Reduction in post-natal weight-bearing does not alter the trajectory of murine meniscus growth and maturation

Abstract

The early post-natal period represents a critical window for the maturation and development of orthopedic tissues, including those within the knee joint. To understand how mechanical loading impacts the maturational trajectory of the meniscus and other tissues of the hindlimb, perturbation of post-natal weight bearing was achieved through surgical resection of the sciatic nerve in neonatal mice at 1 or 14 days old. Sciatic nerve resection (SNR) produced significant and persistent disruptions in gait, leading to reduced tibial length and reductions in Achilles tendon mechanical properties. However, SNR resulted in minimal disruptions in morphometric parameters of the menisci and other structures in the knee joint, with no differences in Col1a1-YFP or Col2a1-CFP expressing cells within the menisci. Furthermore, micromechanical properties of the meniscus and cartilage (as assessed by atomic force microscopy-based nanoindentation testing) were not different between experimental groups. In contrast to our initial hypothesis, reduced hindlimb weight bearing via neonatal SNR did not significantly impact the growth and development of the knee meniscus. This unexpected finding demonstrates that the input mechanical threshold required to sustain meniscus development may be lower than previously hypothesized, though future studies incorporating skeletal kinematic models coupled with force plate measurements will be required to calculate the loads passing through the affected hindlimb and precisely define these thresholds. Collectively, these results provide insight into the mechanobiological responses of the meniscus to alterations in load, and contribute to our understanding of the factors that influence normal post-natal development.

INTRODUCTION:

The meniscus is a complex fibrocartilaginous tissue within the knee joint that serves a critical role in the transfer and re-distribution of mechanical loads.^1,2 In the adult, regional specialization of the cellular and extracellular matrix (ECM) components of the meniscus enables appropriate distribution of mechanical forces and is crucial for maintaining joint health and function throughout life. Specifically, the high proteoglycan content and abundance of type II collagen in the inner meniscus enables the tissue to withstand compressive forces, while the circumferentially aligned type I collagen fibers that predominate in the outer meniscus function to resist circumferential hoop stresses.^3,4 Regional complexity within the mature meniscus arises from fate decisions that occur during embryonic and early post-natal development after the meniscus forms from the interzone early in embryogenesis during knee joint formation.^5,6 Prior to birth, the fetal murine meniscus already shows a distinct inner to outer regional variation, as evidenced by cellular and ECM organization. After birth, rapid growth occurs with increases in meniscus cross sectional area and tibial plateau coverage, decreases in cellularity, and increases in mechanics.⁷ Additionally, the regional specification continues with divergent expression of genes that will come to define the ECM of the outer and inner zone in the adult. These changes in morphometry, collagen and proteoglycan content, fiber arrangement, and regional expression become even more prominent through the first month of life.^6,8

Given the marked specialization that occurs during late embryogenesis (during which time muscle forces and contraction are operative) and early weight-bearing life, many have postulated that load-bearing plays a critical role in defining meniscus regional phenotype. This hypothesis is based, in part, on observations of regional specialization in a variety of species, ranging from rabbit to human, early in postnatal development.^9–11 In other load-bearing tissues, changing mechanical inputs during early postnatal growth alters the trajectory of tissue maturation and specialization. For example, blockade of muscle contraction with botulinum toxin in the developing enthesis in mice results in changes in both soft tissue and bone organization and mineralization.¹² Consistent with the important role of the meniscus in the transfer of mechanical forces within the knee joint, muscle paralysis of chick embryos in ovo results in initial meniscus formation followed by its subsequent degeneration. This finding suggests that continued loading of the knee joint via skeletal muscle contraction is an important factor for maintaining normal meniscus development.¹³ In mature mice, the meniscus degenerates as a result of decreased load-bearing via hindlimb suspension or microgravity conditions during spaceflight.¹⁴ Similarly, knee joint cast-immobilization and prevention of weight-bearing via a sling also causes degenerative changes in the meniscus.^15,16 However, little is known regarding the effect of altered mechanical loading specifically within the early post-natal phase of meniscus development, during which its rapid growth and specialization occurs.

