A Dataset of Simple 1-D and 2-D NMR Spectra of Peptides, Including all Encoded Amino Acids, for Introductory Instruction in Protein Biomolecular NMR Spectroscopy

Abstract

NMR spectroscopy is the most important technique for understanding the structure of peptides and proteins in solution. Contemporary publications in the interpretation of NMR spectra of peptides and proteins generally focus on advanced techniques and complex spectra, with a lack of simple spectra and guides available for beginning students. A dataset of ¹H NMR spectra was generated from a series of simple peptides that include all canonical amino acids (X) [Ac-X(S/pS)-NH₂, Ac-X(T/pT)-NH₂, and Ac-XPPGY-NH₂; pS=phosphoserine, pT=phosphothreonine)]. The characterization of each peptide includes 1-D and TOCSY spectra, with both raw and processed data available. The spectra can be used for instructional applications, including analysis of regions of the spectra (e.g. amide, aromatic, Hα, and aliphatic); identification of spin systems and residue assignment via TOCSY spectra; analysis of conformational features including amide H_N chemical shift dispersion and changes due to hydrogen bonding or post-translational modifications; the ³JαN coupling constant that reports on the ϕ torsion angle and on order versus disorder at a residue; conformational preferences at Hα via chemical shift index analysis; understanding of diastereotopic hydrogens; dynamic processes, including hydrogen exchange; and identification of proline cis-trans isomerism. In addition, for a limited number of peptides, NOESY spectra are included to allow sequential resonance assignment and for assignment of trans versus cis proline conformations. Spectra from closely related peptides allow the analysis of the relative effects of single amino acid changes. The paper is written to be directly accessible to students in research labs or in the classroom as a tutorial guide.

Graphical Abstract

Introduction

NMR spectroscopy is the primary high-resolution technique to understand the structure of peptides and proteins in solution.¹ NMR spectroscopy provides details on structure at the single-residue level and indeed at the single-atom level, as each resonance indicates the environment at a single hydrogen. In contrast, bulk techniques such as circular dichroism spectroscopy give global details about secondary structure, while providing no information at the individual residue level.² NMR spectroscopy also provides information about the dynamics of molecules in solution that is not typically possible via techniques such as X-ray crystallography, which normally exhibits a single conformation that may or may not be the dominant conformation in solution.³ NMR spectroscopy is particularly valuable in understanding structure in intrinsically disordered regions of proteins, which are typically not amenable to X-ray crystallography.^4–14 Importantly, nearly all students in chemistry, the biological sciences, and biomedically oriented fields in engineering have a basic understanding of NMR spectroscopy due to its applications in introductory organic chemistry courses and associated laboratory courses, as well as from dedicated courses in spectroscopy.

However, despite the broad importance of NMR spectroscopy in the analysis of proteins, the technique in practice often has a significant barrier to entry. Most importantly, organic chemistry classes do not typically significantly cover NMR spectroscopy as applied to peptides and proteins. Indeed, peptide structure is typically a late or optional topic in organic chemistry classes. In addition, TOCSY spectra that are central to peptide and protein resonance assignments are not commonly taught in the introductory organic chemistry spectroscopy curriculum. As such, students are often dependent on specialized classes or research to have access to instruction in this technique, inherently and unnecessarily limiting its applications. Indeed, much instruction in these techniques is passed on to students in their individual research groups. However, because essentially all students in the biological sciences have early instruction in the basics of NMR spectroscopy, this technique should inherently be highly accessible.

Wüthrich wrote a landmark book on using NMR spectroscopy to understand structure and dynamics in proteins.¹ In addition, many subsequent textbooks, review articles, and original research articles have been written on the application of NMR spectroscopy to understanding protein structure.^{3,5,7,13,15,16} However, early articles include representations of the spectra that lack either modern graphical analysis and/or lack access to the underlying spectra. Recent articles, in contrast, focus on larger proteins, whose NMR spectra are functionally inaccessible to beginning students due to spectral complexity. While an exceptional database of NMR data is available at BioMagResBank,¹⁷ these data are better appreciated by advanced students and researchers than they are accessible to students in their introductory work in NMR spectroscopy of peptides and proteins.

In our instruction both of undergraduate students in the teaching laboratory and of starting undergraduate students and graduate students in the research laboratory, we realized that additional teaching resources with more manageable NMR spectra and instructional guides could simplify instruction and accelerate student understanding of the basic elements of peptide and protein NMR spectroscopy. We recently developed a teaching laboratory module for introductory (second-semester) organic chemistry laboratory students to learn about techniques in solid-phase peptide synthesis, NMR spectroscopy on the peptides students synthesized, and conformational analysis.¹⁸ In the context of that work, the students synthesized a series of peptides Ac-X(S/pS)-NH₂, Ac-X(T/pT)-NH₂, and Ac-XPPGY-NH₂, where X = all 20 canonical amino acids, S = serine, pS = phosphorylated serine, T = threonine, and pT = phosphorylated threonine (Figure 1). That work was focused on the research questions of understanding either (a) the effects of local amino acids on phosphorylation-induced structure in peptides or (b) the propensities of amino acids for nucleation of polyproline II helix (PPII) structure.^19–22 The discussion of those residue-specific effects on peptide structure will be reported separately.

Figure 1. Peptide series investigated and nomenclature.

(a) N-acetylated, C-terminal carboxamide peptides with serine (Ser/S), phosphoserine (pSer/pS), threonine (Thr/T), and phosphothreonine (pThr/pT). R indicates the Xaa/X amino acid side chain. (b) Backbone of the peptide Ac-Xaa-Pro-Pro-Gly-Tyr-NH₂ (Ac-XPPGY-NH₂). (c) Greek letter nomenclature for amino acid side-chain carbons (left) and hydrogens (right) with lysine (Lys/K) are shown as an example. H_N = amide hydrogen, R^N = N-terminal amino acid sequence, R^C = C-terminal amino acid sequence.

However, the simple NMR spectra from these peptides represent a potentially useful general education tool in biochemistry, biophysics, chemical biology, and molecular biology. Herein, we report a dataset of simple NMR spectra for instruction in peptide and protein NMR spectroscopy, along with general guidelines for their interpretation. These data include all canonical (encoded) amino acids, with each amino acid present in at least 5 different peptide spectral contexts. Because the peptide sequences are quite short, the NMR spectra are simpler than typical peptide and protein NMR spectra. These spectra can be used in introductory classroom and laboratory contexts, where the simple spectra can be readily interpreted by students. This manuscript is also intended to function as a basic tutorial guide for undergraduate and graduate students in research labs in this instruction, which can be complemented by traditional in-class teaching, textbooks, and laboratory instruction manuals, as well as by various web-based resources.

Experimental

Peptide synthesis.

All peptides were synthesized by standard methods in solid-phase peptide synthesis. The synthesis and characterization of the peptides Ac-APPGY-NH₂ and Ac-CPPGY-NH₂ was described previously.¹⁹ The synthesis, purification, and characterization data for all other peptides will be described elsewhere.¹⁸

1-D ¹H NMR spectroscopy.

