Coordination Environment of Cu(II) Ions Bound to N-Terminal Peptide Fragments of Angiogenin Protein

Angiogenin (Ang) is a potent angiogenic factor, strongly overexpressed in patients affected by different types of cancers. The specific Ang cellular receptors have not been identified, but it is known that Ang–actin interaction induces changes both in the cell cytoskeleton and in the extracellular matrix. Most in vitro studies use the recombinant form (r-Ang) instead of the form that is normally present in vivo (“wild-type”, wt-Ang). The first residue of r-Ang is a methionine, with a free amino group, whereas wt-Ang has a glutamic acid, whose amino group spontaneously cyclizes in the pyro-glutamate form. The Ang biological activity is influenced by copper ions. To elucidate the role of such a free amino group on the protein–copper binding, we scrutinized the copper(II) complexes with the peptide fragments Ang(1–17) and AcAng(1–17), which encompass the sequence 1–17 of angiogenin (QDNSRYTHFLTQHYDAK-NH2), with free amino and acetylated N-terminus, respectively. Potentiometric, ultraviolet-visible (UV-vis), nuclear magnetic resonance (NMR) and circular dichroism (CD) studies demonstrate that the two peptides show a different metal coordination environment. Confocal microscopy imaging of neuroblastoma cells with the actin staining supports the spectroscopic results, with the finding of different responses in the cytoskeleton organization upon the interaction, in the presence or not of copper ions, with the free amino and the acetylated N-terminus peptides.


Introduction
Human Angiogenin (Ang) is a 14 kDa protein belonging to the ribonucleases family, with a RNase catalytic activity about 10 6 lower than pancreatic RNase [1].
For both peptides, the two lowest pKs refer to the carboxylic group of the two aspartic residues. Since their deprotonation occurred with overlapping titration curves, the first and second protonation constants have not been assigned to the specific residue, Asp-2 or Asp-15, respectively [43]. The next two deprotonation steps involve the imidazole nitrogen atoms of the two His residues; also in this case hence each protonation constant has been considered as a macroconstant value, due to the partially overlap of the titration curves,. Until the deprotonation reaction step of histidine (pH < 7), the behavior of the two peptides is similar. At higher pH values, the deprotonation processes of Tyr-6, Tyr-14 and Lys-17, measured only for AcAng (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17), resulted in the assignment of three macroconstant values, in good agreement with those observed for analogous peptide fragments [39,44].
The CD difference spectra (inset in Figure 2a) do not indicate an increase of specific conformational structure at the different pH values, but the presence of a turn structure can be figured out at the highest pH values of 9 and 10, as reported for other linear peptides [46].
The enhancement of the peptide turn conformation, induced by the pH increase, can be interpreted as due to the involvement of backbone amide nitrogen atoms in Cu 2+ coordination (see Section 2.3). The far-UV CD spectra of Cu-peptide system are governed by the amide chromophore, but may contain a contribution of aromatic side chains, in particular the band at 224 and 245 nm might be related to the tyrosine phenolic group deprotonation [47]. copper(II) induces a decrease for the signal at 198 nm in the pH ranges of 4-5 and 7-9, whereas a opposite trend is observed at pH 6 and 10, accompanied also by a red-shift at pH = 6 and a blue-shift at pH = 10, respectively ( Figure 2a). Moreover, at the highest pH values of 9 and 10, two new maxima and one minimum are clearly visible, at 214, 245 and 224 nm, respectively. The CD difference spectra (inset in Figure 2a) do not indicate an increase of specific conformational structure at the different pH values, but the presence of a turn structure can be figured out at the highest pH values of 9 and 10, as reported for other linear peptides [46].

Speciation and Characterization of Copper Complexes with Ang
The stability constant values are reported in Table 2 and the corresponding distribution diagrams in Figure 3, respectively.
UV-vis and CD parameters can discriminate the actual copper(II) coordination environment ( Table 3). The UV-vis parameters of [CuLH] species (λmax = 628 nm ε = 90 M −1 ·cm −1 , see Table 3) rule out the formation of a macrochelate, involving two imidazole nitrogen atoms and one carboxylate, which would exhibit the absorption at higher wavelength [49]. Our data are indeed very similar to On the other hand, by addition of one equivalent of copper to AcAng(1-17), a general decrease and slight shifts of the bands at 198 and 228 nm in the whole pH range of 4-9 are observed ( Figure 2b). The difference spectra indicate, already at acidic pH, the formation of turn structures that increase at basic pH. The intensity of signals is much lower than that observed for Cu 2+ -Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17). Indeed, in the acetylated peptide, the imidazole nitrogen atoms of His-8 and His-13 are the potential metal anchoring sites; hence, the formation of a macrochelate, resulting in a turn structure, might include primarily the central portion of the peptide (residues 8-13). Otherwise, in Ang(1-17), the free amino terminal amino group is a further potential copper anchoring site; the stronger dichroic effect observed suggests the formation of a greater macrochelate involving amino group of first residue and imidazole nitrogen of His-8 and/or His-13.

