Biomedical Engineering Reference
In-Depth Information
functions of Plexin B3 to be associated with SEMA5A. We modeled the structure
of the semaphorin domain of Plexin B3 and found that it shares similarity in
the ligand-binding region with that of SEMA5A. Furthermore, we observed co-
expression of SEMA5A and Plexin B3 in aggressive pancreatic cancer cell lines.
Interestingly, phylogenetic analysis has shown that SEMA5A and Plexin B3 co-
evolved. We coni rmed the interaction of SEMA5A and Plexin B3 in normal
tissue samples using co-immunoprecipitation (Sadanandam
et al
. 2008b). h is
demonstrates that integrated methods can be used at the genome level to discover
protein interactions of many unknown CAMs expressed on the surface tumor and
secondary tissues using ligand-binding information of CAMs.
GENE EXPRESSION MICROARRAY ANALYSIS AND
BIOINFORMATICS
h e advent of DNA microarrays revolutionized many aspects of biology by allowing
investigators to quantitatively study gene expression on a genome-wide scale. h e
applications of microarrays are many, but for this review we will be describing
microarray technologies that measure gene expression in tissues or cells. h ere are
two primary types of gene expression microarrays, cDNA and oligonucleotides
arrays. Both of these arrays work on the principle of hybridization between nucleic
acids to measure the mRNA abundance of tissues or cells. In the case of cDNA arrays,
polynucleotides corresponding to 3ยข end of RNA transcripts are immobilized on a
solid support, whereas the probes on oligonucleotide arrays are built up one base
at time using, for example, photo-masking technologies. Although experimental
objectives and resources typically determine the most suitable platform, some
investigators have suggested that the choice of platform can signii cantly inl uence
experimental outcomes. In response to this criticism, more recent comparisons of
microarray technologies have found good cross-platform agreement, but in order
to achieve this level of consistency, experimental and analytical protocols must be
carefully observed (Kuo
et al
. 2006). All of the microarray platforms have at least
one thing in common: they generate large volumes of data that are best analyzed
using appropriate statistical and bioinformatics methods.
Gene expression microarrays can be used to identify CAMs responsible for
organ-specii c metastasis of tumor cells (Kakiuchi
et al
. 2003). In this vein, Kakiuchi
et al
. used a metastatic model by injecting SBC-5 lung cancer cell line intravenously
into mice lacking tumor-associated natural killer cells. By performing permutation
tests of the metastatic foci of the lung cancer cells developed in lung, liver, kidney
and bone, they identii ed 435 genes that are dif erentially expressed between
dif erent organ-specii c metastatic foci. h ey observed that the dif erentially
expressed genes were enriched for CAMs. In particular, they observed that lectin
family of proteins such as LGALS1 in pulmonary metastases and LGALS9 in
renal metastases are highly expressed. Lectin, a family of b-galactoside-binding
