Biomedical Engineering Reference
In-Depth Information
to stored matrices (signatures) for known targets (chemicals). The mathematical methods
for carrying out signature analysis are well known [110-113], and in cases where each
measurement is given equal weight, can be represented numerically as a least-squares
sum over the deviations of the observed parameters from the values measured for each
specific target. Signature analysis gains comparative advantage by allowing weighting of
components based on iterative optimization to enhance both sensitivity and accuracy.
Implementation of the sensor architecture described above requires a platform that can
simultaneously measure both the protein's photochromokinetic response and photoelec-
tric effect. Design of such an architecture poses a challenge, in that the photoelectric effect
can be much more difficult to detect than the protein's photokinetics. To register both
responses, use of a concentrated sample deposited on a conducting, and perhaps trans-
parent, substrate will be necessary, with a second conducting lead in contact with the
sample surface. The photokinetic response is relatively easy to monitor, given the appro-
priate optics and circuitry; a fairly basic flash photolysis set-up monitoring the protein's
response at two wavelengths (M and O) is all that is needed. However the photoelectric
effect, whether measuring light-induced current or voltage, requires that the sample
essentially be sandwiched in a capacitor that can be used to measure light-induced dis-
placement currents. Measuring the photocurrent component may require a hydrated sam-
ple to allow the protein to pump protons. However, the studies of Xu et al. [64] and Bryl
et al. [88] detected photo-induced current in dried samples, indicating that ambient
humidity provided enough moisture to elicit a signal. There are several examples in the
literature where dried biological preparations (including BR) have proven to be stable and
maintain high levels of activity. Greenbaum et al. [114], demonstrated the use of dried tis-
sues for detection of chemical antagonists; samples consisted of immobilized photosyn-
thetic algae deposited on filter paper disks, and utilized a fluorescence detection scheme.
BR is nonliving and extremely stable and robust when dried into films (as required for the
photoelectric effect). Such films can be deposited in a variety of ways, including elec-
trophoresis, simple evaporation, or chemical ligation to modified surfaces. As discussed
above, additional stability can be achieved by polymer or sol gel encapsulation, which
maintains protein immobilization and stability, while providing a porous structure to facil-
itate diffusion of volatile agents. Ultimately, sensor miniaturization onto a semiconductor
platform will be required. Issues with integration of the protein-based component of the
sensor have been covered above, and a discussion of overall miniaturized sensor design is
beyond the scope of this chapter.
14.3.2.2.1 Preliminary Results
Currently, work is progressing along the lines of characterizing the response of the wild-
type protein and a selection of genetically engineered BR variants to a variety of chemicals
known to elicit a photochromokinetic response [115]. Preliminary results from those trials
are presented here. A bench-scale flash-photolysis prototype has been developed in-house
that is capable of measuring the rise and decay kinetics of the M and O states. The
prototype was designed to demonstrate proof-of-principle, and does not fully represent
the technology ultimately needed for an in-field sensor platform. The actinic beam is pro-
vided by a 532 nm and 20 mW laser that initiates the BR photocycle. Blue (450 nm) and red
(650 nm) light emitting diodes are used to monitor the photokinetics of the M and O states,
respectively. A pair of photo detectors records the resulting data traces, which are fit to
exponential curves to determine the M- or O-state lifetimes and decay rates. The current
prototype lacks the capability to measure the photoelectric effect. Aqueous samples of the
wild-type protein, as well as a number of randomly selected genetically engineered vari-
ants, were prepared carefully with a series of chemicals known to modulate photocycle
kinetics [115]. Included were 1,2-diaminopropane (1,2-DAP) and triethylamine (TEA),
