Insights into the Recognition of Phosphate Groups by Peptidic

Subscriber access provided by TRINITY COLL

Article

Insights into the Recognition of Phosphate Groups by Peptidic Arginine from Action Spectroscopy and Quantum Chemical Computations Juan Ramon Aviles-Moreno, Giel Berden, Jos Oomens, and Bruno Martínez-Haya J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.9b06201 • Publication Date (Web): 15 Aug 2019 Downloaded from pubs.acs.org on August 27, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

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

The Journal of Physical Chemistry

Insights into the Recognition of Phosphate Groups by Peptidic Arginine from Action Spectroscopy and Quantum Chemical Computations Juan Ramón AvilésMoreno,† Giel Berden,‡ Jos Oomens,‡ and Bruno MartínezHaya∗,† †Department

of Physical, Chemical and Natural Systems, Universidad Pablo de Olavide, 41013 Seville, Spain

‡Radboud

University, Institute for Molecules and Materials, FELIX Laboratory, Toernooiveld 7c, 6525ED Nijmegen, The Netherlands

E-mail: [email protected]

1

ACS Paragon Plus Environment

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

Abstract The side group of the amino acid arginine is typically in its guanidinium protonated form under physiological conditions, and participates in a broad range of ligand binding and charge transfer processes of proteins. The recognition of phosphate moieties by guanidinium plays a particularly key role in the interactions of proteins with ATP and nucleic acids. Moreover, it has been recently identied as the driving force for the inhibition of kinase phosphorilation activity by guanidinium derivatives devised as potential anticancer agents. We report on a fundamental investigation of the interactions and coordination arrangements formed by guanidinium with phosphoric, phosphate and pyrophosphate groups. Action vibrational spectroscopy and ab initio quantum chemical computations are employed to characterize the conformations of benchmark positively-charged complexes isolated in an ion trap. The multidentate structure of guanidinium and of the phosphate groups gives rise to a rich conformational landscape with a particular relevance of tweezerlike congurations, where phosphate is eectively trapped by two guanidinium cations. The pyrophosphate complex incorporates a Na+ cation, which serves to compare the interactions associated with the localized versus diuse charge distributions of the alkali cation and guanidinium, respectively, within a common supramolecular framework.

INTRODUCTION The dierentiated interaction of guanidinium with molecular anions lies at the heart of the biochemical activity displayed by arginine-rich proteins in a broad range of enzymatic and metabolic processes.

1

The planar geometry of guanidinium and the delocalization of positive

charge among its ve NH

δ+

bonds, are particularly suitable for coordination with the diuse

anionic electronic structure of phosphates, carboxylates or sulfates. In fact, in most events of binding of these ions, guanidinium competes favourably with ubiquitous cationic species

+ + + such as Na , K or NH4 and protonated amines.

2

These features have served as inspira-


Page 2 of 27

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


tion for the development of synthetic guanidinium-based molecular materials for sensing and catalysis.

24

Recent studies in our group have corroborated the favourable interactions of

guanidinium with carboxylic side groups in a cyclic polyether.

57

In those systems, guani-

dinium displayed a capability for binding both to the ether and to the carboxylate groups of

+ the host macrocycle. Nevertheless, when guanidinium and protonated amine -NH3 groups were both present in the guest cation (as in the free amino acid arginine), guanidiniumcarboxylate and ammonium-ether interactions prevailed. Here, we focus on guanidinium-phosphate interactions, a topic of particular biochemical relevance. The ecient binding of guanidinium to phosphates promotes the recognition of nucleic acids and has led to important developments in areas of biotechnology related to sensing and transmembrane transport of biomolecular and genetic material.

816

Guanidinium-

phosphate interactions have gained particular recognition due to their key role in the control of ATP-mediated phosphorylation reactions, with remarkable implications in the design of novel anti-tumour agents. Guanidinium derivatives have been shown to be capable of acting as inhibitors of kinase enzymatic action, which is often over-activated in cancer cells.

17

Moreover, the phosphorilation of arginine residues has been identied as a key regulatory factor for protein degradation in Gram-positive bacteria.

18

This investigation exposes the versatile conformational landscape spanned by the lowenergy coordination arrangements of guanidinium with phosphoric, phosphate and pyrophosphate moieties in isolated supramolecular complexes. Fig. 1 represents the structure of the benchmark species involved in this investigation. Favourable ionic interactions largely follow from the matching between the spacing of the NH/NH2 groups of guanidinium and those

− of the P=O, POO , POH and POP groups.

This leads to a rich ensemble of low-energy

congurations that includes concerted and bifurcated coordination arrangements. The in-

+ corporation of Na cations to the coordination structure is as well considered in order to assess the preferential binding sites of the 'soft' guanidinium cation and the 'hard' alkali metal cation with the phosphate moieties.

3



Page 4 of 27

In this study, cationic complexes with an overall single charge are mass-selected and isolated in an ion trap at room temperature for interrogation by means of vibrational action spectroscopy. Details about the experimental and computational methodologies are provided below. It will be shown that this approach provides well-dened molecular benchmarks for the quantum chemical models, leading to fundamental insights into the intrinsic relative anities and the electronic structures responsible for guanidinium-phosphate interactions. The characterization of the preferred coordination arrangements in isolated complexes is expected to guide predictions of supramolecular behavior in solution. anion recognition actually constitute an active eld of research,

3,12,13

Solvent eects on and a controversial

issue that demands insights from experimental and computational contributions linking gasphase complexation with anion binding in solution and crystal phases.

