NIH Public Access

Author Manuscript
Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.
Published in final edited form as:
NIH-PA Author Manuscript

Osteoarthritis Cartilage. 2015 March ; 23(3): 405–413. doi:10.1016/j.joca.2014.11.021.

Optical Clearing in Collagen- and Proteoglycan-Rich

Osteochondral Tissues
Corey P. Neu, Ph.D.*, Tyler Novak, Kateri Fites Gilliland, M.S., Peter Marshall, and Sarah
Calve, Ph.D.*
Weldon School of Biomedical Engineering, Purdue University, West Lafayette, IN 47907

Abstract
Objective—Recent developments in optical clearing and microscopy technology have enabled
the imaging of intact tissues at the millimeter scale to characterize cells via fluorescence labeling.
While these techniques have facilitated the three-dimensional cellular characterization within
NIH-PA Author Manuscript

brain and heart, study of dense connective tissues of the musculoskeletal system have been largely
unexplored. Here, we quantify how optical clearing impacted the cell and tissue morphology of
collagen-, proteoglycan-, and mineral-rich cartilage and bone from the articulating knee joint.

Methods—Water-based fructose solutions were used for optical clearing of bovine osteochondral
tissues, followed by imaging with transmission and confocal microscopy. To confirm preservation
of tissue structure during the clearing process, samples were mechanically tested in unconfined
compression and visualized by cryoSEM.

Results—Optical clearing enhanced light transmission through cartilage, but not subchondral
bone regions. Fluorescent staining and immunolabeling was preserved through sample
preparations, enabling imaging to cartilage depths 5 times deeper than previously reported, limited
only by the working distance of the microscope objective. Chondrocyte volume remained
unchanged in response to, and upon the reversal, of clearing. Equilibrium modulus increased in
cleared samples, and was attributed to exchange of interstitial fluid with the more viscous fructose
solution, but returned to control levels upon unclearing. In addition, cryoSEM-based analysis of
cartilage showed no ultrastructural changes.
NIH-PA Author Manuscript

Conclusion—We anticipate large-scale microscopy of diverse connective tissues will enable the
study of intact, three-dimensional interfaces (e.g. osteochondral) and cellular connectivity as a
function of development, disease, and regeneration, which have been previously hindered by
specimen opacity.

© 2014 Osteoarthritis Research Society International. Elsevier Ltd. All rights reserved.
*
Corresponding Authors: Corey Neu Tel: (765) 496-1426, Fax: (765) 494-0902, [email protected]; Sarah Calve Tel: (765) 496-1768,
Fax: (765) 494-0902, [email protected].
AUTHOR CONTRIBUTIONS
CPN, TN, and SC designed the studies. All authors performed experiments, analyzed data, and prepared the manuscript.
CONFLICT OF INTEREST
There are no conflicts of interest.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our
customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of
the resulting proof before it is published in its final citable form. Please note that during the production process errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
 Neu et al. Page 2

Keywords
confocal microscopy; articular cartilage and bone; SeeDB; chondrocyte
NIH-PA Author Manuscript

INTRODUCTION
Articular cartilage and bone exhibit a gradient in structural heterogeneity through the depth
of the layered osteochondral tissue that gives rise to unique functional and biological
properties [1]. At the articular surface, where cartilage counterfaces contact in the joint to
permit normal daily activities like walking and running, type II collagen fibrils run parallel
to the surface, and are covered by a molecularly (i.e. nanometer) thin boundary lubricating
material, including superficial zone protein (aka SZP/lubricin/PRG4), that contributes to low
friction and wear [2]. Beneath this superficial zone, the extracellular matrix (ECM) of the
middle zone is characterized by randomly oriented type II collagen fibrils and localization of
cartilage oligomeric protein (COMP; [3]). Type X collagen is a characteristic marker of
hypertrophic chondrocytes at or near the deep zone [4, 5], while type II collagen fibrils
orient perpendicular (and anchor) to the calcified cartilage and subchondral bone. Cell and
pericellular matrix morphology, proteoglycan (e.g. aggrecan) expression, and water content
NIH-PA Author Manuscript

all vary by tissue depth from the surface, giving rise to zonal cellular subpopulations within
a load-bearing viscoelastic and anisotropic tissue. Importantly, cartilage tissue structure is
altered – degraded, worn, and softened – during aging and through the progression of
osteoarthritis (OA), a debilitating disease affecting tens of millions of people in the United
States alone [6]. Moreover, recapitulation of the normal osteochondral tissue structure
remains an elusive target for scientists designing regenerative medicine and tissue
engineering strategies [7].

