The LG/J Murine Strain Exhibits Near-Normal Tendon Biomechanical Properties Following a Full-Length Central Patellar Tendon Defect

Abstract

Purpose of the study

Identifying biological success criteria is needed to improve therapies, and one strategy for identifying them is to analyze the RNA transcriptome for successful and unsuccessful models of tendon healing. We have characterized the MRL/MpJ murine strain and found improved mechanical outcomes following a central patellar tendon injury. In this study, we evaluate the healing of the LG/J murine strain, which comprises 75% of the MRL/MpJ background, to determine if the LG/J also exhibits improved biomechanical properties following injury and to determine differentially expressed transcription factors across the C57BL/6, MRL/MpJ and the LG/J strains during the early stages of healing.

Materials and Methods

A full-length, central patellar tendon (PT) defect was created in 16 to 20 week old MRL/MpJ, LG/J, and C57BL/6 murine strains. Mechanical properties were assessed at 2, 5, and 8 weeks post surgery. Transcriptomic expression was assessed at 3, 7, and 14 days following injury using a novel clustering software program to evaluate differential expression of transcription factors.

Results

Average LG/J structural properties improved to 96.7% and 97.2% of native LG/J PT stiffness and ultimate load by 8 weeks post surgery, respectively. We found the LG/J responded by increasing expression of transcription factors implicated in the inflammatory response and collagen fibril organization.

Conclusions

The LG/J strain returns to normal structural properties by 8 weeks, with steadily increasing properties at each time point. Future work will characterize the cell populations responding to injury and investigate the role of the differentially expressed transcription factors during healing.

INTRODUCTION

Musculoskeletal injuries are a frequent cause of health care visits in the United States, amounting to 99 million visits to ambulatory care facilities from 2009 to 2011¹. Tendon and ligament injuries account for 30% of musculoskeletal health care visits². Due to the aging population, tendon and ligament injuries will only become more prevalent, resulting in higher direct medical costs and lost productivity. Unfortunately, natural tendon healing fails to restore native mechanical and biological properties following injury^3–6. Our laboratory seeks to better understand tendon healing to create new treatment strategies with the goal of improving clinical outcomes.

Recently, we have evaluated the Murphy Roths Large (MRL/MpJ) murine strain’s response to a full-length, full-thickness central patellar tendon defect. The MRL/MpJ mice have been shown to have a superior healing capacity following injury in a number of tissue types including the heart⁷, skin⁸, cornea⁹, spinal cord¹⁰, and articular cartilage¹¹. We found that the biomechanical properties reached near normal levels by 8 weeks post injury¹². However, our gene expression analysis following tendon injury in the MRL/MpJ and C57BL/6 model system at early time points (3, 7, and 14 days following injury) revealed no significant differences between the two strains¹². Unfortunately, real-time PCR and microarrays are limited because these measures can only assess expression of known targets that have been previously identified. These technologies cannot evaluate expression of novel targets or other unidentified genetic changes. RNA-seq sequences and measures cDNA reads that are then mapped against a reference genome producing a quantitative transcriptional read-out of the entire genome for a given sample¹³. Further, RNA-seq is highly replicable and unique in that it provides absolute values as opposed to relative gene expression values, offering a more precise assessment of transcript level expression compared to real-time PCR and microarray experiments¹⁴.

In an effort to further our understanding of tendon healing, we are proposing to study the LG/J murine strain because they have shown improved healing of articular cartilage and ear-hole punches compared to wild type^15–17. In addition, the LG/J strain comprises 75% of the MRL/MpJ background (75% LG/J, 12.6% AKR, 12.1% C3H, and 0.3% C57/B6 strains)¹³ and may provide insight towards the molecular mechanisms regulating the improved healing phenotypes. We sought to determine the tendon healing properties of the LG/J mouse and to then compare the transcriptome of the LG/J, MRL/MpJ and C57BL/6 strains in an effort to identify genes that are potentially responsible for the superior healing phenotype. We hypothesized that the LG/J strain would exhibit similar biomechanical outcomes to the MRL/MpJ strain, with significant improvements over the C57BL/6 strain.

