Introduction

Molecular level investigation of cellular processes is of great importance in understanding life. Raman microspectroscopy is one of the most informative methods for in vivo studies of living cells, providing sub-μm space-resolved molecular information without any sample pretreatment. However, this method is not readily applicable to photosensitive / photolabile organisms because of the following two difficulties. One is that such organisms often emit strong autofluorescence that easily interferes with the Raman measurements. The other difficulty is that photolabile systems are easily photodamaged by visible excitation. Some photosensitive organisms emit strong fluorescence and/or are photodamaged even with near-infrared (NIR) excitation at ~800nm. For the purpose of further reducing the fluorescence interference, I have newly developed a 1064nm deep NIR Raman system with the sensitivity high enough for in vivo measurements of single living cells.

This new apparatus has been applied to the structure/function analysis of photosynthetic pigments in cyanobacterial cells, which are prototype model organisms for photosynthesis research. Because cyanobacteria contain several kinds of photosynthetic pigments, they emit strong autofluorescence even with ~800nm excitation. By using 1064nm deep NIR excitation, it has become possible to obtain high SIN Raman spectra from carotenoid, chlorophyll a and phycobilins. Thus, their distribution, structure and function analysis within a single cell become possible under physiological conditions.

1064 rim excited multichannel Raman microspectrometer (Chapter 2)

When constructing a Raman microspectrometer with deep NIR excitation, the detection efficiency is a peculiar problem. A CCD camera, which is widely used in the visible excitation Raman system, cannot be used in the deep N1R region because of its low quantum efficiency. I here use an InP/InGaAsP NIR image intensifier (NIR-II), whose quantum efficiency is approximately 4% in the 900-1400nm wavelength region, which covers the overall fingerprint region when excited at 1064nm.

The developed apparatus is shown in Fugure 1. By using electrical gating synchronized with a Q-switched Nd:YAG laser (1064nm, 30 ns pulse, 10 kHz), the thermal noise from the NIR-II is considerably lowered. The laser beam is focused on the sample with a x 1 00/ 1.3 NA microscope objective. A 50 wir) cross slit is used as the entrance slit of a polychromator as well as the confocal pin hole, in order to reduce the number of optics used. The lateral spatial resolution of 0.7 1.1.in and the depth spatial resolution of 3.1 μm are obtained. Because of the multichannel detection with N1R-II, approximately 1000cm(-1) wavenumber region is simultaneously detected with 10cm(-1) spectral resolution.

Deep NIR Raman measurements of cyanobacteria (Chapter 3)

As the sample, 1 here use cyanobacteria Thermosynechococcus elonguius (T elongatus), which is a well-known model organism for the photosynthesis research. For the preparation of sample, T elongatus cells are fixed between a slide glass and a cover glass by controlling the amount of sample. For avoiding the photodamage to the sample, laser power is lowered to less than I mW, typically 0.5 mW.

The new deep NIR excited microspectrometer enables in vivo Rman measurement of single living cyanobacterial cells. Figure 2 shows the comparison between the Raman spectra of T elongatus cells measured with 785nm excitation and that measured with 1064nm excitation. With 785nm excitation, fluorescence from photosynthetic pigments strongly interferes. It is also seen that, after the start of laser irradiation, the spectral profile of fluorescence background changes with time, indicating that the cell is photodamaged. On the other hand, 1064nm excitation drastically decrease the autofluorescence and shows several distinct Raman bands. Furthermore, the spectra do not show time dependence, indicating that the photodamage or photobleaching does not occur. The present deep NIR excitation is suitable for the Raman measurement of photo-labile living cells.

Raman band assignments (Chapter 4)

For the purpose of band assignments, Raman spectrum of T elongatus is compared with those of three photosynthetic pigments, β-carotene, chlorophyll a and phycobilin (Figure 3). From the comparison, Raman bands at 1008, 1157 and 1523cm(-1) are assigned to carotenoids. The 1225, 1325cm(-1) bands and 1279, 1369, 1586, 1635cm(-1) bands are assigned to chlorophyll a and phycobilin, respectively. In this way, almost all bands are assignable to the three types of pigments. These three photosynthetic pigments are selectively observed in the 1064nm excited Raman measurements of cyanobacteria, because of the pre-resonance Raman effect due to their strong absorptions in the visible region.

Raman imaging of photosynthetic pigments in T elongates (Cheater 4 and 5)

Chemical specificity shown above enables the Raman imaging of photosynthetic pigments distributions inside the cyanobacterial cells. For the Raman mapping experiment, the T elongatus cell is translated by a piezoelectric stage horizontally with a step of 0.3 μm. At each sample point, Raman spectra are measured in 10 seconds. After using singular value decomposition in order to reduce the noise, the Raman images are constructed from the band intensity at each sample point. As shown in Figure 4, distribution information of each photosynthetic pigment is clearly visualized.

Images belonging to the same pigments show identical distribution patterns, supporting the band assignments given above. It is obvious that the distributions of the three pigments are distinct from one another. In particular, carotenoids show a quite different distribution from those of chlorophyll a and phycobilins.

