2-D Wavelet Packet Analysis of Structural Self-Organization and Morphogenic Regulation in Filamentous Fungal Colonies

Cameron L. Jones

This preprint should be referenced as follows:

Abstract

Introduction

Modification of Protein Expression in Fungi:

• Cell wall biochemistry

• cytosol molar ratios of transport proteins which are involved with protein synthesis or,

• the cell-mycelium morphology by enhancing or repressing hyphal branching.

Our interest in paramorphogen stimulation developed from an earlier study which established a correlation between hyphal branching and the number density of active tip production per unit area and extracellular oxidase expression levels (Jones et al., 1996b). Image-analysis of mycelial morphology was used to quantitatively measure the macroscopic branching index of isolated whole fungal colonies growing on solid surface culture (SSC). SSC can be on a large industrial scale or 'bench scale' using agar filled Petri-plates. The Fractal Dimension, D was shown to be a sensitive numerical estimate of branching behaviour (Jones et al., 1993, 1994). This index estimates the amount of surface coverage by the fungal mycelium. For two dimensional (2-D) objects in the plane, 1<D<2. We consider fungal colonies which grow across the agar surface or on a membrane covering the agar, to be two dimensional since we view them as a projection against a flat 2-D plane using transmitted illumination. Straight lines have a D value=1 and a plane filling curve has a D value=2. Values of D within this range therefore characterize the extent of space filling. Highly branched fungal colonies have a 'high' fractal dimension compared against sparsely branched colonies. All measurements must be carried out at a standardized time, and all culture conditions must be well defined. Obviously, measuring various D values for discrete colonies growing on SSC is interesting, but of greater import is the possible relationship between growth in 2- D surface culture and growth in unrestricted 3-D liquid culture. Parallel experiments in liquid culture have established a positive correlation between the Fractal dimension of branching and extracellular oxidase secretion into the medium (Jones et al., 1996b). Over time, the fungal mycelium secretes exo-enzymes and the kinetic 'efficiency' of their production is of considerable industrial importance (Lonergan et al., 1995; Hatakka, 1985) . The activity of the crude enzyme (i.e. laccase in Pycnoporus cinnabarinus) for performing useful substrate transformation (oxidation) was determined by measuring the decolorization (degradation) of the aminoanthraquinone dye, Remazol Brilliant Blue R. The crude extracellular fluid was analyzed daily and shown to contain (i) oxidase enzyme activity and that (ii) oxidase enzyme could decolorize the dye. Of most importance was the fact that the fractal dimension measured in very young colonies (24hrs.) was positively correlated with future enzyme secretion potential.

Global Objective:

The chosen paramorphogen - cellobiose is a carbohydrate-linked glucose disaccharide. The literature documents that cellobiose generally causes a restriction in fungal growth rate and may induce profuse branching, compared against untreated or wild-type growth. Its mode of action is thought to act upon cell wall metabolism, and we review this evidence below and then report on our results with cellobiose induction in P. cinnabarinus.

Several studies have confirmed that cellobiose can induce both macroscopic changes to structural morphology, and microscopic changes at the level of intracellular metabolism (Wilson & Niedepruem,(1967a),1967b; Wilson, 1970). A substance called R-glucan (present in the cell wall) is thought to provide a structural shape preserving action to the fungal cell. Since cellobiose is known to effect cell wall metabolism, and the R-glucan fraction maintains hyphal cell shape (Wessells, 1965) a practical experiment carried out by Wilson (1970) sought to evaluate the changes in R-glucanase activity in response to cellobiose induction. Measurement of R-glucanase (the enzyme against R-glucan) therefore provides an index of potential hydrolytic action against this particular cell wall fraction. It was confirmed (Wilson, 1970) that when several basidiomycete fungi were grown on cellobiose containing media, the levels of several wall-hydrolysing enzymes (including R-glucanase) were all increased. It was concluded that elevated levels of cell wall hydrolysing enzymes provide multiple regions of wall softening, thereby resulting in increased branching per unit area. It has also been noted that hydrolytic enzymes can participate in transfer reactions which suggests the possibility that cell wall synthesis may also be a factor in morphogenesis (Wilson, 1970). They concluded that the elevation in R-glucanase activity could physiologically increase cell wall plasticity and result in an increased branching response.

