Ungerminated Rice Grains Observed by Femtosecond Pulse Laser Second-Harmonic Generation Microscopy
Yue Zhao, Shogo Takahashi, Yanrong Li, K. T. T. Hien, Akira Matsubara, Goro Mizutani, Yasunori Nakamura
aa r X i v : . [ phy s i c s . b i o - ph ] A ug Ungerminated Rice Grains Observed by Femtosecond Pulse LaserSecond-harmonic Generation Microscopy ⊲ Yue Zhao ⊲ Shogo Takahashi ⊲ Yanrong Li ⊲ K. T. T. Hien ⊲ Akira Matsubara ⊲ Goro Mizutani ⋆ ⊲ Yasunori Nakamura School of Materials Science, Japan Advanced Institute of Science and Technology, Asahidai 1-1 Nomi, 923-1292, Japan Faculty of Bioresource Sciences, Akita Prefectural University, Akita City, Akita 010-0195, Japan
J. Phys. Chem. B , As a demonstration that second-order nonlinearoptical microscopy is a powerful tool for ricegrain science, we observed second-harmonicgeneration (SHG) images of amylose-free glutinous rice and amylose-containingnonglutinous rice grains. The images obtainedfrom SHG microscopy and photographs of theiodine-stained starch granules indicate that thedistribution of starch types in the embryo-facingendosperm region (EFR) depends on the type ofrice and suggests that glucose, maltose, or bothare localized on the testa side of the embryo. Inthe testa side of the embryo, crystallized glucoseor maltose are judged to be detected by SHG.These monosaccharides and disaccharides playan important role, as they trigger energy inthe initial stage of germination. These resultsconfirm SHG microscopy is a good method tomonitor the distribution of such sugars andamylopectin in the embryo and its neighboringregions of rice grains.1 Introduction
Rice is an important food. The endosperm of riceis mostly composed of starch, which is an energysource for most organisms. A previous study using sum ⋆ Goro Mizutani: [email protected] frequency generation (SFG) microscopy reported thestarch distribution in the cross section of a glutinous ricegrain. The SFG signal is enhanced in the crush cell layer,which is an ∼ μ m wide zone between the endospermand the embryo [1].Cells in the crush cell layer and the size of thestarch granules are smaller than those in the center ofthe endosperm [2]. In the nuclear stage of endospermformation, the endosperm nucleus grows remarkably inthe vicinity of the embryo [3, 4]. The embryo and theendosperm must interact and communicate at this stagebecause angiosperm seeds of overlapping fertilizationensure cooperative growth [5]. The embryo-facingendosperm region (EFR) develops differently from otherendosperm tissues via programmed cell death [6].Endosperm cells closest to the embryo do not accumulatestarch granules. The next-nearest neighbor endospermcells have special slender shapes, and the stored starchgranules are small sized [6]. Previous research hassuggested that the construction of EFR and the sizeratio of the embryo and endosperm are closely related [6].Therefore, morphological analysis in EFR is importantfor breeding science. On the other hand, during thestage of water absorption of germination, the crush celllayer acts as a water reservoir and absorbs water first[7]. Then, starch is digested from the EFR near thescutellum as the endosperm develops [8].However, the origin of the SFG enhancement inthe crush cell layer [1] remains unclear. Two originshave been proposed for the enhanced large second-ordernonlinear effect. One is the high density of amylopectinnear the crush cell layer. The other is a high degree oforientation of amylopectin near the crush cell layer. Toidentify the origin, herein we observe the crush cell layersof two kinds of rice grains using SHG microscopy andmap the broken symmetry of the molecular structure.Optical second harmonic generation (SHG) occursin non-centrosymmetric media within an electric dipoleapproximation, and its frequency is double that ofthe incident one. Hence, optical SHG microscopycan map asymmetrically oriented structures. Themonosaccharide α -D-glucopyranose is a constituent unit Yue Zhao et al. ♦ of saccharide chains in starch. Since its structure isnon-centrosymmetric, it displays a nonzero second-order nonlinear susceptibility and SHG activity. Theanisotropic hydroxide and