Letter pubs.acs.org/ac
Quantitative Imaging of Single Unstained Magnetotactic Bacteria by Coherent X‑ray Diffraction Microscopy Jiadong Fan,† Zhibin Sun,† Jian Zhang,† Qingjie Huang,‡ Shengkun Yao,† Yunbing Zong,† Yoshiki Kohmura,§ Tetsuya Ishikawa,§ Hong Liu,† and Huaidong Jiang*,† †
State Key Laboratory of Crystal Materials, Shandong University, Jinan 250100, China School of Information Science and Engineering, Shandong University, Jinan 250100, China § RIKEN SPring-8 Center, 1-1-1, Kouto, Sayo, Hyogo 679-5148, Japan ‡
S Supporting Information *
ABSTRACT: Novel coherent diffraction microscopy provides a powerful lensless imaging method to obtain a better understanding of the microorganism at the nanoscale. Here we demonstrated quantitative imaging of intact unstained magnetotactic bacteria using coherent X-ray diffraction microscopy combined with an iterative phase retrieval algorithm. Although the signal-to-noise ratio of the X-ray diffraction pattern from single magnetotactic bacterium is weak due to low-scattering ability of biomaterials, an 18.6 nm halfperiod resolution of reconstructed image was achieved by using a hybrid input-output phase retrieval algorithm. On the basis of the quantitative reconstructed images, the morphology and some intracellular structures, such as nucleoid, polyβhydroxybutyrate granules, and magnetosomes, were identified, which were also confirmed by scanning electron microscopy and energy dispersive spectroscopy. With the benefit from the quantifiability of coherent diffraction imaging, for the first time to our knowledge, an average density of magnetotactic bacteria was calculated to be ∼1.19 g/cm3. This technique has a wide range of applications, especially in quantitative imaging of low-scattering biomaterials and multicomponent materials at nanoscale resolution. Combined with the cryogenic technique or X-ray free electron lasers, the method could image cells in a hydrated condition, which helps to maintain their natural structure.
A
conditions.15,16 Although imaging of hydrated cells is a new research highlight in the future, there are still lots of challenges and so far most bioimaging with CDI method is performed in dehydrated conditions. Here, taking advantages of good penetration and quantifiability of X-rays, we report quantitative, high-resolution, and high-contrast imaging of whole unstained low-scattering magnetotactic bacteria (MTB) with coherent Xray diffraction microscopy. MTB are aquatic prokaryotes and widely distributed in the aquatic environment.17 Since the bacteria with motility directed by the local geomagnetic field was observed in marine sediments,18 MTB have attracted a lot of research interests. The intracellular catenulate magnetosome (MS) particles inside the MTB, which are composed of magnetic iron-bearing inorganic nanocrystals, have potential applications in many fields due to the novel and special magnetic, physical, and perhaps optical properties.19 Some imaging techniques such as transmission electron microscopy (TEM), magnetic force
n accurate understanding of structure and function of biomaterials needs powerful microscopy that could provide not only images with high resolution but also quantitative images in situ. With the benefits of long penetration depth and short wavelength, X-ray microscopy is considered as an ideal tool to achieve high-resolution images without destruction in dehydrated and hydrated conditions. However, for conventional X-ray microscopy, resolution is limited by precise focusing devices for X-rays although significant progress has been made. Coherent X-ray diffraction imaging (CDI) is a promising lensless approach currently under rapid development to achieve high resolution. By recording coherent diffraction patterns in the far field, images of noncrystalline specimens can be retrieved using iterative phase retrieval algorithms from the patterns. The idea of CDI was first suggested by Sayre and demonstrated with X-rays by Miao et al.1,2 In the past decades, CDI has been widely applied to quantitatively determining microstructures at nanoscale resolution.3−12 For biological specimens, nondestructive imaging with high resolution and high contrast is crucial for revealing the relationship between structure and biological function in organisms. Efforts in imaging biomaterials under hydrated conditions were also undertaken with CDI, including frozen-hydrated13,14 and living © 2015 American Chemical Society
Received: February 24, 2015 Accepted: May 26, 2015 Published: May 26, 2015 5849
DOI: 10.1021/acs.analchem.5b00746 Anal. Chem. 2015, 87, 5849−5853
Letter
Analytical Chemistry
were illuminated by plane wave X-rays. In the far field, the diffracted signals from the AMB-1 cells were recorded by the CCD detector. Though the highly sensitive CCD was utilized, due to low scattering ability of biomaterials, the diffraction pattern of a single cell has a low signal-to-noise ratio, especially at high frequencies. To enhance the dynamic range of diffraction intensity and increase signal-to-noise ratio,26 lowresolution (LR) and high-resolution (HR) diffraction patterns were measured from the AMB-1 sample, separately. The total data acquisition time was about 3 h including 1 h for LR data and 2 h for HR data. By merging LR and HR patterns, the missing data due to the beamstop was confined in the centrospeckle of diffraction patterns, which facilitates reliability of image reconstruction.27 Figure 2a shows the diffraction pattern
microscopy (MFM), and scanning transmission X-ray microscopy (STXM) have been used to investigate the structure of MTB and MS.20−22 For instance, MFM with magnetic probe was used to study the extracted MSs and observed the MS trains structures.22 With the benefit from the high resolution of TEM, magnetosome vesicles were usually observed in the ultrathin section in MTB.21 To obtain a thorough understanding on the architecture of whole MTB, a quantitative analysis of MSs in situ is important and meaningful for determination of the arrangement and distribution of MS chains. However, for low-scattering unstained biomaterials, it is still a challenge to get quantitative, high-resolution, highcontrast images of whole cells. In this letter, we investigated the quantitative high-contrast imaging of single unstained Magnetospirillum magneticum AMB-1 with CDI at the nanoscale.23 Cells of Magnetospirillum magnetotacticum AMB-1 were cultured in an enriched magnetic spirillum growth medium (EMSGM) at 28 °C as described in previous study.24 The mutants of AMB-1 generated by a hyperactive mariner transposon was achieved and then picked.25 To fix the mutants of AMB-1, the transposon mutants were washed twice with sterile water and treated with 3.2% formaldehyde for 1 min, followed by adding 12% of glutaraldehyde to 0.25% of the final concentration. After fixation, the mutants of AMB-1 were stored in ethanol. To obtain single isolated AMB-1 cells, the sample solution was diluted with ethanol to an appropriate concentration and deposited on 30 nm thick Si3N4 membranes. The well-isolated AMB-1 cells were selected under an optical microscope and used for the coherent X-ray diffraction experiment. The CDI experiment was carried out on an undulator beamline at a third generation synchrotron radiation facility (SPring-8,BL 29XU). During the experiment, 5 keV monochromatic X-ray was defined by a 20 μm pinhole that was placed in front of the sample chamber. Two silicon apertures were mounted downstream of the pinhole to clean stray scattering from the pinhole and other upstream optical devices. In order to release the dark noise, a liquid nitrogen cooled CCD detector with 1340 × 1300 pixels was used to record diffraction patterns from samples. The distance between sample and CCD was set to 105 cm. A beamstop was mounted just in front of the CCD to block the direct X-ray beam. All these devices were in a vacuum condition of ∼10−6 Torr to eliminate absorption and scattering of the air. Figure 1 shows a schematic layout of a powerful X-ray diffraction microscope we used. The unstained AMB-1 cells
Figure 2. (a) Typical diffraction pattern of a single magnetotactic bacterium. Insert is the central part with missing data due to a beamstop. Parts b and c are two independent reconstructed images with different random initial phases from the diffraction pattern (a). Line scans (d) along the white dashed lines in parts b and c show a good consistency of reconstruction and inner density distribution.
