Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Journal of Molecular Structure, 240 (1990) 19-29 ElsevIer SCIence Pubhshers B V , Amsterdam 19 HYDROGEN-BONDING EFFECT ON 15N NMR CHEMICAL SHIFTS OF THE GLYCINE RESIDUE OF OLIGOPEPTIDES IN THE SOLID STATE AS STUDIED BY HIGH-RESOLUTION SOLID-STATE NMR SPECTROSCOPY SHIGEKI KUROKI, SHINJI ANDO, ISAO ANDO Department of Polymer Chemistry, Tokyo Institute of Technology, Ookayama, Meguro-ku, 152, Tokyo (Japan) AKIRA SHOJI and TAKUO OZAKI Department of Industnal Chemistry, College of Technology, Gunma Unwersity, TenJm-cho, KiryU, 376, Gunma (Japan) GRAHAM A WEBB Department of Chemistry, Unwersay of Surrey, Gwldford, Surrey (Gt Bntam) (ReceIved 19 February 1990) ABSTRACT HIgh-resolutIOn I5N NMR spectra of a vanety ofsohd ohgopeptIdes (X-Gly-Gly) contammg glycme resIdues have been measured, of whIch the crystal structures had already been determmed by X-ray dIffractIOn The expenmental '5 N chemIcal shIfts of the glycme reSIdues were plotted agamst the N"'O hydrogen bond length and the N-H bond length m the :-C=O··· H-N; type hydrogen bond form, respectively It was found that the decrease of the N - H bond length leads to a hnear mcrease m ION shIeldmg, but there IS no clear relatIOnshIp between the I5N chemIcal shIfts and the N· .. 0 separatIOn Further, I5N chemIcal shIft calculations were carned out usmg a model compound, by the FPT -INDO method, m order to further understand the nature of the hydrogen bond The calculated results reasonably explam the expenmental ones INTRODUCTION Recently, high-resolution 15N NMR spectroscopy in the sohd state has been mcreasingly applied to the investigation of polypeptides, protems and biopolymers [1-12]. In a previous paper [13], we demonstrated that the isotropic 15N chemical shifts of a number of homopolypeptides in the solid state, as determined by the cross polarization-magic angle spinning (CP -MAS) method, are significantly displaced according to their particular conformations such as the a-helix and fJ-sheet forms. Moreover, it was found that the 15N chemical shift dIfference between the a-helix andfJ-sheet forms for such homopolypeptides in the solid state varies by as much as 1.2-10.0 ppm, depending mamly 0022-2860/90/$0350 © 1990 - ElsevIer SCIence Pubhshers B V 20 on the nature of the amino acid residue [13] Such a large chemical shift dIfference may come from changes m the hydrogen bond through changes in the dihedral angles (rp, !If). We have also measured 13C CP-MAS NMR spectra of some oligopeptides contammg glycine residues in the solid state [14], in order to obtam mformation about the relationshIp between the carbonyl carbon chemIcal shift and hydrogen bond length. It was found that the 13C sIgnals ofthe carbonyl carbons in the ~C=O··· H-N( type hydrogen bond form are deshieided wIth a decrease in the hydrogen bond length but those in the ~C=O··· H-N+E type hydrogen bond form are shielded with a decrease in the hydrogen bond length Quantum chemIcal calculations of the 1,C shiel dings for model compounds were performed and found to reproduce reasonably the experimental results, taking into account the hydrogen bond and conformatIOnal effects [14]. In this work, in order to obtain and accumulate further knowledge of the hydrogen bonding in peptides, we attempt to measure 15N CP-MAS NMR spectra of the oligopeptIdes containing glycine residues in the solId state and to clarify the origm of the relatIOnship between the 15N chemical shift and the manner of the hydrogen bond. Further, m an attempt to obtain a deep msight mto the nature of the hydrogen bond, we calculate the 15N shieidings and tensor components of the glycine amide nitrogens by employmg quantum-chemIcal methods. EXPERIMENTAL SECTION Matenals A series of oligopeptides containing glycine resIdues, except for sarcosylglycylglycme (Sar-Gly-Gly), were purchased from Sigma Co. and were recrystallized according to the same procedures as those used in the X-ray dIffractIOn studies on them. Sar-Gly-Gly was synthesIzed by stepwise elongatIOn of Nhydroxysuccinimide active esters and amino acids [15]. The N -terminal ofthe active ester was protected by the o-mtrophenylsulfenyl (Nps-) group. This sample was purified and recrystallized from aqueous solution. Glycylglycine nitrate (Gly-Gly· HN0 3 ) was obtained by slow evaporatIOn from an equimolar mixture of glycylglycine and nitric acid m water. Glycylglycine monohydrochloride monohydrate (Gly-Gly· HCI· H 20) was obtained by slow evaporation from an equimolar mixture of glycylglycine and hydrogen chlonde in water. Measurements 15N CP-MAS NMR spectra were recorded at room temperature and 27.25 MHz with a JEOL GSX-270 NMR spectrometer equipped wIth a CP-MAS accessory. Field strength of the IH decoupling was 1.2 mT. The contact time was 5 ms, and the repetition time was 10 s. Spectral width and data points were 21 20 kHz and 8k, respectIvely. Samples were placed in a cylindrical rotor and spun at 4-5 kHz. Spectra were usually accumulated 100-600 times to achieve a reasonable signal-to-noise ratio. 15N chemical shIfts were calibrated mdIrectly through external glycine- 15 N (6 = 11.59 ppm; line wIdth = 17Hz) relatIve to saturated 15NH4N03 (6=0 ppm) solution in H 2 0. Theoret~cal calculatwn We employed the fimte perturbation theory (FPT) wIthin the INDO framework for calculating the 15N shIeldmgs. The FPT INDO theory has the advantage of permitting the calculation of the paramagnetic term wIthout reqmrmg the explicit wavefunctIOns of the excited states and the one-electron excitation energies, whIch are difficult to obtam wIth high accuracy by the usual semIempIrical MO approxImations. This approach reproduces reasonably the expenmental 13 C chemIcal shifts of L-alanme residues in peptides [16]. According to the FPT INDO framework [17-19], (Jdap(A) (diamagnetic term) and (Jp(tP(A) (paramagnetic term) are expressed by a,p=x,Y,Z where the gauge ongin of the vector potential IS set at the posItIOn of nucleus A The vectors r v and r A are the position vectors of the electron consIdered from the nucleus of the atom containing the atomIC orbital (AO) X,, and from the nucleus A, respectively. P J1V (B) and PI'" (0) are the elements of the density matnx with and witlJ.out the perturbation due to the magnetic field, B, respectively. The 15N shielding calculatIOn was carned out by a similar procedure to that reported previously for a 13C shielding calculation [13]. In the calculation, we adoptedN-acetyl-N' -methylglycme amide as a model compound. The bond lengths and bond angles proposed by Momany et al. [20] were used. A HITAC M780H computer at the Computer Center of the Tokyo Institute of Technology was used for the calculation The N - H bond length and N· . ·0 hydrogen bond length, optimIzed by an ab imtio STO-3G MO calculation, were determmed for N-methylacetamide as a model compound. The calculatIOns were performed on a HITAC M680H computer at the Computer Center of the Institute for Moleculer Science, Okazaki. 22 RESULTS AND DISCUSSION 15N NMR chem~cal shifts of glycme amide mtrogens of pept~des m the solld state A 27.25 MHz 15N CP-MAS NMR spectrum ofL-alanylglycylglycine (L-AlaGly-Gly) in the solid state is shown as a typical example in Fig. 1. 