FULL PAPER

MR Elastography for Evaluating Regeneration ofTissue-Engineered Cartilage in an Ectopic Mouse Model

Vahid Khalilzad-Sharghi,1 Zhongji Han,1 Huihui Xu,2 and Shadi F. Othman1*

Purpose: The purpose of the present study was to apply non-invasive methods for monitoring regeneration and mechanical

properties of tissue-engineered cartilage in vivo at differentgrowth stages using MR elastography (MRE).Methods: Three types of scaffolds, including silk, collagen,

and gelatin seeded by human mesenchymal stem cells, wereimplanted subcutaneously in mice and imaged at 9.4T where

the shear stiffness and transverse MR relaxation time (T2) weremeasured for the regenerating constructs for 8 wk. An MREphase contrast spin echo–based sequence was used for col-

lecting MRE images. At the conclusion of the in vivo study,constructs were excised and transcript levels of cartilage-specific genes were quantitated using reverse-transcription

polymerase chain reaction.Results: Tissue-engineered constructs showed a cartilage-like

construct with progressive tissue formation characterized byincrease in shear stiffness and decrease in T2 that can be cor-related with increased cartilage transcript levels including

aggrecan, type II collagen, and cartilage oligomeric matrix pro-tein after 8 wk of in vivo culture.

Conclusion: Altogether, the outcome of this research demon-strates the feasibility of MRE and MRI for noninvasive monitor-ing of engineered cartilage construct’s growth after

implantation and provides noninvasive biomarkers for regener-ation, which may be translated into treatment of tissue

defects. Magn Reson Med 000:000–000, 2015. VC 2015 WileyPeriodicals, Inc.

Key words: magnetic resonance elastography; cartilage tissueengineering

INTRODUCTION

Cartilage abnormalities such as degeneration of thejoint’s cartilage due to primary osteoarthritis, injury tothe articular cartilage, inflammation, and genetic disor-ders affect almost 30% of people in the United States(1). Cartilage is a form of connective tissue that consistsmainly of water, collagen, proteoglycans, and cells (2).Articular cartilage, specialized to provide a smooth sur-face for joints, is composed of 65%–75% water, 14%–

20% collagen, and 5%–15% proteoglycans (2). Repair ofcartilage defects remains a challenging problem due tothe biological features of cartilage tissue, such as limitedblood supply, lack of self-repair capacity, and defectiveimmobilization (3). Tissue engineering and stem celltherapy hold great potential for the restoration, repair,and replacement of damaged or lost cartilage tissue.Human mesenchymal stem cells (hMSCs) have beenused successfully to differentiate into a wide range ofcells, including osteoblasts (bone cells) (4), chondrocytes(cartilage cells) (5), and adipocytes (fat cells) (6), depend-ing on culture conditions. To regenerate functional tis-sues and organs, three-dimensional scaffolds have to bedesigned to prepare attachment sites and bioactive sig-nals for growth and differentiation of cells into a desiredlineage. Collagen (the main constituent of skin, cartilage,bone, and connective tissue) and gelatin, a derivative ofcollagen, have been widely employed in cartilage tissueengineering as scaffolds (7). Another scaffold that hasbeen used for encapsulating chondrocytes in cartilagetissue engineering is highly porous silk scaffold devel-oped by an aqueous process as described by Rockwoodet al. (8). Recent works have assessed the efficacy of dif-ferent scaffolds containing embedded chondrocytes orstem cells for cartilage repair. For example, Park et al. (9)reported cartilage regeneration using biodegradable oxi-dized alginate/hyaluronate hydrogels and observed effec-tive cartilage regeneration after 6 wk of transplantationbased on histological analysis, substantial secretion ofsulfated glycosaminoglycans, and expression of chondro-genic marker gene (collagen II) compared with nonde-gradable oxidized alginate/hyaluronate hydrogels.

There is a great need for the development of effectivemethods for monitoring the composition, structure, andfunction of tissue-engineered cartilage in vivo. Thesetechniques should be noninvasive and should providequantitative information at different growth stages,repair, and regeneration. Conventional characterizationof engineered constructs is performed with a variety oftools, including microscopy, biochemical assays, andhistologic stains. Although these methods provide criti-cal information on tissue’s regeneration, they aredestructive and result in the sacrifice of the tested sub-ject (10). This results in the need to test larger numbersof animals to compensate for data scattering to derivemeaningful conclusions. Therefore, noninvasive andnondestructive techniques that both analyze and quan-tify constructs in vitro and regenerated tissues in vivomust be investigated. The materials and specimens canbe evaluated longitudinally in the same animal by meansof the noninvasive techniques and analyzed by a sta-tistical method for repeated measurements. This reducesthe number of animals and the costs of

1Department of Biological Systems Engineering, University of Nebraska–Lincoln, Lincoln, Nebraska, USA.2School of Engineering and Computer Science, University of the Pacific,Stockton, California, USA.

*Correspondence to: Shadi F. Othman, Translational and RegenerativeMedicine Imaging Laboratory, 249 L.W. Chase Hall, East Campus, Univer-sity of Nebraska-Lincoln, Lincoln, NE 68583. E-mail: sothman2@unl.edu

Received 1 December 2014; revised 30 March 2015; accepted 31 March2015

DOI 10.1002/mrm.25745Published online 00 Month 2015 in Wiley Online Library (wileyonlinelibrary.com).