after assembling LR and HR patterns. The zoomed in view of the central part (Figure 2a, insert) indicates that the LR and HR patterns were merged precisely without obvious trace, which implies the diffraction patterns are quite consistent at low frequencies. To reconstruct real-space images from the far-field diffraction patterns, Fienup’s hybrid input-output (HIO) algorithm combined with the oversampling method was used to conduct the phase retrieval.28 Without any prior information about the sample, a random phase was added to the measured diffraction pattern and a rectangular support was enforced during the reconstruction. To obtain accurate results, 16 replicated reconstructions were conducted for one run. Each reconstruction was iterated back and forth between real and reciprocal space with constraints. For example, in real space, the calculated positive electron density inside the support was reserved. While the electron density outside the support and negative electron density inside the support were gradually pushed to zero as described in eq 1,29 ⎧ r∈S ⎪ ρn ′(r ) ρn (r ) = ⎨ ⎪ ⎩ ρn − 1(r ) − βρn ′(r ) r ∉ S
Figure 1. Schematic of a coherent diffraction microscope.
and or
ρn ′(r ) ≥ 0 ρn ′(r ) < 0 (1)
5850
DOI: 10.1021/acs.analchem.5b00746 Anal. Chem. 2015, 87, 5849−5853
Letter
Analytical Chemistry where S represents the finite support, β a constant, and ρn the electron density at the nth iteration. During reconstruction, a Gaussian distribution kernel was used to smooth the noise out of support, which obviously reduced the noise effect. In order to evaluate the quality of the reconstructions, an error function Rerr was defined as eq 2, where |Fexp| and |Fcal| are amplitude of the measured and calculated diffraction patterns, respectively (2). R err =
materials, determination of density has significant biological and diagnostic applications, including intracellular organelles identification, segmentations, in situ location, etc. On the basis of eq 4,30 I(0, 0) = I0re 2 |F(0, 0)|2
(2)
To evaluate the reliability of image reconstructions, we carried out two independent HIO runs with different initial phases. The two independent reconstructed images (Figure 2b,c) show a good consistency both in cell profile and intracellular structures. The difference of the two images was calculated to be R1,2 = 3.13% according to eq 3, where ρ1(x,y) and ρ2(x,y) represent the electron densities of the independent reconstructed images. Line scans in Figure 2d along the white dashed lines (Figure 2b,c) also indicate the good consistency of the inner structure. R1,2 =
∑x , y |ρ1(x , y) − ρ2 (x , y)| ∑x , y |ρ1(x , y) + ρ2 (x , y)|
(4)
the electron density of AMB-1 can be calculated, where I(0,0) is the number of diffracted X-ray photons in the forward direction (i.e., within the central pixel of the CCD), I0 the incident X-ray flux per unit area, re the classical electron radius, | F(0,0)| the total number of electrons in the AMB-1, Δs the area of the central pixel, and r the distance from the sample to the CCD. Although I(0,0) was missed because of the beamstop, it can be recovered during the reconstruction.30 The total electrons of the single AMB-1 in Figure 3a were calculated to be 1.253 × 1011 electrons. According to the formula of MTB31 and the volume estimated from the morphology by AFM, we for the first time determined an average electron density of single AMB-1 cell to be ∼3.95 × 1011 electrons/μm3, corresponding to a mass density of ∼1.19 g/cm3 which is a little higher than Escherichia coli.31,32 The slight difference of the mass densities may be due to the different types of bacteria and irons in AMB-1. On the basis of the calculated high-contrast density image, the architecture of the AMB-1 could be quantitatively analyzed. As shown in Figure 3a, the cell wall (arrow A) with low density can be easily distinguished from surroundings due to the high image contrast. According to the density distribution in Figure 3a, the ribosome-rich cytoplasm has a relatively high mass density. However, there are not many kinds of organelles in the uniform cytoplasm because AMB-1 is a kind of Gram-negative bacteria. On the basis of the analysis of shape, size, mass density, and previous study on nucleoid of D. radiodurans,33 the region with high density (arrow B) could be the nucleoid. To further confirm it, energy dispersive spectroscopy (EDS) experiment was performed in this region. The result indicates that this region is rich in phosphorus which is the main element of DNA and RNA. The relatively low density region surrounded by high density cytoplasm (arrow D) is the polyβ-hydroxybutyrate (PHB) granule according to the morphology and density. However, because of the limitation of 2D image and resolution, some small organelles cannot be clearly identified. For MTB, MSs mineralized inside cells are composed of crystals of magnetite (Fe3O4) or greigite (Fe3S4).34,35 The AMB-1 cells used in our experiment usually have long Fe3O4 MS chains,23 and the isolated MS is about 35−120 nm enveloped with a membrane of lipid bilayer.36,37 As shown in Figure 3a, the chain-like regions with high density (arrows C and E) were selected to perform the EDS experiment. The result indicates that iron is a main element in these regions (Figure 4b, insert). Thus, considering the mass densities, element composition, and chain-like structure, the regions C and E were identified as two separated MS chains along the long axis of the cell, which were formed in the mutants of AMB-1 generated by a hyperactive mariner transposon.25 The isolated dark spots with high density in each chain are about 2− 4 pixels corresponding to 37−74 nm, which is in good agreement with the size of MS in previous reports.38 Moreover, dual-energy STXM imaging of another similar magnetotactic bacterium prepared in same conditions was performed at below and above Fe L3 absorption edge separately (Supporting
∑ ||Fcal| − |Fexp|| ∑ |Fexp|
Δs r2
(3)
In order to further verify the image reconstruction of the AMB-1 cell (Figure 3a), a scanning electron microscopy (SEM)
Figure 3. (a) Reconstruction image of a single bacterial cell from the diffraction pattern. The color represents electron density distribution of the cell. SEM image (b) shows the surface morphology of the same sample. (c) AFM image of a similar bacterial cell prepared at the same condition.
