Red-light stimulation of prolamellar body recrystallization and thylakoid formation in barley etioplasts

When seedlings of most higher plants are grown in total darkness the proplastids develop into well denned organelles of considerable structural complexity. These organelles have been named ‘etioplasts’ (Kirk & Tilney-Bassett, 1967) and their fine structure in Avena saliva L. has been intensively investigated by Gunning and co workers (Brown & Gunning, 1965; Gunning, 1965 a,b; Gunning & Jagoe, 1965; Gunning, Steer & Cochrane, 1968). The most conspicuous feature of the etioplast is the prolamellar body, the typical form of which is a regularly arranged lattice of mem branes and tubules which appear paracrystalline in the electron microscope.

Gunning (1965a) considers the prolamellar body to be a simple cubic lattice of tubules in which the tubules are swollen at the points of fusion. In Avena sativa 99% of prolamellar bodies exhibited a regular crystalline appearance although non crystalline forms were found in young leaves (Gunning & Jagoe, 1965). In Hordeum vulgare L. however, a different situation exists. Von Wettstein (1958) observed that crystalline prolamellar bodies only rarely could be found. This work utilized per manganate as a fixative, in contrast to the glutaraldehyde-osmic fixations employed by Gunning. The work reported here was initiated as a routine investigation of barley etioplast structure to determine whether the differences in interpretation between von Wettstein and Gunning were real species differences, or were simply due to the methods used.

The effects of short periods of illumination on the structure of the prolamellar body have also been investigated by Virgin, Kahn & von Wettstein (1963), Gunning & Jagoe (1965) and Henningsen & Boynton (1969). The immediate effect of a brief illumination is the disruption of the crystallinity of the prolamellar bodies, such that the tubules become loosely and irregularly interconnected. In oat and wheat leaves (Gunning & Jagoe, 1965) the prolamellar bodies reform during the subsequent 2–3 h of darkness, and regain the virtual 100% crystallinity observed in fully etiolated leaves. Our preliminary observations had shown that only a small proportion of the prolamellar bodies in etiolated barley leaves were of the crystalline form, and this paper reports investigations into the reformation of barley prolamellar bodies following dis ruption with red light.

A further feature of the photocontrol of plastid development investigated here is the effect of pretreatment with red light on subsequent granum formation under con tinuous white light. Such pretreatment is known to remove the lag phase in chloro phyll synthesis (Mitrakos, 1961), and Henningsen (1965) has shown that in the absence of red-light pretreatment granum formation is directly correlated with chlorophyll synthesis. It therefore is of interest to determine whether such pretreatment with red light accelerates the rate of granum formation under subsequent high intensity light.

Hordeum vulgare var. Impala seeds were obtained from Carter’s Seeds, Ltd., London, soaked in running water for 4 h and sown in vermiculite. Plants were grown in a dark room at 25 ± 0·5 °C for 6 days before use.

The red-light source consisted of 3 ‘Double Life Red’ fluorescent tubes filtered through 2 layers of Cinemoid (Strand Electrics, Kingsway, London) No. 14 Ruby+ 1 layer of No. 1 Yellow (transmitting wavelengths from 600 to 690 nm). The far-red source consisted of 5 × 500-W tungsten bulbs with a 10-cm water heat filter, and Cinemoid filters as follows: 1 layer No. 5 A Deep Orange and 1 layer No. 20 Deep Blue Primary (685 nm to 1160 nm, with peak transmission at approximately 800 nm). All manipulations were performed in the dark room using a green safelight consisting of 1 × 15-W daylight fluorescent tube filtered through 3 layers of Cinemoid No. 39 Primary Green (500 nm to 575 nm with a peak at 530 nm). Energy distribution of the light sources was measured using an Isco spectroradiometer.

All fixations were carried out in the darkroom under the green safelight until com pleted. Pieces of leaf tissue 1 mm2 were excised, under the fixative, from the apical 1 cm of the 6-day-old barley leaf. For permanganate fixation the tissue pieces were maintained in 1 % KMnO4 in water for 60 min at o °C. They were then washed and stored in 70% ethanol. Glutaraldehyde fixation was in 5 % glutaraldehyde in 0 1 M sodium phosphate buffer, pH 70, containing 5 % sucrose, for 3 h at o °C. Osmium tetroxide fixation was performed by main taining the tissue pieces in 2 % osmium tetroxide containing barbiturate buffer, pH 7 · 0, and containing 5 % sucrose for 3 h at 4 °C, followed by washing and storing in 70 % ethanol. Tissue pieces were dehydrated, embedded in Araldite and sectioned using an LKB Ultrotome 3 micro tome. Post-staining was in uranyl acetate or lead citrate for 10 min. Electron microscopes used were the Siemens Elmiskop 1A and the Zeiss EM9. Films used were Kodak B4 in the Siemens microscope and Agfa/Gevaert Scientia in the Zeiss.

