Gene expression differences associated with intrinsic hindfoot muscle loss in the jerboa, Jaculus jaculus

Vertebrate animals that run or jump across sparsely vegetated habitats, such as horses and jerboas, have reduced the number of distal limb bones, and many have lost most or all distal limb muscle. We previously showed that nascent muscles are present in the jerboa hindfoot at birth and that these myofibers are rapidly and completely lost soon after by a process that shares features with pathological skeletal muscle atrophy. Here, we apply an intra- and inter-species approach, comparing jerboa and mouse muscles, to identify gene expression differences associated with the initiation and progression of jerboa hindfoot muscle loss. We show evidence for reduced Hepatocyte Growth Factor (HGF) and Fibroblast Growth Factor (FGF) signaling and an imbalance in nitric oxide signaling; all are pathways that are necessary for skeletal muscle development and regeneration. We also find evidence for phagosome formation, which hints at how myofibers may be removed by autophagy or by non-professional phagocytes without evidence for cell death or immune cell activation. Last, we show significant overlap between genes associated with jerboa hindfoot muscle loss and genes that are differentially expressed in a variety of human muscle pathologies and rodent models of muscle loss disorders. All together, these data provide molecular insight into the mechanism of evolutionary and developmental muscle loss in jerboa hindfeet.


Introduction
Skeletal muscles produce force, pulling on the levers of bone to move the vertebrate body.Since locomotion is diverse across species (e.g., flying, running, jumping, swimming), so too are the sizes, shapes, and numbers of muscles that control bone movements.Many species that run or jump, such as large hooved mammals and saltatorial rodents, have substantially reduced the number of distal limb muscles that are no longer necessary for grasping and climbing.We previously showed that the three-toed jerboa (Jaculus jaculus), a desert adapted bipedal rodent, has lost all intrinsic muscles of the hindfoot over both evolutionary and developmental timescales (Tran et al., 2019).Although newborn jerboas have formed nascent myofibers of a single m.flexor digitorum brevis and three pinnate m. interossei, these myofibers begin to disappear by postnatal day 4 (P4) and are entirely absent in adults (Figure 1A, B).Surprisingly, we found no evidence of apoptotic or necrotic cell death and no stimulation of a local immune response during stages of peak myofiber loss, countering well-supported models of developmental tissue remodeling (Tran et al., 2019).Instead, it appears that the immature contractile apparatus is disassembled in a stereotyped manner with Desmin being the earliest protein to become disorganized.The step-wise disassembly of the sarcomere, which is similar to its orderly disassembly during skeletal muscle atrophy, was associated with upregulation of E3 ubiquitin ligases that are also a hallmark of atrophy, MuRF1 and Atrogin-1.
However, skeletal muscle atrophy is typically considered a pathology associated with disuse, injury, starvation, or disease and typically causes a reduction in the size of individual myofibers but not their number (Moschella and Ontell, 1987).
Here, we implement an intersectional cross-species differential RNA-sequencing approach to broaden our understanding of molecular mechanisms that might be important for initiating and driving the unusual 'atrophy-like' process of muscle loss in the jerboa hindfoot.We use the laboratory mouse (Mus musculus) as a reference species; mice and jerboas diverged from a last common ancestor about 50 million years ago, and mice retained intrinsic hindfoot musculature typical of most other rodents.To account for gene expression divergence over such a long timescale that is likely unrelated to the mechanism of muscle loss in jerboas, we also sequenced RNA extracted from an analogous forelimb muscle that is retained in both species.
By intersecting gene expression differences within and between species at two timepoints, we identified sets of genes associated with the initiation and progression of jerboa hindfoot muscle loss.Among the significantly enriched genetic networks and pathways, we find evidence for lower Hepatocyte Growth Factor (HGF) and Fibroblast Growth Factor (FGF) signaling in jerboa hindfoot muscle than in other muscles that are retained.There is also evidence for an imbalance in the nitric oxide/arginine cycle suggesting lower nitric oxide signaling in jerboa hindfoot muscle.In addition to these pathways, which are known to be critical for muscle development, maintenance, and/or repair, we find evidence for phagosome formation suggesting a mechanism whereby remnants of nascent muscle may be removed either by autophagy or by non-professional phagocytic cells.Finally, we show significant overlap between our dataset and several human muscle degenerative disorders and rodent models of muscle disease lending further support to suggest that evolutionary muscle loss resembles a pathological state.