1-D ¹H NMR spectra were collected on either a Brüker 600 MHz NMR spectrometer equipped with a 5-mm Brüker broad band observed (BBO) probe or a Brüker 600 MHz NMR spectrometer equipped with a 5-mm cryogenic QCI probe. Experiments were conducted with water suppression using an excitation sculpting pulse sequence with gradients (zgesgp) and a relaxation delay of between 1.7 and 2.0 s.²³ 32,768 or 65,536 data points in the time domain (TD) were collected. Peptides were allowed to dissolve in buffer containing 5 mM phosphate, 25 mM NaCl, 90% H₂O/10% D₂O, and 10 μM TSP. Spectra were recorded on solutions at pH 4.0 (+/− 0.1) unless otherwise specified. 1-D spectra were recorded at 298 K unless the amide hydrogens did not resolve, in which case a temperature of 278 K was used. Spectra were typically zero-filled to twice the size of the TD. Apodization was applied with combined exponential and gaussian functions, with the Hz values for each of these functions varied by spectrum, which ranged from 0–0.3 Hz for the exponential function and from 0–2.5 Hz for the gaussian function.

TOCSY NMR spectroscopy.

Most TOCSY spectra were collected using a Brüker 600 MHz spectrometer equipped with a cryogenic QCI probe. Experiments were conducted using homonuclear Hartman-Hahn transfer using the MLEV17 sequence for spinlock, using two power levels for excitation and spinlock, phase sensitive method. Water suppression was accomplished using a WATERGATE pulse sequence with gradients, with 25% nonuniform sampling, and a relaxation delay of 1.7–2.0 s.^24–26 256 data points were collected in the F1 (indirect, evolution) dimension and 2048 data points were recorded in the F2 (direct, acquisition) dimension. A smaller subset of TOCSY spectra was collected using a Brüker 600 MHz spectrometer equipped with a 5-mm Brüker SMART probe. Experiments were conducted using homonuclear Hartman-Hahn transfer using the DIPSI2 sequence for spinlock. Water suppression was achieved using excitation sculpting with gradients, with 50% nonuniform sampling, and a relaxation delay of 1.7–2.0 s.^23,27 512 data points were recorded in the F1 dimension and 2048 data points were collected in the F2 dimension. TOCSY experiments were conducted at 298 K with a spinlock mixing time of 80 ms.

NOESY NMR spectroscopy.

NOESY spectra were collected on a Brüker 600 MHz spectrometer equipped with a cryogenic QCI probe. Experiments were conducted using the homonuclear correlation via dipolar coupling, phase-sensitive method. Water suppression was achieved using a WATERGATE pulse sequence with gradients, with 50% nonuniform sampling, and a relaxation delay of 2.0 s.^25,26 400 data points were recorded in the F1 dimension and 2048 datapoints were collected in the F2 dimension. Spectra were recorded at 298 K with a NOESY mixing time of 200 ms. NOESY spectra were collected for the peptides Ac-ST-NH₂, Ac-DT-NH₂, Ac-PS-NH₂, Ac-QS-NH₂, Ac-TS-NH₂, Ac-WS-NH₂, Ac-SpT-NH₂, Ac-ApS-NH₂, Ac-WPPGY-NH₂, and Ac-NPPGY-NH₂.

Results and Discussion

1-D ¹H NMR spectra: regions of the NMR spectrum, amide H_N chemical shifts, and ³J_aN coupling constants.

One challenge of peptide and protein NMR spectroscopy is the preponderance of peaks in the NMR spectra, in contrast to NMR spectra of simple organic molecules. The most basic task in any NMR analysis is peak identification and resonance assignment. The NMR spectra of peptides and proteins have a number of key regions that can be readily identified by experienced practitioners, and that students can be taught to identify and interpret. The starting point here is to make sense of the general spectrum, in order to reduce the complexity to individual spectral regions that are manageable with guidance.

The key spectral regions are the amide H_N resonances (~7–9 ppm), aromatic H resonances (~6.5–8.5 ppm), Hα (~4–5 ppm), and resonances from aliphatic hydrogen C–H bonds (~1–4 ppm). These spectral regions are broadly annotated in Figure 2, while also emphasizing the resonance from H₂O at 4.8 ppm and the overlap that occurs between these spectral regions.

Figure 2. 1–D ¹H NMR spectra of selected peptides, with key spectral regions indicated and all identifiable resonances annotated.

(a) NMR spectrum of the dipeptide Ac-serine-phosphothreonine-NH₂ (Ac-Ser-pThr-NH₂/Ac-SpT-NH₂) at pH 6.8. The following regions are shown above the 1–D spectrum: amide hydrogens (red), aromatic hydrogens (magenta), Hα hydrogens (purple), Hβ hydrogens, and other aliphatic hydrogens (green), represented as H_N, H_aromatic, H_α, H_β, and H_aliphatic respectively. Peaks from water and from 3-(trimethylsilyl)propionic-2,2,3,3–2-d₄ acid sodium salt (TSP), which is used for referencing to 0.00 ppm, are also annotated. (b) Spectrum of Ac-alanine-phosphoserine-NH₂ (Ac-Ala-pSer-NH₂/Ac-ApS-NH₂) at pH 4.0. (c) Spectrum of Ac-phenylalanine-serine-NH₂ (Ac-Phe-Ser-NH₂/Ac-FS-NH₂) at pH 4.0. (d) Spectrum of Ac-glutamate-proline-proline-glycine-tyrosine-NH₂ (Ac-Glu-Pro-Pro-Gly-Tyr-NH₂/Ac-EPPGY-NH₂) at pH 4.0. All NMR spectra were obtained using a solution of the peptide in 90% H₂O/10% D₂O containing 5 mM sodium phosphate pH 4.0 or 6.8 and 25 mM NaCl at 298 K, with 100 μM TSP as an internal reference.

Several key NMR parameters can be obtained from simple 1–D ¹H NMR spectra. First, amide chemical shift dispersion (the overall range of the amide chemical shifts) is inherently related to structure in peptides and proteins (Figure 3). Thus, disordered proteins exhibit relatively little amide chemical shift dispersion (a small range of amide chemical shifts). In contrast, folded proteins and ordered regions of peptides exhibit much greater amide chemical shift dispersion (a wider range of amide chemical shifts). In related peptides, differences in H_N chemical shift dispersion can be readily identified. In particular, differences are observed herein in related peptides as a function of phosphorylation.

Figure 3. Amide hydrogen chemical shift dispersion.