Apical 1 -cm leaf sections were weighed and ground in a pestle and mortar in 80% aqueous acetone. After 30 min extraction the homogenates were filtered, centrifuged and the absorption spectra of the supernatants recorded on a Unicam SP 800 spectrophotometer. The amounts of chlorophyll were calculated from the absorbancy values at 645 nm and 663 nm by the method of Arnon (1949).

Although glutaraldehyde-osmium tetroxide fixation is currently considered preferable to permanganate fixation for the preservation of cellular structure, it was nevertheless of interest to determine whether the differences in prolamellar body dimensions observed in barley by von Wettstein (1958) and in oat and wheat by Gunning (1965b) were due to bonafide species differences or were due merely to fixation differences. In Figs. 2–5 prolamellar bodies of barley etioplasts fixed in glutaraldehyde-osmium tetroxide are shown, whilst Figs. 11,12 contain similar sections fixed in permanganate. In both fixatives, the 2 major categories of prolamellar bodies, crystalline, and non crystalline, can be recognized.

A distinctive feature of 6-day dark-grown barley leaves compared to oat, is the relatively very low proportion of crystalline prolamellar bodies. In this communication the term crystalline is restricted to those prolamellar bodies having at least part of their structures composed of a regular tetrahedral lattice, as described by Gunning (1965a). Henningsen & Boynton (1969), also working with barley, used a much looser definition of crystallinity.

The mean proportion of crystalline prolamellar bodies in 6-day dark-grown leaves is 16–20%. In glutaraldehyde-osmium tetroxide-fixed material (Figs. 2–4) the typical formations of the crystalline prolamellar bodies as described by Gunning (1965a), namely, hexagons, waveforms, tubules, and rectangles, can be discerned. A further similarity with the structures found by Gunning (1965a) in oat and wheat is the loca tion of an electron-dense particle in the centre of the space between the tubules and considered by Gunning to be a ribosome. A notable feature of the glutaraldehyde osmium tetroxide-fixed material is the large osmiophilic particles or globules localized normally within the prolamellar body and often deforming the crystal lattice (Figs. 2,4).

Bundles of fine tubules which have been seen in both transverse and longitudinal section are another interesting feature only visible in glutaraldehyde-osmium tetroxidefixed material (Figs. 6–10). Similar structures have been described by Henningsen & Boynton (1969) and appear to be commonly associated with the developing thylakoids (Fig. 10). These structures have been termed tubules since in cross-section they present a hollow, hexagonal appearance (Figs. 6, 9). The resolution obtained in longitudinal sections, however, makes it difficult to exclude the possibility that the structures are linear aggregations of hollow particles (Figs. 7, 8), unlike other microtubules seen in plant preparations. The cross-sectional diameter of the structures is 16–19 nm, con siderably larger than the 6–9 nm shown by Gunning (1965b) for stromacentre material in oats. The exact nature of these particles or tubules is not known, although it is possible that they represent aggregates of Fraction I protein of a different form from the oat stromacentre.

In permanganate-fixed sections (Fig. n), the details of the prolamellar body lattice structure are not clearly visible although the general pattern is not inconsistent with that in glutaraldehyde-osmium tetroxide fixations. The unit cell length of the crystal line prolamellar body was found to be 58 nm in glutaraldehyde-osmium tetroxide fixations and 53 nm in permanganate fixations. These values are close to the 59 nm obtained for oats (Gunning, 1965a) but more than twice the 25 nm reported by von Wettstein (1958) for permanganate-fixed barley. The diameter of the tubules in both fixatives was 25–30 nm, compared with the 21 nm for oats (Gunning, 1965a) and the 6–9 nm for permanganate-fixed barley (von Wettstein, 1958). Thus the structure and dimensions of the crystalline prolamellar body in barley as visualized by glutaraldehyde-osmium tetroxide fixation and by permanganate fixation are con sistent with the model proposed for the oat prolamellar body by Gunning (1965a). No explanation can be offered for the disparity between these measurements and those made earlier by von Wettstein (1958).

