Research ArticleBREAKTHROUGH TECHNOLOGIES
Open Access

Soft X-Ray Imaging of Cellular Carbon and Nitrogen Distributions in Heterocystous Cyanobacteria

Takahiro Teramoto, Chihiro Azai, Kazuki Terauchi, Masashi Yoshimura, Toshiaki Ohta

Published May 2018. DOI: https://doi.org/10.1104/pp.17.01767

Abstract

Soft x-ray microscopy (SXM) is a minimally invasive technique for single-cell high-resolution imaging as well as the visualization of intracellular distributions of light elements such as carbon, nitrogen, and oxygen. We used SXM to observe photosynthesis and nitrogen fixation in the filamentous cyanobacterium Anabaena sp. PCC 7120, which can form heterocysts during nitrogen starvation. Statistical and spectroscopic analyses from SXM images around the K-absorption edge of nitrogen revealed a significant difference in the carbon-to-nitrogen (C/N) ratio between vegetative cells and heterocysts. Application of this analysis to soft x-ray images of Anabaena sp. PCC 7120 revealed inhomogenous C/N ratios in the cells. Furthermore, soft x-ray tomography of Anabaena sp. PCC 7120 revealed differing cellular C/N ratios, indicating different carbon and nitrogen distributions between vegetative cells and heterocysts in three dimensions.

X-ray microscopes that use soft x-rays as an illumination source are promising tools for microscopic observation in the material and biological sciences (Kirz et al., 1995; Attwood, 1999; McDermott et al., 2009; Chapman and Nugent, 2010; Sakdinawat and Attwood, 2010; Chen et al., 2016). X-ray photons induce the electronic transition of inner-shell core electrons of specific atoms depending on the wavelength. The wavelength range from 2.4 to 4 nm of soft x-rays is called the water window, where the absorbance of water is much less than the C1s, N1s and O 1s inner-shell absorbance of biomolecules. Therefore, such soft x-rays are suitable for imaging hydrated cells in near-native conditions.

Soft x-ray microscopy (SXM) is advantageous over conventional optical or electron microscopy for several reasons. It provides greater spatial resolution (better than 200 nm) because of the shorter x-ray wavelength of 0.1 to 10 nm, and its penetration power (typically 10 μm) is greater than that of electron microscopy (less than 1 μm). Therefore, cellular structures such as organelles can be visualized in vivo or in situ by SXM in the water window without any troublesome preparation. In addition, autofluorescence, which can cause unwanted background noise during fluorescence imaging in the visible region, does not occur in the soft x-ray region because only light with soft x-ray wavelengths is imaged to the x-ray CCD camera with x-ray optical elements. Researchers who use SXM frequently highlight its technical advantages, such as higher spatial resolution, but its biological applications also are valuable. In some cases, dried cells are kept in a vacuum chamber during SXM observation, although there is a potential applicability for biological observations under physiological conditions with conventional SXM.

Filamentous cyanobacteria are multicellular oxygenic phototrophs, some of which are composed of two morphologically and physiologically distinct cell types: vegetative cells and heterocysts when nitrogen is limited. When these two cell types are connected in a filament, they mutually exchange metabolites for their growth and survival (Kumar et al., 2010; Muro-Pastor and Hess, 2012). Photosynthetic vegetative cells provide neighboring heterocysts with sugars and organic acids, while the nitrogen-fixing heterocysts transfer amino acids to vegetative cells (Popa et al., 2007). Metabolites are transported between cells through a cellular junction (Kumar et al., 2010; Muro-Pastor and Hess, 2012). These processes lead to a lower C/N ratio in vegetative cells and a homogenous distribution of carbon and nitrogen in the filaments of Anabaena sp. PCC 7120.

Heterocyst differentiation is tightly regulated by the intracellular signaling system (Wei et al., 1994; Böhme, 1998; Yoon and Golden, 1998; Adams, 2000; Golden and Yoon, 2003; Zhang et al., 2006; Mullineaux et al., 2008; Flores and Herrero, 2010; Muro-Pastor and Hess, 2012; Omairi-Nasser et al., 2015; Muñoz-García and Ares, 2016). The primary cellular signal is the accumulation of 2-oxoglutarate produced by the deamination of Glu (Laurent et al., 2005; Zhao et al., 2010), which is the main cause of the cellular C/N ratio increase (Muro-Pastor and Hess, 2012).

Many microscopic studies of heterocystous cyanobacteria have detailed the morphological changes and physiological functions during heterocyst differentiation. Kumazaki et al. (2013) investigated the dynamics of the thylakoid membrane during heterocyst differentiation in Anabaena variabilis using fluorescence and absorption microspectroscopy. Fluorescence microscopy with GFP-labeled HetR was used to reveal cell lineage dependency (Asai et al., 2009) as well as temporal and spatial fluctuations in key proteins (Corrales-Guerrero et al., 2015) during heterocyst differentiation. In addition, the gram-negative cell wall structures of heterocysts also were investigated (Kangatharalingam et al., 1992; Hoiczyk and Hansel, 2000; Flores et al., 2006; Nicolaisen et al., 2009; Magnuson and Cardona, 2016). Furthermore, intercellular dynamics, such as cell-to-cell communication, revealed that channels with a diameter of less than 20 nm enabled the transport of photosynthetic products from vegetative cells to heterocysts (Omairi-Nasser et al., 2014, 2015; Flores et al., 2016). Electron microscopy revealed detailed structures of cell walls and channels. Energy-filtered transmission electron microscopy imaging is a powerful technique to map out the element distribution in the cell. However, since cells are sliced for observation, it is not applicable to in vivo investigations. Because the techniques mentioned above cannot identify the elements in the cell, no microscopic studies have shown either an increase in C/N ratio or the accumulation of 2-oxoglutarate during heterocyst differentiation in a single cell. The threshold that triggers the signaling cascade has remained unclear, and differentiating a potential heterocyst within a living vegetative cell population was not possible.

The C/N ratio is one characteristic of biological samples that reflects physiological and metabolic activities (Redfield, 1958; Geider and La Roche, 2002). For example, the ratio affects growth and cell differentiation in Arabidopsis (Arabidopsis thaliana). To estimate the cellular C/N ratio, combustion analysis often is used (Lee and Fuhrman, 1987; Fukuda et al., 1998), which quantifies amounts of CO2 and N2 by burning dried cells or have controlled the concentration of sugar (or CO2) and ammonia in a culture (Maekawa et al., 2014). However, combustion is destructive and requires a large amount of sample to obtain reliable estimates; it is not practical for single-cell experiments.