Magnetic Resonance in Medicine 00:00–00 (2015)

VC 2015 Wiley Periodicals, Inc. 1

experimentation and allows early intervention and modi-fication of experimental protocols to obtain superiorengineering outcomes. Biomedical imaging methodssuch as CT (11), ultrasound (12), and MRI (13,14) arebeing applied as noninvasive means of assessment.Among the imaging technologies, MRI has been shownto potentially play a significant role in tissue engineer-ing, providing nonionizing radiation, high spatial resolu-tion, high penetration depth, and multiple contrastmechanisms (15,16). Researchers have employed MRI-based techniques to correlate various MR-derived param-eters (eg, transverse MR relaxation time [T2], apparentdiffusion coefficient, shear stiffness, and magnetizationtransfer ratio) with mineral deposition and molecularcontent (16–19). One recently developed contrast mecha-nism, of importance to the tissue engineering commu-nity, employs a phase contrast–based technique termedMR elastography (MRE) for elasticity imaging (20). MRErequires three steps to generate quantitative stiffnessmaps. First, a mechanical actuator is coupled to the tis-sue of interest to generate propagating shear wavesthrough the tissue. Then, a pulse sequence, including amotion-encoding gradient (MEG), is used to encode themotion as phase in the MR image. Finally, an inversionalgorithm is employed to recover the shear stiffness map(or elastogram) from the spatio-temporal data. MRE hasbeen developed extensively to characterize the mechani-cal properties of biological soft tissue and also has beenused for preliminary research in tissue engineering (21).MRE was transferred to the microscopic scale usinghigh-field MRI and was employed to monitor tissue-engineered constructs regeneration in vitro and in vivo(22–24). MRE provides tissue engineers with a high-resolution tool to address an understudied area: ensuringthe engineered substitute exhibits the strength of theoriginal. Mechanical properties acquired using MRE inaddition to conventional MRI data, such as T2, apparent

diffusion coefficient, magnetization transfer ratio, and23Na MRI can be expected to play a leading role in

assessing engineered cartilage (25,26). These anatomical,

biochemical, and biomechanical features can be

extracted at different time points of cartilage regeneration

in preclinical models to generate sensitive biomarkers,

which can help steer preliminary clinical trials.In the present feasibility study, high-field MRE and

MRI were applied as noninvasive techniques to charac-terize the biomechanical and biological features oftissue-engineered cartilage after implantation in an 8-wkin vivo study. Three types of scaffolds, including silk,collagen, and gelatin seeded by hMSCs and treated withchondrogenic media, were cultured for 3 wk. As the firststep, the constructs were implanted subcutaneously intoan ectopic animal model. The mice were under study for8 wk and were imaged at four time points. MRE- and

MRI-derived parameters, including the MR transverse

relaxation time T2 and the shear stiffness (m), were meas-

ured. At the conclusion of the study, the expressions of

cartilage-specific genes were quantitated for the excised

constructs using reverse-transcription polymerase chain

reaction (RT-PCR) and compared with the results of con-

trol constructs cultured in vitro for 3 wk without

differentiation.

METHODS

Cell Culture and Preparation of Different Tissue-Engineered Cartilage Constructs Using Different Scaffolds

Healthy hMSCs isolated from fresh bone marrow cellswere provided commercially (Lonza, Walkersville, Mary-land, USA). Cells were cultured in a basic culturemedium composed of Dulbecco’s Modified Eagle’sMedium (Gibco, Carlsbad, California, USA) supple-mented with 10% fetal bovine serum (Gibco) and 1%antibiotics/antimycotics (Invitrogen, Carlsbad, California,USA). The chondrogenesis culture medium was createdby adding 10 nM dexamethasone, 40 mg/mL L-proline, 50mg/mL ascorbic acid-2-phosphate (Sigma, St. Louis, Mis-souri, USA), 10 ng/mL transforming growth factor b3(Peprotech, Rocky Hill, New Jersey, USA) and ITSTM pre-mix (containing 6.25 mg/mL insulin, 6.25 mg/mL transfer-rin, 5.33 mg/mL linoleic acid, and 1.25 mg/mL bovineserum albumin; BD Biosciences, Bedford, Massachusetts,USA) to the basic culture medium (27,28).

The engineering outcome will depend on the scaffoldmaterial and physical properties, including the pore sizeand the scaffold composition. Three different scaffoldswere selected for cartilage tissue engineering: gelatinsponges with a pore size of 250 mm (Pharmacia & Upjohn,Kalamazoo, Michigan, USA), fabricated protein silk with apore size of 500 mm (Department of Biomedical Engineer-ing, Tufts University, Medford, Massachusetts, USA), andcollagen constructs with a pore size of 350 mm (KenseyNash, Exton, Pennsylvania, USA). Constructs were biopsy-punched to a 5-mm diameter. Gelatin and collagen con-structs were seeded, in 50 mL Dulbecco’s Modified Eagle’sMedium, at 0.5 � 106 cells/scaffold with the assistance of avacuum using a 20-mL syringe (29,30). They then weremoved to 24-well plates after 2 h. For silk scaffolds, the pro-tocol presented by Rockwood et al. (31) was adopted.Briefly, silk scaffolds were placed in 24-well plates, andtheir pores evacuated with a Pasteur pipette; they werethen seeded at 0.5 � 106 cells/scaffold, which were deliv-ered in equal portions to the top and bottom surfaces. After15 min in the incubator, scaffolds were rotated 180�, and10 mL of cell-free medium was added to maintain hydra-tion. This process was repeated four times. The constructswere cultured for 3 wk before implantation.