The non-crystalline prolamellar bodies shown in Figs. 4 and 12 are similar, in both fixatives, to the ‘atypical’ forms of Gunning (1965a). In barley, it does not seem reasonable to describe this form as ‘atypical’, since it comprises the predominant proportion of all prolamellar bodies in dark-grown material. It should also be pointed out that Henningsen & Boynton (1969) consider this form to be a crystalline form with wide tube spacing. We prefer to restrict the term crystalline to cover only prolamellar bodies with regular lattices and the reasons behind this definition will become apparent when the effects of light are considered.

It is well known that light treatment of etiolated leaves leads to a rapid disruption of the prolamellar body followed by a slower, light-dependent formation of lamellae (Virgin et al. 1963; Henningsen, 1965). If however, only a short period of light is given, the prolamellar bodies disperse, only to reform over a period of 1–2 h in the subse quent darkness. The structural changes accompanying dispersal were thought by Virgin et al. (1963) to involve the conversion of the membrane material into vesicles, on the basis of permanganate-fixed sections of barley. With glutaraldehyde-osmium tetroxide fixations of oat leaves, however, Gunning & Jagoe (1965) have shown that dispersal is followed, in continuous light, by the formation of intergranal thylakoids directly from the prolamellar body material without an intervening vesicular stage. With only a short period of light, the prolamellar body becomes atypical (i.e. non crystalline), yet vesicles are not formed in the ensuing darkness.

Fig. 13 shows a barley prolamellar body fixed in permanganate 30 min after a 2-min exposure to red light. The prolamellar body has dispersed into apparent vesicles, which probably result from sectioning through perforated sheets, or grids. Figs. 14 and 15 show, respectively, glutaraldehyde-osmium tetroxide and permanganate fixa tions performed immediately after a 2-min red-light treatment. It is noteworthy that in Fig. 14 the prolamellar body still has an organized tubular structure although it is no longer crystalline, whilst in Fig. 15 all tubular structure has been lost and the perforated sheets have appeared. Although glutaraldehyde-osmium tetroxide fixation is more likely to retain plastid structure in its native form, the use of permanganate fixation here demonstrates that the brief light treatment must have brought about an immediate change in the chemical constitution of the membrane making up the pro lamellar body. It should be restated that tube structure in both crystalline and non crystalline prolamellar bodies can be observed in fully dark-grown leaves using permanganate fixation (Figs. 11, 12). Light must therefore change the membrane com position so that perforated sheet formation occurs on exposure to the permanganate fixative. The possible nature of these changes is considered below.

When leaves are given red-light treatment and then returned to darkness, pro lamellar body reformation occurs, culminating in oats in the reattainment of almost 100% crystalHnity as is typical of fully dark-grown leaves (Gunning & Jagoe, 1965). It was of interest, therefore, to determine to what degree crystalHnity was achieved in barley etioplasts treated in a similar fashion.

In permanganate-fixed barley leaves, reformation of the prolamellar body was first observed 90 min after exposure to 2 min of red light, and appeared to be complete within 3 h. Extension of the dark phase did not lead to a further increase in percentage crystallinity. Further experiments utilizing up to 4 repeated treatments of 10 s red light each followed by 3 h of darkness yielded marked ultimate increases in per centage crystallinity. This effect was not reversed by irradiating the leaves with 10 s of far-red light immediately following the red-light irradiations, nor did far-red on its own cause prolamellar body disruption. Maintenance of the leaves at o °C in ice during the dark period prevented the recrystallization of the prolamellar body. These results are presented in Tables 1 and 2 and representative electron micrographs of the extreme treatments are presented in Figs. 16, 17.

The effects of red light on prolamellar body structure are therefore 2-fold. There is an immediate disorganization of the crystalline structure followed by its slow reformation. The second process of induced crystallization appears to be potentiated in a cumulative manner by pretreatment with red light but is not regulated by phyto chrome. The interpretation of these changes appears to reside in the as yet incom pletely understood changes occurring in the prolamellar material upon light treatment. The immediate change in the membranes under light treatment results in lability of the regular crystalline lattice, particularly striking in permanganate fixation, yet the long term effect is to change the membrane in such a way that it becomes more con ducive to ultimate crystallization.

