Introduction

Several species of green algae, designated siphonous green algae, are a deep drab color and utilize green light for photosynthesis (Anderson 1983). This coloration and wavelength utilization are caused because the carotenoid siphonaxanthin (Sx) and its ester (Sn) take the place of lutein in LHCII-type photosynthetic antenna complexes (Wang et al. 2013). These siphonous green algae inhabit the subtidal zone of seashores, where irradiance conditions fluctuate drastically because of changes in water column depths. The adaptations of siphonous green algae to these fluctuations in irradiance conditions have not been studied closely because these algae normally grow into large and spongy bodies; consequently, they are difficult to culture. Recently, we developed a culture method for propagation of a filamentous form of Codium (C.) intricatum OKAMURA, a siphonous green alga (Oka et al. 2012). This culture method is optimized to obtain higher Sx/Sn yields in lyophilized cells. During the optimization process, we found that the pigment composition is affected only when extremely high light intensities were used during culture.

Sx-chlorophyll a/b protein (SCP) constitutes the light-harvesting complexes of siphonous green algae. SCPs have been isolated, and the pigment composition has been investigated. The chemical structures of all the carotenoids of C. intricatum are pictured in Fig. 1. The characteristic carbonyl carotenoid, Sx, Sn, or both are the major light-harvesting carotenoids; moreover, α-carotene takes the place of β-carotene (Anderson 1983; Chu and Anderson 1985). Xanthophyll cycle carotenoids violaxanthin (Vx) and neoxanthin (Nx) are also present (Yokohama 1983). The Nx in natural oxygenic photosynthetic organisms takes the 9′-cis Nx (9cNx) conformation (Takaichi and Mimuro 1998). Notably, an additional pigment is evident when intact cultivated filamentous C. intricatum cells are grown under extremely high light intensities, and this pigment is probably all-trans Nx (tNx) based on its elution time and absorption spectrum.

Fig. 1
figure 1

Chemical structures of the carotenoids in C. intricatum

Full size image

Resonance Raman spectroscopy is a powerful method for distinguishing between cis and trans conformations of carotenoids in solution (Koyama and Fujii 1999) and in pigment-protein complexes (Robert 2009; Andreeva et al. 2004). In the LHCII of plants, one 9cNx and two all-trans lutein are bound with an unstoichiometric number of all-trans Vx (tVx), and the key Raman signals of 9cNx can be clearly identified (Ruban et al. 2000, 2001; Snyder et al. 2004). Excitation wavelength dependences of the resonance Raman signals have been used to assign the absorptive transitions of each xanthophyll in LHCII complexes. Moreover, Raman signals of lutein have been empirically assigned in relation to the oligomeric/trimeric/monomeric LHCII in plants (Ruban et al. 2000, 2007).

In this study, we used resonance Raman spectroscopy to examine intact filamentous C. intricatum cells cultivated under high- and low-light intensities to determine the conformations of the carotenoids in these cells.

Experimental

Culture of filamentous C. intricatum

C. intricatum OKAMURA was collected at Okinawa prefecture, Japan, and the clonal algal cultures were established via the micropipette-washing method (Aruga et al. 2000). For one week, filamentous bodies were pre-cultured at 30 °C under a 12–12 h light cycle with 60 µmol m−2 s−1 photosynthetic photon flux density (PPFD) in filtered, sterilized sea water with 0.05 % (v/v) TSP29 medium (South Product, Inc., Okinawa, Japan) and continuous bubbling aeration. The only difference between the culture conditions for the two types of cultures (high versus low light intensity) was the irradiances: high (500 PPFD) versus low (20 PPFD) irradiance. Cells were harvested by centrifugation (Kubota 7780) and frozen at −20 °C until analysis.

Pigment extraction

Frozen filamentous cells were lyophilized in a FDU-2100 freeze dryer with a dry chamber (TOKYO RIKAKIKAI CO, LTD, Tokyo, Japan) before pigment extraction. Lyophilized whole cells were homogenized in dimethyl formamide (DMF, Merck) on ice, and homogenates then were subject to centrifugation. The supernatant (whole extracts) was immediately injected for HPLC analysis.