In Vivo Construct Implantation

The proposed animal work was approved by the AnimalCare Committee at the University of Nebraska–Lincoln.Constructs were implanted subcutaneously in five 8-wk-old male nude immunodeficient mice (nu/nuJ; The Jack-son Laboratory, Bar Harbor, Maine, USA). For the implan-tation surgery, the mice were anesthetized with 5%isoflurane. The surgical site was disinfected with betadineand isopropyl alcohol. A 25-mm mid-sagittal incision wasmade across the dorsum in the prescapular region, wherea subcutaneous pocket was created on back of the midline.The constructs were implanted and fixed in location byplacing a suture through the muscle. The animals wereallowed to heal for 2 wk before removing the suture andconducting MRE/MRI measurements. The mice wereexamined for up to 8 wk following implantation.

2 Khalilzad-Sharghi et al.

MRI System and Measurement of MR Parameters

All MRI acquisitions were conducted at 9.4T (400 MHz forprotons) using an 89-mm vertical bore magnet equippedwith triple axis gradients (maximum strength, 100 G/cm)(Agilent, Santa Clara, California, USA). Measurementswere acquired using a 4-cm Millipede RF imagingprobe (Agilent, Santa Clara, California, USA) to transmitand receive the nuclear magnetic resonance signals.T2-weighted images were acquired from sagittal sectionsfor the silk construct, axial sections for the collagen con-struct, and coronal sections for the gelatin construct using aspin echo multislice sequence with the following parame-ters: repetition time (TR)/echo time (TE)¼ 1000/13.82 ms;number of averages¼2; matrix¼ 128 � 128 pixels; field ofview (FOV)¼ 14–20 mm2; slice thickness¼ 1 mm. T2 timeswere measured by applying multiple-echo spin echo imag-ing sequences to acquire 32 echoes (TR¼ 4000 ms; TE¼ 10ms; number of echoes¼ 32; number of averages¼ 2;FOV¼ 14–20 mm2; matrix¼ 128 � 128 pixels; slicethickness¼ 1 mm). T2 values were extracted from theexperimental data using a least-squares single exponentialfitting, and a low-pass filter was applied to improve thereadability of the maps (16).

MRE Acquisition

A modified phase contrast pulse sequence integratedwith the imaging software VnmrJ 3.1 (Agilent, SantaClara, California, USA) was used to acquire MRE data

(Agilent, Santa Clara, California, USA). The MRE inter-face allows the user to select the MRE parameters,including gradient amplitude (0–100 G/cm), actuator fre-quency, delay between the mechanical actuator and thebipolar gradient, MEG direction, and number of bipolarpairs. The design and setup of the MRE system havealready been demonstrated by Othman et al. (17). Anactuator was designed to provide sufficient displacementinto the cartilage construct as shown in Figure 1a. Thedesign was achieved through the use of a piezoelectricbending motor (Piezo System, Woburn, Massachusetts,USA) secured to a curvature around the mouse. A plasticcap attached to the other side of the piezoelectric motorwas coupled to the mouse’s body adjacent to the tissueconstruct. The actuator was driven by a signal generator(Tektronix, Beaverton, Oregon, USA) followed by anamplifier (Piezo System). Before placement in the scan-ner, the actuator was characterized while the mouse isplaced on the holder using a Laser Doppler Vibrometerand its resonance frequency providing the largestdynamic amplitude is identified as represented in Figure1b. The mice were placed in ventral recumbency onto acustom-designed animal holder and were maintainedanesthetized with 1.5%–2% isoflurane (Molecular Imag-ing Products Company, Bend, Oregon, USA) and moni-tored (Small Animal Instruments, Inc., Stony Brook,New York, USA). MRE acquisitions were performed withthe addition of external respiration gating to reducemotion artifacts.

FIG. 1. Experimental setup of in vivo MRE. (a) The actuator was designed to be placed adjacent to the tissue for maximum wave propa-gation into the construct. The arched suspension design permitted the actuator to be rotated over the curvature of the mouse’s body tooptimize positioning. The device was adapted for the operation of 400–1600 Hz piezoelectric actuators. (b) Actuator characterization.

The frequency response of the actuator was determined by sending a white noise to the system and performing Fourier transform. Acontinuous sinusoid signal at the resonance frequency was delivered to the sample to ensure optimal actuation.

MRE in Tissue-Engineered Cartilage 3

MRE scans were conducted using a spin-echo basedphase contrast sequence with the following parameters:TR¼ 1000 ms; TE¼ 19–21 ms; acquisition matrix¼ 128 �128; FOV¼ 14–20 mm2; mechanical frequency (f)¼ 750–800 Hz; number of gradient bipolar pulses¼ 2; and maxi-mum MEG¼95 G/cm. MRE acquisitions were performedwith six phase-offsets (between mechanical waveformand MEG) equally spaced across each cycle. The result-ant six wave images were processed using a MATLAB(MathWorks, Natick, Massachusetts, USA) code based ona previously developed spatiotemporal filtering approachto solve the inverse problem and calculate the shear stiff-ness of the constructs (32).

