Multiphoton fluorescence lifetime imaging ... - Wiley Online Library

8 downloads 8790 Views 1017KB Size Report
Jan 3, 2012 - ∗School of Botany, University of Melbourne, Australia. †School of .... squares iterative reconvolution routine to individually fit the fluorescence ...
Journal of Microscopy, Vol. 247, Pt 1 2012, pp. 33–42

doi: 10.1111/j.1365-2818.2011.03593.x

Received 30 August 2011; accepted 3 January 2012

Multiphoton fluorescence lifetime imaging shows spatial segregation of secondary metabolites in Eucalyptus secretory cavities A . M . H E S K E S ∗ , C . N . L I N C O L N †, ‡, J . Q . D . G O O D G E R ∗ , I . E . W O O D R O W ∗ & T . A . S M I T H †, ‡ ∗ School of Botany, University of Melbourne, Australia

†School of Chemistry, University of Melbourne, Australia ‡ARC Centre of Excellence for Coherent X-Ray Science, University of Melbourne, Australia

Key words. Essential oil, eucalypt, FLIM, monoterpene, oil gland, oleuropeic acid.

Summary Multiphoton fluorescence lifetime imaging provides an excellent tool for imaging deep within plant tissues while providing a means to distinguish between fluorophores with high spatial and temporal resolution. Ideal candidates for the application of multiphoton fluorescence lifetime imaging to plants are the embedded secretory cavities found in numerous species because they house complex mixtures of secondary metabolites within extracellular lumina. Previous investigations of this type of structure have been restricted by the use of sectioned material resulting in the loss of lumen contents and often disorganization of the delicate secretory cells; thus it is not known if there is spatial segregation of secondary metabolites within these structures. In this paper, we apply multiphoton fluorescence lifetime imaging to investigate the spatial arrangement of metabolites within intact secretory cavities isolated from Eucalyptus polybractea R.T. Baker leaves. The secretory cavities of this species are abundant (up to 10 000 per leaf), large (up to 6 nL) and importantly house volatile essential oil rich in the monoterpene 1,8-cineole, together with an immiscible, non-volatile component comprised largely of autofluorescent oleuropeic acid glucose esters. We have been able to optically section into the lumina of secretory cavities to a depth of ∼80 μm, revealing a unique spatial organization of cavity metabolites whereby the non-volatile component forms a layer between the secretory cells lining the lumen and the essential oil. This finding could be indicative of a functional role of the non-volatile component in providing a protective region

Correspondence to: A.M. Heskes. School of Botany, The University of Melbourne, Parkville, Victoria 3010, Australia. Tel: +61-3-83445218; fax: +61–3-93475460; e-mail: [email protected]

 C 2012 The Authors C 2012 Royal Microscopical Society Journal of Microscopy 

of low diffusivity between the secretory cells and potentially autotoxic essential oil. Introduction Fluorescence microscopy has become an indispensible tool in the botanical sciences, providing a highly sensitive means of elucidating tissue structure (Wuyts et al., 2010), localizing metabolites (Hutzler et al., 1998; Lakowicz et al., 1992) and monitoring dynamic processes (e.g. diffusion of molecules; Lippincott-Schwartz & Patterson, 2009). Multiple fluorescence properties of emitting species can be utilized to provide enhanced image contrast including emission spectra, fluorescence polarization and fluorescence lifetime (Levitt et al., 2009). In particular, fluorescence lifetime imaging microscopy (FLIM) is one of the most informative techniques due to the high sensitivity of a fluorophore’s lifetime to its surrounding microenvironment. This property has made FLIM amenable to diverse applications at the subcellular level including studies on viscosity (Suhling et al., 2004), pH and ion concentration (Babourina & Rengel, 2009; Vroom et al., ¨ 1999) and enzyme interactions (Bucherl et al., 2010). An extension of this technique which has gained popularity in the life sciences (Periasamy & Clegg, 2010; Smith et al., 2009) is FLIM using multiphoton excitation (MP-FLIM). Multiphoton excitation uses near infrared wavelengths, which are able to penetrate deeper into biological tissue than UV wavelengths due to the relative lack of endogenous absorbers and the reduced scattering of near infrared light compared to UV (Helmchen & Denk, 2005). In addition, the nonlinear dependence of multiphoton absorption on the excitation light intensity results in inherent optical sectioning (Denk et al., 1990), often resulting in higher quality images obtained from deeper within materials, particularly biological tissue, than

