- Email: [email protected]
Imaging sclera with hard X-ray microscopy
Micron 42 (2011) 506–511
Contents lists available at ScienceDirect
Micron journal homepage: www.elsevier.com/locate/micron
Imaging sclera with hard X-ray microscopy Nigel J. Fullwood a,∗ , Francis L. Martin b , Adam J. Bentley a , Jin Pyung Lee c , Sang Joon Lee d a
Division of Biomedicine and Life Sciences, School of Heath and Medicine, Lancaster University, Lancaster, UK Centre for Biophotonics, Lancaster Environment Centre, Lancaster University, Lancaster, UK School of Environmental Science and Engineering, Pohang University of Science and Technology, Republic of Korea d Department of Mechanical Engineering, Division of Integrative Biosciences and Biotechnology, Pohang University of Science and Technology, Republic of Korea b c
a r t i c l e
i n f o
Article history: Received 30 August 2010 Received in revised form 27 January 2011 Accepted 29 January 2011 Keywords: X-ray microscopy Synchrotron Sclera Collagen Proteoglycans
a b s t r a c t In this study, the organization of collagen fibrils within the sclera of the eye was investigated using the 7 keV hard X-ray microscope of the Pohang light source and compared to images from electron and atomic force microscopy. From the captured X-ray images, individual collagen fibrils were observed clearly in a spatial resolution much better than 100 nm, both in longitudinal sections and in transverse sections. Some of the collagen fibrils showed evidence of axial periodicity. In some regions of the samples, we could see cross-bridge like structures between adjacent collagen fibrils. The X-ray microscope also allowed the observation of keratocytes and the lamella structure of the scleral stroma. The X-ray microscope has some unique advantages in the nano-scale imaging of bio-samples relative to other established imaging techniques. © 2011 Elsevier Ltd. All rights reserved.
1. Introduction The X-ray microscope has great potential for the imaging of biological samples. Spatial resolution of less than 100 nm is possible, far better than a conventional light microscope (Chao et al., 2005; Rodenburg et al., 2007). However, this statement does not apply to the new generation of super-resolving light microscopes which offer a similar level of resolution for certain applications (Nugent and Bellair, 2003; Wessels et al., 2010). In addition some near-field optical and spectroscopic techniques have sub-100 nm resolution (Kim and Song, 2007; Martin et al., 2010). The advantages of X-ray microscopy over electron microscopy include the fact that samples can be imaged with less sample preparation and without being subjected to vacuum (Fullwood, 2008). In addition, due to enhanced phase-contrast, it is not necessary to stain the samples with heavy metals as in electron microscopy. It also avoids the steps of fixation, dehydration and specimen embedding that are proven to significantly disrupt the ultra-structures of biological samples (Fullwood and Meek, 1993). Although it should be mentioned that the use of cryo-microscopical techniques for the electron microscope can avoid these artifacts in some tissues (Massover, 2011). The spatial resolution of the hard X-ray microscope is slightly less than that of the soft X-ray microscope. However, it has the
∗ Corresponding author at: Division of Biomedicine and Life Sciences, School of Heath and Medicine, Lancaster University, Lancaster, UK. Tel.: +44 1524 593474. E-mail address: [email protected] (N.J. Fullwood). 0968-4328/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.micron.2011.01.012
advantage of being able to image thicker samples (Kagoshima et al., 2001). Previous work has shown that most biological specimens have little contrast with the absorption-contrast hard X-ray imaging method. However, by adopting the Zernike’s method (Zernike, 1935), the phase-shift cross-section of these samples was estimated to be around 1000 times greater than the absorption cross-section of hard X-ray imaging (Momose and Fukuda, 1995). Therefore, the phase-contrast hard X-ray microscope based on the Zernike’s method has been shown to give far better contrast on biological samples, compared to the absorption contrast method (Kagoshima et al., 2003; Yokosuka et al., 2003). Some workers have reported that low-power interference X-ray microscopy can also provide good results (Bonse and Beckmann, 2010). Youn et al. (2005) set up a hard X-ray microscope at the Pohang light source (PLS). Using a Zernike phase plate, Youn and Jung (2005) were able to image a human hair shaft at a spatial resolution less than 100 nm. This phase-contrast hard X-ray microscope has been used to image biological samples successfully. Recently, it was used to observe various intracellular organelles of human cells and mouse tumor tissues with a resolution less than 100 nm (Kim et al., 2008). In that study, the optimum preparation methods for biological samples were explained and the ideal imaging conditions for the X-ray microscope were discussed. In this paper, we have applied the phase-contrast hard X-ray microscope to sclera for the observation of associated structures within the scleral stroma. Sclera has been previously studied using conventional microscopical techniques such as electron microscopy (Fullwood and Meek, 1993). Sclera is a very interesting tissue to study. Sclera or the white of the eye is a typical example of an
N.J. Fullwood et al. / Micron 42 (2011) 506–511
507
Fig. 1. Schematic of the hard X-ray microscope at 1B2 microprobe beam line of Pohang light source.
