The light scattering microscopy of metallic or organic NPs has been well investigated for decades. Theoretical and experimental studies have provided established fundamental concepts of light scattering dynamics, microscopy, and imaging of NPs. One of interesting features of this technique is that the scattering spectrum can give us information on the shape, size, refractive index, electron density of the target particles, as well as the refractive index of the medium. Thus, by utilizing such a technique with grana of plant cells being the target particles, we study optically and spectroscopically the configuration and size of in vivo grana.
For this purpose, we have used Egeria densa Planchon (Hydrocharitaceae), commonly known as Brazilian elodea or common waterweed, as the target. This submersed perennial
plant species native to south-eastern America was selected due to its well-known “ecosystem engineer,” given its role in stabilizing sediment and reducing turbidity and its important role
in trophic dynamics[26]. The alga with a 2 cell thick leaf was cultivated in an aquarium in our lab (Fig. 4.1). To maintain the healthiness of the alga, visually indicated by its growth and green leaves, we set the aquarium in light area and exchange the water medium once a week.
28
Figure 4.1 A picture of Egeria densa in the aquarium.
4.1 Sample preparation
The sample was prepared by sandwiching a section of 1 mm × 3 mm of green leaf of E. densa with two cover slips (18 mm × 18 mm and 24 mm × 40 mm, Mastsunami). The gap
of the cover slips was filled with water to keep it wet, and it was sealed with a nail polish to avoid evaporation. The typical sample cell containing the green cell of E. densa is shown in Fig. 4.2.
29
Figure 4.2 A sample in dimension of 1 mm × 3 mm was cut by scissor. The upper and top
glass substrates were 24 mm × 40 mm and 18 mm × 18 mm, respectively. Additional water and nail polish to seal glasses were used to keep plant alive.
Under bright and dark field (with objective lens 100×, N.A. =1.4), the plant cells in the leaf are observed as “islands” with white and dark-green area respectively (Fig. 4.3). More specifically, by dark-field imaging, chloroplasts are easily distinguished and each cell is bordered by cell wall which can be seen obviously. Dark-field image also shows higher contrast that could be a great help of observation.
Figure 4.3 (A) Bright field and (B) dark field images of plant cell of Egeria densa
A B
30
4.2 Results and discussion
4.2-1 Three dimensional morphological analysis of Egeria densa
By changing detection system to confocal microscopy, with the objective lens 100×, N.A.= 1.4, we are able to observe a single granum in more detail. Granum, single grana, is composed of stacked dicoidal membranous system called thylakoids where all the molecular complexes that drive the light-induced reaction and provide a medium for energy transduction are located.
Though such a 3-D organization of granum has been revealed by high resolution electron tomography [16, 18, 32], in this study we show that our optical, spectroscopic and non-destructive method can be used to explore the characteristic of grana. The architecture of grana provides insights into their formation and function to clarify light-harvesting and electron transport.
Thylakoid membranes could not be recognized by our system. However, distinguishing each single granum in vivo is quite easy. Confocal light scattering spectroscopic images of a living plant cell with different magnifications are shown in Figs. 4.4 ~ 4.6. Fig. 4.4 shows the structure of the cells; the rectangles represent the cell walls, and many bright spots forming a
31
circle is regarded as a chloroplast, hence those gathering spots are grana while other isolated bright spots are organelles. In some cells, only few chloroplasts were found, which suggests the organelles moved freely due to living cell. Also, it depicts the construction of chloroplasts and the different brightness of spots shows the spatial distribution of grana.
By scanning microscope objective along the z-axis from the bottom to the top within a cell, we obtained two dimensional image stack with different height and constructed three dimensional images of chloroplasts (Fig. 4.7). Three dimensional images not only reflect the depth of a cell but illustrate the arrangements of grana. With confocal light scattering images, we found that chloroplasts are oval disc-shape structure and diverse size, which have been confirmed by electron microscopy.
32
Figure 4.4 Light scattering image of plant cell of Egeria densa without zooming in.
Figure 4.5 Light scattering image of plant cell of Egeria densa with three times zoom.
33
Figure 4.6 Light scattering image of plant cell of Egeria densa with ten times zoom.
Figure 4.7 Confocal light scattering spectroscopic images of Egeria densa at different vertical
positions, from the bottom to the top of the leaf with 0.2 μm per step.
