Introduction

Energy for maintaining life is produced by oxidative breakdown of food in mitochondria (cellular respiration). Mitochondrion possess an electron transport system in which electrons are successively transferred to molecules of higher redox potentials to generate a membrane potential that eventually facilitates the ATP production. Cytochromes of type a, b and c are key components of the electron transport system involved in respiration. The process of electron transfer leads to the formation of oxidized and reduced forms of cytochromes. Resonance Raman spectroscopy can detect these heme proteins in different redox states with high specificity. Raman spectroscopy generally has the advantage of being a non-destructive and non-staining technique that can be applied to living cells.

The present study demonstrates the possibility of simultaneous observation by resonance Raman spectroscopy of cytochrome b and c (cyt b and cyt c) from mitochondria in a single living yeast cell and also from isolated mitochondria in vitro. Figure 1 shows the Raman spectra of cyt b and cyt c in the reduced and oxidized forms in PBS (phosphate buffered saline (pH 7.2)) of the same concentration. These Raman spectra are obtained with the 532 nm excitation which selectively enhance the cytochrome band intensities. Two characteristic bands in Fig. 1 can be used for the identification and quantification of the cytochrome species. The 1638 cm-1 band is a marker of the oxidized form of cyt b and cyt c, and the 604 cm-1 band is characteristic to the reduced form of cyt c. Using these marker bands, the redox states of cyt b and c have been quantitatively determined under different experimental conditions. In particular, the cytochrome redox states are shown to depend on the mitochondrial respiration activity. In addition, the spatial distribution of cyt b and c redox states in a living animal cell has been quantitatively visualized.

Mitochondrial respiration and the cytochrome redox states

The relationship between the cytochrome redox states and the mitochondrial respiration activity has been studied for living cells as well as for isolated mitochondria. Precultured budding yeast (S. cerevisiae) cells in the mid-log phase were used. Budding yeast cells were cultured in lactate medium for 13 hours and mitochondria were isolated by two different speed centrifugation and density gradient combination.

A representative Raman spectrum from mitochondria in living S. cerevisiae cells is shown in Fig. 2 in comparison with that of isolated mitochondria suspended in a non-nutrient buffer. The living cell spectrum (Fig. 2(a)) corresponds well with that of the isolated mitochondria spectrum (Fig. 2(b)). However, the oxidation marker band at 1638 cm-1 is observed only in isolated mitochondria and not in living cells. The oxidized form of cyt b and/or cyt c exists only in the isolated mitochondria. Respiration in isolated mitochondria can be induced by adding ADP and succinate to the medium. The isolated mitochondrial spectrum after addition of ADP and succinate is shown in Fig. 2(c). Small but significant changes are observed at 1638 cm-1 and 604 cm-1 on going from Fig. 2(b) to 2 (c). In order to clarify these small changes, the difference spectrum between the respiration inactive (Fig. 2 (b)) and active (Fig. 2(c)) mitochondria spectra is calculated (Fig. 3(a)). The difference spectrum shows more clearly that the intensity of the 1638 cm-1 marker band for oxidized cytochromes decreases while the 604 cm-1 marker band specific for reduced cyt c increases.

A more quantitative analysis of these changes is carried out by fitting this difference spectrum with a linear combination of the four Raman spectra of cyt b and c in the oxidized and reduced forms. The best least-squares fitted spectrum (Fig.3 (b)) is obtained with the following coefficients, +10 for cyt b oxidized, -8 for cyt b reduced, -71 for cyt c oxidized and +13 for cyt c reduced. These coefficients indicate that mitochondrial respiration activity induced by ADP and succinate promotes the oxidation of cyt b (+10 for the oxidized form and -8 for the reduced) and the reduction of cyt c (-71 for the oxidized form and +13 for the reduced). Although these coefficients need further quantitative confirmation, the change of the cytochrome redox state is successfully correlated with the mitochondrial respiration activity.

Spatial distribution of cyt b and c redox states in a living animal cell

Spatial distribution of cyt b and c redox states has been studied in a living L929 (NCNC) cell. L929 cells were cultured in DMEM (Dulbeco's modified eagle medium) supplemented with 10% fetal bovine serum for 4 days on a glass bottom dish. Figure 4 (a) shows a representative Raman spectrum obtained from the mitochondria in living L929 cells. In order to extract the spectral component derived from cytochromes, I have performed the following spectral analysis. The Raman spectrum of bovine albumin (Fig.4 (b)) as a standard protein spectrum is subtracted from the raw spectrum in order to eliminate the spectral contribution from non-heme proteins. The Raman band at 1003 cm-1 due to the phenylalanine residue is used as an intensity standard. Figure 4 (c) shows the spectrum obtained after subtracting the albumin spectrum and the fluorescence background. The Raman band at 1638 cm-1, which is a marker band of oxidized cytochromes, is clearly observed. This is the first detection of oxidized cytochromes in a living cell by Raman microspectroscopy.

The spectrum in Fig. 4 (c) is again fitted with a linear combination of the standard Raman spectra of cytochrome b/c reduced/oxidized forms. The best fit model spectrum (Fig. 4 (d)) is obtained with coefficients 14 for reduced cyt b, 39 for oxidized cyt b, 27 for reduced cyt c and 20 for oxidized cyt c. Since the concentrations of the four cytochrome standard solutions are the same, the determined coefficients correspond to the relative molecular abundance of the four cytochrome species. Fig. 4 (e) shows the residue spectrum (Fig. 4 (c) - Fig. 4 (d)). It shows no spectral features corresponding to cytochrome species and is likely to be due to lipids. The fitting analysis seems to be performed adequately.