opaque white photonic solid. The sclera is composed of collagen fibrils in a highly disordered state so that all wavelengths of visible light are scattered (Vaezy and Clark, 1991). As a result, the sclera is both opaque and brilliantly white. In contrast, the cornea which is biochemically almost identical to the sclera, has collagen fibrils which are arranged in a highly-ordered lattice, resulting in constructive scattering of light and almost perfect transparency (Hart and Farrell, 1969). At the limbus (edge of the cornea), the outer tunic of the eye changes abruptly from the opaque white sclera to the clear transparent cornea. The limbus is also the location for the corneal stem cells (Zhao et al., 2008; Nakamura et al., 2010). This transition from a disordered opaque white photonic solid to a highly-ordered transparent photonic solid is achieved simply by the elegant rearrangement of collagen fibrils within these tissues. Due to the importance of these tissues, sophisticated models of light scattering have been developed by the use of X-ray diffraction, biochemical and ultrastructural studies (Fullwood and Meek, 1994; Meek and Fullwood, 2001). The sclera has been less intensively studied than cornea. However, as it has a much wider size range of collagen fibrils than cornea it is more suitable to evaluate the performance of a new imaging technology such as X-ray microscopy. Therefore, the sclera is an ideal sample to evaluate the performance and usefulness of hard X-ray microscopy. We compared the results of this new technology with well-established imaging techniques including transmission electron, scanning electron and atomic force microscopy. 2. Materials and methods 2.1. Samples In this study, we used bovine cornea. The bovine eyes were obtained from a local abattoir within 2 h of death and immediately transported to the laboratory at 4 ◦ C. 2.2. Hard X-ray microscopy The sclera was embedded in Tissue-Tek (Agar Scientific, Stansted, UK) and instantly frozen with liquid nitrogen in the laboratory. Cryosections (10 m thick) were cut and mounted on silicon nitride windows (Norcada, Edmonton, Canada). A schematic diagram of the X-ray microscopy system is shown in Fig. 1. Full details of the optical set up are given in previous papers (Yokosuka et al., 2003; Youn et al., 2005). The X-ray microscopy system was installed at the bending magnet beamline of 1B2 at the PLS, which is a third-generation synchrotron radiation facility with an operating energy of 2.5 GeV. The X-rays pass through several optical components including a double crystal monochromator with Ge(1 1 1) double crystal monochromator (DCM), a condenser zone plate (a 4 mm diameter, a 100 nm outermost zone width), an annular aperture, a pinhole, an objective zone plate and a phase
ring. The photon energy was set at 6.95 keV to image the test specimens. The condenser zone plate demagnifies the X-ray source at the sample position by a factor of 8.85. The objective zone plate magnifies the X-ray image of a sample by 50 times. The numerical aperture of the condenser lens is 0.000885. Both of the condenser and micro zone plates are made of gold. Their outermost zones have a width of 100 nm and a nominal thickness of 1.6 m. The depth of focus of the micro zone plate is 200 m at 6.95 keV. For the Zernike phase-contrast imaging, an annular aperture is installed 3 cm downstream from the condenser zone plate. A gold phase plate of 2.1 m thickness is located at the back focal plane of the micro zone plate. The outer and the inner diameter of the phase ring are 114.6 and 85.6 m, respectively. Each nano-scale X-ray image of a test specimen is converted into a visible image on a thin X-ray scintillator crystal CsI (TI). The visible image that appeared on the scintillation crystal is magnified by a 20× objective lens. The images are captured by a digital CCD camera with an exposure time of about 60 s. The CCD camera (Redlake ES310/T) has a spatial resolution of 1300 × 1300 pixels and the field of view is about 21 m × 21 m. 2.3. Atomic force microscopy (AFM) Atomic force microscopy (Topometrix Explorer AFM) operating at contact mode was employed to image the partially hydrated samples of bovine sclera. Sections of approximately 100 m thick sclera were cut from fresh bovine eyes by hand, and attached to sample stubs. The sections were then squashed between two polished steel surfaces in a vice to produce a flat stable surface for AFM. All sections were cut parallel to the surface of the sclera. Topographs were obtained using a silicon nitride cantilever with an integrated pyramidal tip. A tube scanner of 2 m × 2 m was used for the high-resolution imaging. Some basic image processing routines were carried out such as leveling, histogram manipulation and shading. Multiple scanning of the same area and changing the scanning direction were used to ensure that similar features for the same specimen were observed each time to reduce the chance of artifacts. Samples were examined in air. 2.4. Low-temperature scanning electron microscopy (SEM) The specimens were mounted onto specimen stubs using a mixture of Tissue-Tek and colloidal carbon, and then rapidly frozen. Freezing was carried out by plunging the specimen into liquid nitrogen slush or by slam freezing onto a copper block cooled by liquid nitrogen. Once frozen, the tissue was placed in a Reichert-Jung FC 4E cryoultramicrotome at −140 ◦ C. The specimens could then be stored indefinitely under liquid nitrogen until they were transferred. The samples were transferred onto the cold stage of a JEOL 840 using a Hexland CT1000 cryo-trans system. For examination, they were uncoated at 2 kV. Ice was sublimated from the cell surface at −90 ◦ C until the cell surface was exposed. This step usually took 3–4 min. The sample was then moved back to the CT1000 cold stage
508
N.J. Fullwood et al. / Micron 42 (2011) 506–511
Fig. 2. TEM image of sclera showing the arrangement of collagen fibrils of varying diameter arranged in lamellae. The collagen fibrils of the lamellae to the left are in cross-section showing the variation in diameters of scleral fibrils. The fibrils in the lamellae on the right are in longitudinal-section clearly showing the D-period banding. (Scale bar = 200 nm).
where the sample was sputter-coated with gold. The sample was then returned to the cold stage inside the SEM and examined at 15 kV at −160 ◦ C. 2.5. Transmission electron microscopy (TEM) In order to visualize proteoglycans, the samples were fixed overnight in 2.5% glutaraldehyde in 25 mM sodium acetate buffer (pH 5.7), with 0.1 M magnesium chloride and 0.05% cupromeronic blue. After this fixation, the specimens were washed in buffer solution, followed by three 15-min washes of the fixed samples, first in aqueous and then in 50% ethanolic 0.5% sodium tungstate solutions (Fullwood et al., 1993). Specimens were dehydrated through an alcohol series, transferred to propylene oxide and embedded in Araldite resin. Ultrathin sections were collected onto copper grids and examined in a 10-10 JEOL TEM. The sections were not counterstained to visualize the cupromeronic stain under the TEM. Conventional TEM images were also acquired.
Fig. 3. TEM image of the sclera with the proteoglycans stained with cupromeronic blue. The dense filaments appear to form cross-bridge structures between adjacent filaments (arrows). (Scale bar = 200 nm).
A great deal of work has been carried out on the arrangement of proteoglycans in the cornea, using the stain cupromeronic blue (Fullwood et al., 1993) to visualize them with TEM and the crossbridge structures are easily observed (Fig. 3). Our results using cupromeronic blue and TEM show these structures clearly in Fig. 3, as electron-dense filaments running between collagen fibrils. Both the TEM and AFM images show the 65 nm D-periodicity along the collagen fibrils very clearly (Figs. 2–4). The low-temperature SEM images resolve the scleral collagen fibrils clearly (Fig. 5), but the D-periodicity is not noticeably visible. Because the “gap” region of the collagen fibril does not shrink to a smaller diameter than the rest of the fibril in a hydrated state, it cannot be so easily resolved. These figures show instances of cross-bridge structure between fibrils (Fig. 6), but it appears very large, compared to the TEM images. However, with this technique, the glycosaminoglycans should be in a frozen but hydrated state. Although ice crystal formation is always a problem with cryo-microscopy, previous work with this low-temperature SEM
3. Results and discussion This study was carried out on a specimen of sclera. Sclera is the tough and white outer tunic of the eye and 90% of its thickness is composed of the scleral stroma. The scleral stroma is composed of type I collagen fibrils of various diameters arranged in a series of layers or lamellae each of which is a few microns thick (Fig. 2). The sclera has been previously studied using conventional electron microscopical and immunolabelling techniques as reviewed by Ihanamäki et al. (2004). Sclera contains the proteoglycans; aggrecan, lumican, biglycan and decorin. Biglycan and decorin are small dermatan sulphate proteoglycans and lumican a small keratan sulphate proteoglycan. These small proteoglycans are believed to form cross-bridge structures between collagen fibrils in the cornea. It seems likely that they have a similar functional role in sclera. Aggrecan is a much larger keratan sulphate proteoglycan and its role in sclera is much less well understood. Although it has been known to interact both with collagen fibrils and also to interact with lumican (Dunlevy and Rada, 2004), it may also be involved in the formation of collagen associated structures (Hedlund et al., 1999). The glycosaminoglycan component of the proteoglycans can be visualized in the TEM using the “critical electrolyte” staining technique (Fullwood et al., 1993).