(a) Amide regions from the ¹H NMR spectra of Ac-Arg-pSer-NH₂ (Ac-RpS-NH₂) and Ac-Asp-pSer-NH₂ (Ac-DpS-NH₂) at pH 6.8, with amide resonances identified by amino acid. These spectra demonstrate large chemical shift differences between the amide hydrogens of the phosphoserine and the N-terminal amino acid residue, which is associated with order at phosphoserine. (b) Amide regions from the ¹H NMR spectra of Ac-Trp-Thr-NH₂ (Ac-WT-NH₂) and Ac-Trp-pThr-NH₂ (Ac-WpT-NH₂) at pH 4.0 and 6.8, respectively. Amide resonances are labeled with their corresponding amino acid. These spectra demonstrate smaller chemical shift dispersion between the amide hydrogens, but also indicate how amide hydrogen chemical shifts can change due to changes in local structure (e.g. as a result of changes to the adjacent amino acid).

More generally, in most of the peptides, the distinct amide hydrogens are all well resolved. In these H_N, the coupling constant between H_N and Hα (³JαN) can be determined directly from the 1-D ¹H NMR spectrum (Figure 4). ³JαN correlates with the ϕ main chain torsion angle, via a parametrized Karplus relationship.¹⁵ In disordered peptides, it is unlikely that a single conformation in ϕ is adopted, and ³JαN broadly represents the population-weighted average of all conformations present in solution. However, ³JαN correlates overall with ϕ, with values around 7 Hz indicating random coil conformations. Smaller values of ³JαN (especially < 6 Hz) indicate more compact conformations in ϕ (e.g., more α-helical), while larger values of ³JαN (especially > 8 Hz) indicate more extended conformations in ϕ (e.g., more β-like) Most generally, deviations from random coil values of ³JαN (6–8 Hz) indicate significant ordering in peptides. In addition, changes in ³JαN as a result of variations in peptide sequence or post-translational modification more generally can be correlated with changes in biases in peptide main-chain conformation, including the propensities of amino acids for different regions of Ramachandran space. Shifts to smaller values of ³J_αN indicate adoption of more compact conformations, while shifts to larger values of ³J_αN indicate changes to more extended conformations.^15,28–32 For example, in Ac-XPPGY-NH₂ peptides, ³J_αN = 6.0 Hz for Ala, but 7.7 Hz for Val, consistent with Ala preferring more compact conformations, but Val preferring more extended conformations.^20,33 As an example of how post-translational modifications can change protein structure,^{19,21,22,34–37 3}J_αN changes from 8.3 Hz for Thr to 6.4 Hz for pThr in Ac-Y(T/pT)-NH₂ peptides, indicating here that threonine phosphorylation leads to a change from more extended conformations to a more compact overall conformational ensemble. Notably, Gly H_N resonances are distinct due to the appearance of a pseudo-triplet (doublet of doublets) due to the presence of two Hα (Figure 4ab).

Figure 4. H_N–H_α ³J_αN coupling constants.

(a) Amide-aromatic regions of the ¹H NMR spectrum of Ac-Ala-Pro-Pro-Gly-Tyr-NH₂ (Ac-APPGY-NH₂) with the chemical shifts of all amide hydrogens annotated. H_α–H_N coupling constants (³J_αN) are listed, with brackets depicting how this coupling constant was measured. Diagrams shown demonstrate the source of ³J_αN coupling and explain the multiplicities observed for each signal. ³J_αN values serve as an indication of average ϕ angle values at a particular residue, via a parametrized Karplus relationship (Figure S2).¹⁵ (b) Spectrum of Ac-Val-Pro-Pro-Gly-Tyr-NH₂ (Ac-VPPGY-NH₂). This peptide exhibits distinct signals from both the trans-proline and cis-proline amide bond conformational isomers, resulting in two peaks corresponding to each amide hydrogen found in the structure. (c) Spectrum of Ac-Tyr-Thr-NH₂ (Ac-YT-NH₂). (d) Spectrum of Ac-Tyr-pThr-NH₂ (Ac-YpT-NH₂). In conjunction with (c), this spectrum serves to highlight the effect of phosphorylation on ³J_αN values and therefore the role of phosphorylation in changing protein structure.

Identification of amino acid spin systems and resonances in NMR spectra using TOCSY spectra.

Even in simple peptide NMR spectra, the identity of the amide H_N and other chemical shifts cannot typically be definitively determined solely from the 1-D spectra. TOCSY (TOtal Correlation SpectroscopY) spectra are the predominant method to assign resonances in peptide and protein NMR spectra.^24,38 TOCSY spectra are, to a first approximation, similar to COSY (COrrelation SpectroscopY) spectra that also exhibit longer-range through-bond coupling, albeit without providing coupling constants. COSY spectra are commonly taught in organic spectroscopy classes, and indicate which resonances have short-range (2-bond or 3-bond) through-bond (scalar, spin-spin) coupling (²J or ³J). In contrast, in TOCSY spectra, longer-range correlations are observed, in addition to the correlations seen in COSY spectra. However, in TOCSY spectra, no longer-range correlations are observed that extend through amide bonds, and thus the TOCSY spectra represent correlations within single residues of a peptide or protein. Peaks in TOCSY spectra also exhibit “self-correlation” peaks, that is, crosspeaks of a resonance with itself, which are along the “diagonal” that extends from the lower left to the upper right (i.e. the ⁄ pattern that appears to divide the spectrum in half) of the TOCSY spectrum. The off-diagonal signals indicate long-range (TOCSY) coupling as well as short-range (COSY) coupling.

TOCSY spectra exhibit correlations within the entire spin systems of single amino acids. As such, in the TOCSY spectra, correlations are typically observed between H_N and all of Hα, Hβ, and (if present) even Hγ, Hδ, and Hε (Figure 5). The “fingerprint region” of TOCSY spectra (upper left quadrant) thus connects all of these hydrogens from a single residue in a single vertical slice centered on the chemical shift of H_N. Because every amino acid has distinct sets of hydrogens, with characteristic typical ranges of chemical shifts,^1,28,39–47 the amino acid identity of any given amide hydrogen resonance can often be clearly determined directly from the fingerprint region of the TOCSY spectrum (Figure 6, Table 1). This assignment then inherently includes the more difficult assignments of (often overlapping) peaks in the aliphatic region of the NMR spectra. Each crosspeak in the TOCSY spectrum indicates the correlation between any given two hydrogens in the same amino acid, and thus also provides the chemical shifts of each of those hydrogens.

Figure 5. Interpretation of total correlation spectroscopy (TOCSY) spectra.

(a) Full TOCSY spectrum of Ac-Met-Thr-NH₂ (Ac-MT-NH₂), with the 1-D ¹H NMR spectrum annotated on the x-axis. Boxes indicate regions of the spectrum further analyzed in subsequent parts of this Figure. (b) Carboxamide hydrogens in the TOCSY spectrum, showing the characteristic pattern of these hydrogens. (c) Fingerprint region of the TOCSY spectrum, showing each amide hydrogen and the crosspeaks corresponding to the side chains of each amino acid. (d) Aliphatic region of the TOCSY spectrum, with crosspeaks from each side chain indicated in separate boxes.

Figure 6. Side-chain correlations to the amide hydrogen observed in the fingerprint region of TOCSY spectra.