34

A. M. HESKES ET AL.

those obtained through conventional laser scanning confocal microscopy (Centonze & White, 1998; Feij´o & Moreno, 2004; Helmchen & Denk, 2005). Benefits have also been reported in terms of reduced photobleaching and increased cell viability due to both the limited excitation volume and less harmful near infrared wavelengths used in multiphoton excitation (Feij´o & Moreno, 2004). In the botanical sciences MP-FLIM has been taken up relatively slowly compared to the other areas of life science such as the biomedical field. The majority of studies on plant systems have used MP-FLIM to detect the close interaction of fluorescently labelled proteins in protoplasts and leaf epidermal cells via F¨orster resonance energy transfer (Aker et al., 2007; Bernoux et al., 2008; Boutant et al., 2010). In this application the benefits of MP-FLIM are the increase in signal to noise ratio which raises the detection limit of the labelled proteins as well as offering improvements in cell viability. However, as these cells are on the surface of plants and lack strongly absorbing pigments, the full power of multiphoton excitation’s increased depth penetration of plant tissue has not been realized. Nonetheless, there are limited plant studies that have used this technology to obtain detailed lifetime information from fluorophores embedded deep within intact plant tissue. For example, MP-FLIM has been used to successfully determine the excited state kinetics of lightharvesting complexes in chloroplasts situated throughout the leaf lamina of Alocasia wentii (Broess et al., 2009) and also used to quantify the flux of pH and ions with high spatial resolution in intact roots of Arabidopsis (Babourina & Rengel, 2009; Guo et al., 2009). Prime targets for the application of MP-FLIM in plants are the secretory cavity complexes found in many plant families, such as the essential oil cavities of Rutaceae (e.g. Citrus spp.) and Myrtaceae (e.g. Eucalyptus and Melaleuca spp.) and the resin ducts of conifers. Secretory cavities are structurally and metabolically complex comprising a large extracellular lumen (in which the secretory products, e.g. essential oils, are stored), encased by single or multiple layers of specialized secretory cells (Bosabalidis & Tsekos, 1982b; Carr & Carr, 1970; Kalachanis & Psaras, 2005; Nagy et al., 2000; Rodrigues et al., 2011). Their embedded nature has traditionally meant that the only way to investigate their structure and function has been through tissue sectioning resulting in the loss of lumen contents and, particularly in the case of essential oil cavities, disorganization of the bounding secretory cells (Kalachanis & Psaras, 2005; Turner et al., 1998; Yamasaki & Akimitsu, 2007). MP-FLIM, with its potential for superior penetration of plant tissue, provides the opportunity to overcome these problems and moreover, to study the organization of the intact metabolome of cavity lumina. These structures are known to produce and store complex mixtures of secondary metabolites, many of which exhibit autofluorescence and the use of timeresolved emission detection can potentially enable fine scale discrimination between different fluorescent species.

This latter advantage shows particular promise for studies on species of Eucalyptus, where the secretory cavities have recently been shown to house a highly fluorescent non-volatile component (NVC) comprised largely of oleuropeic acid glucose esters in addition to 1,8-cineole rich essential oil (Goodger et al., 2009). Nothing is known of the biosynthesis, transport or functional role of the NVC in eucalypt secretory cavities, but the application of MP-FLIM has the potential to aid in characterizing this metabolically complex system. Herein we report for the first time the successful application of MPFLIM with time-resolved emission detection to visualize the lumen of Eucalyptus secretory cavities, and use the technique to not only differentiate between the different classes of secondary metabolites housed within lumina, but to also show a unique level of spatial organization within these extracellular domains. Materials and methods Plant material Eucalyptus polybractea R.T. Baker leaves were harvested from plantation grown saplings (for plantation details see Goodger & Woodrow, 2009). Fully expanded leaves were used in all cases as they are known to contain high levels of both essential oil and an NVC in large, isolatable secretory cavities (Goodger et al., 2009, 2010). Sample preparation Secretory cavities were isolated from leaf tissue using a pectinase leaf digestion protocol as described in Goodger et al. (2010). Isolated cavities were mounted in buffer (500 mM sorbitol, 5 mM MES-KOH, 1 mM CaCl2 , pH 5.5) on a concave glass microscope slide and covered with a cover slip, which was then sealed around the edges with VALAP (Vaseline, lanolin and paraffin wax in 1 : 1: 1 ratio). Non-enzymatically isolated cavities were viewed by dissecting away as much surrounding leaf tissue as possible without rupturing the cavity and mounting them as for enzymatically isolated cavities. NVC was extracted from fully expanded leaves directly from cavities with the use of a microprobe with 1 μm tip (World Precision Instruments Pty Ltd., Sarasota, FL, U.S.A.). The extract was smeared directly onto a cover slip and imaged. Essential oil was steam distilled from a bulk leaf sample (for extraction details see Goodger et al., 2007). Fluorescence spectrometry Steady-state fluorescence excitation and emission spectra were collected for extracted NVC and essential oil using a Cary Eclipse fluorescence spectrophotometer (Varian, Inc., Palo Alto, CA, U.S.A.) with a spectral bandwidth of 5 nm. Excitation and emission spectra were collected from samples dissolved in  C 2012 The Authors C 2012 Royal Microscopical Society, 247, 33–42 Journal of Microscopy 

