SHG and Optical Clearing - Second Harmonic Generation Imaging - page 177

8.5 optical clearing in collagenous tissues

Analogous experiments coupled with Monte Carlo simulations showed even more pronounced reduction
of scattering in murine tendon ex vivo . The treatment with 50% glycerol decreased the scattering coefficient
μ s from 510 to 27 cm −1 at SHG wavelength, and from 399 to 3 cm −1 at 890 nm excitation, Table 8.2 [24]. The
reduction of scattering is achieved by replacing interfibril water with index-matching glycerol solution,
which increases the spacing between the fibrils and results in swelling of the tissue. Interestingly, in human
skin samples treated with high glycerol concentration (75%), electron microscopy revealed that diameters
of collagen fibers become significantly smaller [19]. This particular finding helps us to understand better
the F / B SHG curves for the tendon treated with 25%, 50%, and 75% glycerol [6]. In the previous section, we
discussed muscle, where there was no difference in the attenuation observed between 50% and 75% glyc-
erol solutions, as the resulting bulk optical properties were also similar. In contrast, in tendon samples, F / B
ratio drops from 5 to 3.5 when glycerol concentration is increased from 50% to 75%, causing the individual
fibrils to dehydrate and decrease in size. This suggests that SHG-producing domains can now give rise to
a significant share of SHG emitted backward directly. Consequently, collagen will be a much better target
for in vivo imaging with SHG, where only the backward-propagating signal can be detected. Recall that the
emission in muscle was highly forward directed even in the cleared case.
It is important to further determine if there are any detrimental effects that accompany optical clear-
ing. One of the controversial subjects of this quest is to understand the deterioration or complete dis-
appearance of backward-detected SHG signal from collagenous tissues treated with high-concentration
glycerol solution, as there have been conflicting arguments made by different research groups. Initially,
Yeh et al. [25] showed decreased SHG signal from RAFT (real architecture for three-dimensional tissue)
models and skin, and suggested the mechanism of reversible collagen dissociation, where clearing agents
reversibly affect higher-order collagen structures. Specifically, the interaction between collagen and optical
clearing agents screens noncovalent attractive forces, inhibiting collagen self-assembly in the solution and
destabilizing higher-order collagen structures at microscopic and ultrastructural levels. In this eventual-
ity, the asymmetry of dipoles arrangement would be broken and no SHG would be produced. Yeh et al. [33]
reported a correlation between collagen solubility and optical clearing, where they studied a self-assembly
of solubilized collagen from murine tail tendon incubated for 24 h with highly concentrated clearing agent.
In a similar study, both dimethyl sulfoxide (DMSO) and glycerol were shown to exhibit a weak
concentration-dependent inhibition of the self-assembly of collagen molecules into fibrils in vitro [34].
Previously, the efficiency of collagen fibrillogenesis was shown to decrease with the length of molecular
chain of clearing agents (e.g., as shown in Figure 8.1) [35,36]. Most recently, Yeh and coworkers [37] pub-
lished additional experimental and theoretical arguments to support the adverse effects of sugars and
sugar alcohols on higher-order collagen structures. The interactions between the clearing agents with
collagen are described as destabilizing, but are nonreactive and reversible upon agent removal. These
interactions were investigated using a combination of optical spectroscopy, integrating sphere mea-
surements of bulk optical parameters, and molecular dynamics simulations. The simulations revealed
a theoretical correlation between the rate of optical clearing and the affinity to form hydrogen bond
bridges between a clearing agent and collagen molecules. The study suggested that such a bridge forma-
tion disturbs the collagen hydration layer and facilitates water replacement by an OCA. According to

TABLE 8.2 Bulk Optical Parameters for Murine Tendon at the Fundamental and SHG Wavelengths

λ 457 nm 890 nm

% Glycerol Control 50% Control 50%

g 0.96 0.96 0.96 0.96

μ s (cm −1 ) 510 ± 300 27 ± 7.1 399 ± 44 3.1 ± 2.1

μ a (cm −1 ) 5.0 ± 1.7 2.1 ± 1.1 2.2 ± 0.3 1.5 ± 0.7

Source: Reproduced from LaComb, R. et al. 2008. J. Biomed. Opt . 13:021108.