Lecture by Anne Condon.
MASSIVELY PARALLEL SIGNATURE SEQUENCING (MPSS)
Source paper: Brenner et al.
"Gene Expression Analysis ...", Nature Biotech, Vol. 18, June 2000
There are 2 more important technologies used for gene expression analysis:
Use of microarrays for gene expression analysis assumes that one knows a superset of the genes to be expressed.
An advantage is that one can detect a wide range of expression levels, which means that rare mRNA's can be detected.
Limitations of microarray technologies (particularly cDNA
microarrays) include the variability in measurements, due to probe hybridization
differences and cross-reactivity, element-to-element differenceswithin
micro-arrays, and microarray-to-microarray differences.
SAGE is considered to be statistically more robust than
microarrays (because counting statistics are well modelled by the Poisson
distribution) but is less scalable. It is based on beads counting.
New technolgy: MPSS
Brenner et al. proposed a method for gene expression analysis, which they call massively parallel signature sequencing, that addresses the limitations of the above two technologies. The method is as follows:
a) Copies of cDNA templates are attached to glass
microbeads, with a unique template on each bead. (If the template library
has multiple identical strands, each gets its own bead.)
b) The beads are arrayed in a flow cell.
c) Bases of the templates are read off, one by
one, using fluorescence-based methods
Attaching templates to beads
In the reported experiment, the template library contained 3-4 x 104 mRNA strands. Strands are converted to cDNA strands using tailed poly-T primers, and the ends of the cDNA strands are restricted. The cDNA strands are then inserted into a cloning vector. The cloning vector contains a set of 1.67 x 107 32-mer oligo address tags, thus forming a set of 5-7 x 1011 conjugates.
Select a small random sample of about 1.5 x 105 conjugates. With high probability, the sample has the following properties:
- each cDNA is represented at least once in the sample
- tach tag is represented at most once
Amplify this sample using PCR, render single-stranded, and add a fluorescent tag.
Prepare a population of microbeads, each with about 106 copies of the complement of an address tag (anti-tag) attached to its surface. The beads were prepared in 8 rounds, with one of 8 4-mers subunit attached to each bead at each round. (Result: 88 distinct tags = 16.7 million tags.)
Combine beads and conjugates (under stringent hybridization
conditions). Conjugates attach to beads, with 1% of the beads getting
about 104-105 copies of a single kind
of template molecule. Using a fluorescence-activated cell sorter, obtain
the loaded beads.
Reading off bases
Bases are read off from the free end in cycles, 4 bases per cycle. Initially, three bases at the 3' end of the molecule are exposed (made single-stranded) using the enzyme DpnII, so the first cycle actually reads off 3 rather than 4 bases. In what follows, we describe the more general cycle, where 4 bases are read off.
a) 16 classes of adaptor molecules are used:
o 4 represent possible bases at position 1
o 4 represent possible bases at position 2
o 4 represent possible bases at position 4
There are 43 possible adaptors in each class, all of which have a common decoder at one end. The 43 adaptors correspond to the possibilities for the three free bases among the four bases at the free end.
Picture for class 2A: the *'s indicate the double stranded molecule that is to be read off, attached to the bead. Four bases at the free end are being read off in this cycle. The dashed part is the same in all adaptors. Decoders will be explained later.
The dashed part is: ACGAGCTGCCAGTC
positions to be read off:
4 3 2 1
(bead)* * * * * * * * * * * * * * * * * * 3'----------------decoder----3'
* * * * * * * * * * * * * * N N A N-----------------anti-decoder-5'
There are 16 times 43 = 210 adaptors in total.
b) Wash all 210 adaptors over the beads and ligate. 4 adaptors will ligate to the templates on each bead.
c) In 16 separate steps, wash one fluorescently labeled decoder over the beads. This decoder hybridizes to the beads with matching anti-decoder. There is a unique anti-decoder per class. Read off which beads light up, using fluorescence imaging, augmented by software to track any movement of the beads in the flow cell.
d) Add type IIs restrition endonuclease BbvI. This
restriction enzyme recognizes the sequence
- GCTGC -
- CGACG -
in the fixed part of the adaptor. This endonuclease chews off the end of the strand attached to the bead, so that the next four bases of the cDNA template are exposed!
Repeat from step a) until the desired number of
bases are read off.
MODULE 7: BIOMOLECULAR COMPUTING
Adleman's Hamiltonian Path Experiment
Input: generate random paths
Biomolecular Computation Research
Advantages (over "solution phase" chemistry):
Process: has 3 steps
Eg. Satisfiability problem