Synthesis and structure–activity relationship studies of IgG-binding peptides focused on the C-terminal histidine residue

A novel IgG-binding peptide with a strong affinity to human IgG₁ was discovered and efficiently synthesized overcoming serious synthesis problems.

Abstract

Currently, IgG-binding peptides are widely utilized as a research tool, as molecules that guide substrates to the Fc site for site-selective antibody modification, leading to preparation of a homogeneous antibody–drug conjugate. In this study, a structure–activity relationship study of an IgG-binding peptide, 15-IgBP, that is focused on its C-terminal His residue was performed in an attempt to create more potent peptides. A peptide with a substitution of His17 by 2-pyridylalanine (2-Pya) showed a good binding affinity (15-His17(2-Pya), K_d = 75.7 nM). In combination with a previous result, we obtained 15-Lys8Leu/His17(2-Pya)-OH that showed a potent binding affinity (K_d = 2.48 nM) and avoided three synthetic problems concerning the p-hydroxybenzyl amidation at the C-terminus, the difficulty associated with coupling at the His7 position and the racemization of 2-Pya.

Introduction

Antibody–drug conjugates (ADCs) are attractive biomedicines that can selectively deliver drugs to target cells.1 This kind of compound with selective cytotoxicity can be used to reduce side effects in cancer chemotherapy. In the classical method of ADC preparation, bivalent crosslinkers were used to modify lysine residues present on the antibody surface.2 However, due to the nonspecific nature of this reaction, it is possible that the modification will be localized on the antigen binding (Fab) site, which influences antigen recognition of the original antibody. In addition, it is generally believed that heterogeneous ADCs have narrow therapeutic indices, because a difference in the drug–antibody ratio (DAR) and modification sites of ADCs negatively affects in vivo pharmacokinetic, efficacy and safety profiles.3 Therefore, site-specific antibody modification methods have been developed to obtain homogeneously modified ADCs. These methods have been developed on the basis of genetic engineering techniques4 and chemical methods5 and among them, an attractive method using peptides that selectively bind to the Fc site of an antibody has emerged in recent years.6,7

In 2019, Ito et al. developed a method involving chemical conjugation by affinity peptides (CCAP) for the Fc site-selective crosslinking using the IgG-binding peptide (1) (K_d = 225 nM),6 that binds to the region overlapping with CH2 and CH3 at the Fc site (PDB: ; 1DN2 (ref. 8) and ; 6IQH (ref. 6)). This method first reacts peptide 1 at the ε-amino group of Lys at its 8th residue with one site of a bivalent crosslinker such as disuccinimidyl glutarate (DSG) to afford a peptide with a reactive succinimidyl ester on the other side of the linker. Subsequently, upon reacting this peptide with an antibody, a crosslinking reaction occurs selectively at the antibody Fc site, influenced by a proximity effect. This reaction involves Lys248 of the antibody, and is thus near to the binding site and it is possible to prepare a homogeneous ADC retaining its binding affinity to the target antigen. Since the application of IgG-binding peptides involves not only chemical conjugation for the ADC preparation but also noncovalent-type use,9 development of potent and characteristic peptides can lead to a wide variety of applications.

Previously, we performed a structure–activity relationship (SAR) study of peptide 1, a 17mer with K_d = 225 nM, to identify the structural requirements for its binding to antibodies (Fig. 1).10 We found that the shortened peptide 15-IgBP, a 15mer with K_d = 267 nM (Fig. 1), retained the binding affinity in spite of the deletion of two N-terminal residues, Gly1 and Pro2. In further derivatization focusing on Lys8, its replacement by Leu greatly improved the antibody binding affinity (15-Lys8Leu, K_d = 8.19 nM), although this peptide cannot be used in the CCAP method due to the deletion of the amino group in the crosslinking site. On the other hand, the substitution or truncation of C-terminal residues (including His17) shows decreased antibody binding affinity. This means that C-terminal residues were important for the antibody binding. Therefore, we performed an SAR study and kinetics analysis focusing on one of the important residues, the C-terminal His17, with the aim of obtaining a more potent IgG-binding peptide. We also focused during the structure derivatization on the synthetic problems of 15-IgBP derivatives which include by-product generation and racemization, and finally produced a potent IgG-binding peptide which can overcome the synthetic problems.

