Guard cells of onion irradiated with broad-band blue light display a green intrinsic fluorescence. The fluorescence has been found in eleven species of Allium, but it has not been observed in any other monocot or dicot examined. The fluorescence occurs only in guard cells and is absent in neighbouring epidermal cells. During development it is first apparent in guard mother cells soon after the asymmetric division. Microscopic observation reveals that the fluorescence is associated with the vacuole and examination of vacuoles isolated from guard cell protoplasts suggests that it may be localized on the tonoplast. Microspectrophotometric analysis of single cells reveals an emission peak at around 520 nm. Our results are consistent with the view that this blue light receptor is a flavin or flavoprotein and that it might be related to the blue light-enhanced stomatal opening observed in onion.
During the course of our studies on the development and physiology of stomata we discovered that when guard cells of onion (Allium cepa L.) were irradiated with blue light they fluoresced green. This observation seems important because it directly indicates the presence of a blue-light receptor in guard cells and thus might help us to understand the basis for the blue light-mediated stomatai opening reported in many species (see Hsiao, 1976, for review). Blue light has been reported to enhance stomatai opening in onion (Meidner, 1968), and recent studies in our own laboratory show that blue light also causes ion and water influx into guard cell protoplasts (Zeiger & Hepler, 1977).
The presence of the intrinsic fluorescence may be another manifestation of a basic blue light-dependent system capable of driving ion transport in the guard cells (Zeiger, Moody, Hepler & Varela, 1977) and thus may provide new avenues for experimentation on its mechanism of action. In this paper we report on the localization of the fluorescence in onion guard cells, on some of its spectral properties, and on its intra- and intergeneric distribution.
MATERIALS AND METHODS
Seeds of Allium cepa cv. Cima Hybrid (Keystone, Co., Ca., U.S.A.) were germinated as described previously (Zeiger & Hepler, 1976). Sets from A. sativum were bought from a local supplier and grown in vermiculite in a greenhouse; leaves of A. bivalve were obtained from the collection of alpine plants at the University of Massachusetts. Seeds of other Allium species were kindly supplied by Dr G. D. McCollum (U.S. Department of Agriculture, Beltsville, Md., U.S.A.). Leaves from mature plants of Tradescantia fluminensis, Agapanthus sp., Iris sp. and Zantedeschia aethiopica were dissected from ornamental plants growing in the surroundings of the laboratory. A mature specimen of Aspidistra eliator was bought from a commercial nursery. Hordeum vulgare cv. ‘Early Bonus’, Zea mays cv. Bear Hybrid, Viciafaba and Nicotiana tabacum var. Mammoth were grown from seeds in Petri dishes layered with moistened filter paper at room temperature.
Peels from leaves were obtained by tearing off small portions of the epidermis with a pair of fine forceps. Peels largely free of underlying mesophyll tissue were required to avoid interference by the intense red fluorescence of the chlorophyll in mesophyll cells. The peels were uncurled and mounted in a drop of mannitol solution (0·15 to 0·23 M) on a glass slide under a No. 1 coverglass. The osmoticum was used because, in contrast to preparations in distilled water the protoplasm of guard cells kept close to isotonic conditions continues to stream for several hours (Zeiger & Hepler, 1976). Peels and protoplasts of A. cepa used for quantitative measurements were kept in Sykes Moore culture chambers (Bellco Glass Inc., Vineland, New Jersey, U.S.A.) as described previously (Zeiger & Hepler, 1976). The solutions were perfused through the chambers with syringe needles inserted through the o-ring.
Protoplasts were obtained from slices of 6-to to-day-old cotyledons by digestion with 4% Cellulysin (Calbiochem, Los Angeles, Ca.) in 0·23 M mannitol for 10–12 h (Zeiger & Hepler, 1976). At the end of the digestion, the enzyme solution was replaced by 0·4–0·5 M mannitol and o·5 mM CaCl2. Changing solutions washed away most of the mesophyll protoplasts and thus eliminated background red fluorescence. For the release of isolated vacuoles, chambers with the isolated protoplasts were perfused with 5 to 10 ml of a 0·23 M mannitol solution. Swelling and bursting of the protoplasts occurred within a few minutes.