Vertical slices at the H_N resonance from the TOCSY spectra of all 19 amino acids bearing an amide hydrogen. These slices show the characteristic crosspeak patterns observed in the fingerprint region of TOCSY spectra for each non-proline amino acid.

Table 1.

Observed chemical shift ranges (ppm) for all amino acids in this library of peptide spectra.

residue	δ HN	δ Hα	δ Hβ	δ Hγ	δ Hδ	δ Hε	other
A, Ala	8.5–8.1	4.6–4.2	1.5–1.2	–	–	–	–
C, Cys	8.5–8.4	4.6–4.3	3.0–2.7	–	–	–	–
D, Asp	8.5–8.2	4.9–4.6	2.9–2.3	–	–	–	–
E, Glu	8.4–8.2	4.7–4.2	2.2–1.7	2.5–2.0	–	–	–
F, Phe	8.4–8.1	4.8–4.6	3.3–2.7	–	–	–	–
G, Gly	8.6–8.2	4.1–3.8	–	–	–	–	–
H, His	8.5–8.0	4.9–4.6	3.3–2.7	–	7.4–6.9	8.6–7.6	–
I, Ile	8.5–8.0	4.4–4.1	2.0–1.7	1.6–1.1	1.0–0.8	–	CH3 γ 1.0–0.8
K, Lys	8.6–8.1	4.6–4.3	1.9–1.4	1.7–1.4	1.9–1.5	3.9–3.0	NH3 ζ 7.6–7.4
L, Leu	8.4–8.2	4.4–4.3	1.7–1.4	1.7–1.4	1.0–0.8	–	–
M, Met	8.5–8.2	4.8–4.4	2.2–1.9	2.7–2.5	–	2.6–2.0	–
N, Asn	8.5–8.3	5.0–4.7	3.0–2.7	–	–	–	NH2 δ 7.6–6.8
P, Pro	–	5.0–4.3	2.5–1.8	2.2–1.5	4.0–3.5	–	–
Q, Gln	8.4–8.2	4.6–4.3	2.2–1.8	2.5–1.9	–	–	NH2 ε 7.6–6.8
R, Arg	8.4–8.1	4.6–4.2	2.0–1.6	1.8–1.5	3.3–3.0	–
S, Ser	8.8–7.9	4.8–4.2	4.4–4.0	–	–	–	–
T, Thr	8.4–7.9	4.7–4.1	4.4–4.0	1.3–1.0	–	–	–
V, Val	8.3–7.8	4.5–4.0	3.1–2.1	1.0–0.9	–	–	–
W, Trp	8.3–7.9	4.8–4.6	3.4–3.1	–	7.3–7.2	–	Haromatic 7.7–7.1 Hindole 10.3–10.0
Y, Tyr	8.3–7.9	4.7–4.4	3.2–2.6	–	7.2–7.1	6.9–6.8	–
pS, pSer	9.3–8.0	4.9–4.3	4.4–4.0	–	–	–	–
pT, pThr	9.5–8.0	4.8–4.3	4.4–4.0	1.4–1.2	–	–	–
Ac-							2.1–1.9
-NH2							7.9–6.9

Considering the peptide Ac-MT-NH₂ (Figure 5), the amide resonance at 8.41 ppm can be identified as the H_N of Met via its correlation to peaks at 4.52 ppm (Hα), 2.64 ppm (Hγ), and 2.03 ppm (terminal CH₃ Hε), which represents a characteristic pattern for Met. In contrast, the amide resonance at 8.16 ppm can be identified as Thr due to its correlation with hydrogens at 4.36 ppm (Hα), 4.28 ppm (Hβ), and 1.21 ppm (CH₃ Hγ). Notably, the side-chain hydrogens are not typically identified via their coupling constant patterns, the approach that is typical in NMR spectroscopy of organic compounds, due both to complex splitting patterns (multiple adjacent hydrogens coupled) and to spectral overlap in the aliphatic region. In addition, Hα and other peaks near the water resonance are often greatly reduced in magnitude or not observed in 1-D spectra due to water suppression techniques (i.e., the methods necessary to permit the observation of peptide peaks present at μM–mM concentrations in 55 M H₂O), but are readily identifiable in TOCSY spectra.

The simplest TOCSY correlations for residues with aliphatic side chains are those of Ala, Cys, Asp, Asn, and Ser (Figure 6, which includes an example of the correlations observed for each amino acid). In their TOCSY fingerprint regions, these residues exhibit correlations between H_N, Hα, and Hβ. In each case, the chemical shift of Hβ correlates with the electron-withdrawing nature of the group attached to Hβ (Ser Hβ ~ 4 ppm; Cys Hβ ~ 3 ppm; Asp/Asn Hβ ~ 2.5–3 ppm; Ala Hβ ~ 1.3 ppm; see also discussion of diastereotopic hydrogens, below). Gly resonances are similarly simple, with correlations of H_N to the two diastereotopic Hα. Thr, Glu, and Gln fingerprint regions are also fairly straightforward to identify, with correlations out to Hγ. Importantly, the fingerprint regions of Asp versus Asn, or of Glu versus Gln, are not readily distinguished via the TOCSY spectra, because the patterns of chemical shifts are similar, and no correlations are typically observed from H_N to the carboxamide hydrogens of Asn or Gln. However, peptides with Asn and Gln will have an additional set of carboxamide resonances compared to equivalent peptides with Asp or Glu (see below).

More complex fingerprint regions are observed for amino acids with longer side chains (Ile, Lys, Leu, Met, Arg, Val) (Figure 6, Figure 7). In particular, correlations with the hydrogens most distant from H_N, or with 3° methine (C–H) hydrogens, are sometimes weak or not observed to be above the noise in the fingerprint region. In these cases, the aliphatic region (upper right quadrant) of the TOCSY spectrum can be used to assign all resonances within the spin system of these amino acids. For example, in Ac-IpS-NH₂ (Figure 7ad), correlations from H_N are only seen to some of the aliphatic side-chain hydrogens of Ile. However, all of these hydrogens exhibit strong correlations to each other, with several also correlating to Hα, as observed within the upper right (aliphatic quadrant), confirming the resonances of the entire Ile spin system.

Figure 7. Identification and differentiation of amino acids with aliphatic side chains.

(a-c) Full TOCSY spectra with side-chain crosspeaks identified by amino acid. Spectra were recorded at pH 6.8 for (a,d) Ac-IpS-NH₂ and (b,e) Ac-VpT-NH₂, or at pH 4.0 for (c,f) Ac-LpT-NH₂. (d-f) Expansion of the fingerprint regions to highlight individual correlations. Ac-IpS-NH₂ (d) shows a peak from H_γ1 at 1.47 ppm in the aliphatic region of the TOCSY spectrum, but this hydrogen is not observed as a crosspeak with the isoleucine amide hydrogen. Leucine crosspeaks (e) show overlap of peaks from H_γ and H_β, resulting in a complex pattern of peaks at this chemical shift in the ¹H NMR spectrum.

