| |
Analysis of Confocal Microscope Images of Microbial Biofilms |
Xavier, J.B.(1,3), R. Malhó(2), A.M.Reis(1), J.S.Almeida(1,3)
|
    |
Confocal Scanning Laser Microscopy (CSLM) allows the optical slicing of translucent material by filtering out the light outside the focal plane. This technique is normally non-destructive which enables the in vivo reconstitution of 3D structure of microbial biofilms. Therefore, biofilm formation can be described in detail by scanning the same site repeatedly. The formation of a mixed culture denitrifying biofilm was analyzed by CSLM. The tri-dimensional reconstitutions were analyzed in order to access the evolution of biofilm volume, thickness, interface area, porosity, density and turtuosity. It was observed that biofilm formation proceeded mostly by decrease of biofilm porosity and increase in the interface area, rather than by changes in apparent biofilm thickness. The calculation of geometric turtuosity was used to develop a measure of effective thickness that evaluates the mass transfer resistance offered by the biofilm. The interface area measured is also dependent on the resolution. Therefore, the same image was analyzed with different resolutions in order to determine its fractal dimension, which was observed to be constant and above 2.
The growth of a mixed species biofilm was monitored non-destructively by confocal
scanning laser microscopy, CSLM (Fig.1). A mineral medium supplemented with nitrate and
acetate was circulated through the flow cell, previously inoculated with a mixed species
inoculum.
Figure 1 - Experimental layout, click on image to enlarge
Biofilm is stained with Syto 9 stain (Molecular Probes TM) immediately before image
acquisition. Tri-dimensional SCLM Images were acquired at regular intervals, from 0 h
until 76 h after inoculation.
A series of pre-processing steps involving normalization of tilting, position and contrast
was used to convert the two dimensional image stacks of confocal images into binary
three-dimensional structural reconstitution (Fig.2). In situ image acquisition at regular
intervals allowed the non-destructive in situ visualization and study of biofilm structure
development. This approach allowed the 4-dimensional (time is the 4th dimension)
characterization of biofilm formation.
Figure 2 - Pre-processing steps, click on image to enlarge
Analysis of data allows time-evolution study of several biofilm properties:
Biofilm volume was observed to increase exponentially (Fig. 3) after a initial period of adaptation (first 24h). The exponential growth was maintained for different biofilm thickness which implies that all biofilm is growing, not just cellular material located at the interface. This observation leads to the conclusion that the mass transfer characteristics associated with the biofilm structure are scale independent. The 3D reconstitution of biofilm structure revealed that biofilm interfacial area also increases exponentially and at the same rate as biofilm volume (not shown). Consequently, the biofilm volumetric interfacial area (interfacial area/biovolume ratio) is kept constant during biofilm growth (Fig.4). This suggests that biofilm structure is constantly optimized in order to maintain stable growth conditions, namely in what respects nutrient mass transfer. Further analysis revealed that biofilm thickness is not a relevant indicator of growth stage as biofilm porosity and morphology changes continuously (Fig. 5).
 |
|
Figure 3 - Exponential increase in biofilm volume |
Figure 4 - Volumetric interface area kept constant during growth |
|
|
Figure 5 - Biofilm filling at different depths, click on image to enlarge |
Figure 6 - Effective thickness (arrow) determined by the maximum cumulative turtuosity, determined heuristically: |
The 3D reconstitution was also used for simulation of molecular diffusion (Fig. 8), in
order to determine structural turtuosity (the ration between effective and molecular
diffusion after correction for porosity). Tortuosity is an empirical measure that
describes the effect of geometry on mass transfer and, consequently, was used to determine
the effective biofilm thickness (Fig. 6). The fractal analysis of biofilm morphology
confirmed the geometric scale independence: the volumetric interfacial area has a constant
fractal dimension (Fig.7), with values around 2.5, well above the Euclidean dimension for
area, which is 2.
|
|
Figure 7 - The fractal dimension of volumetric interfacial area is higher than the Euclidean dimension (2) |
Figure 6 - Effective thickness (arrow) determined by the maximum cumulative turtuosity, determined heuristically: |
This work is financially supported by the Foundation for Science and Technology/ M.C.T., PORTUGAL, through the grant PRAXISXXI/BD/18285/98.
Click here to go to the top of the page
Click here to go to the ITQB homepage
© Copyright 1999 ITQB
Last Modified: 17.06.1999
jxavier@itqb.unl.pt