BioMEMS for Cell Biology - Introduction to BioMEMS - page 331

Signal
gradient

a

Bacteria inlet

Signal inlet
(low)

Signal inlet
(high)

1000 µm

Gradient generator

100

0 µM 100 µM 0 µM 225 µM

b

Ni 2+

c

80

Aspartate

60

40

Live cells
Dead cells

20

100 µm

0 0

0.2 0.4 0.6 0.8
Normalized position across
chemotaxis chamber

1

FIGURE 6.33 Bacterial. chemotaxis. in. a. Dertinger. gradient. generator.. (From. Derek. L.. Englert,.
Michael. D.. Manson,. and. Arul. Jayaraman,. “Flow-based. microluidic. device. for. quantifying. bacte-
rial.chemotaxis.in.stable,.competing.gradients,”. Appl. Environ. Microbiol. .75,.4557-4564,.2009..
Figure.contributed.by.Arul.Jayaraman.)

One of the most powerful attributes of microluidics is that it allows for mimicking the physi-
cochemical properties of cellular microenvironments. Martin Polz and coworkers at MIT used
microluidic devices to simulate the tracks of dissolved organic matter, such as coming from snow
particles or lysed algae, that slowly settle by gravity in the ocean, and accumulate feeding bac-
teria around them. hese “bacterial hot spots,” simulated in the microluidic devices as plumes
ejected from a hole ( Figure 6.34 ), are able to attract the marine bacterium Pseudoalteromonas
haloplanktis within tens of seconds, more than 10 times faster than E. coli (leading to twice the
nutrient exposure). In other words, the rapid chemotactic response of P. haloplanktis substan-
tially enhances its ability to exploit nutrient patches before they dissipate.
Roseanne Ford's group at the University of Virginia has used microluidic devices (contain-
ing 200-μm-diameter tightly packed cylindrical posts) to simulate the percolating low through
dirt ( Figure 6.35 ). he average low velocity in the channel (5-20 m/day) was used to match that
of groundwater lows. Gradients of α-methylaspartate were formed with a T-mixer upstream
and cells counted within each pore (between the posts). he observed chemotactic response
( Figure 6.35b ) was larger than predicted, which could be due to (a) within-pore dynamics, or
(b) hydrodynamic efects (e.g., trapping) confounding some of the bacteria-surface interactions.
Bacterial cells are swimmers; they do not attach. Hence, in bacterial chemotaxis experiments,
low is a dangerous proposition (unless it is one of the biological variables under investigation). he
biggest problem is that, because of the parabolic low proile, the cells are not all being carried at
the same low velocity; if the cells are moving quickly across the velocity proile ield, the problem
becomes even more complicated. A secondary problem is that, when there are concentration ields
superimposed with low, dispersion phenomena such as the butterly efect (see Figure 3.73 ) will
occur; cells that are at diferent z distances from the surface will experience diferent gradients.
Norman Stocker's group at MIT was the irst to apply microluidic design principles to reduce
the exposure of bacteria to low in bacterial chemotaxis experiments ( Figure 6.36 ). A major
advantage of “stopping low” is that microscopic observation becomes straightforward, and