Fig. 1. Structure derivatization of IgG-binding peptide performed in the previous study.10.

Results and discussion

Synthesis of peptide derivatives of 15-IgBP

The standard Fmoc (9-fluorenylmethyloxycarbonyl)-based solid-phase peptide synthesis (SPPS) method11 was applied to the synthesis of peptide derivatives (see the Experimental section for details). In the synthesis of 15-IgBP, the purity of the crude product before disulfide bond formation was low. The crude product was prepared from the resin under the cleavage and deprotection conditions of TFA : TIS : H₂O : 1,2-EDT = 40 : 1 : 1 : 0.4, (peak area of HPLC = 46%, Table 1, entry 1 and Fig. 2). From the analysis of the by-products, it could be seen that two kinds of undesired peptides had been produced. One is a peptide with a molecular weight (MW) 106 Da higher than that of 15-IgBP (MW + 106, peak c, peak area = 25% in Fig. 2 left), and the other is a truncated peptide, peptide (8–17), formed by the termination of chain elongation at Lys8 (peak b, peak area = 11%), suggesting difficulty associated with the coupling reaction at His7. Regarding the formation of the MW + 106 by-product, a similar side reaction using the Rink amide resin involving p-hydroxybenzyl amidation was reported in 2006 by Stathopoulos et al.12 This p-hydroxybenzyl by-product was derived from the incomplete decomposition of the Rink amide linker during TFA cleavage. This side reaction had been reported to be sequence-independent, but in our study, we found that the C-terminal amino acid residue significantly affects the production of the MW + 106 product. When the C-terminal His was replaced by an Ala residue (15-His17Ala), the MW + 106 formation was suppressed to a peak area of 12% (Table 1, entry 2) compared to 15-IgBP, whose corresponding peak area was 25%. With replaced aromatic residues, 15-His17Phe and 15-His17Trp, moderately suppressed the by-product formation (entries 3 and 4, both with peak area = 17%). In peptide 15-IgBP-Gly which has an additional Gly residue at the C-terminus of 15-IgBP, the production of the MW + 106 compound is significantly suppressed (entry 5, peak area = 6%). These results suggest that particular care is required when a His residue is placed at the C-terminus. In the previous study,12 it was also reported that the formation of the by-product can be suppressed in the presence of 1,3-dimethoxybenzene (1,3-DMB) which acts as a scavenger at the cleavage step. In our experiment, this greatly improved the HPLC purity of 15-IgBP (entry 6, peak area = 62%) with the reduction of the production of the MW + 106 compound (peak area = 4%) under the cleavage conditions of TFA : TIS : 1,3-DMB = 40 : 1 : 2 (Table 1, entry 6).

Table 1. HPLC purity of the crude peptide before disulfide formation.

Entry	Peptide	Cleavage a	Product (%)	By-product (%)
				MW + 106	8–17
1	15-IgBP	A	46	25	11
2	15-His17Ala	A	73	12	7
3	15-His17Phe	A	60	17	12
4	15-His17Trp	A	48	17	22
5	15-IgBP-Gly	A	61	6	19
6	15-IgBP	B	62	4	14
7	15-His17(2-Pya) b	A	53	30	1
8	15-His17(2-Pya) b	B	60	11	1
9	15-His17(2-Pya)-OH c	B	77	0	0

^aCleavage conditions: A = TFA : TIS : H₂O : 1,2-EDT = 40 : 1 : 1 : 0.4 for 1 h, B = TFA : TIPS : 1,3-DMB = 40 : 1 : 2 (ref. 12) for 1 h.

^bPeak areas including the epimerized peptide at the 2-Pya residue.

^cC-terminal carboxylic acid peptide synthesized by using the Cl-Trt resin.

Fig. 2. Crude HPLC chromatograms of 15-IgBP and 15-His17(2-Pya)-OH after cleavage from the resin. Peak a: desired peptide, peak b: peptide (8–17), peak c: MW + 106.

