the peptide backbone direction inducing protein
tertiary structure. Even if loops are often called random coils, they exhibit a significant amount of
regular structural patterns involving polypeptide segments of varying length.[2] Considering the
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
enormous variety of sequences that can form such motifs, researches focused on their identification
are gaining increasing attention.[2b]
[
3]
Small-molecules capable of mimicking peptide turn structures are particularly useful in validating
new drug targets, and in the development of new therapeutics. They are indeed valuable in defining
the residues and the secondary structures responsible for the binding recognition and affinity, and,
of particular relevance, they can induce an increased stability to proteolysis. One of the most
promising applications of turn mimics is addressed to the modulation of protein-protein interactions
(PPIs). Turn region is indeed often involved in the protein interaction interface. An important
feature to be taken into account when designing PPIs modulators is the flexibility of the proteins.
[
4]
So, their adaptability in the formation of the complex with the protein is of relevance. The use of
flexible turn tools that can accommodate themselves to the target in a better way with respect to the
rigid turn mimics can represent an effective tool to approach this issue. Another important feature is
the type of functional groups of the turn region responsible for PPI. An aromatic residue could be
considered a key element in stabilizing the complex through non covalent bonds.[5]
[
6]
[6c,i,l]
The synthesis of non natural amino acids, and of scaffolds to be used as turn-mimics
as well
as their use to prepare peptide model sequences[6a-c,e] and PPI inhibitors[4a,7] is one of our main
research fields.
2
Recently, we envisaged the tetrahydroisoquinoline-4-carboxylic acid (1, Scheme 1), a β -amino
acid, as a potential turn mimetic satisfying the above requirements. According to the nomenclature
[
8]
reported for the corresponding 3-carboxylic acid, a -amino acid, we named compound 1 -TIC.
-TIC is a very expensive commercially available compound whose preparation in racemic form

[
9]
has already been reported in literature. Here, we revisited the synthetic protocol, reducing the
synthetic steps, allowing us to obtain ethyl ester of 1 in gram scale and excellent yields.
Furthermore, we attempted a biocatalytic route leading to the unavailable enantiopure -TIC.
Biocatalysis is indeed a very efficient tool for performing a chemical resolution in the frame of
green chemistry.[10] The obtainment of enantiopure -TIC places several issues. First, H- proton in
benzylic position is very acidic, favouring its potential racemization in non-mild conditions. To our
3
2
knowledge, in contrast with their β homologs, few enzymatic resolutions of β -amino acids have
been described in literature giving poor results in terms of enantioselectivity and efficiency.[11]
More recently, oxazinones were used as suitable precursors for the enzymatic dynamic kinetic
2
[11b]
resolution of non cyclic β -amino acids.
The most similar analogues on which a kinetic
resolution have been carried out, although with scarce success, have been reported in a paper by
Kanerva et al. [11c] Despite these not so encouraging results, we considered it would have be worth
attempting an enzymatic resolution, firstly for its simplicity, but also for its potential success.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
Due to the importance of β-amino acid residues in controlling peptide shape and function,[12] in the
second part of our researches, we focused on the use of -TIC in combination with β-alanine, as an
effective tool for flexible turn mimics. The use of -TIC for the preparation of peptides has been
claimed on several patents, although no literature data are reported so far.
Gellman et al.[3d,e,g] used the parent nipecotic acid (Nip) as a turn inducer when inserted in -
peptide sequences as an homo-dipeptide. Curiously, R-Nip-S-Nip dipeptide segment or the S, R
counterpart, gives -peptide reverse-turns that promote the hairpin formation. On the other hand,
dipeptides containing Nip in the same stereochemistries are ineffective,[3e] just like when a single
[3g]
Nip unit is matched with an acyclic -amino acid.
In our case, we envisaged that the aromatic part of -TIC could induce a further constrain in the
dipeptide architecture, better orienting the groups involved in a H-bond. A turn conformation could
be stabilized, also in the presence of an acyclic -amino acid. The combination of the constrained

-TIC with the highly conformational flexible β-alanine might constitute the molecular trick for
modulating rigidity and adaptability.[4] This construct provides a simple model system to be used
for future applications such as in building peptidomimetics addressed to PPI interaction. As added
value, the presence of the aromatic ring could favour non-covalent interactions, as point out before.
Finally, the use of the -TIC/achiral β-alanine, instead of two chiral amino acids of opposite
conformation, makes easier the synthetic approach of the corresponding turn.
Short model peptide sequences containing both L--amino acids and -amino acids, i.e. the
tetrapeptides Fmoc-L-Ala--TIC--Ala-L-ValOBn, were prepared. In order to prove the correlation
between the conformation of the peptides and -TIC stereochemistry, both -TIC-enantiomers were
tested. Both computational and spectroscopic studies were performed to prove the ability of -TIC-

-Ala construct to generate a turn structure.
Results
[
9]
Synthesis of amino acid 1 and derivatives. As reported in literature, -TIC 1 is prepared from 3-
bromo-isoquinoline in 4 steps and 45% overall yield. Here we revisited the synthetic procedure,
decreasing the synthetic steps, and focusing on the synthesis of ethyl ester of -TIC, the key reagent
for the enzymatic resolution. Ester 3 (98%) was directly prepared from 4-bromo-isoquinoline (2)
using a recent procedure reported for the corresponding methyl ester,[13] consisting of a
carbonylation reaction in the presence of EtOH/TEA, Pd(OAc)
2
and xantphos as ligand. The
reduction of pyridine ring[14] with H
in the presence of Pt O in AcOH gave the expected
2
2
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
compounds 4 (85%) (Scheme 1). Amino acid 1 (HCl salt, quantitative yield) was obtained by
hydrolysis of esters 4 with 6N HCl (100 °C, 13h; Scheme 1).
It should be stressed that, starting from the cheap commercially available 4-bromoisochinoline (2),
amino acid 1 was prepared in gram sale (10 g) in three steps and 83% overall yield.
The preparation of the Boc-derivative 5 (60%) was achieved starting from 4 by reaction with Boc-
anhydride (dioxane/H
2
O/NaHCO
3
, 80 °C). The yield of 5 was dramatically increased (96%) using
(25 °C, 26 h) (Scheme 1).
TEA (2.2 eq.) as the base in CH
2
Cl
2
Enzymatic resolution of N-Boc derivative 5. NBoc-protected ethyl ester 5 was chosen as the
substrate for the enzymatic resolution. This ester was more stable with respect to the corresponding
methyl derivative to the spontaneous hydrolysis in the enzymatic resolution conditions (buffer at pH
8
, 37 °C). Moreover, the hydrogen at C-4 is sufficiently acidic to permit the continuous
5
]
racemization of the substrate. This would give a base for trying to work under dynamic kinetic
resolution conditions.[16]
An intensive screening involving about 50 commercial hydrolases belonging to various subclasses
(esterases, lipases, proteases, peptidases) was carried out (Table TS1 and Figure 1; for details see
Supporting Information). Pronase, a mixture of proteases extracted from the extracellular fluid of
Streptomyces griseus, proved to be the most convenient catalyst in terms of activity and specificity.
