Synthesis and Structural Study of a New β-Turn Inducer Peptidomimetic Incorporating 1-Amino-1-Aminomethylcyclohexane ^†

Abstract

A new β-turn inducer peptidomimetic incorporating 1-amino-1-aminomethylcyclohexane (compound 3) was synthetized and structurally studied. It was established that there is a rapid conformational equilibrium between the expected β-turn inducer structure and other extended conformations.

1. Introduction

The search for structures that fold in the form of peptide turns, in which α-amino acids are totally or partially replaced by restrictions made up of β- or γ-amino acids, is of great interest. However, so are those structures in which α-amino acids are replaced by non-amino acid molecules and which also give rise to peptide turns. The interest, in both cases, is due to the limitations of α-peptides: low structural diversity, due to the limited number of proteinogenic α-amino acids, as well as remarkable conformational flexibility and low metabolic stability.

Some examples of peptidomimetics, including β-turn-inducing structures containing two urea fragments connected by the D-prolyl-cis-1,2-diaminocyclohexane dimer (D-Pro-DACH) [1], are listed in Figure 1. This cluster adopts a well-defined β-turn or reverse-turn conformation.

It is a turn-inducing cluster consisting of two urea fragments linked by a subunit that combines D-proline with a 1,2-diaminocyclohexane. This allows the two urea units to be close enough in space to form two hydrogen bonds between them, defining 10- and 12-membered rings, which stabilize the turn.

Our interest in the generation of β-turns with modulable polarity led us to propose the synthesis and structural study of β-turn 3 (Figure 2), analogous to those in Figure 1, but with modulable polarity.

The structural similarity of our compound 3 with compounds 1 and 2 (Figure 1) led us to think that it would also adopt a β-turn structure. The additional contribution in our case would be that the β-turn thus generated would foreseeably have the ability to modulate its polarity, depending on the substitution at the N atom of the piperidine ring.

The synthesis plan for our diurea 3 is shown in Scheme 1.

This plan involves carrying out the following synthetic operations:

• Coupling of α-aminonitrile 7 with commercial N-Cbz-L-proline 6, to form the strategic bond “a” of compound 5.
• Reduction of the cyano group to amine and subsequent coupling with ethyl isocyanate to form the strategic bond “b” to generate compound 4.
• Finally, removal of the Cbz protecting group and coupling with benzyl isocyanate, leading to the formation of the strategic bond “c” and thus generating the desired diurea 3.

2. Results and Discussion

2.1. Synthesis of Compound 3

According to our synthetic approach, as shown in Scheme 2, aminonitrile 7 was obtained from commercial cyclohexanone, following a procedure described in the literature. The coupling of this aminonitrile with commercial N-Cbz-L-proline (6) was then carried out, using isobutyl chloroformyate as coupling agent and NMO as base, leading to the expected aminonitrile 5, with a yield of 73%.

Next (Scheme 3), aminonitrile 5 was subjected to a reduction reaction of its cyano group to amine, using NaBH₄ as reducing agent in the presence of CoCl₂. The amine 9, obtained at 68%, was treated directly with ethyl isocyanate, leading, in 70% yield, to the expected compound 4.

Following our synthetic approach, the Cbz-protecting group was removed from 4 by Pd-C catalyzed hydrogenation. The amine 10 thus obtained was directly reacted with benzyl isocyanate, yielding the desired peptidomimetic 3 (48%).

2.2. Structural Study of Compound 3

Once we had synthesized peptidomimetic 3 (Figure 3), we decided to carry out a structural study to check whether it presents a β-turn structure in solution.

The study was started by acquiring NMR experiments (¹H, ¹³C and TOCSY) leading to the assignment of the signals of the compound (Figure 4 and

Table 1. Assignment of the ¹H NMR signals of compound 3 (+25 °C, CDCl₃, 500 MHz).

Atom	δ1H (ppm)
Bn: Ph	7.22–7.42 (m, 5H)
Bn: CH2	4.31–4.51 (mAB, 2H)
Bn: HN	4.85 (t, J = 5.7 Hz, 1H, NH)
Pro: α	4.22 (dd, J = 7.8, 3.7 Hz, 1H)
Pro: β	1.81–2.23 (m, 2H)
Pro: H	1.81–2.23 (m, 2H)
Pro: δ1	3.32 (q, J = 7.3 Hz, 1H)
Pro: δ2	3.43–3.63 (m, 1H)
Cy: β	1.29–1.61 (m, 2H)
Cy: β	1.81–2.23 (m, 2H)
Cy: γ	1.81–2.23 (m, 2H)
Cy: γ	1.81–2.23 (m, 2H)
Cy: δ	1.81–2.23 (m, 2H)
Cy: ɛ	3.43–3.63 (m, 2H)
Cy: α-HN	5.87 (s, 1H, NH)
Cy: ɛ-HN	5.82 (s, 1H, NH)
UREA: Et-CH2	3.12 (dt, J = 11.7, 6.0 Hz, 2H)
UREA: Et-CH3	1.00 (t, J = 7.2 Hz, 3H)
UREA: Et-HN	4.61–4.74 (m, 1H)

).