Cartilage and bone microstructure has historically been visualized in situ by optical
microscopy. Unfortunately, the penetration depth of light in confocal microscopy is
influenced by the tissue structure of the collagen-, proteoglycan-, and mineral-rich tissue,
which limits the absorption of excitation energy and increases scattering of excitation and
emission fluorescent photons, and therefore also limits the understanding of native
osteochondral tissue structure. In cartilage, chondrocytes have been visualized in 3D by
confocal microscopy, although only to depths of less than ~100 μm [8, 9]. Two-photon
NIH-PA Author Manuscript

microscopy is capable of penetrating deeper in tissue owning in part to the use of longer
wavelength excitation light [10, 11]. In a direct comparison of optical imaging of cartilage,
two-photon microscopy provided visualization of calcein-loaded chondrocytes to a depth of
~300 μm (or approximately a two-fold increase over one-photon excitation), depending on
the wavelength and setup of the optical system [12]. Nonlinear optical microscopy,
including second harmonic generation, provides increased depth of imaging in thick tissues,
to ~400 μm, though collagen and extracellular matrix are typically visualized without
revealing detail of embedded cells or cellular substructures [13, 14]. Moreover,
histomorphometry techniques are widely used to study osteochondral tissues, but they
require invasive (~5 μm thick) sectioning that limit a full three-dimensional characterization
of cells throughout the complex osteochondral tissue, and in the context of disease and
regeneration.

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 3

Optical clearing by sugar-based and hydrogel replacement techniques has emerged as a

powerful tool to characterize cellular volume, morphology, and connectivity in a range of
complex tissues, including brain and heart [15–18]. Moreover, optical clearing by refractive
NIH-PA Author Manuscript

index matching using BABB or glycerol-based solutions in conjunction with second

harmonic generation has imaged tendon [19], skin [20], muscle [21], cartilage and bone [22]
at the tissue scale. However, the previous studies utilized reagents that precluded the ability
to identify and resolve individual cells within the musculoskeletal system due to significant
specimen shrinkage (BABB) or disruption of extracellular matrix morphology (glycerol). In
this study, we explore the possibility that optical clearing of dense connective tissues will
reveal unique microstructures and cell characteristics that are not easily observed or
quantified using standard microscopy or histomorphometry techniques. Our primary
objective is to determine the effect of morphology-preserving and fructose-based optical
clearing agents [15] in situ on collagen-, proteoglycan-, and mineral-rich osteochondral
tissues. Secondarily, we characterize the connectivity of chondrocytes throughout cartilage
and near the complex osteochondral interface within tissues in situ. Multiple assays,
including mechanical testing and scanning electron microscopy, are used to verify the
reversibility and morphology-preserving characteristics of optical clearing for osteochondral
tissues.
NIH-PA Author Manuscript

MATERIALS AND METHODS

Tissue Acquisition
Osteochondral tissues (5–7 month-old) were harvested from bovine knee (stifle) joints
obtained from a local abattoir (Dutch Valley Foods, Holland, IL) within 48 hours of
slaughter. Cylindrical samples (ø = 5.0 mm, 5.0 mm height) were excised from medial
femoral condyle regions and trochlear groove that naturally experience relatively high and
low in vivo contact pressures (i.e. load-bearing (LB) and non-load-bearing (NLB) regions,
respectively; Figure 1A) [23]. Samples were immediately placed in 4% paraformaldehyde
(PFA) in phosphate buffered saline (PBS) or PBS alone for subsequent comparative
analyses. Samples in PFA or PBS were placed on a rocker for 1–2 days at 4°C, then rinsed
2× 1 hour with PBS at room temperature with gentle rocking

Optical Clearing
NIH-PA Author Manuscript

Osteochondral tissues were cleared using a fructose-based optical clearing agent [15]
(Figure 1D; sample sizes for each assay defined subsequently). Clearing solutions were
prepared using D-(−)-fructose dissolved in ultrapure (milliQ) water of increasing
concentrations (i.e., 20%, 40%, 60%, 80%, 100% and 115% wt/vol), with 0.5% α-
thioglycerol added to prevent browning. Tissues were placed in ~1 mL of each concentrated
fructose solution overnight at room temperature and under a gentle rocking motion.
Unclearing of osteochondral tissues was achieved by reversing the clearing process, through
daily equilibration in tissues in fructose solutions of decreasing concentrations, to PBS.

Transmission and Confocal Microscopy

Cartilage and chondrocyte morphology were imaged by light microscopy. Light
transmission was assessed through the bulk (control, cleared, and uncleared) tissue (n=3)

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 4

using a stereo dissecting microscope (Leica M80, Wetzlar Germany). Tissues were placed
on a printed grid pattern (with grid spacing=2.1 mm), and images were acquired with front
and back lighting to visualize both the tissue and grid pattern.
NIH-PA Author Manuscript

Chondrocytes in cleared tissues were labeled and imaged by confocal microscopy to

visualize morphology (n=3) and quantify volumes. To image chondrocytes deep within
cleared cartilage, cell membranes (DiI, Life Technologies, Grand Island, NY) and nuclei
(Hoechst, Life Technologies) were labeled and visualized on a confocal microscope (Zeiss
LSM 710) equipped with a 25× LD LCI Plan-Apochromat multi-immersion lens (NA=0.8,
working distance≈500μm, immersion oil n=1.518). Imaging parameters were: excitation
wavelengths= 561 (DiI) and 405 (Hoechst 34580) nm; field of view= 340.1×340.1 μm2;
matrix= 512 × 512 pixels2; number of slices=479; interslice spacing=1.0 μm. To determine
the influence of fructose-based clearing on chondrocyte volume, cell membranes (DiI) were
labeled in separate samples and imaged following control, cleared, and uncleared treatments.
Confocal imaging parameters were identical to those described previously. Cell volumes
(n=10) were computed using image thresholding and region filling routines written in
MATLAB (Mathworks, Natick, MA).
NIH-PA Author Manuscript