It was able to mediate the stereoselective hydrolysis of the precursor with a not outstanding, yet
significative, enantioselectivity (E = 17). As a result, it is possible to isolate the slow-reacting
enantiomer (i.e. (-)-5) in high enantiomeric purity by pushing the conversion over 65% (see Figure
FS1 in Supporting Information).
The scale up of the enzymatic resolution was performed starting from racemic 5 (1.5 g). The
hydrolysis was carried out in a water/DMSO mixture kept at pH = 8 by means of an automatic
titrator: when the conversion reached about 65-70%, we were able to isolate the starting ester (-)-5
with a very high enantiopurity (e.e. > 99%) (Scheme 2). Enantio-enriched acid of opposite
configuration (+)-6 (e.e. ≈ 48%) was isolated and transformed back into the corresponding ester
(
+)-5 (DMAP/CH
material for the successive batch taking advantage of the easiness of the racemization of C-4 in
basic conditions (DBU/CH Cl ; Scheme 2).
2
Cl /EtOH, 0 °C, then DCC; e.e. ≈ 48%). Ester (+)-5 was thus recycled as starting
2
2
2
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
Starting from (-)-5, amino ester (-)-4 TFA was obtained in 100% yield operating in standard
conditions (TFA, CH Cl , 0 °C, 1h) (Scheme 2).
2
2
Preparation of tetrapeptide models Fmoc-L-Ala--TIC- Ala-L-Val-OBn. In order to evaluate the
ability of -TIC to induce a turn and to understand the effect of its chirality, two tetrapeptide
models, i.e. Fmoc-L-Ala--TIC--Ala-L-Val-OBn, were prepared containing the R- or S--TIC
stereoisomer, respectively, (Scheme 3).
The synthesis of dipeptide  Ala-L-Val-OBn (7) is not reported in literature. The preparation of 7
.
from Boc- Ala and L-valineOBn HCl using a standard protocol [HOBT (1.2 equiv.)/EDC (1.2
equiv.), DIPEA (3 equiv.), CH
2
Cl ] gave poor yields due to the formation of a 2,6-
2
dioxotetrahydropyrimidine by-product. An increase of the yield of 7 was achieved by transforming
first Boc-Ala into its corresponding anhydride [iBuOCOCl, N-methylmorpholine (NMM, 1
.
equiv.), CH
2
Cl
2
, 0 °C, 1h] that was directly coupled with L-valineOBn HCl (NMM, 1 equiv.). Boc-

Ala-L-Val-OBn (84%) was isolated and deprotected at nitrogen atom (TFA, CH
2
Cl , 0 °C, 1h)
2
affording dipeptide 7 (76%; Scheme SS1).
.

-TIC ester (-)-4 TFA was treated with Fmoc-L-Ala (HOBT/EDC/CH
2
Cl2, 25 °C, 8h) giving (-)-8
in 90% yields. To avoid the racemization of -TIC during the hydrolysis of the ester function,
different conditions were tested. The acid (-)-9 (89%) was successfully obtained without
racemization from (-)-8 using Me
3
SnOH[17] (CH
2
Cl , 50 °C, 10 h). The reaction of 9 with -Ala-L-
2
ValOBn (7) (HOBT/EDC, DMF, 25 °C, 2 h) gave (+)-10 (71%). Peptide (+)-10 was obtained in
excellent overall yields (57%) starting from (-)-4.
The whole protocol could be performed from the racemic compound 4, without separation of any
intermediate. Tetrapeptides (+)-10 and (-)-11 were obtained in 43% overall yield and were partially
separated by column chromatography.
1
NMR peptide characterization. H NMR spectra of dipeptides (-)-8 and (-)-9 showed the presence
of two conformers (1:1). To confirm the presence of two rotames, studies at variable temperatures
were performed for (-)-9 (DMSO-d
6
) indicating a temperature of coalescence of 50°C (Figure FS3).
1
13
Both tetrapeptides (+)-10 and (-)-11 were fully characterized by NMR ( H, C, HMQC, HMBC,
TOCSY and ROESY experiments in CDCl
resonances were assigned.
3
, 20 Mm, 500 MHz). As reported in Tables TS3-TS6 all
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
1
H NMR spectrum of peptides (+)-10 showed the presence of two rotamers (10 and 10’, 1:1) at -
TIC tertiary amide bond.[3e] According to the computational models, a gauche and a staggered
conformation are proposed for -alanine in 10 and10’, respectively.
All NH signals in both rotamers are well dispersed indicating defined conformations. Furthermore,
NH resonances of the same amino acid in the two rotamers fall in the same region. Valine NHs
resonate at lower fields (7.88 and 7.16, respectively) with respect to NHs of -Ala (6.62 and
6
.10) and Ala (5.70 and 5.96).
The solvent titration of the NH chemical shifts was performed (Figure 2). The polar solvent DMSO
added to the CDCl solution induces similar chemical shift changes in both rotamers. A pronounced
downfield shift with increasing concentrations of DMSO was observed for NHAla (NH: 0.66 and
.02, respectively) indicating its solvent exposure. The intermediate value for NHAla (NH:0.47
3
1
and 0.69, respectively) suggests that this NH is partially solvated. The chemical shift of NHVal is
insensitive to the solvent (NH: 0.14 and 0.19, respectively), confirming its shielding from the
solvent. This datum, together with its resonance at low field, confirms that NH of valine is involved
in a strong intramolecular hydrogen bond. Comparing the stability of the H-bond network of the
two rotamers, it is evident that rotamer 10 is less sensitive to DMSO with respect to 10’, indicating
a stronger H-bond stability.
NOESY/ROESY experiments (see Tables TS3 and TS4 for detail in SI and Figure 3) at different
mixing times (from 100, 200, 300 ms) gave interesting information on the secondary structure of
peptide (+)-10 (Figures FS4, FS5 and FS6) and, together with J values of the -TIC protons,
allowed us to assign the R-stereochemistry to -TIC (see Supporting Information for discussion).
According to the ROEs (Figure 3) and the computational models, it was possible to assign
unequivocally the pro-R and pro-S chemical shifts to methylene groups of -alanine of both
conformer 10 and 10’.
A significant ROE for rotamer 10, supporting the formation of a turn, is that of H-3-TIC in
equatorial position with NHVal (Figure 4A). As shown in Figure 4B, a very week ROE was detected
between H-3pro-S of -ALA and H-3-TIC in axial position. According to the computational models
(see below), this ROE is not consistent with conformer 10, neither with 10’, but it is probably due to
a transition conformation between 10/10’. Both CHAla and MeAla showed proximity with H-1-TIC
(Figure 4C) confirming that the carbonyl function of Ala is oriented inside the turn (E rotamer).
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
Furthermore, a complete set of CH/NH(i,i+1) is present (Figure FS5).
Rotamer 10’ (Figure 3) does not present intrastrand NOEs and showed spatial proximity between
H- and Me of alanine with H-3-TIC, confirming the presence of the Z-rotamer at tertiary amide
(Figures 4D and FS 6).
