DOI: 10.1002/chem.201500279
Communication
& NMR Spectroscopy
Labeling Strategy and Signal Broadening Mechanism of Protein
NMR Spectroscopy in Xenopus laevis Oocytes
Yansheng Ye,[a, b] Xiaoli Liu,[a] Yanhua Chen,[a, b] Guohua Xu,[a] Qiong Wu,[a] Zeting Zhang,[a]
Chendie Yao,[a, b] Maili Liu,[a] and Conggang Li*[a]
Chem. Eur. J. 2015, 21, 1 – 6
These are not the final page numbers! ÞÞ
1
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
Communication
Abstract: We used Xenopus laevis oocytes, a paradigm for
a variety of biological studies, as a eukaryotic model
system for in-cell protein NMR spectroscopy. The small
globular protein GB1 was one of the first studied in Xenopus oocytes, but there have been few reports since then
of high-resolution spectra in oocytes. The scarcity of data
is at least partly due to the lack of good labeling strategies and the paucity of information on resonance broadening mechanisms. Here, we systematically evaluate isotope enrichment and labeling methods in oocytes injected
with five different proteins with molecular masses of 6 to
54 kDa. 19F labeling is more promising than 15N, 13C, and
2
H enrichment. We also used 19F NMR spectroscopy to
quantify the contribution of viscosity, weak interactions,
and sample inhomogeneity to resonance broadening in
cells. We found that the viscosity in oocytes is only about
1.2 times that of water, and that inhomogeneous broadening is a major factor in determining line width in these
cells.
Figure 1. 15N-1H HSQC spectra of 15N-enriched PDI and SYN in buffer (top
row) and Xenopus oocytes (bottom row).
resolution spectrum in oocytes. The UBQ spectrum, however, is
nearly undetectable because of the interactions between the
hydrophobic surface patch and other cytoplasmic molecules.[3b]
CaM displays broad cross-peaks. PDI, the largest protein examined in oocytes, shows a few cross-peaks, mainly from its disordered C terminus. The disordered protein SYN, which is studied in Xenopus oocytes by using NMR spectroscopy for the first
time here, gave a spectrum similar to that obtained in dilute
solution, except that its resonances were broadened. No protein cross-peaks were detected in solutions taken from above
the oocytes (ca. 200 mL) after the experiment, indicating that
protein leakage did not occur.
Deuteration reduces dipolar interactions, which should enhance the resolution of 15N-1H HSQC spectra.[4] Injection of deuterated, 15N-enriched CaM gave improved sensitivity and resolution (Figure 2 a and 2 b), but some cross-peaks that were
present in buffer, were absent in oocytes (Figure 2 c). The residues with missing resonances are presumably located near the
Ca2 + binding sites, in which exchange between the Apo- and
Mg2 + -bound states might make a large contribution to linebroadening. The cross-peaks from the 2H,15N-enriched CaM in
oocytes nearly overlay those from the Ca2 + -free (Apo) form in
buffer (Figure 2 C), indicating that the injected CaM mainly
exists as Ca2 + -free form in oocytes.
Having shown that 13C-methyl enrichment of Ile, Leu, Val,
and Ala (ILVA enrichment) is beneficial for detecting proteins in
E. coli,[4] we acquired 1D 13C spectra and 2D 13C-1H HMQC spectra of ILV-enriched CaM in oocytes (Figure 3). In contrast to E.
coli, strong background signals in the methyl region were present. In both the 1D and 2D spectra, most of target protein resonances are below or overlapped with the background, but
a few CaM resonances peaks can be distinguished in 1D spectrum from oocytes. The few resolved peaks (red arrows in Figure 3 C) from injected Apo-CaM have almost the same chemical
shifts as those from Apo-CaM in buffer (Figure 3 a), indicating
that the Ca2 + -free form of CaM exists in the oocytes, consistent
with the 15N-1H HSQC results.
