Photosynthesis Research 25: 299-307, 1990. © 1990 Kluwer Academic Publishers. Printed in the Netherlands.

Regular paper

Time-resolved spectroscopy of the blue fluorescence of spinach leaves Yves Goulas 1, Ismael Moya 2 & Guido Schmuck 3

ID~partement de Bioclimatologie, Institut National de la Recherche Agronomique, 84140 Montfavet, France; 2Laboratoire pour l' Utilisation du Rayonnement Electromagnetique, Universit~ de Paris XI, 91405 Orsay, France (for correspondence and~or reprints); 3joint Research Center, Ispra Establishment, 21020 Varese, Italy Received 16 August 1989; accepted in revised form 18 May 1990

Key words: algae, blue fluorescence, ferulic acid, flavonoid, photon counting, polyphenolics, spinach, time-resolved fluorescence Abstract

When excited by ultraviolet radiation, leaves of a great number of species of higher plants exhibit emission of blue fluorescence, comparable in intensity to the red emission of chlorophyll. The fluorescence decay of the blue emission of spinach leaves recorded by single photon counting techniques is decomposed into exponential components and it is shown that at least three different components are present. The lifetime of the three components does not show significant variations with the excitation or emission wavelengths. The excitation and emission spectra of each component were determined. The nature of the chemical compounds which cause this emission is discussed in relation to these spectra.

Introduction

In recent years, in vivo chlorophyll fluorescence has become a useful investigation tool in photosynthesis research, plant physiology and early stress detection. It also seems to have promising applications in remote sensing of vegetation by use of Laser Induced Fluorescence. However, chlorophyll fluorescence is not the only fluorescence signal exhibited by photosynthetic systems such as higher plants, algae or photosynthetic bacteria. When excited by ultra-violet light between 220 and 400 nm these systems also show a blue fluorescence emission around 440 nm. Several authors have established that under UV excitation, a fluorescence band with an emission maximum near 440 nm is detectable in baker's yeast, photosynthetic bacteria and algae, more probably due to bound reduced pyridine nucleotide (NADH, NADPH) (Duysens and Amesz 1957, Duysens and Sweep 1957, Olson et

al. 1959, Olson and Amesz 1960). They also showed that this fluorescence band has an excitation maximum at 280 and 340 nm. More recent experiments, using a nitrogen pulsed laser at 337 nm for the excitation, showed that the fluorescence spectrum of leaves and algae over the entire visible wavelength range, depends strongly on plant type (Chappelle and Williams 1987). These experiments showed that monocots can be distinguished from dicots by the ratio between the fluorescence at 440 and 685 nm and that hardwoods and conifers have an additional fluorescence band at 525 nm. These authors also claimed that a correlation between some kind of stress and the fluorescence spectra between 400 and 800 nm could be made in some cases. However, exploitation of the blue signal for physiological investigation or stress detection requires a better understanding of its nature and origin. In the present work, time resolved excita-

300

Materialsandmethods

tion and emission spectra of the blue fluorescence of spinach leaves have been obtained by means of synchrotron radiation. The light emission of the storage ring of Super-Aco (L.U.R.E., Orsay France) provides short pulses (400ps of duration) of white light at the frequency of 8 MHz. Such a source of sub-nanosecond pulses is well suitable to investigate the fluorescence properties of green plants using excitation ranging from 200 to 1000 nm.

Figure 1 shows a schematic representation of the apparatus used for recording both excitation and emission spectra and for single photon counting measurements under low light intensity. The experiments were performed on intact, attached spinach leaves (Spinacia oleracea L.). Some measurements were performed on the green algae Chlamydomonas reinhardtii in a quartz

plane \ mirror~

magnet A4

synchrotron ~diation

~

-------~,,1= ~ - ]

UV Filter excitation

~

synchronization.

photodiode

monochromator 220-400 nm . . .~l.

sampleholder '~'~visible ~ light f "

shutter

t

" mirror

;

filter detection

monochromator 400-800 m

Cooling system for photomultiplier

~d

, L t

J r

II:'="::°° I I

o,,,oo.oj

fast photomultiplier

Delay line iscriminators

converter

~

Lcompute~

I multichannel

I analyzer

Fig. i. Schematic diagram of the instrumentation used for time-resolved fluorescencespectroscopyon leaves and algae.