High-performance liquid chromatography (HPLC)

A reversed-phase HPLC column (Inertsil ODS-P (5 μm), 4.6 × 150 mm, GL Science Inc.) was equilibrated with eluent A (methanol, acetonitrile, and water at a 9:7:4 (v/v/v) proportional relationship). Samples (20 μl) were injected to the HPLC equipped with a photo-diode array detector (SPD-M20A, Shimadzu, Kyoto, Japan). A linear gradient of eluent B (acetonitrile and ethyl acetate at a 3:7 (v/v) ratio) was then run through the column with flow rate of 0.8 ml/min from 5 to 25 min after injection. Absorption data for a 300-800 nm region were recorded during the HPLC run via both the D2 and W lamps. LC–APCI–MS spectrometry was performed with a quadrupole MS spectrometer (JMS-700T, JEOL, Japan) equipped with an atmospheric-pressure chemical ionization (APCI) ion source (4.0 kV, 400 °C at needle) and the HPLC system. Peaks were assigned chemical identities based on the results of the LC–APCI–MS spectrometry, HPLC elution pattern, and absorption spectra (see Fig. 2). For each culture, three independent extractions followed by HPLC analyses were used to calculate pigment composition.

Fig. 2
figure 2

a HPLC profiles of all pigments from intact filamentous C. intricatum cells cultivated under HL (black solid line) or LL (red solid line with hash). Profiles were normalized relative to the Sn peak at 22 min. b Absorption spectra of 9cNx (dashed line for HL cell and solid line for LL cell), tNx (solid line) and Vx (dashed line) in HPLC eluent. (See text) (Sx siphonaxanthin, 9cNx 9′-cis neoxanthin, tNx all-trans neoxanthin, Vx violaxanthin, Sn siphonaxanthin ester, α, α-carotene)

Full size image

Absorption spectra at 77 K

Frozen filamentous cells were suspended in 60 % glycerol–buffer, and absorption spectra at 77 K were recorded with a Varian Cary E5 spectrophotometer. The spectra were normalized relative to the glycerol–buffer standard; following removal of the glycerol–buffer baseline. The resulting spectra were plotted. The second derivatives of the absorption spectra with respect to wavenumber was computed using the Igor Pro 6.33 J program package followed by spectral smoothing (see Fig. 3b).

Fig. 3
figure 3

a 77 K Absorption spectra of samples from intact filamentous C. intricatum cells cultivated under HL or LL (black solid and dashed red lines, respectively), b their second derivatives (× -100), and c difference spectrum (× -300) of the second derivatives (HL minus LL) (See Experimental section for further information). Carotenoid regions of (b) and (c) are magnified relative to the chlorophyll region by the factors indicated. Excitation wavelengths for the Raman spectra are indicated by numbered arrows (See Fig. 5)

Full size image

Resonance Raman spectroscopy

Frozen filamentous cells were placed on the sample plate and inserted into a liquid nitrogen flow cryostat (Air Liquide). Excitation lights (457.9, 476.5, 488.0, 496.5, 501.7, 514.5, and 528.7 nm) were generated with a 24 W Ar+ Sabre laser (Coherent, Palo Alto, CA); the 441.6 nm excitation light was generated with a Liconix He–Cd laser. The 90° scattering from each sample was focused onto a Raman spectrometer (U1000, Jobin–Yvon, France) equipped with a blue-sensitive back-illuminated liquid nitrogen-cooled CCD camera (Spectrum One, Jobin–Yvon, France).

Results and discussion

Pigments in intact filamentous C. intricatum cells

The natural irradiance conditions for Codium species fluctuate drastically because they inhabit the subtidal zone. We successfully cultivated a filamentous form of one Codium species, C. intricatum OKAMURA, to study pigment composition under multiple controlled light conditions. The filamentous form is rich in chloroplasts and lacks the tough outer structure that protects against dehydration; therefore, it is ideal for the direct detection of intact thylakoid membranes.

HPLC elution profiles of whole pigments from cultivated C. intricatum grown under high-light (HL) and low-light (LL) conditions are shown (Fig. 2a). Each HPLC peak associated with samples extracted from LL cells was assigned a definite pigment identity based on retention time and absorption spectrum; these samples contained Sx, 9cNx, Vx, Sn, Chl b, Chl a, and α-carotene (α). With the HL samples, an additional peak was evident just after the 9cNx peak; this peak had absorption maxima at 472, 443, and 415 nm in HPLC eluent (Fig. 2b). This peak with three absorption maxima was assigned to tNx based on previous studies (Takaichi and Mimuro 1998; Snyder et al. 2004). LC–APCI–MS signals for tNx and 9cNx contained identical bands (MH+, 601 m/z; MH+–H2O, 583 m/z; and MH+–2H2O, 565 m/z), confirming this assignment. The tNx peak increased when extracts from intact cells were incubated in light for a few hours (data not shown). This means isomerization from 9cNx to tNx may occur under ambient condition. Therefore the reproducibility was precisely investigated under strict extraction conditions, i.e., dim light, ice temperature, and quick extraction, to draw this result. Thus, we concluded that tNx and 9cNx accumulated during cultivation under HL condition, but only 9cNx accumulated under LL conditions.