A by-product of peptide (8–17) was significantly suppressed (Table 1, entry 7, peak area of HPLC = 1%) in the synthesis of a peptide substituted with 2-pyridylalanine (2-Pya), 15-His17(2-Pya), which has a potent binding affinity, as discussed below. While position 17 is remote from position 8 in the primary structure of the peptide, substitution of His17 by 2-Pya affects the coupling efficiency at the His7 position, possibly through the secondary structure of the protected peptide on the resin. The exact reason is however unclear, but as a result, we successfully suppressed the production of peptide (8–17). In the synthesis of 15-His17(2-Pya), the formation of the MW + 106 compound was suppressed in the presence of the scavenger, 1,3-DMB (entry 7 vs. entry 8).

Racemization of a 2-pyridylalanine derivative (2-Pya)

In the synthesis of 15-His17(2-Pya), two by-products were successfully suppressed as described above, but other synthetic problems with the racemization at 2-Pya appeared during the coupling reaction. In the crude analysis of 15-His17(2-Pya), two products with the same molecular weight were observed. In the case of 15-His17(2-Pya), these peaks could be separated and purified by RP-HPLC after disulfide bond formation. When d-2-Pya was reacted at the first residue, the abundances of the two peaks were inverted (Fig. S1 in the ESI†), indicating that the racemization occurred during the coupling reaction. Two racemization mechanisms, shown in Fig. 3, can be considered. In the first one, similar to His,13 the nitrogen atom at the δ-position, which corresponds to the N^π atom in the His residue, extracts an α-proton, resulting in an sp² α-carbon (route A). The second one involves a nucleophilic attack on the activated ester by the coupling reagent (route B). Route B is perhaps more likely, because the reaction mixture turned deep red after addition of the coupling reagent. This suggests that an intermediate with an extended conjugated system has been produced during the reaction. Products of this sort, involving an N-acylpyridyl structure, have been reported to have a red color.14 Although this problem cannot be resolved by the consideration of reaction conditions such as the coupling reagent and solvent used in this study (data not shown), it can be avoided by using a Cl-Trt(2-Cl) resin in the synthesis of 15-His17(2-Pya)-OH, because the first reaction then excludes the activation step of the carboxylic acid (see the Experimental section). This result also suggests that racemization progressed in the activation step of carboxylic acid. Finally, 15-His17(2-Pya)-OH (Fig. 2 and Table 1, entry 9) was the peptide which overcame all three synthetic problems, formation of the MW + 106 adduct, coupling difficulty at the His7 position and racemization of 2-Pya, resulting in a 77% pure crude peptide.

Fig. 3. Possible mechanisms of racemization at 2-pyridylalanine. OAt = 1-hydroxy-7-azabenzotriazole ester.

Evaluation of the binding affinity of IgG-binding peptides

The binding affinity of the synthetic peptides was evaluated by a surface plasmon resonance (SPR) assay using Herceptin (human IgG₁). The values of the dissociation constant (K_d), the association rate constant (k_on) and the dissociation rate constant (k_off) were calculated by curve fitting on the basis of a Langmuir 1 : 1 binding model or steady state analysis. The typical sensorgrams are shown in Fig. S2 in the ESI.† 15