15N CPMAS spectra of the remaining samples were also obtained with sImilar resolutions. SIgnals were also assigned wIth reference to prevIOUS 15N CP-MAS NMR and solution state 15N NMR data [21]. We use selected glycylglycine (X-Gly-Gly) sequence ohgopepbdes in this work. It is well known that the 15N chemical shifts of pep tides vary with the amino acid sequence [22]. For example, 15N chemical shift of the glycine residue depends upon the ammo acid residue linked to the N-termmal ofthe glycine residue. Consequently, we use the glycylglycine sequence ohgopeptides, in order to neglect the sequence-dependent 15N chemical shifts. All the 15N chemical shift values of oligopeptides determined from the observed spectra are tabulated in Table 1, together with the geometrical parameters obtained by X-ray dIffraction studies. Some of the geometrical parameters were calculated by usmg the unit cell parameters and fractIOnal coordmates given in the literature [23-31]. The 15N chemical shifts and the geometncal parameters of tert-butyloxycarbonyl-glycylglycylglycme-benzylester (Boc-G lyGly-Gly-OBzl) are inferred from the results of Hlyama et al. [31]. Figure 2 shows the plot of the observed 15N chemIcal sh1fts ofGly NH against the hydrogen bond length N· .. 0 (R N 0)' However, it is found that there is no clear relationship between RN 0 and 15N chemical shifts. This is dIfferent Ala GliGly2 Flg 1 A typlcal 27 25 MHz !5N CP -MAS NMR spectrum of L-alanylglycylglycme m the crystallme state 23 TABLE 1 15N chemIcal shIft of glycme resIdue amIde mtrogens for ohgopeptJdes contaInmg glycme as determmed by 15N CP-MAS NMR (± 02 ppm from 15NH4N03) and the geometrIcal parameters of X-Gly-Gly peptIdes 15N chemIcal shIft", t5 (ppm) Sample GlyGly*OH GlyGlY*'H 2 0 HCI GlyGly*OH' HN0 3 VaiGlyGly*OH ProGlyGly*OH AlaGlyGly*OH SarGlyGly*OH TyrGlyGly*OH BocGlyGlyGly*OBzl 9995 9095 8971 9443 8944 8855 8900 9230 9275 c GeometrIcal parameters DIhedral angle (deg) ¢ If! 1571 -796 1656 -1466 -1105 1709 -854 -1039 -778 107 38 1769 -43 1751 1752 -177 5 -1526 -1784 N· H"'O N-H (A) (A) N-H ·0 (deg) Ref (A) 0 294 330 b 312 305 289 300 306 288 292 197 230 238 219 207 219 223 213 210 102 079 076 093 085 084 085 086 092 158 162 165 152 165 160 167 144 157 23 24 25 26 27 28 29 30 31 "These values are for the Gly resIdues wIth an asterIsk bThe glycIne resIdue partIcIpates In an H type hydrogen bond cThls value IS converted to the 15NH4N01 reference from ref 31 0 0 -0 ~90 'CI 0 0 Go L U 0 0 Z 0 '" 0 95 0 28 29 30 31 32 33 N 0 length(A) FIg 2 Plots of the observed 15N shleldmg of oilgopeptIdes m the crystallme state agamst the N· . ·0 hydrogen bond length from the case of 13C chemical shifts of the carbonyl carbons assocIated with the hydrogen bond which we reported previously [14] where the 13C sIgnals of the carbonyl carbons are linearly desh18lded WIth a decrease in RN o. Figure 3 shows the plot of the observed 15N chemical shifts (JobsN) of the 24 ~ 85 2 ::C :f- g 0' to 90 0. 0. \ o 0 0 \0, , , 0'\0 E , if) -0 u ~ .J:: \ 0\, , 95 u , \0 :f- 06 07 08 09 10 11 N- H lergth(A) FIg 3 Plots of the observed N-H bond length 15N shIeldmg of ohgopeptldes m the Cf) stalhne state agamst the glycme resIdue in X-Gly-Gly against the N-H bond length (R N - H ) assocIated WIth the hydrogen bond. It is found that there IS a clear relatIOnship between these parameters and the decrease of RN - H leads to a linear increase in shIelding. The expression for this relationshIp is Jobs N = 39 32R N _H + 57.73 Such a trend is very dIfferent from that obtamed from 1JC NMR. Amide 15N chemIcal shifts are closely related to the length of the N-H bond but are not related to the N· . ·0 hydrogen bond length. This implies that the ION chemical shift value gIves useful information about the N-H bond length m the hydrogen bond. It seems that the hydrogen bond angle (L N - H .. ·0) IS also related to the 15N chemIcal shift Ab m~tw MO calculatwn relatwnsh~p between the N-H bond length and the N···O hydrogen bond length The structure of the two hydrogen-bonded N-methyl acetamides used as a model system is shown in Fig. 4. At first, the geometrical parameters of an Nmethyl acetamide molecule were optimized usmg the ab ImtIO STO-3G MO method Next, for two hydrogen-bonded N-methyl