During endochondral ossification, chondrocytes undergo maturation and biochemically modify the collagenous extracellular matrix of cartilage. Similar modifications to cartilage proteoglycans (PGs), which are predominantly chondroitin sulfate PGs, have not been characterized. Using synchrotron X-ray fluorescence imaging, we demonstrated that PG sulfation significantly decreased during cartilage maturation of chick embryos. Laser-capture microdissection and RNAseq revealed upregulation of Arylsulfatase I (Arsi) in mature cartilage of mouse. ARSI protein also increased in mature cartilage of mouse and chick in vivo and during maturation of ATDC5 chondrocytes in vitro, whereas expression of the two known chondroitin sulfate PG sulfatases (ARSB and GALNS) was not specific to mature cartilage. Colocalization studies suggested that ARSI is lysosomal, and lysosome homeostasis was altered in ARSI loss of function chondrocytes. Biochemical analyses of ARSI gain and loss of function cell lines and isolated cell-free systems revealed that ARSI is a novel chondroitin endosulfatase, specifically desulfating chondroitin-4-sulfate at pH 4.5. Finally, Arsi knockout in RCS chondrocytes caused increased expression of maturation genes, such as Col10a1 and Mmp13. In total, these data identify ARSI as a novel PG sulfatase regulating endochondral ossification.

The role of proteoglycans (PGs) in regulating growth factor signalling during endochondral ossification, or formation of bone around a cartilage template, is established. However, whether PGs regulate other skeletal processes, like intramembranous ossification or fin regeneration, has not been studied extensively. Given that endochondral ossification in PG-deficient fam20b mutant zebrafish is altered due to increased Bmp signalling, we hypothesized that PGs also normally inhibit Bmp signalling during intramembranous ossification and fin regeneration. To test this hypothesis, the functional relevance of Bmp signalling during craniofacial dermal bone formation and caudal fin regeneration was examined in wild type and fam20b mutant zebrafish. Bmp responsiveness of cells around wild-type dermal bones was verified by phospho-Smad1/5/9 immunoreactivity. Dermal bones in fam20b mutants were generally unaffected, apart from early formation of the dentary and quadrate, which actually initiate in the perichondrium of Meckel's cartilage and the palatoquadrate, respectively. Treatment of wild-type or fam20b-/- embryos with the Bmp inhibitor DMH1 did not yield clear differences in the size or morphology of most dermal bones, but did rescue early dentary and quadrate formation in fam20b mutants. During adult fin regeneration, Bmp signalling was confirmed by careful temporal analyses of GFP expression of Tg(5xBmpRE-Xla.Id3:GFP)ir1189 and Tg(BmpRE:EGFP)pt510 zebrafish, each expressed in different tissue compartments of the regenerating fin. fam20b-/- adults did not show fin regeneration defects. Treatment with DMH1, however, significantly reduced outgrowth of regenerating fins in distinct anatomical regions (the most dorsal and ventral principal rays) in adult zebrafish. Bony ray differentiation also was inhibited by DMH1 treatment. In total, these data do not support a major role of Bmp signalling or fam20b-dependent PGs in craniofacial intramembranous ossification, but do show that Bmp signalling is required for outgrowth of regenerating fins.