In this study, we describe the use of SXM to derive the single-cell C/N ratio through spectroscopic imaging of nitrogen-starved Anabaena sp. PCC 7120 cells. Spectroscopic and statistical analyses from the absorbance images helped to determine the C/N ratios of both vegetative cells and heterocysts in a single cell in vivo. The cellular C/N ratio was not homogenous, and C and N distributions in vegetative cells and heterocysts were visualized from the soft x-ray images using spectroscopic analysis. Vastly different 3D mappings of subcellular C/N ratio, C, and N distributions also were obtained from vegetative cells and heterocysts.

RESULTS

Soft X-Ray Imaging of Anabaena sp. PCC 7120

Nitrogen-starved cells of Anabaena sp. PCC 7120 were packed tightly between two Si3N4 membranes for SXM. The position and type of each cell in the membranes were determined by optical and fluorescence microscopy (Fig. 1, A and B; see “Materials and Methods”). Heterocysts were identified from the optical images, whereas vegetative cells were observed in fluorescence images because heterocysts lack PSII and do not produce chlorophyll fluorescence (Fig. 1B, cells in the white circle).

Figure 1.
Figure 1.

Soft x-ray images of Anabaena sp. PCC 7120. A and B, Heterocysts and vegetative cells in Anabaena sp. PCC 7120 observed using optical microscopy (A) and fluorescence microscopy (B). Heterocysts are identified in white circles. C, Photoabsorption cross section around the N1s absorption edge. D and E, Soft x-ray images of Anabaena sp. PCC 7120 observed at a photon energy of 398 eV (D; below the N1s absorption edge) and 416 eV (E; above the N1s absorption edge). Bars = 10 μm (A and B) and 2 μm (D and E).

Photon energies of 398 and 416 eV (below and above the N1s edge) were selected for soft x-ray imaging of Anabaena sp. PCC 7120 (Fig. 1, D and E). According to the photon energy dependence of photoabsorption cross sections of the C and N elements in the soft x-ray region (Henke et al., 1993; Fig. 1C), the C1s absorption edge is at a much lower photon energy (∼285 eV) than that used in this study; the tailing part of the C1s absorption continues around the N1s absorption edge. The N1s absorption profile shows an absorption edge at ∼411 eV; above this threshold, the absorption cross section gradually decreases. The bright-field SXM images were collected with a spatial resolution of 110 nm and a magnification of 600×. The field of view was 21 × 21 μm2 (see “Materials and Methods”). As absorbance of soft x-rays increases, the area of absorbance in the images becomes darker. Since the area of the Si3N4 membrane (250 × 250 μm2) was wide enough to collect transmittance images of both cells and blank areas in the same sample (Supplemental Fig. S1), the transmittance image could be converted to absorbance images of the cells, according to the Lambert-Beer law (see “Materials and Methods”).

Statistical Analysis of Soft X-Ray Absorbances of Anabaena sp. PCC 7120 Cells

The C and N element-specific absorbances of each Anabaena sp. PCC 7120 cell were determined by averaging the absorbances of each cell area on the soft x-ray absorbance images at 398 and 416 eV. A total of 131 vegetative cells and 54 heterocysts were investigated for statistical analysis of the absorbance data and summarized in histograms at photon energies of 398 and 416 eV (Fig. 2, A and B). The results showed clear differences between heterocysts and vegetative cells in both histograms; absorbance was greater in the heterocysts than in vegetative cells at both photon energies. Since the concentrations of C and N elements are the main contributors to cellular absorbance near the N1s region, the greater absorbance at 398 and 416 eV in heterocysts indicated that they had greater cellular C and N content compared with the vegetative cells. The C/N ratios of all cells in a single cell layer also were calculated from area-averaged absorbances at the two photon energies (see “Materials and Methods”; Fig. 2C). Differences were found between heterocysts and vegetative cells for mean value (μ) and variance (σ2) in the histogram of C/N ratios. The μ values for heterocysts and vegetative cells were 2.42 and 4.56, respectively, demonstrating that the C/N ratio of heterocysts was less than that of vegetative cells. The σ2 values were 1.38 and 10.7 for histograms of heterocysts and vegetative cells, respectively. The C/N ratio was as high as 18 for vegetative cells but less than 6 for heterocysts. These results suggest that vegetative cells accumulated less N-containing organic compounds than heterocysts.

Figure 2.
Figure 2.

Histograms of absorbances and C/N ratios of heterocysts and vegetative cells. A and B, Histograms of absorbances of heterocysts and vegetative cells with photon energies of 398 eV (A) and 416 eV (B). C, Histogram of C/N ratios of heterocysts and vegetative cells. Red and black bars correspond to heterocysts and vegetative cells, respectively. arb., Arbitrary.

Two-Dimensional Mapping of Intracellular C/N Ratio, C, and N Distributions

The same procedure used to convert soft x-ray absorbances to C/N ratios described above was applied to soft x-ray images to visualize C/N ratios in a single cell. A transmittance image collected at 398 eV (Fig. 3A) was converted to a corresponding C/N ratio image (Fig. 3B; Supplemental Fig. S2), which uses color to indicate the C/N ratio. Distributions of the C/N ratio were not uniform in each cell and showed some subcellular localization. Both heterocysts and vegetative cells had C/N ratios greater than 7 near the center of the cell, but this was more obvious in vegetative cells. The distribution pattern in the vegetative cells did not form a smooth circle but instead was amoeboid in shape (Fig. 3B; Supplemental Fig. S2). At the cell periphery, including cell walls, the C/N ratio was approximately 1 to 2 in heterocysts (Fig. 3B; Supplemental Fig. S2); this same area was thick and dense in soft x-ray images at 398 eV (Fig. 3A).

Figure 3.
Figure 3.

Inhomogenous distributions of C/N ratio and carbon and nitrogen concentrations in the cells. A, Soft x-ray imaging of Anabaena sp. PCC 7120 measured at a photon energy of 398 eV. B, C/N ratio mapping corresponding to the soft x-ray image of A. C and D, Reconstructed reductant carbon (C) and nitrogen (D) concentration distribution corresponding to the soft x-ray image of A. Units in the color bar for A and B are arbitrary units but correspond to pg μm−2 in C and D. Bar in A = 1 μm.