Taken together these data confirm that peptide (+)-10, containing (R)--TIC enantiomer, gave two
turn conformations that are in equilibrium. (Figure 3). The formation of a strong H-bond between
C=O of alanine with NHVal (12-membered ring) is proposed for the E-rotamer 10. The intermediate


NH value of NHAla found in titration experiment suggests that an equilibrium between a H-bonded
and not-bonded could occur between NHAla and C=O of valine. On the other hand, the Z-rotamer
0’ forms a strong H-bond between NHVal and C=O of the Fmoc group. In this case a bigger turn
15-membered ring) was formed (Figure 3).
1
(
Peptide (-)-11, containing S--TIC, is present as a mixture of three conformers in a 1:0.6:0.1 ratio.
Only the two main isomers, named 11 and 11’, were fully characterized (Tables TS5 and TS6).
Differently from (+)-10, NH signals are not dispersed and the DMSO titration gave NH values
larger then 0.7, except for NHVal (NH 0.5) (Figure FS10). A complete set of CH/NH(i,i+1) ROEs
were detected exclusively at C-termini arm for 11’ (Figures FS7 and FS8). MeAla shows spatial
proximity with H-1-TIC in both conformers (11, m; 11’, w; Figure FS9). These data support the
formation of E-rotamers. In both cases, intrastrand ROEs are absent. These data indicate that all the
isomers of peptide (-)-11 did not assume a preferred conformation.
Based on NOE data, a computational refinement of 10 and 10’ structures was attempted by mean of
molecular dynamic (MD) simulations. Distance restraints derived from NMR experiments (see
Experimental), shown in Table 1, were used to drive a Simulated Annealing (SA) experiments
consisting in a quick heating phase followed by a slow cooling phase.
The simulation was repeated three times starting from randomly chosen conformations, and full
convergence was observed. The final geometries obtained from the SA using the restraint set
derived from the NOE of conformers 10 and 10’ are depicted in Figure 5.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
Both 10 and 10’ structures fit well with the geometries proposed in Figure 3 and can be considered
as models of peptide turns. Conformation 10 is characterized by a peptide bond between Ala1 and
β-TIC in pseudo-cis configuration. Indeed, the Ala1 C=O points “inward” with respect to the turn
curvature and is involved in a rather strong H-bond with Val4 NH (O···H distance = 1.88 Å) making
this turn rather similar to a type II β-turn. Conversely, in conformation 10’ the Ala1-β-TIC peptide
bond is in pseudo-trans configuration and an H-bond can be observed between the Fmoc-carbonyl
and Val4 NH. Selected torsional angles, reported in Table 2, show that the β-TIC geometry is rather
similar in both conformations. Principal differences in the turn core, in addition to the configuration
of the Ala1-β-TIC peptide bond, are found in the β-Ala dihedrals (φi+2, θi+2 and ψi+2). Indeed, a
gauche conformation is observed for the two β-Ala methylene groups in 10, in accordance with
experimental data, while a staggered conformation (corresponding to a fully extended backbone) is
found for 10’.
Another important difference between the two conformations is observed for the benzyloxy
protected Val that terminates the turn core. Indeed, in conformer 10 the benzyl group is oriented
toward the β-TIC scaffold, while in 10’ it points toward the Fmoc (fluorene moiety). It is possible
that turn conformations are stabilized by the hydrophobic collapse of aromatic rings in polar
solvents.
To evaluate the degree of such a contribution, we performed Replica Exchange Molecular
Dynamics (REMD) simulations on peptide 10 and on model peptide m10, differing from the former
in the nature of protecting groups [acetyl and NHMe (m10) instead of Fmoc and OBn (10)].
Representative geometries obtained from the cluster analysis of the REMD trajectory at about 300
K are shown in Figure 6, together with the populations of the corresponding conformational cluster.
First of all, it can be observed that the two principal geometries of peptide 10 (#1 and #2, Figure 6,
panel A) matches those obtained by the computational refinement of NMR, even if geometry #1,
which resembles conformation 10’, resulted the most abundant. This might be due to the use of a
solvent model for water, as acetonitrile cannot be simulated by the Generalized Born solvent model
implemented in Amber.[18,19] However, since both conformation are well represented, we can
consider the results of the simulation sufficiently accurate for our purpose. Geometry #3 (Figure 6,
panel A) represents about the 12% of the conformational population and is characterized by a γ turn
involving β-TIC and β-alanine.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
Considering model m10 (Figure 6, panel B), similar geometries were obtained, but with an inverted
stability order. Indeed, the most abundant conformation of m10 is comparable to conformation #3
of peptide 10. Conversely, the turn conformation #2 of m10 is superimposable to conformation #1
of 10 with a root mean squared displacement (RMSD) of 0.05 Å, calculated on α carbons, meaning
that the two geometries are identical.
In conclusion, computational results suggest that β-TIC/β-Ala is intrinsically able to stabilize a turn
conformation whose type is influenced by the N- and C-protecting groups.
Conclusion
The synthesis of -TIC 1 was revisited affording the ester 4 in gram scale (10 g) and excellent
overall yields. The enzymatic kinetic resolution of 4 was performed in gram scale too (1.5 g),
2
representing one of the few successful examples of enzymatic resolution of β -amino acids.
Furthermore, this allowed an enantiopure material (R isomer: e.e. > 99%) to be provided for the
first time. This has been employed in the synthesis of the target peptide 10. The racemization of the
enantio-enriched S-isomer was achieved in order to recycle the material and improve the yield in
the view of a possible scale-up. Finally, R--TIC is the right stereoisomer providing a peptide
reverse-turn when linked to -alanine, as demonstrated by computational and NMR studies for
tetrapeptide model Fmoc-L-Ala-(R)--TIC --Ala-L-Val-OBn.
Experimental Section
Ethyl isoquinoline-4-carboxylate (3). Compound 3 was prepared according to the known
1
3
procedure, used for the methyl ester, starting from 4-bromo-isoquinoline (2) (10 g, 48 mmol) in
the presence of EtOH (70 mL), TEA (70 mL), Pd(OAc) (430 mg, 1.96 mmol) and xantphos (555
2
mg, 0.96 mmol) under CO atmosphere. Pure ethyl ester 3b (9.5 g, 98%) was obtained after column
1
chromatography on silica gel (nhexane/AcOEt, 1:0 to 0:1). M.p. 48 °C (MeOH/Et
2
O); H NMR
(
CDCl , 200 MHz): δ = 9.37 (s, 1H), 9.18 (s, 1H), 8.34 (d, J 8.4 Hz, 1H), 8.03 (d, J = 8.1 Hz, 1H),
3
1
3
7
.84 (t, J 7.0 Hz, 1H), 7.67 (t, J 7.0 Hz, 1H) 4.50 (q, J 7.3, 2H), 1.48 (t, J 7.3, 3H); C NMR
(CDCl
3
, 300 MHz): δ = 14.5, 61.5, 120.9, 125.3, 127.8, 128.4, 128.7, 132.4, 134.1, 147.0, 157.0,
+
1
5
66.7; MS (ESI): m/z (%): 202.1 [M+1] ; elemental analysis calcd (%) for C12
H
11NO : C, 71.63; H,
2
.51; N, 6.96; found C, 71.47; H, 5.63; N, 6.67.
.