Having carried out the assignment of the ¹H NMR signals of compound 3, we turned our attention to the key experiments, in order to determine the existence of a β-turn structure.

• Serial dilution

In order to find out whether our compound 3 aggregates in solution, which could interfere with studies of its possible intramolecular interactions, ¹H NMR experiments were carried out in CDCl₃ at +25 °C and at different concentrations (1, 2, 5, 10 and 20 mM). These experiments are shown in Figure 5, and from them it can be concluded that:

• There is little change with concentration.
• Between 1 and 5 mM there are hardly any changes, so it can be concluded that at these concentrations there is no aggregation and the molecules are found as monomers in solution.
• When going from 5 to 10 and 20 mM, some signal shifts, which may mean that above 10 mM aggregation does occur.

Therefore, temperature coefficients (T-coef) were determined with 2 mM concentration samples and 2D spectra (TOCSY and NOESY) with 5 mM concentration samples.

• Temperature coefficients (T-coef)

To evaluate the existence of intramolecular hydrogen bonds that can stabilize a β-turn secondary structure for our compound 3, we now determine the T-coef (−Δδ/ΔT). For this purpose, ¹H NMR experiments were performed using water- and acid-free CDCl₃, at a concentration of 2 mM and in a temperature range from −10 to 45 °C (with 5 °C increments). The results obtained are shown in Figure 6 and

Table 2. Temperature coefficients of the H^N protons of compound 3.

Name	δ-max (+45C)	δ-min (−10C)	range T/°C	T-coef (ppb/K)
H4	4.49	4.63	55	−2.5
H3	4.69	4.77	55	−1.5
H2	5.67	5.86	55	−3.5
H1	5.80	5.69	55	+2.0

.

The T-coefs (−Δδ/ΔT) of the four nitrogen-bonded protons (H^N) in our molecule are listed in Table 2 and Figure 7. The values of the temperature coefficients are less than 5 ppb/K for the four protons studied, which indicates that all of them may be either forming hydrogen bonds or not exposed to the solvent.

• Titration with DMSO-d₆

In order to shed light on the possibility that the HNs of our compound are forming intramolecular hydrogen bonds, we decided to perform a titration experiment with DMSO-d6. This experiment will allow us to observe the interaction of the HN protons of our compound with DMSO-d6 and, therefore, how engaged they are in the formation of intramolecular hydrogen bonds [2].

For this, to a 2 mM solution of 3 in CDCl₃ (600 μL), amounts of DMSO-d6 were gradually added, making additions of 5 μL, 10 μL and finally 20 μL. After each addition, the ¹H NMR spectrum was measured to see the change in HN proton shift (Figure 8).

As can be seen in Figure 8, which plots the chemical shift value of the HN protons of 3 with respect to the ratio of DMSO-d6, as DMSO-d6 is added, H3 undergoes a large shift at low field, whereas for H4 and H1 the shift is much milder and practically non-existent for H2. These results suggest that H2 is less accessible to hydrogen bond formation with DMSO-d6, which would support it being part of intramolecular hydrogen bonds. The large variation in the δ of H3 with the addition of DMSO-d6 is consistent with this HN being exposed or accessible to the solvent and therefore also accessible to DMSO-d6. The H₁ and H₄ protons are in an intermediate situation, possibly also forming intramolecular hydrogen bonds.

• NOESY

To confirm the above indications of the presence of intramolecular hydrogen bonds in our compound and thus of a β-turn secondary structure, 2D-NOESY spectra were analyzed.

The experiments were performed in CDCl₃ and the parameters used were: 500 Mz, a mixing time of 600 ms, a concentration of 5 mM and +25 °C. The results are shown in Figure 9 and Figure 10.

No long-range (LR = “Long Range”) contacts are observed between the two ends of the hairpin (residues Bn and Et), which could justify the existence of a β-turn. The only NOE contacts observed are short-range:

• H2/H4 and H2/CH2 NOE of the ethyl chain.
• Sequential H2/CH2 exocyclic NOE of the Cy unit.

Chemical exchange (EXSY) is also observed between HNs and solvent water (δ = 1.6 ppm), except H1.

Given the results obtained, which do not support the presence of a β-turn conformation, it was decided to repeat the NOESY experiment by lowering the temperature to 10 ºC in order to reduce the exchange rate of the HNs, and to reduce the rate of conformational exchange between the β-turns (if any) and other extended conformations. The experiment (Figure 11 and Figure 12) was successful and allowed confirmation of the β-turn (“hairpin”) conformation. This confirmation is based on the observation of long-range contacts, namely NOE, which were observed between:

• The aromatic protons of phenyl/CH₃ of ethyl urea.
• Benzyl CH2/CH3 of ethyl urea.
• Benzyl CH2/H4.

Comparison of NOE at −10 °C and +25 °C suggests a rapid conformational equilibrium between the β-turn and other extended conformations.