Cartilage and chondrocyte morphology was further visualized with selective chemical labels
and by immunofluorescence. Tissue discs (n=3) were cut in hemicylinders, allowing for an
increased number of labels per sample, with four available excitation laser wavelengths
(405, 488, 555, 633 nm) on a confocal microscope system (Nikon A1Rsi, Melville, NY),
and a 20× multi-immersion objective (Plan Fluor, NA=0.75, working distance≈300μm).
Staining reagents were diluted in 0.2% bovine serum albumin in PBS and obtained from
Life Technologies unless otherwise noted. Samples were stained with primary antibodies for
72hrs at 4°C with gentle rocking, rinsed with PBS 3 × 30 min at room temperature, stained
with fluorescently-labeled secondary probes for 72hrs at 4°C with gentle rocking, rinsed
again with PBS 3 × 30 min at room temperature before gradual equilibration in concentrated
fructose. For half of the samples, primary labels were superficial zone protein (aka SZP/
lubricin/PRG4; 1:500; mAb S6.79 [24]; 555 nm) and fibrillin-2 (1:200; kindly provided by
Prof. R. Mecham, Washington University, St. Louis; 633 nm) and counterstained with
Hoechst 34580 (1:500; to label DNA), phalloidin (1:200; to label actin), donkey anti-mouse
AF555 (1:500; to label SZP) and donkey anti-rabbit AF647 (1:500; to label fibrillin-2). For
the other half of the samples, primary labels were perlecan (1:100; Santa Cruz
NIH-PA Author Manuscript

Biotechnology, Dallas, TX, sc33707), type II collagen (1:200; II-II6B3, Developmental

Studies Hybridoma Bank, Iowa City, IA), and type VI collagen (1:200; ab7821, Millipore,
Billerica, MA) and counterstained with Hoechst 34580, donkey anti-rat AF488 (1:500; to
label perlecan), donkey anti-mouse AF555 (to label type II collagen) and donkey anti-rabbit
AF647 (to label type VI collagen). After equilibration in 115% fructose solution, images
were acquired at a depth of 200μm from the imaging (cut) surface.

Mechanical Testing
Mechanical testing was performed in unconfined compression on cartilage samples to
determine the effect of both paraformaldehyde (PFA) fixing and optical clearing on material
properties. Specifically, cartilage samples (n=6, ø = 5.0 mm, 1.6 mm height) were assigned

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 5

to treatment groups in a two-factor design, with fixation (+/− paraformaldehyde) as one

factor, and optical clearing (control, cleared, or uncleared) as a second factor. In addition, to
isolate the effects of time-dependent (e.g. proteolytic) or clearing-related changes in
NIH-PA Author Manuscript

cartilage over the time duration required for chemical exchanges (Figure 1D), a separate
(control) experiment was performed with samples (n=3) incubated under identical
conditions as optically cleared samples, but without undergoing the clearing process.
Unfixed samples were supplemented with 0.02% sodium azide and protease inhibitor
cocktail (Sigma, St. Louis MO) to prevent proteolytic degradation. Cartilage explants were
tested in unconfined compression using a servoelectric materials testing system
(TestResources, Shakopee, MN). Samples were compressed in 5% nominal strain
increments at a physiological strain of 17.9% s−1 coupled with a 10-minute hold (stress
relaxation) [2]. All tests were performed in a 1× PBS bath at 25°C. Relaxation period was
set to ensure that stress equilibrium was achieved (no significant force slope). The solid
platen was lowered until a prestress of 0.3 N was achieved, ensuring full contact with the
sample. Engineering stress was determined using initial sample geometry and a 50-lbf load
cell (LPM 512, Cooper Instruments, Warrenton VA) at a sampling rate of 100 Hz.
Instantaneous modulus was calculated based on the largest (compressive) stress experienced
during each strain step, from which equilibrium modulus was calculated as an average of
NIH-PA Author Manuscript

100 data points at the end of each relaxation period, as previously described [25]. Data was
analyzed using a custom MATLAB algorithm.

Ultrastructure analysis via cryo-SEM

Cartilage was visualized using cryo-scanning electron microscopy (cryo-SEM, Nova
NanoSEM, FEI, Hillsboro OR) to qualitatively analyze the tissue ultrastructure and fibril
organization throughout the clearing process. To examine the effects of the clearing process
on ultrastructure, both control and uncleared samples were analyzed using fixed and unfixed
samples (n=3). Briefly, samples were flash frozen in critical point liquid nitrogen slush.
Samples were then fractured at the imaging plane and sublimated for 18 minutes. Prior to
imaging, samples were platinum sputter coated in the presence of argon for 120 seconds. All
samples were imaged under vacuum in the middle zone. Cleared samples were unable to be
imaged due to the inability to sublimate fructose for image clarity. Images were
quantitatively analyzed for porosity to examine changes in gross morphology. Briefly,
images were thresholded and binarized, and cumulative pixels of fibers and pores were
NIH-PA Author Manuscript

utilized to calculate porosity via pore area fraction.