Spin systems in aromatic amino acids.

The side chains of all of the aromatic amino acids are distinct (Figure 8 for a comparison of the aromatic regions of 1-D spectra; Figure 9 for TOCSY spectra). The usual correlation observed in the fingerprint region is between H_N, Hα, and Hβ. This pattern is similar to those of Cys, Asp, Asn, and Ser, with Hβ on aromatic amino acids typically ~ 3 ppm (similar to that of Cys). While the aromatic hydrogens do not typically exhibit a cross-correlation to the amino acid H_N in the TOCSY spectra, the aromatic hydrogens sometimes exhibit TOCSY cross-peaks to the Hβ of the same residue (Figure 9c), which can then be correlated to H_N. Independently, the aromatic side-chain spin systems of Phe, Tyr, Trp, and His are readily differentiated, and identifiable in both the 1-D spectra (Figure 8) and in the cross-correlations in the TOCSY spectra (Figure 9). Tyr exhibits a simple pair of apparent doublets, representing the Hδ (H_ortho) and Hε (H_meta) that are coupled to each other, each with a multiplicity of 2. Phe exhibits a more complicated spectrum due to Hζ (H_para) being coupled to two Hε (H_meta), observed as a triplet, and due to the Hε (H_meta) being coupled to both Hδ (H_ortho) and Hζ (H_para). Trp is distinct due to its indole H at 10.0–10.5 ppm, in addition to 5 other aromatic hydrogens with different coupling patterns. All of these aromatic resonances can be correlated to each other in the aromatic region of the TOCSY spectra.

Figure 8. Aromatic regions of the ¹H NMR spectra for all four aromatic amino acids, showing the typical chemical shifts and distinct coupling patterns observed for each.

At pH 6.8, His is present as a mixture of cationic and neutral forms, while at pH 4 His is almost exclusively in the cationic form; the differences in chemical shifts of the hydrogens are consistent with changes in populations of the His ionization states.

Figure 9. TOCSY spin systems of aromatic amino acid side chains.

(a,b) Aromatic regions from the TOCSY spectra of Ac-FpT-NH₂ and Ac-YpT-NH₂ (pH 4.0). The aromatic region of the TOCSY spectrum demonstrates the signature crosspeak patterns for these amino acids. (c) Spectra corresponding to Ac-WpT-NH₂ (pH 4.0) showing the crosspeaks between H_β and H_δ (left), allowing for definitive determination of aromatic amino acids via correlation to peaks from the TOCSY fingerprint region. (d,e) Aromatic regions from the TOCSY spectra of Ac-HpT-NH₂ at pH 6.8 and 4.0, respectively, which show the effects of changes in ionization state at His residues on the chemical shifts. Carboxamide peaks often overlap with aromatic peaks; as such, the carboxamide resonances have been indicated by gold labels in the spectra, showing one side of the box-like signals observed from these hydrogens.

Diastereotopic hydrogens.

In achiral compounds, the individual hydrogens of each methylene (–CH₂–) group are enantiotopic, and thus appear as a single hydrogen in the NMR spectrum. In contrast, in compounds with at least one stereocenter, the individual hydrogens of each methylene are diastereotopic, and thus may have distinct chemical shifts. As such, within peptides and proteins, the hydrogens on all methylene groups are diastereotopic (Figure 10). The chemical shift dispersion between any diastereotopic hydrogens can directly indicate order at those atoms, as it indicates different electronic and chemical environments of those hydrogens. These differences can occur as a result of defined conformational preferences, instead of all conformations being of equivalent energy, which would generate similar time-weighted chemical environments.

Figure 10. Diastereotopic hydrogens and diastereotopic methyl groups.

(a) H_β of selected amino acids, demonstrating (top) two distinct signals at Ser and Asp, with the diastereotopic H_β readily distinguishable for each. However, the H_β of tryptophan (Trp) (bottom) are indistinguishable from one another. These spectra highlight the two limiting cases observed for any set of diastereotopic hydrogens. (b) The ¹H NMR spectrum of (top) Ac-Val-Thr-NH₂ shows the methyl groups of valine as one doublet. However, in the spectrum of (bottom) Ac-Val-pSer-NH₂, these diastereotopic methyl groups show distinct signals as two separate doublets. (c) The H_α of glycine (Gly) are also diastereotopic and can present as a single doublet (in Ac-Gly-Ser-NH₂, top), or as a pair of doublets overlapping (in Ac-Gly-pSer-NH₂, bottom); in both of these spectra, second-order effects are also observed. The H_β of Ser are distinct in both glycine-containing spectra (c). The NMR spectra of Ac-Asp-Ser-NH₂, Ac-Trp-Ser-NH₂, Ac-Val-Thr-NH₂, Ac-Gly-Ser-NH₂, and Ac-Gly-pSer-NH₂ were recorded at pH 4.0, while the spectrum of Ac-Val-pSer-NH₂ was recorded at pH 6.8. Assignments of H₁ and H₂ correspond to the signal with the most downfield and most upfield chemical shift, respectively, and are not an indication of stereochemistry.

Diastereotopic hydrogens are an inherently difficult concept for students to understand. This challenge is due in part to a limited number of examples in their coursework. Independent of this difficulty, the diastereotopic hydrogens in peptides and proteins often have distinct chemical shifts (but also frequently do not). Examples of diastereotopic hydrogens that potentially could exhibit distinct chemical shifts include Hα of Gly; Hβ of all amino acids except Ala, Ile, Thr, and Val; Hγ of Glu, Lys, Met, Pro, Gln, and Arg; Hδ of Lys, Pro, and Arg; and Hε of Lys. In general, the closer a pair of diastereotopic hydrogens is to the peptide backbone, the more likely that different chemical shifts will be observed, although any ordering in the predominant 3-D structure present in solution can lead to greater differences in the chemical shifts of diastereotopic hydrogens. In addition, the methyl (–CH₃) groups of Val and Leu are diastereotopic, and can exhibit distinct chemical shifts.

In the fingerprint region spectral slices of Figure 6, differences in chemical shifts of the diastereotopic Hβ can readily be seen for Asp, Glu, Phe, Ser, and Gln. These can also be observed in well-resolved 1–D spectra (Figure 10). For example, in Ac-DS-NH₂ (Figure 10a, top), the diastereotopic Hβ both of Ser and of Asp exhibit distinct chemical shifts. In Ac-WS-NH₂ (Figure 10a, bottom), the diastereotopic Hβ of Ser are distinct, while the diastereotopic Hβ of Trp have the same chemical shift. Similarly, the diastereotopic γ-CH₃ groups of Val are distinct in Ac-VpS-NH₂, but have the same chemical shift in Ac-VT-NH₂ (Figure 10b).

Carboxamide hydrogens.