Sample selection and experimental design
We showed previously that there was no significant difference in the number of myofibers located within the third interosseous muscle between postnatal day 0 (P0, birth) and P2 in either jerboa or mouse (Tran et al., 2019).However, whereas there is a significant increase over two-day intervals from P2 to P8 in mice, there is substantial variance between individual jerboas at P4 and a subsequent decrease until almost all myofibers are lost by P8.
We further showed that the largely nascent and immature structure of the skeletal muscle sarcomere is most similar in the intrinsic hindfoot muscle of mouse and jerboa at P0, preceding degeneration in the jerboa.We therefore chose to isolate and sequence mRNA of the intrinsic hindfoot muscles at P0 and at P3 to capture molecular events at the initiation and during the process of degeneration but prior to tissue loss.
Unlike the larger muscles of the proximal limb, the intrinsic hindfoot muscles are very small making it extremely difficult to manually dissect tissue for transcriptome analysis.We therefore used laser capture microdissection (LCM) to isolate and enrich the intrinsic muscles from sections of P0 and P3 jerboa and mouse hindfeet (Figure 1C-F).To obtain sufficient material for sequencing and analyses, we pooled samples collected from the right and left hindfeet of six individuals for each of three biological replicates of each species and time point.
We then developed an experimental design to identify gene expression differences that might provide molecular evidence in support of a mechanism of muscle loss.As we showed previously for limb growth cartilages (Saxena et al., 2022), direct comparison of the homologous intrinsic hindfoot muscles of jerboa and mouse will identify the plethora of expression differences that accumulated since the two species diverged from their last common ancestor about 50 million years ago, most of which are likely unrelated to muscle loss in jerboas.Yet substantial expression diversity among different healthy skeletal muscles within an individual mouse or rat (Terry et al., 2018) suggests it would also be difficult to rely solely on direct comparison to a 'typical developing' jerboa muscle.Our approach therefore uses both withinspecies and between-species comparisons of jerboa hindfoot muscle that will be lost to muscles that will be retained in order to identify gene expression differences that are robustly associated with muscle loss.
We first sought an analogous forelimb muscle that is retained in both species.The intrinsic muscles of the hand are even smaller than in the hindfoot and thus more difficult to isolate in sufficient quantity.We therefore chose the flexor digitorum superficialis (FDS), which originates in the forelimb autopod during embryogenesis and later translocates to the fetal forearm (Huang et al., 2013).It is therefore evolutionarily and developmentally analogous to the intrinsic hindfoot muscles and also much larger and easy to manually dissect.We extracted mRNA from the FDS of one individual per three biological replicates of stage-matched (P0 and P3) jerboas and mice.We then processed all samples using the Illumina TruSeq Stranded mRNA Library Preparation Kit with polyA-enrichment and sample indexing and sequenced pools of indexed libraries using the Illumina HiSeq 4000 High Output platform.

Differential expression analyses and filtering
For differential expression analyses that compare species transcriptomes directly, it is important to use a strongly supported 1:1 orthologous index of transcripts.We therefore applied TOGA, a method that uses a whole genome alignment to annotate coding genes, identifies (co)orthologous genes, and detects genes with reading frame inactivating mutations (Kirilenko et al., 2023).Using the Mus musculus genome (mm10) as reference and the revised Jaculus jaculus genome (mJacJac1.mat.Y.cur) as query, we annotated 16,667 1:1 orthologous transcripts in the two genomes from which we selected the longest isoform as representative of the gene body.We mapped sequenced reads from each biological replicate to the respective indexed genome.Principal component analysis (PCA) and sample-to-sample distance show segregation between experimental groups first by species and then by muscle type: jerboa hindfoot (jHF), jerboa FDS (jFDS), mouse hindfoot (mHF), mouse FDS (mFDS) (Figure 1G).
We then used DESeq2 to quantify differential expression between the hindfoot and FDS muscles of the jerboa at each stage (intra-species).We also quantified differential expression between jerboa and mouse hindfoot muscles and between jerboa and mouse FDS (interspecies) at each stage accounting for species specific transcript length normalization (Saxena et al., 2022).Statistically significant differentially expressed 1:1 orthologous transcripts in each pairwise analysis are defined as those with a p-adjusted (padj) value less than 0.05 (Supplementary Table 1).We did not apply a fold-change threshold, because genes with different functions (e.g., transcription factors versus enzymes) are likely differentially sensitive to altered expression levels.