RNA Extraction and Complementary DNA Synthesis

Fresh constructs were transferred into 1.5-mL centrifugetubes. Samples were homogenized in 1.5 mL Trizol (LifeTechnologies, Grand Island, New York, USA) using mor-tar plus liquid nitrogen, and RNA was extracted accord-ing to the single-step acid-phenol-guanidinium method(33). The RNA samples were reverse-transcribed intocomplementary DNA using the QuantiTect Reverse Tran-scription Kit according to the manufacturer’s protocol(Qiagen, Hilden, Germany). A 300-ng total RNA samplewas used for the single-strand complementary DNA syn-thesis. The reverse-transcription reaction was incubatedat 42�C for 30 min and was terminated at 95�C for 3 min.

Real-Time RT-PCR

Aggrecan (Agg), type I collagen (collagen I), type II colla-gen (collagen II), type X collagen (collagen X), and carti-lage oligomeric matrix protein (COMP) transcript levelswere quantified using Fast SYBR Green Master Mix (LifeTechnologies) and the ABI Prism 7000 real-time PCR sys-tem (Applied Biosystems, Carlsbad, California, USA). Thetranscript data were normalized to the housekeeping gene,glyceraldehydes-3-phosphate-dehydrogenase (GAPDH).Reactions were performed in triplicate. Expression of tar-get genes was normalized to GAPDH and was expressed asthe fold ratio relative to the control group, using the2�DDCT method (34). Specific genes and primer sequencesare listed in Table 1.

Statistical Analysis

All stiffness and transverse relaxation values areexpressed as the mean 6 standard deviation over theentire region of interest. Statistical analysis for geneexpression data was performed by one-way analysis of

variance in conjunction with Tukey’s post hoc compari-sons for multiple comparisons, with P< 0.05 consideredsignificant.

RESULTS

The in vivo results indicate variations of imaging parame-ters during regeneration and differentiation of the chon-drogenic constructs toward cartilage. Figure 2demonstrates magnitude MRI images, T2 relaxation maps,displacement fields, and stiffness maps generated for thesilk scaffold at different time points (week 2, week 3,week 4, and week 8 after implantation). The average of T2

relaxation times calculated inside the regions of interestreduced from 91.2 6 7.6 ms at week 2 after implantationto 71.6 6 6.1 ms at week 4 and 67.6 6 3.1 ms at week 8.Average stiffness values increased from 7.6 6 2.0 kPa atweek 2 after implantation to 13.7 6 4.3 kPa at week 4 and17.2 6 3.1 at week 8. Measured values for the collagenconstruct are presented in Figure 3. Average T2 timesdecreased from 75.2 6 18.4 ms to 61.3 6 7.1 ms at week 4and 58.4 6 4.2 ms at week 8. Average stiffness values forthe collagen constructs increased from 4.6 6 1.7 kPa to12.0 6 3.7 kPa at week 4 and to 14.7 6 3.8 kPa at week 8.There are major limitations in estimating the stiffnessusing MRE. One major requirement for MRE is to ensurevisualizing half a wavelength within a homogeneousvoxel to construct are liable stiffness map which mightrequire increasing MRE resolution requires increasing theexcitation frequency (22,35).

The shear stiffness and T2 relaxation time dependedheavily on the type of implanted scaffolds. Figure 4ademonstrates a comparison of transverse MR relaxationtimes measured for the three implanted constructs (silk,collagen, and gelatin). At week 2, T2 times measured forthe silk, collagen, and gelatin constructs were 91.2 6 7.6ms, 87.4 6 9.2 ms, and 75.2 6 10.3 ms, respectively. Atthe end of the study, T2 decreased to 67.9 6 3.1 msfor the silk, 61.1 6 3.8 ms for the gelatin, and 58.4 6 4.2ms for the collagen constructs. The shear stiffness datafor all three constructs are provided in Figure 4b. Thegelatin construct showed the most shear stiffnessincrease, from 6.4 6 1.0 kPa to 22.6 6 4.1 kPa. Shear stiff-ness of the collagen construct increased moderately,from 4.6 6 1.7 kPa to 14.7 6 3.8 kPa from week 2 to week8. Shear stiffness of the silk construct increased less thanthe others, from 7.6 6 2.0 kPa to 17.2 6 3.1 kPa. It shouldbe mentioned, however, that the silk construct is theonly one that maintained its original anatomical shapeduring the entire study.