The carotenoid composition of each culture type was calculated using values reported for the molar extinction coefficients and absorption spectra in HPLC eluent (Table 1). Note that in the binary HPLC system the HPLC eluent means a mixture of eluent A and B at the retention time. All values are in relative and in arbitrary units; the value for Sn was set to 1. For HL cells, both 9cNx and tNx were calculated as 9cNx. The mixing of the HPLC peaks representing 9cNx and tNx were simulated; the peak shape of 9cNx in the HPLC profile of LL cells was used as that of both 9cNx and tNx, and magnitude and retention time were adjusted to fit real HPLC peaks of HL cells. As a result, the relative area of 9cNx/tNx in the HL profile was found to be 2.2/1.0. The composition of 9cNx (tNx) in HL cell was calculated accordingly to be 0.8 ± 0.1 (0.4 ± 0.1). Therefore, the amount of tNx in LL was clearly lower than that in HL cells, but the amount of 9cNx was the same in the two cell types. Additionally, accumulations of Vx and α-carotene were also higher in HL cells than in LL cells.

Table 1 Carotenoid compositions of intact filamentous C. intricatum cells cultivated under HL or LL conditions
Full size table

Low-temperature absorption spectra

The absorption spectra of intact filamentous cells were recorded at 77 K (Fig. 3). The absorption spectra of intact cells (or chloroplasts) represent the characteristics of the major constituent, the SCP (Anderson 1983). Absorption spectra of HL (upper solid black line) and LL (lower dotted red line) cells were very similar, but the scattering backgrounds differed. The second derivatives were useful for identifying peak positions. Figure 3b shows the second derivatives of the absorption spectra. Importantly, the sign of data was flipped, hence the peak of the second derivative indicates the peak of original absorption spectrum. The spectrum representing the difference of the second derivatives is also shown to emphasize the meaningful differences in the original spectra (Fig. 3c).

In the near-infrared region, second derivative peaks were observed at 676.5, 661.0, and 649.5 nm (Fig. 3b). The 676.5 nm peak can be assignable to the characteristic peak of photosystems reported at 675.0 nm in intact chloroplast (Anderson 1983). The shoulder around 670.0 and 661.0 nm peak can be assignable to the characteristic peaks of the SCP reported at 669 and 660 nm for isolated SCP (Anderson 1983). The spectrum representing the difference of the second derivatives shows two significant features: one feature is a first-derivative type sigmoid shape at 678.0 nm, and the other is positive enhancement at 661.0 nm. The sigmoid shape can be explained as a blue shift at 678.0 nm peak in the original absorption spectra. The positive enhancement can be explained as an increase in intensity at 661.0 nm peak in the original absorption spectra. Near-infrared absorption peaks were sensitive to the assembly of Chl a pigment in the pigment–protein complexes. Thus, both an increase in the magnitude of the 661.0 nm peak and the blue shift at the 678.0 nm peak with HL samples versus LL samples can be explained as an increase in the number of SCPs relative to the number of photosystems under HL conditions.

In the visible-light region, a peak characteristic of SCPs was clearly observed at 542.0 nm in the second derivative spectra (Fig. 3b). This peak could be definitely assigned to Sn/Sx binding (Yokohama 1983). The 487.5, 476.5, and 465.0 nm bands in the second derivative spectra could be assigned to a mixture of the component carotenoids, 9cNx, Sn/Sx, Vx, and α-carotene. The carotenoid peaks from plant LHCII have been clearly documented by Ruban et al. 2000, and the complete absence of a 495 nm peak from our spectrum was consistent with their conclusion that the 495 nm peak represents lutein. In the spectra representing the difference of the second derivatives (Fig. 3c), two clear peaks were evident at 486.5 and 453.5 nm. These peaks indicated notable increases in intensity at both wavelengths in HL cells, and they had an interval of 1,544 cm−1, suggesting 0–0 and 0–1 vibrational transitions of a single carotenoid. Consequently, both peaks can be originated to a single carotenoid that increases in HL cells. The absorption spectra of tNx and Vx were almost identical (Fig. 2b); therefore, the increasing carotenoids represented by the 486.5 and 453.5 nm bands were assigned to both tNx and Vx (See Table 3). Notably, the absorptive transitions were determined based on the π electron-conjugated systems; consequently, we could not distinguish between tNx and tVx with this spectroscopy.