As shown in Table 2, 15-His17Ala shows a higher K_d value (841 nM) than 15-IgBP, whose K_d = 267 nM, the same value as was obtained in the previous Ala scan study of peptide 1.10 Both 15-His17Phe and 15-His17Trp have decreased binding affinity (K_d = 506 and 1220 nM, respectively) even though these substitutions retain the aromatic properties of the original His residue. This result suggests that the His side chain is important for the binding affinity. Then, the five peptides with a basic side chain at a position similar to that of His, i.e. a primary amine (Dab and Orn) or pyridine (2-Pya, 3-Pya and 4-Pya), were synthesized. 15-His17Dap and 15-His17Orn showed a weak binding affinity (K_d = 1020 and 710 nM, respectively). This suggested that ionized nitrogen atoms at δ and ε-positions were not suitable in the antibody binding, since primary amine (pK_aH = 10.34)16 exists as a protonated form at neutral pH while imidazole (pK_aH = 6.45)16 does not. On the other hand, 15-His17(2-Pya) with a basic aromatic pyridine ring structure (pK_aH = 5.25)17 shows a 4-fold improvement in the binding affinity (K_d = 75.7 nM), while 15-His17(3-Pya) and 15-His17(4-Pya) have lower binding affinities (K_d = 550 and 533 nM, respectively). This difference indicated that an orientation of the nitrogen atom on the pyridine ring was important for the binding affinity. The nitrogen atom might interact with the antibody as a hydrogen-bond acceptor. Considering this, the imidazole of His in 15-IgBP could also work as the hydrogen-acceptor rather than the hydrogen-donor for the binding. Interestingly, peptides with a phenyl or pyridyl ring showed similar kinetic properties in the antibody binding (K_d = 506–579 nM, k_on = 0.713–0.772 s^–1 μM^–1 and k_off = 0.361–0.425 s^–1), the exception being 15-His17(2-Pya) which has a slower k_off value (0.0579 s^–1). These results suggest that the interaction of the nitrogen atom at the δ-position on a basic aromatic ring contributes significantly to the slow dissociation, resulting in the high binding affinity. The peptide in which His17 has been replaced by 2-thienylalanine (2-Thi), 15-His17(2-Thi), has a moderate binding affinity (K_d = 220 nM) with a little slow dissociation property (k_off = 0.169 s^–1), probably because of the relatively electronegative sulfur atom at the δ-position. 15-His17(d-2-Pya), which was an unexpected product generated from the racemization, has a lower binding affinity (K_d = 579 nM), suggesting that position 17 is preferably composed of an l-form amino acid. Additionally, the non-aromatic peptide, 15-His17Cha, showed a slightly better binding affinity (K_d = 165 nM) than 15-His17Phe, evidencing that this position may not always require an aromatic structure. Interestingly, the k_on value of every derivative with the substitution at His17 was similar to that of 15-IgBP (k_on = 0.567–0.776 s^–1 μM^–1), while the k_off value varied over a broad range, suggesting that the His17 position significantly affects the dissociation step, but not the association step of the IgG binding.

Table 2. Binding affinity evaluation of 15-IgBP derivatives.

Peptide	K d (nM)	k on (s–1 μM–1)	k off (s–1)
15-IgBP a	267 ± 4	0.741 ± 0.003	0.198 ± 0.001
15-His17Ala	841 ± 6	0.567 ± 0.003	0.477 ± 0.001
15-His17Phe	506 ± 7	0.713 ± 0.004	0.361 ± 0.001
15-His17Trp b	1220 ± 50	—	—
15-His17(2-Pya)	75.7 ± 4	0.764 ± 0.003	0.0579 ± 0.0000
15-His17(3-Pya)	550 ± 4	0.772 ± 0.003	0.425 ± 0.001
15-His17(4-Pya)	533 ± 14	0.730 ± 0.007	0.389 ± 0.004
15-His17Dab	1020 ± 10	0.731 ± 0.006	0.749 ± 0.000
15-His17Orn b	710 ± 50	—	—
15-His17(2-Thi)	220 ± 3	0.771 ± 0.002	0.169 ± 0.002
15-His17(d-2-Pya)	579 ± 20	0.593 ± 0.009	0.344 ± 0.005
15-His17Cha	165 ± 4	0.776 ± 0.003	0.128 ± 0.002
15-IgBP-Gly	338 ± 4	0.615 ± 0.002	0.208 ± 0.000
15-IgBP-pHB	203 ± 11	0.603 ± 0.007	0.122 ± 0.000
15-His17(2-Pya)-OH c	97.6 ± 3.8	0.690 ± 0.002	0.0674 ± 0.0001

^aFrom ref. 10.

^bSteady state analysis was applied because of the low reliability of a fitting curve (U-value > 14).

^cC-terminal structure is carboxylic acid.