Statistics
All data were analyzed using Minitab v.16 Statistical Software (Minitab Inc., State College
PA). Cell volumes were compared using a repeated-measures analysis of variance
(ANOVA). Instantaneous and equilibrium modulus were compared using a two-factor
ANOVA, with fixation and optical clearing as fixed main effects, with post-hoc tests to
compare differences between cleared and uncleared samples. Equilibrium modulus was not
found to have a normal distribution, and was transformed to normality via a log transform
prior to ANOVA analysis. Mechanical properties of samples that did not undergo the
clearing process were compared to uncleared samples via one-way ANOVA, with fixation
as the main effect. Porosity was compared using a nested ANOVA, with fixation and optical

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 6

clearing as fixed main effects, sample as a random effect, and the image locations nested
within samples. The residuals of each model were inspected and met the assumptions of
ANOVA (i.e. homogeneity of variances and normality) or were transformed as described
NIH-PA Author Manuscript

such that they did meet the assumptions of ANOVA. All data is presented as mean ±
standard deviation.

RESULTS
Transmission and Confocal Microscopy Enable Deep Microscopy in Cartilage
Optical clearing reveals light transmission through cartilage, but not blood-rich subchondral
bone regions (Figure 2). In 2.5 mm thick samples from both load-bearing and non-load-
bearing joint regions, cartilage became translucent following graded immersion in
concentrated fructose solutions, revealing printed grid patterns on an underlying substrate.
Translucent tissues regained normal (control) opacity following reversible graded immersion
in PBS.

Fluorescent labeling was preserved through sample preparations, enabling standard confocal
imaging to depths approaching 500 μm (Figure 3). The outlines of individual chondrocytes
NIH-PA Author Manuscript

(DiI, red) and their nuclei (Hoechst, blue) could be imaged through the superficial zone (z=1
μm) to deep within the cartilage middle zone (z=440 μm), which was limited by the working
distance of the microscope objective employed. Moreover, there was no significant change
in average cell volume (n=10) in control (399.6±218.9 mm3), cleared (390.6±218.4 mm3),
and uncleared samples (402.4±216.8 mm3) (p>0.050) (Figure 4). Normalized cell volume on
average was slightly lower in cleared samples, though not significant, and qualitatively
demonstrated improved definition of membrane morphology (Figure 4A).

Multiple fluorescent labels in cartilage were imaged by confocal microscopy. Nuclei

(Hoechst, blue) and actin filaments (phalloidin, green) were observed through the full
thickness of cartilage, and showed characteristic morphology of zone-specific chondrocytes.
Immunolabeling of perlecan was observed through the full thickness of the tissue, and
superficial zone protein and fibrillin-2 were observed at the articular surface, but
immunolabeling for types II and VI collagen could not be observed at 200 μm (Figure 5).
Additionally, comparison of actin networks within chondrocytes suggested more cellular
connectivity and protrusions in samples from load-bearing regions of lateral condyles
NIH-PA Author Manuscript

compared to non-load-bearing regions (Figure 6).

Mechanical Properties are Recovered Following Unclearing

Cartilage instantaneous modulus depended on optical clearing (p=0.013), but not fixation
(p=0.855) or their interaction (p=0.831) (Figure 7). Instantaneous moduli for unfixed/fixed
samples were 2.6±0.9/2.5±0.8 MPa (control), 3.3±1.2/3.5±1.4 MPa (cleared), and
2.3±0.3/2.1±0.8 MPa (uncleared). Similarly, equilibrium modulus was also dependent on
clearing (p<0.001), but not fixation (p=0.849) or their interaction (p=0.908). Equilibrium
moduli for unfixed/fixed samples were 0.4±0.2/0.5±0.2 MPa (control), 0.8±0.5/0.9±0.9 MPa
(cleared), and 0.2±0.1/0.2±0.1 MPa (uncleared). Post-hoc analysis for instantaneous
modulus revealed significant differences between cleared/control samples and uncleared

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 7

samples (p<0.050). All other comparisons for instantaneous modulus were not significantly
different. Post-hoc analysis of equilibrium moduli showed that cleared samples (fixed and
unfixed) were significantly higher than uncleared-fixed samples. When the time-dependent
NIH-PA Author Manuscript

or clearing-related changes in cartilage over the time duration required for chemical
exchanges was assessed, instantaneous and equilibrium moduli unfixed/fixed samples were
found to be 2.6±0.3/2.3±0.8 MPa (control) and 0.2±0.0/0.2±0.1 MPa (uncleared),
respectively, with no statistical difference for any main effects or interactions (all p>0.300).