Two optical systems were used for the fluorescence microscopy: (1) a Reichert Zetopan microscope with a dark-field fluorescence condenser, a 200-W Hg lamp, a BG12 exciting filter (maximum at 410 nm) and a GG9/OG530 barrier filter, 50% cutoff point at 500 nm; and (2) an American Optical H1oTU-VF4 microscope with an AO 2070C vertical illuminator for incident fluorescence and a 50-W Hg lamp, with either (a) a BG12 exciting filter, a 500-nm cutoff dichroic beamsplitter and a OG515 barrier filter, or (b) a 436-nm blue exciting filter and a 450-nm cutoff dichroic beamsplitter with or without an OG515 barrier filter. Transmission curves for the latter combination are shown in Fig. 8, p. 6. Light intensities under the conditions described in (a), measured with the blue-light window of a Plant Growth Photometer (International Light, Newburyport, Massachusetts) at the level of the preparation were 0·55 mW cm−2 with the 40 × objective and 3 mW cm−2 with the 100 × objective.
Emission spectra and fluorescence decay of single cells were studied with a NanoSpec/10 microfluorospectrophotometer (Nanometrics, Sunnyvale, Ca., U.S.A.) attached to the AO optical system. The NanoSpec has a high sensitivity gallium-arsenide photomultiplier and a motor-driven diffraction grating monochromator allowing readings to the nearest 01 nm. The relative fluorescence intensity is displayed in a digital read-out and a chart recorder. Preparations on microscope slides or in chambers were placed on the stage of the microscope and observed under dim green light (green interference filter, λ = 546 nm, Reichert) using the built-in halogen lamp. Cells selected for active protoplasmic streaming and normal cytological appearance were centred in the optical field of the slit viewing system of the spectrophotometer head and the slit was adjusted to expose about 50 μm* of cell area. The halogen lamp was then switched off and the exciting beam from the Hg lamp was unblocked. Emission spectra were obtained by mechanically advancing the motor-driven monochromator between 2 preselected wavelengths. The relative fluorescence intensity at each wavelength was registered by a chart recorder. For the measurement of fluorescence decay, the monochromator was set at 530 nm, and the changes in intensity with time were followed by the chart recorder with zero time set at the instant the cells were exposed to the exciting beam.
Cytology of fluorescing guard cells and their isolated protoplasts and vacuoles
Onion guard cells show a green, intrinsic fluorescence (Fig. 2) when excited with broad-band blue light. Examination of epidermal peels and paradermal slices reveals that the green fluorescence is restricted to the guard cells; epidermal cells do not fluoresce (Figs, 1, 2). In paradermal slices that include mesophyll cells there is a pronounced red fluorescence due to chlorophyll. Red-fluorescing chloroplasts are also observed in the guard cells (Fig. 2).
In order to determine the intracellular location of the fluorescing compound, cells were examined at high magnification, and direct comparisons were made on the same cells with both fluorescence and Nomarski differential interference-contrast optics. From these observations it is readily apparent that the green fluorescence does not emit uniformly from throughout the cell; rather it is confined to those regions occupied by the vacuole. The nucleus, usually centrally located, is always opaque. The cytoplasm and plasmalemma also appear to lack the green fluorescence since the chloroplasts, which can be distinguished by their characteristic red fluorescence, are observed at the outer boundary of the green emission (Fig. 2). Fully mature guard cells show, in addition, a faint green fluorescence in the ridge surrounding the stomatai pore. Its slightly different colour from the fluorescence of the vacuole and its position in the ridge suggest to us that it originates from a different molecular species, probably lignin within the thickened cell wall.