Ethyl 1,2,3,4-tetrahydroisoquinolin-4-carboxylate AcOH (4). In a round bottomed flask Pt
2
O (1
atmosphere
for 15 min. and then the isoquinoline-4-carboxylate 3 (8.8 g, 44 mmol) was added. The mixture was
g, 4.41 mmol) was suspended in acetic acid (45 mL). The solution was stirred under H
2
stirred under H
2
atmosphere at 25 °C for 24 h. After filtration over a celite plug, the solvent was
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
removed under reduce pressure. The crude mixture was purified by column chromatography
1
(
CH
NMR (CDCl
s, 2H), 3.73, 3.55, 3.15 (ABX system, J 13.6, 4.8, 2.9 Hz, 3H), 1.97 (s, 3H), 1.24 (t, J 7.3 Hz, 3H);
13C NMR (CDCl
, 500 MHz): δ = 14.3, 22.2, 42.1, 44.6, 46.2, 61.6, 126.8, 127.0, 127.9, 129.8,
30.6, 134.2, 173.6, 176.1; IR (NaCl) νmax = 3400, 1728 cm ; MS (ESI): m/z (%): 206.2 [M+1] ;
2
Cl
2
/MeOH, from 80:1 to 40:1) giving pure compound 4 as a pale yellow oil (9.9 g, 85%). H
3
, 200 MHz): δ = 7.64 (s, 2H, exch.), 7.33-7.04 (m, 4H), 4.16 (q, J 7.3 Hz, 2H), 4.07
(
3
-1
+
1
elemental analysis calcd (%) for C14
H19NO
4
: C, 63.38; H, 7.22; N, 5.28; found C, 69.92; H, 7.56; N,
4
.87.
.
1
,2,3,4-Tetrahydroisoquinoline-2-carboxylic Acid HCl (1). Operating in a sealed tube,
compound 4 (11 g, 41 mmol) was suspended in HCl (6 N, 25 mL) and the mixture was heated at
1
20°C for 9 h. The solvent was removed under reduced pressure and the residue was crystallized
1
affording pure 1 (8.8 g, quantitative yield). Mp 237 °C (MeOH/Et
2
O); H NMR (DMSO-d
6
, 200
MHz) δ = 11.0-8.5 (brs, 3H), 7.21-7.45 (m, 4H), 4.23 (s, 2H), 4.10 (t, J 5.5 Hz, 1H), 3.14-3.61 (m,
1
3
2
1
H); C NMR (DMSO-d
73.2; IR (KBr) νmax 3030, 1725, 1609 cm ; MS (ESI): m/z (%): 178.1 [M+1] ; elemental analysis
: C, 56.21; H, 5.66; N 6.56; found C, 55.83; H, 5.85, N, 6.02.
6
, 50 MHz) δ = 41.1, 42.7, 44.2, 127.7, 128.2, 128.3, 129.6, 129.8, 130.3,
-1
+
calcd (%) for C10
H
12ClNO
2
Ethyl N-Boc-1,2,3,4-tetrahydroisoquinolin-4-carboxylate (5). In a three necked round bottom
flask equipped with a magnetic stirrer, nitrogen inlet, and thermometer, compound 4 (3 g, 11.3
mmol) was dissolved in dry CH
5.1 mL, 32.2 mmol) was added. After 10 min., di-tert-butyl dicarbonate (3.8 g, 17.5 mmol) was
slowly added and the stirring was continued for 2 h at 25 °C. The reaction mixture was washed with
2
Cl
2
(15 mL) and the solution was cooled to 0 °C. Triethylamine
(
KHSO
4
(pH = 3, 0.5 N, 5 mL) and then dried over Na
2
SO
4
. The solvent was removed under
1
reduced pressure and product 5 was obtained as a colourless oil (4.5 g, 96%). H NMR (CDCl
3
, 200
MHz): δ = 7.27-7.12 (m, 4H), 4.73, 4.48 (AB system, J 17.2 Hz, 2H), 4.24, 3.81, 3.59 (ABX
1
3
system, J 13.2, 5.1, 4.4 Hz, 3H), 4.17 (q, J 7.0 Hz, 2H), 1.48 (s, 9H), 1.25 (t, J 7.0, 3H); C NMR
CDCl , 200 MHz): δ = 14.4, 28.6 (x3), 44.1, 45.1, 46.0, 61.3, 80.3, 126.7 (x2), 127.7, 129.1, 131.9,
33.9, 154.9, 172.2; IR (NaCl) υmax = 1736, 1699 cm ; MS (ESI): m/z (%): 328.2 [M+Na] ;
elemental analysis calcd (%) for C17 : C, 66.86; H, 7.59; N, 4.59; found C, 66.58; H, 7.80; N,
.43.
(
3
-1
+
1
H23NO
4
4
Enzymatic resolution of compound 5 with Pronase. A 250 mL three necked round-bottomed
flask was immersed in a thermostatic oil bath, equipped with a magnetic stirrer and connected to an
automatic titration device through the necks (burette, electrode and temperature probe,
respectively). The bath temperature was set to 40°C and the titrator reservoir was filled with 0.5 M
NaOH. In the flask, TRIS base (75 mg, 0.6 mmol) was dissolved in water (120 mL) together with
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
Pronase (1.00 g), and the pH was adjusted to 8 with 1M NaOH. Compound 5 (1.50 g; 5.2 mmol)
was dissolved in DMSO (12 mL) and dropped to the solution under stirring. The titrator was
regulated for keeping the reaction mixture at pH = 8, and the data were collected. At fixed intervals,
®
aliquots were taken from the solution and analysed by HPLC [Chiralcel OD column (250 x 4.6
mm), 25 °C, nhexane/2-propanol, 9:1, flow 1 mL/min; 210 nm.]. When e.e. of substrate 5 was
found to be over 99% (usually around 65-70% of conversion, after 6-8 days), the titration was
switched off and the stirring was interrupted. The reaction mixture was extracted twice with Et
The organic layers were brought together and dried on anhydrous Na SO . The solvent was
removed leaving an oily residue corresponding to ester (-)-5 (432 mg, 29%, e.e. 99%): [α]
6.4 (c = 1 in CHCl ).
The aqueous layer was acidified to pH 3 with 6M HCl and extracted twice with Et
layers were brought together, dried over Na SO and the solvent was removed. An oily residue was
2
O.
2
4
25
D
= -
3
3
2
O. The organic
2
4
obtained corresponding to the acid (+)-6 (823 mg, 60 %, e.e. ≈ 48%). NMR data of compound 6
are in agreement with those reported in the literature.[20]
Recycle of the enantio-enriched acid (+) 6 to racemic 5. Acid (+)-6 (e.e. ≈ 48%; 1 g, 3.6 mmol)
was dissolved in dry CH
mmol), EtOH (1.5 mL) were added and then DCC (819 mg, 4 mmol) dissolved in CH
was slowly dropped. The reaction was stirred at 25 °C for 1 h. The solvent was removed. The crude
reaction mixture was taken up with Et O. The solid was filtered and the crude was purified by
2
Cl
2
(33 mL) and the solution was cooled to 0 °C. DMAP (44 mg, 0.36
2
Cl (5 mL)
2
2
column chromatography affording pure ester (+)-5 (618 mg, 56%).