34
The average size of the grana is examined by the line profile method on the basis of the obtained 3-D light scattering images. This practical way is useful for examining the relative behavior of all variables in a multivariate data set. The line profile plot consists of a sequence of equi-spaced vertical spikes with each spike representing a different variable in the multivariate data set. In this case, each spike is corresponding to the pixel intensity of the granum in the X-Y image as shown in Fig. 4.8 (A). By fitting the light intensity profile of granum with the Gaussian fit, we can obtain the size which is derived from the FWHM, full width at half maximum, of the Gaussian curve. The grana we estimated were on the focal plane. In conclusion, the average size of the grana was approximated as 273 nm for 1000 numbers of grana. In order to determine the sensitivity and lateral resolution, we used small Au NPs. The cross section of a 40 nm Au NP allowed us to determine the lateral point spread
function (PSF) with FWHM of the PSF of light, ∆ [27]. ∆ is calculated by ∆= √∆12− ∆22, where ∆1 is the FWHM of Gaussian function which was used to fit the intensity profile from confocal image of a 40 nm diameter Au NP, and ∆2 is that of a model Gaussian function of the particle. In this experiment, ∆1 and ∆2 were 256 nm and 20 nm respectively, so ∆ can be estimated to be 255 nm, which is better than λ/2 for the excitation wavelength, about 300 nm.
By the size analysis, not only the size distribution (Fig. 4.10) but also the spatial population could be resolved. We also found that on average, there are 100 grana distributed
35
randomly in the chloroplast of E. densa. From statistical analyses, both grana of the small size and the large one (Fig. 4.9) can be found in the center or on the surface of the chloroplast. The distribution of grana randomly in the chloroplast by this optical method is consistent with the direct observation by TEM imaging.
Figure 4.8 (A) A chloroplast image where a granum examined by us is marked as the red
circle. (B) A measured scattering intensity profile (•) of (A) and its fitting curve by a Gaussian function (－). The FWHM calculated from Gaussian is used as the estimated size of granum.
Since the distance between the close vicinity grana is small, we could not fit it with wider range, and could not have a clear zero background.
Figure 4.9 The definitions of small and large size. Grana below -2S are small size and beyond
+2S are large size.
(A) (B)
36
Figure 4.10 The statistic chart depicts the size distribution estimated for around 1000 grana,
giving 273 nm as average size.
4.2-2 Confocal light scattering microspectroscopic analysis of grana in Egeria densa
The light scattering spectra show obvious Soret band located from 490~530 nm and the weak signal Q band can be found around 680 nm. Both bands are related to the absorption spectrum of chloroplasts, which contains two main bands; the Soret band is around 450 nm and the Q band is approximately at 650 nm. The difference is due to the relationship between absorption and scattering spectra. Fig. 4.11 illustrates the typical scattering spectrum of a single granum. The different size results in some side bands and relative intensity fluctuation,
200 300 400 500
0 100 200
Counts
FWHM
Statistics
N 1038
Mean 272.8 Standard deviation 41.1
37
indicating different assorted components inside the granum. By combination of light scattering images and scattering spectra to analyze the relationship between size and spectra, we found that the Soret band of the spectrum shifts to longer wavelength with the increasing size (Fig. 4.12 (B)). The range of the size starts from 200 nm to 460 nm and the Soret band of spectra shifts from 490 nm until 530 nm. Fig. 4.13 shows three types of the Soret band. When the band splits into two peaks, we considered both peaks and plotted the peak position of the Soret band against size (Fig. 4.14). It is clearly shown that the size and the Soret band wavelength presents a good correlation. The correlation implies that the red-shift is observed as the size is enlarged.
Figure 4.11 The light scattering spectrum of a granum.
38
500 550 600 650 700
335 nm
Scattering efficiency/ a.u.
Wavelength / nm
226 nm 281 nm
Figure 4.12 Light scattering image and spectra of single grana are shown. (A) Light
scattering images of chloroplasts with 10 times magnification and (B) three spectra in different colors are related to single granum in (A), which differs from granum to granum.
(A) (B) (C)
Figure 4.13 Light scattering spectra of grana. (A) The Soret band shows single peak. (B) The
side peak of Soret band is small so we assumed this kind of peak as single peak. (C) The Soret band splits. We assumed the case as split peaks.
(A) (B)
39
Figure 4.14 The plot of the peak wavelength of the Soret band against size. The split peaks
correspond to the case C in Fig. 4.13.It is clearly shown that the Soret band shifts to longer wavelength as the size of grana increase.