To identify gene expression differences between species (inter-species) that are associated with jerboa hindfoot muscle loss, we first selected all genes that are significantly differentially expressed between jerboa and mouse in the hindfoot but not in the FDS (1,632 at P0; Figure 2A).We then plotted the log2 fold-change values for all genes that are significantly differentially expressed between species in both muscles (Figure 2B).The slope of the linear regression is 0.96 (R 2 =0.86), suggesting these genes have expression differences between species that are largely the same in both locations and likely unrelated to muscle loss specific to the jerboa hindfoot.However, 306 genes lie outside the 99% confidence interval; we therefore consider their expression differences between species to be 'disproportionate' in the two muscles.Combining the genes that are differentially expressed at P0 in hindfoot but not FDS with those that are disproportionately differentially expressed in hindfoot compared to FDS gives us 1,938 genes associated with muscle loss after the interspecies comparison.An identical filtering of samples collected from P3 muscles reveals 2,184 significantly differentially expressed genes are associated with muscle loss after the interspecies comparison at this later stage (Figure 3A, B).Intersection of all genes that are differentially expressed between jerboa and mouse hindfoot muscle and between jerboa and mouse FDS.The orange sliver of 'disproportionately differentially expressed' genes lie outside of the 99% confidence interval of the regression of jerboa:mouse FDS versus hindfoot shown in (B).(C) The intersection of interspecies and intraspecies expression differences reveals genes that are differentially expressed in both comparisons.(D) A majority of differential expression correlates in the two comparisons; anti-correlated genes (gray dots) were removed.
We next used the difference in developmental outcome of muscles within jerboas as a 'second pass' filter to identify genes that are also differentially expressed between jerboa hindfoot muscles that will be lost and FDS muscles that are retained (Figure 2C,3C).We found correlations between the inter-and intraspecies expression differences with slope 0.86 (R 2 =0.42) at P0 and 0.84 (R 2 =0.48) at P3 (Figure 2D, 3D).We then selected all genes with consistent expression differences in the same direction in jerboa hindfoot muscle that is lost compared to mouse hindfoot and jerboa FDS muscles that are retained.Altogether, these interand intraspecies analyses identified 1162 genes associated with jerboa hindfoot muscle loss at P0 and 1382 genes at P3 (black dots in Figure 2D and 3D; Supplementary Table 2), which we used for all subsequent candidate gene and network and pathway analyses.Comparing the two timepoints, we find that 749 genes are differentially expressed only at P0, 969 are differentially expressed only at P3, and 413 genes are differentially expressed at both timepoints.Among these that are consistent, all but two differ in the same fold-change direction at both stages (Supplementary Figure 1).

Mechanistic insights from gene expression differences and pathway enrichment analyses
These gene sets provide an opportunity to explore possible mechanisms of evolutionary and developmental muscle loss in the jerboa hindfoot.We first implemented a network and pathway enrichment analysis of all genes associated with jerboa hindfoot muscle loss at P0 and at P3 using Ingenuity Pathway Analysis (IPA, Qiagen).Canonical pathway analysis of wellcharacterized metabolic and cell signaling pathways in IPA showed significant enrichment [log(p-value) >1.3] for 118 pathways at P0 and 32 pathways at P3.The 20 most significantly enriched pathways at each time point are presented in Figure 4, and all significant pathways are in Supplementary Table 3.Here, we focus on a few notable differentially expressed genes and pathways that functionally relate to muscle development, regeneration, and/or maintenance, providing insight into the possible molecular mechanisms of jerboa hindfoot muscle loss.
We previously observed no evidence of cell death by a variety of markers and no macrophages in the vicinity of jerboa hindfoot muscles during degeneration (Tran et al., 2019).It is therefore unclear how nascent myofibers disappear after showing signs of 'atrophy-like' degeneration.Here, we show that IPA calls the 'Phagosome Formation' canonical pathway as significantly enriched at both P0 and P3, and the 'Unfolded Protein Response' pathway as enriched at P3. Absence of evidence for professional phagocytic cells (e.g., macrophages and dendritic cells) in our previous work suggests that phagosomes might form in another cell type.It is possible the enriched phagosome formation pathway reflects myofiber autophagy, whereby muscle cells may degrade and recycle their own damaged proteins (Xia et al., 2021), which could be consistent with the Unfolded Protein Response.Alternatively, phagosomes may form within fibroblasts that we previously observed intermingled with highly degenerating muscle by electron microscopy and immunofluorescence (Tran et al., 2019).If so, this would suggest these are non-professional phagocytic cells that might consume the remains of myofibers.