Materials and Methods

Powdered newspaper at 0.2% w/v was solidified with 2% w/v Technical Grade Agar No. 3 (Oxoid). The sole carbon source was therefore newspaper only (N) for the control, while the cellobiose treated plates were prepared by adding 2% w/v cellobiose (Sigma). Media was poured into 9cm diameter Petri plates at 25-30mL final volume. Colonies of P. cinnabarinus were inoculated following the previously described protocol for the germination of discrete, single colonies on microporous polycarbonate membrane overlays (Jones et al.,1994). The colonies were also contrast stained with Coomassie Brilliant Blue and examined using an Olympus CH- 2 optical microscope set up for bright field transmission. The methods used to apply 2-D WPA are essentially the same as those used for 1-D WPA discussed in detail in Jones et al. (1996a). However we have used the 2-D symmetrical (s.12) wavelet filter with 12 vanishing methods, using input images of size: 512x512. More complete details will shortly be made available when the software is lodged with the StatLib software repository for S-Plus written software (Carnegie Mellon University Statistics Department). A brief preprint detailing step-by-step instructions for operating the software is available here (Jones, 1996c). Source code (in Microsoft Word ver 6.x for Windows format) is available individually for the following 2-D image sizes: 128x128, 192x192, 256x256, 384x384, 300x200, 512x512.

To examine the effect of cellobiose supplementation versus newspaper only controls, the following experiment was also conducted. Liquid media were prepared using the same newspaper and cellobiose concentrations as for solid media in Petri plates, however technical agar was not added. Static batch flasks (250mL volume) containing 25mL of liquid media were inoculated with 2x5mm diam. plugs of 5 day old stock cultures of P. cinnabarinus taken from the exponentially growing frontier. All flasks were incubated under humidified conditions at 37°C for five days, by incubating the flasks in trays covered with plastic wrap. Aliquots (700µl volume) were removed daily under aseptic conditions for spectrophotometric analysis of oxidase activity. Phenol oxidase activity of laccase was measured at 415nm over a period of 3 minutes at room temperature. Oxidase enzyme activity was measured as activity against ABTS (Niku-Paavola et al., 1990; Bourbonnais and Paice, 1992). This spectrophotometric assay used a reaction mixture containing 0.5mL of 100mM sodium succinate buffer at pH 4.5; 0.1mL of 2.92mM ABTS; 0.3mL of distilled water and 0.1mL of extracellular culture fluid. One unit (U) of enzyme was defined as the amount which caused a change in absorbance (415nm) of 1.0 per minute at room temperature. All activities were reported as U mL-1 (Table 1).

Although we have monitored only oxidase enzyme, the fate of other extracellular proteins have been established in two ways: Total protein was measured on Day 5 run-out samples and could be compared against oxidase activity to provide a specific activity index on each day. This led to further enzyme expression profiles following cellobiose induction (Table 2). Specifically, the Day 5 extracellular fluid was concentrated by Amicon Ultracentrifugation using Centriprep-10 and Centricon-10 filtration units (10,000 dalton M.W. cut-off) for three hours. This reduced the total volume from approximately 20mL to 40µL of concentrated sample. Proteins were separated by electrophoresis on native PAGE (12%) by loading 20µL of sample and 5µL of buffer. Activity staining using guaiacol established the band position of oxidase, and the gels could be further stained with Coomassie Brilliant Blue R or silverstain enhanced to reveal additional protein bands.

Extracellular protein was assayed at Day 5 for both media treatments using a Bio-Rad Microassay Kit for protein, using bovine serum albumin as standard. The specific enzyme activity (U/mg) was estimated by dividing the oxidase activity (U/mL) by the protein concentration (g/mL). This provides a semi-quantitative biochemical index of the amount of extracellular proteins which are oxidases.

Results and Discussion