acetamides, the bond length N - H was optimized as a function of the hydrogen bond length N· . ·0 FIgure 5 shows the relationshIp between the mmimized bond length N-H and the hydrogen bond length N· . ·0, as determined by ab mitio MO calculatIOns It IS shown that at RN 0 < 2.97 A, an mcrease of RN 0 leads to an mcrease 25 ~ H H \ N-C----H / I H _C'--- / \ C H H II ? : H H / H-C I~ I H /N""",\ C-H H C II o / H FIg 4 Molecular structure of the two hydrogen-bonded N-methylacetamldes used as model compounds 10200 N-H "4 :<: g ~ 10100 :J: I Z 10000 24 26 N 28 30 o length (A) 32 34 FIg 5 Plots of the calculated N-H bond length agamst the hydrogen bond length (N·· 0) obtamed usmg the ab InItIO STO-3G MO method of RN- H • However, at RN 0> 2 97 A, an Increase of RN 0 leads to a decrease of RN-H • In the samples used in thIS work, the RN 0 values are between 2.85 and 3.30 A. Therefore, it can be said that the R N - H values decrease wIth an Increase of RN o. On the other hand, X-ray dIffractIOn studIes, have shown that the R N - H values decrease with an Increase of the RN 0 values [32] From the above results, It can be SaId that there is an apparent relationship between the hydrogen bond length RN 0 and the bond length RN- H • 26 15N shieldmg calculatwns FIgures 6 and 7 show the calculated isotropIc shieldings (alSO) and the paramagnetic terms of the tensor components (all> a2b a3 .3, from downfield to upfield) ofGly NH using the model compoundN-acetyl-N' methylglycine (Fig. 8). The calculated values are all expressed in parts per millIon (ppm) with an OpposIte sign to those in Table 1. Note that the negative sign for the calculated E -300 (l ~ C .9 -305 ~ (3 ~ -310 ~ §, -315 f2 ;'§ -320 -325 N- H length(A) FIg 6 Plots of the calculated 15N shIeldmg agamst the N-H bond length from the FPT INDO method E 0. 6 -490 -500 -510 E0. E0. ':'250 -:140 tS-260 t5-150 -270 160 -280 -170 -290 180 -300 -190 0. 2- ~ 0. -310 08 09 10 11 N-H length (A) 08 09 10 11 N-H length (A) 08 09 10 11 N-H length (Ai FIg 7 Plots of the calculated 15N shIeldmg tensor components agamst the N - H bond length from the FPT INDO method 27 FIg 8 Molecular structure of N-acetyl-N' -methylglycme amIde used as the model compound FIg 9 OnentatlOn of the prmclpal axes of the mtrogen as determmed m the lIterature [31] 15N shleldmg tensors of the glycme resIdue amIde shieldmg denotes deshieldmg, which is similar to the positive sign of the experimental chemical shift values. A shIelding value, or tensor component, is usually represented as the sum of the diamagnetic and the paramagnetic terms. However, the amsotropic behavIOur of the shielding tensor can be predominantly explained by the paramagnetIc term, smce the diamagnetic term is isotropic. Figure 6 shows the R N - H dependence of the calculated isotropic 15N shIelding (a,SO) of Gly NH. It IS shown that a decrease of R N - H leads to a very large mcrease of a,SO; for example, a decrease of 0.4 A m R N - H leads to a calculated shieldmg mcrease of about 30 ppm. This agrees wIth the observed results. ThIs ImplIes that the observed shielding increase is a consequence of changes m the electromc structure through a decrease of the N - H bond length. Figure 7 shows the R N - H dependence of the calculated principal values of the 15N shielding of Gly NH. It IS shown that an decrease of R N - H leads to an increase in shielding of all> a 22 and a33, respectively. The magmtudes for changes of the shieldmg occur m the order a22> a 33 > all' The dIrection of the prmcipal axes of the Gly NH 15N shielding tensor components, as determmed by the NMR study of a Boc-Gly-Gly-Gly-OBzl smgle crystal [31], are shown m FIg. 9. The all component lies approximately along the N-H bond, and the a 33 component lies approxImately along the N-C' bond. The a22 component IS aligned m the dI- 28 rectlOn