Specific character traits of mineralized endoskeletal tissues need to be clearly defined and comprehensively examined among extant chondrichthyans (elasmobranchs, such as sharks and skates, and holocephalans, such as chimaeras) to understand their evolution. For example, tiles of mineralized polygonal structures called tesserae occur at cartilage surfaces in chondrichthyans, but recent studies showing trabecular mineralization at elasmobranch cartilage surfaces suggest that tesserae are not as common as previously thought. Also, while areolar mineralized tissue in elasmobranchs is generally considered a unique, shared chondrichthyan feature, some chondrichthyan species demonstrate bone-like tissues in both a specific region of tesserae termed the cap zone and continuous (not tiled) mineralized neural arches. To clarify the distribution of specific endoskeletal features among extant chondrichthyans, adult skeletal tissues in a holocephalan chimaera (spotted ratfish) and two elasmobranchs (small-spotted catshark and little skate) were characterized using synchrotron radiation and desktop micro-CT imaging, and histological and immunofluorescent assays. Endoskeletal mineralization in the ratfish, catshark, and little skate varied both quantitively in tissue mineral density (TMD) and qualitatively in the morphology and localization of mineralized structures and tissues. For example, TMD of several skeletal elements was significantly lower in ratfish, compared to catshark and little skate. Trabecular and areolar mineralization were shared among these extant chondrichthyan species, but tesserae and bone-like tissues were not. Interestingly, three separate analyses argued that the adult chimaera endoskeleton has features of the embryonic little skate endoskeleton. Generally, this study proposes specific terminology for character states of the extant chondrichthyan endoskeleton and infers those states in ancestral chondrichthyans with reference to fossil data.

The mechanisms by which crucial transcription factors of a gene regulatory network (GRN) interact continue to be revealed. In the vertebrate skeleton, SOX9 and RUNX2 combine to specify three different cell types. Sox9 drives immature chondrocyte differentiation, Runx2 regulates osteoblast differentiation, and both Sox9 and Runx2 are somehow required for mature chondrocyte formation. To elucidate mechanisms of GRN regulation in mature chondrocytes, transcriptomic data were examined from all three skeletal cell types isolated by laser capture microdissection of embryonic mouse. Multiple bioinformatic analyses supported the hypothesis that SOX9 and RUNX2 operate two cross-inhibitory subnetworks of the skeletal cell GRN during immature chondrocyte and osteoblast formation, but mature chondrocyte differentiation involves cooperation between these subnetworks. Several mature chondrocyte gene clusters had expression levels that represented an averaging of SOX9 and RUNX2 subnetworks, while one cluster, containing the hallmark mature chondrocyte genes collagen type 10a1 and Indian hedgehog, suggested a synergistic interaction between subnetworks. Generally, this in vivo LCM-RNA-seq approach enabled new understanding of interactions between distinct GRN subnetworks during cell differentiation and can similarly reveal regulatory control of any developmental process.

Adolescent idiopathic scoliosis (AIS) is the most prevalent pediatric spine disorder, developing in the absence of obvious physiological defects. Genome sequencing and functional studies have demonstrated association of musculoskeletal collagen variants and cartilaginous extracellular matrix (ECM) defects in a subset of patients. However, the underlying biological causes of AIS remain poorly understood, limiting treatment options. Using multiple zebrafish AIS models, we demonstrate that reduction-oxidation (redox) imbalances induce cell stress and collagen remodelling within intervertebral segments of the developing spine. Mutant spines are consequently stiffer, as measured by shear wave elastography, and exhibit deformations of intervertebral structures. Remarkably, elevated stiffness and intervertebral ECM phenotypes are detectable prior to scoliosis onset, suggesting a causal relationship, and can be suppressed by antioxidant treatment. Together, our preclinical studies implicate oxidative stress-induced intervertebral deformations in the pathogenesis of AIS and identify elevated spine stiffness and redox imbalance as plausible first-in-kind prognostic biomarkers and therapeutic targets.

During endochondral ossification, chondrocytes secrete a proteoglycan (PG)-rich extracellular matrix that can inhibit the process of cartilage maturation, including expression of Ihh and Col10a1. Because bone morphogenetic proteins (BMPs) can promote cartilage maturation, we hypothesized that cartilage PGs normally inhibit BMP signalling. Accordingly, BMP signalling was evaluated in chondrocytes of wild-type and PG mutant (fam20b-/-) zebrafish and inhibited with temporal control using the drug DMH1 or an inducible dominant-negative BMP receptor transgene (dnBMPR). Compared with wild type, phospho-Smad1/5/9, but not phospho-p38, was increased in fam20b-/- chondrocytes, but only after they secreted PGs. Phospho-Smad1/5/9 was decreased in DMH1-treated or dnBMPR-activated wild-type chondrocytes, and DMH1 also decreased phospho-p38 levels. ihha and col10a1a were decreased in DMH1-treated or dnBMPR-activated chondrocytes, and less perichondral bone formed. Finally, early ihha and col10a1a expression and early perichondral bone formation of fam20b mutants were rescued with DMH1 treatment or dnBMPR activation. Therefore, PG inhibition of canonical BMP-dependent cartilage maturation delays endochondral ossification, and these results offer hope for the development of growth factor therapies for skeletal defects of PG diseases.