image of the same sample was taken (Figure 3b). To eliminate the charging effect, a 5 nm thick Au film was deposited on the sample. By comparison, the reconstructed image is in excellent agreement with the SEM image. Figure 3c shows an image of a similar AMB-1 on the same membrane using atomic force microscope (AFM). In comparison with Figure 3b,c, the benefit from the longer penetration length of X-rays and the result of CDI reconstruction in Figure 3a provides not only the sample profile but also buried intracellular structure information with high contrast. As X-rays interact with electrons of atoms, absolute electrons densities of reconstructed image could be calculated by CDI. Combining with composition elements information, an accurate mass density of specimen could be determined, which makes quantitative analysis possible. For biological 5851
DOI: 10.1021/acs.analchem.5b00746 Anal. Chem. 2015, 87, 5849−5853
Letter
Analytical Chemistry
resolution of 25 nm. Considering the overall signals and signalto-noise ratio, finally a 700 × 700 matrix of diffraction pattern with valid signals was selected for reconstruction. According to the equation of d = λ/2 sin(θ), where λ is the wavelength and θ half of the diffraction angle, the resolution was calculated to be 37.2 nm (i.e., pixel size of 18.6 nm). In order to confirm the reliability of the resolution and quality of the reconstructed image, we quantitatively analyzed the phase retrieval transfer function (PRTF) of the diffraction pattern in reciprocal space.39 Because phase errors cause image aberrations during reconstructions, the actual spatial resolution of the image could be worse than the calculated resolution if the reconstruction is incorrect. Therefore, we calculated the PRTF curve (Figure 5b) by comparing the Fourier amplitudes calculated from the reconstructed image to the measured ones. On the basis of the criterion of PRTF = 1/e,5,40 the PRTF curve indicates a reliable resolution of 37.2 nm. The area on the left of the red dash-dot line represents the missing data because of the beamstop. To further illustrate the resolution, line scans across the typical MSs (i.e., along black and blue dash lines shown in Figure 4a) were taken. The line scan results indicate that the typical MSs could be identified with good contrast. As shown in Figure 5c, the MSs edges were presented in one to two pixels, corresponding to an 18.6 nm half period resolution or a 37.2 nm full period resolution of the MTB image in physiological conditions. In addition, for biomaterials, radiation damage restricts the achievable resolution of diffraction patterns before degeneration emerging. An appropriate exposure less than the maximum tolerable dose, is a main issue for a successful experiment on low-scattering biomaterial imaging. The radiation dose deposited in the AMB-1 during our experiment was calculated to be ∼3.68 × 107 Gy. On the basis of a previous study,41 such radiation dose would not cause radiation damage of bacteria. In summary, we demonstrated quantitative imaging of single unstained low-scattering AMB-1 with high spatial resolution and high contrast using coherent X-ray diffraction microscopy. On the basis of the quantitative reconstructed image, the average mass density of AMB-1 was determined to be ∼1.19 g/ cm3. A spatial resolution of 37.2 nm (18.6 nm half-period resolution) was calculated and confirmed by both diffraction pattern in reciprocal space and reconstructed image in real space. Taking advantage of the high-contrast density distribution, we quantitatively analyzed the structure of AMB-1 and identified the location of nucleoid, PHB granules, and separated MS chains inside the cytoplasm. This work hence paves a way for unveiling the buried structures of a wide range of biological specimens at the nanoscale. Although imaging of hydrated biological samples currently faced some challenges,13,14 by more efficient combination of X-ray diffraction microscopy with the frozen-hydrated technique in the future, 5−10 nm resolution imaging of whole frozen-hydrated cells will be achievable considering the radiation damage.42 Besides, combining single-shot X-rays free electron lasers (XFEL), the CDI method has demonstrated imaging of small living cells with about 70 nm resolution.15,16 With the rapid development of the most promising X-ray source, high-resolution quantitative imaging of hydrated biological samples will be achieved.
Figure 4. Zoomed in view (a) of a rectangular region in Figure 3a shows some dark spots with high densities. (b) Line scan along the white dashed line and EDS spectrum at the position around the arrow in part a indicate a chain of magnetosomes.
Information). Some isolated magnetosomes and a magnetosome chain were observed, which is consistent with the CDI result. To further quantitatively analyze the MS chains, a line scan along the white dashed line in Figure 4a was performed. The line scan result (Figure 4b) shows that four MSs are contained in the chain due to the high contrast of mass density. By removing the sample thickness effect, the mass density of MS was estimated to be about 4.6−5.3 g/cm3, which is well consistent with a Fe3O4 mass density of 5.18 g/cm3. High spatial resolution imaging plays an increasingly important role in that it can provide powerful insights into the nanoworld. For CDI in biomaterials, although the spatial resolution of the reconstructed image is principally limited by the X-ray wavelength, at present, radiation damage restricts the achievable resolution. For the butterfly like diffraction pattern of the AMB-1 cell (Figure 2a), in comparison with the vertical direction, the signals in the horizontal direction are much stronger. The power spectral density (PSD) curve (Figure 5a) indicates the diffraction signals actually extend beyond a spatial
Figure 5. (a) Power spectral density of the diffraction pattern indicates a valid special frequency of 26.9 μm−1, which corresponds to 37.2 nm resolution. (b) PRTF curve of the diffraction pattern shows that the resolution of the reconstructed image is estimated to be 37.2 nm, based on the criterion of PRTF = 1/e. (c) Line scans along the black and blue dashed lines across the typical MSs shown in Figure 4a. A 37.2 nm full period resolution (two pixels) and an 18.6 half period resolution (one pixel) is achieved. 5852
DOI: 10.1021/acs.analchem.5b00746 Anal. Chem. 2015, 87, 5849−5853
Letter
Analytical Chemistry
■
