Proton-detected polarization optimized experiments (POE) using ultrafast magic angle spinning solid-state NMR: Multi-acquisition of membrane protein spectra

Abstract

Proton-detected solid-state NMR (ssNMR) spectroscopy has dramatically improved the sensitivity and resolution of fast magic angle spinning (MAS) methods. While relatively straightforward for fibers and crystalline samples, the routine application of these techniques to membrane protein samples is still challenging. This is due to the low sensitivity of these samples, which require high lipid:protein ratios to maintain the structural and functional integrity of membrane proteins. We previously introduced a family of novel polarization optimized experiments (POE) that enable to make the best of nuclear polarization and obtain multiple-acquisitions from a single pulse sequence and one receiver. Here, we present the ¹H-detected versions of POE using ultrafast MAS ssNMR. Specifically, we implemented proton detection into our three main POE strategies, H-DUMAS, H-MEIOSIS, and H-MAeSTOSO, achieving the acquisition of up to ten different experiments using a single pulse sequence. We tested these experiments on a model compound N-Acetyl-Val-Leu dipeptide and applied to a six transmembrane acetate transporter, SatP, reconstituted in lipid membranes. These new methods will speed up the spectroscopy of challenging biomacromolecules such as membrane proteins.

Keywords: H-DUMAS; H-MAeSTOSO; H-MEIOSIS; Membrane proteins; Multi-acquisition; Polarization optimized experiments (POE); SIM-CP; Solid-state NMR; Ultra-fast magic angle spinning.

Figure 1:

(A) CP and SIM-CP pulse sequences used for quantifying relative signal intensities. For SIM-CP, ¹³C and ¹⁵N spectra were acquired in separate experiments with a single receiver. (B) and (C) Comparison of CP and SIM-CP on SatP membrane protein obtained from integrated intensities of ¹³C and ¹⁵N spectra acquired at 65 kHz and 12 kHz MAS rates, respectively.

Figure 2:

(A) Pulse sequence for two-dimensional H-DUMAS (proton detected DUMAS) experiment for simultaneous acquisition of ¹³C- and ¹⁵N-edited experiments, (H)CH and (H)NH. (B) 1D (H)CH and (H)NH spectra of SatP membrane protein obtained from first increment (t1=0) of H-DUMAS (shown in blue and red), and the corresponding spectra (shown in black) obtained from single acquisition methods using 65 kHz MAS rate. For comparison the 1D spectra shown in blue and red are slightly shifted to the right. (C) Two-dimensional (H)CH and (H)NH spectra of SatP protein acquired using H-DUMAS pulse sequence with n=2 loops for ¹⁵N t1 evolution.

Figure 3:

MEIOSIS experiment at fast MAS rates. (A) Pulse sequence for 1D MEIOSIS that records four polarization pathways (CC_res, NCA_trans, NN_res, and CAN_trans) resulting from the first SPECIFIC-CP period τ. (B) Normalized integrated intensities of all four pathways measured for SatP protein by varying the τ mixing times. The polarization in each pathway is normalized with respect to initial value at τ=0. (C) 1D spectra of CC_res, NCA_trans, NN_res, and CAN_trans (color coded accordingly) polarization pathways obtained at τ= 0 and 6.5 ms. (D) 1D spectra at τ=6.5 ms with corresponding intensities normalized with respect to highest intensity spectrum of CC polarization pathway.

Figure 4:

(A) Two-dimensional H-MEIOSIS (proton detected MEIOSIS) pulse sequence for simultaneous acquisition of two pairs of ¹³C- and ¹⁵N-edited experiments, (H)CH, (H)N(CA)H, (H)NH and (H)CA(N)H. (B) 2D H-MEIOSIS spectra of SatP protein acquired at 65 kHz MAS rate using two loops (n=2) for ¹⁵N t1 evolution. A pair of 2D spectra in the 1^st and 2^nd acquisitions are obtained by adding and subtracting the two data sets recorded with ϕ=x and −x.

Figure 5:

(A) Pulse sequence for evaluating the residual ¹³C and ¹⁵N polarization that remains on ¹³C and ¹⁵N at the end of back CP periods (T^C_CP and T^N_CP). (B) Membrane protein SatP spectra acquired in 2^nd and 4^th acquisition periods by varying T^C_WALTZ and T^N_WALTZ decoupling periods. (C) Transferred and residual polarization of ¹³C–¹H and ¹⁵N–¹H back CP periods measured by varying T^C_CP and T^N_CP back CP periods.

Figure 6:

Normalized integrated intensities of transferred and residual polarization pathways of ¹³C–¹H and ¹⁵N–¹H back CP obtained from the spectra reported in Figure 5C using the pulse sequences of Figure 5A.

Figure 7:

(A) Pulse sequence for the H-MAeSTOSO-4 experiment for simultaneous acquisition of (H)CHH, (H)NHH (RFDR=1.47 ms), H(N)HH (RFDR=0.74ms), and (H)NH experiments. Corresponding 2D spectra on SatP membrane protein are shown in (B). Suppression of ¹H signals from water and lipids is obtained by using t_supp=200ms. Prior to 3^rd and 4^th acquisitions 0.25*t_supp (=50 ms) period was used for eliminating waters signals. (C) 2D spectra (H)NHH and (H)NH obtained from H-MAeSTOSO-4 pulse sequence in the absence of 0.25*t_supp period, showing the residual water signals.

Figure8:

(A) Pulse sequence for the H-MAeSTOSO-10 experiment for simultaneous acquisition of ten 2D spectra. (B) Application of H-MAeSTOSO-10 on the NAVL (N-acetyl-Val-Leu) dipeptide.