Microarray Comparative Genomic Hybridization for the Analysis of Bacterial Population Genetics and Evolution - Bacterial Pathogenesis: Methods and Protocols - page 49

intensity of the slide. The median ratio of normalized test strain fluorescence to
normalized control strain fluorescence is used to identify gene differences. Genes
are typically considered lacking or highly variable in the test vs. control strain
below a 0.5 test/control ratio and considered present in greater copy number in
the test vs. control strain above a 2.0 test/control ratio ( see Note 9 ).
3. Use GeneSpring software (Agilent Technologies) for further statistical analysis
of the data.
3.6. Phylogenetic Analysis
1. To determine the evolutionary relatedness of strains, use the Cluster program (11)
based on the presence or absence of all open reading frames (ORFs) represented
on the microarray.
2. The output can be represented in a phylogenetic tree constructed using the program
TreeView ( see Note 10 ).
4. Notes
1. ORFs defined as “absent” either are not present in the genome that is being tested
or have sufficient sequence divergence that hybridization will not occur with the
ORFs represented on the microarray under conditions of high stringency. For
each strain, two or three hybridizations are carried out.
2. The reference strain is representative of one of the sequenced strains encoded
on the microarray, and the test strain represents a strain of unknown genomic
content.
3. To validate the microarray hybridization process, use an array validation system
such as the Spot report oligo array validation system (Stratagene, La Jolla, CA).
4. The unlabeled dNTPs are prepared as outlined in the Promega protocol. Briefly,
mix 1 μL of each non-isotopically labeled dNTP to yield 3 μL of a premix
containing the three unlabeled dNTPs, each at 500 μ M . The volume of aqueous
labeled dNTPs should not exceed 50% of the total reaction volume.
5. As suggested in the Promega protocol, unincorporated, labeled nucleotides may
be removed by size-exclusion chromatography using Sephadex® G-50 spin
columns or by selective precipitation of the labeled DNA. This step is unnec-
essary if dye incorporation is above 60%.
6. Streaks of labeled product from strong spots (“comet tails”) may be observed
after scanning. Steps can be taken to reduce this effect including denaturing of
the slides at 95 °C for 2 min before hybridization, reducing the amount of label
used in the reaction, or diluting the PCR products before printing the microarray.
7. To minimize high background fluorescence, one may increase the spacing
between the spots and lower the photomultiplier tube gain (PMT) during the
scanning process. Only genomic DNA of >1.8 A 260/280 ratio should be used for
labeling and hybridization.
8. The occurrence of intense spots in the background of the slide can be minimized
by the use of powder-free gloves and working in a dust-free environment. It is