(15) Kimura, T.; Joti, Y.; Shibuya, A.; Song, C.; Kim, S.; Tono, K.; Yabashi, M.; Tamakoshi, M.; Moriya, T.; Oshima, T. Nat. Commun. 2014, 5, 3052. (16) van der Schot, G.; Svenda, M.; Maia, F. R.; Hantke, M.; DePonte, D. P.; Seibert, M. M.; Aquila, A.; Schulz, J.; Kirian, R.; Liang, M. Nat. Commun. 2015, 6, 5704. (17) Faivre, D.; Schüler, D. Chem. Rev. 2008, 108, 4875−4898. (18) Blakemore, R. Science 1975, 190, 377−379. (19) Sun, J. B.; Duan, J. H.; Dai, S. L.; Ren, J.; Guo, L.; Jiang, W.; Li, Y. Biotechnol. Bioeng. 2008, 101, 1313−1320. (20) Kalirai, S. S.; Bazylinski, D. A.; Hitchcock, A. P. PLoS One 2013, 8, e53368. (21) Komeili, A.; Vali, H.; Beveridge, T. J.; Newman, D. K. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 3839−3844. (22) Suzuki, H.; Tanaka, T.; Sasaki, T.; Nakamura, N.; Matsunaga, T. Jpn. J. Appl. Phys. 1998, 37 (11A), L1343. (23) Matsunaga, T.; Sakaguchi, T.; Tadakoro, F. Appl. Microbiol. Biotechnol. 1991, 35, 651−655. (24) Yang, C.-D.; Takeyama, H.; Tanaka, T.; Matsunaga, T. Enzyme Microb. Technol. 2001, 29, 13−19. (25) Bo, T.; Wang, K.; Ge, X.; Chen, G.; Liu, W. Curr. Microbiol. 2012, 65, 98−107. (26) Xu, R.; Salha, S.; Raines, K. S.; Jiang, H.; Chen, C.-C.; Takahashi, Y.; Kohmura, Y.; Nishino, Y.; Song, C.; Ishikawa, T. Synchrotron Radiat. 2011, 18, 293−298. (27) Miao, J.; Nishino, Y.; Kohmura, Y.; Johnson, B.; Song, C.; Risbud, S. H.; Ishikawa, T. Phys. Rev. Lett. 2005, 95, 085503. (28) Fienup, J. R. Opt. Lett. 1978, 3, 27−29. (29) Miao, J.; Ishikawa, T.; Anderson, E. H.; Hodgson, K. O. Phys. Rev. B 2003, 67, 174104. (30) Miao, J.; Amonette, J. E.; Nishino, Y.; Ishikawa, T.; Hodgson, K. O. Phys. Rev. B 2003, 68, 012201. (31) Naresh, M.; Das, S.; Mishra, P.; Mittal, A. Biotechnol. Bioeng. 2012, 109, 1205−1216. (32) Godin, M.; Bryan, A. K.; Burg, T. P.; Babcock, K.; Manalis, S. R. Appl. Phys. Lett. 2007, 91, 123121. (33) Giewekemeyer, K.; Thibault, P.; Kalbfleisch, S.; Beerlink, A.; Kewish, C. M.; Dierolf, M.; Pfeiffer, F.; Salditt, T. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 529−534. (34) Frankel, R. B.; Blakemore, R. P.; Wolfe, R. S. Science 1979, 203, 1355. (35) Mann, S.; Sparks, N. C.; Frankel, R. B.; Bazylinski, D. A.; Jannasch, H. W. Nature 1990, 343, 258. (36) Bazylinski, D. A.; Garratt-Reed, A. J.; Frankel, R. B. Microsc. Res. Technique 1994, 27, 389−401. (37) Gorby, Y. A.; Beveridge, T. J.; Blakemore, R. P. J. Bacteriol. 1988, 170, 834−841. (38) Devouard, B.; Posfai, M.; Hua, X.; Bazylinski, D. A.; Frankel, R. B.; Buseck, P. R. Am. Mineral. 1998, 83, 1387−1398. (39) Chapman, H. N.; Barty, A.; Marchesini, S.; Noy, A.; Hau-Riege, S. P.; Cui, C.; Howells, M. R.; Rosen, R.; He, H.; Spence, J. C. J. Opt. Soc. Am.: A 2006, 23, 1179−1200. (40) Seibert, M. M.; Ekeberg, T.; Maia, F. R.; Svenda, M.; Andreasson, J.; Jö nsson, O.; Odić, D.; Iwan, B.; Rocker, A.; Westphal, D. Nature 2011, 470, 78−81. (41) Shen, Q.; Bazarov, I.; Thibault, P. J. Synchrotron Radiat. 2004, 11, 432−438. (42) Howells, M. R.; Beetz, T.; Chapman, H. N.; Cui, C.; Holton, J.; Jacobsen, C.; Kirz, J.; Lima, E.; Marchesini, S.; Miao, H. J. Electron Spectrosc. Relat. Phenom. 2009, 170, 4−12.
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b00746.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected]. Author Contributions
The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We thank Prof. Weifeng Liu for providing the AMB-1 samples used for this study, Prof. Jianwei Miao and Prof. Adam P. Hitchcock for stimulating discussion, and the staff at Shang-hai Synchrotron Radiation Facility for assistance with data acquisition. This work was supported by the Major State Basic Research Development Program of China (Grant 2014CB910401), the National Natural Science Foundation of China (Grants 31430031, 21390414, and U1332118), the Natural Science Foundation of Shandong Province (Grant JQ201117), and the Program for New Century Excellent Talents (Grant NCET-11-0304). Use of the RIKEN beamline (BL29XUL) at SPring-8 was supported by RIKEN.
■
REFERENCES
(1) Sayre, D.; Schlenker, M. Lect. Notes Phys. 1980, 112, 229. (2) Miao, J.; Charalambous, P.; Kirz, J.; Sayre, D. Nature 1999, 400, 342−344. (3) Jiang, H.; Ramunno-Johnson, D.; Song, C.; Amirbekian, B.; Kohmura, Y.; Nishino, Y.; Takahashi, Y.; Ishikawa, T.; Miao, J. Phys. Rev. Lett. 2008, 100, 038103. (4) Dierolf, M.; Menzel, A.; Thibault, P.; Schneider, P.; Kewish, C. M.; Wepf, R.; Bunk, O.; Pfeiffer, F. Nature 2010, 467, 436−439. (5) Shapiro, D.; Thibault, P.; Beetz, T.; Elser, V.; Howells, M.; Jacobsen, C.; Kirz, J.; Lima, E.; Miao, H.; Neiman, A. M. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 15343−15346. (6) Jiang, H.; Song, C.; Chen, C.-C.; Xu, R.; Raines, K. S.; Fahimian, B. P.; Lu, C.-H.; Lee, T.-K.; Nakashima, A.; Urano, J. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 11234−11239. (7) Nam, D.; Park, J.; Gallagher-Jones, M.; Kim, S.; Kim, S.; Kohmura, Y.; Naitow, H.; Kunishima, N.; Yoshida, T.; Ishikawa, T. Phys. Rev. Lett. 2013, 110, 098103. (8) Nishino, Y.; Takahashi, Y.; Imamoto, N.; Ishikawa, T.; Maeshima, K. Phys. Rev. Lett. 2009, 102, 018101. (9) Song, C.; Jiang, H.; Mancuso, A.; Amirbekian, B.; Peng, L.; Sun, R.; Shah, S. S.; Zhou, Z. H.; Ishikawa, T.; Miao, J. Phys. Rev. Lett. 2008, 101, 158101. (10) Miao, J.; Chen, C.-C.; Song, C.; Nishino, Y.; Kohmura, Y.; Ishikawa, T.; Ramunno-Johnson, D.; Lee, T.-K.; Risbud, S. H. Phys. Rev. Lett. 2006, 97, 215503. (11) Jiang, H.; Xu, R.; Chen, C.-C.; Yang, W.; Fan, J.; Tao, X.; Song, C.; Kohmura, Y.; Xiao, T.; Wang, Y. Phys. Rev. Lett. 2013, 110, 205501. (12) Pfeifer, M. A.; Williams, G. J.; Vartanyants, I. A.; Harder, R.; Robinson, I. K. Nature 2006, 442, 63−66. (13) Huang, X.; Nelson, J.; Kirz, J.; Lima, E.; Marchesini, S.; Miao, H.; Neiman, A. M.; Shapiro, D.; Steinbrener, J.; Stewart, A. Phys. Rev. Lett. 2009, 103, 198101. (14) Lima, E.; Wiegart, L.; Pernot, P.; Howells, M.; Timmins, J.; Zontone, F.; Madsen, A. Phys. Rev. Lett. 2009, 103, 198102. 5853
DOI: 10.1021/acs.analchem.5b00746 Anal. Chem. 2015, 87, 5849−5853
Quantitative Imaging of Single Unstained Magnetotactic Bacteria by Coherent X‑ray Diffraction Microscopy Jiadong Fan,† Zhibin Sun,† Jian Zhang,† Qingjie Huang,‡ Shengkun Yao,† Yunbing Zong,† Yoshiki Kohmura,§ Tetsuya Ishikawa,§ Hong Liu,† and Huaidong Jiang*,† †
State Key Laboratory of Crystal Materials, Shandong University, Jinan 250100, China School of Information Science and Engineering, Shandong University, Jinan 250100, China § RIKEN SPring-8 Center, 1-1-1, Kouto, Sayo, Hyogo 679-5148, Japan ‡
S Supporting Information *
ABSTRACT: Novel coherent diffraction microscopy provides a powerful lensless imaging method to obtain a better understanding of the microorganism at the nanoscale. Here we demonstrated quantitative imaging of intact unstained magnetotactic bacteria using coherent X-ray diffraction microscopy combined with an iterative phase retrieval algorithm. Although the signal-to-noise ratio of the X-ray diffraction pattern from single magnetotactic bacterium is weak due to low-scattering ability of biomaterials, an 18.6 nm halfperiod resolution of reconstructed image was achieved by using a hybrid input-output phase retrieval algorithm. On the basis of the quantitative reconstructed images, the morphology and some intracellular structures, such as nucleoid, polyβhydroxybutyrate granules, and magnetosomes, were identified, which were also confirmed by scanning electron microscopy and energy dispersive spectroscopy. With the benefit from the quantifiability of coherent diffraction imaging, for the first time to our knowledge, an average density of magnetotactic bacteria was calculated to be ∼1.19 g/cm3. This technique has a wide range of applications, especially in quantitative imaging of low-scattering biomaterials and multicomponent materials at nanoscale resolution. Combined with the cryogenic technique or X-ray free electron lasers, the method could image cells in a hydrated condition, which helps to maintain their natural structure.