Speciation and Characterization of Copper Complexes with Ang(1-17) and AcAng(1-17) Peptides
The stability constant values are reported in Table 2 and the corresponding distribution diagrams in Figure 3, respectively.
UV-vis and CD parameters can discriminate the actual copper(II) coordination environment ( Table 3). The UV-vis parameters of [CuLH] species (λ max = 628 nm ε = 90 M´1¨cm´1, see Table 3) rule out the formation of a macrochelate, involving two imidazole nitrogen atoms and one carboxylate, which would exhibit the absorption at higher wavelength [49]. Our data are indeed very similar to those reported for a peptide binding Cu 2+ by means of the terminal amino group, the deprotonated amide nitrogen atom and the oxygen of a carboxylate group of contiguous aspartic residue [50].      Accordingly, the CD spectrum carried out at pH = 5.5 shows the presence of a band at 299 nm ascribable to a N-amide ÑCu 2+ charge transfer, confirming the presence of a deprotonated amide nitrogen in metal coordination environment. In addition, a band at 265 nm is found, assigned to either imidazole or amino groups. Hence, the [CuLH] species has to be considered instead as a [CuLH´1(H) 2 ] species, in which one imidazole nitrogen is still protonated and the actual chromophore is Cu(NH 2 , N´, O COO´, O water ).
The next species formed, [CuL], is predominant at physiological pH. The stepwise constant logK 110 (logK 110 = logβ 110´l ogβ 111 = 6.35) is compatible with a deprotonation of another amide nitrogen atom. The 25 nm blue shift in the UV-vis λ max absorption confirms the involvement of a further nitrogen atom in the metal coordination environment.
In Cu 2+ 3N1O coordination mode two or more isomers may exist; among them the most likely are the following: (i)-form where a second deprotonated amide atom is coordinated to Cu 2+ (NH 2 , 2N´, O COO´) ; and (ii)-form, with a deprotonated imidazole ring (N im ) in the equatorial plane (NH 2 , N´, N Im , O COO´) . The presence in the CD spectra recorded at pH = 7 of a dichroic band centered at 336 nm (diagnostic of N-imidazole ÑCu 2+ charge transfer) point to the predominant (ii)-form. Indeed, the UV-vis parameters (λ max = 605 nm ε = 95 M´1¨cm´1) confirm this coordination mode, as found in literature for complex species formed by peptides with a similar primary sequence [49].
Increasing the pH, [CuLH´1] complex species is formed; this species displays its maximum percentage of formation at pH = 8.5. The logK value (logK = logβ 11-1´l ogβ 110 = 8.14) indicates deprotonation and a further coordination of an amide nitrogen atom. This hypothesis is confirmed by UV-vis spectrum where a 35 nm blue shift of the d-d band is observed (Table 3), and by the intensity increase of the dichroic band relative to N-amide ÑCu 2+ charge transfer, at 322 nm.
Above pH = 10, the Tyr and Lys deprotonation occurs, with the formation of [CuLH´3], [CuLH´4] and [CuLH´5] species. Spectroscopic data do not evidence any change, indicating that Tyr and Lys residue are not involved in metal ion binding [48].
As to AcAng(1-17), Figure 3b shows that the peptide forms the [CuLH 4 ] species at pH = 4, which reaches its maximum percentage of formation (around 25%) at pH = 5. The calculated stability constant (logK = logβ 114´l ogβ 014 = 3.99) is higher respect to the typical stability constant of a copper complex in which the metal ion is coordinated to only one imidazole nitrogen of a histidine [51]. Accordingly, this complex species involves a carboxylate group of one aspartate residue, whereas the other aspartate is deprotonated but not coordinated to the metal. Therefore, four isomers with [CuLH 4 ] stoichiometry can exist, formed by His-8, His13, Asp-2 and Asp-15. The smaller macrochelate formed by His13 and Asp-15 is the most thermodynamically favored isomer form.
Increasing the pH, [CuLH 3 ] is formed and its stability constant (logK = logβ 131´l ogβ 031 = 5.95) is indicative of a macrochelate formation with coordination to Cu 2+ of two imidazole rings and one COO´group of aspartic. This stability constant value is in a good agreement with the value reported for the corresponding complex species formed by other similar peptide fragments [39].
