Application of TaqMan RT-PCR for Real-Time Semiquantitative Analysis of Gene Expression in the Striatum - Drugs of Abuse: Neurological Reviews and Protocols - page 234

correct for this, each value can be divided by the corresponding copy number
obtained for an appropriate housekeeping gene determined in a replicate
plate, and data plotted graphically as a proportion of that reference gene. The
appropriate housekeeping gene for normalization is dependent on the model
system being investigated, and is best determined by running at least three
different genes across samples to fi nd the one most consistently expressed. In
some cases, there is little difference between data when expressed as copies
per nanogram or relative to a housekeeping gene as seen with the D 2 receptor
enrichment in caudate and putamen ( Fig. 5 ).

4. Notes

1. We do not routinely treat our RNA samples with DNase, as experiments per-
formed in our laboratories showed little effect of DNase treatment on TaqMan
PCR signals, refl ecting the minimal presence of contaminating genomic DNA
following the modifi ed Trizol RNA extraction protocol. Instead we used the
subtraction method to correct for minimal genomic DNA contamination as
described in Subheading 3.4. It is also important to be careful which DNase
system is used. For example, DNase reactions run at 37°C for 60 min with high
concentrations of MgCl 2 in the buffer result in Mg 2+ -dependent RNA hydrolysis
as demonstrated by increased C t values obtained using TaqMan RT-PCR analysis.
When required, we have used DNase I amplifi cation grade (Life Technologies)
successfully, which involves only 15 min of incubation at room temperature,
followed by termination of the reaction with EDTA, which did not cause any
RNA hydrolysis or affect TaqMan signals downstream.

2. When large numbers of cDNA samples were compared across different plates
plates, four samples in triplicate were included on both plates. Comparison of
the mean values from these four parallel samples allowed the calculation of a
correction factor between plates, which took into account any variation between
different TaqMan runs. If two identical plates are run for the same gene in
parallel, we observed similar expression profi les across the plates, but some
difference in actual fluorescence levels between plates, hence the use of a
correction factor.

3. When aliquoting our cDNA into 96-well plates using the Hydra96 (Robbins
Scientifi c), we fi rst dilute the 20 L of sterile water

to each well. This allows an increase in dispensing volume to 4

µ L of cDNA by adding 60 µ

µ L per well.

This routinely gives approx 18-19 plates from 20 µ L of cDNA. We have also

Fig. 5. (opposite page) Classical enrichment of D 2 receptor expression in human
caudate and putamen. Data are expressed as (A) arbitrary units, (B) normalized to
GAPDH, and (C) normalized to cyclophilin. With a discrete distribution like this one,
the data is not really affected by normalization to the housekeeping genes.