Cartilage tissue engineering (CTE) with the help of engineered constructs has shown promise for the regeneration of hyaline cartilage, where fibrocartilage may also be formed due to the biomechanical loading resulting from the host weight and movement. Previous studies have primarily reported on hyaline cartilage formationin vitroand/or in small animals, while leaving the fibrocartilage formation undiscovered. In this paper, we, at the first time, present a comparison study on hyaline cartilage versus fibrocartilage formation in a large animal model of pig by using two constructs (namely hydrogel and hybrid ones) engineered by means of three-dimensional (3D) bioprinting. Both hydrogel and hybrid constructs were printed from the bioink of alginate (2.5%) and ATDC5 cells (chondrogenic cells at a cell density of 5 × 106cells ml-1), with the difference in that in the hybrid construct, there was a polycaprolactone (PCL) strand printed between every two bioink strands, which were strategically designed to shield the force imposed on the cells within the bioink strands. Both hydrogel and hybrid constructs were implanted into the chondral defects created in the articular cartilage of weight-bearing portions of pig stifle joints; the cartilage formation was examined at one- and three-months post-implantation, respectively, by means of Safranin O, Trichrome, immunofluorescent staining, and synchrotron radiation-based (SR) inline phase contrast imaging microcomputed tomography (inline-PCI-CT). Glycosaminoglycan (GAG) and collagen type II (Col II) secretion were used to evaluate the hyaline cartilage formation, while collagen type I (Col I) was used to indicate fibrocartilage given that Col I is low in hyaline cartilage but high in fibrocartilage. Our results revealed that cartilage formation was enhanced over time in both hydrogel and hybrid constructs; particularly, the hydrogel construct exhibited more cartilage formation at both one- and three-months post-implantation, while hybrid constructs tended to have less fibrocartilage formed in a long time period. Also, the result from the inline-PCI-CT revealed that the inline-PCI-CT was able to provide not only the information seen in other histology images, but also high-resolution details of biomaterials and regenerating cartilage. This would represent a significant advance toward the non-invasive assessment of cartilage formation regeneration within large animal models and eventually in human patients.

Hydrogels show promise in cartilage tissue engineering (CTE) by supporting chondrocytes and maintaining their phenotype and extracellular matrix (ECM) production. Under prolonged mechanical forces, however, hydrogels can be structurally unstable, leading to cell and ECM loss. Furthermore, long periods of mechanical loading might alter the production of cartilage ECM molecules, including glycosaminoglycans (GAGs) and collagen type 2 (Col2), specifically with the negative effect of stimulating fibrocartilage, typified by collagen type 1 (Col1) secretion. Reinforcing hydrogels with 3D-printed Polycaprolactone (PCL) structures offer a solution to enhance the structural integrity and mechanical response of impregnated chondrocytes. This study aimed to assess the impact of compression duration and PCL reinforcement on the performance of chondrocytes impregnated with hydrogel. Results showed that shorter loading periods did not significantly affect cell numbers and ECM production in 3D-bioprinted hydrogels, but longer periods tended to reduce cell numbers and ECM compared to unloaded conditions. PCL reinforcement enhanced cell numbers under mechanical compression compared to unreinforced hydrogels. However, the reinforced constructs seemed to produce more fibrocartilage-like, Col1-positive ECM. These findings suggest that reinforced hydrogel constructs hold potential for in vivo cartilage regeneration and defect treatment by retaining higher cell numbers and ECM content. To further enhance hyaline cartilage ECM formation, future studies should focus on adjusting the mechanical properties of reinforced constructs and exploring mechanotransduction pathways.