19

In this context, the

ensemble of supramolecular conformations derived from this study should conform a valuable reference for the rationalization of guanidinium-phosphate recognition.

d´+

d+

G+

dPa0 d´´+

dP-

-

-

pP2-

-

+ Figure 1: Schematic representation of methyl guanidinium (G ) and of the benchmark phos0 phate species involved in the present investigation, diethyl-phosphoric acid (dPa ), diethyl − 2− phosphate (dP ) and pyrophosphate (pP ).

4


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


METHODS IRMPD spectroscopy The infrared multiple photon dissociation (IRMPD) spectroscopy experiments were carried out in a quadrupole ion trap mass spectrometer (Bruker AmaZon Speed), coupled to the free-electron laser FELIX.

20

IRMPD is a type of action spectroscopy that is based on the

detection of photofragments resulting from the sequential absorption of infrared photons at wavelengths resonant with vibrational transitions of the interrogated molecular ion.

21

The ionic complexes were produced by means of electrospray ionization of water/methanol solutions of methyl guanidinium hydrochloride and either one of the phosphate presursors, namely diethyl-phosphoric acid or pyrophosphoric acid.

The resulting ions were pulse in-

jected into the ion trap for storage at room temperature.

Mass selection of the following

three complexes with net charge +1 was performed in the present study:

1) dPa

0

·G+

2) dP

−

3) pP

2−

(m/z= 228)

·(G+ )2

(m/z= 301)

·(G+ )2 ·Na+

(m/z= 347)

+ 0 − 2− Here, G , dPa , dP and pP stand for methyl guanidinium, diethyl phosphoric acid, diethyl phosphate and pyrophosphate, respectively (see Fig. 1). In the IRMPD experiments, the mass-selected complexes are stored in the ion trap and thermalized with He background gas at room temperature. For this study, it will be assumed that thermal equilibrium is reached in the ion trap and Boltzmann population ratios associated with the energetics predicted by the quantum chemical computations will be indicated when discussing the potential contribution of the dierent conformers to the experimental

5



Page 6 of 27

0

dPa G

+

228

off resonance

G

+

74 at band C

50

100

150

200

-

250 +

dP (G )2 301

off resonance 0

dPa G 228

at band C

150

+

200

250

300

2-

+

+

pP (G )2Na

347

off resonance

-

+

+

pP G Na 274

at band C

200 Figure 2:

250

m/z

300

350

Illustrative IRMPD mass spectra obtained with the laser frequency tuned at

the vibrational band C of each of the three complexes object of this study (see Figs.3 + 5). The most intense product peaks registered during the scans, namely G (m/z= 74), 0 + − + + dPa ·G (m/z= 228) and pP ·G ·Na (m/z= 274), are related to guanidinium loss from the complexes (see text for details).

6


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


signal. It has been pointed out that solvent eects or operating conditions of the ion source may have an inuence on the relative population of conformers lying close in energy, occasionally demanding a statistical analysis of the computational structures more extended than the one performed here.

22

The stored ions were irradiated with 2 free-electron laser infrared macro-pulses.

Each

macropulse is approximately 5 microsecond long, has an energy of 20 to 70 mJ, and consists of a train of micropulses with a repetition rate of 1 GHz. The spectral bandwidth of the FELIX radiation is 0.5% of the central IR frequency. The IRMPD spectrum is generated from the normalized IR dissociation yield (Y), as determined from six averaged mass spectra, by plotting the fragment uence (dened as -ln(1-Y)) as a function of the IR frequency, with linear corrections of the fragment uence to account for changes in the laser pulse energy.

23

Fig. 2 depicts illustrative IRMPD mass spectra recorded for each of the three complexes. Fragment peaks are observed only when the laser frequency matches a vibrational transition of the complex. The dominant fragmentation products are in the three cases related to the release of guanidinium from the complex. The G

+

cation (m/z= 74) was observed as dissoci-

ation product for the three systems, although it could be monitored in the spectral scans only for the dPa

0

·G+

complex. For the two heavier complexes, dP

−

·(G+ )2

2− + + and pP ·(G )2 ·Na ,

the spectrometer required a low mass cut-o for optimal operating conditions that precluded a systematic registration of the G

+

cation. In these cases, the most intense IRMPD prod-

ucts involve a net loss of a G unit from the complex (i.e. the product retains a proton from guanidinium upon dissociation), leading to the cationic fragments dPa

0

·G+

(m/z= 228) and

− + + pP ·G ·Na (m/z= 274), respectively.

Quantum Chemistry calculations Ab initio

quantum chemical computations at the MP2/6-311++G(d,p) level were employed

to characterize the low energy conformations of the ionic complexes. The calculations were carried out with the Gaussian 09 code.

