Biomedical Engineering Reference
In-Depth Information
of amplification [4, 5]. Further, in one study it was shown that the yield from FNAs was
highly variable, ranging from 30 000 to 2 580 000 cells [5]. Assersohn
et al
. were only able
to process 15% of the total number of patient samples due to the limited RNA obtained
from the majority of FNAs.
As a greater understanding of the complexity of diseases, and cancers in particular, has
developed it has become increasingly clear that the study of pure cell populations is critical.
More researchers are turning to methods such as laser capture microdissection (LCM),
fluorescence-activated (assisted) cell sorting (FACS), micromanipulation and paramagnetic
bead-based separation. While each of these technologies allows for the purification of more
homogeneous cell populations, it is often difficult to obtain completely pure cell populations
without resorting to single-cell analysis.
The need for technologies that can assay samples at the single-cell level is becoming
increasingly apparent. During development, it has been shown that many cells may look
identical morphologically, but that the cells have already started down their path of differen-
tiation and, thus, are showing discrete gene expression profiles [6]. In order to characterize
these stages of development fully, it is necessary to look at individual cells. Neurological
systems are classical examples of heterogeneous systems, which benefit greatly from the
study of single cells [7-9]. In addition, there is increasing interest in rare cell populations,
such as stem cells, cancer stem cells, precancerous lesions and circulating tumor cells, often
comprising a very small number of cells or even a single cell. Despite the large number
of publications which have used microarrays to study various cancers, there have been few
true breakthroughs that have resulted. One possible explanation for this is that while the
studies conducted to date have contributed greatly to the understanding of cancer biology,
they have in fact missed what is perhaps the most important message: that of the cancer
stem cell or cancer initiating cell. With a frequency of only 1 in 100 to 1 in 100 000 cells
[10-12], clearly the signatures of these cells would be lost among those of the more abundant
cell types.
Thus, it is clear that in order to push toward more meaningful analysis of the biology of
development, physiology and disease, improvements in the microarray technique that allow
for greatly reduced amounts of input material are required. While the sensitivity of the
technology itself has been increased somewhat through increased array densities, improved
substrates and better detection technologies, the gains in overall sensitivity have been modest
at best. The greatest advancements have come from amplification technologies that have
in many ways opened the door to analysis of minute RNA samples, including single-cell
analysis.
5.1.2 Amplification approaches
The restrictive nature of the relatively high sample size requirement was recognized with
the initial development of microarray technology, and companies such as Affymetrix, Agi-
lent and Illumina have all incorporated amplification as a standard part of their procedures.
Despite this, the sample requirements still remain relatively high, with between 0.1 and
8
g of total RNA (10 000-20 000 cells) being required (Table 5.2). Most of these vendors
have concentrated on isothermal, linear amplification strategies based on
in vitro
transcrip-
tion (IVT) reactions. While many researchers heralded these improvements, the sample
requirements are still too restrictive for many important biological questions.
μ
