Polymerase chain reaction and agarose gel electrophoresis of DNA

Electrophoresis run (blank) without PCR product, only control ('ladder') DNA.

Primer-dimers; note everything is less than 100 bp in size

Monitoring the progress of electrophoresis with bromophenol blue, which migrates at the same rate as a 150-300 bp DNA fragment

Seen on an UV transilluminator (blue band is bromophenol)

Often, one is not sure of the optimal temperature of a primer to anneal; in which case, trial and error at several temperatures is done. Modern thermocyclers allow different temperature for each row of wells (gradient PCR). The temperature with most product formation (thickest band) is selected.

Quantitative PCR for gene expression

Testing for gene expression is tricky; one needs to extract mRNA, reverse-transcibe it to cDNA and then run quantitaive (Q) PCR to determine copy numbers, with respect to some housekeeping gene (which is assumed to be equally expressed in healthy as well as diseased cells).

Here's interferon gamma cDNA expression from a healthy lymphocyte; note that the cycle threshold (Ct) is 28th cycle.

However, in a sick subject, IFNG is overexpressed, and determined from 22nd cycle onwards.

Determining difference in expression

To put a number on the mRNA expression between the two samples. one also needs to run a housekeeping gene like glyceraldehyde-3-phosphate dehydrogenase GAPDH. This would produce a table like the following:

Sample	Ct_IFNG	Ct_GAPDH	Difference in Ct
Healthy	28	22	6
Sick	23	21	2
Difference of differences in Ct			6-2 = 4

Thus, the sick sample has 2⁴ = 16 times more IFNG mRNA expression than healthy. (Remember that PCR cycles produce 2ⁿ copies after n cycles).

Is the product pure?

This particular run was carried with SYBR green, a dye which preferentially binds to double stranded DNA. Once we start heating the PCR products, at a certain temparature, the products denature, and SYBR green falls off, causing a drop of fluorescence. The sudden change in fluorescence is plotted in the melting curve.

Note that the y-axis is (negative) derivative of fluorescence, i.e. rate of change; thus the sudden drop is reflected in the peak

The melting curve from both samples show a single peak (pure product) which occur at the same temparature (indicating the same product in both runs).

Thanks: Dr Michael John,Sanskriti Rai, Pankaj Kumar & Uddep Chowdhury.