24

An initial ensemble of candidate molecular struc-

7



Page 8 of 27

tures was produced by means of simulated annealing with the Universal force eld, as implemented in the Materials Studio suite v7.0 (Accelrys Software). About 120150 non-redundant structures were produced for each complex, which were initially optimized with density functional theory at the B3LYP-D3/6-311++G(d,p) level (B3LYP functional with Grimme's D3 dispersion correction). The around 30 to 40 most stable conformers were subsequently reoptimized with the MP2 method. The conformers were ranked according to their relative congurational free energies (zero-point corrected electronic energies with thermal and vibrational contributions to entropy at T=298K). Natural bond orbital (NBO) analysis

25

was employed for a detailed characterization

of the electronic orbitals involved in the redistribution of charge associated with the ionic interactions in the dierent coordination arrangements of the complexes. The main focus of the NBO analysis is set on the

σ∗

anti-bonding orbitals of the -NH groups of guanidinium

and on the lone-pair orbitals of the oxygen atoms of the P=O, POH and POC groups, that are involved in the coordination arrangements. Two qualitative types of prbitals are found among the oxygen lone-pairs, namely a set of globular shaped orbitals (labelled for discussion) and a set of lobular shaped ones (labelled

n2 , n02 ,

n1 , n01 ,

etc,

etc).

The theoretical IR spectrum of each conformer was produced by convoluting the normal modes of vibration obtained in the MP2 computation with a line broadening of 25 cm

−1

(full

width at half maximum) and a scaling of the MP2 harmonic vibrational frequencies by a factor 0.97.

RESULTS The most salient conformational features of each of the complexes are discussed below, in the light of the experimental and computational information obtained in our study. The multipodal structure of guanidinium has the potential to build concerted interactions with several oxygen atom sites. This yields a rich conformational landscape and a challenging scenario for

8


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


the microscopic approach employed in this investigation. The ve conformations of lowest energy obtained for the three benchmark complexes investigated in this work are depicted in Figs. 3, 4 and 5. Those same Figures also show the IRMPD spectrum measured for each complex, along with the computational IR spectra associated with the dierent low-energy conformers. Moreover, an illustration of the NBO analysis of the main electronic orbitals involved in the most relevant supramolecular coordination arrangements is also provided.

Table 1: Qualitative assignment of the main bands observed in the IRMPD spectra of the three complexes investigated in this work, as derived from the dominant vibrational motions predicted by the MP2 computation. The bands are labelled AG, as indicated in Figs. 35. Band dPa0 ·G+ complex

mode assignment

A

+ C-N stretching, NH2 scissoring

B

CH3 umbrella, CH2 scissoring, N-H wagging

C

P=O and C-N stretching, NH2 rocking

D

stretching of P-O-C-C backbone

E

POH bending, O-P-O and C-C stretching

F

P-OH stretching, C-N3 symmetric stretching

G −

P-O-C stretching, CH3 rocking

dP ·(G )2 complex +

A

same as above

B

same as above

C

asymmetric POO stretching

D

P-O-C and symmetric POO stretching

E

stretching of P-O-C-C backbone, NH wagging

F 2−

P-O-C stretching, NH/NH2 wagging +

pP ·(G )2 ·Na complex A

+

same as above

B

same as above

C

asymmetric POO stretching

D

C-N and symmetric POO stretching, N2 rocking

E

POH bending

F

stretching of HO-P-O-P-OH backbone, NH wagging

9



Page 10 of 27

+0

Figure 3: Summary of the results obtained in this study for the dPa

dPa1dPa5

0

·G+

complex.

Top-

left: Conformations of lowest energy , with relative free energies indicated in −1 kJ·mol . Top-right: IRMPD spectrum measured for the complex (upper black trace) and IR spectra predicted by the MP2 computation for the dierent conformers (blue traces). See Table 1 for a qualitative assignment of the vibrational bands AG. Bottom: Dominant natural bond orbitals involved in the coordination arrangement of the complex (the relative contributions to the stabilization energy are indicated). Charge transfer occurs from globular (n1 ) and lobular (n2 ,n3 ) lone-pair orbitals of the oxygen atoms to the antibonding of the NH bonds of guanidinium.

10


σ∗ orbitals

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


+0

Figure 4: Summary of the results obtained in this study for the dP

−

·(G+ )2

complex. The

experimental and computational information shown is as described in the caption of Fig. 3.

11



Page 12 of 27

+0

Figure 5: Summary of the results obtained in this study for the pP

2−

·(G+ )2 ·Na+

complex.

The experimental and computational information shown is as described in the caption of Fig. 3.

12


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


dPa0 ·G+ complex 0 + The characterization of the dPa ·G complex serves to assess the preferred coordination achieved by the guanidinium N-H

δ+

bonds with the oxygen atom of the neutral P=O group,

in concurrence with the neighbouring oxygen atoms of the P-O-H and P-O-C moieties. Fig. 3 depicts the conformations of lowest energy predicted by the MP2 computation for the complex. The most stable conguration,

dPa1,

corresponds to a concerted coordination of

two NH bonds with the P=O group of the phosphoric acid. The conformers next in energy,

dPa2 and dPa3, are associated with individual coordinations of NH bonds with the oxygen atoms of the P=O group and of either the POH group or the phosphoester POC backbone. Finally, conformers

dPa4 and dPa5 represent cases in which guanidinium coordinates with

the oxygen atoms of the phosphoester and POH groups, with negligible interactions with the P=O group.

These two latter arragements are not favored by the MP2 computation

which assigns to them energies higher than 35 kJ·mol

−1

relative to

dPa1.

