Labeling strategies for 13C-detected aligned-sample solid-state NMR of proteins

Abstract

(13)C-detected solid-state NMR experiments have substantially higher sensitivity than the corresponding (15)N-detected experiments on stationary, aligned samples of isotopically labeled proteins. Several methods for tailoring the isotopic labeling are described that result in spatially isolated (13)C sites so that dipole-dipole couplings among the (13)C are minimized, thus eliminating the need for homonuclear (13)C-(13)C decoupling in either indirect or direct dimensions of one- or multi-dimensional NMR experiments that employ (13)C detection. The optimal percentage for random fractional (13)C labeling is between 25% and 35%. Specifically labeled glycerol and glucose can be used at the carbon sources to tailor the isotopic labeling, and the choice depends on the resonances of interest for a particular study. For investigations of the protein backbone, growth of the bacteria on [2-(13)C]-glucose-containing media was found to be most effective.

Figure 1

Analysis of fractional uniform ¹³C labeling at the C_α sites in the polypeptide backbone. A. Schematic chemical structure of an isolated ¹³C_α site bonded only to ¹²C that would contribute a high resolution NMR signal. B. Schematic chemical structure of a non-isolated ¹³C_α site bonded to ¹³C_β and/or ¹³C_O sites that would contribute a broadened or undetectable NMR signal. C. Schematic chemical structure of an unlabeled ¹²C_α site that would not contribute a signal. D. The probability (p) of occurrence of an isolated ¹³ sitesC_αthat would contribute high resolution NMR signals is predicted by the function f(p) = p(1-p)². The arrows denote the experimentally tested labeling ratios of 15%, 25%, 35%, and 45%.

Figure 2

Analysis of 13C labeling patterns of proteins obtained from E. coli by solution NMR of fd coat protein in micelles. A. and B. are one-dimensional direct-detected 13C NMR spectra obtained by direct excitation. A. The protein was obtained from bacteria grown on 25% uniformly 13C labeled media. B. The protein was obtained from bacteria grown on 2–13C glucose-containing media. C. Overlay of two-dimensional ¹³C_O-edited ¹H/¹⁵N correlation spectra of the proteins whose spectra are shown in panels A. and B. Grey indicates 25% uniformly labeled with ¹³C and black indicates labeled by growth on [2-¹³C]-glucose containing media.

Figure 3

Analysis of ¹³C labeling using ¹³C-edited two-dimensional ¹H/¹⁵N correlation spectra. Left column: Topology of ¹³C labeling (black) from a sample obtained from bacteria grown on [2-¹³C] containing media. A. Isoleucine 22 linked to glycine 23. B. Lysine 40 linked to leucine 41. C. Leucine 41 linked to phenylalanine 42. Right column: Experimental two-dimensional ¹H/¹⁵N projections of three-dimensional HNCO, HNCA, HNCOCA protein backbone spectra. Spectra from the sample obtained by growth on [2-¹³C] glucose containing media (thick lines) is compared with a [25 % uniformly 13C labelled sample (thin dashed lines).

Figure 4

Experimentally observed ¹³C labeling of fd coat protein obtained from E. coli grown on [2-¹³C]-glucose-containing media as measured from intensities in ¹H-¹³C projections of HNCO and HNCA spectra. A. Ca sites. B. C_O sites. The dashed lines mark 18% labeling, which is the level used to designate those sites that are likely to provide observable signals in solid-state NMR experiments.

Figure 5

Comparison of NMR spectra of 100% uniformly ¹⁵N labelled and fractional uniformly ¹³C labelled samples of Pf1 coat protein obtained from P. aeruginosa. Column A: ¹⁵N-edited ¹H solution NMR spectra of the protein in micelles. Column B: Direct-detected solution ¹³C NMR spectra obtained by direct-excitation of the protein in micelles. Column C: ¹³C-detected cross-polarization solid-state NMR spectra obtained with both {¹H} and {¹⁵N} decoupling during acquisition. D. 100% uniformly 13C labelled. E. 45% uniformly 13C labelled. F. 25% uniformly 13C labelled. G. From growth on 2–13C glycerol-containing media. H. From growth on 1, 3 13C glycerol-containing media. I. From growth on 2–13C-glucose containing media

Figure 6

Experimentally observed ¹³C labeling at the Pf1 coat protein obtained from P. aeruginosa grown on three different types of ¹³C-containing media. The intensity ratios are measured from ¹H,¹³C-projections of HNCO and HNCA spectra compared to those from a 100% uniformly ¹³C sample. A. C_α sites of a 25 % uniformly ³C labeled protein sample. The dashed line marks 25% labeling. B. C_α sites of a protein sample obtained from growth on [2-¹³C]-glucose containing media. C. C_o sites of a protein sample obtained from growth on [2-¹³C]-glucose containing media. In B. and C. the dashed lines marek 18% labeling, as in Figure 4.

Figure 7

Amino acids with isolated C_α sites in the polypeptide backbone with > 18% labeling. The amino acids present in Pf1 coat protein are marked by asterisks; for those amino acids that do not occur in the protein, the labeling anticipated from the Entner-Doudoroff pathway is included for completeness. The presence or absence of significant ¹³C labeling in the C_α, C_β, and C_O sites of the specified amino acids is designated in the horizontal rows. If the labeling at a C_α site is > 18% and the labeling at the bonded C_β and C_O sites is < 18%, then the C_α site is marked as + isolated. A few marginal cases are marked +/−. A. Protein obtained from E. coli grown on media containing [2-¹³C]-glucose. B. Protein obtained from P. aeruginosa grown on media containing [2-¹³C]-glucose. C. Protein obtained from P.aeruginosa grown on media containing [2-¹³C]-glycerol.

Figure 8

Comparison of one-dimensional solid-state ¹⁵N NMR NMR spectra of 100% uniformly ¹⁵N labelled and fractional uniformly ¹³C labelled samples of Pf1 coat protein obtained from P. aeruginosa. A., C., E., and G. were obtained with both ¹H and ¹³C decoupling. B., D., F., and H. were obtained with only ¹H decoupling. A. and B. 45% uniformly 13C labelled. C. and D. 25% uniformly ¹³C labelled. E. and F. Protein obtained from bacteria grown on 1,3 -¹³C glycerol-containing media. G. and H. Protein obtained from bacteria grown on I, 2-¹³C glucose- containing media.

Figure 9

Comparison of two-dimensional ¹H-¹³C PISEMA spectra of aligned Pf1 bacteriophage samples. A. 100% uniformly ¹³C labelled. B. - E. Fractional uniformly ¹³C labelled at the indicated percentages. F. From bacteria grown on [2-¹³C] glycerol-containing media. G. From bacteria grown on [1,3-¹³C] glycerol-containing media. H. From bacteria grown on [2-¹³C] glucose-containing media.

Figure 10

Comparison of single-to-noise ratios and resonance linewidths in the ¹H-¹³C PISEMA spectra of aligned Pf1 bacteriophage. The error bars represent the estimated uncertainty in the measurements made on the experimental spectra in Figure 9. A. Signal-to-noise ratios of the one-dimensional solid-state NMR spectra. B. Full width at half height of the ¹³C chemical shift dimension. C. Full width at half height of ¹H-¹³C dipolar coupling dimension. Left panels: The percentages listed on the bottom correspond to the extent of fractional uniform ¹³C labeling. Right panels: ¹³C labeling from bacteria grown in media containing D. [2-¹³C] glycerol. E. [1,3-¹³C] glycerol. F. [2-¹³C] glucose.