The 'HGF Signaling' pathway appears to be significantly inhibited (z-score <-2) in jerboa hindfoot muscle at both P0 and P3 based on differential expression of networked molecules.This result stands out as notable due to the well-established importance of the HGF ligand and c-Met receptor to multiple aspects of muscle cell biology.Homozygous c-met loss-of-function mice lack all limb muscle, as well as muscles of the diaphragm and tip of the tongue, due to defective muscle precursor migration (Bladt et al., 1995).In embryonic chickens, exogenous HGF is sufficient to stimulate muscle precursor migration and also prevents myogenic differentiation (Scaal et al., 1999).The importance of HGF/c-Met signaling is not limited to embryogenesis; HGF is released by injured adult muscle and stimulates the c-Met receptor expressed by satellite cells (Tatsumi et al., 1998;Miller et al., 2000).Satellite cells are quiescent muscle stem cells nestled between the muscle and basal lamina, which are activated by HGF signaling to re-enter the cell cycle and become migratory.These cells then fuse to one another to form new myofibers or to injured myofibers for repair.Thus, HGF signaling is also essential to the earliest stages of muscle regeneration after injury.Furthermore, evidence suggests that HGF can inhibit or reverse skeletal muscle atrophy induced by denervation and that cMet inhibition after nerve injury further increases expression of the E3 ubiquitin ligases, Murf1 and Atrogin1 (Choi et al., 2018).Altogether, these findings suggest that evidence for HGF pathway inhibition is consistent with a putative role in the rapid loss of jerboa hindfoot muscle.
'FGF Signaling' is the most significantly enriched pathway at P3 with a z-score trending toward significant inhibition.At least six ligands and two receptors of this highly pleiotropic growth factor pathway are expressed in the skeletal muscle lineage of mouse and/or rat (Hannon et al., 1996;Kästner et al., 2000), though most loss-of-function mice have normal or minimally affected skeletal muscle possibly due to redundancies (Pawlikowski et al., 2017).An exception, Fgf6, appears to be necessary for an early postnatal expansion of the muscle stem cell pool, which may affect muscle regeneration after injury (Floss et al., 1997;Zofkie et al., 2021).Notably, Fgf6 ligand expression in jerboa hindfoot muscle is 4.7-fold lower than in mouse hindfoot and 6.8-fold lower than in jerboa FDS.Although Fgf6 is also significantly differentially expressed in jerboa hindfoot at P0 (3.3-fold and 4.1-fold lower than mouse hindfoot and jerboa FDS, respectively), the FGF signaling pathway is not significantly enriched at this earlier stage.
Considering also the largest fold-change differences in each dataset reveals that Nitric oxide synthase 1 (Nos1/nNos) is expressed a hundred to a thousand-fold lower at P0 in jerboa hindfoot muscle when compared to either mouse hindfoot muscle (log2 fold-change = -10.4;padj = 1.7E-17) or jerboa FDS (log2 fold-change = -7.0;padj = 2.7E-07) (Figure 2; Supplementary Table 2).IPA calls the 'nNos Signaling in Skeletal Muscle Cells' canonical pathway as significantly enriched at P0 (-log10 p-value=1.33).Nitric oxide (NO) is a gaseous molecule with important functions in many tissues (Lundberg and Weitzberg, 2022).In skeletal muscle, NO regulates key aspects of cell biology and physiology, including early stages of myogenesis, muscle force production, metabolism, and repair after muscle injury (Stamler and Meissner, 2001).Nitric oxide is produced by the catalytic activity of Nos1, which converts L-Arginine to NO and L-Citrulline.As evidence of the importance of NO signaling in muscle maintenance and repair, Nos1 activity is reduced in multiple muscle degenerative disorders with a variety of genetic underpinnings (Brenman et al., 1995;Chao et al., 1996;Crosbie et al., 2002).
Furthermore, Nos1 -/-knockout mice have a smaller myofiber cross-sectional area, reduced force production, and show ultrastructural damage to the sarcomere after exercise (De Palma et al., 2014).Interestingly, Argininosuccinate synthase 1 (Ass1) has one of the largest fold-change differences among genes that are expressed higher in jerboa hindfoot muscle at both P0 (log2 fold-change >4.7) and at P3 (log2 fold-change >5.6) when compared to muscles that are retained in each species.Ass1 catalyzes a key step in the biosynthesis of cellular L-Arginine from L-Citrulline, the secondary product of Nos1 activity (Wu and Morris, 1998).Together, this suggests that an imbalance in the nitric oxide/arginine cycle might also contribute to jerboa hindfoot muscle loss.