Fig. 4. High resolution AFM image of collagen fibrils within the sclera, the Dperiodicity is clearly visible. (Scale bar = 400 nm).
N.J. Fullwood et al. / Micron 42 (2011) 506–511
509
Fig. 5. Low-temperature SEM image of the sclera, the collagen fibrils are easily resolved and are associated with globular structures, some of which appear to form cross-bridges between adjacent fibrils. (Scale bar = 600 nm).
method has shown good results on extracellular matrix tissues such as cartilage (Gardner et al., 1997). Aggrecan is known to have a highly extended “spaced filling” role in cartilage and sclera. The size of the hydrated proteoglycans in their native state is known to be much larger than observed in the collapsed condensed state when using TEM or conventional SEM imaging. It seems most likely that the large cross-bridge structures observed with the low-temperature SEM in this study must be formed from the large proteoglycan aggrecan which has been shown to associate with collagen fibrils and with lumican (Dunlevy and Rada, 2004; Hedlund et al., 1999). Previous work with AFM on partially hydrated samples of bovine sclera has also shown evidence of large cross-bridge structures (50–80 nm) in sclera (Fullwood et al., 1995). These structures appeared less condensed than the images obtained with TEM after staining with cupromeronic blue, this was attributed to the fact that the structures were partially hydrated when imaged with AFM in air. Comparing the results obtained by the X-ray microscopy with those captured by conventional imaging techniques, it is evident that individual type I collagen fibrils can be resolved easily down to a diameter of less than 100 nm. The fibrils can be seen both in longitudinal sections (Fig. 7) and in cross-section (Fig. 8). Interestingly, some fibrils exhibit evidence of axial periodicity (Fig. 9). This suggests that the microscopical technique might be capable of resolving structures which are less than 70 nm in size. Although periodicity is certainly not as clear as that seen by the AFM or
Fig. 6. Low-temperature SEM image of the sclera, the collagen fibrils do not show any clear evidence of periodicity along their length. Cross-bridges (arrows) can clearly be seen between adjacent fibrils (×2 insert) (Scale bar = 400 nm).
Fig. 7. X-ray microscope image of fibrils within the sclera, the collagen fibrils running parallel to each other in longitudinal orientation can be resolved (arrows). (Scale bar = 300 nm).
Fig. 8. X-ray microscope image of collagen fibrils in cross-section (arrows). The fibrils are of various diameters. (Scale bar = 300 nm).
Fig. 9. X-ray microscope image of adjacent collagen fibrils with some indication of axial periodicity, or at least variations in density along their length. Arrows indicate bands of higher or lower density along their length. (Scale bar = 100 nm).
510
N.J. Fullwood et al. / Micron 42 (2011) 506–511
Fig. 11. X-ray microscope image of a scleral fibroblast (F), the characteristic shape of the cell is evident, but little detail of the interior of the cell is visible. (Scale bar = 2 m). Fig. 10. X-ray image of longitudinal scleral fibrils, in some regions cross-bridge like structures can be resolved between adjacent fibrils. These can be seen clearly in the ×2 insert (arrows) (Scale bar = 200 nm).
with a stained TEM, TEM requires the use of heavy metal stains for providing contrast enough to image the D-periodicity. In addition, conventional TEM processing results in considerable shrinkage of collagen as well as changes to fibril spacing (Fullwood and Meek, 1993). Although AFM avoids many processing steps necessary for TEM imaging, it does involve physical interaction with the collagen fibrils via the AFM tip which can cause artifacts (Fullwood et al., 1995). Thus at a lower spatial resolution, the images of collagen fibrils captured by the X-ray microscope might be more authentic than those with other microscopical methods. This is supported by the fact that the fibrils look similar to the low-temperature SEM images of the same sample which are in a frozen but hydrated state. Very significantly the X-ray images show what appear to be cross-bridge like structures in some regions (Fig. 10). These are towards the limit of resolution of the X-ray microscope. It is interesting that cross-bridge structures are resolved with the X-ray microscope, because the size of these cross-bridges shown on the TEM is too small to be resolved by X-ray microscopy. However, further reflection and deliberation give a possible explanation on why these show up on the X-ray microscope. In the hydrated state (Fig. 6), the cross-bridge structures are very large and in the range of 100 nm. Furthermore, the glycosaminoglycan chains of the proteoglycans have a significantly higher atomic density than the surrounding molecules. Glycosaminoglycans are heavily sulphated and in addition their negative charge attracts counter ions such as calcium. The high atomic density, relative to the lighter elements present in the collagen fibrils, causes these cross-bridges to exhibit contrast under the X-ray microscope. It seems likely that the structures seen in Fig. 10 are composed of the proteoglycan aggrecan in a partially hydrated state. Other structures which can be identified from the X-ray images include what appear to be whole keratocyte cells (Fig. 11). The shape of the cells is clearly outlined. However, little internal detail is visible within the keratocyte cells. What are the actual advantages of the hard X-ray microscope? Many papers report that the main one is the absence of additional processing routines as is associated with TEM. This study suggests that in addition the hard X-ray microscope is capable of providing unique information. In this study the hard X-ray microscope seemed able to resolve the cross-bridge structures between fibrils, it is significant that these images appear different to conventional TEM images, but similar to low-temperature SEM images. This result suggests that the X-ray images are closer to the in vivo structure of the tissue than conventional TEM images.