To understand how altered mechanical loading affects the trajectory of post-natal meniscus maturation, this study employed surgical resection of the sciatic nerve to alter hindlimb loading in neonatal mice. The sciatic nerve originates from the ventral rami of spinal nerves L4, L5, S1, S1, and S3 and supplies motor innervation to the posterior and distal hindlimb musculature. Previous studies have shown that sciatic nerve resection (SNR) in mice eight days after birth results in significant disruptions in gait and decreases post-natal tibial growth plate organization and tibial lengthening.¹⁷ However, the impact of SNR on post-natal maturation specifically of the meniscus and other soft tissues of the knee has not been evaluated. To delineate the effects of altered mechanical loading on meniscus development at different stages within the early post-natal period, we applied SNR on post-natal day 1 (SNR P1) or post-natal day 14 (SNR P14) (Figure 1). We hypothesized, given the marked changes in meniscus structure, function, and specialization during these critical time windows, and the potentially important role that mechanical loading might play in this process, that SNR would result in significant alterations in meniscus development in both groups of mice, with more severe developmental abnormalities seen in SNR P1 mice compared to SNR P14 mice.

Figure 1:

Experimental design and timing of outcome analyses along with intraoperative images of sciatic nerve resection (SNR) at post-natal days 1 (P1) and 14 (P14).

MATERIALS AND METHODS:

Animal care and use:

All animals were cared for in accordance with the guidelines of the National Institutes of Health, and procedures were approved by the University of Pennsylvania Institutional Animal Care and Use Committees (Protocol #806669). Experiments were performed using the Col1-YFP/Col2-CFP double transgenic line in male and female mice on a CD1 background.^8,18,19

Sciatic nerve resection:

Unilateral sciatic nerve resection (SNR) was performed with microscopic guidance on post-natal day 1 (SNR P1) or post-natal day 14 (SNR P14) mouse pups (Figure 1). SNR P1 mice were anesthetized via hypothermia, while SNR P14 mice were anesthetized with inhaled isoflurane (1–4%, titrated to effect in O₂). Following induction of a surgical plane of anesthesia, the surgical site was aseptically prepared, and a linear skin incision was made parallel with and caudal to the left femur on the lateral aspect of the proximal left hindlimb. The sciatic nerve was exposed through gentle blunt dissection between the vastus lateralis and biceps femoris. A portion of the nerve representing 50–75% of the femur length was resected proximal to its bifurcation. Following nerve resection, the skin incision was closed with 6–0 suture. Sham sciatic nerve resection was performed on post-natal day 14 (sham P14) for CatWalk XT gait analysis; sham-surgery involved all of the aforementioned steps without resection of the nerve. After surgery, pups were returned to their cages and monitored closely throughout the remainder of the weaning stage, with regular assessment of body mass. Non-operated control mice served as littermate controls within the same cage of operated mice. Comparisons between littermates was performed whenever possible; however, to ensure equal sex distributions, mice from other litters were included in some analyses.

CatWalk XT gait analysis:

Gait analysis was performed at post-natal day 42 in male and female SNR P1 (n=14), SNR P14 (n=14), sham P14 (n=7), and control (n=12) mice via the CatWalk XT gait analysis system (version 10.6; Noldus Information Technology, Wageningen, The Netherlands). Prior to testing, mice were habituated to the room for 30 minutes. All settings were consistent between treatment groups and sexes (minimum run duration: 0.5 seconds; maximum run duration: 15 seconds; maximum allowed speed variation: 60%; Green Intensity Threshold: 0.21; Red Ceiling Light: 17.7; Green Walkway Light: 16.5; Camera Gain: 20). A compliant run on the CatWalk XT platform was defined as a run in which the animal continued moving forward along the 1.3 m glass platform within the run duration and detection settings without turning or changing direction. For each animal, 3–5 compliant runs were used for analysis of each parameter. Automatic classification of the right forelimb, right hindlimb, left forelimb, and left hindlimb pawprints was verified manually. Measured parameters included speed, paw print width and contact area, maximum and mean paw print intensities, and percentage of left (operated) hindlimb paw prints above the detectable threshold of the CatWalk XT system. Due to the low percentage of left (operated) hindlimb paw print detection in SNR mice, some parameters (limb swing speed, stride length, and time between two consecutive paw contacts) were not analyzed.