Comparison of genes associated with evolutionary muscle loss and models of pathological muscle loss
Pathological muscle loss can result from disease causing mutations or in response to denervation, disuse, cancer, fasting, or aging.Are the molecular mechanisms of muscle loss in the jerboa hindfoot broadly similar to pathological muscle loss or similar to a narrower subset of disorders?To answer this question, we selected twenty-four publicly accessible differential expression (RNA-Seq or microarray) datasets that compared human biopsies or mouse or rat models of pathological muscle loss to healthy control skeletal muscle.Datasets were included only if the full list of differentially expressed genes reported were accessible without requiring reanalysis of the raw data.
We first identified differentially expressed genes (p-adj<0.05)within each of the mouse, rat, and human datasets that were assigned the same name (and unique ENMUSG for mouse genes) as in our 1:1 jerboa/mouse orthologous reference set.Three datasets were excluded at this point because fewer than 50 differentially expressed genes remained after filtering for jerboa/mouse orthologs.Using the 16,667 jerboa and mouse 1:1 orthologs as the total number of genes, we performed a Fisher's exact test with Benjamini-Hochberg multiple hypothesis correction to identify significant overlap between each disease/pathology dataset and the sets of genes associated with jerboa hindfoot muscle loss at P0 and at P3 (Table 1).Of the 21 pathology datasets, we found that four overlap significantly with jerboa hindfoot muscle loss at P0, five overlap significantly with jerboa hindfoot muscle loss at P3, and fourteen do not significantly overlap with jerboa hindfoot muscle loss at either developmental stage.
Two pathology models overlap significantly at both timepoints: human critical illness myopathy (CIM) and the mdx mouse model of Duchenne's muscular dystrophy (Llano-Diez et al., 2019;Ralston et al., 2021).Critical illness myopathy (CIM), also known as acute quadriplegic myopathy, is the significant depletion of skeletal muscle mass and compromised performance in individuals receiving intensive care (Latronico et al., 1996;De Jonghe et al., 2002).The underlying mechanisms of CIM are not fully understood but involve processes such as activation of protein degradation pathways, decreased expression of myofibrillar proteins, reduced excitability of cell membranes, mitochondrial dysfunction, and altered excitationcontraction coupling (Shepherd et al., 2017).Duchenne's muscular dystrophy (DMD), on the other hand, is one of the most well-characterized and severe forms of hereditary muscular dystrophy.DMD is caused by mutations in Dystrophin, a large protein component of the complex linking and stabilizing the myofiber cytoskeleton to the extracellular matrix.The mdx mouse model has a spontaneous mutation that prematurely terminates Dystrophin translation (Bulfield et al., 1984;Ryder-Cook et al., 1988;Sicinski et al., 1989), and it is one of the most widely used rodent models of human DMD.At P0 but not P3, we see overlap with a botulinum toxin rat model of atrophy one week after treatment, and with a rat skeletal muscle injury model (Mukund et al., 2014;Ren et al., 2021).Injection of BT was used to inhibit motor neuron activity, thus mimicking conditions of muscle inactivity often seen in multiple neuromuscular disorders or bed-ridden patients (Mukund et al., 2014).The authors performed a long-term study of BT-induced muscle loss from one week to up to a year after BT injection and reported that the most dramatic transcriptome changes (1989 genes) occurred within one week compared to four weeks or longer.The transcriptional differences reported after mechanical injury to the rat tibialis anterior were observed within hours of wounding (Ren et al., 2021).At P3 but not P0, we see significant overlap between genes associated with jerboa hindfoot muscle loss and differentially expressed genes in two independent mouse models of cancer cachexia (Blackwell et al., 2018;Hunt et al., 2021) and a symptomatic model of spinal muscular atrophy (Doktor et al., 2017).Cancer-induced cachexia is a highly complex metabolic syndrome characterized by progressive muscle wasting (Fearon et al., 2012).Notable clinical manifestations of cachexia include loss of weight, inflammation, resistance to insulin, and heightened breakdown of muscle proteins (Fearon et al., 2012).In both mouse models, cancerinduced muscular atrophy was caused by injection of Lewis lung carcinoma (LLC) mouse tumor cells, which lead to the reduction in size of type 2B myofibers with no change in the number of myofibers or the relative distribution of different myofiber types (Hunt et al., 2021).