301 cuvette. The fluorescence decays were deconvoluted by means of least-squares program using the Marquardt search algorithm for non-linear parameters (Moya et al. 1986). Time-resolved emission and excitation spectra were determined using a m e t h o d previously described (Hodges and Moya 1986). Excitation spectra were corrected for the instrumental response by dividing each spectrum with the excitation spectrum measured on a saturating solution of rhodamine B (Lakowicz 1983). Emission spectra were not corrected for the instrumental response. Fluorescence was measured at 90 ° to the excitation beam. The incidence of the excitation beam to the plane of the leaf was about 30 ° in order to minimize the contribution of reflected light. Figure 2 shows the fluorescence decay of a spinach leaf recorded at 460 nm after excitation at 340 nm (Fe). The curve labeled G is the decay recorded when the scattered light is observed instead of the fluorescence (i.e., the instrumental response function). The time resolution in the time correlated photon counting technique, is usually considered to be about 1/10 of the width of the instrumental response function, which is about 800 ps in our experimental set-up, so components as fast as 80 ps can be detected. When the fluorescence lifetimes are not long compared to the duration of the instrumental

response function, we have to consider Fe(t) as the convolution of G(t) by the unknown fluorescence decay F(t): Fe(/) = G(/)*F(t) where t is the time and * designates the convolution product (i.e., S F(u). G ( t - u ) . du). F(t) is usually modeled by a sum of exponential components: F(t)=

'~

Fp(t)

p=l,n

where n is the number of elementary components. In the case of a simple dye, n = 1. The quality of the fit was estimated by the reduced Ki z which is defined by: Ki 2 = ~'~ [Fe(ti) - F(ti)]2/[N

where N designates the degrees of freedom ( = n u m b e r of points) (Bevington 1969) and t i is the sampled time. Because of the Poissonian nature of the noise in photon counting, this parameter must have a value close to 1, when the model is correct. In addition we plot the weighted residual R(t) defined by: R(t) = [Fe(t) - F(/)]/X/[Fe(t)],

o

,..,~

• Fe(/i)]

i

IX

5ooo

,,[

,.,--g

z

.~

u. tL2

rr -5

10

20

30

(ns}

Fig. 2. Fluorescence decay of a spinach leaf at 430 nm after excitation by a light pulse from Super-ACO (wavelength 300 nm). The figure shows the excitation flash (G), the experimental decay (Fe) and the recalculated fluorescence decay (F). The residual function (R) between the experimental curve and the calculated curve is also plotted.

302 which must display a fiat noise distribution over the whole decay (see Fig. 2). This has been checked using oxazine in methanol which exhibits a lifetime of about 800 ps as well as a reduced Ki 2 = 1 (not shown). In the case of blue fluorescence, models with one or two exponentials components were unable to fit data with the expected noise (Ki 2 ~ 1 and an uniform distribution of R(t), not shown). Three exponential components at least were necessary to get the best fit of the experimental data. Models with four exponential components do not give much better results than models with three exponential components and lead to addition of a long component of about 20 ns without changing the characteristics of the three other components.

0.8

.~

o.6 0.4 0.2

0 400

500 (nm)

550

600

(a)

0.8

~: o.6 m .~ 0.4 0.2

Results and discussion

450

F

u

\ •

250

-

-

300

350

400

(nm)

Figure 3a shows the fluorescence emission spectrum of the green algae C. reinhardtii in a quartz cuvette under excitation at 350 nm. This emission spectrum is similar to the fluorescence emission spectrum of Chlorella vulgaris recorded previously (Duysens and Amesz 1957). It was reported that the blue fluorescence of C. vulgaris and other photosynthetic systems increased under photosynthetic illumination (Olson and Amesz 1960). The excitation spectrum of this fluorescence increase was very similar to the fluorescence excitation spectrum of N A D P H bound to enzymes such as yeast alcohol dehydrogenase, liver alcohol dehydrogenase, and lactic dehydrogenase (maximum excitation wavelength near 280 and 345 nm). The excitation spectrum of C. reinhardtii (Fig. 3b) shows two maxima located near those of NADPH. If we take into account a modification of the shape of the spectrum caused by a contamination by other fluorescent materials in the algae as well as in the medium, blue fluorescence of C. reinhardtii is presumably due to N A D P H for a large part. This evidence for the observation of the fluorescence of N A D P H in green algae by these authors suggested to us that changes in N A D P H concentration could also be detected in leaves of higher plants. But our attempts to detect any

(b) Fig. 3. (a) Fluorescence emission spectrum of C. Reinhardtii excited at a wavelength of 350 nm. T h e fluorescence of the m e d i u m has been subtracted from the total fluorescence. (b) Excitation spectrum of the fluorescence at 420 n m of green algae C. Reinhardtii.

change in the blue fluorescence emission of leaves upon illumination with actinic light were unsuccessful. Possible unfavorable experimental conditions could explain these negative results. But since the ratio of the blue emission upon the red emission of the chlorophyll is much greater in leaves than in algae (Fig. 4a and b, Chappelle and Williams 1987) we suppose that the blue fluorescence of N A D P H in green leaves is hidden by a strong emission of at least one other fluorescent compound. Time-resolved spectroscopy The fluorescence emission of a particular compound in defined conditions is characterized by several parameters including the excitation spectrum, the emission spectrum and the fluorescence lifetime. A complex emission of fluorescence created by the superposition of several emissions coming from different compounds having almost the same excitation and emission

303

0.8 J2 .= o.6

s. o,4

Time-resolved spectroscopy of the blue fluorescence of spinach leaves.

When excited by ultraviolet radiation, leaves of a great number of species of higher plants exhibit emission of blue fluorescence, comparable in inten...
594KB Sizes 2 Downloads 0 Views