BACKGROUND:Mouse, chick, and zebrafish undergo a highly conserved program of cartilage maturation during endochondral ossification (bone formation via a cartilage template). Standard histological and molecular features of cartilage maturation are chondrocyte hypertrophy, downregulation of the chondrogenic markers Sox9 and Col2a1, and upregulation of Col10a1. We tested whether cartilage maturation is conserved in an amphibian, the western clawed frog Xenopus tropicalis, using in situ hybridization for standard markers and a novel laser-capture microdissection RNAseq data set. We also functionally tested whether thyroid hormone drives cartilage maturation in X tropicalis, as it does in other vertebrates.RESULTS:The developing frog humerus mostly followed the standard progression of cartilage maturation. Chondrocytes gradually became hypertrophic as col2a1 and sox9 were eventually down-regulated, but col10a1 was not up-regulated. However, the expression levels of several genes associated with the early formation of cartilage, such as acan, sox5, and col9a2, remained highly expressed even as humeral chondrocytes matured. Greater deviances were observed in head cartilages, including the ceratohyal, which underwent hypertrophy within hours of becoming cartilaginous, maintained relatively high levels of col2a1 and sox9, and lacked col10a1 expression. Interestingly, treating frog larvae with thyroid hormone antagonists did not specifically reduce head cartilage hypertrophy, resulting rather in a global developmental delay.CONCLUSION:These data reveal that basic cartilage maturation features in the head, and to a lesser extent in the limb, are not conserved in X tropicalis. Future work revealing how frogs deviate from the standard cartilage maturation program might shed light on both evolutionary and health studies.

AbstractScaffolds play a key role in bone tissue engineering (BTE) as they provide a mechanically and biologically supportive template to treat bone defects. Yet, the ideal scaffold structures are far from certain, leaving a lot to be discovered in terms of the scaffold structure–performance relationships. In this study, we investigated the influence of pore size and internal structure on osteoblast performance in vitro. Three‐dimensional (3D) scaffolds were printed from polycaprolactone (PCL) reinforced with 30% (wt.) nano‐hydroxyapatite (nHAp), with two different internal structures (lattice and staggered) and four pore sizes (0.280, 0.380, 0.420, and 0.550 mm). Scaffolds were seeded with pre‐osteoblast cells (MC3T3‐E1). Metabolic activity of cells, osteoblast differentiation, and capability of osteoblasts to deposit mineralized matrix were examined in vitro. Staggered scaffolds better supported cellular performance. The pore size of 0.280 mm was more favorable to support cell proliferation while the pore size ≥0.420 mm was more effective to promote osteoblast differentiation and mineralization. Findings revealed that osteoblast activities were affected differently by the pore size. Our study further suggests that the structure with a gradient pore size would be better than a single size in terms of supporting cell proliferation, differentiation, and secretion of a mineralized matrix.

The goal of cartilage tissue engineering (CTE) is to regenerate new hyaline cartilage in joints and treat osteoarthritis (OA) using cell-impregnated hydrogel constructs. However, the production of an extracellular matrix (ECM) made of fibrocartilage is a potential outcome within hydrogel constructs when in vivo. Unfortunately, this fibrocartilage ECM has inferior biological and mechanical properties when compared to native hyaline cartilage. It was hypothesized that compressive forces stimulate fibrocartilage development by increasing production of collagen type 1 (Col1), an ECM protein found in fibrocartilage. To test the hypothesis, 3-dimensional (3D)-bioprinted hydrogel constructs were fabricated from alginate hydrogel impregnated with ATDC5 cells (a chondrogenic cell line). A bioreactor was used to simulate different in vivo joint movements by varying the magnitude of compressive strains and compare them with a control group that was not loaded. Chondrogenic differentiation of the cells in loaded and unloaded conditions was confirmed by deposition of cartilage specific molecules including glycosaminoglycans (GAGs) and collagen type 2 (Col2). By performing biochemical assays, the production of GAGs and total collagen was also confirmed, and their contents were quantitated in unloaded and loaded conditions. Furthermore, Col1 vs. Col2 depositions were assessed at different compressive strains, and hyaline-like cartilage vs. fibrocartilage-like ECM production was analyzed to investigate how applied compressive strain affects the type of cartilage formed. These assessments showed that fibrocartilage-like ECM production tended to reduce with increasing compressive strain, though its production peaked at a higher compressive strain. According to these results, the magnitude of applied compressive strain governs the production of hyaline-like cartilage vs. fibrocartilage-like ECM and a high compressive strain stimulates fibrocartilage-like ECM formation rather than hyaline cartilage, which needs to be addressed by CTE approaches.