MATERIALS AND METHODS

Experimental Design

PT dimensions, along with structural and material properties, were assessed following a full-length, full thickness central PT injury in 16 to 20 week old male LG/J mice. Breeding pairs (stock number: 000675) were obtained from The Jackson Laboratory (Bar Harbor, ME) and bred in-house. Biomechanical data was obtained after 2, 5, and 8 (n= 8–10 per time point) weeks of natural healing post surgery. Comparisons were made to uninjured PTs from 16 to 20 week old LG/J mice (n=12) and to data previously obtained in our laboratory for C57BL/6 and MRL/MpJ strains¹². RNA transcriptomic analysis was performed on LG/J, MRL/MpJ and C57BL/6 mice at 3, 7 and 14 days post injury by pooling 3 tendons per time point.

Surgical Procedure

A central-third PT defect was created using a previously described surgical technique⁴. The University of Cincinnati Institutional Animal Care and Use Committee approved all procedures prior to the study. Animals were anesthetized through inhalation of 3% isoflurane and both hind limbs were shaved and aseptically prepared using ethanol and betadine washes. Mice were placed on a sterile surgical table and incisions were made on each limb to expose the PT. Medial and lateral longitudinal incisions were created on either side of the PT and then jeweler’s forceps were slipped under the tendon to isolate it from surrounding tissue. Next, a longitudinal incision was made along the length of the tendon approximately one-third of the distance from the lateral edge of the PT creating the lateral strut. Using a pair of modified jeweler’s forceps, one prong was placed anterior to the lateral strut and pushed through posteriorly to isolate the central portion of the PT. The central portion was removed by cutting at both the patellar and tibial insertions, and a modified jigsaw blade was used to disrupt the tibial insertion. Skin incisions were closed using 5-0 prolene suture (Ethicon, Somerville, NJ) and animals were allowed unrestricted cage activity following the procedure. At appropriate time points post-injury, mice were euthanized by carbon dioxide asphyxiation and cervical dislocation. Limbs were harvested for biomechanical analysis or RNA-seq experiments with qRT-PCR validation.

Biomechanical Testing

All samples were stored at −20°C until the day of testing, when they were removed and thawed at room temperature. The limbs were dissected and the gross PT morphological properties were recorded. Muscle and surrounding tissue were removed from the tibia, and the PT struts were cut away, leaving only the central third. The tibia was potted in PMMA (Dentsply International, York, PA) and secured with a staple to prevent slippage inside a grip that was mounted to the material testing system (100R, TestResources, Shakopee, MN). The patella was lowered into a specially designed, conical-shaped grip. The test was conducted in a 0.9% PBS bath at 37°C. A 0.02 N preload was applied to the PT and thickness was measured using digital photography. The tendon was preconditioned for 25 cycles between 0% and 1% strain and failed in uniaxial tension at 0.1% of total tendon length/second. Load and displacement were recorded during the test. Ultimate load, failure displacement, stress, and strain were also recorded, and linear stiffness and elastic modulus were calculated from the load-displacement and stress-strain curves.

Total RNA Isolation, Quantification, cDNA Library Creation and Sequencing

Following sacrifice at 3, 7, and 14 days following injury, the tendon midsubstance, including the repair tissue, paratenon, and medial/lateral struts of the defect samples were isolated and 3 tendons were pooled per sample. Native, uninjured tendon midsubstances were isolated to serve as a control. A total of n=1 was evaluated for each condition including native, 3 days post injury, 7 days post injury, and 14 days post injury for each strain¹⁸. Tissue samples were placed in RNAlater® (Invitrogen) and stored at −20°C. The RNAeasy® Mini Kit (74104; Qiagen, Venlo, Limburg) was used to isolate RNA from each sample following the manufacturer’s protocol. The tissue was removed from RNAlater® (Invitrogen), weighed, and then placed in an RNAse-free tube partially submerged in liquid nitrogen. The tissue was disrupted using a pestle, vortexed, and centrifuged. The supernatant was then passed through a spin column and an on-column DNase digestion was performed following the RNase-Free DNase Set protocol (79254; Qiagen), followed by subsequent washes/spins. The RNA was eluted and quantified using a NanoDrop ND-1000 Spectrophtometer (NC9904842, NanoDrop Technologies, Inc, Wilmington, DE).