Whether the site of granum in the chloroplast is one of the factors determining the scattering spectrum or not was confirmed by measuring the spectra of grana located at various site. First, we checked the position of chloroplast; we found that wherever chloroplast is either located, at the center, close to the cell wall, or in the alternative space in the leaf, the spectra of grana are not distinguishable. Moreover, the distribution of grana inside the chloroplast is not important. Fig. 4.15 (A) ~ (C) illustrate the position dependence of grana. It implies that the spectra of grana are independent on the position, although grana at the margin and at the center of chloroplast still show size dependence.
480
200 250 300 350 400 450
w a v el eng th o f So ret ba nd / nm
size / nm
single peak split peak
40
Figure 4.15 Light scattering image and spectra of grana are shown. (A) By yellow circle, we
roughly defined the border of center and surface of the chloroplast. (B) The spectra of grana in the center of the chloroplast. (C) The spectra of grana at the surface of the chloroplast.
Figure 4.16 Polarizer is set in front of the confocal microscope. Each time we started from
same angle as 0 degree; the polarization angle is relative.
(A)
41
The dependence of scattering spectrum of an individual granum on polarization of the probe light gives important information on the optical anisotropy of grana in the chloroplast.
Such a method has been employed, for instance, in the study of optical anisotropy of chlorophyll molecules within in vivo chloroplasts [28]. Differential polarization imaging and circular dichroism provide information on long-range chiral organization of the pigment-protein complexes in mature granal chloroplasts [29]. In this study, the polarization measurement was operated by setting a polarizer in front of the microscope (Fig. 4.16). In other words, we expect that the polarized light interacts with molecules directly, and then the scattered light was transformed into signal in this measurement. Circular polarized supercontinuum passed through the polarizer and became a linear polarized input. The polarized angle is relative with respect to a certain angle in laboratory axis. To ensure the reliability of the polarization measurement of grana, at first we measured the scattering spectrum of 200 nm Au NP as a standard with essentially the same optical probe light and experimental setup.
The Au NPs are optically isotropic and exhibit high scattering efficiency in visible region due to the resonant plasma oscillation of conduction electrons. To minimize the contribution of the index-mismatched reflection, Au NPs are casted on the glass substrate and dipped in the immersion oil (n = 1.5) [30]. To analyze the polarization measurement data, we
42
plotted the scattering efficiency as a function of polarization angle (Fig. 4.17). The scattering efficiency was chosen from the fixed wavelength which was the Soret band of the unpolarized spectrum. Because Au NP is optical isotropic, polarization dependence of Au NP shows in Fig. 4.17 can be attributed to instrumental function. If we use depolarizer instead of instrumental function, we will not have any information on anisotropy of the grana or molecular alignment in the grana. By this instrumental function, we can therefore correct the raw data of polarization dependent measurement of grana.
Fig 4.18 showed the spectra of single granum at polarization angles of 45, 90 and 135 degree. The spectra changed dramatically, especially at the Soret band. With the same analytical procedure with the case of Au NP, we found it hard to determine at which wavelength the scattering efficiency of granum should be analyzed, because the spectral shape also changed with polarization angle. As a consequence, we selected the wavelength from the unpolarized spectrum at which the scattering efficiency is maximum. After being corrected by instrumental function, we obtained the relative scattering intensity as a function of polarization angle, as shown in Fig. 4.19 (A). The polarization measurements of grana exhibit a band at 135 degree and a valley at 45 degree; that is to say that the Soret band appears and disappears depending on the angle of incident light with 90° symmetry, indicating optical anisotropy of grana. In Fig. 4.19 (B), we used polar coordinate system to depict the anisotropy
43
of grana efficiently. The solid and dotted lines represent experimental and expectation result, respectively, since the motion of living grana made measurement difficult.
Further analyses are expressed in Fig. 4.20 (A) and (B). For easier comparison, the maximum and minimum values of the polarization plot are defined as b and a, and the ratio of b and a (R), calculated as a/b, represents the change of light interacted by granum.
Fig. 4.20 (A) illustrates the parameter R as a function of the polarization angle at the peak wavelength of the Soret band, which shows clear correlation.
The FWHM which also can stand for the effect of light interacted by pigment molecules is used as a parameter of polarization dependence. Fig. 4.20 (B), which plots the FWHM of the primary band from polarization plot versus the polarization angle, trends down with correlation coefficient r = -0.702. The detail will be considered in discussion.
44
0 20 40 60 80 100 120 140 160 180 200 2
4 6 8
scattering efficiency / abr. unit
polarization / degree
200 nm Au
Figure 4.17 The polarization measurement of 200 nm Au NP. We measured the scattering
efficiency of the resonance band at the wavelength giving its maximum intensity for each 15 degree. The dependence can be used as an instrumental function because Au NP is optically isotropic.