Video tracking gait analysis:

Custom gait analysis was performed at post-natal day 14 in SNR P1 and control mice (n=4–5/group) to compare the knee joint angles in operated, contralateral control, and non-operated control hindlimbs during the gait cycle (both stance and swing phases). A custom-made walkway with a transparent bottom was enclosed by 2 mirrors on either side to obtain video recordings from three different views (ventral and left and right sagittal views). A digital camera was placed underneath the glass runway and videos were recorded. Mice walked across the 20 cm walkway at least 5 times. MaxTRAQ software (Innovision Systems Inc.) was used for 2D manual tracking of bony landmarks. Videos were converted to JPEG image sequences and imported into the MaxTRAQ software, followed by identification and marking of the hip, knee and ankle joint centers, from which the knee joint flexion angle was measured. Measurements made at initial contact used the first frame where the paw touched the ground (confirmed using both ventral and sagittal views). Knee angles at paw lift-off used the last frame at which the paw was in contact with the walkway before entering the swing phase. Knee joint range of motion was defined as the difference between the maximum and minimum knee flexion angles across a complete gait cycle. For analysis of each parameter, average values of knee flexion angle were obtained from 3–4 strides per animal.

Tissue harvest:

All mice were euthanized prior to tissue harvest via carbon dioxide exposure. For micro-computed tomographic imaging and histological analyses, disarticulated hindlimbs were fixed in 10% neutral buffered formalin for 2 days at 4° prior to a 1-day incubation in 30% sucrose at 4°. Following this, hindlimbs were stored at −20° until micro-computed tomographic imaging and/or embedding in OCT and stored at −80°. For atomic force microscopy nano-indentation testing and biomechanical testing of the Achilles tendons, whole mice were stored at −20° following euthanasia until dissection of unfixed tissues for analysis.

Multiplexed cryohistology:

Cryosectioning (8 μm slices) of OCT-embedded knee joints was performed in the coronal place from the level of the patellar tendon through the posterior meniscus horn using cryofilm 2C (Section-lab Co).^8,20 Sections were stored at 4° prior to slide fixation. Sections were fixed to glass slides using a Chitosan adhesive. Knee joints were imaged with a Zeiss Axio Scan.Z1 slide scanner using a 10X objective. To assess fluorescent reporters, sections were counterstained with TO-PRO-3 (ThermoFisher, Cat#: T3605) and then imaged under darkfield, polarized light, and in the Cy5, CFP, YFP, and RFP channels. Analysis of each channel was performed separately. Next, sections were stained with Toluidine Blue (0.025% solution, Sigma 364-M) and imaged under brightfield.^8,20

Meniscus histomorphometry measurements:

Height, length, area, and cellularity of the medial meniscus mid-body adjacent to the medial collateral ligament was measured using Fiji image analysis software in SNR P1, SNR P14, and control mice (2–3 sections per mouse). Meniscus height was determined by measuring the outermost edge of the meniscus body within the body of the meniscus. Meniscus length was obtained by measuring a point midway of the meniscus height to the innermost point of the meniscus. Cross-sectional area was obtained from drawing an outline around the meniscus surfaces. The number of nuclei within this region was counted to determine cellularity. Medical meniscus height, length, area, and cellularity were measured at post-natal day 42 (n=3–4 per group) as well as post-natal days 14 and 28 (n=5–6 per group). For fluorescent reporter quantification, the nuclei were segmented via automated thresholding using Fiji image analysis software. The mean intensity value was then recorded for Col1-YFP and Col2-CFP within each nucleus. A consistent minimum intensity threshold was applied to quantify the percentage of Col1-YFP+ and Col2-CFP+ cells.

