Spatially Resolved Photoacoustic Spectroscopy

We are working on devising a simple setup to perform gel—phase photoacoustic spectroscopy with unidimentional position resolution, to monitor the movements of molecular shuttles between members of bacterial colonies, or between bacteria and biochemically active substances. We are generally interested in the study of slow chemical transformations, with both time and position resolution, between interacting (living) organisms in their native state. This new technique will be described in the context of an experiment devised to study the role of extracellular electron transfer in the respiration of Schewanella Putrefaciens [1,2].

The principal motivation for the development of this experimental project is the study of the diffusive behavior of molecular shuttles regulating bacterial metabolism, and the interactions of bacteria with other organisms. To allow maximum flexibility, this new technique relies exclusively on the absorption spectrum of the chromophore, and not on fluorescence, so low quantum yields requiring dye—tagging are no longer an issue (thus allowing the observation of biological systems in their native state). In most cases, biomolecules feature large absorption shifts dependent upon their oxidation state. This important feature combined with the use of narrow band lasers allows to observe specific chemical changes with high selectivity which, when coupled to a high sensitivity detection technique, permits tracking of the formation and diffusion of minute concentrations of bioactive chemicals. The detection method must be somewhat insensitive to the experimental conditions (cell culture medium, presence of buffers to perfectly mimic natural conditions, etc.), and the excitation process must be immune from scatterers, or other absorbers, surrounding the analyte.

To allow careful determination of the relevant lengths in the system under study by means of probing variations in the spectral density of chromophores, the bacteria and other bioactive material must be immobilized. However, the substrate used to hold the bacteria must not hinder the diffusion of molecular shuttles. The use of a photoacoustic method of detection [3] allows to minimize the negative effects of the gel in terms of light propagation: because of the possibly very large scattering effects induced by the matrix, a method like (classical) absorption spectroscopy might turn out to be problematic. In our case, light does not need to propagate into the culture medium over extended distances: it only needs to reach the chromophores. The main drawback of this method is that the absorption strength might be affected by the vertical position ("depth" inside the matrix) of the absorber. Keeping the diffusion layer thin will help to minimize this effect.

Not unlike traditional gas—phase photoacoustic spectroscopy, the observation of the chemically relevant species will be performed via pickup of acoustic waves produced by the relaxation of vibrationally excited chromophores. In our case however, acoustic waves will be transferred from the gel (culture/immobilization matrix) to the surface of the substrate where some fraction of the initial acoustic amplitude will couple to surface modes propagate as surface acoustic waves (SAWs) [4]. These waves will then be converted into electrical signal by a network of interdigitated transducers (IDTs, see Figure 1) [5] and measured via phase sensitive detection performed at the resonance frequency of the IDT.

Our experimental system can be separated into 4 different parts which are somewhat independent of each other:

1. A piezoelectric substrate with relatively high efficiency to convert acoustic amplitude from the SAWs into electric current. Our setup uses lithium niobate (LiNbO₃), which features a large piezoelectric coefficient and is readily available, in many different forms (cuts, shape, etc.), from various supplier.
2. A network of IDTs with well defined resonance frequency. Using gold, on a chromium adhesion layer, insures chemical inertness.
3. A thin layer of low concentration gel, in which electron shuttles can move freely between the bacteria and the terminal electron acceptors. It must be thin to minimize scattering effects.
4. - An immobilization matrix to carefully position the colony and the oxidizing agent: a gel of some sort that features a rather high stiffness so that the bacteria cannot diffuse through the matrix. It must be thin, such that molecular shuttles can move to the diffusion layer easily.

Figure 1.

Top Panel: Schematics of the detector, side view. The first structure (dark grey) is the immobilization matrix. The second structure (light grey) represents the fluid diffusion layer. Just underneath it is the IDT (black line), resting on top of the piezoelectric wafer (white rectangle). The two circles represent possible locations for the bacteria and oxidizing agents. The dotted line represents the zone where shuttle molecules will be probed.

Bottom Panel: Expanded view of the IDT, now viewed from the top. Each finger (and gap) measures 50 microns in width, and is 1.45 mm long. The whole structure measures 3.95 mm in length, and is 3.0 mm wide. The two black pads above and below the fingers are electrical contacts where wires can be ultrasonically welded.