hydrogen bonds of alignedwater have been reported as the origin of SHG in glucoseand amylopectin hydrates [9]. Additionally, D-glucose,D-fructose, and sucrose have been reported to show sumfrequency generation (SFG) signals [10].Starch, which is comprised of various componentssuch as amylopectin and amylose, exhibits SHG activity[11, 12]. The origin of SHG in dry starch is amylopectin[13–15]. Amylopectin has a tandem cluster structurewhere the sugar chains on small branches have a doublehelix structure [16]. It is hardly soluble in water.Since its higher-order cluster structure has a macro-scopic asymmetry, the large nonlinear polarizationin monosaccharide α -D-glucopyranose generates strongSHG light. In contrast, the amylose structure isspeculated to be mostly amorphous in the starchgranule of rice with a single helical structure [17–21]. The nonlinear polarization of the monosaccharide α -D-glucopyranose is thought to be canceled by theglucopyranose unit in the single helical structure. Thus,dry amorphous amylose does not display macroscopicnonlinear polarization, and has a very weak SHG [15,22].On the other hand, amylose can be crystallized in bothsingle (V) and double (A and B) crystal forms [23, 24].The ordered hydrogen and hydroxyl bond networks inthe hydrates of these crystals are reported to allow SHG[25].Generally, a starch iodine reaction is an easyand useful method to observe the starch distribution.However, the sample may permeate, resulting in anirreversible change. Although the method of F. Vilaplana et al. can quantify amylopectin and amylose in starch[26], it cannot map their densities.Second-order nonlinear optical microscopy isa non-destructive method. Amylopectin in starchgenerates an SHG signal due to macroscopicasymmetry [11,13,14], but amorphous amylose does notshow SHG [15,22]. Thus, the distribution of amylopectinin the endosperm of amylose-free glutinous-type ricecultivar can be examined using SHG microscopy. Inthe initial stage of germination of a rice seed, themorphological structure of the crush cell layer, whichis a special structure between the embryo and theendosperm that serves as a water reservoir, and thedistribution of monosaccharides and disaccharides,which trigger energy in the embryo, have yet to beaccurately examined.Herein, SHG images of a typical nonglutinousrice cultivar Koshihikari, which includes bothamylose and amylopectin, and a typical glutinousrice cultivar Shintaishomochi, which includes onlyamylopectin, are obtained using SHG microscopy witha femtosecond pulse laser and a morphological analysis.The comparison of the two cultivars confirms that amylose does not affect the experimental results. To ourknowledge, previous research has not directly comparedthese two cultivars, demonstrating that SHG microscopyis a powerful tool to evaluate the distribution of sugarsand amylopectin in the embryo and its neighboringregions of rice grains. In the experimental setup (see Fig. 1), the lightsource was a titanium sapphire femtosecond pulsedlaser (Spectra-Physics: Tsunami) with a regenerativeamplifier (Spectra-Physics: Spitfire) with a repetitionfrequency of 1 kHz, pulse width of 120 fs, and wavelengthof 800 nm. The beam directly irradiated the samplewithout a focusing lens. The irradiated area on thesample stage with the sample was about 170 mm . Theincident light energy density per unit area on the samplewas 0.4 nJ/ μ m , and the incident angle with respectto the normal to the sample stage was 60 ◦ . A CMOScamera (Lumenera corporation, Lu135M) was usedto monitor the sample surface using an incandescentlamp. The images of the scattered SHG light wereobserved by microscope optics (OLYMPUS BX60) withan image intensified charge coupled device (II-CCD)camera (HAMAMATSU, PMA-100). The power of theexcitation light on the sample surface was controlled byusing ND (Neutral Density) filters. Fig. 1
Experimental setup for SHG microscopic observationusing an II-CCD camera, femtosecond Ti:sapphire pulselaser, and regenerative amplifier. Details are described in themain text. M1, M2: Mirror. L.P.F.: Long wavelength passfilter. S.P.F.: short wavelength filter. B.P.F.: bandpass filter.
