Attachment of Escherichia coli on plant surface structures built by microfabrication

Consumption of fresh fruits and vegetables contaminated with human pathogenic bacteria has caused numerous outbreaks of food-borne disease in recent years. Understanding the nature of bacterial colonisation on plant surfaces could lead to more effective practices to prevent contamination. The goal of our study was to determine the effect on bacterial attachment of structures mimicking those that occur on plants. Plant surfaces have microstructures in a wide range of sizes and shapes, causing difficulty in separating and understanding individual effects. Microfabrication methods were used to build surface structures on silicon to mimic stomata, trichomes, and grooves between plant epidermal cells. These structures were subjected to a 48-h culture of Escherichia coli tagged with green fluorescent protein. Bacterial attachment characteristics were determined by observation under a confocal laser scanning microscope. The number of bacteria in an area with an array of trichomes was about half of that for the stomata or grooves. Within an array of trichomes, the bacteria level around the base of the trichomes was 4.7 times higher than areas further away. Within an array of stomata, the level of bacteria attaching to the area 2.5–5.0 μm away from the stomatal opening was 35–59 percent higher than the area next to the opening or areas further away. Within an array of grooves, there were no significant location effects. Microstructure geometry and position clearly affected bacterial attachment in our model system.

Introduction

Fresh fruit and vegetable consumption has increased over the past several decades due to greater health awareness by consumers, but outbreaks of food-borne disease have also increased. Intensification and centralisation of crop production, wider distribution over longer distances, introduction of minimal processing, and globalisation of supply and distribution of raw produce increase the risk of outbreaks (Tauxe et al., 1997). Pathogenic Escherichia coli and Salmonella enteric are the two most frequent etiologic agents of fresh produce contamination in the United States (Lynch, Painter, Woodruff, & Braden, 2006). Preventing contamination is still preferable to existing decontamination methods such as washing with chlorinated water (Beuchat, 1998). Understanding bacterial attachment to plant surfaces may lead to more effective practices for preventing contamination.

All aerial parts of higher plants such as stems, leaves, petals, and fruits are covered by a continuous layer called the cuticle, with the exception of stems that have undergone secondary growth (Müller & Riederer, 2005). The only gaps in the cuticle are the stomatal pores. In addition to the cuticle, plant tissue has evolved to include other structures, such as trichomes and stomata. Trichomes reduce water loss, restrict entry of microorganisms and toxic substances, mitigate radiation, and strengthen and support the plant. They also reduce the plant’s susceptibility to damage from rapid air movement (Beck, 2005, pp. 138–153). Stomata allow carbon dioxide used in photosynthesis to enter the plant, but the plant loses water vapour as a consequence. The plant tissues represent both structural and functional compromises between protecting from and interacting with their surroundings. Leaf surface topography of different plants comes in various appearances and has been documented in several references (Harr and Guggenheim, 1995, Troughton and Donaldson, 1972).

Many factors can introduce bacterial pathogens into field crops, including fertilisation with manure, irrigation with contaminated water, and nearby animal husbandry. One field-scale investigation showed that E. coli O157:H7 persisted on lettuce for more than 77 days and on parsley for more than 177 days (Islam, Doyle, Phatak, Millner, & Jiang, 2004). These plants were grown in fields treated with contaminated manure compost or contaminated irrigation water. Another study of cut lettuce leaves immersed in a suspension of E. coli O157:H7 showed that the bacteria attached predominantly to the cut edges, and fewer bacteria attached to the intact cuticle region (Seo & Frank, 1999). Some bacteria were also observed near stomata, on trichomes, and on veins. In other studies it was found that E. coli O157:H7 colonised stomata, broken trichomes, or cracks in the plant surface, and these bacteria were protected from sanitation with chlorine (Takeuchi and Frank, 2000, Takeuchi and Frank, 2001). Less waxy surfaces also attract bacteria (Solomon, Brandl, & Mandrell, 2006), probably due to the bacteria surface being more hydrophilic than the epicuticular wax on the outermost plant surface. Human pathogens localise in higher numbers on broken plant tissues, most probably because of leakage of water and nutrients (Lindow & Brandl, 2003). Recent reviews (Berger et al., 2010, Warriner and Namvar, 2010) showed that most research emphasised the biomolecular aspects of plant–microbe interactions. Human pathogens apply cell surface components including bacterial cellulose, flagella, and pilus curli for successful attachment.

In addition to the biological factors, bacteria may be attracted to specially shaped surfaces (i.e., microstructures) and preferentially attach there. On actual plant surfaces, these microstructures come in diverse shapes, sizes, and distributions, causing difficulty in understanding the exact influence of microstructure on bacterial attachment. To study this problem, one approach is to build microstructures with fixed shapes and dimensions by microfabrication. The concept of employing micro-devices to understand biological problems has become more common, but only a few studies have employed microstructures imitating those in nature to observe microbial behaviour. One group used silicon surfaces containing arrays of micro-pillars with different contact area sizes on top and looked at the continuous contact area requirement for attachment of the plant pathogenic fungi Colletotrichum graminicola (Apoga, Barnard, Craighead, & Hoch, 2004). They also looked at the attachment behaviour of Xylella fastidiosa in microfluidic flow chambers mimicking plant xylem (De La Fuente et al., 2007). The chambers provided a system which could measure the adhesion forces of type I and type IV pili used by the bacteria for attachment.

Bacteria appear to localise selectively on plant surfaces and thus must be affected by certain surface characteristics, including possibly surface geometry. Our goal in the first phase of this research was to determine the effects of microstructure on bacterial attachment. The levels of bacteria found attached to the different microstructure types are reported here.