Anterior cruciate ligament (ACL) injuries are prevalent among athletes and the general public, resulting in socioeconomic stress, loss of performance, and decreased quality of life. Currently, surgical repair is the gold standard. However, these interventions are expensive and associated with limited improvement in quality of life due to the increased risk of post-traumatic osteoarthritis (PTOA). PTOA is due in part to the lack of recapitulation of the enthesis between the ligament and bone. Therefore, there is a need to develop alternative ways to improve ligament-to-bone grafting and promote the development of a functional enthesis. Tissue engineering techniques have emerged as one way to generate these tissues. Current tissue-engineered ligament grafts fail to mature completely, and usually fail at the ligament-bone interface. The purpose of this study was to use a two-step modular approach to promote enthesis formation and thereby improve the mechanical integrity of engineered ligaments. First, an enthesis construct was formed around individual brushite anchors. Second, a linear bone-to-bone ligament was formed using two enthesis anchors. Control constructs and constructs that controlled for the increased cell density (400k) were formed in parallel around two bare brushite anchors. Neither interface strength (MTL), ultimate tensile strength, collagen percentage, nor modulus was significantly different between the groups. Although, we observed a trend in interface strength, collagen content, and failure at the midsubstance with the modular constructs. Tissue infiltration at the anchor was observed, suggesting the initial formation of a tidemark. We also observed the presence of collagen X near the anchor. Although, it was less intense in the modular group.

Abstract: Manipulating Skeletal Muscle Acetylation to Enhance Healthspan and Lifespan

Human aging is marked by a progressive loss of skeletal muscle mass and cognitive function—two primary determinants of healthspan and lifespan. Even though there are many studies characterizing this association, the mechanisms underlying this relationship remain unclear. In contrast to aging, endurance exercise has been shown to improve skeletal muscle mass and function and slow age-related cognitive decline 1,2. Further, animals bred for low levels of physical activity show impaired muscle and neurocognitive function 3. The main adaptations seen with endurance exercise training are heightened fat oxidation, and increased mitochondrial mass and activity downstream of a transcriptional co-factor called the peroxisome-proliferator activates receptor 1α (PGC-1α), which is regulated by acetylation 4,5. Roberts et al (2017) demonstrated that the use of a ketogenic diet (KD), increased protein acetylation and activated PGC-1α in muscle concomitant with improved markers of skeletal muscle strength, mitochondrial mass and activity, endurance, neurocognitive function, and extended lifespan 13.6% in mice 6. Along with enhanced healthspan and lifespan, mice on a KD demonstrated improved muscle mitochondrial biogenesis, muscle mass, and oxidative fiber cross-sectional area; all things associated with endurance exercise training 7. However, the molecular mechanisms underlying the benefits seen with increases in either exercise or a KD on skeletal muscle and neurocognitive function remains largely unexplored. Our global hypothesis for this body of work was that increasing protein acetylation plays a central role in improving skeletal muscle mass and function, mitochondrial biogenesis, and neurocognitive function. In the present work, we identify novel mechanisms that underlie the ability of skeletal muscle to increase muscle mass and function, and describe the role of skeletal muscle mitochondria in augmenting cognitive performance in response to manipulation of acetylation either through natural compounds, pharmaceuticals, or a KD. Our approach to understand how acetylation both with and without a KD enhances healthspan, and lifespan was 4-fold: 1) Determine the effect of a cocktail of natural products, selected to inhibit the NAD-dependent deacetylase Sirtuin 1 (SIRT1) in rats following ablation of the gastrocnemius and soleus muscles (Functional Overload—FO), on the resulting increase in plantaris muscle fiber cross sectional area, 2) Establish how the class II histone deacetylase (HDAC) inhibitor, scriptaid, affects skeletal muscle and neurocognitive function, and 3) Determine whether a 2-month KD alters skeletal muscle function, mitochondrial mass, and cognitive behavior in middle-aged female mice, Muscle mass and strength are important aspects of human health. Low levels of muscle mass and strength are directly correlated with lifespan. In Chapter 2, we optimized a cocktail of naturally