The sample was placed on a silicon wafer substrate.The SHG intensity of the oxidized Si surface is veryweak compared to the SHG of bulk amylopectin riceseeds, and thus, it is negligible. Before the sample, a longwavelength pass filter (L.P.F.) was placed to block lightwith wavelengths shorter than 780 nm while allowingthe 800 nm wavelength beam to pass. The scatteredSHG light from the sample passed through an objectivelens with a × ngerminated Rice Grains Observed by Femtosecond Pulse Laser Second-harmonic Generation Microscopy 3 filter (S.P.F.) with a transmission wavelength range of350785 nm. This short pass filter blocked the 800 nmwavelength light. Finally, the SHG signal was selectedby Semrocks (FF01-395/ 11) band-pass filter (B.P.F.)with a center wavelength of 395 nm (the transmittanceat 400 nm wavelength is over 90%), and two-photonexcitation fluorescence (2PEF) was selected by Semrocks(FF02-438/24) band-pass filter with a center wavelengthof 438 nm. Both the S.P.F. and B.P.F. were tilted 5 ◦ withrespect to the wavefront of the parallel ray to eliminate“ghost”signals. The spatial resolution of the microscopewas determined by the chip size of the II-CCD camera(11 μ m × μ m). In this case, the resolution was 2.6 μ mfor a 5 × magnification. The imaging integration timewas 60 s. The polarization of the incident light wasparallel to the horizontal direction of the microscopicimages of Figs. 2(e), 3(e), 4(c) and 5(c). The polarizationof the detected SHG was not chosen.The SHG signal was identified from the differencein the intensity distributions between 400 and 438 nm.There are two possible origins for the 400 nm signalin an image with a 400 nm wavelength bandpass filter.One is an SHG signal. The other is multiphoton excitedfluorescence. In the latter case, the sample is excited bya multiphoton transition and emits luminescence pulsesat a photon energy lower than the multiphoton energy.If the signal is dominated by the latter, observationwavelengths of 400 and 438 nm should provide similarresults, and the images for both B.P.F.s should besimilar, because the spectra of multiphoton-excitedfluorescence should be continuous near the two-photonwavelength [27]. If the spatial distributions of the signalsclearly differ, then the signal from the 400 nm B.P.F. isSHG, whereas that for the 438 nm B.P.F. is 2PEF.In this study, the maximum incident light energydensity per unit area on the sample was 0.4 nJ/ μ m .No damage was observed at this incident light energydensity. When observing starch by a scanning type SHGmicroscopy [9, 25], the excitation beam was focusedon the sample via an objective lens. In this case, theirradiation area should be 0.23 μ m for a wavelength of800 nm and a numerical aperture (NA) of the focusinglens of 0.75. When the energy of one pulse was 2 nJ,the incident light energy density per unit area was8.7 nJ/ μ m [9, 25]. This comparison shows that thecurrent method is much less likely to damage the samplethan typical scanning SHG microscopy. The nonscanningtype setup can give a wider field of view than thescanning type SHG microscope, so it is convenient forsamples with a large size such as the rice seeds. OurSHG microscope can acquire images in a short timesuch as several seconds to several minutes. This isbecause the excitation light energy density at the samplesurface is made comparable to that of the scanningtype microscope by using the regenerative amplifier. Thescanning type laser microscope can give a diffraction limit spatial resolution, and it is better than the typicalresolution of our type. First, the grain cross section of ungerminated brownrice was observed. Figures 2 and 3 show grains fromamylose-free glutinous rice cv. Shintaishomochi andamylose-containing nonglutinous rice cv. Koshihikari,respectively. The photographs before and after thestarch-iodine staining of the cross section of glutinousrice are shown in Fig. 2(b) and (d), while parts (b)and (d) of Fig. 3 show those of nonglutinous rice,respectively. Figures 2(a), 2(c), 3(a) and 3(c) are theirexpanded images taken by the CMOS camera. The SHGimages of the corresponding samples are shown in Figs.2(e) and 3(e). The SHG image intensity is pseudocoloredby the intensity scale bar on the right and superimposedon the images in Figs. 2(c) and 3(c). Because theiodine solution does not affect the SHG intensity underthe current conditions [11], the SHG images after theaddition of iodine solution are shown.Cisek et al. [25] reported that a hydration treatmentalters the SHG in glucose, amylose, and starch. However,we found that the water absorption has a negligibleimpact on the relative SHG intensity in the images.Kong et al. [28] reported that amylose in the presence ofsmall guest molecules of iodine forms 6-fold left-handedsingle helices packed in an antiparallel arrangementand shows SHG activity. They oriented the amylosemolecules by pulling the amylose film before and duringexposure of the film to iodine gas [29]. This study doesnot apply a mechanical force on the sample. Thus, theamylose remains amorphous in the rice grains, and thecontribution of amylose to the SHG response is regardedas negligible. The signals in Figs. 2(e) and 3(e) are SHGbecause the images in Fig. 3(e) and (g) differ from eachother.The images in Figs. 2(c) and 3(c) show the two kindsof rice cross sections expanded by water absorption inthe starch iodine test. These images slightly differ fromthose in Figs. 2(a) and 3(a). First, the iodine reagent,which reacts more strongly to amylose than amylopectin,colors the endosperm of both the glutinous rice and thenonglutinous rice (Figs. 2(d) and 3(d)). Figures 2(c) and3(c) are black and white images, but the yellow hollowarrows denote the colored parts. In the glutinous ricegrains (
Oryzaglutinosa cv. Shintaishomochi) in Fig. 2(d),the color of the endosperm near the crush cell layer isweaker, suggesting that the crush cell layer either lacksor has a small amount of amylopectin. On the otherhand, in the nonglutinous rice grains (
Oryzaglutinosa cv.Koshihikari) in Fig. 3(d), the coloring of the endospermnear the crush cell layer is stronger than the other part,indicating that amylose is present at higher densities inthat region.
