Biology Reference
In-Depth Information
human bone marrow at single-cell resolution for a combined
total of 52 cell surface and intracellular proteins
[2]
. Such
studies are essentially ex vivo, as they are done on non-
separated primary cells that have been stimulated and
fixated such that the effects of the local milieu on cells is, to
a large extent, maintained. Cells were measured not only
at rest, but also when stimulated with a variety of stimuli
and inhibitors, and the abundance of phosphorylated
proteins was measured at a single-cell level. The increased
dimensionality into the immune system is so high that
standard methods for manual cell subset identification (also
known as 'gating') are inadequate, and novel automatic
gating and visualization strategies had to be developed
[1,26]
(see
Figure 25.2
A,B). Furthermore, the ex vivo
profiling methodology used and the orders-of-magnitude
higher resolution attained are assured to advance the
clinical goals of improved diagnostics and personalized
healthcare
[27
is illustrated by data showing that different stimulations of
these cells yield different responses, and that T cells
specific for different viruses express both distinct and
overlapping sets of cytokines
[1]
. This suggests that the
immune system can take a combinatorial approach to its
functional responses in addition to the genetic diversity
encoded in its antigen receptors. Alternatively, this high
amount of variation may be quiescent with respect to
cellular output or yield a uniform response across cells.
Single cells studies of NF
k
B response to TNF-
a
stimula-
tion indicate that cells respond heterogeneously and in
a binary manner (i.e., either respond or not respond) as
a function of stimulus concentration
[32]
, suggesting that
initial conditions would play a significant role in deter-
mining which cells respond and the nature of the response.
The large degree of functional diversity, both in cells of
the same specificity as well as across the entire hematopoi-
etic lineage
[2]
, provides the immune system with remark-
able flexibility and indicates a need for mechanisms that
enable intercellular communication and convergence of
responses between cells. Supporting this, Love and
colleagues recently showed the existence of an intricate
regulatory program for cellular cytokine release
[6]
. Using
a novel microfluidic device that allows for serial, time-
dependent single-cell analysis
[33]
(see
Box 25.2
), they
showed that polyfunctional human T cells (i.e., T cells
releasing multiple cytokines) release their cytokines
predominantly in a sequential manner, which begins asyn-
chronously from cell to cell
[6]
. As virtually all previous
measurements were made at a specific single endpoint, this
temporal complexity was previously masked and just inte-
grated sum over time of cytokine secretion noted.
e
31]
. Simultaneously, they are expected to
reveal
important new insights
into immune system
heterogeneity.
Recently, Newell et al. analyzed antigen-specific CD8
þ
T cells to map the expression landscape of 17 cell surface
proteins and nine intracellular cytokines known to be var-
iably expressed in CD8
þ
T cells. Through the use of
principal component analysis, a mathematical trans-
formation which converts a set of possibly correlated
variables (i.e., 17 cell surface proteins) into a set of linearly
uncorrelated ones (i.e., a combination of said proteins), this
large phenotypic landscape was projected onto a set of
vectors that best explain the variation observed in the data.
Plotting the first three principal components on a three-
dimensional axis showed a continuum of cells, reproduc-
ible between individuals, with 'hotspots' in established
'phenotypic clusters' such as na¨ve, central memory and
effector memory (
Figure 25.2
B). In total, the first three
components explained 60% of the observed variation.
Additionally, the data showed more than 200 functional
phenotypes (different cytokine signatures) just within the
CD8
þ
T-cell subsets, showing a much greater complexity
than previously appreciated. How this diversity is generated
Limitations of a Cell-Focused Approach
With advances in automation, a scaling-up of methodologies
is likely to occur such that ultimately we will be able to
delineate regulatory networks for a large variety of cell
types. Immune cells are highly affected by the complex local
microenvironment they experience. Thus, how much of the
BOX 25.2 Empowering single-cell measurements through microfluidics devices
Much of our knowledge in immunology comes from bulk
measurements of many cells together. Owing to problems of
averaging and noise, the behavior of cells as inferred from
average measurements often drowns cell-to-cell differences
and may not reflect the behavior of any single cell. Microfluidic
devices can control the flow of minute amounts of liquids or
cells and may be designed for sensitive detection of analytes at
a single-cell level. Detection (often light) is usually set up in an
automated manner such that many cells may be measured. For
example, cells may be moved through the device by capillary
flow, measured and imaged, or placed stationary into an array
of serially micro-engraved wells, each of which is imaged. A
recent newcomer to immunology, microfluidic devices are
already being utilized for single-cell gene expression studies
run in multiplex (across multiple cells and genes)
[5]
, the
measurement of single cells cytokine secretion
[6]
and cell
counting
[7,8]
, to name but a few. The miniaturization of these
devices and their relative low cost and transportability are
promising for the future development of microfluidics-based
diagnostics.