Atomic force microscopy (AFM)-based nanoindentation:

AFM-nanoindentation was applied to freshly dissected menisci and femoral condyle cartilage from SNR P1 and control mice (n=5–6 per group) at post-natal days 14 and 28, following established procedures.^21,22 In brief, freshly harvested cartilage and meniscus tissues were mounted onto AFM sample disks using a cyanoacrylate adhesive gel, and then, tested using polystyrene microspherical tips (R ≈ 5 μm, nominal k ≈ 8.9 N/m, HQ:NSC36/tipless/Cr-Au, cantilever C, NanoAndMore, Watsonville, CA) and a Dimension Icon AFM (BrukerNano, Billerica, MA, USA) in 1× PBS with protease inhibitors. For each sample, at least 15 different indentation locations were tested on the load-bearing regions up to ∼1 μN force at 10 μm/s AFM z-piezo displacement rate. Care was taken to test only within the meniscal body to avoid areas of mineralization in the anterior and posterior horns. The effective indentation modulus, E_ind, was calculated by fitting the entire loading portion of each force-displacement curve to the Hertz model, assuming the Poisson’s ratio (ν) of 0.1 for cartilage²³ and 0 for the meniscus.²⁴

Achilles tendon mechanical testing:

Achilles tendons were dissected from the left hindlimb of SNR P1 and control mice (n=5–6 per group) at post-natal day 42. All tendons were stamped into a dog bone shape (equal width per sample) for mechanical testing. Sandpaper was placed 5 mm proximal to the calcaneal insertion and the calcaneus was placed in custom grips. Samples were loaded into a PBS bath on an Instron 5542 universal testing system. All tendons were preconditioned (10 cycles between 0.02 N and 0.04 N) followed by extension to failure at 0.1%/sec. Structural and material properties were calculated using custom MATLAB scripts.

Micro-computed Tomographic Imaging:

Excised left hindlimbs of SNR P1, SNR P14, and control mice (n=4 per group) at post-natal day 42 were imaged via a Scanco μCT45 scanner (voxel size=10.5 μm, energy=55 kV). 3D reconstruction scans were created using Dragonfly 2020.2 software (Object Research Systems Inc., Montreal, Canada). Tibial length was determined by drawing a line between the intercondylar eminence and the medial malleolus. Trabecular parameters in the proximal metaphysis of the tibia were quantified: bone volume/total volume (BV/TV), trabecular spacing, and trabecular thickness.

Statistics:

Quantitative data were processed with GraphPad Prism (version 9.0; GraphPad Software, San Diego, CA, USA) using student’s t-test, or one-way ANOVA followed by Tukey-Kramer post-hoc multiple comparison to correct for family-wise type I errors. In all the tests, the significance level was set at α = 0.05.

RESULTS:

Sciatic nerve resection (SNR) in the early postnatal period results in prolonged gait deficits.

Sciatic nerve resection was carried out on neonatal mice at P1 and P14. All animals recovered from surgery well and were successfully reintroduced to their litters. These mice exhibited normal activity levels and weight gain, with body weights obtained at the time of gait analysis not differing between experimental groups (Figure 2B). Severe alterations in gait were observed in the SNR limbs when evaluated on post-natal day 42 via the CatWalk XT gait analysis system. Over 50% of left (operated) hindlimb strides in 9 mice in the SNR P1 group (n=14) and 12 mice in the SNR P14 group (n=14) were not detected by the CatWalk XT system (Figure 2A,C), compared to 100% detection of the left hindlimb in the control and sham P14 groups. Of the detectable strides, there was a marked reduction in paw print width and maximum contact area in the left (operated) hindlimb of SNR P1 and SNR P14 groups, consistent with the inability to spread the toes as a result of sciatic denervation (Figure 2D–F). Significant differences were also observed in the measured stride intensities of the left (operated) hindlimb in SNR P1 and SNR P14 mice compared to control and sham surgical mice (Figure 2A,G; Supplemental Figure 1A–D). These differences spanned multiple parameters output from the gait analysis software, including Maximum Intensity of the complete paw; Maximum Intensity at the time of maximum paw contact; Mean Intensity of the complete paw; Mean Intensity at the time of maximum paw contact; and Mean Intensity of the 15 most intense pixels of the paw. Minimal differences were observed in the average body speed and mean body speed within a left hindlimb step cycle between experimental groups, with the only significant difference found between SNR P1 and sham P14 mice (average body speed: p=0.022; mean body speed LH: p=0.033) (Supplemental Figure 1E,F). No differences were observed between the sham operated limb and non-operated control limb for any measured parameter. Additionally, no significant differences were observed in the parameters of any other limb (i.e., right forelimb, left forelimb, or right hindlimb) across any experimental group. Together, these data indicate that SNR at P1 or P14 results in persistent alterations in gait.