Yue Zhao et al. ♦ Fig. 2
Ungerminated glutinous rice (
Oryzaglutinosa cv. Shintaishomochi) grain images. (a) Microscopic image of the glutinousbrown rice cross section and (b) a photograph of the whole grain cross section. (c) Microscopic image and (d) macroscopicphotograph after starch iodine reaction for 1 h. (e) SHG image of the same grain as in part c. (f) SHG intensity distributionwithin the yellow frame in part e. (g) 2PEF image of the same grain as in part c. The cyan frames in photographs b and dshow the field of view of the microscope images of parts a, c, e, and g. Linear images a and c are illuminated by white light.SHG (e) and 2PEF (g) images are overlaid on the corresponding linear microscopic images. It is pseudocolored by the intensityscale bar at the right. In part f, each intensity is the summation of the corresponding vertical pixel SHG intensities. Note thestrong SHG spots in the embryo near the embryo testa and the enhanced SHG at the crush cell layer.
Fig. 3
Ungerminated nonglutinous rice (
Oryzaglutinosa cv. Koshihikari) grain images. (a) Microscopic image of thenonglutinous brown rice cross section and (b) a photograph of the whole grain cross section. (c) Microscopic image and(d) macroscopic photograph of 5 min starch iodine reaction. (e) SHG image of the same grain as in part c. (f) SHG intensitydistribution within the yellow frame in part e. (g) 2PEF image of the same grain as in part c. The cyan frames in photographsb and d show the field of view of the microscope images of parts a, c, e, and g. Linear images a and c are illuminated by whitelight. SHG (e) and 2PEF (g) images are overlaid on the corresponding linear microscopic images. It is pseudocolored by theintensity scale bar at the right. In part f, each intensity is the summation of the corresponding vertical pixel SHG intensities.Note the strong SHG spots in the embryo near the embryo testa and a gradual change of the SHG intensity at the crush celllayer.
The SHG signals are observed from the endospermsof both the glutinous and the nonglutinous rice (Figs.2(e) and 3(e)). In the crush cell layer, the SHG signal isenhanced in the glutinous rice (Figs. 2(e, f)), but not inthe nonglutinous rice (Figs. 3(e, f)).SHG signals along the testa edge of the embryoexhibit a very strong intensity (Figs. 2(e) and 3(e),empty red arrow heads). To further investigate the strong SHG signals from the testa edge of the embryo inFigs. 2(e) and 3(e), the outer walls on the testa side ofthe embryo of the two brown rice grains were observedfrom the outer side (Figs. 4 and 5). Very strong SHGspots are seen along the hypocotyl in Figs. 4(c) and5(c). A 2PEF signal is observed (Figs. 4(d) and 5(d)),but its distribution differs from that of SHG (Figs. 4(c) ngerminated Rice Grains Observed by Femtosecond Pulse Laser Second-harmonic Generation Microscopy 5
Fig. 4
Images of ungerminated glutinous rice (
Oryzaglutinosa cv. Shintaishomochi) grains. (a) Macroscopic photograph fromthe embryo side. (b) Linear microscopic image. (c) SHG image. (d) 2PEF image. Linear images are obtained by illuminatingthe sample by white light. SHG (c) and 2PEF (d) images are overlaid on the respective linear microscopic images. The SHGimage is pseudocolored according to the intensity scale bar at the bottom.
Fig. 5
Images of ungerminated nonglutinous brown rice (
Oryzaglutinosa cv. Koshihikari) grains. (a) Macroscopic photographfrom the embryo side. (b) Linear microscopic image. (c) SHG image. (d) 2PEF image. Linear images are obtained byilluminating the sample by white light. SHG (c) and 2PEF (d) images are overlaid on the respective linear microscopicimages. The SHG image is pseudocolored according to the intensity scale bar at the bottom.
