How an Autosequencer Works

    This information is intended to be supplemental to that presented in the lab manual.  Please read the section about the Sanger method of DNA sequencing in the background information in the manual before viewing this page.

Before sequencing can begin, template DNA must be made.  This is usually done by amplifying a small sample of DNA using PCR.  In PCR, the part of the DNA sequence to be amplified is determined by the choice of primers.  In the example in the manual, the primers are called tab-a and tab-b.  They bind to opposite strands of DNA separated by several hundred base pairs. 

The resulting PCR product contains numerous copies of the part of the sequence between the two primers.  These strands serve as the templates for the Sanger reaction described in the manual. 

Because an autosequencer is intended to separate DNA fragments that are much smaller (500-600 bp) than those we have been separating with agarose gels, a polyacrylamide gel is used.   An autosequencer can simultaneously run a number lanes of samples, so the gel itself is wide.  In addition, to achieve separation of the fragments the gel must also be long.  Before forming the gel, the plates must be cleaned meticulously.  Spacers are placed between the glass plates and the plates are clamped together to form the gel cassette.  Polyacrylamide is mixed and injected between the plates before it polymerizes. 

After the gel polymerizes, the cassette is placed vertically inside the autosequencer.  An apparatus with wells is inserted on top of the gel and the plates are covered with a heater that maintains the gel at a uniform temperature (below).  The high temperature keeps the DNA single stranded and the uniformity maintains a consistent size/velocity relationship for all lanes on the gel.

The gel is run in a different manner than the agarose gels that we did in lab.  Instead of turning off the electric field after a certain amount of time and observing the position of the bands on the gel, the gel is run until all of the bands run off the bottom of the gel.  Each band consists of a nucleotide chain that was terminated by one of four dNTPs tagged with a dye that fluoresces at a characteristic frequency (one for each of the four types of nucleotides).  As each chain passes out of the gel (in order of size), its fluorescent dye is excited by an argon laser.  The amount of light at the four frequencies is measured and recorded by a computer.  

The computer monitor in the image above is displaying the intensities of the fluorescence measured for the samples in an autosequencer run.  At the left, the intensities of the four frequencies (colors) for a single sample are overlaid in a single graph.  Computer software analyzes this trace and assigns a nucleotide identity to each peak depending on which frequency is the most intense.   Sometimes the software is unable to decide and assigns "unresolved" (N) to that position.  The quality of the signals also declines as the fragment sizes get larger and the software begins to misidentify nucleotides.  For this reason, a long sequence is usually derived from sequencing smaller sequences that overlap each other by about 300 bp.  The reverse strand is usually sequenced as well so that it can be compared to the forward strand.  In projects where the highest sequence quality is required (e.g. the Human Genome project), a stretch of DNA is sequenced several times before a consensus sequence is created.