Resonance Raman spectra

Resonance Raman spectra of filamentous whole cells of C. intricatum grown under HL and LL conditions excited at 496.5 nm are shown in Fig. 4. Both spectra showed Raman lines characteristic of carotenoids: C=C stretching (str.) mode (ν1) appears around 1,500 cm−1, C–C str. mode (ν2) appears around 1,100–1,240 cm−1, methyl rocking mode (ν3) appears around 1,000–1,010 cm−1, and C–H out-of-plane wagging mode (ν4) appears around 950–970 cm−1 (Robert 2009). These bands are sensitive to both conformation and structure of carotenoids; therefore, the excitation wavelength dependences of the Raman spectra are useful for assigning absorptive transitions of carotenoids bound to pigment–protein complexes.

Fig. 4
figure 4

77 K Resonance Raman spectra of samples from intact filamentous C. intricatum cells cultivated under HL or LL (black and red solid lines, respectively) both excited at 496.5 nm

Full size image

The ν1 ~ ν4 region of the Raman spectra of HL samples were compared with those of LL samples (Fig. 5); all samples were excited with eight different laser-generated wavelengths (441.6, 457.9, 476.5, 488.0, 496.5, 501.7, 514.5 and 528.7 nm, see arrows in Fig. 3a). Figure 5d shows the ν1 Raman lines which are sensitive to both conjugation lengths (Mendes-Pinto et al. 2013) and to cis–trans configurations (Koyama and Fujii 1999; Robert 2009). The ν1 Raman lines of xanthophylls bound to higher plant-type LHCII have been precisely assigned, and 9cNx results in lines at 1,532–1,535 cm−1 that change slightly in energy depending on excitation wavelengths (Ruban et al. 2001). The ν1 Raman lines of xanthophylls in the LHCII are close enough to result in one peak when present together; therefore, the plot of the peak frequencies against excitation wavelength can be helpful for evaluating the contribution of each xanthophyll.

Fig. 5
figure 5

77 K Resonance Raman spectra of samples from intact filamentous C. intricatum cells cultivated under HL or LL (black and red solid lines, respectively). The a ν4, b ν3, c ν2 and d ν1 regions of the spectra are shown separately. Excitation wavelengths were (1) 441.6, (2) 457.9, (3) 476.5, (4) 488.0, (5) 496.5, (6) 501.7, (7) 514.5, and (8) 528.7 nm for each region

Full size image

In C. intricatum, Sx and Sn substitute for lutein, and α-carotene substitutes for β-carotene. Based on the structural similarities and the similarities of the absorption spectra in solution, the ν1 Raman frequency of both Sx and Sn can be quite similar to that of fucoxanthin, i.e., approximately 1,533 cm−1 in solution and approximately 1,529 cm−1 in the pigment–protein complexes, FCP (fucoxanthin–chlorophyll a/c protein) (Premvardhan et al. 2009). Also, α-carotene may have a ν1 Raman frequency very similar to that of lutein; the ν1 Raman frequency is 1,522–1,524 cm−1 (Ruban et al. 2001).