Wall-less protoplasts from onion guard cells, kept in a 0·4 M mannitol and 0·5 mM CaCl2, also exhibit the intrinsic fluorescence (Fig. 3). Because of the rearrangement of the subcellular organelles in the spherical protoplasts, the green fluorescence is sometimes masked by bleached, yellowish chlorophyll.
Exposure of the protoplasts to an hypotonic medium like a 0·23 M mannitol solution causes them to swell and burst. By alternately observing such a preparation under bright-field and incident-fluorescent illumination, it can be seen that the intact swollen vacuole emerges through the broken plasmalemma and that it exhibits all of the green fluorescence (Fig. 4). An isolated fluorescing vacuole is seen in Fig. 5. Prolonged exposure to the hypotonic medium leads, in turn, to the bursting of the isolated vacuole. Unlike the plasmalemma, which seems to collapse upon rupture, the tonoplast reforms into smaller vesicles of varying sizes. Qualitative observations indicate that most, if not all, of the green fluorescence remains confined to the vesicles, but quantitative measurements remain to be done.
The fluorescing properties of developing guard cells
Developing guard cells in epidermal peels from 5-to 8-day-old onion cotyledons (Zeiger & Cardemil, 1973) were tested for their ability to fluoresce when excited with blue light. Guard cells showed the characteristic fluorescence at all stages of development. The fluorescence was faint in newly formed guard cells and increased at later stages of development. Guard cells from peels of leaves dissected from older plants growing in a garden also fluoresced, although the intensity was weaker than the one exhibited by mature guard cells from young cotyledons. The green fluorescence was also seen in a dividing guard mother cell found in prophase; it was found in the corners of the cell, areas where the vacuoles would be expected to be located. At an even earlier developmental stage very young guard mother cells, observed soon after the completion of the asymmetrical division at which they originated (Zeiger & Cardemil, 1973), showed a faint, yet unequivocal green fluorescence. On the other hand, their sister cells which are to remain epidermal, did not fluoresce. These observations indicate that the fluorescing compound was synthesized or became physiologically competent at a very early stage of guard cell differentiation.
Observations made on etiolated preparations suggest that the ability of the guard cells to fluoresce is not dependent upon illumination during growth. Seeds were germinated in complete darkness, and peels from 5-to 8-day-old cotyledons were made under dim green light. The preparation was observed immediately after illumination with blue light, and the fluorescing pattern was indistinguishable from that of peels grown in the light.
Emission spectra and fluorescence decay
The availability of a sensitive microfluorospectrophotometer allowed us to make some quantitative measurements of the green fluorescence in single cells. Cells scanned between 400 and 600 nm showed peaks in the blue and in the green regions of the spectrum. Qualitative observations made with different filter combinations indicated that the blue component resulted from residual transmission of the exciting beam and from light scattered by the cell walls.
Spectra from the 500 to 600 nm region, obtained with a narrow-band, exciting blue filter, a dichroic beamsplitter and a barrier filter, resolved a single peak in the green, with a maximum at 525 to 530 nm (Fig. 7). Exclusion of the barrier filter increased the blue background substantially but also caused a shift of the green peak to around 520 nm (Fig. 7), indicating that the transmission properties of the barrier filter (Fig. 8, inset to Fig. 7) produced an artifactual displacement of the green peak toward a longer wavelength.
It was also found that the relative intensity of the fluorescence decreased with time of exposure of the preparation to the exciting beam. Because of that fluorescence decay, successive spectra from the same cell had decreasing overall intensities, but the position of the green peak was always the same. That the kinetics of decay do not influence the emission spectrum is also apparent from the observation that scans of cells done from 500 to 600 or from 600 to 500 nm gave identical peaks.