Table 1
Results of SHG Microscopy of the Dried Residuesample not washed water washed DMF washed ethanol washedSHG active inactive inactive active and 5(c)). The outer walls of the endosperms displayvery weak SHG signals but not 2PEF signals (data notshown).To assign the origin of the strong SHG spotat the end of the embryo, we examined the SHGresponse of possible SHG active substances in theembryo. Commercially available powders of threesubstances (i.e., α -amylase (Kishida Chemical Co., Ltd.,product code: 260-04412), glucose (Wako Pure ChemicalIndustries, Ltd., product code: 049-00591), and maltose(ChromaDex, Inc., product code: ASB-00013055-001)crystalline powders) as delivered were observed by ourSHG microscope. All of these are SHG active (data notshown).We also examined the SHG response of testa powdertaken from the rice grains by a commercial “rice huller”,which removes only the testa and a part of the embryo.The starch iodine reaction confirms that the peeledpowder lacks starch. However, the peeled powder showsan SHG response (data not shown). We put three dosesof the peeled powder of 2.5 mg each into 50 mL of water, N, N -dimethylformamide (DMF), and ethanol, stirredthem well for 5 min, and filtered them. Three residues onthe filter paper were collected and dried for 24 h. Table1 shows the SHG responses of the dried residues. The SHG activity of the testa powder is lost after washingwith water and DMF but remains after washing withethanol.
A previous study on the cross section of a glutinous ricegrain around the crush cell layer between the endospermand the embryo reported an enhanced SFG signal atwavenumbers of 2920 and 2970 cm − of the IR light[1]. This IR frequency is the resonant C-H stretchingvibration region of the amylopectin molecules. A similarenhanced SHG in the glutinous rice occurs around thecrush cell layer in this study (Figs. 2(e, f)). Additionally,the SHG intensity distribution profile of Fig. 2(f) issimilar to the SFG profile of the previous report [1].There are two possibilities for the enhancedsecond-order nonlinear effect. One is the high density ofamylopectin near the crush cell layer. The other is thehighly oriented amylopectin near the crush cell layer.The iodine starch reaction (Fig. 2(c)) indicates that thedensity of amylopectin is rather constant in this region.Since this study reveals that both SFG and SHG are Yue Zhao et al. ♦ Table 2
Solubilities, Residue Components, and SHG Activities of the Residues for the Three Candidateswater DMF ethanol(i) α -amylase soluble insoluble insoluble(ii) glucose, maltose soluble soluble insoluble(iii) leukoplasts insoluble unknown soluble (only lipids)possible residue (iii) (i), (iii) (i), (ii), (iii) [without lipids]SHG of residue in Table 1 no no yes strong near the crush cell layer, we speculate that theorigin of the enhanced SFG signal in the crush cell layerof the glutinous rice [1] is the high degree of asymmetryin the amylopectin chain structure.On the other hand, the nonglutinous rice does notshow SHG near the crush cell layer (Figs. 3(e, f)),suggesting that crush cell layer of the nonglutinous ricehas a lower concentration of amylopectin. The starch inthe glutinous rice is composed mostly of amylopectin,whereas the nonglutinous rice contains amylose andamylopectin. The starch iodine reaction is rather strongnear the crush cell layer of the nonglutinous rice (Fig.3(d)). Therefore, one possibility is that the endospermnear the crush cell layer of the nonglutinous rice has ahigher concentration of amylose. Amylose has a betterwater absorption than insoluble amylopectin. Becausethe crush cell layer serves as a water reservoir ingermination, the amylose content may be high in thenonglutinous rice. This hypothesis is consistent with theweak SHG in the crush cell layer of the nonglutinous rice.These results suggest that the distribution of starch inthe vicinity of the crush cell layer depends on the typeof rice.In fact, we measured the vibrational spectrum of across section of a rice grain with the Fourier transforminfrared spectroscopy (FT-IR), but the factor analysisof the infrared spectra of amylose, amylopectin, and ricestarch was unsuccessful. FT-IR is a general method todistinguish between molecular species, but it is unsuitedto identify starch types [30]. The molecular units ofthe polysaccharides in starch and cellulose are almostthe same. It has also been reported that the infraredspectroscopy of amylose, amylopectin, and rice starchhave similar factor analysis patterns [31]. On the otherhand, it has been reported that amylose gives a Ramanband at 1657 cm − while amylopectin does not [32].Therefore, Raman spectroscopy and microscopy can beanother way to map amylose in the rice grain. Thecomparison of the performances of SHG and Ramanspectroscopy is our future target.2PEF images of chromophores in a young leaf tissuehave been reported [33,34] and the origin of the resonant2PEF was assigned to proplastid. However, it is difficultto judge that the observed SHG in our study is dueto proplastid because its resonant wavelength does notmatch our light source. The resonance bands of colorlesssaccharides in the visible region are quite unlikely, andthe origin of the 2PEF in this study is not clear. We considered three possible origins of the SHG spotsin Figs. 2(e), 3(e), 4(c) and 5(c) at the ends of theembryos: (i) amylase, (ii) crystals of glucose, maltose,or both, and (iii) leucoplasts. (i) Amylase may be storedin the aleurone layer of ungerminated rice grains [35]. α -Amylase reagent has a detectable SHG signal. Because α -amylase forms a single crystal in the P2 spacegroup [36] with a broken inversion symmetry, it shouldhave a nonzero second-order nonlinear susceptibility.