A
conditions.15,16 Although imaging of hydrated cells is a new research highlight in the future, there are still lots of challenges and so far most bioimaging with CDI method is performed in dehydrated conditions. Here, taking advantages of good penetration and quantifiability of X-rays, we report quantitative, high-resolution, and high-contrast imaging of whole unstained low-scattering magnetotactic bacteria (MTB) with coherent Xray diffraction microscopy. MTB are aquatic prokaryotes and widely distributed in the aquatic environment.17 Since the bacteria with motility directed by the local geomagnetic field was observed in marine sediments,18 MTB have attracted a lot of research interests. The intracellular catenulate magnetosome (MS) particles inside the MTB, which are composed of magnetic iron-bearing inorganic nanocrystals, have potential applications in many fields due to the novel and special magnetic, physical, and perhaps optical properties.19 Some imaging techniques such as transmission electron microscopy (TEM), magnetic force
n accurate understanding of structure and function of biomaterials needs powerful microscopy that could provide not only images with high resolution but also quantitative images in situ. With the benefits of long penetration depth and short wavelength, X-ray microscopy is considered as an ideal tool to achieve high-resolution images without destruction in dehydrated and hydrated conditions. However, for conventional X-ray microscopy, resolution is limited by precise focusing devices for X-rays although significant progress has been made. Coherent X-ray diffraction imaging (CDI) is a promising lensless approach currently under rapid development to achieve high resolution. By recording coherent diffraction patterns in the far field, images of noncrystalline specimens can be retrieved using iterative phase retrieval algorithms from the patterns. The idea of CDI was first suggested by Sayre and demonstrated with X-rays by Miao et al.1,2 In the past decades, CDI has been widely applied to quantitatively determining microstructures at nanoscale resolution.3−12 For biological specimens, nondestructive imaging with high resolution and high contrast is crucial for revealing the relationship between structure and biological function in organisms. Efforts in imaging biomaterials under hydrated conditions were also undertaken with CDI, including frozen-hydrated13,14 and living © 2015 American Chemical Society
Received: February 24, 2015 Accepted: May 26, 2015 Published: May 26, 2015 5849
DOI: 10.1021/acs.analchem.5b00746 Anal. Chem. 2015, 87, 5849−5853
Letter
Analytical Chemistry
were illuminated by plane wave X-rays. In the far field, the diffracted signals from the AMB-1 cells were recorded by the CCD detector. Though the highly sensitive CCD was utilized, due to low scattering ability of biomaterials, the diffraction pattern of a single cell has a low signal-to-noise ratio, especially at high frequencies. To enhance the dynamic range of diffraction intensity and increase signal-to-noise ratio,26 lowresolution (LR) and high-resolution (HR) diffraction patterns were measured from the AMB-1 sample, separately. The total data acquisition time was about 3 h including 1 h for LR data and 2 h for HR data. By merging LR and HR patterns, the missing data due to the beamstop was confined in the centrospeckle of diffraction patterns, which facilitates reliability of image reconstruction.27 Figure 2a shows the diffraction pattern
microscopy (MFM), and scanning transmission X-ray microscopy (STXM) have been used to investigate the structure of MTB and MS.20−22 For instance, MFM with magnetic probe was used to study the extracted MSs and observed the MS trains structures.22 With the benefit from the high resolution of TEM, magnetosome vesicles were usually observed in the ultrathin section in MTB.21 To obtain a thorough understanding on the architecture of whole MTB, a quantitative analysis of MSs in situ is important and meaningful for determination of the arrangement and distribution of MS chains. However, for low-scattering unstained biomaterials, it is still a challenge to get quantitative, high-resolution, highcontrast images of whole cells. In this letter, we investigated the quantitative high-contrast imaging of single unstained Magnetospirillum magneticum AMB-1 with CDI at the nanoscale.23 Cells of Magnetospirillum magnetotacticum AMB-1 were cultured in an enriched magnetic spirillum growth medium (EMSGM) at 28 °C as described in previous study.24 The mutants of AMB-1 generated by a hyperactive mariner transposon was achieved and then picked.25 To fix the mutants of AMB-1, the transposon mutants were washed twice with sterile water and treated with 3.2% formaldehyde for 1 min, followed by adding 12% of glutaraldehyde to 0.25% of the final concentration. After fixation, the mutants of AMB-1 were stored in ethanol. To obtain single isolated AMB-1 cells, the sample solution was diluted with ethanol to an appropriate concentration and deposited on 30 nm thick Si3N4 membranes. The well-isolated AMB-1 cells were selected under an optical microscope and used for the coherent X-ray diffraction experiment. The CDI experiment was carried out on an undulator beamline at a third generation synchrotron radiation facility (SPring-8,BL 29XU). During the experiment, 5 keV monochromatic X-ray was defined by a 20 μm pinhole that was placed in front of the sample chamber. Two silicon apertures were mounted downstream of the pinhole to clean stray scattering from the pinhole and other upstream optical devices. In order to release the dark noise, a liquid nitrogen cooled CCD detector with 1340 × 1300 pixels was used to record diffraction patterns from samples. The distance between sample and CCD was set to 105 cm. A beamstop was mounted just in front of the CCD to block the direct X-ray beam. All these devices were in a vacuum condition of ∼10−6 Torr to eliminate absorption and scattering of the air. Figure 1 shows a schematic layout of a powerful X-ray diffraction microscope we used. The unstained AMB-1 cells
Figure 2. (a) Typical diffraction pattern of a single magnetotactic bacterium. Insert is the central part with missing data due to a beamstop. Parts b and c are two independent reconstructed images with different random initial phases from the diffraction pattern (a). Line scans (d) along the white dashed lines in parts b and c show a good consistency of reconstruction and inner density distribution.
