LIGHT MICROSCOPIES

LM is the oldest of the microstructural analysis techniques applied to food materials and also the technique most likely to be found in a food quality, or new product develop-ment laboratory. Although the variety of configurations for specific contrast mechanisms is large, the basics of LM remain effectively unchanged from its inception. A light source produces visible light that passes through a condenser lens, through a thin sample (unless working in reflectance) and is focused by an objective lens to form an enlarged image.

Resolution of the LM (the smallest distance between two features where the features are still visibly separate) is around 1000 times better than the naked eye.

Table 3.1 Example Structural Features in Food Materials at Different Length Scales Size µm–mm Air cells in bread Texture effects through apparent

Young’s modulus, “resilience,”

Pumpability in plant, shelf-life (ice crystal ripening), texture effects (Bollinger et al. 2000)

Cryo-SEM

µm Fat globules in cheese Baking performance on pizza (Ma et al. 2013)

CLSM, cryo-SEM, XRM

µm Emulsion droplets in

mayonnaise Rheology, texture, nutrition (Stokes

and Telford 2004) CLSM, SEM,

low-vacuum SEM µm–nm Surface fat structure of

chocolate Shelf stability, bloom performance, appearance, texture (Sonwai and Rousseau 2010)

SEM, AFM

µm–nm Casein micelle

network in yoghurt Texture, processing, shelf-stability

possibly nutrition (McCann et al. 2011) SEM, cryo-SEM nm Casein micelle

structure Building block of most dairy-based

systems (Marchin et al. 2007) TEM, scattering techniques

3.2.1 Sample Preparation, LM

Sample preparation for LM can be moderately simple as the samples can be imaged in atmospheric conditions, meaning hydrated samples can be easily examined. An excellent text for anyone starting LM for food structure analysis is Food Microscopy by Olga Flint (1994). Samples must be sectioned to thin slices for transmission LM, and various micro-tomes are available to do this without introducing substantial artifacts. Alternatively, liq-uid samples can be smeared or squashed between a slide and cover slip to create a thin sample. Frequently, samples are stained to improve contrast between components. These histochemical techniques rely on known reactions between the stain and the food com-ponents, for example iodine is a stain used to highlight the presence of starch and protein (Figure 3.2). The choice of stain depends on the sample and the component of interest and, with experience, very subtle details of food structure can be ascertained, as shown in Table 3.2.

Fluorescent stains for specific components are useful for differentiating, for example, cell wall components in plant-based foods. These flurophores are covered in the next sec-tion on confocal laser scanning microscopy (CLSM). Various contrast techniques can also enhance contrast without staining. For example, in differential interference contrast (also called Nomarski contrast), the phase of the light is altered to create contrast. Polarizing filters can be used to track specific changes in birefringent food components; native starch gives a distinctive “maltese cross” contrast that disappears as the starch gelatinizes (Aguilera et al. 2001).

3.2.2 Confocal Laser Scanning Microscopy

Since the 1990s CLSM has become increasingly relevant to food microstructure analy-sis and is now often the “go to” technique before all others. One of the disadvantages of

Figure 3.2 Iodine-stained chewed bolus of biscuit showing staining of starch granules. (Courtesy of Sophia Rodrigues, University of Auckland.)

Food MiCRostRuCtuRe AnAlYsis

standard LM is the need for samples to be thin for good transmittance and high reso-lution (avoiding blurring from features out of the focal plane). Producing thin sections, or smears, of sample can introduce structural artifacts. CLSM instruments use a focused laser, scanned across the specimen, to image a subsurface layer. Information from this focal plane is projected onto a detector, with out-of-plane information blocked by a “pinhole” or confocal aperture. The resolution of any optical microscopy technique can be described by the Rayleigh criterion, and a large amount theory can be accessed elsewhere for readers particularly interested in optics (Mertz 2010). In terms of practical application of CLSM, the parameters of relevance to optimal resolution are the magnification, the numerical aperture (NA) of the objective lens, and the working distance. NA determines how much reemitted light is collected and works in combination with the confocal aperture to deter-mine the thickness of the optical “slice” from which the image is built. The lateral and axial resolution can be estimated by (Lorén et al. 2007):

R^lateral nA

emission

= 0 4.

and

R n

axial = 1 4. nAemission₂

Table 3.2 Stains for Contrast in Light Microscopy

Food Component Stain Result

Muscle fibres Toluidine blue^a Pale blue (fresh meat)

Raw collagen Toluidine blue Pale pink

Cooked collagen Toluidine blue Pale lilac

Elastin fibres Toluidine blue Turquoise

Soya protein Toluidine blue Dark purple-blue

Wheat protein (gluten) Toluidine blue Pale blue-green Lignified cellulose Toluidine blue Dark blue or blue-green Lignified cellulose Trypan blue Dark blue

Cellulose Trypan blue Pale blue

Fat Toluidine blue Unstained (unless acidic)

Fat Sudan red or Oil Red O Red (solid fats can stay unstained but can be identified by birefringence)

Fatty acids Toluidine blue Pale blue

Raw amylose starches Aqueous iodine Blue (birefringent between crossed polars) Raw amylopectin

starches

Reddish (birefringent between crossed polars)

source: Adapted from Flint O. 1994. Food Microscopy. Oxford, UK: BIOS Scientific Publishers Ltd.

a Toluidine blue is a very sensitive stain for gums differentiating gums based on stain color and intensity.

where λemission is the wavelength of the detected light, NA = n sin (opening angle), and n is the refractive index of the mount medium. Typical values give R_lateral of around 150 nm and R_axial of around 550 nm. One of the great strengths of CLSM is the ability to collect a number of optical “slices” and recombine them to generate a three-dimensional (3D) image of the sample. This is frequently referred to as a z-stack.