Total RNA for each sample was then submitted to the Cincinnati Children’s Hospital Medical Center Gene Expression Core Facility for amplification and cDNA library creation. First, RNA quality and quantity were evaluated using the Agilent Bioanalyzer (Agilent Technologies, Inc). The Ovation® RNA-seq System V2 (7102; NuGEN Technologies, Inc., San Carlos, CA) was used to amplify the RNA samples creating 2–4 micrograms of double stranded cDNA. The Nextera DNA Sample Preparation Kit (FC-121–1031; Illumina, San Diego, CA) was then used to create DNA library templates from the double stranded cDNA. Samples were pooled and then submitted to the Cincinnati Children’s Hospital Medical Center Genetic Variation and Gene Discovery Core for sequencing using the Illumina HiSeq2500 (Illumina). The parameters for the sequencing were set to single read, 50 base pair read length, with the number of reads requested per sample set to 25–30 million.

Alignment and Analysis of Illumina Reads

The sequence reads were aligned to the Mouse NCBI37/mm9 (NCBI and the Mouse Genome Sequencing Consortium) reference genome using Avadis NGS software (Strand Genomics, San Francisco, CA). The Cufflinks package was used for transcript identification and assembly.

Pattern Based Clustering

Gene expression was measured in fragments per exon kilobase per million mapped sequence reads (FPKM). All genes on the mm9 reference genome were assigned to a functional category using the gene ontology (GO) classification and functional categories (Gene Ontology Consortium, 2001) and Kyoto Encyclopedia of Genes and Genomes (KEGG) database.

The identified transcripts and expression values were processed through Subsequent analysis following Pattern Based Clustering (SPBC) and visualized using a plot-clustering program. This program combines gene expression profiles, transcription factor binding site analysis, and information derived from the GO and KEGG databases to produce a statistically valid, comprehensive method of evaluating and clustering time-series RNA-seq data sets. GOStat is a software package (Bioconductor) incorporated into the clustering program used to measure GO term enrichments within each of the individual clusters, producing a p-value. Small p-values indicate that the GO terms falling in a particular cluster are unlikely to be grouped into the cluster randomly. First, significantly regulated genes were identified and grouped on fold change patterns by implementing previously characterized statistical methods^19–21. Observed fold change patterns were assigned to a set of values of 0, 1, or 2²⁰. A 0 assignment indicates little to no fold change, 1 indicates a 1.5- to 2-fold change, and 2 indicates >2-fold change. Genes showing little to no fold changes were removed from the analysis to facilitate accelerated processing times of the data sets. Three comparisons were completed within each strain across time points as follows (Figure 18b): 1) C57BL/6 D3, D7, and D14; 2) LG/J D3, D7, and D14; 3) MRL/MpJ D3, D7, and D14.

Real-Time qPCR Validation

First-strand cDNA synthesis was performed with the High Capacity RNA-to-cDNA kit (4387406; Applied Biosystems, Grand Island, NY). Real-time qPCR reactions were run using Taqman® Gene Expression Fast Mastermix (4364103; Applied Biosystems) and Taqman® Gene Expression Assays (Applied Biosystems) for Col1a1, Col3a1, Dcn, Fmod, TnC, Tnmd, Scx, Mkx, Egr1, p21, Myc, Ifng, Pou5f1, and Hoxa13, Table 1. All reactions were run in triplicate to account for technical variability and averaged. Delta CT values were computed and normalized to 18S expression within each sample.

Table 1.

Gene symbols and gene expression assay ID numbers (Applied Biosystems)

Name	Symbol	Gene Expression Assay ID
Eukaryotic 18S rRNA	18s	Hs99999901_s1
Collagen Type I, alpha 1 chain	Col1a1	Mm00801666_g1
Collagen Type III, alpha 1 chain	Col3a1	Mm01254476_m1
Decorin	Dcn	Mm00514535_m1
Fibromodulin	Fmod	Mm00491215_m1
Tenascin-C	TnC	Mm00495662_m1
Tenomodulin	Tnmd	Mm00491594_m1
Scleraxis	Scx	Mm01205675_m1
Mohawk Homeobox	Mkx	Mm00617017_m1
Early Growth Response-1	Egr1	Mm00656724_m1
Cyclin-Dependent Kinase Inhibitor 1A	p21	Mm04205640_g1
Myelocytomatosis Oncogene	Myc	Mm00487804_m1
Interferon Gamma	Ifng	Mm01168134_m1
POU Domain, Class 5, Transcription Factor 1	Pou5f1	Mm03053917_g1
Homeobox A13	Hoxa13	Mm00433967_m1

Statistical Analysis - Biomechanics

Native LG/J PT dimensions and mechanical properties were compared to native C57BL/6 and MRL/MpJ properties using independent Welch’s two-sample t-tests. A one-way ANOVA with Tukey’s Honest Significant Difference was used to test the significance of healing at each time point compared to native LG/J properties. Welch’s two-sample t-tests were used to compare mechanical properties between strains at the same time points. All comparisons were made using a p value of 0.01. All statistical testing was performed using R (version 3.2.3).