19
F NMR spectroscopy is promising because there are few
background signals, and it can be used to study large proteins
in E. coli.[4, 5] However, its applicability to Xenopus oocytes has
NMR spectroscopy can provide conformational and dynamic
information about proteins at the atomic level in living cells.[1]
Bacterial cells, almost exclusively Escherichia coli, are the most
popular system for in-cell protein NMR spectroscopy because
of the ease of isotope enrichment and labeling.[2] Xenopus
laevis oocytes, a model eukaryotic cell system, can be employed through micro-injection of proteins labeled or enriched
with NMR-active nuclei. Unfortunately, few proteins have been
studied in oocytes.[3] One reason for the lack of data might be
the insensitivity of 15N-1H HSQC spectra, the most common experiment for in-cell protein NMR spectroscopy. Here we explore labeling strategies and resonance broadening mechanisms in Xenopus oocytes with the goal of expanding the application of this cell type.
Four globular proteins, the B1 domain of protein G (GB1,
6.3 kDa), ubiquitin (UBQ, 8.5 kDa), calmodulin (CaM, 16.8 kDa),
and protein disulfide isomerase (PDI, 54.4 kDa), and one disordered protein, a-synuclein (SYN, 14.4 kDa), were chosen to
evaluate isotope enrichment and labeling techniques.
First, the HSQC spectra of the uniformly 15N-enriched proteins were acquired in simple buffered solution and in injected
oocytes (Figure 1). GB1, UBQ, and CaM spectra have been reported,[3a, b] and are shown in Figure S1 in the Supporting Information. Consistent with other reports,[3a] GB1 yielded a high[a] Y. Ye, X. Liu, Y. Chen, G. Xu, Q. Wu, Z. Zhang, C. Yao, Prof. M. Liu, Prof. C. Li
Key Laboratory of Magnetic Resonance in Biological Systems State
Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics
Wuhan Center for Magnetic Resonance Department
Wuhan Institute of Physics and Mathematics
Chinese Academy of Sciences, Wuhan, 430071 (P.R. China)
E-mail: [email protected]
[b] Y. Ye, Y. Chen, C. Yao
Graduate University of Chinese Academy of Sciences
Beijing, 100049 (P.R. China)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201500279.
&
&
Chem. Eur. J. 2015, 21, 1 – 6
www.chemeurj.org
2
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
Communication
The width is less than that observed in E. coli, suggesting the
protein experiences weaker interactions in oocytes. CaM contains two tyrosine residues (Y99
and Y138), and two peaks were
observed in oocytes. Their
widths are similar to those observed in E. coli. The 19F NMR
shifts in oocytes are almost the
same as those from 3FY-labeled
Figure 2. 15N-1H HSQC spectra of 15N-enriched CaM a) without and b) with deuteration in Xenopus oocytes. 1D
Apo-CaM in buffer, indicating
cross sections, taken at the position indicated by the dashed line, are shown above the spectra. Spectra comprise
that the injected CaM mainly re2
15
2+
64 transients and are plotted at equal contour levels. c) HSQC spectra of H-, N-enriched Ca -free CaM in vitro
mains in the Ca2 + -free form,
(red) and in oocytes (blue).
consistent with the 15N-1H HSQC
data, the 1D 13C data, and previ[3b]
ous results. SYN contains four tyrosine residues (Y39, Y125,
Y133, and Y136), but only one broad 19F envelope is observed
in oocytes because of overlap, and the line width in oocytes is
similar to that observed in E. coli. PDI contains five tryptophan
residues (W35, W111, W347, W379, and W390). Four resolved
resonances were observed in oocytes, and their widths were
less than those observed in E. coli, in which two strong broad
resonances were observed because of overlap with resonances
from free 5FW and 5-fluoroindole (5FI). In summary, the
19
F NMR resonances of all five proteins can be observed in injected oocytes, and their widths are similar to, or less than,
Figure 3. Spectra of 13C-methyl ILV-labelled a) Ca2 + -free and b) Ca2 + -bound
those observed in E. coli cells. 19F NMR spectroscopy is more
CaM in buffer, Xenopus oocytes injected with c) Ca2 + -free CaM, d) a control
spectrum of oocytes without injected protein, and e) 13C-1H HMQC spectra
promising for studying larger proteins in Xenopus oocytes
of Ca2 + -free (red) and Ca2 + -bound CaM (cyan) in buffer, Ca2 + free CaM in
compared with isotope enrichment techniques.
oocytes (blue), and a control spectrum of oocytes without injected protein
Viscosity, weak interactions, and sample inhomogeneity are
(green).