Accurate evaluation of morphological changes in articular cartilage are necessary for early detection of osteoarthritis (OA). 3T magnetic resonance imaging (MRI) has highly sensitive contrast resolution and is widely used clinically to detect OA. However, synchrotron radiation phase-contrast imaging computed tomography (SR-PCI) can also provide contrast to tissue interfaces that do not have sufficient absorption differences, with the added benefit of very high spatial resolution. Here, MRI was compared with SR-PCI for quantitative evaluation of human articular cartilage. Medial tibial condyles were harvested from non-OA donors and from OA patients receiving knee replacement surgery. Both imaging methods revealed that average cartilage thickness and cartilage volume were significantly reduced in the OA group, compared to the non-OA group. When comparing modalities, the superior resolution of SR-PCI enabled more precise mapping of the cartilage surface relative to MRI. As a result, MRI showed significantly higher average cartilage thickness and cartilage volume, compared to SR-PCI. These data highlight the potential for high-resolution imaging of articular cartilage using SR-PCI as a solution for early OA diagnosis. Recognizing current limitations of using a synchrotron for clinical imaging, we discuss its nascent utility for preclinical models, particularly longitudinal studies of live animal models of OA.

Ancestors of the Antarctic icefishes (family Channichthyidae) were benthic and had no swim bladder, making it energetically expensive to rise from the ocean floor. To exploit the water column, benthopelagic icefishes were hypothesized to have evolved a skeleton with "reduced bone," which gross anatomical data supported. Here, we tested the hypothesis that changes to icefish bones also occurred below the level of gross anatomy. Histology and micro-CT imaging of representative craniofacial bones (i.e., ceratohyal, frontal, dentary, and articular) of extant Antarctic fish species specifically evaluated two features that might cause the appearance of "reduced bone": bone microstructure (e.g., bone volume fraction and structure linear density) and bone mineral density (BMD, or mass of mineral per volume of bone). Measures of bone microstructure were not consistently different in bones from the icefishes Chaenocephalus aceratus and Champsocephalus gunnari, compared to the related benthic notothenioids Notothenia coriiceps and Gobionotothen gibberifrons. Some quantitative measures, such as bone volume fraction and structure linear density, were significantly increased in some icefish bones compared to homologous bones of non-icefish. However, such differences were rare, and no microstructural measures were consistently different in icefishes across all bones and species analyzed. Furthermore, BMD was similar among homologous bones of icefish and non-icefish Antarctic notothenioids. In summary, "reduced bone" in icefishes was not due to systemic changes in bone microstructure or BMD, raising the prospect that "reduced bone" in icefish occurs only at the gross anatomic level (i.e., smaller or fewer bones). Given that icefishes exhibit delayed skeletal development compared to non-icefish Antarctic fishes, combining these phenotypic data with genomic data might clarify genetic changes driving skeletal heterochrony.