The UV-vis parameters of [CuLH 3 ] species (λ max = 650 nm, ε = 60 M´1¨cm´1, see Table 3) indicate a metal coordination to two imidazole nitrogen atoms, one carboxylate group and a water molecule. It is to note that for the corresponding species formed by the unacetylated peptide, λ max value is at the lower wavelength of 628 nm. The CD spectra at pH 5.9 do not show any evidence of the amide band while a band at 331 nm is a fingerprint of imidazole coordination.
The next deprotonation step involves one amide backbone nitrogen atom, but the [CuLH 2 ] is only a minor species; for this reason, no spectroscopic data for this complex species have been obtained.
At physiological pH, the main species for the Cu-AcAng(1-17) system is [CuLH]. The calculated pK value (pK(2/1) = 5.79; see Table 2) indicates the deprotonation of a further amide nitrogen. This finding suggests the simultaneous presence of the imidazole rings and amide nitrogen atoms bound to the copper(II), as confirmed by the CD spectrum recorded at pH = 7.4, showing the presence of two bands centered around 360 nm and 320 nm, respectively. Furthermore, UV-vis parameters (λ = 580 nm and ε = 10 5 M´1¨cm´1) indicate the involvement of four nitrogen atoms in the equatorial plane of the metal.
Increasing the pH, the [CuL] species is formed. In this case, the pK determined for this deprotonation step is 7.38 and is reported in Table 2 as pK(1/0). The coordination of a third nitrogen atom deprotonated is confirmed by the blue-shift of 40 nm observed for the λ max (λ = 540 nm and ε = 125 M´1¨cm´1). Furthermore, the increase of the ε is a clear indication that the coordination to the Cu(II) of this third amide nitrogen atom deprotonated with a release of an imidazole ring, created a more distorted copper coordination polyhedron with a stronger ligand field.
The other complex species formed at basic pH is assigned to the deprotonation of the side chains of the tyrosine and lysine residues that do not influence the complexation of Cu(II), as shown by the unchanged CD ad UV-vis spectral parameters.
Increasing the pH, the [CuL] species is formed. In this case, the pK determined for this deprotonation step is 7.38 and is reported in Table 2 as pK(1/0). The coordination of a third nitrogen atom deprotonated is confirmed by the blue-shift of 40 nm observed for the λmax (λ = 540 nm and ε = 125 M −1 ·cm −1 ). Furthermore, the increase of the ε is a clear indication that the coordination to the Cu(II) of this third amide nitrogen atom deprotonated with a release of an imidazole ring, created a more distorted copper coordination polyhedron with a stronger ligand field.
The other complex species formed at basic pH is assigned to the deprotonation of the side chains of the tyrosine and lysine residues that do not influence the complexation of Cu(II), as shown by the unchanged CD ad UV-vis spectral parameters.
The direct involvment of D15 residue in the Cu 2+ coordination, rather than D2, is also supported by the disappearance in the 1 H-1 H TOCSY spectrum of its Hβ-HN and Hα-HN correlations ( Figure S3), confirming the representation of [CuLH 3 ] coordination mode above described. Figure 5 shows the superposition of 1 H aromatic and aliphatic region for AcAng(1-17) by increasing sub-stoichiometric metal to ligand molar ratio, from 1:0 to 1:0.1, obtained at pH 7.
From the spectra, the gradual but specific decrease in intensity involving mainly proton signals from H8, H13 and D15 and, to a lesser extent, from F9 and Y14 residues is evident. The involvement of H8, H13 and D15 residues in metal binding is further proven by the disappearance or broadening of their protons in the 1 H-13 C HSQC spectra ( Figure S4). The disappearance of signals from Hε1 and Hδ2 of H8 and H13 residues together with Asp-15 residue is visible, supporting the concomitant participation of both histidine residues together with aspartate residue in the coordination to Cu 2+ ions. Moreover, the attenuation of signals intensities owing to residues close to the paramagnetic binding site is detected for the residues F9, L10, Q12 and Y14.
The direct involvment of D15 residue in the Cu 2+ coordination, rather than D2, is also supported by the disappearance in the 1 H-1 H TOCSY spectrum of its Hβ-HN and Hα-HN correlations ( Figure S3), confirming the representation of [CuLH3] coordination mode above described. Figure 5 shows the superposition of 1 H aromatic and aliphatic region for AcAng(1-17) by increasing sub-stoichiometric metal to ligand molar ratio, from 1:0 to 1:0.1, obtained at pH 7.