The hydrogens of carboxamide groups (–C(O)–NH₂), present in Asn, Gln, and at the C-terminal carboxamide in peptides, exhibit distinctive patterns in the NMR spectrum (Figure 11). Because amide bond rotation is slow on the NMR timescale, these hydrogens have distinct resonances. In addition, due to the carbonyl, these resonances do not exhibit coupling that is observable within the fingerprint region with other resonances in the amino acid. Thus, carboxamide resonances appear as a pair of broad singlets at ~ 7 ppm in the 1–D NMR spectrum, with correlation between these resonances observed in the TOCSY spectrum. Correlations are sometimes observed between one of the carboxamide hydrogens and the nearest -CH₂- hydrogens (Figure 11b).

Figure 11. Carboxamide hydrogens in the TOCSY spectrum.

(a) Carboxamide region of the TOCSY spectrum of Ac-Asn-pThr-NH₂, where correlations between both pairs of carboxamide hydrogens can be seen clearly. C-terminal carboxamide peaks are shown in purple. (b) Fingerprint region of the TOCSY spectrum, with crosspeaks between H_α of Asn and its carboxamide hydrogens (green).

Proline resonances.

Unlike other canonical amino acids, proline lacks an amide hydrogen. As such, there are no crosspeaks associated with proline in the fingerprint region of the TOCSY spectrum. Proline resonances are identified in the aliphatic region of the TOCSY spectrum, via a pattern of correlations between Hα (4.0–4.5 ppm), Hβ (2.0–2.5 ppm), Hγ (2.0–2.5 ppm), and Hδ (3.0–3.5 ppm) (Figure 12). This set of cross-correlations is typically distinct from those of any other amino acid. Glu, Gln, and Met can exhibit similar patterns of correlations between Hα and more upfield aliphatic resonances, but each of those will also exhibit a crosspeak with an amide H_N, and thus will be distinct from Pro. Notably, Pro Hβ, Hγ, and Hδ are all diastereotopic, and thus each of these positions can exhibit distinct chemical shifts for the diastereotopic hydrogens.

Figure 12. NMR spectra of proline.

(a) ¹H NMR spectrum of Ac-Pro-Ser-NH₂, showing signals from conformations with both trans-proline and cis-proline. The equilibrium constant K_trans/cis was calculated from the ratio of integration of resolved peaks from trans-proline to peaks from cis-proline. (b) Aliphatic region of the ¹H NMR spectrum of Ac-Pro-Ser-NH₂ with cis (gold) and trans (purple) H_α resonances identified. Crosspeaks allow for the determination of chemical shifts of peaks obscured in the 1-D spectra, as is seen with cis H_α, cis H_γ’ and trans H_γ’.

Proline cis-trans isomerism.

Amide bonds between amino acids are in slow exchange between the trans and cis amide conformations, due to the partial double-bond character of amides (barrier to bond rotation ~ 20 kcal mol^–1, corresponding to half-lives of seconds) (Figure 12a).^48–51 For all amino acids other than proline, only the trans amide bond is typically observed, due to the large steric clash present between the Cα of adjacent residues when a cis amide bond is present. In contrast, the amide bonds at proline can be present as either a trans or a cis amide bond due to similar potential steric clashes of Cα of the pre-proline residue with either proline Cα (Cα•••Cα clash, cis-proline) or proline Cδ (Cα•••Cδ clash, trans-proline, a steric clash that is not present in non-proline amino acids, where H_N is present instead of Cδ). In folded proteins, only a single amide conformation (trans OR cis) is typically observed at each amide bond, with cis-proline observed at approximately 5% of all proline residues, and the tertiary structure stabilizing the observed proline amide conformation. Notably, identifying whether a proline amide bond is trans or cis is not trivial by ¹H NMR spectroscopy (see below), particularly in complex proteins.^13,52–55

In contrast, within peptides and in intrinsically disordered proteins, both trans-proline and cis-proline are commonly observed, leading to more complicated NMR spectra with separate sets of resonances for each conformation, due to the interconversion being in slow exchange (t_1/2 ~ seconds to minutes) on the NMR timescale. Typically, the major species has trans-proline, while the minor species has cis-proline (Figure 13). The identity of local residues or local structure can dramatically impact the amount of cis-proline observed.^48,56–59 Notably, if multiple proline residues are present within a peptide or protein, the total number of possible species observable is 2ⁿ, where n = the number of proline residues.^20,57

Figure 13. Proline cis-trans isomerism.

Amide regions of the ¹H NMR spectra of the peptides Ac-Pro-Thr-NH₂, Ac-Tyr-Pro-Pro-Gly-Tyr-NH₂, Ac-Val-Pro-Pro-Gly-Tyr-NH₂ and Ac-Trp-Pro-Pro-Gly-Tyr-NH₂, showing peaks from both conformational isomers of each proline amide bond (two possible in (a), four possible in (b–d)). These spectra show ranges of values of K_trans/cis that might be observed in NMR spectra of peptides containing proline.

In peptides in this dataset with a proline residue, proline cis-trans isomerism is observable in both the 1-D and TOCSY NMR spectra. Peptides with aromatic-proline, proline-proline, or proline-aromatic sequences are most likely to exhibit increased populations of cis-proline.^{29,48,49,56,58,60–63} The equilibrium constant (K_trans/cis) between trans-proline and cis-proline can be directly determined via integration of relevant peaks. Typically, due to the large differences in local structure between trans-proline and cis-proline, all nearby hydrogens exhibit separate resonances for the species with trans-proline versus the species with cis-proline. Notably, the chemical shifts and ³JαN of all resonances can be quite distinct in the trans-proline versus cis-proline conformations, consistent with the dramatically different nature of these conformations and the ability of proline cis-trans isomerism to act as a functional switch in proteins.^50,51,64

The identities of cis-proline versus trans-proline can be definitively determined via NOEs, through-space correlation of nuclei which occurs via magnetization transfer. In practice, this determination of NOEs is accomplished through 2-D NOESY or ROESY spectra. For trans-proline, Hα of the residue prior to proline is close to the Pro Hδ, and thus a strong NOE is observed between these resonances. In contrast, in cis-proline, Hα of the residue prior to proline is close to the Pro Hα, and thus a strong NOE is instead observed between these resonances. These NOEs are observable in the included NOESY spectra of Ac-PS-NH₂, Ac-WPPGY-NH₂, and Ac-NPPGY-NH₂.

Identification of dynamics: hydrogen exchange.

In addition, the effects of hydrogen exchange can be identified in some of the pH-dependent NMR spectra (Figure 14). The rate of hydrogen exchange on amide hydrogens is greater with increasing pH, due to rate = k_ex[amide][HO^–], with solvent-exposed amides exhibiting broadening and exchange at pH ≥ 7.⁶⁵ Amide exchange is an important technique for understanding structure and solvent accessibility of amides.^66–69 Protected amide hydrogens (e.g. those within regular secondary structures or in the buried core of proteins) are typically observed at pH ≥ 7, but those hydrogens that are solvent-exposed and not in hydrogen-bonded structures exchange at or faster than the NMR timescale, resulting in broadening or disappearance of these amide resonances. In the peptides examined herein, only minimal structure is expected, and thus the NMR spectra at lower pH exhibit all amide hydrogens, while at higher pH the amide hydrogens are partially or fully exchanged, and in the latter case they are thus not observed.