SEM Reveals No Alteration in Microstructure Due to Clearing Process

Microstructure is qualitatively maintained throughout the clearing process (Figure 8). In
control and uncleared samples, ultrastructure is not discernibly different at 20,000×
magnification for all samples (Figure 8B). At higher magnification (40,000×), there is a
noticeable reduction in pore size for samples fixed with PFA. However, there is no apparent
difference between control and uncleared pairs for fixed and unfixed (Figure 8B). Statistical
analysis revealed no difference in porosity (pore area fraction) due to fixation (p=0.921),
optical clearing (p=0.776), or the interaction (p=0.947). Additionally, multiple images
acquired within samples were found to contribute insignificant variation to the results
(p=0.524).
NIH-PA Author Manuscript

DISCUSSION
The purpose of this study was to investigate the effect of fructose-based optical clearing
agents on collagen-, proteoglycan-, and mineral-rich osteochondral tissues, and to
characterize patterns of cellular networks within tissues in situ. We found that optical
clearing permits light transmission through 2.5 mm-thick cartilage tissue, but not bone
regions. Fluorescent imaging was performed using standard (one-photon) confocal imaging
with penetration depths limited only by the working distance of the microscope objective.
Tissue morphology remained unchanged and material properties returned to control levels in
uncleared tissues, indicating that the structural integrity was conserved during the clearing
process

Optical clearing reveals light transmission through cartilage, but not blood-rich subchondral
bone regions. Sample preparations did not alter fluorescent labeling, which enabled imaging
to cartilage depths limited by the working distance of our microscope objectives. Hoechst,
NIH-PA Author Manuscript

DiI and phalloidin were capable of penetrating deep into the cartilage matrix prior to
fixation (Figures 3–6), which may be related to their small molecular weight. In contrast,
some (perlecan), but not all (types II and VI collagen) primary antibodies were able to
penetrate the dense matrix and bind to deep tissue structures. Therefore, care should be
taken to evaluate antibodies and labeling systems for intended visualization of molecular
targets. A distinct disadvantage of current clearing techniques is that they do not permit
visualization of viable cells, which will impact the design of experiments for time-course
studies of cellular dynamics.

Cellular connectivity was observed in some load-bearing cartilage tissues. Cartilage is often
viewed as a matrix-rich tissue containing only a sparse distribution of chondrocytes that are
often separated by both a local pericellular and bulk interstitial matrix. Our data supports

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 8

previous reports of chondrocyte connectivity revealed by visualization of cytoskeletal

connections between cells [26]. However, localization of chondrocyte connections to load-
bearing tissue regions (Figure 6) suggests the interesting possibility that cellular connectivity
NIH-PA Author Manuscript

(e.g., cadherins, connexins) may be involved in the transduction of mechanical force, which
would compliment existing proposed mechanisms based on TGF-β signaling for SZP/
lubricin/PRG4 in the superficial zone of load-bearing cartilage [23]. Additionally, 3D
cellular morphology changes in the progression of osteoarthritis, which suggests that
connectivity may be enhanced during, or involved as a response to, cartilage disease [9, 27–
30]. We expect that the role of cellular connectivity and associated mechanisms may be
further elucidated through application of optical clearing in normal and diseased tissues
from a variety of human, bovine, murine, and related species.

Throughout the clearing process, instantaneous and equilibrium moduli remained within
typical cartilage mechanics for all samples [25, 31]. Both instantaneous and equilibrium
moduli increased when cleared. While there was a significant decrease in the moduli for
uncleared samples, comparison to control samples (that were incubated without undergoing
the clearing process) revealed similar trends, suggesting that the effect was due to time-
dependent changes (e.g. proteolytic breakdown) of the matrix, and not the clearing process.
NIH-PA Author Manuscript

At this stage, the mechanistic basis of increased mechanical properties during cartilage
clearing is unclear. Compressive properties of articular cartilage are largely dependent on
porous resistance to interstitial fluid flow [32, 33]. We believe the increased viscosity of the
fructose solution in the interstitium greatly contributes to the increase of mechanical
properties seen in the results. However, other interesting mechanical observations have been
observed in optical clearing using sugars and sugar alcohols. Specifically, hydrogen bond
dissociation, and related loss of high order fibril structure has been shown to occur with
optical clearing utilizing these agents [34, 35]. It has been postulated that the loss of non-
covalent bonds due to this dissociation causes the immediate engagement of stiffer, covalent
bonds such as those found in collagen crosslinks, resulting in higher tissue stiffness [35].
Within the scope of this work, the independent or interactive effects of these mechanisms
are not able to be uncoupled, but are of significant importance to be studied further.
However, our results, showing no significant change in cell morphology, may suggest
minimal microstructural changes as a result of optical clearing and may further suggest the
bulk of the mechanical differences are a result of interstitial fluid property changes.
NIH-PA Author Manuscript

Considering all of these possibilities, these results still show that the clearing process can be
performed, and combined with a multitude of labeling techniques as shown herein, while
maintaining little fluctuation in the mechanics of the environment.

Clearing process does not impact tissue structural morphology or ultrastructure as assessed
by cryo-SEM. This, along with other data presented herein, indicate that the clearing process
produces no apparent alteration to the overall structure of the tissue. These results are
consistent with previous literature discussing the reversibility of sugar and sugar alcohol
based clearing while recovering original morphology [34, 35]. Previous research has
asserted that the replacement of interstitial fluid with sugar or sugar alcohols disrupts
hydrogen bonding in high order collagen structures, lending to the optical clearing effect
[35]. Unfortunately, we were unable to assess cartilage morphology in the cleared state to
support or refute these results. All samples were assessed via cryo-SEM, which requires an

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 9

aqueous interstitium such that critical point drying can be used to reveal tissue morphology.
Due to the inability of fructose to sublimate under similar conditions, cryo-SEM was not
suitable for analysis. When considered alongside the lack of mechanical difference in
NIH-PA Author Manuscript

control and uncleared cartilage as well as associated fluorescence data, it is apparent that
minimal structural and morphological alterations occur to the tissue as a result of the
clearing process. These results suggest that enhanced light transmission is primarily a result
of solvent exchange.