(ii) Glucose and maltose are believed to be present inungerminated grain embryos [7, 37]. Detection of theSFG signal in the crystalline powder of glucose has beenreported [10]. We have confirmed the SHG responsefrom the crystalline powder of glucose and maltose. (iii)Leucoplast is a generic term for amyloplast, elaioplast,and proteinoplast [38]. Since amyloplast, which cangenerate SHG light, is generally contained in the rootcell of the plant [39], it may also be contained in thehypocotyl and the radicle in the grain in question.Elaioplast is an organelle that stores lipids. Some lipidshave optical nonlinearity [38]. Similarly, proteinoplast isan organelle. It is only found in plant cells, especially inroots and seeds. It contains some protein crystals [40,41].In a protein crystal, the protein molecules may showa strong second-order nonlinear response due to theirwell-oriented structure [42].The SHG observations in Table 1 are used to discussthe feasibility of the three candidates. Table 2 liststhe solubility of the three candidates in the solvent.After washing with water, only (iii) leucoplast remainsin the residue. The SHG signal is not observed inthe water washed residue. Thus, candidate (iii) canbe excluded as the origin of the strong SHG spots inembryos near embryo testa in Figs. 2(e), 3(e), 4(c) and5(c). After washing with DMF, (i) α -amylase remainsin the residue and the DMF-washed testa does notexhibit an SHG signal. Hence, candidate (i) can also beexcluded. Although the solubility of leucoplast in DMF isunknown, it is excluded above. Therefore, the remainingcandidate (ii) of glucose or maltose is considered to bethe origin of the strong SHG spots in the embryo.When the rice seed absorbs water duringgermination, the water passes through the seedcoat covering the boundary between the embryo andendosperm (Figs. 4(a) and 5(a), near the black arrows)to reach the crush cell layer. The water subsequently ngerminated Rice Grains Observed by Femtosecond Pulse Laser Second-harmonic Generation Microscopy 7 passes through the absorption cell layer of the scutellumand moves to embryonic tissue [7]. The embryo breathesby consuming its carbohydrate but not its starch [7, 37].Since the hypocotyl needs energy for growth, glucoseand maltose must move to it to be consumed. In fact,since seeds sprout when they are covered by water inthe ripening season, it is highly likely that glucose andmaltose will move even in ripened seeds.Finally, although our measurement indicates that thereagent of α -amylase is SHG active, it is assumed thatthe amount of α -amylase present in the ungerminatedrice grains is too small for SHG detection because α -amylase in rice is synthesized only during the seedgermination process [43]. The SHG images are obtained from ungerminatedglutinous rice (
Oryzaglutinosa cv. Shintaishomochi)and nonglutinous rice (
Oryzaglutinosa cv. Koshihikari).The observed enhancement effect of the SHG in thecrush cell layer of the glutinous rice is consistentwith the enhancement effect of SFG in the previousstudy [1]. The origin of the enhancement effect isbroken centrosymmetry, not the resonance of molecularvibration. On the other hand, the enhancement effectof SHG is not observed near the crush cell layer of thenonglutinous rice. Considering the results of the starchiodine reaction, amylopectin in the crush cell layer inthe glutinous rice grains may have a highly orientedstructure, while the endosperm near the crush cell layerof the nonglutinous rice may have a higher concentrationof amylose. Hence, the type and distribution of starch inthe vicinity of the crush cell layer may depend on thetype of rice. Additionally, some strong SHG spots areobserved along the testa side on the embryo. However,the testa side lacks starch. Thus, the origin of these SHGspots is likely glucose, maltose, or both. In the testa sideof the embryo, crystallized glucose or maltose is detectedby SHG.This study examined typical rice species. Asthe results demonstrate that SHG can monitor thedistribution of sugars and amylopectin in the embryoand neighboring regions of rice grains, we planto investigate more varieties to verify whether thetendencies observed here apply to rice in general.
Notes
The authors declare no competing financial interest.