The enantio-enriched ester (+)-5 (600 mg, 2.08 mmol) was dissolved in CH
2
Cl (3 mL) and DBU
2
(
300 µ L, 2.01 mmol) was added. The solution was stirred at 25 °C for 3 h. The reaction was
monitored by HPLC [Chiracel OD; nhexane/2-PrOH, 9:1; 1 mL/min; 210 nm]. The organic layer
was washed with 0.1 N HCl solution (3 mL). The aqueous layer was extracted with Et O (3 x 3
affording racemic ester 5 in
2
mL). The organic layers were brought together and dried over Na
2
SO
4
quantitative yield.
.
Ethyl (R)-1,2,3,4-tetrahydroisoquinolin-4-carboxylates CF
3
CO H (-)4. Enantiopure ester (-)-5
2
(140 mg, 0.46 mmol) was dissolved in dry CH
2
Cl
2
(5 mL) and the solution was cooled at 0 °C.
CF CO H (0.5 mL) was added and the solution was stirred for 1.5 h at 25 °C. After solvent
3
2
.
25
evaporation, (-)-4 CF
3
CO
2
H (147.3 mg, quantitative yield) was isolated: [α]
D
= -42.3 (c = 1 in
1
MeOH). H NMR (MeOD, 200 MHz) δ = 7.55-7.24 (m, 4H), 4.38 (s, 2H), 4.23 (q, J 7.2 Hz, 2H),
4
.19, 3.90, 3.52 (ABX system, J 13.1, 5.2, 3.6 Hz, 3H), 1.27 (t, J 7.2 Hz, 3H); elemental analysis
calcd (%) for C14
H
16
F
3
NO
4
: C, 52.67; H, 5.05; N, 4.39; found C, 52.19; H, 5.48; N, 3.91.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
Fmoc-(L)-Ala-(R)--TIC-OEt (-)-8. Fmoc-L-Ala (88.7 mg, 0.28 mmol) was dissolved in CH
3 mL) and DMF (0.5 mL), the solution was cooled to 0 °C and HOBT (42.3 mg, 0.31 mmol) and
EDC (60.1 mg, 0.31 mmol) were added. The mixture was stirred at 0°C for 1h. Then compound (-)-
(100 mg, 0.31 mmol), dissolved in CH Cl (0.5 mL) and DIPEA (97.5µL 0.56 mmol) were added.
The reaction mixture was stirred at room temperature for 8h. A saturated solution of NaHCO (4
mL) was added. The aqueous layer was separated and the organic layer was washed first with a
saturated solution of NH Cl (4 mL) and than with a saturated solution of NaCl (4 mL). The organic
layer was dried over Na SO . The solvent was removed under reduced pressure and he crude
2
Cl
2
(
4
2
2
3
4
2
4
mixture was purified by column chromatography (nhexane/AcOEt, 7:3). Pure compound (-)-8 was
obtained as a colourless oil (130 mg, 90%). [α] 25
= -11 (c = 1 in CDCl
); H NMR (CDCl
1
, 300
D
3
3
MHz): δ (mixture of rotamers: 1:1) = 7.76 (d, J 7.7, 2H), 7.63-7.60 (m, 2H), 7.42-7.15 (m, 8H),
.06 (d, J 7.1 Hz, 0.5H), 5.95 (d, J 7.7 Hz, 0.5H), 5.19, 4.49 (AX system, J 17.6 Hz, 1H), 4.96-4.92
m, 0.5H), 4.78, 4.67 (AB system, J 17.7 Hz, 1H), 4.79-4.62 (m, 0.5H), 4.38-4.09 (m, 6H), 3.89-
.85 (m, 1H), 3.65-3.53 (m, 1H), 1.42+1.37 (d, J 6.7 Hz, 1.5+1.5H), 1.27+1.20 (t, J 7.2, 1.5+1.5H);
13C NMR (CDCl
, 300 MHz): δ = 14.4 (14.3), 19.6 (19.5), 42.2, 44.5, 44.7, 47.2, 47.5 (47.4), 61.9
61.5), 67.2 (67.1), 120.2, 129.5-125.4 (x12), 132.0 (130.8), 132.7 (132.1), 141.5, 144.2 (144.1),
6
(
3
3
(
+
1
44.3 (144.2), 155.8 (155.6), 171.1, 171.9; MS (ESI): m/z (%): 499.2 [M + 1] ; elemental analysis
: C, 72.27; H, 6.06; N, 5.62; found C, 71.96; H, 6.39; N, 5.39.
Fmoc-L-Ala-(R)--TIC-OH (-)-9. Dipeptide (-)-8 (110 mg, 0.22 mmol) was dissolved in DCE (7
mL). Me SnOH (320 mg, 1.8 mmol) was added and the mixture was stirred under reflux at 80°C for
4 h. The solution was than cooled at room temperature, diluted with CH Cl (25 mL) and washed
twice with a solution of KHSO (40 mL, 0.05 M) and then with a saturated solution of NaCl (40
mL). After the separation of the aqueous layer, the organic layer was dried over Na SO . The
solvent was removed under reduced pressure and the crude mixture was purified by column
calcd (%) for C30
H
30
N
2
O
5
3
2
2
2
4
2
4
chromatography (CH
2
Cl
2
/MeOH 20:1). Pure compound (-)-9 was isolated as a colourless oil (92
2
5
1
mg, 89%). [α]
D
= -15.4 (c = 1 in CDCl
3
); H NMR (CDCl
3
, 300 MHz): δ (mixture of rotamers:
1
7
:1) = 7.80-7.76 (m, 2H), 7.63-7.57 (m, 2H), 7.44-7.17 (m, 8H), 6.31 (d, J 8.1 Hz, 0.5H), 6.00 (d, J
.4 Hz, 0.5H), 5.27, 5.11 (AB system, J 17.6 Hz, 1H), 5.07-4.97 (m, 0.5H), 4.84, 4.70 (AB system,
J 16.6 Hz, 1H), 4.82-4.76 (m, 0.5H), 4.48-4.17 (m, 4H), 3.95-3.88 (m, 1H), 3.60-3.58 (m, 0.5H),
1
3
3
.36 (dd, J 13.3, 3.8, 0.5H), 3.10-2.10 (brs, 1H), 1.42+1.37 (d, J 6.7 Hz, 1.5+1.5H); C NMR
CDCl , 300 MHz): δ = 19.5, 42.6, 44.0 (44.4), 44.6 (44.8), 47.1, 44.4, 67.1 (67.2), 120.2, 125.2-
29.9 (x13), 131.4, 132.6, 141.2, 144.0 (m), 156.1, 173.3, 174.4; MS (ESI): m/z (%): 471.2
(
3
1
+
+
[
M+1] , 493.3 [M + Na] ; elemental analysis calcd (%) for C28
H
26
N
2
5
O : C, 71.47; H, 5.57; N, 5.95;
found C, 71.02; H, 5.89; N, 5.54.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
Fmoc-L-Ala--TIC--Ala-L-Val-OBn (+)-10. Dipeptide (-)-9 (80 mg, 0.17 mmol) was dissolved
in CH Cl (2 mL). The solution was cooled to 0 °C and HOBT (27.6 mg, 0.20 mmol) and EDC
39.1 mg, 0.20 mmol) were added. The mixture was stirred at 0°C for 1h. Boc--Ala-L-Val-OBn (7)
51.8 mg, 0.18 mmol) was dissolved in CH Cl (0.5 mL) and DIPEA (59 µL, 0.34 mmol) was
2
2
(
(
2
2
added. The reaction mixture was stirred at 25 °C for 8 h. The organic layer was washed with a
saturated solution of NaHCO (4 mL), saturated solutions of NH Cl (4 mL), and NaCl (4mL) and
then dried over Na SO . The solvent was removed under reduced pressure. The crude mixture was
purified by column chromatography (nhexane/AcOEt, 1:0 to 0:1) affording pure (+)-10 (94.4 mg,
3
4
2
4
7
6%). [α] 25
D
= +31.6 (c = 1 in CDCl
3
); NMR data are reported in Supporting Information; MS
: C, 70.67; H, 6.34; N,
+
(ESI): m/z (%): 731.3 [M+1] ; elemental analysis calcd (%) for C43
H
46
N
O
4 7
7
.67; found C, 70.41; H, 6.59; N, 7.50.