Table 1Amplified Genes and Primer Sequences

Primer Sequence

Gene Forward Reverse

Agg 50-AGGCAGCGTGATCCTTACC-30 50-GGCCTCTCCAGTCTCATTCTCTC-30

Collagen II 50-CGTCCAGATGACCTTCCTACG-30 50-TGAGCAGGGCCTTCTTGAG-30

Collagen I 50-CAGCCGCTTCACCTACAGC-30 50-TTTTGTATTCAATCACTGTCTTGCC-30

Collagen X 50-GCAACTAAGGGCCTCAATGG-30 50-CTCAGGCATGACTGCTTGAC-30

COMP 50-AGGGAGATCGTGCAGACA A-30 50-AGCTGGAGCTGTCCTGGTAG-30

GAPDH 50-TCCACTGGCGTCTTCACC-30 50-GGCAGAGATGATGACCCTTT-30

4 Khalilzad-Sharghi et al.

Transcript levels of cartilage-related extracellularmatrix genes were assessed using RT-PCR to determinechondrogenic differentiation after 3 wk of in vitro cul-ture and 8 wk of in vivo implantation of MSCs seededon silk, collagen, and gelatin scaffolds. As shown in Fig-ure 5, transcript levels of Agg in cells within the gelatin(Fig. 5a, 3.58 folder) and collagen (Fig. 5b, 42.50 folder)constructs of in vivo implantation were significantly up-regulated in comparison with in vitro culture and thecontrol (P< 0.001). The transcript levels of Agg in cellswithin the silk (Fig. 5c, 3.66 folder) constructs of in vivoimplantation were up-regulated in comparison with invitro culture and the control (P¼0.076). The expressionof collagen II in cells within the gelatin (Fig. 5a, 3.75folder), collagen (Fig. 5b, 50.21 folder), and silk (Fig. 5c,14.02 folder) constructs of in vivo implantation were sig-nificantly up-regulated in comparison with in vitro cul-ture and the control (P< 0.001). The expression ofCOMP in cells within the gelatin (Fig. 5a, 7.19 folder),collagen (Fig. 5b, 6.72 folder), and silk (Fig. 5c, 6.18folder) constructs of in vivo implantation were up-regulated in comparison with the control and weredown-regulated in comparison to in vitro culture. Theexpression of collagen I in cells within the gelatin (Fig.

5a, 2.66 folder), collagen (Fig. 5b, 14.77 folder) and silk(Fig. 5c, 5.81 folder) constructs of in vivo implantationwere up-regulated in comparison with the control. Theexpression of collagen X in cells within the gelatin (Fig.5a, 1670.90 folder), collagen (Fig. 5b, 2732.83 folder),and silk (Fig. 5c, 904.85 folder) constructs of in vivoimplantation were significantly up-regulated in compari-son with the control (P<0.001). A correlation mappingtechnique was used to analyze the pattern of collagen II,COMP, and Agg changes based on alterations in theshear stiffness and T2 relaxation time of cartilage con-structs; however, no statistically significant correlationwas observed.

DISCUSSION

The results of this study will provide tissue engineerswith a noninvasive rapid-feedback method for improvingthe tissue-engineered outcome. The developed methodallows sacrificing a smaller number of animals perexperiment. In a longitudinal experiment, each animal isstudied as its own control, thereby enabling repeatedevaluation of regenerated cartilage features and moremeasureable statistics. In this study, longitudinal

FIG. 2. Tissue-engineered cartilage silk–based construct evaluation over an 8-wk study at four time points. Shown from left to right are

the magnitude spin echo images, T2 relaxation maps, displacement fields, and stiffness maps of the construct.

MRE in Tissue-Engineered Cartilage 5

evaluation of hMSCs-derived cartilage constructs, seededinto three types of scaffolds (silk, collagen, and gelatin)and implanted subcutaneously in mice, were performedby measuring T2 and shear stiffness. We explored thepotential of stiffness mapping as well as T2 mapping at9.4T in characterizing cartilage regeneration followingimplantations in an ectopic model. This is the first stepbefore moving to clinically relevant defect models. Thecurrent results confirmed that the expression of collagenII, Agg, and COMP were up-regulated within silk, colla-gen, and gelatin scaffold after 3 weeks in vitro culture,compared with constructs cultured in vitro for 3 wkwithout differentiation (Fig. 5).

In cartilage tissue engineering, several factors must beinvestigated after implantation, such as localization,depth, and diameter of the constructs, the type of scaf-

folds, cells, and growth factors. To date, macroscopicand histological studies remain the gold standard forcharacterizing the type and quality of the repair tissue atdifferent time points, but are invasive in nature. There-fore, the development of noninvasive techniques is nec-essary to evaluate longitudinally the different tissueengineering techniques used for cartilage repair. Numer-ous studies have been performed to study effective imag-ing markers in tissue repair in various experimentalmodels of cartilage defects in humans (36–38) or animals(39,40). Chou et al. (41) temporally characterized the MRrelaxation time of a tri-copolymer sponge in vivo in arodent heterotopic model. They described that T2 of thesponge decreased significantly over time, whereas longi-tudinal relaxation time T1 remained stable comparedwith the control. Ballyns et al. (42) used MRI and CT to

FIG. 3. Tissue-engineered cartilage collagen–based construct evaluation over an 8-wk study at four time points. Shown from left to rightare the magnitude spin echo images, T2 relaxation maps, and stiffness maps of the constructs.

6 Khalilzad-Sharghi et al.

design custom-printed molds that enabled the generationof anatomically shaped cartilage constructs.

In the current study, we noted in all three types ofimplanted cartilage constructs a progressive and signifi-cant decrease of T2 over the 8-wk study. Average trans-verse MR relaxation time decreased 25% for the silk,30% for the gelatin, and 23% for the collagen-based con-

structs. Total differentiated tissue was characterized bythe smallest average T2 value and resultant larger signalloss compared with the baseline. This reduction can bebecause of decrease in the relative water content. Inaddition, formation of cartilage fiber is another dominantfactor for T2 shortening (43,44). We also observed stiffen-ing of the tissue constructs along with the differentiation

FIG. 4. Characterization of T2

relaxation time and shear stiff-ness as imaging markers for

three hMSC-seeded tissue-engi-neered cartilage constructs (silk,gelatin, and collagen). Compari-

son of the (a) T2 relaxation and(b) shear stiffness changes

among the constructs is shown.