after assembling LR and HR patterns. The zoomed in view of the central part (Figure 2a, insert) indicates that the LR and HR patterns were merged precisely without obvious trace, which implies the diffraction patterns are quite consistent at low frequencies. To reconstruct real-space images from the far-field diffraction patterns, Fienup’s hybrid input-output (HIO) algorithm combined with the oversampling method was used to conduct the phase retrieval.28 Without any prior information about the sample, a random phase was added to the measured diffraction pattern and a rectangular support was enforced during the reconstruction. To obtain accurate results, 16 replicated reconstructions were conducted for one run. Each reconstruction was iterated back and forth between real and reciprocal space with constraints. For example, in real space, the calculated positive electron density inside the support was reserved. While the electron density outside the support and negative electron density inside the support were gradually pushed to zero as described in eq 1,29 ⎧ r∈S ⎪ ρn ′(r ) ρn (r ) = ⎨ ⎪ ⎩ ρn − 1(r ) − βρn ′(r ) r ∉ S
Figure 1. Schematic of a coherent diffraction microscope.
and or
ρn ′(r ) ≥ 0 ρn ′(r ) < 0 (1)
5850
DOI: 10.1021/acs.analchem.5b00746 Anal. Chem. 2015, 87, 5849−5853
Letter
Analytical Chemistry where S represents the finite support, β a constant, and ρn the electron density at the nth iteration. During reconstruction, a Gaussian distribution kernel was used to smooth the noise out of support, which obviously reduced the noise effect. In order to evaluate the quality of the reconstructions, an error function Rerr was defined as eq 2, where |Fexp| and |Fcal| are amplitude of the measured and calculated diffraction patterns, respectively (2). R err =
materials, determination of density has significant biological and diagnostic applications, including intracellular organelles identification, segmentations, in situ location, etc. On the basis of eq 4,30 I(0, 0) = I0re 2 |F(0, 0)|2
(2)
To evaluate the reliability of image reconstructions, we carried out two independent HIO runs with different initial phases. The two independent reconstructed images (Figure 2b,c) show a good consistency both in cell profile and intracellular structures. The difference of the two images was calculated to be R1,2 = 3.13% according to eq 3, where ρ1(x,y) and ρ2(x,y) represent the electron densities of the independent reconstructed images. Line scans in Figure 2d along the white dashed lines (Figure 2b,c) also indicate the good consistency of the inner structure. R1,2 =
∑x , y |ρ1(x , y) − ρ2 (x , y)| ∑x , y |ρ1(x , y) + ρ2 (x , y)|
(4)
the electron density of AMB-1 can be calculated, where I(0,0) is the number of diffracted X-ray photons in the forward direction (i.e., within the central pixel of the CCD), I0 the incident X-ray flux per unit area, re the classical electron radius, | F(0,0)| the total number of electrons in the AMB-1, Δs the area of the central pixel, and r the distance from the sample to the CCD. Although I(0,0) was missed because of the beamstop, it can be recovered during the reconstruction.30 The total electrons of the single AMB-1 in Figure 3a were calculated to be 1.253 × 1011 electrons. According to the formula of MTB31 and the volume estimated from the morphology by AFM, we for the first time determined an average electron density of single AMB-1 cell to be ∼3.95 × 1011 electrons/μm3, corresponding to a mass density of ∼1.19 g/cm3 which is a little higher than Escherichia coli.31,32 The slight difference of the mass densities may be due to the different types of bacteria and irons in AMB-1. On the basis of the calculated high-contrast density image, the architecture of the AMB-1 could be quantitatively analyzed. As shown in Figure 3a, the cell wall (arrow A) with low density can be easily distinguished from surroundings due to the high image contrast. According to the density distribution in Figure 3a, the ribosome-rich cytoplasm has a relatively high mass density. However, there are not many kinds of organelles in the uniform cytoplasm because AMB-1 is a kind of Gram-negative bacteria. On the basis of the analysis of shape, size, mass density, and previous study on nucleoid of D. radiodurans,33 the region with high density (arrow B) could be the nucleoid. To further confirm it, energy dispersive spectroscopy (EDS) experiment was performed in this region. The result indicates that this region is rich in phosphorus which is the main element of DNA and RNA. The relatively low density region surrounded by high density cytoplasm (arrow D) is the polyβ-hydroxybutyrate (PHB) granule according to the morphology and density. However, because of the limitation of 2D image and resolution, some small organelles cannot be clearly identified. For MTB, MSs mineralized inside cells are composed of crystals of magnetite (Fe3O4) or greigite (Fe3S4).34,35 The AMB-1 cells used in our experiment usually have long Fe3O4 MS chains,23 and the isolated MS is about 35−120 nm enveloped with a membrane of lipid bilayer.36,37 As shown in Figure 3a, the chain-like regions with high density (arrows C and E) were selected to perform the EDS experiment. The result indicates that iron is a main element in these regions (Figure 4b, insert). Thus, considering the mass densities, element composition, and chain-like structure, the regions C and E were identified as two separated MS chains along the long axis of the cell, which were formed in the mutants of AMB-1 generated by a hyperactive mariner transposon.25 The isolated dark spots with high density in each chain are about 2− 4 pixels corresponding to 37−74 nm, which is in good agreement with the size of MS in previous reports.38 Moreover, dual-energy STXM imaging of another similar magnetotactic bacterium prepared in same conditions was performed at below and above Fe L3 absorption edge separately (Supporting
∑ ||Fcal| − |Fexp|| ∑ |Fexp|
Δs r2
(3)
In order to further verify the image reconstruction of the AMB-1 cell (Figure 3a), a scanning electron microscopy (SEM)
Figure 3. (a) Reconstruction image of a single bacterial cell from the diffraction pattern. The color represents electron density distribution of the cell. SEM image (b) shows the surface morphology of the same sample. (c) AFM image of a similar bacterial cell prepared at the same condition.