perpendIcular to the peptide plane. From this, it can be Said that the a22 and a 33 components, rather than the all component, are sensitive to R N - H changes. This may be due to the fact that the a22 component lies approximately along the direction of the nitrogen lone-pair electrons and the electron density is very high in this direction. The a33 component lies approxImately along the N-C' bond into which lone-paIr electrons of the nitrogen atom transfer, and so the bond order becomes very high. Consequently, the a22 and a33 components are sensitive to changes in R N - H • Finally, we can draw the following conclusions. The observed 15N shIeldmgs of amide nitrogens increase linearly wIth a decrease of the N-H bond length associated with the hydrogen bond; this can be justified by quantum chemical calculatIOns. The 15N shieldmg is applicable as a means for obtaining direct information about the nature of the hydrogen bond in the solId state, m addibon to the 13C shielding of the carbonyl carbons m the hydrogen bonds REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 G A Webb and M WitanowskI, Proc IndIan Acad SCI (Chern ScI) ,94 (1985) 241 G Harblson,J HerzfeldandRJ Gnffm,J Am Chern Soc, 103 (1981) 4752 T A Cross, J A DlVerdl and S T, Opella, J Am Chern Soc, 104 (1982) 1759 T A Cross, M H Frey and S J Opella, J Am Chern Soc, 105 (983) 7471 H G Forster, D Muller and H R Kncheldorf,lnt J BIOI Macromol, 5 (1983) 101 T H Huang, W W Achovchm, R G Gnffm and C M Dobson, BIOchemlstry, 23 (1984) 5933 EO Stejskal, J Schaefer and R A McKay, J Magn Reson, 57 (1984) 471 C N Matthews, R Ludlcky. J Schaefer, EO Stejskal and R A McKay, Ongms LIfe, 14 (1984) 243 T A Cross and S T, Opella, J Mol BIOI, 182 (1985) 367 B S ChOl, J E Roberts, J N S Evans and M F Roberts, BIOchemIstry, 25 (1986) 557 J Schaefer, J R Garbo\\', G S Jacob, T M Forest and G E WIlson, Jr, BIOchem BIOphys Res Commun, 137 {1986) 736 P L Stewart, K G Valentme and S T Opella, J Magn Reson, 71 (987) 45 (a) A ShOJI, T OzakI, T FUJlto, K Deguchl and I Ando, Macromolecules, 20 (1987) 2441 (b) A ShOJI, T Ozaki, T FUJltO, K Deguchl, SAndo and I Ando. Macromolecules, 22 (1989) 2860 SAndo. I Ando, A ShOJI and T OzakI, J Am Chern Soc. 110 (1988) 3380 (a) I., Zervas, D Borovas and E Gazls, J Am Chern Soc, 85 (1963) 3660 (b) E Wunsch, Synthese von Peptlden, Tell 1 and 2, Georg Thieme Verlag Stuttgart, 1974 (c) N IzumIYa, T Kato, M Ohno and H Aoyagl, PeptIde SyntheSIS. Maruzen, Tokyo. 1974 lAndo, H SaIto, R Tabeta. A ShOJI and T OzakI, Macromolecules, 17 (1984) 457 R DltchfIeld. D P MIller and.J A Pople, J Chern Phys. 54 (1971) 4186 P D Elhss, G E MaCiel and J M Mciver. Jr , J Am Chern Soc, 94 (1972) 4069 I Ando and G A Webb, Theory of NMR Parameters, AcademIC Press, London, 1983 FA Momany, R F McGUIre, J F Yan and H A Scheraga,.J Phys Chern, 75 (1971) 2286 M Wnanowsk!. L Stefamak and G A Webb, Annual Reports on NMR Spectroscopy. Vol 18. AcademIC Press, London, 1986 H R Kncheldorf,Org Magn Reson, 15 (2) (1981) 162 29 23 24 25 26 27 28 29 30 31 32 A B Blswas, E W Hughes, B D Sharma and J N WIlson, Acta Crystallogr , Sect B, 24 (1968) 40 T F Koetzle, W C Hamliton and R Parthasarathy, Acta Crystallogr, Sect B.28 (1972) 2083 S N Rao and R Parthasarathy, Acta Crystallogr , Sect B, 29 (1973) 2379 V L Murah and E Subramaman.lnt J Peptlde Protem Res, 27 (1986) 472 V Lahta. E Subramaman and R Parthasarathy,lnt J Peptlde Protem Res. 27 (1986) 223 V Lahtha. E Subramaman and J Bordener, IndIan J Pure Appl Phys. 23 (1985) 506 JP Glusker,HL Carrell,HM Berman,B GallenandRM Peck,J Am Chern Soc,99(2) (1977) 595 W M Carson and M L Hackert, Acta Crystallogr , Sect B, 34 (1978) 1275 Y HlYama, C NlU, J V SIlverton. A Bavoso and D A TorchIa, J Am Chern Soc, 110 (1988) 2378 R Taylor, 0 Kennard and W VersIchel, Acta Crystallogr , Sect B.40 (1984) 280