All the 100 x 100 points Raman spectra from the whole cell are analyzed in the same way as shown above. Figures 5 (a)~(d) show the obtained distributions of reduced cyt b, reduced cyt c, oxidized cyt b and oxidized cyt c, respectively. These four distributions are similar to one another, indicating that they grossly correspond to the distribution of mitochondria, which contain both cyt b and c abundantly. In the case of cyt c, the reduced form is more abundant than the oxidized form all over the cell. On the other hand, in the case of cyt b, the concentrations of both the oxidized and reduced forms are only slightly different from each other. As far as the author is aware, such quantitative information about the redox states of cytochromes in a living cell has been obtained for the first time.

Finally, the cytochrome redox state changes have been observed after 2 hours under a non-nourishment condition. Spectral changes similar to Fig. 3 (c) are observed. Although further experiments under various culture conditions are required to dissect out the spectral changes due solely to respiration, the present resonance Raman approach has an unparalleled potential for accessing the mitochondrial respiration activity in vivo and label free.

Conclusion

In summary, this thesis demonstrates that resonance Raman spectroscopy can assess the mitochondrial respiration activity by quantifying the redox states of cytochrome b and c simultaneously. This new method is applicable in vivo, in vitro and in situ without using any labeling or genetic manipulation and is therefore promising for many biological applications.

Figure 1 Standard resonance Raman spectra for (a): reduced cytochrome b, (b): reduced cytochrome c, (c): oxidized cytochrome b and (d): oxidized cytochrome c .

Figure 2 Raman spectra of (a): mitochondria in living yeast cells, (b): isolated mitochondria without ADP and succinate and (c): isolated mitochondria with ADP and succinate. Fluorescence backgrounds have been subtracted from the Raman spectra by polynomial fitting.

Figure 3 (a): Difference Raman spectrum (Fig. 2(c) - (b)) (b): best fit linear combination of the four authentic cytochrome Raman spectra in Fig. 1 and (c): the residual spectrum. Flat parts were not used for fitting.

Figure 4 (a): typical Raman spectrum from a living cell, (b): Raman spectrum of albumin, (c): albumin and background subtracted spectrum, (d): model spectrum made by standard cytochrome spectra, (e): residual spectrum.

Figure 5 Raman images of (a): reduced cyt b, (b): reduced cyt c, (c): oxidized cyt b and (d): oxidized cyt c. The bar indicates 10 μm.

本論文は、ラマンスペクトルを用いてミトコンドリア内のシトクロムの酸化還元状態を定量的に評価し、ミトコンドリアの呼吸活性とシトクロムの酸化還元状態の相関を定量的に見積もる新しい方法論を提案するもので、全7章から構成される。

第1章では導入として、本研究の対象であるミトコンドリアの重要性と機能、シトクロムの特徴と役割、手法である共鳴ラマン分光の概要が述べられている。また、本論文で提案する手法が、ミトコンドリアの活性を評価する有力な分析法となり得ることが述べられている。

第2章では、細胞培養、ミトコンドリアの単離、及びラマンスペクトルの取得など実験操作について述べられている。シトクロムの酸化型、還元型の標準溶液を市販の精製試料を用いて作成し、標準共鳴ラマンスペクトルを得たこと、またその標準共鳴ラマンスペクトルの特徴的なバンドや強度について述べられている。

第3章では、単離したミトコンドリア内のシトクロムの含量比と成長曲線の関係について述べられている。細胞分裂頻度の最も活発な時にシトクロムの含量比が最も高くなることが見出された。また、得られた共鳴ラマンスペクトルが4つの標準スペクトルで非常によく表現でき、4成分の比を求めることで、ミトコンドリア内のシトクロムの酸化還元状態が定量的に求められることが示された。

第4章では、単離したミトコンドリアと生きた酵母中のミトコンドリア中のシトクロムの酸化還元状態を比較している。生細胞中のミトコンドリアでは観測されなかったシトクロムの酸化型が単離したミトコンドリアでは確認され、呼吸活性を与えると減少する結果が示された。標準スペクトルと比較した結果、呼吸活性を与えるとシトクロムcの酸化型が大きく減少することが確認された。

第5章では、動物細胞中のシトクロムの酸化還元状態の分布イメージを作成し、その栄養素の有無による違いを考察している。分布イメージにより、細胞中に含まれる2種類のシトクロムそれぞれの酸化還元分布が明瞭に可視化され、生細胞中ではシトクロムcは還元型であることが確認された。さらに、無栄養の環境下に放置した場合は、シトクロムcの酸化型が徐々に増加していくことが確認された。

第6章は第4章と第5章の結果の比較と改善点について述べている。両結果共にシトクロムcの電荷の収支が合わないことをとりあげ、この原因として電子リークや中間体の可能性を議論している。まだ不十分ではあるが、提案された手法がミトコンドリア内の電子の流れを定量化し得ることが述べられており、さらなる研究の発展の可能性が示されている。

第7章は以上の研究成果のまとめである。

本研究により、生細胞内及び単離されたミトコンドリア中の異なる2種類のシトクロムの酸化還元状態を定量的に測定することが可能となり、これまで不可能であった生細胞中の酸化還元状態の計測、解析が可能となった。この結果、呼吸活性と酸化還元状態が密接に相関することが明らかとなった。非破壊・非侵襲で、単離したオルガネラから生細胞中までの測定が可能な新しい方法論の開発と、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4章の主要部分は、Journal of Biophotonics誌に公表済み(Venkatesh Kaliaperumal、〓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、論文提出者柿田穣に博士(理学)の学位を授与することが適当であると認める。