The concentrations of C and N elements also were mapped directly from the transmittance images (see “Materials and Methods”; Fig. 3, C and D; Supplemental Fig. S2). Although the C/N ratio of the heterocysts was low (∼1) and more uniformly distributed than in vegetative cells (Fig. 3B), both C and N concentrations were high in the peripheral region of the cell and had inhomogenous distributions in heterocysts (Fig. 3, C and D; Supplemental Fig. S2). In vegetative cells, the C and N concentrations were less than those in the heterocysts. Carbon was condensed slightly in the peripheral region of the cell. Compared with the relatively uniform distribution of the C at the center of the cell, the N produced an amoeba-shaped inhomogenous distribution (Fig. 3D), similar to the C/N ratio distribution (Fig. 3B).

3D Mapping of Intracellular C/N Ratio Distribution

To visualize the inhomogenous distributions of intracellular C and N, tomographic soft x-ray images and their corresponding element-specific maps of the Anabaena sp. PCC 7120 cells were constructed. The tomographic soft x-ray images were reconstructed by inverse radon transform using 45 pieces of transmittance images of cells at different angles, which were collected by rotating the glass capillary containing the cells (see “Materials and Methods”; Supplemental Fig. S3). The positions and types of cells were determined by optical and fluorescence microscopy (Fig. 4A). The 3D absorbance image of the cyanobacterium and its cross-sectional view with a photon energy of 398 eV were reconstructed (Fig. 4, C and D). The color indicates the absorbance of one voxel of the 3D image (one voxel corresponds to 40 × 40 × 40 nm3). Voxels with an absorbance greater than 0.5 were detected in the peripheral region of heterocyst cells (Fig. 4C). The slice images of these cells also were reconstructed (Supplemental Fig. S4).

Figure 4.
Figure 4.

3D distribution of C/N and C and N concentrations in the cells in Anabaena sp. PCC 7120. A, Optical and fluorescence images of Anabaena sp. PCC 7120 in the capillary. B, 3D display of a tomographic image of Anabaena sp. PCC 7120 measured at a photon energy of 398 eV. C, Cross-sectional display of the tomographic image of B. D, Cross-sectional display of C/N tomography. E and F, Cross-sectional display of reductant C (E) and N (F) concentration distribution. Units in the color bar for C and D are arbitrary units but correspond to pg μm−3 in E and F. Bars in A = 10 μm.

Using the same framework of the 2D mapping as described above, the tomographic image was converted into a 3D map of the C/N ratio (Fig. 4D). The results showed that the C/N ratio was not distributed homogenously in either heterocysts or vegetative cells and was 2 to 3 times greater at the cell center compared with the peripheral regions. 3D maps of the C and N concentrations also were reconstructed with tomographic images (Fig. 4, E and F), which displayed denser concentrations in the heterocysts than in vegetative cells. C was more homogenously distributed than N in both heterocysts and vegetative cells (Fig. 4E; Supplemental Fig. S4), while N in the heterocysts was localized in the peripheral region, where the cellular membranes were located (Fig. 4F; Supplemental Fig. S4). These 3D distributions of the cellular C and N (Fig. 4, E and F; Supplemental Fig. S4) were similar to and consistent with the observed cellular distributions in the 2D maps (Fig. 3). The subcellular distributions of the C and N elements suggest that both C- and N-containing compounds were dense near the membrane in the heterocysts, while vegetative cells accumulated less N-containing organic compounds in the membranes and outside of the cells.

DISCUSSION

Determination of Cellular C/N Ratios from Soft X-Ray Images

SXM observation of the intact cells of Anabaena sp. PCC 7120 was achieved without any pretreatments (Fig. 1). The collected images allowed for a determination of the cellular C/N ratios in individual cells and related histograms (Fig. 2). While no research had been conducted on the C/N ratio in Anabaena sp. PCC 7120, the closely related Anabaena cylindrica was found to have a C/N ratio of 4.5 to 8 during heterocyst differentiation (Kulasooriya et al., 1972). However, this ratio was estimated using the classic combustion analysis without separating heterocysts and vegetative cells. To compare this ratio with the SXM estimation in this study, an averaged C/N ratio of the N-starved Anabaena sp. PCC 7120 was estimated using the histogram of the cellular C/N ratios (Fig. 2C). Assuming a 1:10 population ratio of heterocysts and vegetative cells, the weighted mean value of the C/N histogram was 4.37, which is very close to the C/N ratio of A. cylindrica. This indicates that our procedure to determine cellular C/N ratio is compatible with the classical combustion method but is more useful because of its applicability to in vivo single-cell analyses.

Difference in C/N Ratio between Heterocysts and Vegetative Cells

In the filaments of Anabaena sp. PCC 7120, heterocyst differentiation is triggered by the accumulation of 2-oxoglutarate in the parental vegetative cells, which is the main determinant of the cellular C/N ratio. Our SXM estimation demonstrated that the µ value was greater in vegetative cells compared with heterocysts (Fig. 2C). In the histograms, distributions at values less than 6 were similar between heterocysts and vegetative cells, while the histogram of vegetative cells had a much greater distribution value than that of heterocysts (maximum values were 18 versus 6, respectively; Fig. 2C). The greater C/N values in the vegetative cells are likely caused by the deprivation of N-containing organic compounds and the accumulation of 2-oxoglutarate. The population of vegetative cells with C/N ratios greater than 6 was 21% in the histogram, similar to the population ratio of heterocysts and vegetative cells in a filament. Therefore, these vegetative cells may be potential or parental cells that are triggered for heterocyst differentiation. These parental cells should be located between two heterocysts that are separated by 10 to 20 vegetative cells in the same filament because of the regulation system with diffusible HetN and PatS. In this study, since the Anabaena sp. PCC 7120 cells were packed densely and did not maintain the filament structure when observed with SXM, the positions of the vegetative cells with high C/N ratios in the filaments could not be identified. Hence, sparse packing to keep cells in contact within the filament is required for visualization of the C/N ratio distribution.