Some quantitative parameters of the fluorescence decay were also obtained. Guard cells from epidermal peels kept in a 0·23 M mannitol solution in sealed microchambers (chamber volume: 0·65 ml) were irradiated with blue light (BG12 exciting filter 500-nm dichroic beamsplitter and OG515 barrier filter, 100 × oil-immersion objective) and the relative fluorescence intensity was measured at 530 nm. The fluorescence reached a maximum in less than o·5 s, the time lag of the chart recorder, and decayed continuously thereafter with approximately exponential kinetics. Typically, 50% of the initial intensity was lost within 1 min. Exposed cells, kept in the dark for periods of up to 5 h, failed to recover any of the lost intensity in spite of their continuous protoplasmic streaming. The rate of decay decreased substantially with decreased intensity of exciting light attained by using either a 40 × objective or a double BG12 exciting filter.
The intragenic and intergeneric distribution of the green fluorescence
All eleven species of the genus Allium tested (A. ascalonicum, bivalve, cepa,jistulosum, galanthum, pskemense, roylei, sativum, schoenoprasum, vavilovii, and tuberosum) showed the green fluorescence in their guard cells. We therefore conclude that the fluorescence is a common characteristic of the Allium genus.
On the other hand, we were unable to find the green fluorescence in any other species of several monocotyledons and dicotyledons tested. These include: Hordeum vulgare, Zea mays (Gramineae); Zantedeschia aethiopica (Araceae), Tradescantia fluminensis (Commelinaceae; kindly identified by Dr J. Thomas, Stanford University); Aspidistra eliator (Liliaceae), Agapanthus sp. (Amaryllidaceae), Iris sp. (Iridaceae), Vicia faba (Leguminosae) and Nico tian a tabacum (Solanaceae). With the exception of the two Gramineae, all guard cells observed under the established optical conditions lacked the green fluorescence characteristic of Allium while exhibiting large, bright red-fluorescing chloroplasts and a faint fluorescence on the ridges surrounding the pore (Fig. 6). Guard cells from barley and corn, on the other hand, did show an abundant, green fluorescence. It was clear, however, that most, if not all, of the fluorescence originated from the thickened walls of the guard cells. Because of the special morphology of the gramineous stomata, with their thick walls occupying most of the cell volume, it was especially difficult to ascertain whether a protoplasmic green fluorescence was present.
The green, intrinsic fluorescence found specifically in onion guard cells directly indicates the existence of a blue light photoreceptor in these cells. The location of the fluorescent compound in the vacuole, and perhaps in the tonoplast, suggests that it might be associated with membrane-mediated active ion transport in the guard cells and thus offers a new approach to the study of some basic aspects of stomatai function. The spectral properties of the fluorescing compound are suggestive of a flavin, thus raising the possibility that it could be a mediator of electron transport in the guard cells and be connected with blue light-enhanced stomatai opening.
The evidence for the vacuolar localization of the fluorescing compound is compelling. First, direct observations reveal fluorescence in the region occupied by the vacuole, and not within the nucleus, or cytoplasm in general; second, isolated protoplasts fluoresce, thus eliminating the cell wall as a source for the blue light receptor; finally, rupture of the protoplasts shows that the isolated, intact vacuole retains all the fluorescence. The observations showing that burst vacuoles reform into many smaller vesicles all of which appear fluorescent, indicate that the fluorescing compound is likely to be located on the tonoplast.
The cellular specificity of the fluorescence is also clear. The fluorescence is not only absent from neighbouring epidermal cells, but can be detected in guard mother cells just emerging from the asymmetrical division at which they originate. Both the cellular specificity and its appearance at a very early stage of development are suggestive of a physiological role of the fluorescing compound in connexion with stomatai function.