Soft X-ray microscopy also can be used to examine frozen hydrated samples, but the sample preparation and interpretation of these samples is not trivial. Hard X-ray microscopes have the ability to image thicker samples, with increased depth in focus, in a partially hydrated state. We believe that hard X-ray microscopes will in the future make significant contributions to the understanding of the ultrastructure of cells and tissues. Acknowledgments This work was supported by the Creative Research Initiatives (Diagnosis of Biofluid Flow Phenomena and Biomimic Research) of MEST/NRF of Korea. X-ray imaging experiments were performed at 1B2 beamline of Pohang Accelerator Laboratory (PAL), Pohang, Korea. References Bonse, U., Beckmann, F., 2010. X-Ray Imaging with Phase Contrast. AIP Conference Proceedings, 1236 , pp. 189–194. Chao, W., Harteneck, B.D., Liddle, J.A., Anderson, E.H., Attwood, D.T., 2005. Soft X-ray nanoscopy at a spatial resolution better than 15 nm. Nature 435, 1210–1213. Dunlevy, J.R., Rada, J.A., 2004. Interaction of lumican with aggrecan in the aging human sclera. Invest. Ophthalmol. Vis. Sci 45, 3849–3856. Fullwood, N.J., Hammiche, A., Pollock, H.M., Hourston, D.J., Song, M., 1995. Atomic force microscopy of the cornea and sclera. Curr. Eye Res. 14, 529–535. Fullwood, N.J., Meek, K.M., 1993. A synchroton X-ray study of the changes occurring in the corneal stroma during processing for electron microscopy. J. Microsc. 169, 53–60. Fullwood, N.J., Meek, K.M., 1994. An ultrastructural, time-resolved study of freezing in the corneal stroma. J. Mol. Biol. 236, 749–758. Fullwood, N.J., Troilo, D., Wallman, J., Meek, K.M., 1993. Synchrotron X-ray diffraction and histochemical studies of normal and myopic chick eyes. Tissue Cell 25, 73–85. Fullwood, N.J., 2008. Microscopy for life scientists. Science 32, 1445. Gardner, D.L., Salter, D.M., Oates, K., 1997. Advances in the microscopy of osteoarthritis. Microsc. Res. Tech. 37, 245–270. Hart, R.W., Farrell, R.A., 1969. Light scattering in the cornea. J. Opt. Soc. Am. 59, 766–774. Hedlund, H., Hedbom, E., Heinegard, D., Mengarelli-Widholm, S., Reinholt, F.P., Svensson, O., 1999. Association of the aggrecan keratan sulfate-rich region with collagen in bovine articular cartilage. J. Biol. Chem. 274, 5777–5781. Ihanamäki, T., Pelliniemi, L.J., Vuorio, E., 2004. Collagens and collagen-related matrix components in the human and mouse eye. Prog. Retin. Eye Res. 23, 403–434. Kagoshima, Y., Ibuki, T., Yokoyama, Y., Tsusaka, Y., Matsui, J., Takai, K., Aino, M., 2001. 10 keV X-ray phase-contrast microscopy for observing transparent specimens. Jpn. J. Appl. Phys. 40, 1190–1192. Kagoshima, Y., Yokoyama, Y., Niimi, T., Koyama, T., Tsusaka, Y., Matsui, J., Takai, K., 2003. Hard X-ray phase contrast microscope for observing transparent specimens. J. Phys. IV 104, 49–52. Kim, G.B., Yoon, Y.J., Shin, T.J., Youn, H.S., Gho, Y.S., Lee, S.J., 2008. X-ray imaging of various biological samples using a phase-contrast hard X-ray microscope. Microsc. Res. Tech. 71, 639–643. Kim, J., Song, K., 2007. Recent progress of nano-technology with NSOM. Micron 38, 409–426.
N.J. Fullwood et al. / Micron 42 (2011) 506–511 Martin, F.L., Kelly, J.G., Llabjani, V., Martin-Hirsch, P.L., Patel, I.I., Trevisan, J., Fullwood, N.J., Walsh, M.J., 2010. Distinguishing cell types or populations based on the computational analysis of their infrared spectra. Nat. Protoc. 5, 1748–1760. Massover, W.H., 2011. New and unconventional approaches for advancing resolution in biological transmission electron microscopy by improving macromolecular specimen preparation and preservation. Micron 42, 141–151. Meek, K.M., Fullwood, N.J., 2001. Corneal and scleral collagens—a microscopist’s perspective. Micron 32, 261–272. Momose, A., Fukuda, J., 1995. Phase-contrast radiographs of nonstained rat cerebellar specimen. Med. Phys. 22, 375–379. Nakamura, T., Kelly, J.G., Trevisan, J., Cooper, L.J., Bentley, A.J., Carmichael, P.L., Scott, A.D., Cotte, M., Susini, J., Martin-Hirsch, P.L., Kinoshita, S., Fullwood, N.J., Martin, F.L., 2010. Microspectroscopy of spectral biomarkers associated with human corneal stem cells. Mol Vis. 16, 359–368. Nugent, K.A., Bellair, C.J., 2003. Emulated super-resolution using quantitative phase microscopy. Micron 34, 333–338. Rodenburg, J.M., Hurst, A.C., Cullis, A.G., Dobson, B.R., Pfeiffer, F., Bunk, O., David, C., Jefimovs, K., Johnson, I., 2007. Hard-X-ray lensless imaging of extended objects. Phys. Rev. Lett. 98, 34801–34804.
511
Vaezy, S., Clark, J.I., 1991. A quantitative analysis of transparency in the human sclera and cornea using Fourier methods. J. Microsc. 163, 85–94. Wessels, J.T., Yamauchi, K., Hoffman, R.M., Wouters, F.S., 2010. Advances in cellular, subcellular, and nanoscale imaging in vitro and in vivo. Cytometry A 77, 667–676. Yokosuka, H., Watanabe, N., Ohigashi, T., Aoki, S., Ando, M., 2003. Phasecontrast hard X-ray microscope with a zone plate at the photon factory. J. Phys. IV 104, 591–594. Youn, H.S., Baik, S.Y., Chang, C.H., 2005. Hard X-ray microscopy with a 130 nm spatial resolution. Rev. Sci. Instrum. 76, 023702–23704. Youn, H.S., Jung, S.W., 2005. Observations of a human hair shaft with an X-ray microscope. Phys. Med. Biol. 21, 5417–5420. Zernike, F., 1935. Das phasenkontrastverfahren bei der mikroskopischen beobachtung. Z. Tech. Phys. 16, 454–457. Zhao, B., Allinson, S.L., Ma, A., Bentley, A.J., Martin, F.L., Fullwood, N.J., 2008. Targeted cornea limbal stem/progenitor cell transfection in an organ culture model. Invest. Ophthalmol. Vis. Sci. 49, 3395–3401.