In the derivatization of the C-terminal structure, the peptide with an additional Gly residue at position 18 (15-IgBP-Gly) and the p-hydroxybenzyl amidated peptide (15-IgBP-pHB) showed generally similar binding affinities (K_d = 203 and 338 nM, respectively) from 15-IgBP (267 nM). These results suggest that structural changes in the C-terminal structure of the peptide can be highly tolerated. In addition, since the peptide with a C-terminal carboxylic acid in the substitution of His17 to 2-Pya (15-His17(2-Pya)-OH, K_d = 97.6 nM) shows a similar binding affinity to the amide type (15-His17(2-Pya), K_d = 75.7 nM), the difference of the C-terminal structure, i.e., the amide- or acid-form, has little effect on the binding affinity. We finally obtained a highly potent and readily synthesized peptide 15-His17(2-Pya)-OH with reasonably slow dissociation, which can be applied to the CCAP method.

Binding affinity of IgG-binding peptides with Lys8Leu substitution

In a previous SAR study of 15-IgBP,10 the substitution of Lys8 by Leu drastically improved the K_d value (15-Lys8Leu, K_d = 8.19 nM) (Fig. 1). Although the peptide cannot be used in the CCAP method, the strong binding peptide is expected to have many other functions, such as an immunoliposome9a or a non-covalent-type ADC.9c In this study, we synthesized a peptide derivative that combines the substitutions of Lys8Leu and His17(2-Pya). As a result, 15-Lys8Leu/His17(2-Pya) showed an improved K_d value (2.19 nM) with a slower k_off value (0.00326 s^–1, Table 3). The epimerized peptide, 15-Lys8Leu/His17(d-2-Pya), showed a 10-fold weaker binding affinity (K_d = 23.0 nM) compared to the l-form but as expected, the C-terminal carboxylic acid peptide showed the same K_d value (15-Lys8Leu/His17(2-Pya)-OH, 2.48 nM) as the C-terminal amidated peptide. The peptide, 15-Lys8Leu/His17(2-Pya)-OH, has one of the best K_d values among the previously reported IgG-binding peptides including bicyclic scaffolds (FcIII-4C (ref. 18) and FcBP-2,19K_d = 2.45 and 2.2 nM, respectively). In addition, 15-Lys8Leu/His17(2-Pya)-OH overcomes the chemical synthesis issues.

Table 3. Antibody binding affinity of 15-Lys8Leu derivatives.

Peptide	K d (nM)	k on (s–1 μM–1)	k off (s–1)
15-Lys8Leu a	8.19 ± 2.07	1.47 ± 0.00	0.0116 ± 0.0010
15-Lys8Leu/His17(2-Pya)	2.19 ± 1.49	1.49 ± 0.00	0.00326 ± 0.00000
15-Lys8Leu/His17(d-2-Pya)	23.0 ± 2.9	1.32 ± 0.00	0.0304 ± 0.0001
15-Lys8Leu/His17(2-Pya)-OH b , c	2.48 ± 1.42	1.35 ± 0.00	0.00330 ± 0.00000

^aFrom ref. 10.

^bC-terminal structure is carboxylic acid.

^c0.0001% DMSO was added for the analysis at 100 nM and serially diluted.

Secondary structure analysis by circular dichroism spectroscopy

The secondary structure of 15-IgBP derivatives was evaluated by analysis of the circular dichroism (CD) spectra (Fig. 4 and Fig. S3 in the ESI†). As discussed previously,10 despite its reported rigid β-sheet-like structure involving β-bulge,8,19 15-IgBP shows a random coil-like spectrum with a random coil content of 60% calculated on the basis of Reed's study.20 This discrepancy was probably attributed to the lack of CD spectral information of β-bulge. All of the synthetic peptides also showed a similar random coil propensity (51–64%). However, 15-His17Phe and 15-His17Trp showed a tendency different from that of 15-IgBP, even though their high random coil contents (51% and 56%, respectively) are similar. This difference was seen in the decrease of the β-sheet content (15-IgBP vs. 15-His17Phe and 15-His17Trp: 28% vs. 16 and 0%) involving an increase of α-helicity and turn propensities. The antibody binding affinity of these peptides is low (K_d = 506 and 1220 nM, respectively) among the synthesized derivatives. These results would suggest that the secondary structure of these peptides was pre-organised into a less favourable conformation. In contrast, the potent peptides, such as 15-His17(2-Pya)-OH or 15-Lys8Leu/His17(2-Pya), showed almost the same CD spectra as a parent, 15-IgBP. Therefore, it can be seen that the secondary structure of 15-IgBP, which indicates a random coil-like propensity (but not a random coil structure) in the CD spectrum analysis, is one of the important features showing the strong antibody binding.