From the spectra, the gradual but specific decrease in intensity involving mainly proton signals from H8, H13 and D15 and, to a lesser extent, from F9 and Y14 residues is evident. The involvement of H8, H13 and D15 residues in metal binding is further proven by the disappearance or broadening of their protons in the 1 H-13 C HSQC spectra ( Figure S4). Additional data to support the involvement of these residues in the coordination are provided by the 1 H-1 H TOCSY, where the spin system HN from H8, H13 and D15 in the AcAng(1-17):Cu(II) 1:0.05 system disappears ( Figure 6).  Additional data to support the involvement of these residues in the coordination are provided by the 1 H-1 H TOCSY, where the spin system HN from H8, H13 and D15 in the AcAng(1-17):Cu(II) 1:0.05 system disappears ( Figure 6). The disappearance of signals from Hε1 and Hδ2 of H8 and H13 residues together with Asp-15 residue is visible, supporting the concomitant participation of both histidine residues together with aspartate residue in the coordination to Cu 2+ ions. Moreover, the attenuation of signals intensities owing to residues close to the paramagnetic binding site is detected for the residues F9, L10, Q12 and Y14.
The direct involvment of D15 residue in the Cu 2+ coordination, rather than D2, is also supported by the disappearance in the 1 H-1 H TOCSY spectrum of its Hβ-HN and Hα-HN correlations ( Figure S3), confirming the representation of [CuLH3] coordination mode above described. Figure 5 shows the superposition of 1 H aromatic and aliphatic region for AcAng(1-17) by increasing sub-stoichiometric metal to ligand molar ratio, from 1:0 to 1:0.1, obtained at pH 7.
From the spectra, the gradual but specific decrease in intensity involving mainly proton signals from H8, H13 and D15 and, to a lesser extent, from F9 and Y14 residues is evident. The involvement of H8, H13 and D15 residues in metal binding is further proven by the disappearance or broadening of their protons in the 1 H-13 C HSQC spectra ( Figure S4). Additional data to support the involvement of these residues in the coordination are provided by the 1 H-1 H TOCSY, where the spin system HN from H8, H13 and D15 in the AcAng(1-17):Cu(II) 1:0.05 system disappears ( Figure 6).  A relevant line broadening is also detected for F9, T11, Q12 and Y14, indicating that the residues mainly involved in the coordination are those between H8 and H13 ( Figure S5).
The NMR study of Ang(1-17) has been carried out by using the same increasing sub-stoichiometric copper to ligand molar ratios, at both pH of 5.5 and 7.
In Figure 7, the 2D 1 H-13 C HSQC NMR spectra of Ang(1-17) at pH 5.5, in the aromatic and aliphatic region, for the free (in red) and the copper complex systems at 0.02/1 and 0.05/1 molar ratios (in blue) are reported. The most evident difference with respect to AcAng(1-17) is the strong decrease of Q1, D2, N3 and Y6 signals, other than Hβ of D15, Hε1 and Hδ2 of H8 and H13 residues. A relevant line broadening is also detected for F9, T11, Q12 and Y14, indicating that the residues mainly involved in the coordination are those between H8 and H13 ( Figure S5).
The NMR study of Ang(1-17) has been carried out by using the same increasing substoichiometric copper to ligand molar ratios, at both pH of 5.5 and 7.
In Figure 7, the 2D 1 H-13 C HSQC NMR spectra of Ang(1-17) at pH 5.5, in the aromatic and aliphatic region, for the free (in red) and the copper complex systems at 0.02/1 and 0.05/1 molar ratios (in blue) are reported. The most evident difference with respect to AcAng(1-17) is the strong decrease of Q1, D2, N3 and Y6 signals, other than Hβ of D15, Hε1 and Hδ2 of H8 and H13 residues. The 1 H-1 H TOCSY spectra of Ang(1-17), both free and Cu(II)-complexed at pH 5.5, are reported in Figure 8. The proton spin systems of D15 residue totally disappear, whereas those of D2 residue are still present, though decreased in intensity.
The titration of Ang(1-17) with increasing amount of Cu(II), in the molar ratio range from 1:0 to 1:0.05, has been followed at pH 7 and the corresponding 1 H spectra are reported in Figure S6. From the comparison of 1 H-13 C HSQC NMR spectra in the aromatic and aliphatic region of Ang(1-17) free (red) and of Cu(II):Ang system (blue) ( Figure S7) and from the selection of 1 H-1 H TOCSY spectrum of the same systems ( Figure S8), it is evident that the signals from H8 and H13 entirely disappear. Moreover, the H-C correlations from Asp residues, D2 and D15, almost disappeared in the HSQC spectrum, are instead still present in the TOCSY spectrum, suggesting that their involvement in the coordination to Cu(II) ion is reduced, mainly for D15 residue.