Hence, the MP2

computation points to a preferred coordination of guanidinium with the P=O bond, with alternative less stable coordinations with the POH or POC oxygen atom sites. The analysis of the natural bond orbitals involved in the dierent types of NH· · · O coordinations provides an intuitive framework for the rationalization of the stability of the low energy conformations of the complex. Fig. 3 depicts the most relevant NBOs involved in the

dPa1, dPa2, dPa3

conformers and indicates their corresponding relative contribution

to the stabilization energy. The formation of the complex involves the transfer of electronic density from lone-pair orbitals of the oxygen atoms of the phosphoric and phosphoester acid moieties, to the anti-bonding

σ*

orbitals of the N-H

δ+

bonds of guanidinium. Three types

of lone-pair orbitals participate in the coordination of the P=O group with the guanidinium cation, namely one of globular sp-like geometry, denoted n1 , and two virtually p-like lobular orbitals, labelled n2 and n3 . Electronic density in these orbitals partly drag charge from the bonding orbitals of the P=O bond. The interaction of guanidinium with the POH and POC groups receives contributions through the p-like lone pairs exclusively.

13


Interestingly, the


Page 14 of 27

P=O n1 →σ * charge redistribution provides the largest individual contribution to the hostguest interactions in the complex, namely 48% of the total stabilization energy in conformers

dPa1, dPa2,

and 40% in

dPa3.

These values are similar to the joint contributions from

the P=O n2 /n3 →σ * charge redistributions, each of which amounts to 1525% of the total stabilization energy. Fig 3 compares the IRMPD spectrum measured for the complex with the MP2 infrared spectra predicted for the low energy conformers

dPa1dPa5.

spectrum features a progression of bands in the 700-1800 cm

−1

The experimental IRMPD range that are assigned by

the MP2 computation to vibrational modes of the guanidinium and diethyl phosphoric acid frameworks, as described in Table 1. The prominent band A and the weaker band B on the high energy ank of the spectrum correspond to methyl guanidinium vibrations.

Diethyl

phosphoric acid participates in the vibrational motions associated with the remaining bands CG. In bands C and F, stretching vibrations of the P=O bond and of the P-OH bond couple to stretching and rocking motions of the C-N bonds and NH2 groups of guanidinium, respectively. Bands D, E and G are associated to stretching modes of the P-O-C-C backbone and to angular motions of the P-O-H and -CH3 groups. This assignment is supported by the excellent matching found between the vibrational bands of the IR spectrum of conformer

dPa1

and those of the IRMPD measurement. The spectrum of

dPa1

provides in fact the

best agreement with experiment among the low energy conformers of the complex. tributions to the recorded IRMPD signal from the

dPa2

and

dPa3

Con-

conformations are also

plausible and would in fact serve to explain the appreciable broadening of some of the bands bands, in particular bands A, C and D. The overall qualitative picture emerging from these results is a preferential coordination of guanidinium with the P=O group (conformer

dPa1), and a potential migration of one of the

guanidinium NH bonds to the oxygen atoms of the POH group (

dPa3)

(

dPa2) or the POC moiety

−1 with a moderate endothermicity (MP2 energies of 5 and 7 kJ·mol , respectively;

these relative energies would correspond to predicted Boltzmann population ratios of 0.13 and

14


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


0.06 for these conformers with respect to

dPa1 at the room temperature of our experiments.

It seems timely to point out that the position of band C predicted by the MP2 computation is somewhat red-shifted (ca. 3040 cm

−1

) with respect to the IRMPD experiment. This

band is closely related to vibrational modes involving the stretching of the P=O bond of the phosphoric acid group (Table 1). Interestingly, previous studies on related systems have reported similar red-shifts of the P=O stretching vibrational modes, leading to the conclusion that the computed frequencies of this type of mode did not require any scaling factor for comparison with experiment.

26

The use of a single eective scaling parameter for the

full vibrational spectrum, as applied here, is often reasonable within the moderate spectral resolution attained with a free-electron laser (in our case, up to 5 cm

−1

). Nevertheless, the

computed frequency of particular vibrational modes may deviate from the overall scaling, as it appears to be often the case for the P=O stretching vibration in phosphate and related groups. Within the scope of this study, such deviation is found for the dPa

2− + as just discussed, and also for the pP ·(G )2

·Na+

0

·G+

complex,

complex (see below). In contrast, the

computation of the P=O stretching modes for the dP

−

·(G+ )2

complex leads to a band C in

agreement with experiment within the framework of a general scaling factor. These ndings suggest that the accurate modelling of the vibrational modes of phosphate moieties demands specic approaches, depending on the molecular system and possibly on its environment. In order to further assess the perfomance of the B3LYP and MP2 quantum chemical methods, a comparison of the IR spectra predicted by the two types of computations has been included as Electronic Supplementary Information (Fig. S1) for all the systems and conformers discussed in the manuscript.