Cell fate during differentiation depends on the concentration of key molecules, time, and the cell’s environment, including its pH and temperature (Zeng et al., 2010). Since the potential in the cell fluctuates, interstate transitions during differentiation can occur stochastically. This fluctuation is frequently called noise and can be used to quantify the cell differentiation process (Raser and O’Shea, 2005; Arias and Hayward, 2006; Süel et al., 2007). Interestingly, Corrales-Guerrero et al. (2015) investigated the noise during cell differentiation in cyanobacteria. They quantified the noise (cell-to-cell fluctuation of the number of HetR) in filamentous Anabaena sp. PCC 7120, with several mutated species to identify key proteins. In that study, the accumulation of 2-oxoglutarate occurred after nitrogen deprivation, and NtcA (which is a receptor of 2-oxoglutarate; Zhao et al., 2010) bound to DNA for the promotion of heterocyst differentiation. Therefore, the C/N ratio in cells also can be considered a key factor in heterocyst differentiation. The noise σ2/µ2 of the C/N ratio was determined to be 0.24 for heterocysts and 0.51 for vegetative cells. Because the noise of the C/N ratio in this study originated from cell-to-cell fluctuation and the heterocysts were finished differentiating, the 2-fold greater value of the noise in the vegetative cells can be regarded as the possibility of cellular division.

The earth mover’s distance (EMD) is a measure of the distance between two probability distributions, and the EMD-based similarity analysis can be used as an effective tool in pattern recognition (Rubner et al., 2000; Corrales-Guerrero et al., 2015), as already applied in several biological studies (Bernas et al., 2006; Eyiyurekli et al., 2008; Corrales-Guerrero et al., 2015; Orlova et al., 2016). Here, we used the EMD to quantify the dissimilarity of the histograms between heterocysts and vegetative cells (see “Materials and Methods”). The EMD values of the histograms of absorbances below and above the N1s threshold between heterocysts and vegetative cells were 0.2 ± 0.07 and 0.064 ± 0.02, respectively, where the ± values indicate sd. However, the EMD value was estimated to be 2.14 ± 0.79 from the histograms of the C/N ratios between heterocyst and vegetative cells, indicating that the C/N ratio is more sensitive to the dissimilarity than the absorbance data. (The sd in the EMD values were estimated by a bootstrap test of 1,000 iterations.)

Subcellular Distributions of C and N in Anabaena sp. PCC 7120

SXM imaging revealed that the subcellular C/N ratio and C and N distributions in heterocysts and vegetative cells of Anabaena sp. PCC 7120 (Figs. 3 and 4) were not uniform in both cell types. In vegetative cells, the center had an amoeboid-like distribution of the C/N ratio (greater than 7) and C concentration, which were greater than those in the peripheral region (Fig. 3). Since vegetative cells have a thylakoid-like inner membrane structure, which is absent in heterocysts, the amoeboid-like distribution suggests an accumulation of C-rich compounds in the cytosol or cytosolic organelle.

In contrast, heterocysts had N localized in the membrane-like peripheral region of the cell (Figs. 3 and 4; Supplemental Fig. S4). Since the C compounds showed similar peripheral localization and a rather uniform distribution of C/N ratio (Fig. 3), the peripheral localization of the N-containing compounds suggests an accumulation of N-containing organic compounds in the heterocyst membrane or its cell wall. The heterocyst is a gram-negative cell, and its cell wall is composed of three layers consisting of (from the inner to the outer membrane) peptidoglycans, glycolipids, and polysaccharides (Fay, 1992; Kumar et al., 2010; Muro-Pastor and Hess, 2012). These compounds contain C, O, and H elements but lack N. Peripheral localization with a high C concentration can be attributed to these N-lacking polymers of the cell wall structure. Based on previous reports using immunoelectron microscopy (Sherman et al., 2000) and energy-filtered transmission electron microscopy images (Koop et al., 2007), the N distribution around the cell wall in heterocysts can be assigned to cyanophycin grana proteins that act as a nitrogen reservoir at the grana membrane.

The cellular structure of 3D images of Anabaena sp. PCC 7120 can be clarified without the need for slicing with a microtome (Fig. 4, B and C). In addition, the elemental distribution of C, N, and the C/N ratio can be visualized in 3D (Fig. 4, D–F; Supplemental Fig. S4). The center of both cell types contained a high C/N ratio with a high concentration of C. In addition, a high C/N ratio occurred at the connection between the heterocyst and vegetative cells, with a greater C distribution and lower N distribution. Cell-to-cell communication between vegetative cells and heterocysts is well known; vegetative cells transport by-products from photosynthesis (such as Suc) to the heterocysts, and heterocysts transport the N-containing compounds (including amino acids) to the vegetative cells (Böhme, 1998; Golden and Yoon, 2003; Zhang et al., 2006; Mullineaux et al., 2008; Flores and Herrero, 2010; Muro-Pastor and Hess, 2012; Omairi-Nasser et al., 2015; Muñoz-García and Ares, 2016). In our research here, although the N amounts were small, the transportation of Suc from vegetative cells to heterocysts was visualized. The C and N distribution images (Fig. 4, D and E; Supplemental Fig. S4) allowed for visualization of the localization of membrane polymers and cyanophycin at the membrane in heterocysts.

We succeeded in packing the cells in the capillary while keeping the filament structures intact during experiments. The capillary diameter used for tomographic data collection narrowed toward the tip end (Fig. 4A). Therefore, the optical path of the soft x-ray, which is proportional to the capillary diameter at each observed position, increases from left to right (Fig. 4A). If it exceeds the penetration depth of the soft x-ray, then it is not possible to image the cell. In our experiments, we observed cells in the SXM, as displayed in Figure 4B. Future research will encompass the design and optimization of the packing material as well as develop techniques to keep cells and filaments intact.

CONCLUSION

SXM images of nitrogen-starved cells of the filamentous cyanobacterium Anabaena sp. PCC 7120 were visualized without staining or other treatments. Application of the Lambert-Beer law to the soft x-ray transmittance images at photon energies below and above the N1s absorption edge produced maps of the C/N ratio and C and N elements in single cells and at subcellular levels in both 2D and 3D. Statistical analyses of the cellular C/N ratio and the C and N contents in the cell revealed that 21% of the vegetative cells had high C/N ratios, indicating that they may differentiate into heterocysts. These element-specific microscopic maps showed that heterocysts accumulated N-containing organic compounds in the peripheral region of the cell. Temporal change in the distribution of the cellular C/N ratio is a potential and promising target for elucidation of the mechanism of heterocyst differentiation. This can be visualized and traced using several heterocyst mutant strains of Anabaena sp. PCC 7120 as well as with artificial triggering using 2-oxoglutarate.