Microscopy of diverse collagen- and proteoglycan-rich tissues will enable the study of
intact, 3D interfaces in situ in cartilage and related (e.g., ligamentous) connective tissues.
Moreover, we anticipate widespread use of optical clearing for the investigation of cellular
connectivity and tissue structure as a function of development, disease, and regeneration,
which have been previously hindered by specimen opacity.

Acknowledgments
The authors are grateful for technical assistance provided by Aaron Taylor, Ph.D., in the Multi-Scale Imaging
Center at Purdue University, and for use of the SZP and fibrillin-2 antibodies, provided by Thomas Schmid, Ph.D,
and Robert Mecham, Ph.D, respectively. This research was supported in part by NIH grants R01 AR063712, R21
AR064178, R21 AR066230 (CPN), R03 AR065201 (SC), NSF CAREER 1349735 (CPN), and an NSF Graduate
NIH-PA Author Manuscript

Research Fellowship (TN).

References
1. Mow VC, Ratcliffe A, Poole AR. Cartilage and diarthrodial joints as paradigms for hierarchical
materials and structures. Biomaterials. 1992; 13:67–97. [PubMed: 1550898]
2. Neu CP, Komvopoulos K, Reddi AH. The interface of functional biotribology and regenerative
medicine in synovial joints. Tissue Eng Part B Rev. 2008; 14:235–247. [PubMed: 18601586]
3. Hedbom E, Antonsson P, Hjerpe A, Aeschlimann D, Paulsson M, Rosa-Pimentel E, et al. Cartilage
matrix proteins. An acidic oligomeric protein (COMP) detected only in cartilage. The Journal of
biological chemistry. 1992; 267:6132–6136. [PubMed: 1556121]
4. Schmid TM, Linsenmayer TF. Immunoelectron microscopy of type X collagen: supramolecular
forms within embryonic chick cartilage. Developmental biology. 1990; 138:53–62. [PubMed:
2307289]
5. Gannon JM, Walker G, Fischer M, Carpenter R, Thompson RC Jr, Oegema TR Jr. Localization of
type X collagen in canine growth plate and adult canine articular cartilage. Journal of orthopaedic
research: official publication of the Orthopaedic Research Society. 1991; 9:485–494. [PubMed:
2045975]
NIH-PA Author Manuscript

6. Lawrence RC, Helmick CG, Arnett FC, Deyo RA, Felson DT, Giannini EH, et al. Estimates of the
prevalence of arthritis and selected musculoskeletal disorders in the United States. Arthritis and
rheumatism. 1998; 41:778–799. [PubMed: 9588729]
7. Hunziker EB. The elusive path to cartilage regeneration. Advanced materials. 2009; 21:3419–3424.
[PubMed: 20882507]
8. Hambach L, Neureiter D, Zeiler G, Kirchner T, Aigner T. Severe disturbance of the distribution and
expression of type VI collagen chains in osteoarthritic articular cartilage. Arthritis and rheumatism.
1998; 41:986–996. [PubMed: 9627008]
9. Bush PG, Hall AC. The volume and morphology of chondrocytes within non-degenerate and
degenerate human articular cartilage. Osteoarthritis and cartilage/OARS, Osteoarthritis Research
Society. 2003; 11:242–251.
10. Niesner R, Andresen V, Neumann J, Spiecker H, Gunzer M. The power of single and multibeam
two-photon microscopy for high-resolution and high-speed deep tissue and intravital imaging.
Biophysical journal. 2007; 93:2519–2529. [PubMed: 17557785]

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 10

11. Wong BJ, Wallace V, Coleno M, Benton HP, Tromberg BJ. Two-photon excitation laser scanning
microscopy of human, porcine, and rabbit nasal septal cartilage. Tissue engineering. 2001; 7:599–
606. [PubMed: 11694193]
NIH-PA Author Manuscript

12. Bush PG, Wokosin DL, Hall AC. Two-versus one photon excitation laser scanning microscopy:
critical importance of excitation wavelength. Frontiers in bioscience: a journal and virtual library.
2007; 12:2646–2657. [PubMed: 17127269]
13. Werkmeister E, de Isla N, Marchal L, Stoltz JF, Dumas D. Interest of second harmonic generation
imaging for diagnosis in thick and opaque tissue. Biorheology. 2008; 45:375–383. [PubMed:
18836238]
14. Yeh AT, Hammer-Wilson MJ, Van Sickle DC, Benton HP, Zoumi A, Tromberg BJ, et al.
Nonlinear optical microscopy of articular cartilage. Osteoarthritis and cartilage/OARS,
Osteoarthritis Research Society. 2005; 13:345–352.
15. Ke MT, Fujimoto S, Imai T. SeeDB: a simple and morphology-preserving optical clearing agent
for neuronal circuit reconstruction. Nature neuroscience. 2013; 16:1154–1161.
16. Chung K, Wallace J, Kim SY, Kalyanasundaram S, Andalman AS, Davidson TJ, et al. Structural
and molecular interrogation of intact biological systems. Nature. 2013; 497:332–337. [PubMed:
23575631]
17. Erturk A, Becker K, Jahrling N, Mauch CP, Hojer CD, Egen JG, et al. Three-dimensional imaging
of solvent-cleared organs using 3DISCO. Nature protocols. 2012; 7:1983–1995.
18. Goergen CJ, Radhakrishnan H, Sakadzic S, Mandeville ET, Lo EH, Sosnovik DE, et al. Optical
coherence tractography using intrinsic contrast. Optics letters. 2012; 37:3882–3884. [PubMed:
NIH-PA Author Manuscript