occurring compounds (epicatechin, epicatechin-gallate, and celestrol) that were believed to inhibit sirtuin 1. This work shows that optimal inhibition of SIRT1 increases global protein acetylation (Figure 2.3C), and the increase in fiber cross sectional area (fCSA) in response to FO 61.5% more than vehicle dosed counterparts (Figure 2.2H). One proposed limit to skeletal muscle hypertrophy in both mice and man is the capacity for protein synthesis (i.e., ribosome mass)8–10. In response to an optimal dose of the SIRT1 inhibitor cocktail, ribosomal RNA (which comprises ~80% of the total RNA) relative to muscle mass decreased in the optimal group (Figure 2.4B). This was supported by lower markers of ribosome biogenesis (ITS1 and 5’ETS RNA levels) indicating that the benefit of the cocktail is not the result of increased translational machinery (Figure 2.4C&D). One possible explanation for the apparent increase in hypertrophy could be an increase in translational efficiency secondary to acetylation of ribosomes 11. In support of this, immunoprecipitation of ribosomal subunits showed greater associated acetylated lysine levels in the optimal group when compared to control animals (Figure 2.6A-C). Though a long term KD has been shown to help maintain oxidative fCSA and improve muscle and cognitive metrics, there is still uncertainty surrounding the mechanism underlying the changes caused by a KD 6,7. To address this gap in knowledge, in Chapter 3 we employed an isocaloric (11.2 kcal/day) KD in 23-month-old mice (Table 3.2) together with concomitant injection of either the class II HDAC inhibitor scriptaid (SCRIPT) or a vehicle (VEH) control. Our findings demonstrated that a one-month KD intervention increases mitochondrial mass (Figure 3.2D) and improved grip strength (Figure 3.1G) in skeletal muscle preferentially while not affecting fCSA (Figure 3.4A&B). As an explanation for the significant improvement on grip strength in the KD fed animals, we are the first to report that an acute KD increases dystrophin protein levels (Figure 3.4F). These data indicate that KD muscle is better able to laterally transmit force and prevent contraction induced muscle damage than their CD fed counterparts. To understand to what degree a short-term KD started late in life influences cognitive behavior the Y-Maze, Novel Object Recognition, and Open Field assays were conducted (Figure 3.1H, 3.1J&K). We report no changes between the CD and KD fed animals. These findings indicate that an acute KD first effects skeletal muscle biochemistry and functionality prior to that of neurocognitive function. Interestingly, all of the biochemical and functional benefits seen with an acute KD were blunted when paired with SCRIPT treatment. This blunting of adaptation coincides with KD SCRIPT animals’ inability to produce adequate ketones in the fasted state. This finding implies that ketone production may be a crucial component of maintaining skeletal muscle adaptation and functionality in aged female mice. Even though positive effects of a KD had been demonstrated in old and diseased male mice, whether the diet could benefit females was uncertain. To test the hypothesis that female mice would thrive on a KD, in Chapter 4 middle-aged female mice were put on a two-month isocaloric KD (Table 4.1) and upon collection, skeletal muscle and liver mitochondrial mass and biogenesis, cognitive function, and serum were analyzed. Our results show that a two-month KD is sufficient to improve mitochondrial mass (Figure 4.4A) and PGC-1 protein level in the total and nuclear muscle fractions (Figure 4.4C&E) of skeletal muscle. PGC-1α being responsible for the expression of kynurenine aminotransferases (KAT) in muscle 12, enzymes that convert a potential neurotoxin kynurenine (KYN) into kynurenic acid (KYNA) that cannot enter the brain, therefore we measured KAT protein levels in muscle after a 2-month KD. Two KAT proteins, KAT1 and KAT4 increased in the GTN muscle (Figure 4G&I). The increase in muscle KAT proteins was associated with a significant decrease in circulating KYN and strong trend for KYNA to increase (Figure 4F&G). The increase in KAT protein and decrease in circulating KYN were associated with improvements in Barnes maze and rearing assays, indicating greater spatial learning and memory capacity along with decreased anxiousness (Figure 4J&K). We are the first to propose a mechanism that connects muscle metabolism to brain function in response to a dietary intervention. Within this dissertation, I discuss mechanisms whereby altering acetylation in skeletal muscle improves translational efficiency resulting in greater load-induced skeletal muscle hypertrophy; however, acetylation is not sufficient to induce mitochondrial biogenesis or improve cognitive function.