Figure 2:

CatWalk XT gait analysis system demonstrating significant gait alterations in SNR P1 and SNR P14 mice compared to control and sham surgical mice at post-natal day 42. (A) 2D footprint plots illustrating recordings of the right forelimb (RF), left forelimb (LF), right hindlimb (RH) and left hindlimb (LH). Significant decreases in paw print detectability and intensity of the operated hindlimb (LH) observed in SNR mice. (B) No significant differences in body weight between treatment groups. (C) Significant decreases in the percentage of operated hindlimb (LH) strides reaching above the detectable threshold were noted in SNR mice, with data points in red corresponding to the 2D footprint plots of mice represented in (A). Significant decreases in paw print width (D) and contact area (E) were seen in the operated hindlimb (LH) in SNR mice, corresponding with an inability to spread the toes due to sciatic denervation (F). (G) Maximum intensity measurements of the operated hindlimb (LH) were significantly decreased in SNR mice. ***p<0.005, ****p<0.0001, ns = not significant.

Abnormal gait patterns following SNR are evident in young animals.

While the CatWalk XT gait results demonstrated persistent gait deficits at P42 that culminated in reductions of hindlimb growth and development, our previous work identified postnatal day 1 through 21 as the time during which the meniscus changes most in terms of cellularity, expression patterns, and overall morphology.⁸ Our initial attempts to measure gait in such young animals did not yield usable data with the CatWalk XT system. In addition, the CatWalk XT system focuses on the stance phase of gait, which will likely miss gait alterations found in the swing phase. Therefore, we developed a gait system with angled mirrors such that both ventral and sagittal views of these young mice could be visualized simultaneously. Using this system, we found marked differences in knee joint flexion angles throughout the gait cycle in SNR P1 mice compared to control mice, with the SNR limbs being more extended overall. At post-natal day 14, knee flexion angles at both the time of initial paw contact and lift off were significantly decreased in operated SNR P1 hindlimbs compared to the contralateral controls (Figure 3A–C). Interestingly, these changes in motion about the knee coincided with an increase in knee joint range of motion (ROM) throughout the gait cycle, with the maximum flexion angle being lower in SNR P1 mice (Figure 3D–F). Overall, the SNR limbs were more extended during stance phase with similar flexion angles (data not shown) occurring during the swing phase compared to the contralateral limbs. Comparison between contralateral control limbs and non-surgical control limbs from unoperated mice revealed no significant differences (Supplemental Figure 2).

Figure 3:

Video gait analysis on post-natal day 14 comparing SNR P1 operated limbs to contralateral control limbs. (A) Gait analysis images with labeled hip (purple), knee (green), ankle (blue), and metatarsophalangeal (red) joints. Knee flexion angle (yellow in panel A) in the operated limbs was significantly decreased during initial paw contact (B) and paw lift off (C), corresponding with deficits in knee flexion associated with sciatic denervation. Operated limbs demonstrate a significant decrease in maximum knee flexion angle (D) and an increase in knee joint range of motion (E), with data points in red corresponding to the knee angles of an SNR P1 and control limb throughout the entire gait cycle in (F). **p<0.01

Sciatic nerve resection (SNR) in the early postnatal period compromises hindlimb growth and development.

