Biomedical Engineering Reference
In-Depth Information
(a)
(b)
2
1.5
Fiber diameter
1
SHG(
P
ICS,
W
ICS
)
SHG(
P
PA,
d
SHG
)
m
= 2.2
R
2
= 0.68
m
= 0.84
R
2
= 0.93
0.5
0
Mesh size
-0.5
-1
-4
Log [(
L
m
estimate
-14/5
)(
d
estimate
-4/5
)]
-3
-2.5
-2
-1.5
-1
-3.5
FIgurE 11.8
(
See color insert.
) (a) Combined SHG + TPF (SHG, blue signal; TPF, green signal) image of acel-
lular collagen hydrogel showing examples of measurements of collagen fiber diameter and network mesh size. (b)
Log-log plot of the scaling relationship of the storage modulus versus mesh size,
L
m
, and fiber diameter,
d
, estimated
from SHG images of collagen gels polymerized at temperatures 4°C, 14°C, 24°C, and 37°C. The estimates were from
image correlation spectroscopy (
P
ICS
for mesh size;
W
ICS
for fiber diameter), or particle analysis and manual image
measurements (
P
PA
for mesh size;
d
SHG
for manual image measurements). The linear best-fit slopes and
R
2
values
are indicated in the figure. (Reprinted from
Biophys J
., 94, Raub, C. B. et al., Image correlation spectroscopy of
multiphoton images correlates with collagen mechanical properties, 2361-2373, Copyright 2008, with permission
from Elsevier.)
11.3.8 Summary: SHG imaging of Acellular collagen Gels
In the studies described above, acellular collagen gels were polymerized at a range of concentrations
(1.5-9 mg/mL), temperatures (4-37°C), and pH values (5.5-8.5). The segmented signal intensity,
image area fraction, average fiber diameter, and average mesh size were measured from SHG images
of the gels. In some cases, a scaling relationship from semiflexible network theory was applied,
showing good correspondence to experimental data inputs (average fiber diameter, mesh size from
SHG images,
G
′ from rheology). While this correspondence shows the sensitivity of SHG imaging
to mechanical properties, there are limits to the microstructural information captured by SHG sig-
nal. Most notably, the smallest structures that can be measured from SHG images are determined
by the optical resolution of the system. Structural information on the order of single pixels can still
be obtained through careful measurement of optical parameters—such as signal orientation depen-
dence with respect to polarizers, or forward-to-backward signal ratios. Second, SHG signal decays
exponentially with penetration depth into the tissue [1,2,96]; so, the interrogated microstructure
must be within a resolvable distance (typically ~150 μm to several mm). Third, SHG is specific only
for certain structural proteins, most notably collagen and myosin [3,97-100]; so, SHG imaging will
fail to capture mechanically relevant matrix components that do not emit, such as elastin, pro-
teoglycans, cells, and noncollagenous tissue structures. The matrix components that do not emit
second harmonic may be characterized through other imaging modalities and signals, such as TPF
[21,22,26,59,101,102]. Acellular collagen gels, however, possess a uniform microstructure and are
free of noncollagenous matrix components, which allow for very thorough structural characteriza-
tion of these gels from SHG images alone. Tissue-engineering experiments utilize collagen gels as
matrix scaffolds for cells, which typically remodel the gel through stress generation, protease activ-
ity, and new matrix deposition. SHG imaging, especially in conjunction with other nonlinear optical
signals, can provide detailed information about the dynamic changes during culture of cellularized
collagen gels.