Musculoskeletal injuries account for the highest percent of time away from work and affect over 100 million people in the United States [1,2]. Ligament and tendon injuries account for approximately 50% of all musculoskeletal injuries [3]. Despite the prevalence of these injuries, current tendon and ligament research is lacking, with little advancement in the treatment of tendinopathies for decades. Tendinopathy describes an injured or diseased tendon/ligament. Underneath this umbrella term, tendinitis is used to describe an injured tendon/ligament that has signs of inflammation. There are many molecular pathways which have been explored in great lengths in muscle and bone; however, the same signals in tendon/ligament have significantly less research devoted to them – including the pro-inflammatory pathways upregulated in tendinitis. In this dissertation, we characterize an engineered ligament model, determine the effects of pro-inflammatory cytokines on engineered ligaments, and explore possible interventions to treat tendinitis such as anti-inflammatories and isometric loading.

The in vitro engineered ligament model used throughout this dissertation was first developed by the Baar lab group [4]. In order to further validate the use of this model, we characterized the effect of passage number on cell gene expression and ligament function, as well as the ligament development over time. Despite previous studies suggesting multiple passages of fibroblasts led to decreased gene expression of typical tenoblast markers, our results found that there was no significant change in gene expression across multiple passages. Furthermore, from passage 4 to passage 11 there was no observed change in mechanical function or matrix composition of engineered ligaments. These results validate the use of this in vitro engineered ligament model using both early and later passage cells. We then characterized the development of engineered ligaments over the course of 5 weeks. The results showed that from Day 7 to Day 14 there was an exponential increase in both collagen content and percent collagen, while these measures began to plateau after Day 14. There also was a linear increase in mechanical function from Day 7 to Day 21. Taken together, this data suggests that Day 7 to Day 14 best represents a developing or regenerating tissue, while after Day 14 the plateau of collagen content best represents a more mature tissue. In total, the work in Chapter 2 demonstrates the efficacy of the model used throughout this dissertation and provides multiple timepoints for interventions based on desired modeling of a regenerating vs mature tissue.

To screen for potential therapeutics for tendinitis, we first sought to develop a model for tendinitis using engineered ligaments. Treating ligaments with pro-inflammatory cytokines TNF-, IL-1, and IL-6 individually led to decreased mechanical function and reduced collagen content. Using a Box-Behnken design of experiments, we developed an optimal combination of all three cytokines to decrease ligament mechanics by 50%. Treatment with this cytokine cocktail led to impaired mechanical function and decreased collagen, resulting in the first multi-cytokine in vitro model for tendinitis. Utilizing this reproducible model, we targeted the nuclear factor kappa-light-chain-enhancer of activated B cells (NF-B) pathway since it had been identified as a potential target for treatment of chronic tendinopathies [5]. Surprisingly, treatment with a known NF-B inhibitor alone actually decreased the mechanical function of ligaments, while treatment alongside the cytokine cocktail had little to no positive effects. We then targeted another proinflammatory pathway – Janus kinase 1 (JAK1) and signal transducer and activator of transcription (STAT)3. Inhibition of JAK1, either alone and in the presences of the cytokine cocktail, resulted in increased mechanical function and increased collagen content in engineered ligaments. Furthermore, the quality of the tissue was improved by JAK1 inhibition, as demonstrated by an increase in enthalpy when undergoing differential scanning calorimetry. In total, the work presented in Chapter 3 resulted in the first multi-cytokine model for tendinitis in an in vitro engineered ligament. Additionally, inhibition of JAK1 has been identified as a possible new treatment for tendinitis.