SXM near the N1s absorption edge developed in this study is a powerful method for the observation and analysis of inhomogeneity within cell populations and subcellular structures. Since a high C/N ratio can be used for signatures of high-carbon compounds in the cell, the SXM visualization of the subcellular C/N ratio distribution can be applied to other physiological or cytological studies. Examples of such studies include oil production by microalgae (Suzuki et al., 2013; Khatri et al., 2014; Cornejo-Corona et al., 2016) and the CO2-concentrating processes in the algal pyrenoid (Freemen Rosenzweig et al., 2017) and the cyanobacterial carboxysome (Turmo et al., 2017).

MATERIALS AND METHODS

Bacterial Strain and Culture

Anabaena sp. PCC 7120 was routinely streaked and grown under nitrogen-depleted conditions on agar plates of BG-110 (Rippka et al., 1979). Plates were incubated at 30°C and irradiated by fluorescent lamps at the photosynthetically active radiation of 30 µE m−2 s−1.

Sample Preparation for Soft X-Ray Imaging

Colonies of cell filaments of Anabaena sp. PCC 7120 grown on the BG-110 plate were collected, suspended in pure water, and tightly packed between two silicon nitride membranes (250 μm × 250 μm) with a thickness of 100 nm. The cells of multiple filaments on the same culture plate were used for statistical analyses of SXM images. The absolute position of each cell in the layers was determined by optical and fluorescence microscopy (Nikon Optiphot-2) with a magnification of 200× (Supplemental Fig. S1).

SXM

SXM observations were performed at the full-field transmission SXM beamline (BL-12) at the SR Center of Ritsumeikan University. X-ray energies selected were 398 and 416 eV (above and below the N1s absorption threshold, respectively). The x-ray was monochromatized by chromatic aberration of a Fresnel zone plate using a combination of a condenser Fresnel zone plate and a pinhole with a 20-μm radius. The energy resolution (Ε/ΔΕ) was 200. Samples were imaged with an objective Fresnel zone plate and a back-illuminated CCD camera (C4880-21-24WD; Hamamatsu Photonics). The pixel size and number of pixels of this camera are 24 μm × 24 μm and 512 × 512, respectively. Spatial resolution was 110 nm at a magnification of 600×, estimated from the knife-edge method. Thus, the corresponding area size in the image was 40 × 40 nm2 per 1 × 1 pixel. The field of view was 21 μm × 21 μm. A sample sandwiched between two silicon nitride membranes was held on a sample holder. The sample holder was mounted on a four-axis PC-controlled stage to adjust the microscopic observation in the SXM beamline. Rough alignment of the imaging was done using an optical microscope equipped with the beamline. Next, the soft x-ray imaging was set up, and a final optimal adjustment was conducted with imaging acquisition of the actual target. Exposure time was set to 3 min for the experiments at room temperature.

Soft X-Ray Tomography

3D transmission imaging of the cyanobacterium Anabaena sp. PCC7120 was conducted using a computer tomography method. A fine-tipped glass capillary with an i.d. of about 5 μm was used to hold a filamentous cyanobacterium. The distance from beam stop, which is used to block unwanted x-rays, to the sample is typically less than 1 mm. For tomographic imaging, images of samples rotated from 0° to 180° in the normal plane to the illuminated soft x-ray must be acquired. However, because the size of the Si plate attached to Si3N4 membranes used for 2D imaging restricted the rotational angle within several degrees (5 × 5 mm, a typical size in an SXM study), we instead used a glass capillary to hold the cyanobacteria. The capillary was mounted on a rotational stage to rotate around the capillary axis. Tilt-series images were taken every 4° between 0° and 180°. The exposure time was 3 min per image. Tomographic image acquisition was conducted at two photon energies (398 and 416 eV). Samples were kept cool at a temperature near 165 K with liquid nitrogen to reduce radiation damage during the acquisition time. A tomographic image was reconstructed from 45 tilt-series images with inverse radon transform using MATLAB source code (Supplemental Fig. S3).

Derivation of the C/N Ratio from Soft X-Ray Images

Because the SXM used in this study was full-field bright-field microscopy, the observed images of cells correspond to the transmittance image at a certain soft x-ray photon energy. Transmitted intensity (I or I0), which is the light intensity with or without a cell, respectively, was acquired directly from soft x-ray images. Due to the large size of the Si3N4 membrane used as a sample cell in this study (250 × 250 μm), a blank area exists without a cell, and this image was I0. Transmitted images were converted to the absorbance (A) images of cells according to the Lambert-Beer law: Embedded Image)]

In this study, soft x-rays were used with photon energy below and above N1s k-edge absorption. Around these photon energies, the main contribution to the absorbance from cells are the C and N components. The absorption of O components, including water, was negligible, because the O1s absorption edge is at a far higher photon energy and the absorbance cross section is too low near the N1s edge.

Thus, the total absorbance at a certain photon energy can be defined as follows:Embedded Imagewhere Embedded Image is absorbance at photon energy Embedded Image, Embedded Image and Embedded Image are absorption cross section and concentration of the i-th element (C or N), respectively, and Embedded Image is the optical path length that was assumed to be the thickness of cell.

The relationship between absorbance and the C/N concentration ratios was derived from the absorbance at two different wavelengths (Embedded Image), as follows:Embedded ImageEmbedded Imagewhere Embedded Imageis the C/N ratio of the cell. The absorption cross sections Embedded Image of C and N components are listed in Table I.

View this table:
Table I. Photoabsorption cross section of C and N components

The atomic photoabsorption cross section, Embedded Image, was obtained from the imaginary part of the atomic scattering factor Embedded Image using the relationEmbedded Image, where Embedded Image is the classical electron radius and Embedded Image is the wavelength. The calculated photoabsorption cross sections were compared with experimental data (Henke et al., 1993). The database of photoabsorption cross sections (https://www.cxro.lbl.gov/) also is available.

Reconstruction of C and N Concentrations

From Equations 1 and 2, the concentrations of C and N can be derived as follows:

Embedded ImageEmbedded Image

Knowing (or estimating) the optical lengthEmbedded Image allows the calculation of C and N concentrations directly from soft x-ray images. However, if the path length is unknown or uncertain, this method is not applicable. Therefore, in this study, instead of Equations 4 and 5, the reductant concentration (density multiplied by path length) was examined using the following equations:

Embedded ImageEmbedded Image

In the 2D images, the dimension of Embedded Image becomes [g m−2], because 2D images are a projection of the accumulated concentration of each element normal to the image plane. For 3D images of C and N concentration, path length Embedded Image was defined in pixel size from the camera and the dimension defined as [g m−3].