Synthesis of Fmoc-L-Ala--TIC--Ala-L-Val-OBn (+)-10 and (-)-11. Starting from racemic
4.
AcOH (500 mg, 1.88 mmol), the mixture of (+)-10 and (-)-11 was obtained using the same
reaction conditions described for the transformation of (-)-4 into (+)-10, without separation of
intermediates of the R and S series. A crude mixture of diastereoisomeric peptides were obtained in
4
3% overall yield (170 mg). It is possible to partially separate compounds (+)-10 and (-)-11 by
column chromatography (nhexane/AcOEt, 1:0 to 0:1).
(
-)-11: [α] 25
D
= -38.0 (c = 1 in CDCl
3
); NMR data are reported in Supporting Information. MS
: C, 70.67; H, 6.34; N,
+
(ESI): m/z (%): 731.3 [M+1] ; elemental analysis calcd (%) for C43
H
46
N
O
4 7
7
.67; found C, 70.35; H, 6.63; N, 7.47.
ROESY distances determination. ROESY data were collected with roesyph and roesyetgp pulse
program with 2048 data points in F2 and 512 data points in F1. For roesy spinlock, 200 ms time
was used (16 scans).
Theoretical calculations. The computational refinement of 10 and 10’ geometries was performed
using Amber14[19] by following the procedure described in the Amber documentation. Charge
parameterization of amino acids β-TIC and β-Ala and of protecting groups Fmoc, benzyloxy and
methoxy was done using the RED method to derive RESP charges which are both conformation and
orientation independent,[21] as described previously for other non-natural amino acids.[22] Distance
restraints for both 10 and 10’ conformations were generated from NOE-derived distances using the
“makeDIST_RST” command and peptide 10 was then constructed using tleap. The ff96 force field,
coupled with the Generalized Born model OBC(II) was chosen,[18] accordingly to suggestions
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
reported in the literature by us and by others. [23] Two simulated annealing (SA) cycles, starting
from different conformations generated by a random variation of φ, θ and ψ dihedrals, were
performed by following this protocol: a) geometry minimization by 500 cycles of steepest
descendent followed by 500 cycles of conjugated gradient; b) simulated annealing, for a total of
2
000000 steps (2 ns of simulation); the system was heated up to 600 K during the first 500 ps (heat
bath coupling constant tautp = 0.4 ps), then slowly re-cooled to 100 K from 501 to 180000 ps (tautp
4.0 ps) and finally to 0 K with tautp gradually reducing from 1.0 to 0.05 (see Supporting
Information for detailed input files). NMR restraints were initially applied with a weight of 0.1
from 0 to 300 ps), then with a weight of 1. Langevin dynamics (ntt = 3) with a collision frequency
=
(
-1
of 1.0 ps were used here, as well as in simulation described below.
24
REMD simulations of peptides 10 and m10 were performed accordingly to a protocol described
previously. [4a,23a] Briefly, simulations were conducted using the ff96 force field, coupled with the
Generalized Born model OBC(II). A temperature range spanning from 260.0 and 591.4 K was used
to run 12 replica of 400 ns each (see Supporting Information). Convergence was evaluated by
performing on the 300 K trajectory a cluster analysis every 50 ps. The simulation was considered
converged when the population of the two principal clusters differed less than 5% between the two
5
0 ns batches. Both simulations (10 and m10) were converged after 300 ns. Cluster analysis was
made by using cpptraj, sampling one of every four frames, using the average-linkage algorithm and
the pairwise mass-weighted RMSD on backbone heavy atoms.
Supporting Information
1
Further information on enzymatic resolution of ester 5. General synthesis of dipeptide 7. H NMR
studies at variable temperature for (-)-9. NMR data for peptides (+)-10 and (-)-11. Determination of
1
the stereochemistry of R--TIC in peptide (+)-10 through H NMR data. NMR spectra for
compounds 1,3,4,5,8-11. Additional computational details.
References
[
1] a) K. R. Rajashankar, S. Ramakumar, Protein Sci. 1996, 5, 932-946; b) L. E. Donate, S. D.
Rufino, L. H. Canard, T. L. Blundell, Protein Sci. 1996, 5, 2600-2616; c) G. D. Rose, L. M.
Gierasch, J. A. Smith, Adv. Protein Chem. 1985, 37, 1-109.
[
2] a) Y. Zhu, T. Li, D. Li, Y. Zhang, W. Xiong, J. Sun, Z. Tang, G. Chen, Amino Acids 2012, 42,
749-1755; b) A. Bax, J. Biomol. NMR 2012, 52, 211-232. c) P. Petersen, C. Lundegaard, T. N.
Petersen, PLoS One 2010, 5, e15079; d) B. Dasgupta, P. Chakrabarti, BMC Struct. Biol. 2008, 8,
1
3
9; e) A. J. P. Fuchs. A. Alix, Proteins Struct. Funct. Bioinform. 2005, 59, 828-839. f) H. Kaur, G.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
P. S. Raghava, Protein Sci. 2003, 12, 627-634; g) S. Datta, N. Shamala, A. Banerjee, A. Pramanik,
S. Bhattacharjya, P. Balaram J. Am. Chem. Soc. 1997, 119, 9246-9251; h) J. S. Richardson, Adv.
Protein Chem. 1981, 34, 167-339.
[
3] a) A. A. Fuller, D. Du, F. Liu, F.; J. E. Davoren, G. Bhabha, G. Kroon, D. A. Case, H. J. Dyson,
E. T. Powers, P. Wipf, M. Gruebele, J. W. Kelly, PNAS 2009, 106, 11067-11072; b) S. Oishi, K.
Miyamoto, A. Niida, M. Yamamoto, K. Ajito, H. Tamamura, A. Otaka, Y. Kuroda, A. Asai, N.