References
1. H. Li, Y. Miyauchi, N. A. Tuan, G. Mizutani, andM. Koyano, “Optical sum frequency generation image of rice grains”,
J. Biomater. Nanobiotech. , 286–291(2012).2. K. Hoshikawa, “Development of starch parenchymatoustissue”, in “T. Matsuo, et al., Eds., Science of the RicePlant,” , vol. I, Morphology (Rural Culture Association,Tokyo, 1990), pp. 295–302. [in Japanese].3. K. Hoshigawa, “Studies on the development ofendosperm in rice : 1. Process of endosperm tissueformation [in Japanese]”, Nippon Sakumotsu Gakkai Kiji , 151–161 (1967).4. K. Hoshigawa, “Structure of the endosperm tissue of riceand its development [in Japanese]”, Biol. Sci. , 66–76(1972).5. C. Lafon-Placette and C. K¨ohler, “Embryo andendosperm, partners in seed development”, Curr. Opin.Plant Biol. , 64–69 (2014).6. K. Hibara, “Molecular genetics of the events of earlydifferentiation in rice [in Japanese]”, Breed. Res. ,115–119 (2015).7. N. Takahashi, “Physicogenetical Studies on Germinationof Rice Seeds with Special Reference to its GeneticalFactors [in Japanese]”, Bull. Inst. Agr. Res. TohokuUniv. , 1–87 (1962).8. K. Hoshikawa, “Digestion of the endosperm”, in “T.Matsuo, et al., Eds., Science of the Rice Plant,” , vol. I,Morphology (Rural Culture Association, Tokyo, 1990),p. 75. [in Japanese].9. R. Cisek, D. Tokarz, S. Krouglov, M. Steup, M. J.Emes, I. J. Tetlow, and V. Barzda, “Second harmonicgeneration mediated by aligned water in starchgranules”, J. Phys. Chem. B , 14785–14794 (2014).10. H. C. Hieu, H. Li, Y. Miyauchi, G. Mizutani,N. Fujita, and Y. Nakamura, “Wetting effect onoptical sum frequency generation (SFG) spectra ofD-glucose, D-fructose, and sucrose”,
Spectrochim. ActaMol. Biomol. Spectrosc. , 834–839 (2015).11. G. Mizutani, Y. Sonoda, H. Sano, M. Sakamoto,T. Takahashi, and S. Ushioda, “Detection of starchgranules in a living plant by optical second harmonicmicroscopy”,
J. Lumin. , 824–826 (2000).12. G. Cox, N. Moreno, and J. Feij´o, “Second-harmonicimaging of plant polysaccharides”,
J. Biomed. Opt. ,24013–240136 (2005).13. S. Psilodimitrakopoulos, I. Amat-Roldan,P. Loza-Alvarez, and D. Artigas, “Estimating thehelical pitch angle of amylopectin in starch usingpolarization second harmonic generation microscopy”, J. Opt. , 084007 (2010).14. Z. Y. Zhuo, C. S. Liao, C. H. Huang, J. Y.Yu, Y. Y. Tzeng, W. Lo, C. Y. Dong, H. C.Chui, Y. C. Huang, H. M. Lai, and S. W. Chu,“Second harmonic generation imagingA new method forunraveling molecular information of starch”, J. Struct.Biol. , 88–94 (2010).15. R. Cisek, D. Tokarz, L. Kontenis, V. Barzda,and M. Steup, “Polarimetric Second HarmonicGeneration Microscopy: An Analytical Tool for StarchBioengineering”,
Starch-St¨arke , 1700031 (2018).16. K. Kainuma and D. French, “Naegeli amylodextrin andits relationship to starch granule structure. II. Role ofwater in crystallization of B-starch”, Biopolymers ,2241–2250 (1972). Yue Zhao et al. ♦
17. Y. Takeda, S. Hizukuri, and B. O. Juliano, “Purificationand structure of amylose from rice starch”,
Carbohydr.Res. , 299–308 (1986).18. S. Hizukuri, Y. Takeda, N. Maruta, and B. O. Juliano,“Molecular structures of rice starch”,
Carbohydr. Res. , 227–235 (1989).19. D. J. Manners, “Recent developments in ourunderstanding of amylopectin structure”,
Carbohydr.Polym. , 87–112 (1989).20. A. Imberty, A. Bul´eon, V. Tran, and S. P´erez,“Recent advances in knowledge of starch structure”, Starch-St¨arke , 375–384 (1991).21. T. A. Waigh, I. Hopkinson, A. M. Donald, M. F.Butler, F. Heidelbach, and C. Riekel, “Analysis of thenative structure of starch granules with X-ray microfocusdiffraction”, Macromolecules , 3813–3820 (1997).22. Y. Miyauchi, H. Sano, and G. Mizutani, “Selectiveobservation of starch in a water plant using opticalsum-frequency microscopy”, J. Opt. Soc. Am. A ,1687–1690 (2006).23. A. Imberty, H. Chanzy, S. P´erez, A. Bul´eon, and V. Tran,“The double-helical nature of the crystalline part ofA-starch”, J. Mol. Biol. , 365–378 (1988).24. D. Popov, A. Bul´eon, M. Burghammer, H. Chanzy,N. Montesanti, J.-L. Putaux, G. Potocki-Veronese, andC. Riekel, “Crystal structure of A-amylose: A revisitfrom synchrotron microdiffraction analysis of singlecrystals”,