23041891]
19. Hovhannisyan V, Hu PS, Chen SJ, Kim CS, Dong CY. Elucidation of the mechanisms of optical
clearing in collagen tissue with multiphoton imaging. Journal of biomedical optics. 2013;
18:046004. [PubMed: 23552636]
20. Wen X, Mao Z, Han Z, Tuchin VV, Zhu D. In vivo skin optical clearing by glycerol solutions:
mechanism. Journal of biophotonics. 2010; 3:44–52. [PubMed: 19937846]
21. Plotnikov S, Juneja V, Isaacson AB, Mohler WA, Campagnola PJ. Optical clearing for improved
contrast in second harmonic generation imaging of skeletal muscle. Biophysical journal. 2006;
90:328–339. [PubMed: 16214853]
22. Vesuna S, Torres R, Levene MJ. Multiphoton fluorescence, second harmonic generation, and
fluorescence lifetime imaging of whole cleared mouse organs. Journal of biomedical optics. 2011;
16:106009. [PubMed: 22029356]
23. Neu CP, Khalafi A, Komvopoulos K, Schmid TM, Reddi AH. Mechanotransduction of bovine
articular cartilage superficial zone protein by transforming growth factor beta signaling. Arthritis
Rheum. 2007; 56:3706–3714. [PubMed: 17968924]
24. Schumacher BL, Block JA, Schmid TM, Aydelotte MB, Kuettner KE. A novel proteoglycan
synthesized and secreted by chondrocytes of the superficial zone of articular cartilage. Archives of
biochemistry and biophysics. 1994; 311:144–152. [PubMed: 8185311]
25. Korhonen RK, Laasanen MS, Toyras J, Rieppo J, Hirvonen J, Helminen HJ, et al. Comparison of
NIH-PA Author Manuscript

the equilibrium response of articular cartilage in unconfined compression, confined compression

and indentation. Journal of biomechanics. 2002; 35:903–909. [PubMed: 12052392]
26. Sasazaki Y, Seedhom BB, Shore R. Morphology of the bovine chondrocyte and of its cytoskeleton
in isolation and in situ: are chondrocytes ubiquitously paired through the entire layer of articular
cartilage? Rheumatology (Oxford). 2008; 47:1641–1646. [PubMed: 18796530]
27. Holloway I, Kayser M, Lee DA, Bader DL, Bentley G, Knight MM. Increased presence of cells
with multiple elongated processes in osteoarthritic femoral head cartilage. Osteoarthritis Cartilage.
2004; 12:17–24. [PubMed: 14697679]
28. Langelier E, Suetterlin R, Hoemann CD, Aebi U, Buschmann MD. The chondrocyte cytoskeleton
in mature articular cartilage: structure and distribution of actin, tubulin, and vimentin filaments. J
Histochem Cytochem. 2000; 48:1307–1320. [PubMed: 10990485]
29. Murray DH, Bush PG, Brenkel IJ, Hall AC. Abnormal human chondrocyte morphology is related
to increased levels of cell-associated IL-1beta and disruption to pericellular collagen type VI. J
Orthop Res. 2010; 28:1507–1514. [PubMed: 20872589]

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 11

30. Youn I, Choi JB, Cao L, Setton LA, Guilak F. Zonal variations in the three-dimensional
morphology of the chondron measured in situ using confocal microscopy. Osteoarthritis Cartilage.
2006; 14:889–897. [PubMed: 16626979]
NIH-PA Author Manuscript

31. Mauck RL, Wang CC, Oswald ES, Ateshian GA, Hung CT. The role of cell seeding density and
nutrient supply for articular cartilage tissue engineering with deformational loading. Osteoarthritis
Cartilage. 2003; 11:879–890. [PubMed: 14629964]
32. Levick JR. Flow through interstitium and other fibrous matrices. Q J Exp Physiol. 1987; 72:409–
437. [PubMed: 3321140]
33. Soltz MA, Ateshian GA. Interstitial fluid pressurization during confined compression cyclical
loading of articular cartilage. Ann Biomed Eng. 2000; 28:150–159. [PubMed: 10710186]
34. Hirshburg J, Choi B, Nelson JS, Yeh AT. Correlation between collagen solubility and skin optical
clearing using sugars. Lasers Surg Med. 2007; 39:140–144. [PubMed: 17311267]
35. Yeh AT, Hirshburg J. Molecular interactions of exogenous chemical agents with collagen--
implications for tissue optical clearing. J Biomed Opt. 2006; 11:014003. [PubMed: 16526880]
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 12
NIH-PA Author Manuscript