EMD

The EMD quantifies the dissimilarity among objects being compared (Rubner et al., 2000; Corrales-Guerrero et al., 2015). It is based on minimizing the transportation cost of a product from supplier to customer. In this study, we used the EMD to quantify the dissimilarity among the histograms of C/N ratios between heterocysts and vegetative cells. For our case:Embedded Image and Embedded Image. Here, H and V are the clusters of heterocysts and vegetative cells. hi (vi) and whi (wvi) refer to the value and weight of the i-th bin in the histogram of heterocysts (vegetative cells). Here, we introduce the distance matrix dij and the flow matrix fij to consider the work as described below. This work acts as the minimum cost for the transportation problem between the H and V clusters. dij is the distance (or difference) between hi and vi. fij is related to whi and wvi. Under minimum cost conditions, WORK is defined as below with the clusters Embedded Image:Embedded ImageThis has the following constraints: Embedded ImageEmbedded ImageEmbedded ImageEmbedded ImageUnder these conditions, the EMD is defined as below: Embedded ImageIn the algorithm, WORK was iteratively calculated by changing the combination of i and j iteratively to get the minimum cost for converting H to V. If the optimum cost was determined, WORK is normalized to get EMD.

Supplemental Data

The following supplemental materials are available.

Acknowledgments

We thank Dr. Yuichi Fujita of Nagoya University for the gift of the nitrogen-fixation active culture of Anabaena sp. PCC 7120.

Footnotes

  • www.plantphysiol.org/cgi/doi/10.1104/pp.17.01767

  • The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Takahiro Teramoto (tteramo{at}fc.ritsumei.ac.jp).

  • T.T. and M.Y. performed the majority of the experiments and analyses; C.A., K.T., and T.O. designed and supervised the experiments.

  • ↵1 This study was supported in part by Grants-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (nos. 15H00886, 17H05731 [to C.A.], 23570066, 26102543, 16H00784, 17K19247 [to K.T.], and 16KT0167 [to [T.T.]). This research is partially supported by the Matching Planner Program from Japan Science and Technology Agency (to T.T.).

  • ↵[OPEN] Articles can be viewed without a subscription.

  • Received December 14, 2017.
  • Accepted March 20, 2018.
  • Published March 26, 2018.