Virgin et al. (1963) and Henningsen & Boynton (1969, 1970) have shown strong correlations between the photoreduction of protochlorophyllide to chlorophyllide and the disorganization of the prolamellar bodies in barley. The action spectrum for vesicle formation in permanaganate-fixed material is similar to that for protochloro phyllide conversion (Virgin et al. 1963). Boardman & Anderson (1964) have shown that protochlorophyll is a constituent of the prolamellar body and it is thus reasonable to propose that protochlorophyllide conversions are involved in the disorganization phenomenon. Protochlorophyllide is re-synthesized in darkness after photoreduction by a flash of red light, reaching the regulated dark level within about 20 min (Augus tinussen & Madsen, 1965). It seems unlikely, therefore, that protochlorophyll resynthesis is the only limiting step in the reformation of crystalline prolamellar bodies since reformation does not begin until at least 1 h after protochlorophyll resynthesis has been completed. Furthermore, the protochlorophyll content does not rise above the level found in dark-grown controls even after serial light treatments (Augustinus sen & Madsen, 1965), yet the proportion of crystalline prolamellar bodies reaches more than 3 times the dark level. Protochlorophyll synthesis thus must precede the reformation of the prolamellar body and may well be a prerequisite for it but cannot directly account for the increase in crystallinity. It is possible, however, that the biosynthesis or regeneration of the holochrome and its association with protochlorophyll may be slower processes more intimately involved with the formation of the stable crystalline lattice. Indeed, Khan (1968) has postulated a direct reduction of proto chlorophyllide by the holochrome, associated with structural changes.

It is clear from Table 1 that the process of reformation requires metabolic activity and is presumably energy-requiring. Gunning (1965a) has postulated that the pro lamellar material is laid down around the prolamellar body ribosomes, the regular structure being determined by a regular orientation of the ribosomes in the stroma prior to prolamellar body formation. It is not easy to explain the induced crystalliza tion observed in barley on this basis, since it would be necessary to postulate a red light induced rearrangement of ribosomes. It seems more likely that the capacity to form a regular lattice is a property of the membrane subunits and that the act of crystal lization is a physical process, resembling the assembly of the protein coat of a virus, on a somewhat larger scale. The requirement for energy is likely to be related to the metabolic events preceding crystallization, as perhaps protochlorophyllide and holo chrome synthesis.

Pretreatment of 6-day dark-grown barley leaves with red light leads to a marked diminution of the lag phase in chlorophyll synthesis when the leaves are subsequently transferred to continuous white light (Fig. 1 and Mitrakos, 1961). In Figs. 18, 19 the effect of red-light pretreatment on granum formation in subsequent white light is shown at a point 3 h after the commencement of the white-light treatment. A much higher frequency of recognizable grana is apparent in the red-light treated section, in fact, granum formation does not begin in non-pretreated leaves until 3 h of white-light treatment has been received. In order to quantitate this effect, the frequency of grana containing more than 2 thylakoids was counted per unit area of plastid section. This method was used since some double membranes occurred which did not resemble normal thylakoids (Bachman, Robertson & Bowen, 1969; Henningsen & Boynton, 1969). Such double membranes were thus not scored as grana. The results for the first 12 h of the white-light treatment are presented in Table 3. The difference between the two series is relatively large between 3 and 6 h of light treatment, but is negligible at 12 h. The correlation with the time-course of chlorophyll synthesis is good.

These results are consistent with the generally accepted view that chlorophyll synthe sis and granum formation are closely associated and interdependent processes (Virgin et al. 1963). They also raise the possibility that the red-light sensitive regulatory step in chlorophyll synthesis may determine a change in the structure of the membranes rather than being a direct biochemical step in the biosynthetic pathway of chlorophyll. In view of the postulated light-mediated changes in the membrane structure of the prolamellar body discussed above, it is tempting to explain the acceleration of granum formation on the basis of these changes. Experiments are in hand to assess the effects of serial red-light treatments on subsequent chlorophyll synthesis to determine if a relationship between the proportion of crystalline prolamellar bodies and the time course of synthesis exists. Such a relationship would indicate that the red-light medi ated changes in membrane structure conducive to recrystallization are also conducive to subsequent granum formation.

A difficulty with this interpretation is that the red-light reduction of the lag phase in chlorophyll synthesis is partially reversible by far-red light, indicating the in volvement of phytochrome (Mitrakos, 1961), whereas the effects on prolamellar body recrystallization probably do not involve phytochrome. In the reduction of the lag phase in greening, however, only about 60% reversal is normally achieved with far red light, suggesting that phytochrome is only partially responsible for the red-light effect. The methods used here for estimating the rate of granum formation do not allow us to distinguish reliably such small differences and thus the possibility of a phytochrome involvement in the acceleration of granum formation cannot be ascertained.

It is clear that much yet needs to be learnt about the composition and structure of the etioplast and chloroplast membranes before fully meaningful interpretations of the phenomena described in this communication can be derived.

D.R.B. was a grateful recipient of a Drapers Company Research Studentship. The technical assistance of Mr J. Pacey and the staff of the Biological Electron Microscope Unit at Queen Mary College is gratefully acknowledged. The authors thank Professor E. A. Bevan for his continuing support and encouragement.