Chondrocytes that are impregnated within hydrogel constructs sense applied mechanical force and can respond by expressing collagens, which are deposited into the extracellular matrix (ECM). The intention of most cartilage tissue engineering is to form hyaline cartilage, but if mechanical stimulation pushes the ratio of collagen type I (Col1) to collagen type II (Col2) in the ECM too high, then fibrocartilage can form instead. With a focus on Col1 and Col2 expression, the first part of this article reviews the latest studies on hyaline cartilage regeneration within hydrogel constructs that are subjected to compression forces (one of the major types of the forces within joints) in vitro . Since the mechanical loading conditions involving compression and other forces in joints are difficult to reproduce in vitro , implantation of hydrogel constructs in vivo is also reviewed, again with a focus on Col1 and Col2 production within the newly formed cartilage. Furthermore, mechanotransduction pathways that may be related to the expression of Col1 and Col2 within chondrocytes are reviewed and examined. Also, two recently-emerged, novel approaches of load-shielding and synchrotron radiation (SR)–based imaging techniques are discussed and highlighted for future applications to the regeneration of hyaline cartilage. Going forward, all cartilage tissue engineering experiments should assess thoroughly whether fibrocartilage or hyaline cartilage is formed.

During embryogenesis, skeletal development is tightly regulated by locally secreted growth factors that interact with proteoglycans (PGs) in the extracellular matrix (ECM). Bone morphogenetic proteins (BMPs) are multifunctional growth factors that play critical roles in cartilage maturation and bone formation. BMP signals are transduced from plasma membrane receptors to the nucleus through both canonical Smad and noncanonical p38 mitogen-activated protein kinase (MAPK) pathways. BMP signalling is modulated by a variety of endogenous and exogenous molecular mechanisms at different spatiotemporal levels and in both positive and negative manners. As an endogenous example, BMPs undergo extracellular regulation by PGs, which generally regulate the efficiency of ligand-receptor binding. BMP signalling can also be exogenously perturbed by a group of small molecule antagonists, such as dorsomorphin and its derivatives, that selectively bind to and inhibit the intracellular kinase domain of BMP type I receptors. In this review, we present a current understanding of BMPs and PGs functions in cartilage maturation and osteoblast differentiation, highlighting BMP–PG interactions. We also discuss the identification of highly selective small-molecule BMP receptor type I inhibitors. This review aims to shed light on the importance of BMP signalling and PGs in cartilage maturation and bone formation.

Simulated microgravity (SMG) inhibits osteoblast differentiation (OBD) and induces bone loss via the inhibition of the Wnt/β-catenin pathway. However, the mechanism by which SMG alters the Wnt/β-catenin pathway is unknown. We previously demonstrated that SMG altered the focal adhesion kinase (FAK)-regulated mTORC1, AMPK and ERK1/2 pathways, leading to the inhibition of tumor cell proliferation/metastasis and promoting cell apoptosis. To examine whether FAK similarly mediates SMG-dependent changes to Wnt/β-catenin in osteoblasts, we characterized mouse MC3T3-E1 cells cultured under clinostat-modeled SMG (µg) conditions. Compared to cells cultured under ground (1 g) conditions, SMG reduces focal adhesions, alters cytoskeleton structures, and down-regulates FAK, Wnt/β-catenin and Wnt/β-catenin-regulated molecules. Consequently, protein-2 (BMP2), type-1 collagen (COL1), alkaline-phosphatase activity and matrix mineralization are all inhibited. In the mouse hindlimb unloading (HU) model, SMG-affected tibial trabecular bone loss is significantly reduced, according to histological and micro-computed tomography analyses. Interestingly, the FAK activator, cytotoxic necrotizing factor-1 (CNF1), significantly suppresses all of the SMG-induced alterations in MC3T3-E1 cells and the HU model. Therefore, our data demonstrate the critical role of FAK in the SMG-induced inhibition of OBD and bone loss via the Wnt/β-catenin pathway, offering FAK signaling as a new therapeutic target not only for astronauts at risk of OBD inhibition and bone loss, but also osteoporotic patients.

