Dynamic speckle laser and video analysis to study bacterial chemotaxis

MALDONADO, E.; PASTORE, J.I.; PASSONI, I.; NISENBAUM, M.; MARTÍNEZ ARCA, J.; SENDRA, G.; MURIALDO, S.; GUSMÁN, E.; BOUCHET, A.; TRIVI, M.

La Plata

Workshop; Workshop Imaging Techniques for Biotechnology and Biomedical Applications; 2016

Universidad Nacional de La Plata, CONICET y École Polytechnique de Lausanne, Suiza.

In the last 40 years, bacterial chemotaxis has been an area of increasing attention for researchers. This interest has been powered by increasing experimental insight into the behaviour of bacteria, both at the population and on the individual scales (Tindall et al., 2008). Chemotaxis is a behavioural response exhibited by flagellated bacteria moving directly towards or away from chemicals concentration gradients in the environment (Parales and Harwood, 2002). Some authors attribute to the bacterial chemotaxis towards toxic substances an important role in the fate of pollutants in the environment (Parales and Harwood, 2002).Bacteria use a well-defined biochemical network to communicate the detected temporal difference of the extracellular environment at the membrane receptors to the flagellar motors (Tindall et al., 2008). A combination of runs and tumbles allow bacteria to explore and respond to environmental changes. Therefore, understanding the behaviour of chemotactic bacterial populations is interesting since one bacterium behaves independently, but the population exhibit collective behaviour in the form of colonies and biofilms, which can have substantial impact upon industry and medicine (Davey and O?Toole, 2000). Hence, studying and understanding the morphology and distribution of motility clusters on a nutritive surface would greatly facilitate the prediction of bacterial behaviour in the natural environment.In our work we considered a strain of Pseudomonas sp isolated in our laboratory (Murialdo et al., 2003) and capable of metabolising chlorophenols (Wolski et al., 2006) for chemotactic population studies. These bacteria also showed hydrocarbon degradation and chemotactic responses to these pollutants (Nisenbaum et al., 2013). The chemotactic response is based on the created gradient because of the diffusion or metabolization of the carbon source and the consequent movement of the bacteria (Wolfe and Berg, 1989). Different assays have been developed to show qualitative bacterial chemotaxis in semisolid and liquid mediums. In all of them, the spatial bacteria accumulation is observed at different times using white light. Parales and Harwood (2002) reported that the presence of chemotactic response produces a sharp chemotactic ring of the bacteria that grows and moves outwards from the inoculums which is visualised in the swarming dish. In these assays, low concentrations of bacteria can hinder the visualisation of the chemotactic bacterial accumulation. We have recently developed new methods for chemotaxis determination, one based on dynamic speckle activity segmentation in swim assay (Murialdo et al., 2009) and the other based in video processing analysis applied to agarose-in plug assay. These methods allow differentiation of diverse degrees of motility in bacteria (Murialdo et al., 2009) and discriminate bacteria from filamentous fungi (Murialdo et al., 2012). A dynamic speckle laser method has been recently tested by our group (Nisenbaum et al., 2013) with toxic compounds metabolizable by bacteria. In a recent publication (Nisenbaum et al., 2014) we reported the capability to distinguish motile surface patterns per area of colonisation by applying image processing techniques, employing fuzzy mathematical morphology (FMM; Bloch, 2005) to the laser speckle chemotaxis images. The images of bacterial colonies are usually textured, with fuzzy edges and non-homogeneous grey levels. Hence, conventional image processing methods for shape analysis cannot be applied in these cases (Gonzalez and Woods, 2002). The given approach was effective to segment, detect and also to describe non-completely spatially random colonisation patterns. We could accurately identify the bacterial cluster area of highest motility. Nutrient uptake was shown to depend on the uptake rate and on the fast or slow swimming cells (Tindall et al., 2008). This stresses the importance of putting more emphasis in studying the morphology of chemotactic response, which could be involved in a faster biofilm formation and the consequent biodegradation. The FMM developed algorithms were fast and precise, allowing analysing the pattern spacing and orientation between bacteria groups in order to manipulate biofilms development and thereby control the degradation kinetics.Studying the coordinated temporal and spatial movement of bacterial clusters, combined with previous molecular discoveries, might be useful as predicting the strategies of microorganisms in response to fluctuations in the microenvironment. In our last publication the aim was to design a methodology based on video processing to obtain indicators of bacterial population motility. It allows the quantitative and qualitative analysis and comparison of the chemotactic phenomenon in liquid medium with different attractants. We used an optical microscope and the agarose-in plug bridge method (Yu and Alam, 1997) to study the chemotactic response. Video image sequences were processed applying Shannon´s entropy (Shannon, 1948) to the intensity time series of each pixel, which conducted to a final pseudo coloured image resembling a map of the dynamic bacterial clusters. Processed images could discriminate perfectly between positive and negative attractant responses at different periods of time from the beginning of the assay. Video sequences processing allows discriminating time and space evolution of bacterial population movement and obtaining numerical results as Motility Index and Attraction Index that allow the quantification of the response. With this index, this video processing method allowed obtaining quantitative information of the dynamic changes in space and time from a traditional qualitative assay. We conclude that this computational technique, applied to the traditional agarose-in plug assay, has demonstrated good sensitivity for identifying chemotactic regions with a broad range of motility. This methodology could be broadly used for other microbial chemotactic assays which involve the use of a microscope couple to a computing device.