the main reasons that resonances from inside E. coli cells are
broader than those obtained in simple buffers.[6] Here, we
quantify the source of the broadening in Xenopus oocytes by
using 5FW-labeled GB1 and 3FY-labeled Y3FGB1 (tyrosine
changed to phenylalanine at position 3), and our previously reported methods for assessing cytoplasmic viscosity.[6b] The intracellular viscosity is obtained from the 19F longitudinal relaxation time (T1), based on the linear dependence of T1 on solution viscosity, with the assumption that the internal dynamics
of the labeled residue is the same in cells and buffer, which we
have shown to be true for W43 and Y45.[6b] The viscosity of
Xenopus oocyte cytosol is 1.2–1.3 times that of water (Figure S2 A and B in the Supporting Information), less viscous
Figure 4. 19F NMR spectra of 5FW-labeled GB1 and PDI and 3FY-labeled UBQ,
than the E. coli cytoplasm, which is about twice that of water.
CaM and SYN in buffer, in Xenopus oocytes and overlaid spectra of these
proteins in Xenopus oocytes (black) and E. coli cells (red). The E. coli spectra
The T1 of Y33 is not linearly dependent on viscosity (Figof 3FY-labeled UBQ, CaM, and SYN have been published.[5a] The asterisks inure S2 C in the Supporting Information) possibly because of
dicate resonances from free 5FW (5-fluorotryptophan), 5FI (5-fluoroindole),
the internal dynamics on the 20–100 ps time scale. This residue
and 3FY (3-fluorotyrosine) resonances.
is not a suitable probe as we have discussed previously.[6b] The
19
F transverse relaxation rate (R2) is more sensitive to weak interactions than T1, and the difference in viscosities inferred
not been tested. 19F spectra of 5-fluorotryptophan (5FW)-labeled GB1 and PDI, and 3-fluorotyrosine (3FY)-labeled UBQ,
from R2 and T1 is indicative of these interactions. GB1 experiCaM, and SYN in oocytes, E. coli cells, and buffer are shown in
enced an apparent viscosity of only about twice that of water
Figure 4. GB1 contains one tryptophan residue (W43), and only
in oocytes (Figure S2 D and E in the Supporting Information),
one 19F resonance was observed in oocytes. The line width is
but the value was 6–11 times that of water in E. coli.[6] These
similar to that observed in E. coli. UBQ has a single tyrosine resresults suggest that GB1 experiences weaker interactions in
idue (Y59), and only one resonance was observed in oocytes.
Xenopus oocytes than it does in E. coli and that the oocyte cyChem. Eur. J. 2015, 21, 1 – 6
www.chemeurj.org
These are not the final page numbers! ÞÞ
3
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
Communication
toplasm is less viscous than that of E. coli. The conclusion
seems to contradict our observation that GB1 resonances display similar widths in Xenopus oocytes and in E. coli cells. What
are the reasons for this apparent contradiction?
Homogeneous broadening arises from cellular viscosity and
weak interactions, whereas inhomogeneous broadening arises
from subtle chemical shift differences involving sample inhomogeneity. Inhomogeneous broadening is not a dominant
factor in E. coli cells.[6b] We used 19F NMR spectroscopy and
5FW-labeled GB1 and 3FY-labeled Y3FGB1 to quantify the contribution of homogeneous broadening and inhomogeneous
broadening to the width of resonances in oocytes. The widths
at half-height (W1/2) were obtained from line shape analysis by
using Lorentz/Gauss deconvolution. Contributions from homogeneous broadening (R2/p) were obtained from Carr–Purcell–
Meiboom–Gill (CPMG) experiments. Contributions from inhomogeneous broadening (R2(inho)/p) were then estimated from
the difference between the 19F line width at half-height (W1/2)
and homogeneous broadening (Figure 5). For W43 of GB1, inhomogeneous broadening (13.8 Hz) contributes 15 % to the