The ν1 peak maxima at the eight different excitation wavelengths were plotted for HL and LL samples (Fig. 6a). With LL samples, the resonance Raman profile was similar to that of higher-plant type trimeric LHCII (Ruban et al. 2000); the higher frequency was around 1,529–1,530 cm−1 at both 457.9 and 488.0 nm excitations, and the lower frequency was around 1,527–1,528 cm−1 at 441.6/476.5/501.7 nm excitations. These higher and lower frequencies may have occurred because the contribution of 9cNx to the ν1 Raman line was enhanced with excitation wavelengths of 457.9 and 488.0 nm and the contributions of Sx/Sn to ν1 were enhanced with excitation wavelengths of (441.6)/476.5/501.7 nm. In HL cells, the excitation-wavelength dependences were rather unclear between excitation wavelengths of 441.6 and 488.0 nm. In other words, the contribution of carotenoid species showing higher ν1 frequencies, those around 1,531 cm−1, increased with excitation wavelengths in this region. Table 2 lists the excitation energy dependences of the ν1 Raman frequencies with those bandwidths at the half maxima. Extremely large bandwidths of approximately 18–19 cm−1 were observed when excited at 441.6 and 476.5 nm. The large bandwidth resulting from excitation at 441.6 nm can be explained by the effect of a nearby Raman band of Chl a at 1,560 cm−1; the other large bandwidth resulting from excitation at 476.5 nm may indicate the contribution of many different carotenoids having slightly different ν1 frequencies. Based on the HPLC results, both tNx and tVx were more abundant in HL cells than in LL cells. The linear dependencies of the ν1 Raman lines on absorptive transitions have been reported (Mendes-Pinto et al. 2013); therefore, that both tNx and tVx resulted in the same absorption spectra (as shown in the Fig. 2b) could be attributed to the same ν1 Raman lines. The ν1 Raman line of tVx is reported to be around 1,528 cm−1 in LHCII (Ruban et al. 2000) and around 1,530 cm−1 in thylakoid membranes (Ruban et al. 2001). Consequently, the component that exhibited a ν1 frequency as high as 1,532 cm−1 could have resulted from tNx, tVx, or both having weaker interactions with the proteins, which were more abundant when cells were grown in HL conditions.

Fig. 6
figure 6

a ν1 Raman frequencies and b amplitude of ν4 Raman lines relative to the ν1 Raman lines as a function of the excitation wavelengths for HL (closed circles and solid black line) and LL (open circles and red dashed line) intact filamentous C. intricatum cells

Full size image
Table 2 ν1 Raman frequencies for intact filamentous C. intricatum cells cultivated under HL or LL
Full size table

The ν2 region of the Raman spectra is shown in Fig. 5c; this region is sensitive to the conformation of molecules (Koyama and Fujii 1999; Robert 2009). The Raman line at 1,136 cm−1 (indicated by gray line in Fig. 5c) is highly characteristic of 9′-cis isomer; this fingerprint was evident in both HL and LL spectra at each excitation wavelength (1,132 cm−1 in the reference of Snyder et al. (2004) for the 9′-cis structure of Nx (9cNx in our abbreviation) in the LHCII binding site). A small but unambiguous enhancement of the amplitudes of 1,136 cm−1 bands in LL spectra was observed when samples were excited with longer wavelengths, such as 501.7 and 514.5 nm. In order to quantify the cis contribution to the Raman signals, the relative intensity of the C10–C11 str. mode (at 1,136 cm−1) to the C14–C15 str. mode (at 1,159 cm−1) was calculated to be 0.06 (0.06) and 0.02 (0.05) for HL (LL) samples excited at 501.7 and 514.5 nm, respectively. The relative intensity (1,136 cm−1/1,159 cm−1) can be used as an indicator of the cis conformation of carotenoids, which is to be ~0.2 for the 9′-cis isomer (Koyama and Fujii 1999). In LL cells, the relative intensity was between 0.05 and 0.06 and was slightly smaller than the reported values ~0.2. These discrepancies may be due to the coexistence of all-trans forms of carotenoids, such as Vx, Sx, Sn, and α-carotene. This interpretation is in reasonable agreement with the HPLC results that showed the amount of 9cNx relative to total all-trans carotenoids was smaller in HL cells than in LL cells.

The ν3 regions of the Raman spectra are shown in Fig. 5b. Each band comprised four bands at 1,003, 1,006, 1,008, and 1,010 cm−1. Several frequencies in this region have been empirically assigned to the xanthophylls bound to the LHCII; for example, 1,005–1,006 cm−1 indicates 9cNx, 1,007 cm−1 indicates Vx, 1,003–1,004 cm−1 indicates lutein, and 1,003 cm−1 indicates zeaxanthin (Ruban et al. 2001). We could safely assign the 1,006 cm−1 band to 9cNx and 1,008 cm−1 to Vx. C. intricatum, however, do not contain either lutein or zeaxanthin, but they do contain Sn/Sx and α-carotene. It is rather difficult to predict the frequencies of the methyl-rocking modes of unknown carotenoids only based on the structure; hence we could not assign the origin of either the 1,003 or 1,010 cm−1 band.