Isometric loading has previously been reported to improve tendon function, reduce pain, and even reverse the appearance of a central core patellar tendinopathy on MRI [6–8]. Given the positive effect of isometric loading, we tested the combined effects of isometric loading and JAK1 inhibition on engineered ligaments in Chapter 4. Isometric load by itself tended to increase mechanical functional with no significant effect on collagen, resulting in the first successful reproduction of isometric loading effects in this model. Additionally, isometric load alone increased enthalpy which suggests that load improved mechanical function of the tissue through improved matrix organization. Inhibition of JAK1 again improved mechanical function and collagen content by itself. However, there was no interaction effect between load and JAK1 inhibition, instead there were independent positive effects that together led to an additive increase. Interestingly, the combination of isometric loading and JAK1 inhibition did not increase collagen content relative to load alone, suggesting that isometric load may actually prevent an increase in collagen while simultaneously improving collagen fibril organization, diameter, and/or cross-linking. Overall, the combined intervention of isometric load and JAK1 inhibition improve engineered ligament mechanical and material properties. This further supports the possibility of JAK1 inhibition as a novel therapeutic for tendinopathy and may improve upon the previously reported effective isometric loading protocols.

In sum, this dissertation better characterizes and validates the use of engineered human ligaments to make fundamental discoveries concerning tendon biology. Using this model, I developed a novel multi-cytokine treatment to reliably reproduce the negative effects of pro-inflammatory cytokines on ligaments and screen for therapeutics. This resulted in the identification of a new potential treatment for tendinitis – inhibition of JAK1 using a class of drugs that have FDA approval. Ultimately, this work furthers understanding within the field of tendon/ligament physiology and sets the foundation for future in vivo studies to improve treatment of tendinitis.

The ability to move without pain is essential to many occupations and sports. When chronic tendon pain becomes limiting, it is diagnosed as tendinopathy. At a minimum, this tendon pain limits quality of life. For athletes engaged in professional sport, tendinopathy can be devastating and potentially career ending. The exact etiology of tendinopathy is unknown. However, it is especially common in activities that involve repetitive movements and high jerk (typing, jumping, etc.). Since this describes many sports and professions, it should not be surprising that tendon sprains, strains, and pulls are the leading cause of time away from work. The prevalence of tendon injury is highest in professional basketball with ~75% of players having patellar tendon lesions (Benítez-Martínez et al., 2019). Despite the prevalence and high personal and economic costs, there has been no progress in treatments for chronic tendon injuries over the last 25 years. This is in part due to a lack of understanding about basic tendon biology and adaptations that occur with growth, injury, and repetitive loading. This dissertation investigated post-natal development of the Achilles and patellar tendons, developed a rat model of patellar tendinopathy, determined the molecular response of different types (isometric and dynamic) of loads on a tendinopathic tendon, and characterized spatial gene expression in a healthy adult rat patellar tendon. The molecular and mechanical changes that occur during rat post-natal Achilles and patellar tendon development were characterized at P7 (before walking), P14 (shortly after the onset of walking), and P28 (motor maturity). The results are presented in Chapter 2. From P7 to P28, The Achilles increased 3-fold in length, whereas the patellar tendon increased largely in cross-sectional area. Despite the different modes of growth there was a ~10-fold increase in ultimate tensile strength of both the Achilles and patellar tendons. The increase in Achilles length resulted in a greater increase in modulus, whereas the greater cross-sectional area of the patellar meant a larger increase in maximal tensile load. The tendons shared transcriptional similarity at P7 and P14 but diverged at P28. However, there were still many shared processes indicative of tendon post-natal development. These changes included increased extracellular matrix (ECM) gene expression and decreased cell cycle and