Having established that SNR significantly alters gait and hindlimb loading, we next sought to evaluate the impact this would have on hindlimb growth and development over this 6-week period. We noted marked atrophy of the hindlimb musculature in both SNR P1 and SNR P14 mice at P42 (Figure 4A). When the tibiae of operated hindlimbs in SNR P1 and SNR P14 mice were measured via microCT, we noted a decrease in length compared to non-operated control mice (Figure 4B,C). However, we did not detect significant differences in trabecular parameters of the metaphysis (i.e., BV/TV, trabecular spacing, trabecular thickness) (Supplemental Figure 3). Besides this difference in length, no other gross differences were appreciated in tibial morphology between experimental groups. We also assessed the growth and development of the soft tissues of the lower limb. The Achilles tendon in SNR P1 hindlimbs was smaller in size and showed reduced birefringence. Since the sciatic nerve does not innervate the quadriceps, we did not test the patellar tendon as we would not expect to find a difference in mechanical properties, especially compared to other structures in the limb that would likely be more affected due to less weight-bearing and limited hamstring function. When testing the Achilles tendons uniaxial tensile properties, we found a marked reduction in stiffness and failure force in SNR P1 mice compared to non-operated control mice at P42 (Figure 5). These data indicate that the consequence of altered gait and load-bearing following SNR is a less developed lower extremity.

Figure 4:

SNR P1 and SNR P14 resulted in diminished hindlimb musculature and tibia length. (A) Post-mortem image of an SNR P1 mouse at post-natal day 42 with skin removed demonstrating marked gross muscle atrophy of the operated (left) hindlimb. (B) microCT image of a tibia at post-natal day 42 with red arrows pointing to the intercondylar notch and medial malleolus, which served as landmarks for measuring tibial length (red dashed line). (C) Significant shortening of SNR P1 and SNR P14 operated tibiae compared to control mice at post-natal day 42. ** p<0.01

Figure 5:

Significant effects of SNR P1 on biomechanical properties of the Achilles tendon at post-natal day 42. (A) Post-mortem refracted light image of the distal hindlimbs with skin removed in an SNR P1 mouse, demonstrating reduced size and birefringence of the Achilles tendons in the operated hindlimb (SNR) compared to the contralateral limb (CL). (B) Example force-displacement curve for control and SNR P1 Achilles. Significantly reduced Achilles tendon stiffness (C) and failure force (D) in SNR P1 mice, with data points in red corresponding to the SNR P1 and control samples shown in the example force-displacement curve in (B). ** p<0.01​

Sciatic nerve resection (SNR) in the early postnatal period does not alter knee meniscus growth and development.

Having established that the hindlimb in general was significantly impacted by SNR, we next turned our attention to the knee joint itself, with a focus on the medial meniscus body. Morphological analysis of the medial meniscal body in terms of height, length, area, and cellularity were not significantly different between SNR P1, SNR P14, and non-operative control groups at any time point (Figure 6, Supplemental Figures 4–6). Furthermore, there were no detectable differences in fluorescent reporter expression of Col1a1-YFP or Col2a1-CFP between experimental groups (Supplemental Figure 5). Finally, to evaluate the functional development of the knee joint, we used micromechanical testing by atomic force microscopy to directly measure the cartilage and meniscus at P14 and P28 because of the rapid stiffening that occurs in WT menisci during this time period. No differences were observed in the micro-mechanical properties of whole medial meniscus in SNR P1 mice compared to non-operative controls at post-natal day 14 or 28. Similarly, no differences were observed in the micro-mechanical properties of knee joint cartilage in SNR P1 mice compared to non-operative controls at post-natal day 14 or 28 (Figure 7). Taken together, and counter to our initial hypothesis, these data suggest that SNR does not cause substantial changes in knee joint postnatal growth and development, despite its clear impact on the growth of distal bones and fibrous tissues in the hindlimb.

Figure 6:

No apparent differences in medial meniscus body morphometry between groups at post-natal day 42. (A) Brightfield imaging of control, SNR P1, and SNR P14 sections of the medical meniscus body stained with toluidine blue. Quantifications of meniscus width (B), length (C), and area (D) at P14, P28, and P42 timepoints. ns = not significant.