RESULTS

Native Patellar Tendon Properties

For native PT dimension and mechanical properties, the LG/J exhibited larger thickness, cross sectional area and lower material properties than the MRL/MpJ and C57BL/6 mice, Table 2. The LG/J PT thickness (0.66±0.17) was significantly higher than both the C57BL/6 strain (0.47±0.06, p=0.002) and the MRL/MpJ strain (0.45±0.03, p=0.001). Additionally, the LG/J PT cross-sectional area (0.40±0.12) was significantly higher than both the C57BL/6 strain (0.27±0.04, p=0.003) and the MRL/MpJ strain (0.25±0.03, p=0.001). The LG/J maximum stress (11.49±4.00 MPa) was significantly lower than both the MRL/MpJ strain (17.85±6.72 MPa, p=0.0149) and the C57BL/6 strain (16.20±2.72 MPa, p=0.003). The LG/J elastic modulus (86.00±16.58 MPa) was significantly lower than both the MRL/MpJ strain (130.98±28.03 MPa, p=0.0003) and the C57BL/6 strain (105.30±20.11 MPa, p=0.013). With respect to structural properties, the native linear stiffness for the LG/J (10.88±2.34 N/mm) was not different from the C57BL/6 (9.76±1.30 N/mm) or the MRL/MpJ (9.31±1.99 N/mm), Figure 1. The native ultimate load for the LG/J (4.30±1.46 N) was also not different from the C57BL/6 (4.44±0.75 N) or the MRL/MpJ (4.31±1.53 N), Figure 1

Table 2.

Mechanical Properties for C57BL/6, MRL/MpJ, and LG/J Native and Defects (mean±SD)

	Cross-Sectional Area (mm2)	Ultimate Load (N)	Linear Stiffness (N/mm)	Max Stress (MPa)	Modulus (MPa)
C57BL/6
Native PT (n=14)	0.28±0.04	4.44±0.75	9.76±1.30	16.20±2.72	105.30±20.11
2-week Defect (n=8)	0.38±0.07	2.10±0.25	4.50±0.66	5.80±0.72	39.03±6.45
5-week Defect (n=9)	0.31±0.06	1.87±0.55	4.74±0.79	6.31±2.19	49.53±15.12
8-week Defect (n=7)	0.25±0.03	1.82±0.66	5.72±0.85	7.24±2.47	67.63±11.62
MRL/MpJ
Native PT (n=11)	0.25±0.03	4.31±1.53	9.31±1.99	17.85±6.72	130.96±39.99
2-week Defect (n=8)	0.54±0.12	2.91±1.15	6.50±1.43	5.33±1.64	39.37±7.27
5-week Defect (n=10)	0.39±0.07	1.96±0.29	4.06±1.40	5.12±1.31	38.44±12.87
8-week Defect (n=10)	0.31±0.08	3.48±0.92	7.21±1.00	11.33±3.73	74.49±20.67
LG/J
Native PT (n=12)	0.39±0.12b,c	4.30±1.46	10.88±2.34	11.49±4.00b	86.00±16.58c
2-week Defect (n=10)	0.60±0.35	2.47±1.02a	5.66±2.38a	5.85±4.74	43.92±32.54a
5-week Defect (n=8)	0.45±0.09	3.65±0.74b,c	8.18±1.69b	8.40±2.67	57.25±15.61
8-week Defect (n=8)	0.46±0.16	4.17±1.16b	10.52±3.40b	9.90±4.45	78.61±38.69

^aSignificantly different from respective native

^bSignificantly different from C57BL/6

^cSignificantly different from MRL/MpJ

Figure 1.

The native structural properties of the three strains are not significantly. Data for the C57BL/6 and MRL/MpJ obtained from Lalley et al.¹². Error bars represent SD.