REFERENCES

  1. ↵
    1. Adams DG
    (2000) Heterocyst formation in cyanobacteria. Curr Opin Microbiol 3: 618–624
    OpenUrlCrossRefPubMed
  2. ↵
    1. Arias AM,
    2. Hayward P
    (2006) Filtering transcriptional noise during development: concepts and mechanisms. Nat Rev Genet 7: 34–44
    OpenUrlCrossRefPubMed
  3. ↵
    1. Asai H,
    2. Iwamori S,
    3. Kawai K,
    4. Ehira S,
    5. Ishihara J,
    6. Aihara K,
    7. Shoji S,
    8. Iwasaki H
    (2009) Cyanobacterial cell lineage analysis of the spatiotemporal hetR expression profile during heterocyst pattern formation in Anabaena sp. PCC 7120. PLoS ONE 4: e7371
    OpenUrlCrossRefPubMed
  4. ↵
    1. Attwood D
    (1999) Soft X-Rays and Extreme Ultraviolet Radiation: Principles and Applications. Cambridge University Press, Cambridge, UK
  5. ↵
    1. Bernas T,
    2. Asem EK,
    3. Robinson JP,
    4. Rajwa B
    (2006) Compression of fluorescence microscopy images based on the signal-to-noise estimation. Microsc Res Tech 69: 1–9
    OpenUrlPubMed
  6. ↵
    1. Böhme H
    (1998) Regulation of nitrogen fixation in heterocyst-forming cyanobacteria. Trends Plant Sci 3: 346–351
    OpenUrlCrossRef
  7. ↵
    1. Chapman HN,
    2. Nugent KA
    (2010) Coherent lensless x-ray imaging. Nat Photonics 4: 833–839
    OpenUrlCrossRef
  8. ↵
    1. Chen HY,
    2. Chiang DML,
    3. Lin ZJ,
    4. Hsieh CC,
    5. Yin GC,
    6. Weng IC,
    7. Guttermann P,
    8. Werner S,
    9. Henzler K,
    10. Schneider G, et al.
    (2016) Nanoimaging granule dynamics and subcellular structures in activated mast cells using soft x-ray tomography. Sci Rep 6: 34879
    OpenUrlCrossRef
  9. ↵
    1. Cornejo-Corona I,
    2. Thapa HR,
    3. Browne DR,
    4. Devarenne TP,
    5. Lozoya-Gloria E
    (2016) Stress responses of the oil-producing green microalga Botryococcus braunii Race B. PeerJ 4: e2748
    OpenUrl
  10. ↵
    1. Corrales-Guerrero L,
    2. Tal A,
    3. Arbel-Goren R,
    4. Mariscal V,
    5. Flores E,
    6. Herrero A,
    7. Stavans J
    (2015) Spatial fluctuations in expression of the heterocyst differentiation regulatory gene hetR in Anabaena filaments. PLoS Genet 11: e1005031
    OpenUrlCrossRefPubMed
  11. ↵
    1. Eyiyurekli M,
    2. Manley P,
    3. Lelkes PI,
    4. Breen DE
    (2008) A computational model of chemotaxis-based cell aggregation. Biosystems 93: 226–239
    OpenUrlPubMed
  12. ↵
    1. Fay P
    (1992) Oxygen relations of nitrogen fixation in cyanobacteria. Microbiol Rev 56: 340–373
    OpenUrlAbstract/FREE Full Text
  13. ↵
    1. Flores E,
    2. Herrero A
    (2010) Compartmentalized function through cell differentiation in filamentous cyanobacteria. Nat Rev Microbiol 8: 39–50
    OpenUrlCrossRefPubMed
  14. ↵
    1. Flores E,
    2. Herrero A,
    3. Forchhammer K,
    4. Maldener I
    (2016) Septal junctions in filamentous heterocyst-forming cyanobacteria. Trends Microbiol 24: 79–82
    OpenUrlCrossRefPubMed
  15. ↵
    1. Flores E,
    2. Herrero A,
    3. Wolk CP,
    4. Maldener I
    (2006) Is the periplasm continuous in filamentous multicellular cyanobacteria? Trends Microbiol 14: 439–443
    OpenUrlCrossRefPubMed
    1. Freeman Rosenzweig ES,
    2. Xu B,
    3. Kuhn Cuellar L,
    4. Martinez-Sanchez A,
    5. Schaffer M,
    6. Strauss M,
    7. Cartwright HN,
    8. Ronceray P,
    9. Plitzko JM,
    10. Förster F, et al.
    (2017) The eukaryotic CO2-concentrating organelle is liquid-like and exhibits dynamic reorganization. Cell 171: 148–162.e19
    OpenUrl
  17. ↵
    1. Fukuda R,
    2. Ogawa H,
    3. Nagata T,
    4. Koike I
    (1998) Direct determination of carbon and nitrogen contents of natural bacterial assemblages in marine environments. Appl Environ Microbiol 64: 3352–3358
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Geider R,
    2. La Roche J
    (2002) Redfield revisited: variability of C:N:P in marine microalgae and its biochemical basis. Eur J Phycol 37: 1–17
    OpenUrlCrossRef
  19. ↵
    1. Golden JW,
    2. Yoon HS
    (2003) Heterocyst development in Anabaena. Curr Opin Microbiol 6: 557–563
    OpenUrlCrossRefPubMed
  20. ↵
    1. Henke BL,
    2. Gullikson EM,
    3. Davis JC
    (1993) X-ray interactions: photoabsorption, scattering, transmission, and reflection at E = 50-30,000 eV, Z = 1-92. Atom Data and Nuclear Data Tables 54: 181–342
    OpenUrl
  21. ↵
    1. Hoiczyk E,
    2. Hansel A
    (2000) Cyanobacterial cell walls: news from an unusual prokaryotic envelope. J Bacteriol 182: 1191–1199
    OpenUrlFREE Full Text
  22. ↵
    1. Kangatharalingam N,
    2. Priscu JC,
    3. Paerl HW
    (1992) Heterocyst envelope thickness, heterocyst frequency and nitrogenase activity in Anabaena flos-aquae: influence of exogenous oxygen tension. J Gen Microbiol 138: 2673–2678
    OpenUrlCrossRef
  23. ↵
    1. Khatri W,
    2. Hendrix R,
    3. Niehaus T,
    4. Chappell J,
    5. Curtis WR
    (2014) Hydrocarbon production in high density Botryococcus braunii race B continuous culture. Biotechnol Bioeng 111: 493–503
    OpenUrl
  24. ↵
    1. Kirz J,
    2. Jacobsen C,
    3. Howells M
    (1995) Soft x-ray microscopes and their biological applications. Q Rev Biophys 28: 33–130
    OpenUrlCrossRefPubMed
  25. ↵
    1. Koop A,
    2. Voss I,
    3. Thesing A,
    4. Kohl H,
    5. Reichelt R,
    6. Steinbüchel A
    (2007) Identification and localization of cyanophycin in bacteria cells via imaging of the nitrogen distribution using energy-filtering transmission electron microscopy. Biomacromolecules 8: 2675–2683
    OpenUrlCrossRefPubMed
  26. ↵
    1. Kulasooriya SA,
    2. Lang NJ,
    3. Fay P
    (1972) The heterocysts of blue-green algae. 3. Differentiation and nitrogenase activity. Proc R Soc Lond B Biol Sci 181: 199–209
    OpenUrlAbstract/FREE Full Text
  27. ↵
    1. Kumar K,
    2. Mella-Herrera RA,
    3. Golden JW
    (2010) Cyanobacterial heterocysts. Cold Spring Harb Perspect Biol 2: a000315
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Kumazaki S,
    2. Akari M,
    3. Hasegawa M
    (2013) Transformation of thylakoid membranes during differentiation from vegetative cell into heterocyst visualized by microscopic spectral imaging. Plant Physiol 161: 1321–1333
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Laurent S,
    2. Chen H,
    3. Bédu S,
    4. Ziarelli F,
    5. Peng L,
    6. Zhang CC
    (2005) Nonmetabolizable analogue of 2-oxoglutarate elicits heterocyst differentiation under repressive conditions in Anabaena sp. PCC 7120. Proc Natl Acad Sci USA 102: 9907–9912
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Lee S,