The emission characteristics of the green fluorescence agree with the spectral properties of flavins and flavoproteins. Furthermore, recent experiments have provided information on the excitation properties of the fluorescing compound which show a distinct peak around 450 nm (Zeiger, unpublished). Similar spectra have been reported for the flavoprotein flavodoxin (Ghisla, Massey, Lhoste & Mayhew, 1975) and for presumed flavins from Avena coleoptiles (Zenk, 1967) and the eyespot of Euglena (Sperling-Pagni, Walne & Wehry, 1976). In both Avena and Euglena, the fluorescing compounds have absorption spectra in the blue that closely correlate with action spectra for phototropism (Zenk, 1967) and phototaxis (Sperling-Pagni et al. 1976) indicating that they could be primary photoreceptors. It should be noted, however, that some flavonoids also fluoresce in the yellow green (Geissman, 1955) and glycoside derivatives of the flavonol quercetin have been reported in Allium leaves (Harborne, 1965). However, in contrast to the compound described here, the fluorescing flavonoids do not seem to absorb at wavelengths longer than 400 nm. Furthermore, quercetin can be irradiated for a few hours without considerable photodestruction (Kaneta & Sugiyama, 1971), while the fluorescing substance in onion guard cells decays to about one half of its initial fluorescence intensity within 1 min. This latter characteristic is also consistent with flavins which are known to photodestruct rapidly in the absence of adequate levels of electron donors (Schmidt & Butler, 1976).
If the fluorescing compound is, indeed, a physiologically functional flavin, its presence in the guard cells could have important implications. We have postulated a light-sensitive proton motive force as the basic energy transducing mechanism driving ion transport in the guard cells (Zeiger & Hepler, 1977; Zeiger et al. 1977). Flavins can be reduced by blue light (Schmidt & Butler, 1976) and could mediate the initial photoreduction in a series of electron transfers that would generate an electrochemical gradient driving the uptake of potassium associated with stomatai opening. Such a possibility certainly warrants further investigation of the photochemical properties of the fluorescing compound which, if located in the tonoplast, would also provide another means of studying potential electrogenic activities at that membrane (Moody & Zeiger, 1978).
The restriction of the green fluorescence to the Allium genus is intriguing. The compound might be absent from guard cells of the other genera tested, or the cellular milieu in the latter might preclude its fluorescence under the experimental conditions used. Allium guard cells are unusual in their lack of starch (Schnabl & Ziegler, 1977), their small chloroplasts (compare Figs. 2 and 6) and their seemingly extreme requirement for Cl- as a counter-ion for K+ (Schnabl & Ziegler, 1977). On the other hand, it is well documented that Allium guard cells share many basic physiological properties classically associated with stomatai function, such as sensitivity to light and CO2 (Heath, 1952; Meidner & Heath, 1959), and the use of K+ as the main osmotic cation (Schnabl & Ziegler, 1977). Even in their special response to blue light (Meidner, 1968; Zeiger & Hepler, 1977), a phenomenon to which the green fluorescence could well be related, Allium stomata are not unique, since blue light-enhanced stomatai responses have been reported in a significant number of species, including Vicia (Hsiao, Allaway & Evans, 1973), many grasses (Johnsson, Issaias, Brogårdh & Johnsson, 1976; Skaar & Johnsson, 1978; Raschke, Hanebuth & Farquhar, 1978) and Aspidistra (Voskresenskaya & Polyakov, 1976).
Hence, while it is clear that Allium guard cells show some unique characteristics that probably reflect specific variations in their functioning, the fact that they exhibit many basic responses common to stomata of most of the higher plants suggests to us a single, basic cellular mechanism driving stomatai function in all of them. If that is the case, the further characterization of the green-fluorescing compound in onion provides us with a unique opportunity to enrich our understanding of the photo-biological properties of the guard cells.
We are grateful to Vince Coates of Nanometrics and Steve Westrate of Westrate Scientific for generous help with equipment, to Drs E. J. Stadelmann (University of Minnesota) and S. Malkin (Weizmann Institute, Israel) for valuable suggestions, to Dr G. D. McCollum (U.S. Department of Agriculture at Beltsville, Maryland) for the supply of onion seeds, to Dr J. Brown, Carnegie Institution, Department of Plant Biology, for her generous help with the fluorospectrophotometer and to Eleanor Crump for editing the manuscript. Supported by NSF Grants PCM 74-15245 to P.K.H. and PCM 77-17642 to H. Mooney.