LG/J Mechanical Properties Return to Near Normal Values by 8 Weeks

At 5 and 8 weeks post surgery, the LG/J PT repair tissue was not different from native LG/J PT tissue with respect to structural properties, Figure 2, Figure 3a, Figure 3b. At 5 weeks, the repair tissue reached 84% and 75% of native LG/J ultimate load and linear stiffness, respectively, Figure 3a, Figure 3b. By 8 weeks, the repair tissue reached 97% and 97% of native LG/J ultimate load and linear stiffness respectively, Figure 3a, Figure 3b. For material properties, both the elastic modulus and maximum stress returned to normal levels by 8 weeks post surgery, reaching 92% and 86% of native values, respectively, Figure 3c, Figure 3d.

Figure 2.

Average failure curves for LG/J repaired tendon tissue and 2, 5, and 8 weeks post surgery. At two weeks, the repair is significantly weaker than native, however at 5 and 8 weeks there is no significant difference between the repair tissues and native. Error bars represent SD.

Figure 3.

Structural and material properties of defect tendon tissues plotted as a percent of native for (a) ultimate load, (b) linear stiffness, (c) ultimate stress, and (d) modulus. The LG/J repair tissue exhibited improved structural and material properties between 5 and 8 weeks post surgery compared to both the MRL/MpJ and C57BL/6 murine strains. Error bars indicate SD. The MRL/MpJ data was collected and reported in a previous manuscript¹². * indicates the LG/J strain is different than both the C57BL/6 and MRL/MpJ strains. + indicates the LG/J strain is different than the C57BL/6 strain.

The LG/J ultimate strength was significantly greater than both the MRL/MpJ and the C57BL/6 at 5 weeks (3.65±0.74 N vs. 1.96±0.26 N, p=0.002; 3.65±0.74 N vs. 1.87±0.55 N, p=0.0002). Additionally, at 8 weeks, the LG/J ultimate load was significantly increased compared to the C57BL/6 strain (4.17±1.16 N vs. 1.82±0.66 N, p=0.0006). The LG/J linear stiffness was significantly greater than both the MRL/MpJ strain (8.18±1.69 N/mm vs. 4.06±1.40 N/mm, p=0.024; 10.52±3.40 N/mm vs. 7.21±1.00 N/mm, p= 0.028) and the C57BL/6 strain (8.18±1.69 N/mm vs. 4.74±0.79 N/mm, p=0.0004; 10.52±3.40 N/mm vs. 5.72±0.85 N/mm, p= 0.005) at 5 and 8 weeks respectively.

While the LG/J exhibited increased structural properties compared to the MRL/MpJ and C57BL/6 mice at 5 and 8 weeks, there were no differences in structural, Figure 3a, Figure 3b, or material properties at 2 weeks, Figure 3c, Figure 3d.

RNA-seq Cluster Analysis

Comparison Within Strains

For each comparison, we evaluated Day 3 to Day 7 to Day 14 because we were interested in the genetic changes occurring with respect to time for each strain. We evaluated clusters of the observed values (+2,+2) and (−2, −2) to identify the most differentially expressed genes for each comparison.

C57BL/6

Clustering analysis showed 351 genes comprised the (+2, +2) cluster and 338 genes comprised the (−2, −2) cluster. This analysis indicates the C57BL/6 shows increased expression of genes involved in the muscle system (p=1.22e-09) and decreased activation of cell cycle regulating pathways (p=3.47e-08) over time.

MRL/MpJ

Clustering analysis revealed 384 genes comprised the (+2, +2) cluster and 284 genes comprised the (−2, −2) cluster. The MRL/MpJ showed increased activation of pathways involved in muscle cell differentiation (p=1.82e-04), glucan and glycogen metaboloic processes (p=2.03e-04), developmental processes (p=3.56e-04), and collagen fibril organization (p=0.0019). This strain showed decreased activation for pathways involving the immune response (p=2.33e-07) and inflammation (p=3.61e-04).

LG/J (Figures 4 and 5)

Figure 4.

(+2, +2) GO terms and p-values (LG/J Time Comparison)

Figure 5.

(−2, −2) GO terms and p-values (LG/J Time Comparison)

Clustering analysis indicated 364 genes comprised the (+2, +2) cluster and 140 genes comprised the (−2, −2) cluster. The LG/J shows activation of pathways regulating cell adhesion (p=2.31e-12), inflammatory response (p=1.58e-04), and collagen fibril organization (p=0.0021), and shows decreased activation of peptidyl–serine and –tyrosine phosphorylation (p=2.21e-04 and p=0.0053).