Contents lists available at ScienceDirect
Micron journal homepage: www.elsevier.com/locate/micron
Imaging sclera with hard X-ray microscopy Nigel J. Fullwood a,∗ , Francis L. Martin b , Adam J. Bentley a , Jin Pyung Lee c , Sang Joon Lee d a
Division of Biomedicine and Life Sciences, School of Heath and Medicine, Lancaster University, Lancaster, UK Centre for Biophotonics, Lancaster Environment Centre, Lancaster University, Lancaster, UK School of Environmental Science and Engineering, Pohang University of Science and Technology, Republic of Korea d Department of Mechanical Engineering, Division of Integrative Biosciences and Biotechnology, Pohang University of Science and Technology, Republic of Korea b c
a r t i c l e
i n f o
Article history: Received 30 August 2010 Received in revised form 27 January 2011 Accepted 29 January 2011 Keywords: X-ray microscopy Synchrotron Sclera Collagen Proteoglycans
a b s t r a c t In this study, the organization of collagen fibrils within the sclera of the eye was investigated using the 7 keV hard X-ray microscope of the Pohang light source and compared to images from electron and atomic force microscopy. From the captured X-ray images, individual collagen fibrils were observed clearly in a spatial resolution much better than 100 nm, both in longitudinal sections and in transverse sections. Some of the collagen fibrils showed evidence of axial periodicity. In some regions of the samples, we could see cross-bridge like structures between adjacent collagen fibrils. The X-ray microscope also allowed the observation of keratocytes and the lamella structure of the scleral stroma. The X-ray microscope has some unique advantages in the nano-scale imaging of bio-samples relative to other established imaging techniques. © 2011 Elsevier Ltd. All rights reserved.
1. Introduction The X-ray microscope has great potential for the imaging of biological samples. Spatial resolution of less than 100 nm is possible, far better than a conventional light microscope (Chao et al., 2005; Rodenburg et al., 2007). However, this statement does not apply to the new generation of super-resolving light microscopes which offer a similar level of resolution for certain applications (Nugent and Bellair, 2003; Wessels et al., 2010). In addition some near-field optical and spectroscopic techniques have sub-100 nm resolution (Kim and Song, 2007; Martin et al., 2010). The advantages of X-ray microscopy over electron microscopy include the fact that samples can be imaged with less sample preparation and without being subjected to vacuum (Fullwood, 2008). In addition, due to enhanced phase-contrast, it is not necessary to stain the samples with heavy metals as in electron microscopy. It also avoids the steps of fixation, dehydration and specimen embedding that are proven to significantly disrupt the ultra-structures of biological samples (Fullwood and Meek, 1993). Although it should be mentioned that the use of cryo-microscopical techniques for the electron microscope can avoid these artifacts in some tissues (Massover, 2011). The spatial resolution of the hard X-ray microscope is slightly less than that of the soft X-ray microscope. However, it has the
∗ Corresponding author at: Division of Biomedicine and Life Sciences, School of Heath and Medicine, Lancaster University, Lancaster, UK. Tel.: +44 1524 593474. E-mail address: [email protected] (N.J. Fullwood). 0968-4328/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.micron.2011.01.012
advantage of being able to image thicker samples (Kagoshima et al., 2001). Previous work has shown that most biological specimens have little contrast with the absorption-contrast hard X-ray imaging method. However, by adopting the Zernike’s method (Zernike, 1935), the phase-shift cross-section of these samples was estimated to be around 1000 times greater than the absorption cross-section of hard X-ray imaging (Momose and Fukuda, 1995). Therefore, the phase-contrast hard X-ray microscope based on the Zernike’s method has been shown to give far better contrast on biological samples, compared to the absorption contrast method (Kagoshima et al., 2003; Yokosuka et al., 2003). Some workers have reported that low-power interference X-ray microscopy can also provide good results (Bonse and Beckmann, 2010). Youn et al. (2005) set up a hard X-ray microscope at the Pohang light source (PLS). Using a Zernike phase plate, Youn and Jung (2005) were able to image a human hair shaft at a spatial resolution less than 100 nm. This phase-contrast hard X-ray microscope has been used to image biological samples successfully. Recently, it was used to observe various intracellular organelles of human cells and mouse tumor tissues with a resolution less than 100 nm (Kim et al., 2008). In that study, the optimum preparation methods for biological samples were explained and the ideal imaging conditions for the X-ray microscope were discussed. In this paper, we have applied the phase-contrast hard X-ray microscope to sclera for the observation of associated structures within the scleral stroma. Sclera has been previously studied using conventional microscopical techniques such as electron microscopy (Fullwood and Meek, 1993). Sclera is a very interesting tissue to study. Sclera or the white of the eye is a typical example of an
N.J. Fullwood et al. / Micron 42 (2011) 506–511
507
Fig. 1. Schematic of the hard X-ray microscope at 1B2 microprobe beam line of Pohang light source.