Figure 14. Changes in ionization state and hydrogen exchange dynamics.

(a) Amide-aromatic region of the ¹H NMR spectrum of Ac-His-Pro-Pro-Gly-Tyr-NH₂ at pH 4.0, 6.5, and 8.0. Amide hydrogen exchange is faster with increasing [HO^–] (increasing pH), and results in broadening or disappearance of amide peaks due to intermediate and/or fast exchange of amide hydrogens on the NMR timescale. (b,c) Ionizable side chains, such as carboxylic acids, also exhibit chemical shift changes in all resonances due to changes in ionization state of the side chain. In 100% D₂O, no amide hydrogen resonances are observed, due to hydrogen-deuterium exchange (conversion of the amide N–H to N–D).

In addition, in peptides with Ser, Thr, or Tyr, the side chain hydroxyl H is not observed at any pH, as a result of rapid hydrogen exchange due to hydrogen bonding with water. Similarly, in peptides with the basic residues Lys or Arg, the hydrogens on side-chain nitrogens (ammonium or guanidinium) are typically either absent or broadened in the 1–D NMR spectra due to rapid exchange with water, though crosspeaks to them are sometimes observed in TOCSY spectra. In contrast, aliphatic and aromatic hydrogens are observed under all spectral conditions, due to the low acidity of the C–H bonds. These spectra emphasize the inherently dynamic nature of functional groups in water, with rapid, intermediate, slow, or no hydrogen exchange at a given functional group, depending on its solvent exposure, hydrogen bonding, and acidity.

In order to observe TOCSY correlations with amide hydrogens, NMR spectra are recorded in solutions with 90% H₂O and 10% D₂O, with water-suppression techniques used in order to minimize the signal associated with the water resonance at 4.8 ppm. In 99+% D₂O, all solvent-exposed amide hydrogens exchange with deuterium, resulting in the disappearance of peaks associated with solvent-exposed amide hydrogens (Figure 14c, bottom). Amide H–D exchange is also an important technique for understanding structure in folded proteins, in which amide hydrogens (as well as other exchangeable hydrogens including on alcohols, ammoniums, or guanidiniums) that are solvent-exposed exhibit rapid H–D exchange, while those that are protected from solvent (e.g. those that are part of stably folded secondary structures or otherwise buried) exhibit slower exchange. Amide H–D exchange can be observed quantitatively as a function of time, both by NMR spectroscopy (on a residue-specific basis) and by mass spectrometry (based on bulk changes in numbers of ¹H-hydrogens replaced with ²H-deuterium).^65–69

Sequential resonance assignment using NOESY/ROESY spectra.

In simple peptides, where there is at most one copy of any amino acid type (resonances with similar spin systems and chemical shifts) in the sequence, definitive resonance assignments can typically be made directly from the TOCSY spectra. In contrast, in complex peptides and proteins, the presence of more than one copy of a given amino acid type in the sequence generates ambiguity about, for example, which alanine resonances are associated with any one specific alanine residue in a peptide or protein with multiple alanines. This assignment problem is accomplished using NOESY spectra^70,71 via sequential resonance assignments (a resonance walk), which correlate the resonances of one amino acid with those of the adjacent residues in the sequence. ROESY spectra are used alternatively to NOESY spectra to obtain equivalent through-space correlations of nuclei.^72–74

In order to provide a context for sequential resonance assignments, NOESY spectra were recorded for eight dipeptides and two pentapeptides with clear resolution of all relevant resonances (Figure 15). In sequential assignment, Hα of the first residue is correlated with H_N of the subsequent residue. That H_N then correlates with Hα of the same residue (assignments that are confirmable via the TOCSY spectra), which then correlates with H_N of the subsequent residue. Sequential resonance assignments are broken at proline residues due to the absence of an amide hydrogen H_N. However, the proline Hα can be correlated to H_N of the subsequent residue. Sequential resonance assignments in these peptides can also assign the resonances of residues adjacent to the N-terminal (acetyl, equivalent to an Hα position on the residue prior to the first residue) and C-terminal (carboxamide, equivalent to H_N on the residue after the final residue) capping groups.

Figure 15. Sequential resonance assignment (resonance walks) using Nuclear Overhauser Effect Spectroscopy (NOESY) spectra.

(a) NOESY spectrum of Ac-Asp-Thr-NH₂, where the crosspeaks correspond to through-space couplings that allow for correlations to be made between consecutive residues, which are close in space due to the covalent bonds between residues. These crosspeaks show correlations between the acetyl methyl group and the Asp H_N (green), between the Asp H_N and the Asp H_α (red, also seen in TOCSY spectra), between the Asp H_α and the Thr H_N (purple), between the Thr H_N and the Thr H_α (blue), and between the Thr H_α and the C-terminal carboxamide hydrogens (gold). The intensities of these crosspeaks are also indicative of the proximity of one hydrogen to another and allow for definitive assignments of each carboxamide signal (H_N1 and H_N2, gold). (b) NOESY spectrum of Ac-Ala-pSer-NH_2, showing crosspeaks from the Ac- group to Ala H_N (green), Ala H_N to Ala H_α (red), Ala H_α to pSer H_N (purple), Ser H_N to pSer H_α (blue), and from pSer H_α to the C-terminal carboxamide hydrogens H_N1 and H_N2 (gold), respectively.

Thus, for example, in the data herein, the sequential resonance assignments allow differentiation between isomeric Ac-ST-NH₂ and Ac-TS-NH₂ peptides, confirming or determining peptide sequence (Figure 16). The key correlations for sequential assignment are Hα of one residue (or the acetyl group) to H_N of the next residue. Thus, in Ac-ST-NH₂ the acetyl resonance exhibits an NOE with Ser H_N, while in Ac-TS-NH₂ the acetyl resonance has an NOE with Thr H_N. The NOE between Ser Hα and Thr H_N (Figure 16a) confirms the Ac-ST-NH₂ sequence, while the NOE between Thr Hα and Ser H_N (Figure 16b) confirms the Ac-TS-NH₂ sequence. In both cases, the C-terminal residue shows a strong NOE to one carboxamide H_N1 and a weaker NOE to the other carboxamide H_N2, confirming both the identity of the C-terminal residue and which carboxamide hydrogen is cisoid (stronger NOE, equivalent to the position of an amide hydrogen in a subsequent amino acid) and which is transoid (weaker NOE, equivalent to the position of Cα in a subsequent amino acid).

Figure 16. Differentiation of isomeric peptide sequences via NOESY spectra and sequential resonance assignments.

Through the systematic application of NOESY spectra, the order of amino acids in each peptide is identified, allowing for the differentiation of (a) Ac-Ser-Thr-NH₂ from (b) Ac-Thr-Ser-NH₂ via the crosspeaks observed in each NOESY spectrum.

Inclusion of raw and processed data.