Fig. 4. CD spectrum analysis of 15-IgBP derivatives.

Conclusions

We synthesized peptide derivatives of 15-IgBP (K_d = 267 nM) for an SAR study focused on the C-terminal His17 residue to obtain novel peptides with potent binding affinities. During the peptide synthesis, three problems emerged: p-hydroxy-benzyl amidation at the C-terminus, giving a product (MW + 106), coupling difficulty at the His7 position (peptide (8–17)) and racemization of 2-Pya. It has been shown that these problems can be avoided by substitution of His17 by 2-Pya and using a Cl-Trt(2-Cl) resin for the peptide with a C-terminal carboxylic acid. In the evaluation of antibody binding affinity, 15-His17(2-Pya) showed a 4-fold improved affinity (K_d = 75.7 nM) with a small k_off value. This SAR result reveals that a basic aromatic nitrogen atom at the δ-position of the His or 2-Pya residue at position 17 is an important structural factor probably acting as a hydrogen-bonding acceptor. And the modification of His to 2-Pya suitably affected the dissociation step, resulting in the improved binding affinity. 15-His17(2-Pya) was the best peptide found in this study which can be applied to the CCAP method. Moreover, with the combination of previously reported Lys8Leu substitution, we found that 15-Lys8Leu/His17(2-Pya) is one of the best peptides with a potent binding affinity (K_d = 2.19 nM) among the previously reported IgG-binding peptides. Moreover, the C-terminal structure is either the amide- or acid-form and this does not affect the affinity greatly. We found that the potent peptide, 15-Lys8Leu/His17(2-Pya)-OH (K_d = 2.48 nM), can be efficiently prepared, overcoming the three synthetic issues. These potent peptides can be applied to various research projects such as the preparation of homogeneous ADCs. Further SAR studies of IgG-binding peptides are currently in progress.

Experimental

General procedure

Reagents and solvents were purchased from Wako Pure Chemical Industries (Osaka, Japan), Sigma-Aldrich (St. Louis, MO), Watanabe Chemical Industries (Hiroshima, Japan) and Tokyo Chemical Industries (Tokyo, Japan). All materials were used as received. Herceptin was purchased from the Chugai Co. Ltd. Mass spectra were obtained on a Waters MICRO MASS LCT-Premier.