The fabrication of the SAW detectors is conducted here at Caltech. A 200 nm gold thin film is evaporated on a 20 nm chromium adhesion layer on high purity lithium niobate (LiNbO₃). A thin layer (approximately 2 microns) of photoresist is spin coated on top of the gold layer. Following a short baking step, a transparency mask is applied on top of the photoresist, and the device is irradiated with ultraviolet radiation. The photochemically modified photoresist can then be removed easily with a cleaning solution. The unprotected gold is then dissolved, followed by the chromium. Then, the remaining photoresist is removed and the gold pattern is revealed. Thin wires are then ultrasonically welded to the busbars (the black pads, bottom panel of Fig.1) to allow electrical connection to the IDT's. To further protect the substrate and the IDTs, the whole assembly can be spin coated with a hydrophobic polymer layer prior to laying down the immobilization matrix.

The resulting structure is shown in Figure 2 below. IDT's of several dimension were constructed to determine the smallest feature that our fast and simple fabrication technique could resolve. A finger thickness of 50 microns was determined to be the best compromise in terms of resolution and reproducibility.

Figure 2: Picture of the first SAW detector prototype. Several pairs of IDT's were constructed, with dimension ranging from 20 to 60 microns (from top to bottom, on the picture). Upon observation with an optical microscope at a 20x magnification, the 50 micron "fingers" IDT was chosen. Wires were attached to the busbars with silver epoxy, and the impedance of the device was measured to insure that no structural resonance would hinder the collection of acoustic waves at the resonant frequency of about 9 MHz (see text). To minimize cost, this prototype was built on a quartz substrate which is not piezoelectric.

Considering a surface wavelength of 400 microns (8 times the finger width of 60 microns) and a known wave velocity of 3500—4000 m/s [Ref.], the resonant frequency of the device can be estimated to fall between 8.75 and 10 MHz. By modulating the excitation radiation, maximum sensitivity for pickup of surface acoustic waves can be achieved. Laser modulation at these rather low frequencies is routinely achieved, and can be performed with inexpensive, "off the shelf" components offered by many suppliers.

The proposed experiment focuses on the elucidation of complex respiratory processes in bacteria using insoluble minerals (i.e. that cannot diffuse through the cell membrane) as terminal electron acceptors. Previous study of the bacteria Schewanella by Newman and Kolter [2] has shown that these bacteria do respire using insoluble iron oxides as terminal electron acceptors, and that their respiration relies on the biosynthesis of quinone analogs. These analogs, possibly related to the cells ability to synthesise menaquinone, may also allow them to respire minerals by acting as electron shuttles. So far, very little is known about microbial electron transport to insoluble electron acceptors. Further work by Lower et al [6] has shown that Schewanella oneidensis triggers the production of putative redoxidases when the bacterium recognizes the presence of goethite (a—FeOOH). The living cells were put in direct contact with the surface of the mineral, and sub nanonewton resolution force measurements were analysed, and suggested that a 150 kDalton putative iron reductase is mobilized within the outer membrane to facilitate electron transfer. However, long—range interactions between the bacteria and goethite were not studied.

The present experiment will add to the understanding of interactions between members of a Schewanella bacterial communities and iron mineral by allowing a dimensional study of microbiological activity. For instance, a colony of Schewanella could be immobilized on a gel, in a well defined region. The appropriate oxidant (electron acceptor: in this case an iron oxide like goethite or hematite, a—Fe2O3) would also be immobilized, at a well characterized distance from the colony. Then, using the well known spectroscopic signature of both the oxidized and reduced forms of quinone substituents (previous studies have shown that organic electron shuttles feature a quinone moiety), the diffusion of electron shuttles could be monitored. This way, an accurate depiction of the chemical process underlying the respiration process would be obtained.

Studying the dependence of the shuttle concentration as a function of the separation between the bacteria and the iron would prove useful in elucidating the respiration mechanism in Schewanella. These data will provide invaluable information about the ability of microorganisms to both identify the presence of mineral oxidizing substances, and generate soluble molecular shuttles to transfer electrons to the mineral. These two mechanisms are known to be fundamental to extracellular bacterial respiration in anaerobic environments.

(last updated 25 February 2003)

References