Validation of RNA-seq Results

To validate the RNA-seq expression profiles, quantitative real-time PCR was performed on 14 genes of interest, Table 1, to evaluate overall temporal expression trends. The genes selected for real-time PCR validation were those with a known association with tendon development and/or homeostasis (Col1a1, Col3a1, Dcn, Fmod, TnC, Tnmd, Scx, Mkx, Egr1) and those genes previously implicated in the literature to potentially play a role in the MRL/MpJ and LG/J healing phenotype (p21, Myc, Ifng, Pou5f1, and Hoxa13). For 12 of the 14 genes, qRT-PCR temporal expression patterns were in agreement with the measured RNA-seq data, Figure 6a, b, c, while 2 of the 14 genes (Hoxa13 and Pou5f1) did not match RNA-seq expression profile patterns, Figure 6d. Those genes that did not match RNA-seq results were transcription factors expressed at the upper detection limits of the qRT-PCR assay.

Figure 6.

Figure 6a. Comparison of temporal gene expression patterns obtained from RNA-Seq and qPCR technologies for C57BL/6, MRL/MpJ and LG/J. Expression of collagen type I (Col1a1), collagen type III (Col3a1), scleraxis (Scx), and mohawk homeobox (Mkx) are represented. Overall, the temporal expression patterns measured using qPCR and RNASeq within each strain for each gene were similar. Fragments per exon kilobase per million mapped sequence reads (FPKM) values were plotted against the left y-axis (RNA-Seq output) and 1-Ct values were plotted against the right y-axis (qPCR output). Native tissue and time post-surgery are represented on the x-axis. Figure 6b. Comparison of temporal gene expression patterns obtained from RNA-Seq and qPCR technologies for C57BL/6, MRL/MpJ and LG/J. Expression of early growth response 1 (Egr1), tenascin-C (TnC), tenomodulin (Tnmd), and fibromodulin (Fmod) are represented. Overall, the temporal expression patterns measured using qPCR and RNASeq within each strain for each gene were similar. Fragments per exon kilobase per million mapped sequence reads (FPKM) values were plotted against the left y-axis (RNA-Seq output) and 1-Ct values were plotted against the right y-axis (qPCR output). Native tissue and time post-surgery are represented on the x-axis. Figure 6c. Comparison of temporal gene expression patterns obtained from RNA-Seq and qPCR technologies for C57BL/6, MRL/MpJ and LG/J. Expression of cyclin-dependent kinase inhibitor 1 (p21), v-myc avian myelocytomatosis viral oncogene homolog (Myc), decorin (Dcn), and interferon gamma (IFNG) are represented. Overall, the temporal expression patterns measured using qPCR and RNASeq within each strain for each gene were similar, with the exception of MRL/MpJ’s recorded IFNG expression. This can be attributed to low expression values for this gene with Ct values measured at the upper limits of the Taqman gene expression assay. Fragments per exon kilobase per million mapped sequence reads (FPKM) values were plotted against the left y-axis (RNA-Seq output) and 1-Ct values were plotted against the right y-axis (qPCR output). Native tissue and time post-surgery are represented on the x-axis. Figure 6d. Comparison of temporal gene expression patterns obtained from RNA-Seq and qPCR technologies for C57/BL6, MRL/MpJ, and LG/J. Expression of POU class5 homeobox 1 (Pou5f1) and homeobox A13 (Hoxa13) are represented. The temporal expression patterns for both Pou5f1 and Hoxa13 were different between the qPCR and RNA-Seq measures. This can be attributed to low expression values for these two genes with Ct values measured at the upper limits of the Taqman gene expression assay. Fragments per exon kilobase per million mapped sequence reads (FPKM) values were plotted against the left y-axis (RNA-Seq output) and 1-Ct values were plotted against the right y-axis (qPCR output). Native tissue and time post-surgery are represented on the x-axis.