image of the same sample was taken (Figure 3b). To eliminate the charging effect, a 5 nm thick Au film was deposited on the sample. By comparison, the reconstructed image is in excellent agreement with the SEM image. Figure 3c shows an image of a similar AMB-1 on the same membrane using atomic force microscope (AFM). In comparison with Figure 3b,c, the benefit from the longer penetration length of X-rays and the result of CDI reconstruction in Figure 3a provides not only the sample profile but also buried intracellular structure information with high contrast. As X-rays interact with electrons of atoms, absolute electrons densities of reconstructed image could be calculated by CDI. Combining with composition elements information, an accurate mass density of specimen could be determined, which makes quantitative analysis possible. For biological 5851
DOI: 10.1021/acs.analchem.5b00746 Anal. Chem. 2015, 87, 5849−5853
Letter
Analytical Chemistry
resolution of 25 nm. Considering the overall signals and signalto-noise ratio, finally a 700 × 700 matrix of diffraction pattern with valid signals was selected for reconstruction. According to the equation of d = λ/2 sin(θ), where λ is the wavelength and θ half of the diffraction angle, the resolution was calculated to be 37.2 nm (i.e., pixel size of 18.6 nm). In order to confirm the reliability of the resolution and quality of the reconstructed image, we quantitatively analyzed the phase retrieval transfer function (PRTF) of the diffraction pattern in reciprocal space.39 Because phase errors cause image aberrations during reconstructions, the actual spatial resolution of the image could be worse than the calculated resolution if the reconstruction is incorrect. Therefore, we calculated the PRTF curve (Figure 5b) by comparing the Fourier amplitudes calculated from the reconstructed image to the measured ones. On the basis of the criterion of PRTF = 1/e,5,40 the PRTF curve indicates a reliable resolution of 37.2 nm. The area on the left of the red dash-dot line represents the missing data because of the beamstop. To further illustrate the resolution, line scans across the typical MSs (i.e., along black and blue dash lines shown in Figure 4a) were taken. The line scan results indicate that the typical MSs could be identified with good contrast. As shown in Figure 5c, the MSs edges were presented in one to two pixels, corresponding to an 18.6 nm half period resolution or a 37.2 nm full period resolution of the MTB image in physiological conditions. In addition, for biomaterials, radiation damage restricts the achievable resolution of diffraction patterns before degeneration emerging. An appropriate exposure less than the maximum tolerable dose, is a main issue for a successful experiment on low-scattering biomaterial imaging. The radiation dose deposited in the AMB-1 during our experiment was calculated to be ∼3.68 × 107 Gy. On the basis of a previous study,41 such radiation dose would not cause radiation damage of bacteria. In summary, we demonstrated quantitative imaging of single unstained low-scattering AMB-1 with high spatial resolution and high contrast using coherent X-ray diffraction microscopy. On the basis of the quantitative reconstructed image, the average mass density of AMB-1 was determined to be ∼1.19 g/ cm3. A spatial resolution of 37.2 nm (18.6 nm half-period resolution) was calculated and confirmed by both diffraction pattern in reciprocal space and reconstructed image in real space. Taking advantage of the high-contrast density distribution, we quantitatively analyzed the structure of AMB-1 and identified the location of nucleoid, PHB granules, and separated MS chains inside the cytoplasm. This work hence paves a way for unveiling the buried structures of a wide range of biological specimens at the nanoscale. Although imaging of hydrated biological samples currently faced some challenges,13,14 by more efficient combination of X-ray diffraction microscopy with the frozen-hydrated technique in the future, 5−10 nm resolution imaging of whole frozen-hydrated cells will be achievable considering the radiation damage.42 Besides, combining single-shot X-rays free electron lasers (XFEL), the CDI method has demonstrated imaging of small living cells with about 70 nm resolution.15,16 With the rapid development of the most promising X-ray source, high-resolution quantitative imaging of hydrated biological samples will be achieved.
Figure 4. Zoomed in view (a) of a rectangular region in Figure 3a shows some dark spots with high densities. (b) Line scan along the white dashed line and EDS spectrum at the position around the arrow in part a indicate a chain of magnetosomes.
Information). Some isolated magnetosomes and a magnetosome chain were observed, which is consistent with the CDI result. To further quantitatively analyze the MS chains, a line scan along the white dashed line in Figure 4a was performed. The line scan result (Figure 4b) shows that four MSs are contained in the chain due to the high contrast of mass density. By removing the sample thickness effect, the mass density of MS was estimated to be about 4.6−5.3 g/cm3, which is well consistent with a Fe3O4 mass density of 5.18 g/cm3. High spatial resolution imaging plays an increasingly important role in that it can provide powerful insights into the nanoworld. For CDI in biomaterials, although the spatial resolution of the reconstructed image is principally limited by the X-ray wavelength, at present, radiation damage restricts the achievable resolution. For the butterfly like diffraction pattern of the AMB-1 cell (Figure 2a), in comparison with the vertical direction, the signals in the horizontal direction are much stronger. The power spectral density (PSD) curve (Figure 5a) indicates the diffraction signals actually extend beyond a spatial
Figure 5. (a) Power spectral density of the diffraction pattern indicates a valid special frequency of 26.9 μm−1, which corresponds to 37.2 nm resolution. (b) PRTF curve of the diffraction pattern shows that the resolution of the reconstructed image is estimated to be 37.2 nm, based on the criterion of PRTF = 1/e. (c) Line scans along the black and blue dashed lines across the typical MSs shown in Figure 4a. A 37.2 nm full period resolution (two pixels) and an 18.6 half period resolution (one pixel) is achieved. 5852