FIG. 5. Transcript levels measured by RT-PCR for Agg, collagen I, collagen II, collagen X, and COMP of cartilage-related extracellular

matrix genes in vitro and in vivo within the (a) gelatin, (b) collagen, and (c) silk construct. Data are shown relative to the expression ofthe respective transcript by undifferentiated hMSCs at week 3 (control, CON) and represented as the mean 6 standard deviation.

*P<0.05; **P<0.01; ***P<0.001.

MRE in Tissue-Engineered Cartilage 7

and growth of the cells. MRE measurements revealedthat the shear stiffness increased nearly two-fold for thesilk, three-fold for the collagen, and three-and-a-half-foldfor the gelatin-based cartilage constructs. This increasein shear stiffness of the constructs can be due to forma-tion of their net-like organized structure of the collagentype I and II fibers and increasing concentration of pro-teoglycans (45). In agreement with previous reports(9,46,47), the current results demonstrated that theexpression of collagen II and Agg were up-regulatedwithin silk, collagen, and gelatin scaffold after 8 wkimplantation, compared with in vitro culture (Fig. 5).Agg is an integral part of the extracellular matrix and acritical component for cartilage structure. Collagen II isthe basis for articular cartilage and is a practicallyunique marker for chondrogenesis (48). The expressionof collagen I in cells within three constructs of in vivoimplantation were up-regulated compared with the con-trol. The expression of COMP was down-regulated andcollagen X was up-regulated compared with the in vitroculture. COMP is one of the major noncollagenous pro-teins, Frank et al. (49) reported COMP is down-regulatedfaster than collagen II at the late stage of chondrocytedifferentiation. Collagen X was seen as a marker forhypertrophic cartilage in previous studies; however, anincreasing number of observations of collagen X as acomponent of normal articular cartilage have beenreported (50–52). Mwale et al. (53) even reported colla-gen X was expressed before collagen II in some case. Ourresults suggest week 8 of implantation was at the termi-nal differentiation stage. Additional studies should com-pare histological outcome, PCR, and MRE to betterunderstand the complexity of regenerating engineeredcartilage.

Cartilage is affected in vivo and in vitro by several bio-mechanical forces (eg, direct compression and tensileand shear forces) or the generation of hydrostatic pres-sure and electric gradients, as well as changes in the pH(45). The dynamic processes that occur in cartilage arenecessary to maintain its structure and function andhave to be applied in the tissue engineering of cartilageas well. These effects and their correlation with imagingbiomarkers can be investigated in future studies. Theresults of this research demonstrated that MRI providesthe basis for morphological evaluation of engineered car-tilage tissues, whereas measuring transverse MR relaxa-tion provides deeper insight into the composition of thetissue. In addition, MRE allows observing changes inmechanical properties of the engineered tissue. Theimaging markers achieved in this study through noninva-sive biomedical imaging techniques can improve carti-lage repair and also can be used for the diagnosis ofsurrounding pathologies within defect sites. The methodintroduced in the present study can effectively reducethe number of animals needed for statistical significancein any further in vivo study without compromising reli-ability. This is very significant with regard to reducingthe cost of experiments, avoiding unnecessary animalsuffering, and minimizing experimental variationbetween animals. A combination of all of these factorsmay represent a desirable multimodal approach to followup after cartilage repair procedures and enable the

design of better tissue-engineered cartilage that can betranslated into treatment of tissue defects.

REFERENCES

1. Centers for Disease Control and Prevention. Prevalence and most

common causes of disability among adults—United States, 2005.

MMWR Morb Mortal Wkly Rep 2009;58:421–426.

2. Maroudas A. Biophysical chemistry of cartilaginous tissues with special

reference to solute and fluid transport. Biorheology 1975;12:233–248.

3. Trueta J. Blood supply and the rate of healing of tibial fractures. Clin

Orthop Relat Res 1974:11–26.

4. Bruder SP, Kurth AA, Shea M, Hayes WC, Jaiswal N, Kadiyala S.

Bone regeneration by implantation of purified, culture-expanded

human mesenchymal stem cells. J Orthop Res 1998;16:155–162.

5. Pelttari K, Winter A, Steck E, Goetzke K, Hennig T, Ochs BG, Aigner

T, Richter W. Premature induction of hypertrophy during in vitro

chondrogenesis of human mesenchymal stem cells correlates with

calcification and vascular invasion after ectopic transplantation in

SCID mice. Arthritis Rheum 2006;54:3254–3266.

6. Dicker A, Le Blanc K, Astrom G, van Harmelen V, Gotherstrom C,

Blomqvist L, Arner P, Ryden M. Functional studies of mesenchymal

stem cells derived from adult human adipose tissue. Exp Cell Res

2005;308:283–290.

7. Stone KR, Steadman JR, Rodkey WG, Li ST. Regeneration of meniscal

cartilage with use of a collagen scaffold. Analysis of preliminary

data. J Bone Joint Surg Am 1997;79:1770–1777.

8. Rockwood DN, Gil ES, Park SH, et al. Ingrowth of human mesenchy-

mal stem cells into porous silk particle reinforced silk composite

scaffolds: an in vitro study. Acta Biomater 2011;7:144–151.

9. Park H, Lee KY. Cartilage regeneration using biodegradable oxidized

alginate/hyaluronate hydrogels. J Biomed Mater Res A 2014;102:

4519–4525.