For all spectra within this manuscript, we provide the data in multiple formats. First, and most simply, we provide the processed spectra in a typical manner (available as images), so that students can analyze all of the data described above, for any given peptide, on their own. These spectra could be employed as problem set or exam questions, in which students are asked to make spectral assignments of key resonances, or are asked to identify the amino acids present within a given spectrum. Similarly, if students are given the values of “random coil” chemical shifts, for example as in the widely used tables from Wüthrich,¹ they can identify the presence and nature of any ordering present in the peptides via chemical shifts that deviate substantially from random coil values. ³JαN values can also be employed similarly in instruction to identify order at individual amino acids.^5,15,75

In our experience, one source of confusion for students is that observed chemical shifts normally do not exactly match the very specific random-coil chemical shifts in the Wüthrich table or other data on chemical shifts in random coil or disordered regions.^1,28,39–47 In Table 1, we provide the chemical shift ranges observed for the peptides herein, emphasizing to students that typical chemical shifts in peptides and proteins have a much wider range than indicated single-value random-coil chemical shifts. Figure 6 also serves as a visual complement to Table 1, showing the typical patterns of peaks for each amino acid in the fingerprint region of TOCSY spectra. We emphasize that in more complex peptides and proteins, a substantially wider range of chemical shifts will be observed than are indicated here, due to the impacts of structure and interactions with those functional groups in peptides and proteins.^{8,9,12,76–85}

Alternatively, all spectra are also provided as raw FID (free induction decay) data. These FID data could also be used as data sets to give students practice in topics such as baseline correction, apodization functions, zero filling, referencing, and other aspects of processing of 1-D or 2-D NMR data, followed by analysis of the processed data.

Additional examples of potential problems to be solved and tasks to be conducted by students.

This manuscript was written to be useful as a general tutorial guide for learning the basics of peptide and protein NMR spectroscopy, whether for students in a research lab working independently or as part of a structured classroom setting on NMR spectroscopy. In the manuscript, worked examples are included for major topics in 1–D and 2–D homonuclear ¹H NMR spectroscopy of peptides, while also including numerous additional spectra in the Supporting Information that students can analogously examine independently to achieve understanding. Also included is discussion of other topics relevant to analysis of peptides and proteins by NMR spectroscopy. The manuscript was written with the intent to be modular, with concise introductory discussion of topics that could be read independently, along with key references for more detailed discussion. It also includes discussion of topics that, in our experience, we have found to be confusing to students in both research and classroom settings, including diastereotopic hydrogens and methyl groups, dynamics of hydrogen exchange, and proline cis-trans isomerization. The work is intended to serve as a complement to more advanced textbooks and manuscripts where these topics are more fully and properly addressed.

The peptides herein were synthesized and analyzed by students in a two-week second-semester organic chemistry lab course. The second week of the lab was dedicated to analysis. Each student received the raw 1-D and TOCSY data for their peptide (i.e., each student worked with a different data set), which they then processed. The peptide sequence thus functions as a known, and the students analyzed the spectra to assign all of the resonances (by residue, and by hydrogens within that residue) and to quantify the ³JαN for each non-proline residue. Students also needed to identify any diastereotopic hydrogens and whether they were distinguishable in the NMR spectra. The ³JαN data were used individually and collectively to assign relative order and relative compact versus extended conformation at each residue. Any combination of these tasks could be applied in a formal environment. The most straightforward form, applicable to problem sets or exams, involves providing a student with the processed spectra and the peptide sequence, then asking students to assign the peaks associated with each residue. This approach includes the ability to assign separate peptides to different students, such that each student needs to learn the material independently. This approach can also be broadened by having students compare data between different peptides. Alternatively, as a simpler question, students could be given two TOCSY spectra in which one was misassigned, and be asked to identify the errors. Questions related to all of the above have been used as part of the final exam of the lab course.

A more advanced task could include providing the students with 1–D and TOCSY spectra for a peptide and having the students determine the sequence from a series of possible sequences that would be able to be distinguished from the data. The difficulty of the problem could be modulated by the number of potential choices that students are given for the sequence. For example, identify the dipeptide Ac-XZ-NH₂, where X = one of Ala, Asp, Glu, Phe, Gly, Lys, Leu, Asn, Pro, Gln, Val, Trp, or Tyr, and where Z = one of Ser or Thr. Basic differentiation can be accomplished from the fingerprint region for many of these residues, while additional analysis of the spectrum would be required to differentiate (for example) Asp versus Asn versus aromatic residues, Glu versus Gln, or Phe versus Tyr versus Trp. An exercise of this sort would be analogous to experiments in the determination of the identity of an unknown compound that are typical in second semester organic chemistry labs.

Sequential assignment can be examined via any of the eight dipeptides for which NOESY spectra are included. For each, the students can be provided with the residues present and tasked to identify which isomeric sequence is present (e.g., is this peptide sequence Ac-XZ-NH₂, or Ac-ZX-NH₂?).

Dynamics can be addressed via analysis of the spectra of peptides with exchangeable hydrogens. After resonance assignments, the students are asked why certain hydrogens (e.g., hydroxyls, ammoniums) were not observed in the NMR spectra, while other hydrogens (e.g., amides, C–H bonds) were observed. Dynamics can also be addressed via quantification of the dynamics of proline cis-trans isomerization. Dynamics as a function of enzyme activity (i.e. that would be present in a cellular context) can be addressed via comparison of the NMR spectra of non-phosphorylated and phosphorylated versions of peptides. Collectively, the data herein provide a broad opportunity for instructors to provide students with data sets to analyze in order to address diverse questions in NMR spectroscopy, peptide structure, and dynamics.

Conclusion

Herein, we report sets of 1-D and TOCSY ¹H NMR spectra of simple peptides (2–5 amino acids). NMR data are included for every amino acid within at least 5 different peptide contexts. The simple NMR spectra are intended to allow students to obtain an introductory understanding and familiarity with using NMR spectroscopy for peptides and proteins, using simple spectra that can readily be fully assigned and that lack the complications in NMR spectra of peptides and proteins. This manuscript was written to allow its application both in teaching NMR spectroscopy in the classroom and for students in the research lab, at both the undergraduate and graduate levels. These data were originally collected from NMR spectra of peptides synthesized by students at the University of Delaware in the CHEM334 laboratory class (“Organic Chemistry Laboratory 2”), with NMR data interpretation conducted by students on peptides that they had synthesized. Additional discussion and the interpretation of the data across each peptide series (e.g., what trends were observed and what do those trends mean for protein structure) will be reported elsewhere.

Supplementary Material

Additional representative NMR spectral comparisons; and 1-D ¹H and TOCSY NMR spectra for all peptides. This material is available free of charge via the journal web site. The unprocessed raw NMR data (FIDs) are available from FigShare at http://dx.doi.org/10.6084/m9.figshare.29310029. The full processed NMR spectra are available from FigShare at http://dx.doi.org/10.6084/m9.figshare.29309693.