E. coli, but for Xenopus oocytes, the main causes are weak interactions and inhomogeneous broadening.
In this study, we have demonstrated that 19F labeling is
a good first choice for studying globular and disordered proteins in Xenopus oocytes, especially compared with conventional 15N- or 13C-methyl enrichment. However, if reference
spectrum subtraction were used to reduce the severe background signal in the methyl region, methyl labeling might be
advantageous for larger proteins because of the methyl-TROSY
effect.[7] Uniform 15N enrichment with deuteration is a useful
option when the goal is to study the effect of the intracellular
environment across the entire protein. By using 19F labeling,
we found that, unlike E. coli cells, the viscosity in oocytes is
only about 1.2 times that of water and that inhomogeneous
broadening contributes 60–70 % to the line width. This lower
intracellular viscosity could be a general property of eukaryotic
cells,[8] perhaps as a result of their lower concentration of macromolecules.[1c] Our labeling and broadening mechanism efforts in Xenopus oocytes open additional opportunities for protein NMR spectroscopy in this model system.
Experimental Section
Details are given in the Supporting Information.
Acknowledgements
We thank Dengdi Li (School of Life Sciences, Central China
Normal University) for help with oocytes preparation. This
work is supported by the Ministry of Science and Technology
of China (grant 2013CB910200), the 1000 Young Talents Program, and the National Natural Sciences Foundation of China
(grants 21173258, 21120102038 and 21221064).
Figure 5. Quantification of the homogenous and inhomogeneous broadening to the 19F resonances from 5FW-labeled GB1 and 3FY-labeled Y3FGB1 in
E. coli (red) and Xenopus oocytes (black). Homogenous broadening was estimated from R2/p. Inhomogeneous broadening [R2(inho)/p] was estimated
from the difference between the width at half-height (W1/2) and the homogeneous broadening. W1/2 was obtained from line-shape fitting by using Lorentz/Gauss deconvolution. R2 was measured by using the CPMG experiment.
Values of W1/2, R2/p, R2(inho)/p and percent inhomogeneous broadening are
indicated. For W43 of GB1, inhomogeneous broadening is 13.8 Hz and contributes about 15 % to the width at half-height in E. coli, but contributes
about 73 % to the width in oocytes. For Y45 and Y33 in GB1, inhomogeneous broadening contributes about 23 and about 9 % in E. coli and about 59
and about 69 % in oocytes, respectively.
Keywords: in-cell NMR spectroscopy · NMR spectroscopy ·
proteins · Xenopus laevis oocytes
[1] a) Z. Serber, V. Dçtsch, Biochemistry 2001, 40, 14317 – 14323; b) Y. Ito, P.
Selenko, Curr. Opin. Struct. Biol. 2010, 20, 640 – 648; c) F. X. Theillet, A. Binolfi, T. Frembgen-Kesner, K. Hingorani, M. Sarkar, C. Kyne, C. G. Li, P. B.
Crowley, L. Gierasch, G. J. Pielak, A. H. Elcock, A. Gershenson, P. Selenko,
Chem. Rev. 2014, 114, 6661 – 6714; d) A. E. Smith, Z. Zhang, G. J. Pielak, C.
Li, Curr. Opin. Struc. Biol. 2014, 30C, 7 – 16; e) L. Banci, L. Barbieri, I. Bertini,
E. Luchinat, E. Secci, Y. Zhao, A. R. Aricescu, Nat. Chem. Biol. 2013, 9,
297 – 299; f) E. Luchinat, L. Barbieri, J. T. Rubino, T. Kozyreva, F. Cantini, L.
Banci, Nat. Commun. 2014, 5, 5502.
[2] a) R. Hnsel, L. M. Luh, I. Corbeski, L. Trantirek, V. Dçtsch, Angew. Chem.
Int. Ed. 2014, 53, 10300 – 10314; Angew. Chem. 2014, 126, 10466 – 10480;
b) Z. Serber, A. T. Keatinge-Clay, R. Ledwidge, A. E. Kelly, S. M. Miller, V.
Dçtsch, J. Am. Chem. Soc. 2001, 123, 2446 – 2447; c) Z. Serber, R. Ledwidge, S. M. Miller, V. Dçtsch, J. Am. Chem. Soc. 2001, 123, 8895 – 8901;
d) D. Sakakibara, A. Sasaki, T. Ikeya, J. Hamatsu, T. Hanashima, M. Mishima, M. Yoshimasu, N. Hayashi, T. Mikawa, M. Walchli, B. O. Smith, M. Shirakawa, P. Guntert, Y. Ito, Nature 2009, 458, 102 – 105; e) Y. Ito, T. Mikawa,
B. O. Smith, Methods Mol. Biol. 2012, 895, 19 – 31.
[3] a) P. Selenko, Z. Serber, B. Gadea, J. Ruderman, G. Wagner, Proc. Natl.