Line broadening can be clearly noticed for R5, F9 and Y14 residues. The residues mainly affected by the interaction of Cu(II) ion with Ang(1-17) are evidenced in Figure S9. Conversely to what has been determined with AcAng(1-17), the D15 residue in Ang(1-17) shows a limited participation in complex formation, as demonstrated by its almost unaffected resonances in the TOCSY spectra obtained at pH 7 ( Figure S7). In summary, the NMR data at pH 7 indicate the involvement of both N-terminal residue (Q1, Asp2) together with imidazole, as deduced from potentiometry and UV-vis CD results. The 1 H-1 H TOCSY spectra of Ang(1-17), both free and Cu(II)-complexed at pH 5.5, are reported in Figure 8. The proton spin systems of D15 residue totally disappear, whereas those of D2 residue are still present, though decreased in intensity.
The titration of Ang(1-17) with increasing amount of Cu(II), in the molar ratio range from 1:0 to 1:0.05, has been followed at pH 7 and the corresponding 1 H spectra are reported in Figure S6. From the comparison of 1 H-13 C HSQC NMR spectra in the aromatic and aliphatic region of Ang(1-17) free (red) and of Cu(II):Ang system (blue) ( Figure S7) and from the selection of 1 H-1 H TOCSY spectrum of the same systems ( Figure S8), it is evident that the signals from H8 and H13 entirely disappear. Moreover, the H-C correlations from Asp residues, D2 and D15, almost disappeared in the HSQC spectrum, are instead still present in the TOCSY spectrum, suggesting that their involvement in the coordination to Cu(II) ion is reduced, mainly for D15 residue.
Line broadening can be clearly noticed for R5, F9 and Y14 residues. The residues mainly affected by the interaction of Cu(II) ion with Ang(1-17) are evidenced in Figure S9. Conversely to what has been determined with AcAng(1-17), the D15 residue in Ang(1-17) shows a limited participation in complex formation, as demonstrated by its almost unaffected resonances in the TOCSY spectra obtained at pH 7 ( Figure S7). In summary, the NMR data at pH 7 indicate the involvement of both N-terminal residue (Q1, Asp2) together with imidazole, as deduced from potentiometry and UV-vis CD results.

Neuroblastoma Cells Experiments
The results of thetetrazolium dye 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay ( Figure 9) evidence that all the tested treatments are non-cytotoxic. No significant differences in cell viability are found, except for the incubation with 10 µM of the complex Ang(1-17):Cu(II), where a small increase in neuroblastoma cell viability is measured compared to the control. Preliminary cellular experiments of confocal imaging were performed to scrutinize qualitatively the effects of both peptides and the proteins in terms of actin organization, when supplemented to the cells together with the metal ions.

Neuroblastoma Cells Experiments
The results of thetetrazolium dye 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay ( Figure 9) evidence that all the tested treatments are non-cytotoxic. No significant differences in cell viability are found, except for the incubation with 10 µM of the complex Ang(1-17):Cu(II), where a small increase in neuroblastoma cell viability is measured compared to the control.

Neuroblastoma Cells Experiments
The results of thetetrazolium dye 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay ( Figure 9) evidence that all the tested treatments are non-cytotoxic. No significant differences in cell viability are found, except for the incubation with 10 µM of the complex Ang(1-17):Cu(II), where a small increase in neuroblastoma cell viability is measured compared to the control. Preliminary cellular experiments of confocal imaging were performed to scrutinize qualitatively the effects of both peptides and the proteins in terms of actin organization, when supplemented to the cells together with the metal ions.
The quantitative analysis of fluorescence for the actin staining is shown in Figure 11.  The common patterns for actin microfilament organization in the cell cytoplasm, i.e., peripheral actin bands, actin bundles (stress fibers), and diffuse actin networks, are visible for all the used cell treatments. The actin pattern is brightest for the cells treated with r-Ang (Figure 10f), while it visibly changes in cells treated with r-Ang:Cu(II) complex (Figure 10g). On the contrary, both the treatments with wt-Ang ( Figure 10h) and wt-Ang:Cu(II) complex (Figure 10i) induce comparable actin patterns, with an evident polygonal arrangement of actin filaments.
It is noteworthy that the cells treated with Ang(1-17):Cu(II) complex ( Figure 10b) display a similar actin pattern than those treated with the free-amino r-Ang complexed to copper. On the other hand, for the acetylated peptide, similarities are found between the cells treated with the apo-form ( Figure 10c) and the N terminal amino-blocked wt-Ang (Figure 10h).