Macromolecules , 1167–1174 (2009).25. R. Cisek, D. Tokarz, M. Steup, I. J. Tetlow, M. J.Emes, K. H. Hebelstrup, A. Blennow, and V. Barzda,“Second harmonic generation microscopy investigationof the crystalline ultrastructure of three barley starchlines affected by hydration”, Biomed. Opt. Express ,3694–3700 (2015).26. F. Vilaplana, J. Hasjim, and R. G. Gilbert, “Amylosecontent in starches: Toward optimal definition andvalidating experimental methods”, Carbohydr. Polym. , 103–111 (2012).27. W. R. Zipfel, R. M. Williams, R. Christie, A. Y. Nikitin,B. T. Hyman, and W. W. Webb, “Live tissue intrinsicemission microscopy using multiphoton-excited nativefluorescence and second harmonic generation”, Proc.Natl. Acad. Sci. U.S.A. , 7075–7080 (2003).28. L. Kong, C. Lee, S. H. Kim, and G. R. Ziegler,“Characterization of starch polymorphic structures usingvibrational sum frequency generation spectroscopy”,
J.Phys. Chem. B , 1775–1783 (2014).29. T. L. Bluhm and P. Zugenmaier, “Detailed structureof the Vh-amylose-iodine complex: a linear polyiodinechain”,
Carbohydr. Res. , 1–10 (1981).30. S. Pavlovic and P. R. G. Brand˜ao, “Adsorption of starch,amylose, amylopectin and glucose monomer and theireffect on the flotation of hematite and quartz”, MinerEng. , 1117–1122 (2003).31. K. Iizuka and T. Aishima, “Starch Gelation ProcessObserved by FT-IR/ATR Spectrometry withMultivariate Data Analysis”, J. Food Sci. , 653–658(1999).32. D. L. Phillips, J. Xing, H. Liu, D.-H. Pan, and H. Corke,“Potential use of Raman spectroscopy for determinationof amylose content in maize starch”, Cereal Chem. ,821–823 (1999). 33. J. A. Feij´o and N. Moreno, “Imaging plant cells bytwo-photon excitation”, Protoplasma , 1–32 (2004).34. Z. Benediktyov´a and L. Nedbal, “Imaging of multi-colorfluorescence emission from leaf tissues”,
Photosynth. Res. , 169 (2009).35. K. Okamoto and T. Akazawa, “Enzymic mechanisms ofstarch breakdown in germinating rice seeds 7. Amylaseformation in the epithelium”,
Plant physiology ,336–340 (1979).36. A. Ochiai, H. Sugai, K. Harada, S. Tanaka, Y. Ishiyama,K. Ito, T. Tanaka, T. Uchiumi, M. Taniguchi, andT. Mitsui, “Crystal structure of α -amylase from Oryzasativa: molecular insights into enzyme activity andthermostability”, Biosci., Biotechnol., Biochem. ,989–997 (2014).37. T. Murata, T. Akazawa, and S. Fukuchi, “EnzymicMechanism of Starch Breakdown in Germinating RiceSeeds I. An Analytical Study ”, Plant Physiol. ,1899–1905 (1968).38. D. D´ebarre, W. Supatto, A.-M. Pena, A. Fabre,T. Tordjmann, L. Combettes, M.-C. Schanne-Klein,and E. Beaurepaire, “Imaging lipid bodies in cells andtissues using third-harmonic generation microscopy”, Nat. Methods , 47–53 (2006).39. C. Saito, M. T. Morita, T. Kato, and M. Tasaka,“Amyloplasts and vacuolar membrane dynamics inthe living graviperceptive cell of the Arabidopsisinflorescence stem”, Plant Cell , 548–558 (2005).40. E. L. Vigil and M. Ruddat, “Development and enzymeactivity of protein bodies in proteinoplasts of tobaccoroot cells”, Histochemistry , 17–27 (1985).41. R. R. Wise, “The diversity of plastid form and function”,in “The structure and function of plastids,” (Springer,2007), pp. 3–26.42. S. Fine and W. P. Hansen, “Optical second harmonicgeneration in biological systems”, Appl. Opt. ,2350–2353 (1971).43. S. Miyata and T. Akazawa, “Enzymic mechanism ofstarch breakdown in germinating rice seeds”, PlantPhysiol.70