Fig. 4.18 Scattering spectra of granum without polarization and with 45, 90 and 135 degree.
The spectrum changed as the polarization changed, especially at the Soret band.
500 600 700
scattering efficiency / a.u.
wavelength / nm
45 degree 90 degree 135 degree without polarizer
45
from the scattering efficiency of each spectrum at 533 nm which had the maximum efficiency of the Soret band. The polarization plot then is corrected by instrumental function, and finally the polarization dependence was plotted. The polarization dependence indicated granum is optically anisotropic. (B) Polarization dependence expressed in polar coordinate system. The solid line is the experimental result and the dotted line is the expectation. The symmetry was expected since we used linear polarization.
(B) (A)
46
Fig. 4.20 The analyses of polarization dependence. (A) R is defined by the ratio of the
minimum and maximum the values of the polarization dependence. The inset illustrated the parameters of R. It is clear that R decreases as size of granum enlarges. (B) The y-axis stood for the FWHM shown in the inset. FWHM was calculated by Gaussian fit. In general, FWHM and the size represent negative correlation.
(A)
(B)
47
Chlorophyll fluorescence analysis has become one of the most powerful and widely used techniques, which is available to plant physiologists and ecophysiologists [31]. In this study, we also apply fluorescence measurement to living grana. Sample was directly excited by 488 nm Ar-ion laser. This means that the shoulder of Soret band is excited. Although the total fluorescence yield of chlorophyll is very small (only 1 or 2% of total light absorbed), measurement is quite easy because of strong fluorescence intensity due to high fluorescence cross section [31].Fig. 4.21 (A) and (B) are the fluorescence images with one and ten times zooming respectively. Compared to Fig. 4.4, only chloroplasts were observed and their cell wall was not depicted directly, but it could be distinguished from the distribution of chloroplast. In addition, the 3-D spatial distribution can be differentiated from the brightness of chloroplast. By contrast with Fig. 4.6, the fluorescence image showed lower resolution, only few and unclear grana could be recognized. In contrast to 90 grana observed on average in one chloroplast from the CLSM image, less than 20 grana are clearly observed in the fluorescence image. On the other hand, the granal size of fluorescence image is around 600 nm which is almost 2 to 3 times larger than that of CLSM image. It implies that a granum in the fluorescence image may contain 2 to 3 grana. These demonstrated that CLSM is better than fluorescence measurement in living cell.
48
Chlorophyll fluorescence emission spectra were measured from various sites of the imaged chloroplasts. Results are shown in Fig. 4.21 (C) and (D), with the spectra normalized at 684 nm. Spectra a and b show grana in the same chloroplast, spectra a, c, d and e illustrate the spectral shapes of grana from different chloroplasts. Obviously, the discrepancy among them is quite small. In other words, the fluorescence spectrum is independent on granum size and position in the chloroplast. The results suggest that (i) the grana contain different kinds of molecules with various relative intermolecular orientations and many of them play as a fluorescence quencher with different quenching efficiency, and (ii) some grana may undergo non-radiative decay. Hence, the signal-to-noise ratio of the light scattering is better than that of the fluorescence, so that the spatial resolution of light scattering is apparently better compared to the fluorescence imaging. The consequences indicate that, as compared to fluorescence measurement, light scattering measurement is more suitable for in vivo grana to obtain information about the size and anisotropic structure.
49 a
b c
e d
(A)
(B)
50
620 640 660 680 700 720 740
0.0
620 640 660 680 700 720 740
0.0
Figure 4.21 Fluorescence images and spectra are shown. (A) The fluorescence image of plant
cell without magnification. Different brightness depicts the distance from the focal plane. (B) The fluorescence image of plant cell with 10 times magnification. Only some grana are recognized. (C) The fluorescence spectra of two grana in the same chloroplast. (D) The fluorescence spectra of grana in different chloroplast.
(C)
(D)
51
4.3 Discussion
Fig. 4.22 The analysis of size dependence.
To interpret the granum size-dependence of the Soret band of the spectrum, first we should consider what important factors determine the Soret band. Size dependence of electronic spectra could be discussed based upon optical property. Here we list three possibilities that might be the effective factors of size dependence; chemical composition, optical effect and molecular arrangement (Fig. 4.22). The further detail of each element
will be discussed in the following sections, which includes the computational calculation and experimental evidences.
size dependence of optical properties
optical effect
Reabsorption
by itself
by other particles calculatoin
chemical composition
chemical composition