10. Sabokbar A, Millett PJ, Myer B, Rushton N. A rapid, quantitative

assay for measuring alkaline phosphatase activity in osteoblastic cells

in vitro. Bone Miner 1994;27:57–67.

11. Palmer AW, Guldberg RE, Levenston ME. Analysis of cartilage matrix

fixed charge density and three-dimensional morphology via contrast-

enhanced microcomputed tomography. Proc Natl Acad Sci U S A

2006;103:19255–19260.

12. Klein TJ, Chaudhry M, Bae WC, Sah RL. Depth-dependent biome-

chanical and biochemical properties of fetal, newborn, and tissue-

engineered articular cartilage. J Biomech 2007;40:182–190.

13. Roberts S, McCall IW, Darby AJ, Menage J, Evans H, Harrison PE,

Richardson JB. Autologous chondrocyte implantation for cartilage

repair: monitoring its success by magnetic resonance imaging and

histology. Arthritis Res Ther 2003;5:R60–R73.

14. Neu CP. Functional imaging in OA: role of imaging in the evaluation

of tissue biomechanics. Osteoarthritis Cartilage 2014;22:1349–1359.

15. Neu CP, Arastu HF, Curtiss S, Reddi AH. Characterization of engi-

neered tissue construct mechanical function by magnetic resonance

imaging. J Tissue Eng Regen Med 2009;3:477–485.

16. Xu H, Othman SF, Hong L, Peptan IA, Magin RL. Magnetic resonance

microscopy for monitoring osteogenesis in tissue-engineered con-

struct in vitro. Phys Med Biol 2006;51:719–732.

17. Othman SF, Xu H, Royston TJ, Magin RL. Microscopic magnetic reso-

nance elastography (microMRE). Magn Reson Med 2005;54:605–615.

18. Li W, Hong L, Hu L, Magin RL. Magnetization transfer imaging pro-

vides a quantitative measure of chondrogenic differentiation and tis-

sue development. Tissue Eng Part C Methods 2010;16:1407–1415.

19. Kotecha M, Schmid TM, Odintsov B, Magin RL. Reduction of water

diffusion coefficient with increased engineered cartilage matrix

growth observed using MRI. Conf Proc IEEE Eng Med Biol Soc 2014;

2014:3913–3916.

20. Muthupillai R, Lomas DJ, Rossman PJ, Greenleaf JF, Manduca A,

Ehman RL. Magnetic resonance elastography by direct visualization

of propagating acoustic strain waves. Science 1995;269:1854–1857.

21. Xu H, Othman SF, Magin RL. Monitoring tissue engineering using

magnetic resonance imaging. J Biosci Bioeng 2008;106:515–527.

22. Othman SF, Curtis ET, Plautz SA, Pannier AK, Butler SD, Xu H. MR

elastography monitoring of tissue-engineered constructs. NMR

Biomed 2012;25:452–463.

23. Curtis ET, Zhang S, Khalilzad-Sharghi V, Boulet T, Othman SF. Mag-

netic resonance elastography methodology for the evaluation of tissue

engineered construct growth. J Vis Exp 2012. doi: 10.3791/3618.

8 Khalilzad-Sharghi et al.

24. Othman SF, Curtis ET, Huihui X. In vivo magnetic resonance elastog-

raphy of mesenchymally derived constructs. IEEE 2011;1:621–624.

25. Kotecha M, Klatt D, Magin RL. Monitoring cartilage tissue engineer-

ing using magnetic resonance spectroscopy, imaging, and elastogra-

phy. Tissue Eng Part B Rev 2013;19:470–484.

26. Yin Z, Schmid TM, Yasar TK, Liu Y, Royston TJ, Magin RL. Mechan-

ical characterization of tissue-engineered cartilage using microscopic

magnetic resonance elastography. Tissue Eng Part C Methods 2014;

20:611–619.

27. Mackay AM, Beck SC, Murphy JM, Barry FP, Chichester CO,

Pittenger MF. Chondrogenic differentiation of cultured human mes-

enchymal stem cells from marrow. Tissue Eng 1998;4:415–428.

28. Wang Y, Blasioli DJ, Kim HJ, Kim HS, Kaplan DL. Cartilage tissue

engineering with silk scaffolds and human articular chondrocytes.

Biomaterials 2006;27:4434–4442.

29. Dennis JE, Haynesworth SE, Young RG, Caplan AI. Osteogenesis in

marrow-derived mesenchymal cell porous ceramic composites trans-

planted subcutaneously: effect of fibronectin and laminin on cell

retention and rate of osteogenic expression. Cell Transplant 1992;1:

23–32.

30. Murphy CM, Matsiko A, Haugh MG, Gleeson JP, O’Brien FJ. Mesen-

chymal stem cell fate is regulated by the composition and mechanical

properties of collagen-glycosaminoglycan scaffolds. J Mech Behav

Biomed Mater 2012;11:53–62.

31. Rockwood DN, Preda RC, Yucel T, Wang X, Lovett ML, Kaplan DL.

Materials fabrication from Bombyx mori silk fibroin. Nat Protoc 2011;

6:1612–1631.

32. Boulet T, Kelso ML, Othman SF. Microscopic magnetic resonance

elastography of traumatic brain injury model. J Neurosci Methods

2011;201:296–306.