The ν4 regions of the Raman spectra are shown in Fig. 5a. Based on precise analysis, the ν4 Raman patterns have been assigned to lutein bound to the aggregation/quenched state of the higher-plant type LHCII (Ruban et al. 2001). The ν4 Raman patterns of HL samples were similar to the characteristic ν4 Raman patterns of oligomeric LHCII (Fig. 5 in the reference of Ruban et al. 2000); these patterns include peaks at 953 and 965 cm−1. The ν4 Raman patterns of LL samples were similar to that of trimeric LHCII; in which the peak at 953 cm−1 was eliminated. The excitation energy dependences of the amplitude of ν4 band at 965 cm−1 relative to those of the ν1 band at 1,530 cm−1 were plotted (Fig. 6b). The increases in relative intensities of the excitation at 476.5, 501.7, and 528.7 nm clearly corresponded to the peaks in the second-derivative spectra at 476.5, 500.0, and 542.0 nm, respectively. The differences between these absorptive transitions were ~1,000 cm−1 for 476.5/500.0 nm peaks and ~1,500 cm−1 for 500.0/542.0 nm peaks, indicating that those peaks represented a vibrational structure of the binding Sx/Sn. Table 3 summarizes the assignments of the absorptive transitions. Notably, all the ν1 frequencies were around 1,528 cm−1 when excited at these three excitation wavelengths (476.5/501.7/528.7 nm) (Fig. 6a); these observations confirmed that the ν1 frequencies were indicative of Sx/Sn based on the similarity to the ν1 frequencies of approximately 1,529 cm−1 associated with fucoxanthin bound to FCP (vide infra, Premvardhan et al. 2009). The ν4 patterns of these samples were very similar to those of lutein bound to LHC II (Ruban et al. 2000); consequently, we conclude that Sx/Sn occupies the similar binding site in SCP that is normally occupied by lutein in LHCII. Sx has been considered to be biosynthesized from lutein by two-step oxidations. Actually the skeleton structure of Sx has many similarities with lutein. Therefore, the ν4 modes may originate from the common modes shared by Sx, Sn, and lutein.

Table 3 Assignment of absorptive transitions that occurred in intact filamentous C. intricatum cells
Full size table

Plausible model of pigment regulation in response to HL irradiation

Under HL conditions, filamentous C. intricatum cells accumulate extra tNx and increased tVx and α-carotene. Additionally, low-temperature absorption spectra and the ν4 modes of Raman lines indicated that antenna size increased under HL conditions. Nx is biosynthesized from Vx; therefore, Nx can be accumulated when Vx is accumulated. Vx can be overexpressed in order to enhance the xanthophyll cycle to protect against high light stress. However, tNx, tVx, or both that had accumulated in HL cells exhibited higher ν1 frequencies; these findings indicate that interactions between these pigments and the proteins are weak. Therefore, the extra tNx and tVx may exist in the surface of the SCP to promote oligomerization of the SCP to control the amount of energy transferred from SCP to PSII by adjusting the distance between the energy donor and the energy accepter in order to quench the excess amount of excitation energy.

Conclusion

We used a combination of second-derivative of low-temperature absorption spectroscopy and excitation wavelength dependence of resonance-Raman spectroscopy to investigate the absorptive transitions that occur in an intact filamentous cell of a siphonous green algae, C. intricatum OKAMURA. The absorption spectrum of the intact filamentous cells represents that of the major light-harvesting protein, SCP. Based on these analyses and especially on that of the ν4 Raman patterns, we concluded that the characteristic carbonyl carotenoids, Sx or Sn, may occupy the similar binding site in SCP that lutein normally occupies in LHCII.

Additionally, these cultured cells accumulate tNx as found in higher plant photosynthetic systems.. This accumulation occurred only when cultures were exposed to extremely strong illumination (500 μE). The tNx accumulation that occurred in HL cells was confirmed both by LC analysis of extracted pigments and by resonance Raman spectroscopy analysis of intact cells. The accumulation of tNx in HL cells was accompanied by the accumulation of Vx and α-carotene. Resonance Raman spectroscopy also indicated that the assembly of the SCP differed between LL and HL cells; specifically, the SCP was trimeric in LL cells and oligomeric in HL cells. We proposed a plausible model for adaptation to high-light conditions; nevertheless, the distribution of tNx and excess Vx in the SCPs must be investigated to test the validity of this hypothesis.