Fujii, Tetrahedron 2006, 62, 1416-1424; c) A. J. Souers, J. A. Ellman, Tetrahedron 2001, 57, 7431-
7
448; d) Y. J. Chung, B. R. Huck, L. A. Christianson, H. E. Stanger, S. Krauthauser, D. R. Powell,
S. H. Gellman, J. Am. Chem. Soc. 2000, 12, 3995-4004. e) B. R. Huck, J. D. Fisk, S. H. Gellman
Org. Lett. 2000, 2, 2607-2610; f) R. R. Gardner, G. B. Liang, S. H. Gellman J. Am. Chem. Soc.
1
999, 121, 1806-1816; g) Y. J. Chung, L. A. Christianson, H. E. Stanger, D. R. Powell, S. H.
Gellman, J. Am. Chem. Soc. 1998, 120, 10555-10556; h) P. Wipf, T. C. Henninger, S. J. Geib, J.
Org. Chem. 1998, 63, 6088-6089; i) C. N. Nesloney, J. W. Kelly, Bioorg. Med. Chem. 1996, 4, 739-
7
66; j) U. Nagai, K. Sato Tetrahedron Lett. 1995, 26, 647-650. k) J. P. Schneider, J. W. Kelly,
Chem. Rev. 1995, 95, 2169-2187; l) K. J. Tsang, H. Diaz, N. Graciani, J. W. Kelly, J. Am. Chem.
Soc. 1994, 116, 3988-4005; m) H. Diaz, K. J. Tsang, D. Choo, J. R. Espina, J. W. Kelly, J. Am.
Chem. Soc. 1993, 115, 3790-3791; n) H. Diaz, J. R. Espina J. W. Kelly, J. Am. Chem. Soc. 1992,
1
14, 8316-8318.
[
4] a) S. Pellegrino, N. Tonali, E. Erba, J. Kaffy, M. Taverna, A. Contini, M. Taylor, D. Allsop, M.
L. Gelmi, S. Ongeri, Chem. Sci 2016, 8, 1295-1302; b) E. Ko, A. Raghuraman, L. M. Perez, T. R.
Ioerger, K. J. Burgess, J. Am. Chem. Soc. 2013, 135, 167-173.
[
[
5] P. Chakrabarti, R. Bhattacharyya, Prog. Biophys. Mol. Biol. 2007, 95, 83-137.
6] a) S. Pellegrino, A. Bonetti, F. Clerici, A. Contini, A. Moretto, R. Soave, M. L. Gelmi, J. Org.
Chem., 2015, 80, 5507-5516; b) A. Ruffoni, A. Contini, R. Soave, L. Lo Presti, I. Esposto, I.
Maffucci, D. Nava, S. Pellegrino, M. L. Gelmi, F. Clerici, RSC Adv, 2015, 5, 32643-32656; c) S.
Pellegrino, A. Contini, M. L. Gelmi, L. Lo Presti, R. Soave, E. Erba, J. Org. Chem., 2014, 79,
3
094-3102; d) A. Bonetti, F. Clerici, F. Foschi, D. Nava, S. Pellegrino, M. Penso, R. Soave, M. L.
Gelmi, Eur. J. Org. Chem., 2014, 15, 3203-3209; e) S. Pellegrino, A. Contini, F. Clerici, A. Gori, D.
Nava, M. L. Gelmi, Chemistry-Eur. J. 2012, 18, 8705-8715; f) M. Penso, F. Foschi, S.Pellegrino,
A. Testa, M. L. Gelmi J. Org. Chem. 2012, 77, 3454-3461; g) A. Ruffoni, A. Casoni, S. Pellegrino,
M. L. Gelmi, R. Soave, F. Clerici, Tetrahedron, 2012, 68, 1951-1962; h) F. Gassa, A. Contini, G.
Fontana, S. Pellegrino, M. L. Gelmi, J. Org. Chem., 2010, 75, 7099-7106; i) S. Pellegrino, F.
Clerici, M. L. Gelmi, Tetrahedron 2008, 64, 5657-5665; j) M. L. Gelmi, C. Cattaneo, S. Pellegrino,
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
F. Clerici, M. Montali, C. Martini J. Org. Chem. 2007, 72, 9811-9814; k) C. Cabrele, F. Clerici, M.
L. Gelmi, R. Gandolfi, F. Molinari, S. Pellegrino, Tetrahedron, 2006, 62, 3502-3508.
[
7] a) A. Gori, C.-I A. Wang, P. J. Harvey, K. J. Rosengren, R. F. Bhola, M. L. Gelmi, R. Longhi,
M. J. Christie, R. J. Lewis, P. F. Alewood, A. Brust, Angew. Chem. Int. Ed. 2015, 54, 1361-1364; i)
S. Pellegrino, N. Ferri, N. Colombo, E. Cremona, A. Corsini, R. Fanelli, M. L. Gelmi, C. Cabrele,
Bioorg. Med. Chem. Lett., 2009, 19, 6298–6302;
[
8] G. Valle, W. M. Kazmierski, M. Crisma, G. M. Bonora, C. Toniolo, V. J. Hruby Int. J. Pept.
Prot. Resch. 1992, 40, 222-232.
[
[
9] Bernabeu, M. C.; Diaz, J. L.; Jimenez, O.; Lavilla, R. Synthetic Comm. 2004, 1, 137-149.
10] a) P. Hoyos, V. Pace, M. J. Hernaiz, A. R. Alcantara, A. R. Curr. Green Chem. 2014, 1, 155-
1
81; b) A. Wells, H.-P. Meyer, Chem. Cat. Chem. 2014, 6, 918-920.
[
11] a) A. Liljeblad, L. T. Kanerva, Tetrahedron 2006, 62, 5831-5854; b) A. Berkessel, I.
Jurkiewicz, R. Mohan, Chem. Cat. Chem. 2011, 3, 319-330; c) A. Liljeblad, H.-M. Kavenius, P.
Tähtinen, L. T. Kanerva, Tetrahedron: Asymmetry 2007, 18, 181-191.
[
12] a) C. Cabrele, T.A. Martinek, O. Reiser, Ł. Berlicki, J. Med. Chem. 2014, 57, 9718-9739; b) M.
A. Gelman, S. H. Gellman in Enantioselective Synthesis of β-Amino Acids, Second Edition (eds E.
Juaristi and V. A. Soloshonok), John Wiley & Sons, Inc., Hoboken, NJ, USA.
doi: 10.1002/0471698482.ch22
[
13] J. R. Martinelli, D. A. Watson, D. M. M. Freckmann, T. E. Barder, S. L. Buchwald, J. Org.
Chem., 2008, 73, 7102-7107.
[
[
[
[
14] S. Shotani, T. Hori, K. Mitsuhashi, Chem. Pharm. Bull. 1967, 15, 88-93.
15] D. Tessaro, Cascade Biocatalysis, 2014, 179-198.
16] P. D'Arrigo, L. Cerioli, S. Servi, F. Viani, D. Tessaro, Catal. Sci. Techol. 2012, 2, 1606-1616.
17] K. C. Nicolaou, A. A. Estrada, M. Zak, S. H. Lee, B. S. Safina, Angew. Chem. Int. Ed. 2005,
4
4, 1378-1382.
[
[
18] A. Onufriev, D. Bashford, D. A. Case, Proteins 2004, 55, 383–394.