opaque white photonic solid. The sclera is composed of collagen fibrils in a highly disordered state so that all wavelengths of visible light are scattered (Vaezy and Clark, 1991). As a result, the sclera is both opaque and brilliantly white. In contrast, the cornea which is biochemically almost identical to the sclera, has collagen fibrils which are arranged in a highly-ordered lattice, resulting in constructive scattering of light and almost perfect transparency (Hart and Farrell, 1969). At the limbus (edge of the cornea), the outer tunic of the eye changes abruptly from the opaque white sclera to the clear transparent cornea. The limbus is also the location for the corneal stem cells (Zhao et al., 2008; Nakamura et al., 2010). This transition from a disordered opaque white photonic solid to a highly-ordered transparent photonic solid is achieved simply by the elegant rearrangement of collagen fibrils within these tissues. Due to the importance of these tissues, sophisticated models of light scattering have been developed by the use of X-ray diffraction, biochemical and ultrastructural studies (Fullwood and Meek, 1994; Meek and Fullwood, 2001). The sclera has been less intensively studied than cornea. However, as it has a much wider size range of collagen fibrils than cornea it is more suitable to evaluate the performance of a new imaging technology such as X-ray microscopy. Therefore, the sclera is an ideal sample to evaluate the performance and usefulness of hard X-ray microscopy. We compared the results of this new technology with well-established imaging techniques including transmission electron, scanning electron and atomic force microscopy. 2. Materials and methods 2.1. Samples In this study, we used bovine cornea. The bovine eyes were obtained from a local abattoir within 2 h of death and immediately transported to the laboratory at 4 ◦ C. 2.2. Hard X-ray microscopy The sclera was embedded in Tissue-Tek (Agar Scientific, Stansted, UK) and instantly frozen with liquid nitrogen in the laboratory. Cryosections (10 m thick) were cut and mounted on silicon nitride windows (Norcada, Edmonton, Canada). A schematic diagram of the X-ray microscopy system is shown in Fig. 1. Full details of the optical set up are given in previous papers (Yokosuka et al., 2003; Youn et al., 2005). The X-ray microscopy system was installed at the bending magnet beamline of 1B2 at the PLS, which is a third-generation synchrotron radiation facility with an operating energy of 2.5 GeV. The X-rays pass through several optical components including a double crystal monochromator with Ge(1 1 1) double crystal monochromator (DCM), a condenser zone plate (a 4 mm diameter, a 100 nm outermost zone width), an annular aperture, a pinhole, an objective zone plate and a phase
ring. The photon energy was set at 6.95 keV to image the test specimens. The condenser zone plate demagnifies the X-ray source at the sample position by a factor of 8.85. The objective zone plate magnifies the X-ray image of a sample by 50 times. The numerical aperture of the condenser lens is 0.000885. Both of the condenser and micro zone plates are made of gold. Their outermost zones have a width of 100 nm and a nominal thickness of 1.6 m. The depth of focus of the micro zone plate is 200 m at 6.95 keV. For the Zernike phase-contrast imaging, an annular aperture is installed 3 cm downstream from the condenser zone plate. A gold phase plate of 2.1 m thickness is located at the back focal plane of the micro zone plate. The outer and the inner diameter of the phase ring are 114.6 and 85.6 m, respectively. Each nano-scale X-ray image of a test specimen is converted into a visible image on a thin X-ray scintillator crystal CsI (TI). The visible image that appeared on the scintillation crystal is magnified by a 20× objective lens. The images are captured by a digital CCD camera with an exposure time of about 60 s. The CCD camera (Redlake ES310/T) has a spatial resolution of 1300 × 1300 pixels and the field of view is about 21 m × 21 m. 2.3. Atomic force microscopy (AFM) Atomic force microscopy (Topometrix Explorer AFM) operating at contact mode was employed to image the partially hydrated samples of bovine sclera. Sections of approximately 100 m thick sclera were cut from fresh bovine eyes by hand, and attached to sample stubs. The sections were then squashed between two polished steel surfaces in a vice to produce a flat stable surface for AFM. All sections were cut parallel to the surface of the sclera. Topographs were obtained using a silicon nitride cantilever with an integrated pyramidal tip. A tube scanner of 2 m × 2 m was used for the high-resolution imaging. Some basic image processing routines were carried out such as leveling, histogram manipulation and shading. Multiple scanning of the same area and changing the scanning direction were used to ensure that similar features for the same specimen were observed each time to reduce the chance of artifacts. Samples were examined in air. 2.4. Low-temperature scanning electron microscopy (SEM) The specimens were mounted onto specimen stubs using a mixture of Tissue-Tek and colloidal carbon, and then rapidly frozen. Freezing was carried out by plunging the specimen into liquid nitrogen slush or by slam freezing onto a copper block cooled by liquid nitrogen. Once frozen, the tissue was placed in a Reichert-Jung FC 4E cryoultramicrotome at −140 ◦ C. The specimens could then be stored indefinitely under liquid nitrogen until they were transferred. The samples were transferred onto the cold stage of a JEOL 840 using a Hexland CT1000 cryo-trans system. For examination, they were uncoated at 2 kV. Ice was sublimated from the cell surface at −90 ◦ C until the cell surface was exposed. This step usually took 3–4 min. The sample was then moved back to the CT1000 cold stage
508
N.J. Fullwood et al. / Micron 42 (2011) 506–511
Fig. 2. TEM image of sclera showing the arrangement of collagen fibrils of varying diameter arranged in lamellae. The collagen fibrils of the lamellae to the left are in cross-section showing the variation in diameters of scleral fibrils. The fibrils in the lamellae on the right are in longitudinal-section clearly showing the D-period banding. (Scale bar = 200 nm).
where the sample was sputter-coated with gold. The sample was then returned to the cold stage inside the SEM and examined at 15 kV at −160 ◦ C. 2.5. Transmission electron microscopy (TEM) In order to visualize proteoglycans, the samples were fixed overnight in 2.5% glutaraldehyde in 25 mM sodium acetate buffer (pH 5.7), with 0.1 M magnesium chloride and 0.05% cupromeronic blue. After this fixation, the specimens were washed in buffer solution, followed by three 15-min washes of the fixed samples, first in aqueous and then in 50% ethanolic 0.5% sodium tungstate solutions (Fullwood et al., 1993). Specimens were dehydrated through an alcohol series, transferred to propylene oxide and embedded in Araldite resin. Ultrathin sections were collected onto copper grids and examined in a 10-10 JEOL TEM. The sections were not counterstained to visualize the cupromeronic stain under the TEM. Conventional TEM images were also acquired.
Fig. 3. TEM image of the sclera with the proteoglycans stained with cupromeronic blue. The dense filaments appear to form cross-bridge structures between adjacent filaments (arrows). (Scale bar = 200 nm).
A great deal of work has been carried out on the arrangement of proteoglycans in the cornea, using the stain cupromeronic blue (Fullwood et al., 1993) to visualize them with TEM and the crossbridge structures are easily observed (Fig. 3). Our results using cupromeronic blue and TEM show these structures clearly in Fig. 3, as electron-dense filaments running between collagen fibrils. Both the TEM and AFM images show the 65 nm D-periodicity along the collagen fibrils very clearly (Figs. 2–4). The low-temperature SEM images resolve the scleral collagen fibrils clearly (Fig. 5), but the D-periodicity is not noticeably visible. Because the “gap” region of the collagen fibril does not shrink to a smaller diameter than the rest of the fibril in a hydrated state, it cannot be so easily resolved. These figures show instances of cross-bridge structure between fibrils (Fig. 6), but it appears very large, compared to the TEM images. However, with this technique, the glycosaminoglycans should be in a frozen but hydrated state. Although ice crystal formation is always a problem with cryo-microscopy, previous work with this low-temperature SEM
3. Results and discussion This study was carried out on a specimen of sclera. Sclera is the tough and white outer tunic of the eye and 90% of its thickness is composed of the scleral stroma. The scleral stroma is composed of type I collagen fibrils of various diameters arranged in a series of layers or lamellae each of which is a few microns thick (Fig. 2). The sclera has been previously studied using conventional electron microscopical and immunolabelling techniques as reviewed by Ihanamäki et al. (2004). Sclera contains the proteoglycans; aggrecan, lumican, biglycan and decorin. Biglycan and decorin are small dermatan sulphate proteoglycans and lumican a small keratan sulphate proteoglycan. These small proteoglycans are believed to form cross-bridge structures between collagen fibrils in the cornea. It seems likely that they have a similar functional role in sclera. Aggrecan is a much larger keratan sulphate proteoglycan and its role in sclera is much less well understood. Although it has been known to interact both with collagen fibrils and also to interact with lumican (Dunlevy and Rada, 2004), it may also be involved in the formation of collagen associated structures (Hedlund et al., 1999). The glycosaminoglycan component of the proteoglycans can be visualized in the TEM using the “critical electrolyte” staining technique (Fullwood et al., 1993).