In general terms, the spectra predicted by both

approaches are coincident, notably including the P=O stretching modes that conform band C. The only signicant dierences found between the MP2 and B3LYP spectra are related to the splitting of the two components of band A (guanidinium C-N stretching) which is smaller in the B3LYP computation (in better agreement with experiment). For the dPa

0

·G+

complex, the B3LYP computation does not seem to describe accurately the position of band

15



Page 16 of 27

D (POCC backbone stretching vibrations), which is signicantly red-shifted with respect to the MP2 computation and the experiment. Hence, at least for the dPa

0

·G+

complex, the

MP2 approach is required for a reliable comparison wih the experimental IRMPD spectrum.

dP− ·(G+ )2 complex Deprotonation of the phosphoric acid facilitates the binding of two guanidinium cations, leading in our study to a ternary dP

−

·(G+ )2

complex with a net single positive charge.

In the conformations of lowest energy, depicted in Fig. 4, the phosphate moiety occupies the central part of the complex and the guanidinium cations provide and ecient tweezer like trapping environment for the anion. Fig. 4 shows that stable coordination arrangements arise from a variety of concerted interactions of the phosphate and phosphoester oxygen atom sites with the guanidinium cations. Rearrangements connecting the dierent congurations involve remarkably small net changes in energy, of less than

∼ 2 kJ·mol−1

(i.e., Boltzmann

population ratios of more than 0.45) for the rst four conformers.

− + The IRMPD spectrum of the dP ·(G )2 complex, shown in Fig. 4, shares the main features of the analogous spectrum of the dPa

0

·G+

complex. It is however less congested as a

consequence of the loss of vibrational modes associated with the POH group (see Table 1), in particular in the range below 1100 cm

−1

where isolated bands are neatly observed, in

contrast to the partially overlapping band structures of the phosphoric acid complex. The comparison of the IRMPD spectrum with the computational MP2 spectra sorts out the low energy conformers into two groups, depending on the level of agreement achieved. On the one hand, the MP2 spectra of conformers

dP1, dP3 and dP4 provide a good match to the

experimental bands throughout the whole spectral range investigated. On the other hand, those of conformers

dP2

and

dP5

display appreciable dierences with the experiment, in

particular for bands CF. These two latter conformers have in common a direct interaction of one of the guanidinium moieties with the phosphate POO coordination of two NH

δ+

−

group, through a concerted

bonds. Such coordination arrangement aects substantially the

16


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


stretching modes of the phosphate moiety, and the resulting MP2 spectra display sizeable discrepancies with the IRMPD measurement. This lack of agreement seems to rule out the concerted interaction of either one of the guanidinium moieties with the full POO

−

group.

Such a result is somewhat unexpected, as the concerted guanidinium-phosphate coordination interaction occurring in conformers

dP2 and dP5 could be anticipated to be among the most

stable ones, given the close matching between the distances between the oxygen atoms in the phosphate group (2.6 Å) and between the H atoms of the two NH groups of guanidinium (2.4 Å). In fact, conformation

dP2

is roughly isoenergetic with

the MP2 computation. The large stability predicted for the

dP1

−1 (within 1 kJ·mol ) in

dP2 conformer at the MP2 level

relies on a strong charge transfer interactions between the lone pair orbitals of the oxygen atoms of the phosphate group and the antibonding orbitals of the guanidinium NH bonds. This is shown in Fig. 4 which depicts the natural bond orbitals relevant to the coordination arrangaments of conformers

dP1 and dP2.

In

dP2, a large part of the stabilization energy

arises from the lobular p-like lone pair orbitals of the POO

−

group, which yield particularly

strong interactions with guanidinium. The contribution from the more isotropic sp-like globular lone pair orbitals is somewhat less relevant but also quite signicant. In the light of the comparison of the MP2 infrared spectra with the IRMPD measurement, it appears that the MP2 computation overestimates these interactions. The present results support instead the individual coordination of each of the guanidinium cations with the oxygen atoms of the PO and POC groups of the phosphate anion observed in conformations

dP1, dP3 and dP4.

pP2− ·(G+ )2 ·Na+ complex The extended multipodal structure of the doubly charged pyrophosphate anion pP

2−

is ca-

pable of binding two guanidinium cations in tweezerlike coordination arrangements with distinct features with respect to those just described for the dP

−

anion. Moreover, since our

study probes complexes with net positive charge, the incorporation of a third cation is required for detection. The complex with three guanidinium cations was not observed under the

17



Page 18 of 27

present working conditions; the electrospray ionization of the pyrophosphate/guanidinium mixture did not produce an appreciable signal at m/z= 398.

However, it did yield an in-

tense signal for the complex incorporating two methyl guanidinium units and one Na

+

cation

+ (m/z= 347). Since Na is ubiquitous in biochemical environments, the characterization of the pP

2−

·(G+ )2 ·Na+

complex should be of interest to assess the potential competition be-

tween the alkali cation and guanidinium for binding with the pyrophosphate moiety, possibly weakening the eciency of the peptidic "guanidinium tweezer". From a fundamental point of view, it also serves to compare the diuse interactions of guanidinium with the more

+ localized charge distribution of Na , within a common supramolecular framework. Fig. 5 depicts the ve most stable MP2 conformations of the complex. Favourable coordination arrangements arise from the ecient guanidinium-phosphate and sodium-phosphate intermolecular interactions and the optimum separation of the three cationic moieties in the

pP1, appears to be particularly stable; all other

complex. A singular conguration, namely conformations are more than 15 kJ·mol

−1

higher in energy (hence, a negligible Boltzmann

population ratio, well below 0.01 at room temperature). Closer lying conformers were not found despite the fact that our conformational survey for this complex was particularly thorough (up to 150 and 40 candidate conformations were computed at the B3LYP and the MP2 levels, respectively). In

pP1,

the two guanidinium cations build identical concerted coordinations with PO

bonds from the two phosphate ends.