The quantitative analysis of fluorescence for the actin staining is shown in Figure 11. It is noteworthy that the cells treated with Ang(1-17):Cu(II) complex ( Figure 10b) display a similar actin pattern than those treated with the free-amino r-Ang complexed to copper. On the other hand, for the acetylated peptide, similarities are found between the cells treated with the apo-form ( Figure 10c) and the N terminal amino-blocked wt-Ang (Figure 10h).
The quantitative analysis of fluorescence for the actin staining is shown in Figure 11.  The cells treated with the metal alone at the concentration used with the proteins (100 nM) do not significantly differ with respect to the control untreated cells. On the other hand, the cells supplemented with 10 µM copper (i.e., the concentration used for the peptides) visibly increase the actin fluorescence intensity. The overall trend observed, for both the proteins and the peptides, is a decrease of intensity for cell treatment with their copper complexes in comparison to the treatment with the free ligands. It is noteworthy that such intensity decrease exhibit similarities in wt-Ang and Ang(1-17) (i.e., huge decrease in intensity) and r-Ang and AcAng(1-17) (i.e., small decrease in intensity), which are in contrast to what was expected on the basis of the peptidomimetic design strategy, namely AcAng(1-17) mimicking wt-Ang and Ang(1-17) mimicking the r-Ang. This finding suggest that other factors are likely to be involved in the peptide-metal/cell membrane interaction which cannot simply be ruled out by this experiments. Specifically, the involvement of other protein domains and/or the formation of various Ang-Ang aggregates driven by the presence of the free amino group can be invoked. Once again, these considerations stress the relevant issue of the use of the "proper" protein form, wt-Ang, to scrutinize the pathways involving the protein in angiogenesis processes, especially in the presence of the metal ions.

Discussion
The protein angiogenin (Ang) is a potent angiogenic factor whose activity is influenced by copper ions. Many literature reports on the protein activity have been obtained using the recombinant form (r-Ang), which contains an extra methionine as first residue. In contrast, the protein effectively present in human plasma (wt-Ang), exhibit the amino terminal group blocked, owing to the spontaneously cyclization of the first residue glutamine to pyroglutamate.
Taking into account the biological role of the protein in physiological as well as pathological conditions involving the angiogenic processes, we tested the activity of the two peptides in terms of cell viability and staining of actin, which is one of the potential target receptors driving the Ang-cell membrane interaction.
Cellular experiments in neuroblastoma cell line SH-SY5Y cells demonstrate that, at the used experimental conditions, both the peptides and their correspondent copper complexes are not-cytotoxic. Rather, a small increase of cell viability, which can be explained as proliferative activity [53], is found for the cells treated with free amino peptide Ang(1-17) complexed with Cu(II) ions.
This observation is explained in terms of the different metal coordination modes and binding affinities likely determining different aggregation morphologies that might result in significantly different biological effects [54].
As to the actin staining results for the cells treated with the whole proteins, the recombinant form is more active (i.e., strong fluorescence emission) than the wild-type angiogenin. Analogously to the peptides, the respective copper complexes exhibit a decrease of fluorescence intensity, this effect being larger for r-Ang than wt-Ang.
Since one possible process related to the actin activation mechanism is the occurrence of both intramolecular Ang-Ang and intermolecular Ang-actin interactions [22], the presence of free amino groups might actually affect such a pathway. Again, the strong effect displayed by copper addition in the case of r-Ang can be correlated to the metal-chelating capability of such free amino groups, as demonstrated for short peptide sequences as well as oligopeptides [54,55]. For example, in the case of r-Ang, the formation of dimeric protein structures with bridges through the amino groups could be prompted by the presence of the copper [37].
In conclusion, this work provides experimental proof to support the significance of using the actual angiogenin present in the human plasma, the wild type form (wt-Ang), to scrutinize the protein activity, especially in the presence of the metal ions.

Potentiometric Titrations
Potentiometric titrations were performed with a home-assembled fully automated apparatus sets (Metrohm E654 pH-meter, combined micro pH glass electrode, Orion 9103SC, Hamilton digital dispenser, Model 665) controlled by the appropriate software set up in our laboratory. The titration cell (2.5 mL) was thermostated at 298.0˘0.2 K, and all solutions were kept under an atmosphere of argon, which was bubbled through a solution having the same ionic strength and temperature as the measuring cell. KOH solutions (0.1 M) were added through a Hamilton buret equipped with 1 cm 3 syringe. The ionic strength of all solutions was adjusted to 0.10 M (KNO 3 ). In order to determine the stability constants, solutions of the ligands (protonation constants) or the ligands with Cu 2+ (copper complex constants) were titrated using 0.1 M sodium hydroxide. Ligand concentration ranged from 1.4 to 2.0ˆ10´3 for the protonation and complexation experiments, respectively. A minimum of three independent runs were performed to determine the protonation constants, while four independent experiments were run for the copper(II) complexation constants.