Figure 1. Tissue harvest locations and experimental overview

Osteochondral samples were harvested from the femoral condyles of bovine knee (stifle)
joints. (A) Samples were selected from the load-bearing and non-load-bearing regions of the
medial (MLB and MNLB, respectively) and lateral (LLB and LNLB, respectively) condyles.
For mechanical testing, samples were selected from the trochlear groove. Imaging of
samples was performed from either the articular surface through the depth of the samples (B;
Figures 3 and 4) or transverse to the depth direction (C; Figures 2, 5 and 6) using inverted
confocal microscopy (drawing not to scale). (D) Samples were subjected to equilibration in
NIH-PA Author Manuscript

concentrated fructose-based solutions for optical clearing [15] or PBS in a three-stage

reversible process.
NIH-PA Author Manuscript

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 13
NIH-PA Author Manuscript

Figure 2. Optical clearing is achieved in collagen- and proteoglycan-rich osteochondral tissues

Representative 2.5 mm thick samples from the medial load-bearing (MLB) and medial non-
load-bearing (MNLB) regions became translucent under transmission microscopy following
graded immersion in concentrated fructose solution, which was reversible following graded
NIH-PA Author Manuscript

immersion in PBS. Grid spacing = 2.1 mm.

NIH-PA Author Manuscript

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 14
NIH-PA Author Manuscript

Figure 3. Visualization of fluorescent markers observed through the cartilage depth are limited
by the working distance of the microscope objective
NIH-PA Author Manuscript

(A) Fluorescent labeling of cell membranes (DiI, red) and nuclei (Hoechst, blue) enabled the
visualization by standard confocal microscopy of chondrocytes deep into the middle zone
(MZ), limited by the working distance of the objective. Superficial zone chondrocyte
morphologies were visualized as flat and spread (at 1 μm), whereas MZ cells maintained a
more typical rounded phenotype. (B) Three-dimensional rendering of chondrocytes show
variations in cell patterning from the articular surface (top) to deep tissue regions. Scale bar
= 20 μm.
NIH-PA Author Manuscript

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 15
NIH-PA Author Manuscript

Figure 4. Cell volume does not change through the reversible clearing process
(A) Cell morphology of a representative cell stained with DiI remained qualitatively
unchanged following clearing and unclearing. (B) The volume of individual chondrocytes
was computed across the three conditions. There was no significant change in average cell
volume (n=10) in control (399.6±218.9 mm3), cleared (390.6±218.4 mm3), and uncleared
samples (402.4±216.8 mm3) (p>0.050). Cleared samples typically provided better definition
of membrane structure, with less scatter, as indicated by the improved definition of
membrane morphology and the slight reduction in normalized volume.
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 16
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Figure 5. Fluorescent labeling was observed in cartilage from the medial condyle by confocal
microscopy at a depth of 200 μm
(a) The characteristic morphology of chondrocytes at the superficial articular surface and
deep zones near the osteochondral interface was revealed by staining for nuclei (Hoechst,
blue) and actin filaments (phalloidin, green). Immunolabeling of superficial zone protein
(aka SZP/lubricin/PRG4; white arrows) and fibrillin-2 was observed at the articular surface,
although it is not clear whether penetration of the antibodies through the dense matrix was
achieved. (b) Immunolabeling of perlecan (green), showing increased volume of the
pericellular matrix compared to the actin-labeled cell volume alone (yellow and red arrows).
Immunolabeling for types II and VI collagen could not be observed at the imaging depth of
200 μm. Scale bars = 20 μm.
NIH-PA Author Manuscript

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 17
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Figure 6. Cellular connectivity was observed in the load-bearing regions of articular cartilage
following optical clearing
(A) Chondrocytes in the load-bearing regions of lateral condyles showed connectivity and
protrusions (white arrows) compared to non-load-bearing regions. Cartilage was stained for
SZP/lubricin/PRG4 (red), actin (green), and DNA (blue). Scale bars = 20 μm. (B) Three-
dimensional renderings of chondrocytes confirm cell connectivity. Bounding box
dimensions = 40×40×20 μm3 (l×w×h).
NIH-PA Author Manuscript

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 18
NIH-PA Author Manuscript

Figure 7. Bulk mechanics of cartilage was not changed after reversal of optical clearing
NIH-PA Author Manuscript

Cartilage instantaneous and equilibrium moduli depended on optical clearing (p<0.013), but
not fixation (p>0.855) or their interaction (p>0.831). Post-hoc analysis for instantaneous and
equilibrium modulus revealed significant differences between cleared samples and uncleared
samples (p<0.050).
NIH-PA Author Manuscript

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.

 Neu et al. Page 19
NIH-PA Author Manuscript

Figure 8. Optical clearing does not alter articular cartilage structural morphology
Representative SEM images of articular cartilage ultrastructure in control (pre-cleared) and
uncleared samples. Images at 20,000× (A) and 40,000× (B) reveal no qualitative or
discernible differences in fibril and pore structural morphology. Scale bar (A) 4 μm and (B)
2 μm.
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Osteoarthritis Cartilage. Author manuscript; available in PMC 2016 March 01.