The spacing imposed by the phosphoanhydride P-

O-P backbone leads to a separation between the oxygen atoms of the coordinating PO bonds of 3.1 Å. This distance is somewhat larger than the separation between the NH bonds of guanidinium (2.4 Å) but it leads to favourable coordination bonds with PO· · · H angles of

∼ 125o

and an ecient overlap of the lone pair orbitals of the oxygen atoms with the

antibonding orbitals of the guanidinium NH groups. The natural bond orbital representation of Fig. 5 shows that the charge redistribution induced by these orbitals account for around 75% of the total stabilization energy of the

pP1 conformer. 18


The Na

+

cation accommodates

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


at a position equidistant and coplanar with the other two PO bonds, with a distance 2.2 Å

o from the O atoms and P-O· · · Na bond angles of 125 . Charge transfer occurs in this case dominantly from the lobular lone pair orbitals of the oxygen atoms to the antibonding

σ∗

orbital of the Na

towards the Na

+

+

cation.

The relative contribution of the delocalization of charge

cation accounts for only 10% of the total stabilization energy. This nding

stresses the more favourable interactions arising from the diuse distributions of charge of the molecular ions, guanidinium and pyrophosphate, in comparison to those involving the more localized charge of the alkali cation. It should be noted that the coplanar coordination of Na

+

with the PO bonds is in contrast with the V-shaped and more directional coordination

of the two guanidinium cations. This feature poses an important constraint to the shared coordination of guanidinium and Na

+

with the same pair of PO bonds, due to mutual

repulsion between the two cations. Unlike in

pP1, in none of the higher energy conformers pP2pP5 the two guanidinium

cations bind to coincident oxygen atoms.

In

pP2,

the guanidinium cations coordinate to

complementary pairs of PO phosphate bonds across the POP backbone. In

pP3 and pP4,

one of the guanidinium cations moves partially away from the phosphate moieties and binds to an oxygen atom of either the POP backbone, or the POH bond. In

pP5, the two guani-

dinium cations move to a peripheral position and interact separately with single PO bonds of each of the phosphate groups. In

pP4

and

pP5,

the Na

+

cation displays a coordination

arrangement, coplanar with two PO bonds, similar to the one described above for contrast, in

pP2

and

pP3

the Na

+

pP1.

In

cation interacts with three oxygen atoms across the

backbone of the anion, in a coordination arrangement facilitated by the a relative rotation of the two phosphate groups. The analysis of the IRMPD experiment for the pyrophosphate complex is consistent with the prediction of

pP1 as the most stable conformer.

The computational IR spectrum for this

conformer is in excellent agreement with the band sequence observed in the measurement and it may concluded that

pP1 captures the fundamental structural features of the complex. 19


A


Page 20 of 27

detailed inspection of the IRMPD spectrum also reveals an appreciable broadening of some of the vibrational bands, in particular in bands A and C, which points to the participation of additional conformers. The consideration of

pP2pP5 could in fact account for the observed

broadenings. Unfortunately, some discrepancies are evident between their MP2 spectra and the experiment, for instance band D in conformers

pP3, pP4

and

pP5,

pP2 and pP5, and band E in conformers

do not seem to be properly described. In view of these observations,

and despite the extensive conformational search perfomed in this study, the existence of additional low-energy conformations, that would not have been found in our survey, seems possible. Such conformers would plausibly maintain the general features of the arrangements described here.

SUMMARY AND CONCLUSIONS The microscopic background that supports the recognition of phosphate moieties by guanidinium side groups has been investigated.

For this purpose, the benchmark complexes

formed by methyl guanidinium with diethyl phosphoric acid, dPa

0

·G+ ,

diethyl phosphate,

− + 2− + + dP ·(G )2 , and pyrophosphate, pP ·(G )2 ·Na , have been characterized under isolated conditions with action spectroscopy and quantum chemistry modelling. Favourable coordination arrangements and conformations have been assigned for each of the supramolecular complexes, based on

ab initio MP2 computations for the structure and interactions, and the

comparison of the predicted vibrational spectrum with the IRMPD bands observed experimentally. The study of the binary dPa

0

·G+

guanidinium with the P=O group.

complex has exposed a preferential coordination of

A propensity for conformational transitions, leading

to a partial binding of guanidinium with the oxygen atoms of the POH or POC groups, is predicted at a moderate energetic cost. Such transitions may confer the complex adaptibility to changing environmental conditions and steric eects. The stabilization of these low energy

20


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


coordination arrangements has been found to be sustained by the redistribution of charge between splike lobular and p-like globular lone pair orbitals of the oxygen atoms and the antibonding

σ*

orbitals of the NH bonds of guanidinium.