19] D. A. Case, V. Babin, J. T. Berryman, R. M. Betz, Q. Cai, D. S. Cerutti, T. E. Cheatham, T. A.
Darden, R. E. Duke, H. Gohlke, A. W. Goetz, S. Gusarov, N. Homeyer, P. Janowski, J. Kaus, I.
Kolossvꢀry, A. Kovalenko, T. S. Lee, S. LeGrand, T. Luchko, R. Luo, B. Madej, K. M. Merz, F.
Paesani, D. R. Roe, A. Roitberg, C. Sagui, R. Salomon-Ferrer, G. Seabra, C. L. Simmerling, W.
Smith, J. Swails, Walker, J. Wang, R. M. Wolf, X. Wu, P. A. Kollman, Amber 14. University of
California, San Francisco, 2014.
[
20] A. Bonetti, E. Beccalli, A. Caselli, F. Clerici, S. Pellegrino, M. L. Gelmi, Chem. Eur. J. 2015,
1, 1692-1703.
2
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
[
21] F.-Y. Dupradeau, A. Pigache, T. Zaffran, C. Savineau, R. Lelong, N. Grivel, D. Lelong, W.
Rosanski, P. Cieplak, Phys. Chem. Chem. Phys. 2010, 12, 7821-7839.
22] a) I. Maffucci, S. Pellegrino, J. Clayden, A. Contini, J. Phys. Chem. B 2015, 119, 1350-1361.
b) I. Maffucci, J. Clayden, A. Contini, J. Phys. Chem. B 2015, 119, 14003-14013.
23] a) I. Maffucci, A. Contini, J. Chem. Theory Comput., 2016, 12, 714-727. b) M. K. Robinson,
[
[
J. I. Monroe, M. S. Shell, J. Chem. Theory Comput. 2016, 12, 5631-5642. c) M. S. Shell, R.
Ritterson, K. A. Dill, J. Phys. Chem. B. 2008, 112, 6878-6886.
[
24] Y. Sugita, Y. Okamoto Chem. Phys. Lett. 1999, 314, 141-151.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
Scheme 1. Synthesis of tetrahydroisoquinoline-4-carboxylic acid derivatives
Figure 1. Screening of enzymatic activities on the substrate 5. A negative E value indicates the
enzyme preference towards the opposite enantiomer.
Scheme 2. Enzymatic resolution of -TIC. a) 40 °C, pH 8, 10% DMSO; b) DMAP, CH
°C, then DCC; c) DBU, CH Cl
2
Cl2, EtOH,
0
2
2
Scheme 3. Synthesis of diastereoisomeric peptides (+)-10 and (-)-11. a) HOBT/EDC, CH
2
Cl
2
/DMF,
0
°C, then DIPEA, 25 °C, 8 h. b) Me
3
2
SnOH, CH Cl
2
, 50 °C, 10 h. c) HOBT/EDC, DMF, 25 °C, 2 h.
1
Figure 2. Plots of the NHs in the H NMR spectra of (+)-10 as a function of the increasing
amount of DMSO added to the CDCl solution (v/v) (peptide concentration: 20 mM). Black: 10;
3
Red: 10’.
Figure 3. Proposed conformation for rotamers 10 (a) and 10’ (b). Zoom of the -alanine moiety for
0 (c) and 10’ (d). Dotted lines: H-bond; Arrow: spatial proximities.
1
Figure 4. Zoom of significant ROEs to define turn structures of peptide (+)-10 (CDCl
black: 10; °red: 10’. A), B), D: 200 ms, C) 100 ms.
3
, 20 mM):
*
Figure 5. Structures of 10 and 10’ obtained by restrained MD simulated annealing
Figure 6. Comparison of most representative conformations of peptide 10 (A) and the model
peptide m10 (B), obtained from REMD simulations.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
Table 1. NOE Distances Used to Generate Restraints for Simulated Annealing Calculations.
1
0
10’
Distance (Å) [
a]
Atom2 Distance (Å) ©
Atom1
Atom2
Atom1
MeAla
H1βTIC
H1βTIC
H1βTIC
H8βTIC
H3’βTIC
H4βTIC
H4βTIC
NHβAla
NHVal
3.03±0.22
2.49±0.18
2.58±0.04
2.56±0.04
2.68±0.04
3.2±0.12
MeAla
CHAla
H4βTIC
H4βTIC
H4βTIC
H2βAla
H3βTIC
H3βTIC
H3βTIC
H5βTIC
3.03±0.22
2.13±0.04
2.18±0.03
2.47±0.05
CHAla
H8βTIC
H1’βTIC
H1’βTIC
H3’βTIC
H3βTIC
H3βTIC
H3βTIC
H4βTIC
NHβAla 2.76±0.13
NHVal
2.66±0.02
2.50±0.02
2.91±0.10
2.85±0.11
2.32±0.03
2.46±0.03
2.63±0.03
H5βTIC
H4βTIC
NHβAla
NHβAla
H2βAla
[a]Average of two determinations
Table 2. Principal geometrical parameters of 10 and 10’ geometries from restrained SA calculations
Dihedral
Dihedral value (deg.)
1
0
10’
φ
i
-63.9
140.5
-109.7
68.0
-143.4
145.5
-112.5
71.5
ψ
i
φ
i+1
i+1
i+1
i+2
i+2
i+2
θ
ψ
84.1
82.6
φ
90.1
171.2
178.8
83.5
θ
65.9
ψ
-103.5
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201701045
Text for Table of Contents
eye-catching headline: Flexibility and adaptability
2
Tetrahydroisoquinoline-4-carboxylic acid, a constrained β -amino acid named -TIC, was for the
first time prepared in enantiopure form and used for the synthesis of the model Fmoc-L-Ala--TIC-

-Ala-L-Val-OBn tetrapeptides. R--TIC/-Ala construct represents the first example of a flexible
turn mimetic containing a cyclic and an acyclic -amino acid, that could promote a hairpin
formation
FULL PAPER
Tetrahydroisoquinoline-4-carboxylic acid,
a constrained β -amino acid named -TIC,
Dr. Andrea Bonetti, Raffaella Bucci,
Prof. Francesca Clerici, Donatella Nava,
Dr. Sara Pellegrino, Dr. Davide Tessaro,
Prof. Maria Luisa Gelmi*
Turn region
2
stable to proteolysis
was for the first time prepared in
enantiopure form and used for the
synthesis of the model Fmoc-L-Ala--
TIC--Ala-L-Val-OBn tetrapeptides. R--
TIC/-Ala construct represents the first
example of a flexible turn mimetic
containing a cyclic and an acyclic -
amino acid, that could promote a hairpin
formation
non-covalent
interactions region
O
H
N
N
H
Page No. – Page No.
Me
O
iPr
Tandem
Tetrahydroisoquinoline-4-
O
H
N
N
carboxylic Acid/-Alanine as a New
Construct Able to Generate a Flexible
Turn
O
O
NH
N
O
CO2H
Me
O
iPr
N
O
O
H
NH
HN
b-TIC
Keywords


-amino acids
Enzymatic resolution
Turn mimetic
peptidomimetics
Molecular modelling
This article is protected by copyright. All rights reserved.