Acad. Sci. USA 2006, 103, 11904 – 11909; b) T. Sakai, H. Tochio, T. Tenno, Y.
Ito, T. Kokubo, H. Hiroaki, M. Shirakawa, J. Biomol. NMR 2006, 36, 179 –
188; c) J. F. Bodart, J. M. Wieruszeski, L. Amniai, A. Leroy, I. Landrieu, A.
Rousseau-Lescuyer, J. P. Vilain, G. Lippens, J. Magn. Reson. 2008, 192,
W1/2 in E. coli, but 73 % in oocytes. Similar results were obtained with 3FY-labeled Y3FGB1. For residues Y45 and Y33, inhomogeneous broadening contributes 23 and 9 %, respectively, to the width in E. coli, but 59 and 69 % to the width in oocytes. These results indicate that inhomogeneous broadening
is a major factor affecting protein NMR spectra in oocytes. The
inhomogeneous broadening may be due to the larger size of
oocytes. In summary, cytoplasmic viscosity and interactions are
the main reasons for the low sensitivity of NMR experiments in
&
&
Chem. Eur. J. 2015, 21, 1 – 6
www.chemeurj.org
4
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!
Communication
252 – 257; d) P. Selenko, D. P. Frueh, S. J. Elsaesser, W. Haas, S. P. Gygi, G.
Wagner, Nat. Struct. Mol. Biol. 2008, 15, 321 – 329; e) P. Selenko, G.
Wagner, J. Struct. Biol. 2007, 158, 244 – 253.
[4] G. Xu, Y. Ye, X. Liu, S. Cao, Q. Wu, K. Cheng, M. Liu, G. J. Pielak, C. Li, Biochemistry 2014, 53, 1971 – 1981.
[5] a) C. Li, G. F. Wang, Y. Wang, R. Creager-Allen, E. A. Lutz, H. Scronce, K. M.
Slade, R. A. Ruf, R. A. Mehl, G. J. Pielak, J. Am. Chem. Soc. 2010, 132, 321 –
327; b) J. L. Kitevski-LeBlanc, R. S. Prosser, Prog. Nucl. Magn. Reson. Spectrosc. 2012, 62, 1 – 33.
[6] a) Q. Wang, A. Zhuravleva, L. M. Gierasch, Biochemistry 2011, 50, 9225 –
9236; b) Y. Ye, X. Liu, Z. Zhang, Q. Wu, B. Jiang, L. Jiang, X. Zhang, M. Liu,
G. J. Pielak, C. Li, Chem. Eur. J. 2013, 19, 12705 – 12710.
Chem. Eur. J. 2015, 21, 1 – 6
www.chemeurj.org
These are not the final page numbers! ÞÞ
[7] R. Sprangers, L. E. Kay, Nature 2007, 445, 618 – 622.
[8] a) J. V. Sehy, J. J. Ackerman, J. J. Neil, Magn. Reson. Med. 2002, 48, 42 – 51;
b) S. P. Williams, P. M. Haggie, K. M. Brindle, Biophys. J. 1997, 72, 490 –
498; c) R. E. London, C. T. Gregg, N. A. Matwiyoff, Science 1975, 188, 266 –
268.
Received: January 21, 2015
Published online on && &&, 0000
5
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
Communication
COMMUNICATION
& NMR Spectroscopy
Various isotope enrichment and labelling techniques in protein NMR spectroscopy have been evaluated in Xenopus laevis oocytes and 19F NMR spectroscopy proves to be a promising technique for investigating larger proteins.
The NMR resonance broadening mechanism is also clarified in Xenopus laevis
oocytes.
Y. Ye, X. Liu, Y. Chen, G. Xu, Q. Wu,
Z. Zhang, C. Yao, M. Liu, C. Li*
&& – &&
Labeling Strategy and Signal
Broadening Mechanism of Protein
NMR Spectroscopy in Xenopus laevis
Oocytes
&
&
Chem. Eur. J. 2015, 21, 1 – 6
www.chemeurj.org
6
2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!