Metal to ligand ratios of 1:1 were employed. The initial pH was always adjusted to 2.4. To avoid systematic errors and verify reproducibility, the electromotive force (EMF) values of each experiment were taken at different time intervals.
To obtain protonation and complexation constants, the potentiometric data were refined using Hyperquad [56], which minimizes the error square sum of the measured electrode potentials through a nonlinear iterative refinement of the sum of the squared residuals, U, and also allows for the simultaneous refinement of data from different titrations: where E exp and E calc are the experimental and calculated electrode potentials, respectively. Errors in stability constant values are reported as three times standard deviations.
The formation reaction equilibria of ligands with protons and copper(II) ions are given in Equation (1) As to the actin staining results for the cells treated with the whole proteins, the recombinant form is more active (i.e., strong fluorescence emission) than the wild-type angiogenin. Analogously to the peptides, the respective copper complexes exhibit a decrease of fluorescence intensity, this effect being larger for r-Ang than wt-Ang.
Since one possible process related to the actin activation mechanism is the occurrence of both intramolecular Ang-Ang and intermolecular Ang-actin interactions [22], the presence of free amino groups might actually affect such a pathway. Again, the strong effect displayed by copper addition in the case of r-Ang can be correlated to the metal-chelating capability of such free amino groups, as demonstrated for short peptide sequences as well as oligopeptides [54,55]. For example, in the case of r-Ang, the formation of dimeric protein structures with bridges through the amino groups could be prompted by the presence of the copper [37].
In conclusion, this work provides experimental proof to support the significance of using the actual angiogenin present in the human plasma, the wild type form (wt-Ang), to scrutinize the protein activity, especially in the presence of the metal ions.

Potentiometric Titrations
Potentiometric titrations were performed with a home-assembled fully automated apparatus sets (Metrohm E654 pH-meter, combined micro pH glass electrode, Orion 9103SC, Hamilton digital dispenser, Model 665) controlled by the appropriate software set up in our laboratory. The titration cell (2.5 mL) was thermostated at 298.0 ± 0.2 K, and all solutions were kept under an atmosphere of argon, which was bubbled through a solution having the same ionic strength and temperature as the measuring cell. KOH solutions (0.1 M) were added through a Hamilton buret equipped with 1 cm 3 syringe. The ionic strength of all solutions was adjusted to 0.10 M (KNO3). In order to determine the stability constants, solutions of the ligands (protonation constants) or the ligands with Cu 2+ (copper complex constants) were titrated using 0.1 M sodium hydroxide. Ligand concentration ranged from 1.4 to 2.0 × 10 −3 for the protonation and complexation experiments, respectively. A minimum of three independent runs were performed to determine the protonation constants, while four independent experiments were run for the copper(II) complexation constants.
Metal to ligand ratios of 1:1 were employed. The initial pH was always adjusted to 2.4. To avoid systematic errors and verify reproducibility, the electromotive force (EMF) values of each experiment were taken at different time intervals.
To obtain protonation and complexation constants, the potentiometric data were refined using Hyperquad [56], which minimizes the error square sum of the measured electrode potentials through a nonlinear iterative refinement of the sum of the squared residuals, U, and also allows for the simultaneous refinement of data from different titrations: where Eexp and Ecalc are the experimental and calculated electrode potentials, respectively. Errors in stability constant values are reported as three times standard deviations.
The formation reaction equilibria of ligands with protons and copper(II) ions are given in Equation (1): where L are the peptides under study. The stability constant βpqr is defined in Equation (2): where L are the peptides under study. The stability constant β pqr is defined in Equation (2): β pqr " rCu p H q L r s{rCus p¨r Hs q¨r Ls r (2) The species distribution as a function of the pH was obtained using the computer program Hyss [57].

Ultraviolet-Visible (UV-vis) Measurements
UV-vis spectra were recorded at 25˝C using an Agilent 8453 or a Varian Cary 500 spectrophotometer (Agilent Technologies, Santa Clara, CA, USA). The concentrations of the peptides and copper(II) used to record absorption spectra were the same as those for the potentiometric titrations. Combined spectroscopic and potentiometric metal-complex titrations were performed into a 3 mL quartz cuvette with a 1 cm path length to obtain the spectrum in the Visible region at each pH value simultaneously. These experiments were replicated at least three times for each copper-peptide system. Spectroscopic data were processed by means of HYPERQUAD program [56].