Fig. 4. High resolution AFM image of collagen fibrils within the sclera, the Dperiodicity is clearly visible. (Scale bar = 400 nm).
N.J. Fullwood et al. / Micron 42 (2011) 506–511
509
Fig. 5. Low-temperature SEM image of the sclera, the collagen fibrils are easily resolved and are associated with globular structures, some of which appear to form cross-bridges between adjacent fibrils. (Scale bar = 600 nm).
method has shown good results on extracellular matrix tissues such as cartilage (Gardner et al., 1997). Aggrecan is known to have a highly extended “spaced filling” role in cartilage and sclera. The size of the hydrated proteoglycans in their native state is known to be much larger than observed in the collapsed condensed state when using TEM or conventional SEM imaging. It seems most likely that the large cross-bridge structures observed with the low-temperature SEM in this study must be formed from the large proteoglycan aggrecan which has been shown to associate with collagen fibrils and with lumican (Dunlevy and Rada, 2004; Hedlund et al., 1999). Previous work with AFM on partially hydrated samples of bovine sclera has also shown evidence of large cross-bridge structures (50–80 nm) in sclera (Fullwood et al., 1995). These structures appeared less condensed than the images obtained with TEM after staining with cupromeronic blue, this was attributed to the fact that the structures were partially hydrated when imaged with AFM in air. Comparing the results obtained by the X-ray microscopy with those captured by conventional imaging techniques, it is evident that individual type I collagen fibrils can be resolved easily down to a diameter of less than 100 nm. The fibrils can be seen both in longitudinal sections (Fig. 7) and in cross-section (Fig. 8). Interestingly, some fibrils exhibit evidence of axial periodicity (Fig. 9). This suggests that the microscopical technique might be capable of resolving structures which are less than 70 nm in size. Although periodicity is certainly not as clear as that seen by the AFM or
Fig. 6. Low-temperature SEM image of the sclera, the collagen fibrils do not show any clear evidence of periodicity along their length. Cross-bridges (arrows) can clearly be seen between adjacent fibrils (×2 insert) (Scale bar = 400 nm).
Fig. 7. X-ray microscope image of fibrils within the sclera, the collagen fibrils running parallel to each other in longitudinal orientation can be resolved (arrows). (Scale bar = 300 nm).
Fig. 8. X-ray microscope image of collagen fibrils in cross-section (arrows). The fibrils are of various diameters. (Scale bar = 300 nm).
Fig. 9. X-ray microscope image of adjacent collagen fibrils with some indication of axial periodicity, or at least variations in density along their length. Arrows indicate bands of higher or lower density along their length. (Scale bar = 100 nm).
510
N.J. Fullwood et al. / Micron 42 (2011) 506–511
Fig. 11. X-ray microscope image of a scleral fibroblast (F), the characteristic shape of the cell is evident, but little detail of the interior of the cell is visible. (Scale bar = 2 m). Fig. 10. X-ray image of longitudinal scleral fibrils, in some regions cross-bridge like structures can be resolved between adjacent fibrils. These can be seen clearly in the ×2 insert (arrows) (Scale bar = 200 nm).
with a stained TEM, TEM requires the use of heavy metal stains for providing contrast enough to image the D-periodicity. In addition, conventional TEM processing results in considerable shrinkage of collagen as well as changes to fibril spacing (Fullwood and Meek, 1993). Although AFM avoids many processing steps necessary for TEM imaging, it does involve physical interaction with the collagen fibrils via the AFM tip which can cause artifacts (Fullwood et al., 1995). Thus at a lower spatial resolution, the images of collagen fibrils captured by the X-ray microscope might be more authentic than those with other microscopical methods. This is supported by the fact that the fibrils look similar to the low-temperature SEM images of the same sample which are in a frozen but hydrated state. Very significantly the X-ray images show what appear to be cross-bridge like structures in some regions (Fig. 10). These are towards the limit of resolution of the X-ray microscope. It is interesting that cross-bridge structures are resolved with the X-ray microscope, because the size of these cross-bridges shown on the TEM is too small to be resolved by X-ray microscopy. However, further reflection and deliberation give a possible explanation on why these show up on the X-ray microscope. In the hydrated state (Fig. 6), the cross-bridge structures are very large and in the range of 100 nm. Furthermore, the glycosaminoglycan chains of the proteoglycans have a significantly higher atomic density than the surrounding molecules. Glycosaminoglycans are heavily sulphated and in addition their negative charge attracts counter ions such as calcium. The high atomic density, relative to the lighter elements present in the collagen fibrils, causes these cross-bridges to exhibit contrast under the X-ray microscope. It seems likely that the structures seen in Fig. 10 are composed of the proteoglycan aggrecan in a partially hydrated state. Other structures which can be identified from the X-ray images include what appear to be whole keratocyte cells (Fig. 11). The shape of the cells is clearly outlined. However, little internal detail is visible within the keratocyte cells. What are the actual advantages of the hard X-ray microscope? Many papers report that the main one is the absence of additional processing routines as is associated with TEM. This study suggests that in addition the hard X-ray microscope is capable of providing unique information. In this study the hard X-ray microscope seemed able to resolve the cross-bridge structures between fibrils, it is significant that these images appear different to conventional TEM images, but similar to low-temperature SEM images. This result suggests that the X-ray images are closer to the in vivo structure of the tissue than conventional TEM images.
