We chemically immobilized live, motile Escherichia coli on micrometer-scale, photocatalytically patterned silicon surfaces via amine- and carboxylic acid–based chemistries. Immobilization facilitated (i) controlled positioning; (ii) high resolution cell wall imaging via atomic force microscopy (AFM); and (iii) chemical analysis with time-of-flight-secondary ion mass spectrometry (ToF-SIMS). Spinning motion of tethered bacteria, captured with fast-acquisition video, proved microbe viability. We expect our protocols to open new experimental doors for basic and applied studies of microorganisms, from host-pathogen relationships, to microbial forensics and drug discovery, to biosensors and biofuel cell optimization.

Introduction

Biomolecular placement with controlled spatial resolution is of significant interest to expanding research programs in (i) host-pathogen biology, (ii) biosensors, (iii) and biofuel cells. (i) Pathogen study requires appropriate placement and containment protocols. While micro-injection (1) is a well-controlled technique used by some investigators to deliver pathogens into hosts, more common techniques, such as incubating (2), are not well controlled and pathogen release—whether triggered by accident or intention—can have deleterious health effects. (ii) Biosensors require extremely localized chemical surface modifications to enable associated biological reagent activity, electrical transduction, and readout (3,4,5,6). Biosensor power is typically defined by the density of information on a chip: the tighter and more localized the patterning, the more powerful the chip. (iii) Microbial fuel cells (7,8,9,10) require electron transduction from the organism to an electrode, defining a minute distance for charge to travel. Optimal signal transduction may be accomplished by engineering electron transport paths between microbe and electrode. Our procedures dictate placement and improve containment of microorganisms under investigation, thereby facilitating characterization by complementary analysis techniques.

Other groups have labeled bacteria or dictated bacterial placement. Solution-based fluorescence (11,12) and quantum-dot (13,14) labeling are two such methods that enable simple bacterial visualization. Technetium labeling (15) for medical purposes and covalent labeling of fusion proteins (16) have also been reported, and Da Silva et al. biotinylated bacteria for subsequent avidin-based interactions (17). Large-scale, low-resolution bacterial patterning—historically using replica plating (18,19) and more recently with pin arraying (20,21)—enables microbial genetic and biofilm formation studies. Immobilizing microbes on polymeric hydrogels (22,23) can limit motility and maintain viability. Weibel et al. described the use of micropatterned agarose stamps to print patterns of bacteria on agar plates to study pattern growth and bacterial strain interactions (24). In these experiments, features as small as 200 µm were printed over areas as large as 50 cm². More recently, Keymer et al. constructed habitat microfluidics (linear arrays of coupled, micro-scale patches) to study adaptive dynamics of the bacterial metapopulations (25). Our method of high-resolution microbial patterning incorporates labeling and tethering steps, such that microbe viability is maintained, thereby facilitating analysis.

Our sample preparation protocols pattern and tether live bacteria. Once tethered, the bacteria are available for examination via a number of techniques, or can serve as instrumentation themselves. We first tested our methods by examining bacteria optically with reflectance and fluorescence microscopy, and with atomic force microscopy (AFM); these techniques facilitate in situ examination. We also tested our methods under high-vacuum conditions, with time-of-flight secondary ion mass spectometry (ToF-SIMS). We demonstrate the use of multiple examination techniques on a single sample, in order to gain corroborating information and validate the multiplexing capabilities of our protocols.

We chose Escherichia coli, a Gram-negative bacterium that resides in the gut, as a model microorganism for investigation. E. coli has flagella, each composed of a rotary motor and filament. These flagella enable a cell to swim and confound the ability to study high-resolution aspects of the live microorganism in a native environment. Two similar protocols were used to reduce their movement, exploiting amine or acid chemistry on the bacteria and enabling subsequent investigation.

Materials and methods

Chemical attachment protocols

We developed two methods of chemical attachment to E. coli: the first covalently modifies outer membrane amine groups and allows for chemical and bacterial pattern overlay checking with a fluorescent intermediate, while the second modifies outer membrane carboxyl groups and directly, covalently immobilizes bacteria to silicon. E. coli K12 MG1655 (wild-type) strain was used in all experiments. This type of E. coli has peritrichous flagella (that is, multiple flagella per bacterium that project in all directions).

Protocol 1. Amine-based attachment on patterned, non-adhesive silicon substrates, allowing intermediate fluorescent pattern visualization

We generated 10 µm × 10 µm square-patterned regions by photocatalytically oxidizing an unsaturated silane attached to silicon by shining a red LED light (Lumex, Glenview, IL, USA) through a porphyrin-coated mask in contact with the substrate (26). Briefly, 10 µm × 10 µm square regions of allyltrichlorosilane–coated (ATC; United Chemicals, Bristol, PA, USA) silicon substrates were oxidized away by illuminating the red LED through a porphyrin-coated mask in contact with the substrate for a few seconds, leaving a chemically patterned ATC/SiO₂ checkerboard. A non-fouling interpenetrating network (IPN) chemistry of P(AAm-co-EG) (27) was then covalently grafted to matrix regions retaining the ATC (27). Square, bare silicon regions were then modified with aminopropyl silane (APS; United Chemicals, Bristol, PA, USA) (26).