Biology Reference
In-Depth Information
Promoter from the period gene
Luciferase coding region
FIGURE 11-7.
Schematic diagram of a per-luc construct. The DNA from the controlling region of the period gene
(the period promoter) has been fused to the DNA making up the coding region of the luciferase gene,
producing a luciferase that is then under circadian control.
functionally coupled groups of neurons. But to probe the functional
coupling between rhythms expressed in other non-SCN brain regions
and peripheral tissues and organs, it is important to be able to study
these non-SCN rhythms in vitro, as well as (eventually) in vivo.
Assessment of the regulatory patterning of the period circadian clock
gene (per) became possible through the use of a reporter gene—in this
case, a luciferase cDNA fused to the promoter region of per (see
Figure 11-7). Luciferase is the enzyme responsible for bioluminescence—
when luciferase acts upon its substrate, luciferin, light is produced.
This recombinant reporter gene DNA, called a per-luc construct, was
introduced into rat zygotes to produce what are called transgenic rats,
carrying the per-luc DNA in every cell of their bodies. These rats are
entrained to a circadian rhythm, and then the tissue to be studied is
excised from the animal and put into tissue culture and supplemented
with luciferin. Light emission from the cultured tissue then reveals when
the per gene is being expressed, demonstrating the existence of
a circadian rhythm in the tissue.
F. Technological Challenges in Analyzing Circadian Data
The bioluminescence time series data generated from per-luc
experiments often show patterning in which average bioluminescence
intensity is decreasing with time (drifting downward) as a result of the
depletion of luciferin, a substrate necessary for light production by the
luciferase enzyme (see Figure 11-8). In addition, as in Figure 11-8, a
situation is often encountered in which the oscillatory magnitude
changes with time. This represents a phenomenon referred to as variance
nonstationarity (i.e., the variance exhibited in the time series changes as a
function of time). In such cases, data normalization will be necessary to
reveal the original rhythmic structure. Finally, further problems in the
analysis of time series data come from the presence of noise that may
often obscure the circadian rhythms.
Recall from Chapter 9 that most standard methods designed to
determine periodic components in a time series, such as the fast Fourier
transform (FFT) algorithm, require time series that do not have
a trend. Variance nonstationarities and noise present additional