The characterization of the ternary dP

−

·(G+ )2

complex has revealed a general coordina-

tion landscape dominated by tweezerlike conformations in which the guanidinium cations bind at roughly opposite sites of the phosphate group and provide an ecient trapping environment for the anion. An ensemble of roughly isoenergetic stable conformations are predicted by the MP2 computation for this complex. The best agreement with the IRMPD measurement is found for the conformers involving a coordination of guanidinium with the oxygen atoms of the PO

δ−

and POC groups. Conformations in which a single guanidinum

− cation coordinates with the full POO anionic group in a concerted arrangement of two of δ+ its NH groups, while rendered stable by the MP2 computation, seem to be ruled out due to discrepancies with the IRMPD spectrum, in particular for the vibrational band associated

− with the asymmetric stretching of the POO group. The quaternary pP

2−

·(G+ )2 ·Na+

like coordination congurations.

complex displays as well a variety of stable tweezer

The extended bi-anionic structure of pyrophosphate and

+ the presence of the Na cation generate a rich congurational space in which the three cations compete for coordination while trying to minimize mutual repulsions. The guani-

δ− dinium cations tend to bind to a pair of PO bonds of dierent phosphate groups across the phosphoanhydride P-O-P backbone. The separation between the oxygen atoms of the coordinating PO bonds (3.1 Å) and that between the NH bonds in guanidinium (2.4 Å) lead to favourable PO· · · H coordination angles for the ecient overlap of the lone pair orbitals of the oxygen atoms and the antibonding orbitals of the guanidinium NH groups. This feature should be of general relevance for the coordination of guanidinium and pyrophosphate groups in complex systems. The Na

+

cation may bind either in a coplanar conguration with two

PO bonds or, alternatively, in interaction with three oxygen atoms across the pyrophosphate backbone assisted by a relative rotation of the phosphate groups. The MP2 computation pre-

21



Page 22 of 27

dicts a distinctly stable conformation, associated with a coplanar coordination of Na

+

with

δ− two of the PO bonds, that forces a concerted coordination of the guanidinium cations with the other pair of PO

δ−

bonds in an overall V-shaped local conguration. The infrared spec-

trum of this conformer matches accurately the progression of vibrational bands displayed by the IRMPD measurement, which supports that it captures the main structural properties of the most stable coordination arrangements for the complex. The supramolecular conformations described in this study for isolated complexes should guide the microscopic rationalization of guanidinium-phosphate binding in more complex environments. The assessment of solvent eects in anionic chemistry is however a challenging and still unsolved issue.

19

Interestingly, we nd a qualitative similarity between the coor-

dination arrangements found in our study with those derived from recent experiments on arginine-phosphate recognition in an aqueous environment.

27

It may be tentatively inferred

that, while ion solubility likely aects the equilibrium constants of the complexation process, the preferred guanidinium-phosphate coordination arrangements in solution seem to resemble qualitatively those found in isolated complexes. Work currently in progress in our group seeks to extend the present investigation to progressively more complex benchmark anionic substrates. These will incorporate sugar motifs (ribose or glucose) and will eventually include the AMP/ADP/ATP family. The interaction of methyl guanidinium with these systems should provide valuable information to rationalize the coordination mechanisms and the conformational contraints underlying their recognition by Arg-containing peptides/proteins in biological environments.

Acknowledgement This work constitutes a modest tribute to Prof.

F.J. Aoiz, an outstanding scientist and

a source of continuous inspiration; BMH is particularly thankful for his mentorship dur-

22


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


ing important stages of his research carreer. This study is part of project P12-FQM-4938 of Junta de Andalucia-FEDER (Spain) and has support from the project CALIPSOplus (Grant Agreement 730872, EU Framework Programme for Research and Innovation HORIZON 2020). We thank C3UPO for the HPC facilities and the Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO) for the support of the FELIX Laboratory.

Supporting Information Available The following Supporting Information is provided:

Aviles-ESI-FigS1.pdf: Comparison of the experimental IRMPD spectra with the infrared spectra predicted by the B3LYP and MP2 computations for the ve low energy conformers of the three complexes investigated in this study.

References (1) Chandana, Th.; Venkatesh, Y.P., Occurrence, functions and biological signicance of arginine-rich proteins. Curr. Prot. Pept. Sci.

2016, 17, 507-516

(2) Houk, R.J.T.; Tobey, S.L.; Anslyn, E.V., Abiotic guanidinium receptors for anion molecular recognition and sensing. Top Curr. Chem.

2005, 255, 199-229.

(3) Schug, K.A.; Lindner, W., Noncovalent binding between guanidinium and anionic groups:

focus on biological- and synthetic-based arginine/guanidinium interactions

with phosphonate and sulfonate residues. Chem. Rev.

2005, 105, 67-104.

(4) Blondeau, P.; Segura, M.; Pérez-Fernández, R.; de Mendoza, J., Molecular recognition of oxoanions based on guanidinium receptors. Chem. Soc. Rev.

2007, 36, 198-210.

(5) Avilés-Moreno, J.R.; Berden, G.; Oomens, J.; Martínez-Haya, B. Isolated complexes of

23



Page 24 of 27

the amino acid arginine with polyether and polyamine macrocycles, the role of proton transfer. Phys. Chem. Chem. Phys.

(6) Avilés-Moreno,

J.R.;

Berden,

2017, 19, 31345-31351.

G.;

Oomens,

J.;

Martínez-Haya,

B.

Guani-

dinium/ammonium competition and proton transfer in the interaction of the amino acid arginine with the tetracarboxylic 18-crown-6 ionophore Phys. Chem. Chem. Phys.