Circular Dichroism (CD) Measurements
CD spectra were obtained at 25˝C under a constant flow of nitrogen on a Jasco model 810 spectropolarimeter (Jasco, Easton, MD, USA) at a scan rate of 50 nm¨min´1 and a resolution of 0.1 nm, the path length being 1 cm, in the 280-800 nm range. The spectra were recorded as an average of either 3 or 5 scans. Calibration of the instrument was performed with a 0.06% aqueous solution of ammonium camphorsulfonate. The CD spectra of the copper(II) complexes on varying the solution pH were obtained in both the 190-250 and 250-800 nm wavelength regions. All the solutions were freshly prepared using double distilled water. The copper(II) ion and peptide concentrations used for the acquisition of the CD spectra in the Visible region were identical to those used in the potentiometric titrations. The results are reported as ε (molar adsorption coefficient) and ∆ε (molar dichroic coefficient) in M´1¨cm´1.

Nuclear Magnetic Resonance (NMR) Spectroscopy
NMR experiments were carried out on a Bruker Ascend TM 400 MHz spectrometer (Bruker, Billerica, MA, USA) equipped with a 5 mm automated tuning and matching broad band probe (BBFO) with z-gradients, as previously described [58,59]. NMR measurements were performed by using a concentration of the peptides of 2 mM, in 90/10 (v/v) H 2 O/D 2 O, at 298 K. HSQC (2D 1 H-13 C heteronuclear correlation spectra) were acquired using a phase-sensitive sequence utilizing Echo-Antiecho-TPPI gradient selection with a heteronuclear coupling constant JXH = 145 Hz, and shaped pulses for all 180˝pulses on f2 channel with decoupling during acquisition. Sensitivity improvement and gradients in back-inept were also used. In all of the experiments, relaxation delays of 2 s and 90˝pulses of about 10 µs were used. The solvent suppression in the 1D 1 H, 2D 1 H-1 H TOCSY and 2D 1 H-1 H ROESY experiments was performed by using excitation sculpting with gradients. The spin-lock mixing time of TOCSY experiments was obtained with MLEV17. 1 H-1 H TOCSY spectra were carried out using 60 ms as mixing times. The signals of both free and metal-bound peptides at different pH values and different metal to ligand molar ratios have been assigned by using a combination of 1D, 2D TOCSY, HSQC and ROESY experiments. All NMR results were processed by using TopSpin (Bruker Instruments) software and analyzed using Sparky 3.11 and MestRe Nova 6.0.2 (Mestrelab Research S.L.) programs (Santiago de Compostela, Spain).
Cells were plated in a 48-well plate and grown in DMEM-F12 up to 70%-80% confluency, followed by treatment for 24 h with each of the two peptides at the concentration of 5ˆ10´6, 1ˆ10´5 and 2ˆ10´5 M, as well as the correspondent complexes with CuSO 4 at equimolar concentrations. MTT (1 mg/mL) in PBS was then added to wells and incubated at 37˝C for 2 h to allow for complete cleavage of the tetrazolium salt by metabolically active cells. Next, MTT was removed and 220 µL of dimethyl sulfoxide (DMSO) was added. One hundred microliters of each well was transferred to a 96-well plate followed by colorimetric analysis using a multilabel plate reader at 560 nm of wavelength. Absorbance values plotted are the mean from triplicate experiments. Statistical relevance was calculated by One-way Analysis of variance (ANOVA) test performed with Origin 8.3 software (Northampton, MA, USA).

Confocal Microscopy Imaging
SH-SY5Y cells in culture at 80% of confluence were split on glass bottom Petri dishes (WillCo Wells, glass diameter of 22 mm). The day after cells were treated for 1 h with Ang(1-17), AcAng(1-17), wt-Ang, r-Ang, CuSO 4 , the peptide-copper and the protein-copper complexes at 1:1 mole ratio. The used concentrations were of 1ˆ10´7 M for the peptides and 1ˆ10´5 M for the proteins.
Confocal imaging was performed with an Olympus FV1000 confocal laser scanning microscope (LSM, Olympus, Shinjuku, Japan), equipped with diode UV (405 nm, 50 mW), multiline Argon (457, 488, 515 nm, total 30 mW), HeNe(G) (543 nm, 1 mW) and HeNe(R) (633 nm, 1 mW) lasers. An oil immersion objective (60xO PLAPO) and spectral filtering system were used. The detector gain was fixed at a constant value and images were taken, in sequential mode, for all the samples at random locations throughout the area of the well. Quantitative analysis of fluorescence was performed using the ImageJ software (1.50i version, NIH), in terms of integrated density ID = N¨ [M´B], where N is the number of pixels in the selection, M is the average gray value of the pixels and B is the most common pixel value [60].