DOI: 10.1021/acs.analchem.5b00746 Anal. Chem. 2015, 87, 5849−5853
Letter
Analytical Chemistry
■
(15) Kimura, T.; Joti, Y.; Shibuya, A.; Song, C.; Kim, S.; Tono, K.; Yabashi, M.; Tamakoshi, M.; Moriya, T.; Oshima, T. Nat. Commun. 2014, 5, 3052. (16) van der Schot, G.; Svenda, M.; Maia, F. R.; Hantke, M.; DePonte, D. P.; Seibert, M. M.; Aquila, A.; Schulz, J.; Kirian, R.; Liang, M. Nat. Commun. 2015, 6, 5704. (17) Faivre, D.; Schüler, D. Chem. Rev. 2008, 108, 4875−4898. (18) Blakemore, R. Science 1975, 190, 377−379. (19) Sun, J. B.; Duan, J. H.; Dai, S. L.; Ren, J.; Guo, L.; Jiang, W.; Li, Y. Biotechnol. Bioeng. 2008, 101, 1313−1320. (20) Kalirai, S. S.; Bazylinski, D. A.; Hitchcock, A. P. PLoS One 2013, 8, e53368. (21) Komeili, A.; Vali, H.; Beveridge, T. J.; Newman, D. K. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 3839−3844. (22) Suzuki, H.; Tanaka, T.; Sasaki, T.; Nakamura, N.; Matsunaga, T. Jpn. J. Appl. Phys. 1998, 37 (11A), L1343. (23) Matsunaga, T.; Sakaguchi, T.; Tadakoro, F. Appl. Microbiol. Biotechnol. 1991, 35, 651−655. (24) Yang, C.-D.; Takeyama, H.; Tanaka, T.; Matsunaga, T. Enzyme Microb. Technol. 2001, 29, 13−19. (25) Bo, T.; Wang, K.; Ge, X.; Chen, G.; Liu, W. Curr. Microbiol. 2012, 65, 98−107. (26) Xu, R.; Salha, S.; Raines, K. S.; Jiang, H.; Chen, C.-C.; Takahashi, Y.; Kohmura, Y.; Nishino, Y.; Song, C.; Ishikawa, T. Synchrotron Radiat. 2011, 18, 293−298. (27) Miao, J.; Nishino, Y.; Kohmura, Y.; Johnson, B.; Song, C.; Risbud, S. H.; Ishikawa, T. Phys. Rev. Lett. 2005, 95, 085503. (28) Fienup, J. R. Opt. Lett. 1978, 3, 27−29. (29) Miao, J.; Ishikawa, T.; Anderson, E. H.; Hodgson, K. O. Phys. Rev. B 2003, 67, 174104. (30) Miao, J.; Amonette, J. E.; Nishino, Y.; Ishikawa, T.; Hodgson, K. O. Phys. Rev. B 2003, 68, 012201. (31) Naresh, M.; Das, S.; Mishra, P.; Mittal, A. Biotechnol. Bioeng. 2012, 109, 1205−1216. (32) Godin, M.; Bryan, A. K.; Burg, T. P.; Babcock, K.; Manalis, S. R. Appl. Phys. Lett. 2007, 91, 123121. (33) Giewekemeyer, K.; Thibault, P.; Kalbfleisch, S.; Beerlink, A.; Kewish, C. M.; Dierolf, M.; Pfeiffer, F.; Salditt, T. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 529−534. (34) Frankel, R. B.; Blakemore, R. P.; Wolfe, R. S. Science 1979, 203, 1355. (35) Mann, S.; Sparks, N. C.; Frankel, R. B.; Bazylinski, D. A.; Jannasch, H. W. Nature 1990, 343, 258. (36) Bazylinski, D. A.; Garratt-Reed, A. J.; Frankel, R. B. Microsc. Res. Technique 1994, 27, 389−401. (37) Gorby, Y. A.; Beveridge, T. J.; Blakemore, R. P. J. Bacteriol. 1988, 170, 834−841. (38) Devouard, B.; Posfai, M.; Hua, X.; Bazylinski, D. A.; Frankel, R. B.; Buseck, P. R. Am. Mineral. 1998, 83, 1387−1398. (39) Chapman, H. N.; Barty, A.; Marchesini, S.; Noy, A.; Hau-Riege, S. P.; Cui, C.; Howells, M. R.; Rosen, R.; He, H.; Spence, J. C. J. Opt. Soc. Am.: A 2006, 23, 1179−1200. (40) Seibert, M. M.; Ekeberg, T.; Maia, F. R.; Svenda, M.; Andreasson, J.; Jö nsson, O.; Odić, D.; Iwan, B.; Rocker, A.; Westphal, D. Nature 2011, 470, 78−81. (41) Shen, Q.; Bazarov, I.; Thibault, P. J. Synchrotron Radiat. 2004, 11, 432−438. (42) Howells, M. R.; Beetz, T.; Chapman, H. N.; Cui, C.; Holton, J.; Jacobsen, C.; Kirz, J.; Lima, E.; Marchesini, S.; Miao, H. J. Electron Spectrosc. Relat. Phenom. 2009, 170, 4−12.
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b00746.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected]. Author Contributions
The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We thank Prof. Weifeng Liu for providing the AMB-1 samples used for this study, Prof. Jianwei Miao and Prof. Adam P. Hitchcock for stimulating discussion, and the staff at Shang-hai Synchrotron Radiation Facility for assistance with data acquisition. This work was supported by the Major State Basic Research Development Program of China (Grant 2014CB910401), the National Natural Science Foundation of China (Grants 31430031, 21390414, and U1332118), the Natural Science Foundation of Shandong Province (Grant JQ201117), and the Program for New Century Excellent Talents (Grant NCET-11-0304). Use of the RIKEN beamline (BL29XUL) at SPring-8 was supported by RIKEN.
■
REFERENCES
(1) Sayre, D.; Schlenker, M. Lect. Notes Phys. 1980, 112, 229. (2) Miao, J.; Charalambous, P.; Kirz, J.; Sayre, D. Nature 1999, 400, 342−344. (3) Jiang, H.; Ramunno-Johnson, D.; Song, C.; Amirbekian, B.; Kohmura, Y.; Nishino, Y.; Takahashi, Y.; Ishikawa, T.; Miao, J. Phys. Rev. Lett. 2008, 100, 038103. (4) Dierolf, M.; Menzel, A.; Thibault, P.; Schneider, P.; Kewish, C. M.; Wepf, R.; Bunk, O.; Pfeiffer, F. Nature 2010, 467, 436−439. (5) Shapiro, D.; Thibault, P.; Beetz, T.; Elser, V.; Howells, M.; Jacobsen, C.; Kirz, J.; Lima, E.; Miao, H.; Neiman, A. M. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 15343−15346. (6) Jiang, H.; Song, C.; Chen, C.-C.; Xu, R.; Raines, K. S.; Fahimian, B. P.; Lu, C.-H.; Lee, T.-K.; Nakashima, A.; Urano, J. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 11234−11239. (7) Nam, D.; Park, J.; Gallagher-Jones, M.; Kim, S.; Kim, S.; Kohmura, Y.; Naitow, H.; Kunishima, N.; Yoshida, T.; Ishikawa, T. Phys. Rev. Lett. 2013, 110, 098103. (8) Nishino, Y.; Takahashi, Y.; Imamoto, N.; Ishikawa, T.; Maeshima, K. Phys. Rev. Lett. 2009, 102, 018101. (9) Song, C.; Jiang, H.; Mancuso, A.; Amirbekian, B.; Peng, L.; Sun, R.; Shah, S. S.; Zhou, Z. H.; Ishikawa, T.; Miao, J. Phys. Rev. Lett. 2008, 101, 158101. (10) Miao, J.; Chen, C.-C.; Song, C.; Nishino, Y.; Kohmura, Y.; Ishikawa, T.; Ramunno-Johnson, D.; Lee, T.-K.; Risbud, S. H. Phys. Rev. Lett. 2006, 97, 215503. (11) Jiang, H.; Xu, R.; Chen, C.-C.; Yang, W.; Fan, J.; Tao, X.; Song, C.; Kohmura, Y.; Xiao, T.; Wang, Y. Phys. Rev. Lett. 2013, 110, 205501. (12) Pfeifer, M. A.; Williams, G. J.; Vartanyants, I. A.; Harder, R.; Robinson, I. K. Nature 2006, 442, 63−66. (13) Huang, X.; Nelson, J.; Kirz, J.; Lima, E.; Marchesini, S.; Miao, H.; Neiman, A. M.; Shapiro, D.; Steinbrener, J.; Stewart, A. Phys. Rev. Lett. 2009, 103, 198101. (14) Lima, E.; Wiegart, L.; Pernot, P.; Howells, M.; Timmins, J.; Zontone, F.; Madsen, A. Phys. Rev. Lett. 2009, 103, 198102. 5853
DOI: 10.1021/acs.analchem.5b00746 Anal. Chem. 2015, 87, 5849−5853