There are numerous configurations of CLSM and different laser and filter combina-tions can be selected in almost infinite variety from single photon (point scanning) instru-ments with a single laser and a filter prior to the detector (to collect/exclude wavelengths) to very advanced multi-photon instruments with multiple lasers and spectral imaging detectors. Instruments are developing all the time and a recent summary can be found in the excellent introductory text Basic Confocal Microscopy (Price and Jerome 2011).

3.2.3 Sample Preparation, CLSM

Confocal microscopy is so frequently used in combination with fluorescent stains (fluo-rophores) that the terms are often conflated (it is important to remember that CLSM can also be used in other contrast modes such as brightfield and differential interference con-trast). Staining for specific food components is common, with basic stains such as Nile Red (for lipids), Rhodamine B or Fast Green (for proteins), and so on. Figure 3.3 shows typical images for CLSM of various cheeses stained to separate fat and protein phases, the images are recombinations of the individual excitation signals.

Suitable stains must be selected based not only on the component that needs to be labeled but also giving consideration to the excitation and emission wavelengths of the flu-orescence. These details can be sourced from literature (Auty et al. 2001; Brooker 1995) or suppliers (Invitrogen 2013). Diffusion of stain into solid samples can be time-consuming, although in some cases melting the sample and stirring the dye through prior to resolidi-fication can be suitable. As with all preparation techniques for microstructural character-ization, consideration must be given to the introduction of artifacts. For example, swelling

Figure 3.3 CLSM images of cheese. Samples were sectioned to 50-µm thick using a cryotome and stained with Nile Red and Fast Green. Contrast between fat and protein phases is clear. Samples are (a) cheddar, (b) mozzarella, and (c) provolone. (Image courtesy of Xixiu Ma, University of Auckland.)

Food MiCRostRuCtuRe AnAlYsis

when using liquid stains is possible and this needs to be acknowledged in any subsequent quantification of structural features.

Two particular issues that need to be considered when using CLSM and fluorescent stains are autofluorescence and photobleaching.

3.2.4 Autofluorescence

Fluorophore dyes are unlikely to be the only fluorescent compounds present in a sample of food. Many naturally occurring compounds fluoresce including chlorophyll, carotene, collagen, and elastin. The fixatives often used in sample preparation can also autofluoresce (including formaldehyde and glutaraldehyde). Generally, autofluorescence is an unavoid-able nuisance that can be worked around by optimizing selection of excitation source and filters. Suppression of autofluorescence using heparin has been reported for CLSM of bread doughs (Dürrenberger 2001). Figure 3.4 shows autofluorescence in chocolate, arising from the plant material present in the cocoa nibs.

3.2.5 Photobleaching

Photobleaching occurs as the fluorophores are excited by the laser light. Chemical dam-age to the structure of the fluorophore itself can occur and reactions with surrounding molecules are more likely with the fluorophore in its excited state. This results in fad-ing of the signal and can be a particular issue when conductfad-ing extended observations

Figure 3.4 CLSM image of dark chocolate (72% cocoa solids) showing autofluorescence of the cocoa nibs. Sample was unstained. (Image courtesy of Ying Jing Tan, University of Auckland.)

(e.g., generating a z-stack sequence). Some fluorophores are more susceptible to this prob-lem than others and therefore stain selection can address the issue. Chemicals that limit the phenomenon (antifade reagents such as p-phenylenediamine (PPD), n-propyl gallate (NPG), and 1,4-diazobicyclo(2.2.2)octane (DABCO)) are available, although not all are compatible with every fluorophore (Price and Jerome 2011). An alternative that is gaining increased interest is the use of quantum dots for high-resolution, fade-resistant labeling of cellular components (Weng et al. 2006).

Photobleaching is not all bad news. It has led to a dynamic imaging technique known as “fluorescence recovery after photobleaching.” Using this technique the mobility of mol-ecules within a sample can be monitored by deliberately photobleaching an area of the sample using high intensity excitation then watching the recovery of fluorescence in that region (using less intense excitation) as labeled molecules diffuse back into the area (Braga et al. 2004; Presley et al. 1997).

3.2.6 Dynamic and In Situ Studies Using CLSM

As the sample in CLSM can be, effectively, a bulk sample (rather than a thin specimen), it is feasible to conduct in situ experiments while observing dynamic changes in the specimen.

These experiments require the use of specific sample stages, which are available (or can be made) for mechanical or, rheological testing and heating/cooling studies. The fracture behavior of cheese has been followed during testing using CLSM (Lorén et al. 2007), as have the mechanical and fracture properties of model biopolymer gels (Plucknett et al.

2001). Freezing and thawing can be followed moderately easily using a temperature-con-trolled stage, as the ice crystal formation excludes the fluorophores from the water phase.