MP-FLIM OF EUCALYPTUS SECRETORY CAVITIES

acetonitrile (100%) in a 1 cm path length quartz cell. NVC was extracted directly from secretory cavities and dissolved in acetonitrile and essential oil was steam distilled from a bulk leaf sample and an aliquot then dissolved in acetonitrile. Multiphoton fluorescence lifetime imaging microscopy A schematic of the time-resolved, multiphoton fluorescence microscope is depicted in Fig. 1. A mode-locked Ti : Sapphire laser (Mira 900F, Coherent, Santa Clara, CA, U.S.A.) pumped by a 10 W Nd : Vanadate laser (Verdi V10, Coherent) was used as the ∼800 nm illumination source. This laser produces optical pulses with a temporal pulse width of approximately 100 fs (∼3 nJ pulse energy) at a repetition rate of 76 MHz. The laser beam was coupled to an inverted microscope (IX-71, Olympus, Center Valley, PA, U.S.A.) through a confocal scanning unit (FV-300, Olympus) and a transfer lens, which produces a focussed spot scanned across the normal focal plane of an imaging objective. A dichromatic mirror (UV-Cold Mirror, TFI, Inc., Greenfield, MA, U.S.A.) mounted at 45◦ in the rotating filter turret of the microscope efficiently allows incident illumination to pass up to fill the back aperture of an imaging objective, (Olympus 40× UPlanApo 1.0 NA oil, or 100× UPlanSApo, 1.4 NA oil) while reflecting the emitted visible fluorescence from the sample to a PMT detector (PMC-1–100, Becker & Hickl, Berlin, Germany) masked with a blocking filter (BG39, TFI, Inc.). Time-resolved two-photon fluorescence images were obtained using a complete electronic system for recording fast light signals by time-correlated single photon counting (SPC-830, Becker & Hickl). Synchronized data collection accumulating photon counts in each pixel following multiple scans is achieved with the SPC-830 board using the frame, line and pixel clocks of the FV300 scan unit following necessary hardware modifications. Each image was recorded as scans of 1024 × 1024 pixels that are binned by the SPC-830 data-acquisition software to generate images of 256 × 256 pixels, comprising complete fluorescence decay information in each pixel. The temporal evolution of the emission probability after excitation is described by a histogram of these time spans whereby each counting event is allocated to a temporal bin for each pixel of the image. Intensity images were generated by summing all of the photons in the distribution histogram for each pixel individually using a separate offline image analysis package SPCImage (Becker & Hickl). The fluorescence lifetime in each pixel was calculated using the same software. SPCImage applies a Levenberg–Marquardt nonlinear least squares iterative reconvolution routine to individually fit the fluorescence decay data from each pixel, assuming that the fluorescence decay histogram for each pixel is well described by a multi-exponential decay (Eq. 1), I (t) =

n 

a i exp(−t/τi ) + c

i =1  C 2012 The Authors C 2012 Royal Microscopical Society, 247, 33–42 Journal of Microscopy 

(1)

35

where I(t) is the fluorescence intensity at time t after the excitation pulse, τ i and ai are the fluorescence lifetimes and their fractional contributions and c is a baseline parameter. The SPCImage software derives the instrument response function by approximating the instrument response function as a Gaussian function, the width of which is adjusted to give the best fit to the rising edge of the decays (Becker, 2005). Fluorescence lifetime images were generated by pseudocolour mapping the average weighted fluorescence lifetime (τ m ) of each pixel over the intensity image. τ m values are a weighted average of the different lifetime components and were calculated using Eq. 2. n 

τm =

a i τi

i =1 n 

(2) ai

i =1

The lifetime values of in situ NVC were calculated for defined regions within secretory cavity lumina by averaging pixel values within selected regions (typically of 13 × 13 pixels around the given point where each pixel represents ∼550 nm) containing the brightest pixels. Mean values given are the average of these values for 15 replicate secretory cavities. The mean τ m of extracted NVC was obtained using the same method but is the mean of 400 of the brightest pixels from a single image. In the specific case of secretory cell wall fluorescence, five regions of interest containing only cell wall were selected from a representative image and the combined data (2000 pixels) were fitted with a triple-exponential decay model, as this provided a better fit of the data than a doubleexponential decay model. Imaging parameters The imaging parameters were optimized to obtain the strongest fluorescence signal from isolated secretory cavities whilst minimizing the scanning time for each sample. The laser beam was focused to approximately the centre of the lumen and the wavelength tuned to achieve the strongest fluorescence signal. This was achieved with an excitation wavelength of 760 nm. A compromise between laser power and scan time was reached to minimize laser power yet acquire adequate photon counts in each pixel for accurate fluorescence decay curve fitting. This was achieved with a laser power of ∼14 mW at the sample through the 40× objective (∼180 pJ/pulse) and a total scan time of 3 min. These conditions did not visibly damage the secretory cavities or result in photobleaching with cavities able to be imaged multiple times without any significant decrease in fluorescence intensity (data not shown). The extent to which secretory cavities could be optically sectioned was limited by the focussing depth of the objective (∼80 μm). An image was taken of all secretory cavities at the focussing limit; this varied for different

36

A. M. HESKES ET AL.

Olympus IX-71 microscope

mode-locked Ti:Al2O3 laser

λex

λ = 800 nm, ~100 fs, 76 MHz

λem single photon timing photomultiplier

x-scan galvo Dichroic 420