33. Chomczynski P, Sacchi N. The single-step method of RNA isolation

by acid guanidinium thiocyanate-phenol-chloroform extraction:

twenty-something years on. Nat Protoc 2006;1:581–585.

34. Livak KJ, Schmittgen TD. Analysis of relative gene expression data

using real-time quantitative PCR and the 2(-Delta Delta C(T)) method.

Methods 2001;25:402–408.

35. Lopez O, Amrami KK, Manduca A, Ehman RL. Characterization of

the dynamic shear properties of hyaline cartilage using high-

frequency dynamic MR elastography. Magn Reson Med 2008;59:356–

364.

36. Link TM, Stahl R, Woertler K. Cartilage imaging: motivation, techni-

ques, current and future significance. Eur Radiol 2007;17:1135–1146.

37. Hunter DJ, Li J, LaValley M, et al. Cartilage markers and their associa-

tion with cartilage loss on magnetic resonance imaging in knee osteo-

arthritis: the Boston Osteoarthritis Knee Study. Arthritis Res Ther

2007;9:R108.

38. Gold GE, Burstein D, Dardzinski B, Lang P, Boada F, Mosher T. MRI

of articular cartilage in OA: novel pulse sequences and composi-

tional/functional markers. Osteoarthritis Cartilage 2006;(14 suppl A):

A76–A86.

39. Goebel JC, Pinzano A, Grenier D, Perrier AL, Henrionnet C, Galois L,

Gillet P, Beuf O. New trends in MRI of cartilage: advances and limita-

tions in small animal studies. Biomed Mater Eng 2010;20:189–194.

40. Goebel L, Orth P, Muller A, Zurakowski D, Bucker A, Cucchiarini M,

Pape D, Madry H. Experimental scoring systems for macroscopic

articular cartilage repair correlate with the MOCART score assessed

by a high-field MRI at 9.4 T—comparative evaluation of five macro-

scopic scoring systems in a large animal cartilage defect model.

Osteoarthritis Cartilage 2012;20:1046–1055.

41. Chou CH, Lee HS, Siow TY, Lin MH, Kumar A, Chang YC, Chang C,

Huang GS. Temporal MRI characterization of gelatin/hyaluronic acid/

chondroitin sulfate sponge for cartilage tissue engineering. J Biomed

Mater Res A 2013;101:2174–2180.

42. Ballyns JJ, Gleghorn JP, Niebrzydowski V, Rawlinson JJ, Potter HG,

Maher SA, Wright TM, Bonassar LJ. Image-guided tissue engineering

of anatomically shaped implants via MRI and micro-CT using injec-

tion molding. Tissue Eng Part A 2008;14:1195–1202.

43. Grunder W, Kanowski M, Wagner M, Werner A. Visualization of

pressure distribution within loaded joint cartilage by application of

angle-sensitive NMR microscopy. Magn Reson Med 2000;43:884–891.

44. Shinar H, Seo Y, Ikoma K, Kusaka Y, Eliav U, Navon G. Mapping the

fiber orientation in articular cartilage at rest and under pressure stud-

ied by 2H double quantum filtered MRI. Magn Reson Med 2002;48:

322–330.

45. Schulz RM, Bader A. Cartilage tissue engineering and bioreactor sys-

tems for the cultivation and stimulation of chondrocytes. Eur Bio-

phys J 2007;36:539–568.

46. Tang C, Xu Y, Jin C, Min BH, Li Z, Pei X, Wang L. Feasibility of auto-

logous bone marrow mesenchymal stem cell-derived extracellular

matrix scaffold for cartilage tissue engineering. Artif Organs 2013;37:

E179–E190.

47. Bichara DA, Pomerantseva I, Zhao X, et al. Successful creation of

tissue-engineered autologous auricular cartilage in an immunocompe-

tent large animal model. Tissue Eng Part A 2014;20:303–312.

48. Bosnakovski D, Mizuno M, Kim G, Takagi S, Okumura M, Fujinaga

T. Chondrogenic differentiation of bovine bone marrow mesenchymal

stem cells (MSCs) in different hydrogels: influence of collagen type II

extracellular matrix on MSC chondrogenesis. Biotechnol Bioeng

2006;93:1152–1163.

49. Frank Z, Robert D, Patrik AM, Mats. P. Cartilage oligomeric matrix

protein (COMP) and collagen IX are sensitive markers for the differ-

entiation state of articular primary chondrocytes. Biochem J 2001;

358:17–24.

50. Oettinger HF, Pacifici M. Type X collagen gene expression is transi-

ently up-regulated by retinoic acid treatment in chick chondrocyte

cultures. Exp Cell Res 1990;191:292–298.

51. Sanchez M, Gionti E, Pontarelli G, Arcella A, De Lorenzo F. Expres-

sion of type X collagen is transiently stimulated in redifferentiating

chondrocytes pretreated with retinoic acid. Biochem J 1991;276:183–

187.

52. Barry F, Boynton RE, Liu B, Murphy JM. Chondrogenic differentia-

tion of mesenchymal stem cells from bone marrow: differentiation-

dependent gene expression of matrix components. Exp Cell Res 2001;

268:189–200.

53. Mwale F, Stachura D, Roughley P, Antoniou J. Limitations of using

aggrecan and type X collagen as markers of chondrogenesis in mesen-

chymal stem cell differentiation. J Orthop Res 2006;24:1791–1798.

MRE in Tissue-Engineered Cartilage 9