mitochondrial gene expression. Ribosomal gene expression also significantly decreased in the Achilles tendon and tended to decrease in the patellar tendon. From the genes that were highly expressed at P28 in the patellar tendon, STAT signaling was identified as a downregulated pathway. We hypothesized that these genes contributed to tendon radial growth. STAT inhibition in engineered human ligaments caused an increase collagen content and maximal tensile load (MTL). These results suggest that our developmental transcriptomic data may provide insight into targets to improve tendon function both after injury and during development/training. There are many models of rodent tendon injury; however, no molecular comparison had been made to the human condition. We modeled tendinopathy in the rat using a biopsy punch to remove the central third of the patellar tendon, followed by two weeks of cage activity to allow scar formation. The scar tissue was sequenced with 3’ Tag RNA-seq. Differential gene expression analysis of the healthy versus scar tissues identified an increase in ECM gene expression, decrease in mitochondrial gene expression, and no change in inflammatory pathways. The increase in ECM and decrease in mitochondrial genes meant that the scar was not recapitulating the developmental pattern of gene expression observed in Chapter 2. There was also a visible increase in vascularity and cellularity in the scar tissue. These changes within the tendon are nearly identical to human tendinopathy (Jelinsky et al., 2011). Next, we studied the effect of two distinct types of mechanical load (dynamic and isometric) on gene expression in tendinopathy. The isometric load (4 x 30 s) caused an increased in scleraxis and collagen Ia1 expression relative to the contralateral control leg. In contrast, the time under tension-matched dynamic load (360 x 0.33 s) caused in increase in type II collagen expression relative to the isometric loaded group. These data suggest that isometric loads may be used to treat a tendon injury and that the type of load, rather than the time under tension, is an important consideration when exercising a tendinopathic tendon. The last chapter of the dissertation presents the first complete report of spatial gene expression in tendon. I used the 10X Visium Platform to profile cell type and gene expression location in two adult rat patellar tendon samples. The cell type classification was performed with an anchoring analysis to a single cell RNA sequencing data set from mouse Achilles tendon (de Micheli et al., 2020). There were four cell populations identified: tendon fibroblasts (two main subpopulations), red blood cells, immune cells, and pericytes. The tendon midsubstance contained one tendon fibroblast subpopulation, while the other tendon cell subpopulation was located to the periphery of the main population and within the strip of connective tissue to the side of the tendon. This loose connective tissue also contained red blood cells, immune cells, and pericytes. The top expressed spatially variable genes in both samples were genes with known function in tendon (Col1a1, Dcn, Fmod, Sparc, and Comp). Interestingly, the two most common collagens 1a1 and 3a1 were expressed in separate populations of fibroblasts. I also identified one gene (AABR07000398.1) without a known function that was highly expressed in a pattern opposite to Col1a1. These data are important because they describe the spatial separation of two distinct fibroblast populations and provide the basis for better understanding spatial gene expression following injury. This dissertation provides novel insight into tendon post-natal development and spatial gene expression. It also validates a rodent model of tendinopathy and identifies molecular changes that occur after two different types of exercise on a tendinopathic tendon. Together, these data support a model where force is passed through the center of a healthy tendon and this force is transmitted through a matrix that is maintained by a population of tendon fibroblasts. Following injury, the scarred portion of the tendon is shielded from stress/strain and this may allow the infiltration of the tendon by a second population of fibroblasts. In order to return normal function of the tendon after injury, long holds are necessary to overcome stress shielding. Further, my data suggest that JAK/STAT inhibitors may improve tendon regeneration after injury. These data fill important knowledge gaps in the field of tendon biology, provide evidence for the first drug to treat tendinopathy, and establish the basis for future studies to reverse tendinopathy.