To explain how cartilage appeared in different parts of the vertebrate body at discrete times during evolution, we hypothesize that different embryonic populations co-opted expression of a core gene regulatory network (GRN) driving chondrocyte differentiation. To test this hypothesis, laser-capture microdissection coupled with RNA-seq was used to reveal chondrocyte transcriptomes in the developing chick humerus and ceratobranchial, which are mesoderm- and neural crest-derived, respectively. During endochondral ossification, two general types of chondrocytes differentiate. Immature chondrocytes (IMM) represent the early stages of cartilage differentiation, while mature chondrocytes (MAT) undergo additional stages of differentiation, including hypertrophy and stimulating matrix mineralization and degradation. Venn diagram analyses generally revealed a high degree of conservation between chondrocyte transcriptomes of the limb and head, including SOX9, COL2A1, and ACAN expression. Typical maturation genes, such as COL10A1, IBSP, and SPP1, were upregulated in MAT compared to IMM in both limb and head chondrocytes. Gene co-expression network (GCN) analyses of limb and head chondrocyte transcriptomes estimated the core GRN governing cartilage differentiation. Two discrete portions of the GCN contained genes that were differentially expressed in limb or head chondrocytes, but these genes were enriched for biological processes related to limb/forelimb morphogenesis or neural crest-dependent processes, respectively, perhaps simply reflecting the embryonic origin of the cells. A core GRN driving cartilage differentiation in limb and head was revealed that included typical chondrocyte differentiation and maturation markers, as well as putative novel “chondrocyte” genes. Conservation of a core transcriptional program during chondrocyte differentiation in both the limb and head suggest that the same core GRN was co-opted when cartilage appeared in different regions of the skeleton during vertebrate evolution.

Similarities and differences in the associations of biological entities among species can provide us with a better understanding of evolutionary relationships. Often the evolution of new phenotypes results from changes to interactions in pre-existing biological networks and comparing networks across species can identify evidence of conservation or adaptation. Gene co-expression networks (GCNs), constructed from high-throughput gene expression data, can be used to understand evolution and the rise of new phenotypes. The increasing abundance of gene expression data makes GCNs a valuable tool for the study of evolution in non-model organisms. In this paper, we cover motivations for why comparing these networks across species can be valuable for the study of evolution. We also review techniques for comparing GCNs in the context of evolution, including local and global methods of graph alignment. While some protein-protein interaction (PPI) bioinformatic methods can be used to compare co-expression networks, they often disregard highly relevant properties, including the existence of continuous and negative values for edge weights. Also, the lack of comparative datasets in non-model organisms has hindered the study of evolution using PPI networks. We also discuss limitations and challenges associated with cross-species comparison using GCNs, and provide suggestions for utilizing co-expression network alignments as an indispensable tool for evolutionary studies going forward.

BACKGROUND:Gene co-expression networks (GCNs) are not easily comparable due to their complex structure. In this paper, we propose a tool, Juxtapose, together with similarity measures that can be utilized for comparative transcriptomics between a set of organisms. While we focus on its application to comparing co-expression networks across species in evolutionary studies, Juxtapose is also generalizable to co-expression network comparisons across tissues or conditions within the same species.METHODS:A word embedding strategy commonly used in natural language processing was utilized in order to generate gene embeddings based on walks made throughout the GCNs. Juxtapose was evaluated based on its ability to embed the nodes of synthetic structures in the networks consistently while also generating biologically informative results. Evaluation of the techniques proposed in this research utilized RNA-seq datasets from GTEx, a multi-species experiment of prefrontal cortex samples from the Gene Expression Omnibus, as well as synthesized datasets. Biological evaluation was performed using gene set enrichment analysis and known gene relationships in literature.RESULTS:We show that Juxtapose is capable of globally aligning synthesized networks as well as identifying areas that are conserved in real gene co-expression networks without reliance on external biological information. Furthermore, output from a matching algorithm that uses cosine distance between GCN embeddings is shown to be an informative measure of similarity that reflects the amount of topological similarity between networks.CONCLUSIONS:Juxtapose can be used to align GCNs without relying on known biological similarities and enables post-hoc analyses using biological parameters, such as orthology of genes, or conserved or variable pathways.AVAILABILITY:A development version of the software used in this paper is available at https://github.com/klovens/juxtapose.