Figure 7:

Atomic force microscopy (AFM)-based nanoindentation modulus, E_ind, of meniscus (A-B) and articular cartilage (C-D) in control and SNR P1 mice at post-natal days 14 and 28. ns = not significant.

DISCUSSION:

The early postnatal period is exemplified by rapid growth and development of the appendicular skeleton and specialization of the associated soft tissues. Here, we explored how perturbations in mechanical loading during this period impact this process, with a specific focus on the hindlimb. We found that neonatal sciatic nerve resection resulted in significant and persistent alterations in the growth of osseous and fibrous tissues of the distal limb through post-natal day 42. Consistent with previous results, SNR resulted in decreased growth of tibiae of the operated limb,¹⁷ as well as decreased mechanical properties of the Achilles tendons (Figure 5). However, contrary to our initial hypothesis, SNR did not markedly change development of the knee joint, which showed apparently normal growth and maturation of the meniscus and other joint structures (Figures 6–7 and Supplemental Figures 4–6). Specifically, we did not detect significant differences in meniscus height, length, area, or cellularity between controls and groups in which SNR was performed at postnatal day 1 or day 14. While certain measures may be underpowered, these findings suggest that SNR during the post-natal period did not have a dramatic influence on the trajectory of meniscus growth. Additionally, there were no detectable differences in Col1a1-YFP or Col2a1-CFP expressing cells within the meniscus between experimental groups at P14, suggesting minimal perturbations to cellular phenotype or zonal specification (Supplemental Figure 5). Furthermore, SNR P1 mice did not show any differences in micromechanical properties in cartilage or meniscus, suggesting normal functional maturation of these tissues as well. These findings were all contrary to our initial hypothesis; namely, it appeared that reduced weight-bearing of the hindlimb during early postnatal life did not have significant impact on the growth and development of the knee meniscus.

There are a number of reasons that might underlie this unexpected finding. First, it is possible that while gait was severely compromised, the amount of load passing through the knee joint (while perhaps reduced), may have still been sufficient to elicit load-mediated maturation of the meniscus. We assessed the impact of SNR at post-natal day 42 using the CatWalk XT gait analysis system, which has been used previously as a method to study gait deficits following sciatic nerve injury and repair.^25,26 As expected, given the large nerve resection (>50% of the femur length), SNR resulted in significant gait abnormalities, with many strides on the operated hindlimb being undetectable by the CatWalk XT gait analysis system. This would imply that the paw contact pressure^27–29 was not as high as an unoperated limb. These changes were also not caused by the surgical approach as the sham group did not have altered gait. We only examined shams from the P14 surgical group given that P14 mice inherently experienced greater surgical trauma and were closer in age to the P42 CatWalk testing timepoint. Furthermore, of the several parameters that were captured, including pawprint size and light intensity, significant differences were observed between SNR and control groups, including the mean and maximum light intensity values (which were significantly decreased in SNR P1 and SNR P14 mice). CatWalk XT light intensity values are calculated by the amount of light reflected when the paw contacts the glass plate, with increased light intensity values correlating with increased pressure placed on the plate.^27–29 Light intensity values are also affected by body weight and speed of the mice (which were relatively constant between experimental groups in the current study). Thus, the reduced light intensity values in the SNR mice may correlate with reduced force of the operated hindlimb on the plate, consistent with alterations in hindlimb loading. While these reduced intensity values likely correlate with reduced forces passing through the knee, our meniscus morphometry findings (Figure 6) suggest that these knee loads were sufficient to promote normal, non-pathologic knee development. For instance, due to reduced fine motor control in the distal limb, the hock of the SNR limb would often make impact with the ground at paw strike, which was previously reported.¹⁷ This action may provide an (albeit sharp and abnormal) load through the knee that was sufficient to elicit mechanically-induced maturation of the knee meniscus, but could not be accurately captured in the CatWalk data. Skeletal kinematic models, coupled with force plate measurements, will be necessary to calculate how much (and in what form) this load is passing through the affected hindlimb.