2018, 20, 4067-4073 (7) Avilés-Moreno, J.R.; Berden, G.; Oomens, J.; Martínez-Haya, B. Complexes of crown ether macrocycles with methyl guanidinium: Insights into the capture of charge in peptides. ChemPhysChem

2018, 19, 2169-2175

(8) Wexselblatt, E.; Esko, J.D.; Tor Y. On guanidinium and cellular uptake J. Org. Chem.

2014, 79, 6766-6774 (9) Hargrove, A.E.; Nieto, S.; Zhang, T.; Sessler, J.L.; Anslyn, E.V. Articial receptors for the recognition of phosphorylated molecules. Chem. Rev.

2011, 111, 66036782

(10) Berlicki, L.; Rudzinska, E.; Mlynarz, P.; Kafarski P. Organophosphorus supramolecular chemistry part 1. Receptors for organophosphorus compounds. Curr. Org. Chem.

2006, 10, 2285-2306. (11) He, J.; Hu, P.; Wang, Y.J.; Tong, M.L.; Sun, H.; Mao, Z.W.; Ji, L.N. Double-strand DNA cleavage by copper complexes of 2,2-dipyridyl with guanidinium/ammonium pendants. Dalton Trans.

2008, 24, 3207-3214

(12) Kataev, E.A.; Müller, Ch.; Kolesnikov, G.V.; Khrustalev, V.N. Guanidinium-based articial receptors for binding orthophosphate in aqueous solution. Eur. J. Org. Chem.

2014, 13, 2747-2753. (13) Trujillo, C.; Previtali, V.; Rozas, I. A theoretical model of the interaction between

24


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


phosphates in the ATP molecule and guanidinium systems. Theor Chem Acc

2016,

135, 260 1-12.

(14) Yusufaly, T.I.; Li, Y.; Singh, G.; Olson W.K. Arginine-phosphate salt bridges between histones and DNA: Intermolecular actuators that control nucleosome architecture. J. Chem. Phys.

2014, 141, 165102 1-12.

(15) Herce, H.D.; Garcia, A.E.; Cardoso, M.C. Fundamental molecular mechanism for the cellular uptake of guanidinium-rich molecules. J. Am. Chem. Soc.

2014,

136, 17459-

17467.

(16) Vargas,

J.R.;

Stanzl,

E.G.;

Teng,

N.N.H.;

Wender,

P.A.

Cell-penetrating,

guanidinium-rich molecular transporters for overcoming eux-mediated multidrug resistance. Mol. Pharm.

2014, 11, 2553-2565.

(17) Diez-Cecilia , E.; Kelly , B.; Perez , C.; Zisterer, D.M.; Nevin, D.K.; Lloyd , D.G.; Rozas, I. Guanidinium-based derivatives: searching for new kinase inhibitors . Eur. J. Med. Chem.

2014, 81, 427-441.

(18) Suskiewicz, M.J.; Clausen, T. Chemical biology interrogates protein arginine phosphorylation. Cell Chem. Biol.

2016, 23, 888-890.

(19) Liu, Y.; Sengupta, A.; Raghavachari, K.; Flood, A.H. Anion binding in solution: beyond the electrostatic regime. Chem

2017, 3, 411-427.

(20) Martens, J.; Berden, G.; Gebhardt, Ch.R.; Oomens, J. Infrared ion spectroscopy in a modied quadrupole ion trap mass spectrometer at the FELIX free electron laser laboratory. Rev. Sci. Instrum.

2016, 87, 103108 1-8.

(21) Rijs, A.; Oomens, J. Gas phase IR spectroscopy and structure of biological molecules. Top. Curr. Chem.

2015, 364, 1-42.

25



Page 26 of 27

(22) Ma, L.; Ren, J.; Feng, R.; Zhang, K.; Kong, X. Structural characterizations of protonated homodimers of amino acids: Revealed by infrared multiple photon dissociation (IRMPD) spectroscopy and theoretical calculations. Chin. Chem. Lett.

2018

, 29,

1333-1339.

(23) Berden, G.; Derksen, M.; Houthuijs, K.J.; Martens, J.; Oomens, J. An automatic variable laser attenuator for IRMPD spectroscopy and analysis of power-dependence in fragmentation spectra. Int. J. Mass Spectrom.

2019, 443, 1-8.

(24) Gaussian 09, Revision E.01, M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson, et al, Gaussian, Inc., Wallingford CT, 2009.

(25) Foster, J.P.; Weinhold, F. Natural hybrid orbitals. J. Am. Chem. Soc.

1980,

102,

7211-7218.

(26) Outersterp, R.E; Martens, J.; Berden, G.; Steill, J.D.; Oomens, J.; Rijs, A.M. Structural characterization of nucleotide 5'-triphosphates by infrared ion spectroscopy and theoretical studies Phys. Chem. Chem. Phys.

2018, 20, 28319-28330

(27) Neal, J.F.; Zhao, W.; Grooms, A.J.; Flood, A.H.; Allen, H.C. Arginine-phosphate recognition enhanced in phospholipid monolayers at aqueous interfaces. J. Phys. Chem. C

2018, 122, 26362-26371

26


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


Graphical TOC Entry

27


Insights into the Recognition of Phosphate Groups by Peptidic

Recommend Documents