It is also found that the C=C stretch band of carotenoids shows peak shift from 1526cm(-1) to 1520cm(-1) (Fig. 5a), suggesting their structural variance within the cell. This peak shift is closely correlated with the variance of intensity ratio of carotenoid / phycobilins (Fig 5b, c). In the photosynthesis processes, phycobilins play a role of light harvesting antenna, whereas carotenoids are thought to play a role of antioxidant and photoprotection. This finding indicates that the localization of structurally different carotenoids is likely to be related to the photoprotective process of cyanobacteria

Conclusion

A newly constructed 1064 tun NIR excited Raman microspectrometer using InP/InGaAsP image intensifier is described. In the measurement of a living cyanobacterial cell, bypassing the interference from autofluorescence, varied distribution of photosynthetic pigments, along with the structural variance of carotenoids, are visualized. A new possibility is thus shown for in vivo and molecular level studies of photosynthesis systems.

Figure 1. Schematic diagram of 1064nm excited multichannel Raman spectrometer

Figure 2. 785nm (a), 1064nm (b) excited Raman spectra of T.elongates.

Figure 3. Raman spectra of T.elongates (a) and photosynthetic pigments (b).

Figure 4. Raman mapping images of each photosynthetic nigments obtained from a T.elongatus single living cell.

Figure 5. (a) Peak position mapping image of carotenoid C=C stretch band and (b) intensity ratio mapping image of phycobilin / carotenoid. (c) Peak posision of carotenoid C=C stretch band is plotted against intensity ratio phycobilin / carotenoid.

本論文は、生細胞など光に敏感な生体試料を分子レベルで測定するための近赤外励起顕微ラマン分光装置の開発と、その応用について記述されており、全6章から構成される。

第1章では導入として、本研究の目的が生体機能の分子レベルの理解にあり、それには非破壊、非侵襲な手法を用い、生理的条件下での分析が重要であることが述べられている。既存の近赤外励起ラマン分光器は、感度が低く単一生細胞への応用が困難であることを踏まえ、新規な手法として、より高感度なマルチチャンネル顕微近赤外ラマン分光法が提案されている。また、応用対象であるシアノバクテリアについて、その機能・構造の理解が光合成の総合的な理解には重要であり、本手法が有力な分析手法となり得ることが述べられている。

第2章では、開発された1064nm励起マルチチャンネル顕微ラマン分光装置の詳細及びその性能について述べられている。近赤外イメージインテンシファイアを検出器に用い、また顕微鏡を含めた集光系全体を近赤外仕様に改良することにより、単一の生細胞にも応用可能な装置となった。本装置が面内0.7μm,奥行き3.1μmの空間分解能を有しており、細胞内の空間分解測定に応用可能であることが示された。

第3章では、1064nm励起顕微ラマン分光の応用として、出芽酵母、及びシアノバクテリアのin vivo測定の例が示されている。本装置により、単一の生細胞についても、十分なS/N比で近赤外励起ラマン分光測定が可能となった。また、既存の785nm励起ラマン測定との比較から、シアノバクテリアのような光に敏感な生細胞の測定には、本手法のような近赤外領域の長波長励起が不可欠であることが示された。

第4章では、シアノバクテリア生細胞の空間分解測定に応用した結果が示され、さらに得られたラマンバンドの帰属についての考察が述べられている。シアノバクテリアに含まれる光合成色素である、カロテノイド、クロロフィル、フィコシアニンのラマンスペクトルとの比較、また単離したチラコイド膜のラマンスペクトルとの比較から、シアノバクテリアから得られるラマンバンドの大多数は光合成色素の何れかに帰属されることが確認された。

第5章では、シアノバクテリア生細胞のラマンイメージを構成し、細胞の構造、機能との関連を考察している。ラマンイメージにより、細胞中に含まれる3種類の光合成色素それぞれの分布が明瞭に可視化され、特にカロテノイドが他の色素と異なる分布を示していることが確認された。さらに、ラマンバンドのピークシフトの考察から、異なる分子構造を有したカロテノイドが局在化していることも可視化された。さらに、この局在化はクロロフィルやフィコシアニンの濃度分布とも強く相関があることが確認され、生体内でのカロテノイドの特異な機能との関連について論じられている。

第6章は以上の研究成果のまとめである。

本研究により、生細胞の空間分解測定に応用可能な1064nm近赤外励起顕微ラマン分光手法が開発され、これまで不可能であった光に敏感な単一生細胞の分子レベルでの計測、解析が可能となった。非破壊・非侵襲で分子の分布、構造情報を得ることは、生理的条件下での生体機能の解明には不可欠である。本研究で行ったように、それがシアノバクテリアのような光合成生物でも可能となったことは、本手法が光合成機構の総合的な理解へ向けた新しいアプローチとなることを明瞭に示している。このような新規の観測手法の開発とその有用性を提示した本論文の内容は高く評価できる。

本論文第2-5章の主要部分は、APPliedSpectmsoopy誌に受理済み、公表予定(杉浦美羽、林秀則、渡口宏夫との共著)であるが、論文提出者が主体となって実験および解析を行なっており、その寄与が十分であるので、学位論文の一部とすることに何ら問題はないと判断する。

以上の理由から、論文提出者安藤正浩に博士(理学)の学位を授与することが適当であると認める。