    2. Fuhrman JA
    (1987) Relationships between biovolume and biomass of naturally derived marine bacterioplankton. Appl Environ Microbiol 53: 1298–1303
    OpenUrlAbstract/FREE Full Text
  31. ↵
    1. Maekawa S,
    2. Inada N,
    3. Yasuda S,
    4. Fukao Y,
    5. Fujiwara M,
    6. Sato T,
    7. Yamaguchi J
    (2014) The carbon/nitrogen regulator ARABIDOPSIS TOXICOS EN LEVADURA31 controls papilla formation in response to powdery mildew fungi penetration by interacting with SYNTAXIN OF PLANTS121 in Arabidopsis. Plant Physiol 164: 879–887
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Magnuson A,
    2. Cardona T
    (2016) Thylakoid membrane function in heterocysts. Biochim Biophys Acta 857: 309–319
    OpenUrl
  33. ↵
    1. McDermott G,
    2. Le Gros MA,
    3. Knoechel CG,
    4. Uchida M,
    5. Larabell CA
    (2009) Soft x-ray tomography and cryogenic light microscopy: the cool combination in cellular imaging. Trends Cell Biol 19: 587–595
    OpenUrlCrossRefPubMed
  34. ↵
    1. Mullineaux CW,
    2. Mariscal V,
    3. Nenninger A,
    4. Khanum H,
    5. Herrero A,
    6. Flores E,
    7. Adams DG
    (2008) Mechanism of intercellular molecular exchange in heterocyst-forming cyanobacteria. EMBO J 27: 1299–1308
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Muñoz-García J,
    2. Ares S
    (2016) Formation and maintenance of nitrogen-fixing cell patterns in filamentous cyanobacteria. Proc Natl Acad Sci USA 113: 6218–6223
    OpenUrlAbstract/FREE Full Text
  36. ↵
    1. Muro-Pastor AM,
    2. Hess WR
    (2012) Heterocyst differentiation: from single mutants to global approaches. Trends Microbiol 20: 548–557
    OpenUrlCrossRefPubMed
  37. ↵
    1. Nicolaisen K,
    2. Hahn A,
    3. Schleiff E
    (2009) The cell wall in heterocyst formation by Anabaena sp. PCC 7120. J Basic Microbiol 49: 5–24
    OpenUrlCrossRefPubMed
  38. ↵
    1. Omairi-Nasser A,
    2. Haselkorn R,
    3. Austin J II.
    (2014) Visualization of channels connecting cells in filamentous nitrogen-fixing cyanobacteria. FASEB J 28: 3016–3022
    OpenUrlAbstract/FREE Full Text
  39. ↵
    1. Omairi-Nasser A,
    2. Mariscal V,
    3. Austin JR II.,
    4. Haselkorn R
    (2015) Requirement of Fra proteins for communication channels between cells in the filamentous nitrogen-fixing cyanobacterium Anabaena sp. PCC 7120. Proc Natl Acad Sci USA 112: E4458–E4464
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Orlova DY,
    2. Zimmerman N,
    3. Meehan S,
    4. Meehan C,
    5. Waters J,
    6. Ghosn EEB,
    7. Filatenkov A,
    8. Kolyagin GA,
    9. Gernez Y,
    10. Tsuda S, et al.
    (2016) Earth mover’s distance (EMD): a true metric for comparing biomarker expression levels in cell populations. PLoS ONE 11: e0151859
    OpenUrl
  41. ↵
    1. Popa R,
    2. Weber PK,
    3. Pett-Ridge J,
    4. Finzi JA,
    5. Fallon SJ,
    6. Hutcheon ID,
    7. Nealson KH,
    8. Capone DG
    (2007) Carbon and nitrogen fixation and metabolite exchange in and between individual cells of Anabaena oscillarioides. ISME J 1: 354–360
    OpenUrlCrossRefPubMed
  42. ↵
    1. Raser JM,
    2. O’Shea EK
    (2005) Noise in gene expression: origins, consequences, and control. Science 309: 2010–2013
    OpenUrlAbstract/FREE Full Text
  43. ↵
    1. Redfield AC
    (1958) The biological control of chemical factors in the environment. Am Sci 46: 205–221
    OpenUrlCrossRef
  44. ↵
    1. Rippka R,
    2. Deruelles J,
    3. Waterbury JB,
    4. Herdman M,
    5. Stanier RY
    (1979) Generic assignments, strain histories and properties of pure cultures of cyanobacteria. J Gen Microbiol 111: 1–61
    OpenUrlCrossRef
  45. ↵
    1. Rubner Y,
    2. Tomasi C,
    3. Guibas LJ
    (2000) The earth mover’s distance as a metric for image retrieval. Int J Comput Vis 40: 99–121
    OpenUrlCrossRef
  46. ↵
    1. Sakdinawat A,
    2. Attwood D
    (2010) Nanoscale x-ray imaging. Nat Photonics 4: 840–848
    OpenUrlCrossRef
  47. ↵
    1. Sherman DM,
    2. Tucker D,
    3. Sherman LA
    (2000) Heterocyst development and localization of cyanophycin in N2-fixing cultures of Anabaena sp. PCC 7120 (cyanobacteria). J Phycol 36: 932–941
    OpenUrlCrossRef
  48. ↵
    1. Süel GM,
    2. Kulkarni RP,
    3. Dworkin J,
    4. Garcia-Ojalvo J,
    5. Elowitz MB
    (2007) Tunability and noise dependence in differentiation dynamics. Science 315: 1716–1719
    OpenUrlAbstract/FREE Full Text
  49. ↵
    1. Suzuki R,
    2. Ito N,
    3. Uno Y,
    4. Nishii I,
    5. Kagiwada S,
    6. Okada S,
    7. Noguchi T
    (2013) Transformation of lipid bodies related to hydrocarbon accumulation in a green alga, Botryococcus braunii (Race B). PLoS ONE 8: e81626
    OpenUrl
  50. ↵
    1. Turmo A,
    2. Gonzalez-Esquer CR,
    3. Kerfeld CA
    (2017) Carboxysomes: metabolic modules for CO2 fixation. FEMS Microbiol Lett 364: fnx176
    OpenUrl
  51. ↵
    1. Wei TF,
    2. Ramasubramanian TS,
    3. Golden JW
    (1994) Anabaena sp. strain PCC 7120 ntcA gene required for growth on nitrate and heterocyst development. J Bacteriol 176: 4473–4482
    OpenUrlAbstract/FREE Full Text
  52. ↵
    1. Yoon HS,
    2. Golden JW
    (1998) Heterocyst pattern formation controlled by a diffusible peptide. Science 282: 935–938
    OpenUrlAbstract/FREE Full Text
  53. ↵
    1. Zeng L,
    2. Skinner SO,
    3. Zong C,
    4. Sippy J,
    5. Feiss M,
    6. Golding I
    (2010) Decision making at a subcellular level determines the outcome of bacteriophage infection. Cell 141: 682–691
    OpenUrlCrossRefPubMed
  54. ↵
    1. Zhang CC,
    2. Laurent S,
    3. Sakr S,
    4. Peng L,
    5. Bédu S
    (2006) Heterocyst differentiation and pattern formation in cyanobacteria: a chorus of signals. Mol Microbiol 59: 367–375
    OpenUrlCrossRefPubMed
    1. Zhao M,
    2. Jiang Y,
    3. He Y,
    4. Chen Y,
    5. Teng Y,
    6. Chena Y,
    7. Zhang C,
    8. Zhou C
    (2010) Structural basis for the allosteric control of the global transcription factor NtcA by the nitrogen starvation signal 2-oxoglutarate. Proc Natl Acad Sci USA 107: 12487–12492
    OpenUrlAbstract/FREE Full Text
View Abstract
Back to top

Table of Contents

Print
Download PDF
Email Article
Citation Tools
Request Permissions
Soft X-Ray Imaging of Cellular Carbon and Nitrogen Distributions in Heterocystous Cyanobacteria
Takahiro Teramoto, Chihiro Azai, Kazuki Terauchi, Masashi Yoshimura, Toshiaki Ohta
Plant Physiology May 2018, 177 (1) 52-61; DOI: 10.1104/pp.17.01767
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

In this issue

View this article with LENS