Soft X-ray microscopy also can be used to examine frozen hydrated samples, but the sample preparation and interpretation of these samples is not trivial. Hard X-ray microscopes have the ability to image thicker samples, with increased depth in focus, in a partially hydrated state. We believe that hard X-ray microscopes will in the future make significant contributions to the understanding of the ultrastructure of cells and tissues. Acknowledgments This work was supported by the Creative Research Initiatives (Diagnosis of Biofluid Flow Phenomena and Biomimic Research) of MEST/NRF of Korea. X-ray imaging experiments were performed at 1B2 beamline of Pohang Accelerator Laboratory (PAL), Pohang, Korea. References Bonse, U., Beckmann, F., 2010. X-Ray Imaging with Phase Contrast. AIP Conference Proceedings, 1236 , pp. 189–194. Chao, W., Harteneck, B.D., Liddle, J.A., Anderson, E.H., Attwood, D.T., 2005. Soft X-ray nanoscopy at a spatial resolution better than 15 nm. Nature 435, 1210–1213. Dunlevy, J.R., Rada, J.A., 2004. Interaction of lumican with aggrecan in the aging human sclera. Invest. Ophthalmol. Vis. Sci 45, 3849–3856. Fullwood, N.J., Hammiche, A., Pollock, H.M., Hourston, D.J., Song, M., 1995. Atomic force microscopy of the cornea and sclera. Curr. Eye Res. 14, 529–535. Fullwood, N.J., Meek, K.M., 1993. A synchroton X-ray study of the changes occurring in the corneal stroma during processing for electron microscopy. J. Microsc. 169, 53–60. Fullwood, N.J., Meek, K.M., 1994. An ultrastructural, time-resolved study of freezing in the corneal stroma. J. Mol. Biol. 236, 749–758. Fullwood, N.J., Troilo, D., Wallman, J., Meek, K.M., 1993. Synchrotron X-ray diffraction and histochemical studies of normal and myopic chick eyes. Tissue Cell 25, 73–85. Fullwood, N.J., 2008. Microscopy for life scientists. Science 32, 1445. Gardner, D.L., Salter, D.M., Oates, K., 1997. Advances in the microscopy of osteoarthritis. Microsc. Res. Tech. 37, 245–270. Hart, R.W., Farrell, R.A., 1969. Light scattering in the cornea. J. Opt. Soc. Am. 59, 766–774. Hedlund, H., Hedbom, E., Heinegard, D., Mengarelli-Widholm, S., Reinholt, F.P., Svensson, O., 1999. Association of the aggrecan keratan sulfate-rich region with collagen in bovine articular cartilage. J. Biol. Chem. 274, 5777–5781. Ihanamäki, T., Pelliniemi, L.J., Vuorio, E., 2004. Collagens and collagen-related matrix components in the human and mouse eye. Prog. Retin. Eye Res. 23, 403–434. Kagoshima, Y., Ibuki, T., Yokoyama, Y., Tsusaka, Y., Matsui, J., Takai, K., Aino, M., 2001. 10 keV X-ray phase-contrast microscopy for observing transparent specimens. Jpn. J. Appl. Phys. 40, 1190–1192. Kagoshima, Y., Yokoyama, Y., Niimi, T., Koyama, T., Tsusaka, Y., Matsui, J., Takai, K., 2003. Hard X-ray phase contrast microscope for observing transparent specimens. J. Phys. IV 104, 49–52. Kim, G.B., Yoon, Y.J., Shin, T.J., Youn, H.S., Gho, Y.S., Lee, S.J., 2008. X-ray imaging of various biological samples using a phase-contrast hard X-ray microscope. Microsc. Res. Tech. 71, 639–643. Kim, J., Song, K., 2007. Recent progress of nano-technology with NSOM. Micron 38, 409–426.
N.J. Fullwood et al. / Micron 42 (2011) 506–511 Martin, F.L., Kelly, J.G., Llabjani, V., Martin-Hirsch, P.L., Patel, I.I., Trevisan, J., Fullwood, N.J., Walsh, M.J., 2010. Distinguishing cell types or populations based on the computational analysis of their infrared spectra. Nat. Protoc. 5, 1748–1760. Massover, W.H., 2011. New and unconventional approaches for advancing resolution in biological transmission electron microscopy by improving macromolecular specimen preparation and preservation. Micron 42, 141–151. Meek, K.M., Fullwood, N.J., 2001. Corneal and scleral collagens—a microscopist’s perspective. Micron 32, 261–272. Momose, A., Fukuda, J., 1995. Phase-contrast radiographs of nonstained rat cerebellar specimen. Med. Phys. 22, 375–379. Nakamura, T., Kelly, J.G., Trevisan, J., Cooper, L.J., Bentley, A.J., Carmichael, P.L., Scott, A.D., Cotte, M., Susini, J., Martin-Hirsch, P.L., Kinoshita, S., Fullwood, N.J., Martin, F.L., 2010. Microspectroscopy of spectral biomarkers associated with human corneal stem cells. Mol Vis. 16, 359–368. Nugent, K.A., Bellair, C.J., 2003. Emulated super-resolution using quantitative phase microscopy. Micron 34, 333–338. Rodenburg, J.M., Hurst, A.C., Cullis, A.G., Dobson, B.R., Pfeiffer, F., Bunk, O., David, C., Jefimovs, K., Johnson, I., 2007. Hard-X-ray lensless imaging of extended objects. Phys. Rev. Lett. 98, 34801–34804.
511
Vaezy, S., Clark, J.I., 1991. A quantitative analysis of transparency in the human sclera and cornea using Fourier methods. J. Microsc. 163, 85–94. Wessels, J.T., Yamauchi, K., Hoffman, R.M., Wouters, F.S., 2010. Advances in cellular, subcellular, and nanoscale imaging in vitro and in vivo. Cytometry A 77, 667–676. Yokosuka, H., Watanabe, N., Ohigashi, T., Aoki, S., Ando, M., 2003. Phasecontrast hard X-ray microscope with a zone plate at the photon factory. J. Phys. IV 104, 591–594. Youn, H.S., Baik, S.Y., Chang, C.H., 2005. Hard X-ray microscopy with a 130 nm spatial resolution. Rev. Sci. Instrum. 76, 023702–23704. Youn, H.S., Jung, S.W., 2005. Observations of a human hair shaft with an X-ray microscope. Phys. Med. Biol. 21, 5417–5420. Zernike, F., 1935. Das phasenkontrastverfahren bei der mikroskopischen beobachtung. Z. Tech. Phys. 16, 454–457. Zhao, B., Allinson, S.L., Ma, A., Bentley, A.J., Martin, F.L., Fullwood, N.J., 2008. Targeted cornea limbal stem/progenitor cell transfection in an organ culture model. Invest. Ophthalmol. Vis. Sci. 49, 3395–3401.