DISCUSSION

Identifying and characterizing mammalian models of regeneration has the potential to greatly impact the field of regenerative medicine. Over the past decade, researchers have investigated the MRL/MpJ and LG/J mice strains that exhibit superior healing outcomes in a number of injury models^{7, 9, 16, 17}. Despite this work, it still remains unknown what molecular mechanisms are driving the improved healing response, although many hypotheses have been proposed including mutations producing a p21 deficiency, increased activation of stem-related markers, and abnormalities in the immune and inflammatory response following injury^{9, 22, 23}. Researchers have also determined the healing response may depend on the type and location of the injury²⁴. The goal of this study was to examine the mechanical properties of the LG/J murine patellar tendon before and after injury to determine its biomechanical properties compared to wild-type and MRL/MpJ mice. In addition, this study was designed to determine the differentially activated pathways that are involved in the superior tendon healing following a central patellar tendon injury. Evaluating the early response to injury, measured at 3, 7, and 14 days, could provide information in determining how early signaling events might contribute to long-term repair outcome.

Most of the physical and mechanical properties of native LG/J tendons were not significantly different from either the C57BL/6 or MRL/MpJ native tendons, but the LG/J repair tissue biomechanical properties were significantly greater than the MRL/MpJ and the C57BL/6 at 5 ad 8 weeks post surgery. By 8 weeks, the LG/J repair tissue reached 97% of native for both ultimate load and linear stiffness, exhibiting essentially normal failure curves, compared to 81% of native ultimate load and 77% of native linear stiffness in the MRL/MpJ at 8 weeks. Although the MRL/MpJ had higher values for ultimate load and linear stiffness at 2 weeks compared to the LG/J (yet not statistically significant), the MRL/MpJ’s healing response appears to drop off after the 2-week point, resulting in decreases in ultimate load and linear stiffness at 5 weeks. The LG/J exhibited progressive improvements in biomechanical properties from 2 through 8 weeks. This suggests that different healing pathways are activated starting at 2 weeks post-injury.

To examine the differentially expressed genes between the three strains, we first conducted a series of clustering experiments to understand the global differences in pathway activation existing among the C57BL/6, MRL/MpJ and LG/J strains following tendon injury. Pathway activation assessed over time within each strain showed both the LG/J and MRL/MpJ activated pathways for collagen fibril organization, while the C57BL/6 did not. The early activation of genes in this pathway could help to explain the increased mechanical properties at late post-surgical time points.

This study does have limitations. 1) Our excisional model of tendon injury is not clinically relevant. Most tendon injuries are a result of a chronic condition called tendinopathy, which is linked to aging and excessive exercise²⁵. Tendinopathy, and its associated degenerative phenotype, may also be the cause of ruptures that do not have an identifiable traumatic cause^{25, 26}. 2) We limited our clustering analysis to (+2, +2) and (−2, −2) expression patterns. The Plot Cluster analysis software allows comparisons to be constructed among any of the data sets and provides clustering outputs in the form of (−2, −1, 0, 1, 2). We chose to evaluate the (+2, +2) and (−2, −2) patterns for each comparison because we were most interested in identifying pathways that showed the greatest fold changes between groups, and future work should investigate other possible groups. For the purposes of understanding the global responses to tendon injury across strains, our approach for investigating the (+2, +2) and (−2, −2) patterns is sufficient. 3) We chose to investigate only male mice within these experiments. Prior work suggests there may be gender-related differences with respect to wound healing in a variety of mouse strains, including the MRL/MpJ and LG/J. Blankenhorn et. al. reported that following an ear-hole punch, LG/J females more readily and more completely repaired the injury compared to LG/J males (Blankenhorn et. al., Mamm Genome, 2009). However, Rai et. al. found gender had no effect on healing following an articular cartilage injury in either the MRL/MpJ or the LG/J (Rai et. al., Arthritis Rheum., 2012). Clearly, more investigation is needed to better understand the sex-related differences in regards to this regenerative phenotype.

While we have employed extensive sequencing technologies and novel analysis tools, the molecular mechanisms driving the LG/J and MRL/MpJ healing phenotype remain unknown. However, results from this study suggest an examination of the role of the immune response and collagen fibril assembly. First, we need to identify and characterize the cell populations that express the transcription factors that show differential expression. This will provide us with a better understanding of the cell populations that respond to the tendon injury and allow us to characterize their complete phenotype. Secondly, we need to evaluate the effects of ablating aspects of the healing response to determine how it impacts the ultimate repair outcome. Fully characterizing these model systems has tremendous potential for improving clinical approaches for treatment of tendon injuries.