Because of the low detection level of operated hindlimbs, the inability to characterize gait characteristics at ages earlier than post-natal day 42, and the limited measurements available in swing phase using the CatWalk XT system, we also used a custom video gait analysis method at post-natal day 14 to characterize gait deficits resulting from SNR. This analysis provided additional insight that may have bearing on the reason that the meniscus continued to develop normally despite SNR. We saw at post-natal day 14 that operated limbs of SNR P1 mice exhibited increased ranges of motion of the knee joint compared to the contralateral control limb. During stance phase, the SNR limb was typically more extended than the contralateral unoperated limb. However, the knee in the SNR limb was able to flex to similar levels during the swing phase, resulting in an increased range of motion. Thus, while SNR limits muscular contraction of the knee joint, range of motion is actually increased throughout the gait cycle. In adults, because of the increasing radius of curvature of the femoral condyles, the soft tissues of the knee joint experience increasing loads (even in the absence of a ground reaction force) as the joint goes into full extension.³⁰ It is possible that the exaggerated stance phase seen in our SNR mice results in geometric tensioning of the meniscus as the joint reaches higher extension levels, and that this may have been a mechanism for maintaining normal meniscus development within the post-natal period in the context of SNR. Interestingly, previous studies have shown an anti-inflammatory effect of passive range of motion applied to the injured knee;³¹ similar phenomenon may be occurring here. While previous studies have demonstrated meniscus degeneration following knee joint immobilization in adults,^15,16 immobilizing the knee joints in growing neonatal mice would be technically very challenging. Thus, further gait characterization of SNR mice at different ages (and its impact on geometric load transfer to the meniscus) will be required to elucidate the role of knee joint range of motion in sustaining a normal trajectory of post-natal meniscus maturation.

In addition to possible contributions from aberrant gait, other sources of mechanical load may have had bearing on our findings. That is, while SNR resulted in significant gait alterations, weight-bearing was not completely eliminated and muscular forces on the knee joint were maintained through motor innervation by the femoral nerve. To elucidate the effects of additional denervation on post-natal meniscus maturation, additional methods, such as a femoral nerve resection and/or muscle paralysis via botulinum toxin injection could be considered.^12,32 Additionally, while the above focus was on active muscle driven force generation, growth generated stresses may have played a role here as well.³³ As is evident from Figure 6 and Supplemental Figures 4–6, the knee joint markedly increases in size during the first six weeks of postnatal life. Differential growth of tissues within the knee may place growth generated stresses on adjacent tissues. For instance, if the femoral condyles enlarge at a rate faster than the circumference of the meniscus increases, then this would tend to place a constant tensional stress on the meniscus, absent any additional muscle forces. In early development, these differential growth mechanisms play a role in joint cavitation events,³⁴ and may continue to play a role during early postnatal growth and development of the meniscus. Interestingly, rigid versus flaccid paralysis differentially impacts developing tendons,³⁵ suggesting that fibrous tissues that are not actively loaded, but maintained in a tensed state, may still develop normally. Collectively, these findings demonstrate that even with sustained gait alterations, which suggest reduced weight bearing on the knee, the meniscus continues to grow and mature normally.

Taken together, our findings show that sciatic nerve resection in the neonatal period produced significant gait deficits that altered the normal trajectory of osseous and fibrous tissue growth in the distal hindlimb. However, this perturbation in postnatal loading had minimal effects on the maturation of the knee meniscus. This work provides insight into the possible mechanisms that support normal post-natal meniscus development and contributes knowledge to mechanobiological responses of the meniscus to alterations in mechanical load. Further studies that alter loading through alternative methods and/or further analyze the role of joint range of motion in the neonatal period are needed to further define the role of mechanical forces on post-natal knee joint development. These experiments may be better suited for larger animal models, where we could additionally explore architectural features of the developing meniscus not present in mouse models (i.e. radial tie fibers). Our results indicate that the embryonic development of the meniscus may have established a robust framework for maintaining normal post-natal growth resilient to joint loading disruptions. Investigating the effects of impaired joint loading on embryonic meniscus development could uncover the mechanisms of mechanobiology involved in the initial formation of the meniscus, which is necessary for its rapid post-natal growth, thereby yielding new targets for developing effective regenerative strategies.