Peptide synthesis

IgG-binding peptides were synthesized by the Fmoc based SPPS method11 using an automated protocol (Prelude). For the peptide with a C-terminal carboxylic acid, the Fmoc-amino acid (0.6 eq.) was loaded onto the Cl-Trt(2-Cl) resin (1.60 mmol g^–1) in CH₂Cl₂ with the presence of DIEA (2.4 eq.) under an argon atmosphere. After stirring for 25 min, the resin was washed with CH₂Cl₂ : MeOH : DIEA (17 : 2 : 1) thrice, and then washed with CH₂Cl₂, DMF, and MeOH and dried in vacuo. The loading rate was determined from an absorbance at a wavelength of 301 nm of Fmoc-piperidine [ε = 7800 L mol^–1 cm^–1] after treatment with 50% piperidine/DMF. 40 μmol Fmoc-NH-SAL resin (0.56 mmol g^–1) and Fmoc-2-Pya-Trt(2-Cl) resin were used, and the peptide chains were elongated by the Fmoc-amino acid (5 eq.), HATU (5 eq.) and HOAt (5 eq.) in the presence of DIEA (10 eq.) for 30 min. For the amino acids whose side chain functional groups require protection, the following protecting groups were used: Fmoc-Asp(OtBu)-OH, Fmoc-Cys(Trt)-OH, Fmoc-Tyr(tBu)-OH, Fmoc-His(Trt)-OH, Fmoc-Lys(Boc)-OH, Fmoc-Glu(OtBu)-OH, Fmoc-Trp(Boc)-OH, Fmoc-Thr(tBu)-OH, Fmoc-Dab(Boc)-OH and Fmoc-Orn(Boc)-OH. The Fmoc group was deprotected with 20% piperidine in DMF for 10 min twice. These reactions were repeated to lengthen the desired peptide. Only His7 and Tyr6 were coupled twice because these positions were difficult to couple and the reaction could not be completed. The N-terminal amino group was then capped with acetic anhydride in the presence of DIEA. For the peptide used in SPR and CD analysis, the protected peptide resin (40 μmol) was treated with TFA cocktail (Table 1) to obtain a crude peptide with two unprotected thiol groups. After purification by reverse-phase (RP)-HPLC, an intramolecular disulfide bond was formed by the treatment with methyl 3-nitro-2-pyridinesulfenate (Npys-OMe)21 or DMSO oxidation under a low concentration. The products were then purified again by RP-HPLC. The purity of synthesized peptides was analysed by HPLC with a RP-column (COSMOSIL 5C18 AR-II, 4.6ID × 150 mm) using a binary solvent system with a linear gradient from 10% to 50% CH₃CN in 0.1% aqueous TFA at a flow rate of 0.9 mL min^–1 with UV detection at 230 nm. The peptides used in the analyses had an HPLC purity of >95%. The yield and analytical data are shown in the ESI.†

For the micro-cleavage, 1 μmol of the resin was placed in a 1.5 mL tube. The resins were treated with 500 μL of TFA cocktail. The solution was concentrated with a stream of N₂. Then, the crude products were precipitated with Et₂O and washed twice. After drying, the residual solids were analysed by RP-HPLC using the same aforementioned conditions. The percentage purity of the peptides was calculated from these chromatograms for Table 1.

Evaluation of binding affinity

The binding kinetics were determined on the basis of previously reported conditions using a Biacore T-200 system.10 Herceptin was dissolved in an acetate buffer (pH 5.5) and immobilized with premixed N-hydroxysuccinimide and 1-(3-dimethylaminopropyl)-3-ethyl-carbodiimide hydrochloride (EDC·HCl) onto a CM5 sensor chip. The analytes (IgBP derivatives) were adjusted to the desired concentration by a serial dilution in a running buffer (HBS-EP; 0.01 M HEPES, 0.15 M NaCl, 3 mM EDTA, 0.005% Tween 20, pH 7.4). The sensorgrams were obtained with an association time of 180 s, a dissociation time of 600 s, and a flow rate of 50 μL min^–1. In only the cases of 15-Lys8Leu/His17(2-Pya) and 15-Lys8Leu/His17(2-Pya)-OH, the sensor chip was washed with 10 mM Gly-HCl buffer (pH 2.0) for 5 s because the dissociation time was insufficient. In the analysis of 15-Lys8Leu/His17(2-Pya)-OH, 0.0001% DMSO at a concentration of 100 nM was included and serially diluted, because of the low aqueous solubility. To determine the binding kinetics (k_on, k_off and K_d), the obtained sensorgrams were analyzed using the Biacore T200 evaluation software Ver.1.0, using a 1 : 1 binding model or steady state analysis.

CD spectra

The circular dichroism spectra of the peptide derivatives were measured on the basis of previously described conditions10 using a Jasco J-1500CD spectrometer (JASCO, Japan) and a quartz cell with a 0.5 cm path length. Spectra were collected between 190 and 250 nm with a scan speed of 100 nm min^–1, a response time of 1 s, and a bandwidth of 1 nm. The peptides were dissolved in a 10 mM phosphate buffer (pH 7.4) with a concentration of 2.5 μM.

Conflicts of interest

There are no conflicts to declare.