Spinal muscular atrophy is a neuromuscular disease caused by deficiency of the 'Survival of Motor Neurons' (SMN) protein, which leads to progressive muscle weakness and often causes death in infancy.In a mouse model of severe SMA, animals have a rapid disease progression and a median lifespan of 10 days.Transcriptome analyses at P1 (pre-symptomatic) and P5 (symptomatic) identified hundreds of differentially expressed genes in SMA skeletal muscle compared to control (Doktor et al., 2017).We find significant overlap between our P3 jerboa hindfoot muscle loss dataset and the symptomatic P5 SMA mouse dataset.Together, these intersections lend support to a hypothesis that jerboa hindfoot muscle loss progresses with a gene expression profile similar to pathological atrophy.That only a subset of pathologies overlap with jerboa hindfoot muscle loss is consistent with observations that different causes of skeletal muscle atrophy also have minimal overlap with one another (Hunt et al., 2021).

Conclusions and Limitations
Individual muscles and groups of muscles have been lost repeatedly throughout vertebrate phylogeny as evolution has reshaped the musculoskeletal system to enable a variety of types of locomotion.We previously reported that the cellular mechanisms of intrinsic hindfoot muscle loss in neonatal jerboas, which is histologically similar to intrinsic hindfoot muscle loss in fetal horses (Cunningham, 1883), have atrophy-like characteristics.Here we applied a comparative transcriptomics approach comparing gene expression in jerboa hindfoot muscles to both mouse hindfoot muscles (inter-species) and to jerboa forelimb muscle (intra-species), which are retained to adulthood.Genes we identified with consistent expression differences (same fold-change direction) in these multi-way comparisons are therefore 'associated with jerboa hindfoot muscle loss'.The complete datasets are available in Supplementary Tables 1     and 2.
Correlation of gene expression differences between inter-and intraspecies analyses and between stages of muscle loss (Supplementary Figure 1) supports the logic of our experimental design, but we acknowledge that the datasets are likely incomplete and include false positives.
Furthermore, these are snapshots of gene expression differences that include genes that may be causative and others that are certainly a consequence of the primary mechanism of muscle loss due to the interconnected effects of expression perturbation (Cowen et al., 2017).
Nevertheless, these datasets provide evidence for molecular mechanisms of evolutionary loss of distal limb skeletal muscles.
Although application of pathway enrichment analyses to these datasets is limited by the current knowledge base of gene functions, it does provide valuable insight into the putative molecular mechanisms of a biological phenomenon.Here, we have drawn attention to a few enriched pathways (HGF, FGF, and NO) with well-documented importance to muscle development, maintenance, and/or repair; all enriched pathways are available in Supplementary Table 3.As with all such 'omics' datasets, it will be important to repeat these analyses as the functional annotation of all genes continues to expand, thus enabling broader and deeper insight into putative molecular mechanisms of a variety of biological processes.
Finally, we set out to determine if these differential expression analyses would provide further support for a hypothesis that the mechanism of jerboa hindfoot muscle loss, occurring over both evolutionary and developmental timescales, shares substantial similarities with instances of pathological muscle loss.At one or both stages of jerboa hindfoot muscle loss, we found significant overlap with seven out of 21 of the analyzed disease and injury states, consistent with our prior histological and ultrastructural observations.However, we emphasize that statistically significant overlap is evidence for correlation of the gene expression differences observed in jerboa hindfoot muscle and certain muscle pathologies, but this should not be interpreted to suggest that the causes are the same.Rather, these correlations with specific pathology states demonstrate striking and perhaps surprising similarity between evolutionary and pathological states while also serving as entry points to gain further insight into the mechanisms of muscle loss in a variety of contexts.

Figure 1 :
Figure 1: Sample morphology and preparation for differential mRNA expression analyses (A, B) Transverse section through the mid-foot of adult mouse and jerboa hindfeet with immunofluorescent detection of pro-Collagen I (tendon, green) and skeletal muscle Myosin Heavy Chain (magenta).(C-F) Representative toluidine blue-stained plantar sections of mouse (mHF) or jerboa hindfoot (jHF) at P0 that are intact (C, E) or after laser capture microdissection (LCM) of intrinsic hindfoot muscle (D, F). (G) Principal components analysis of all jerboa and mouse hindfoot and flexor digitorum superficialis (FDS) transcriptomes.

Figure 2 :
Figure 2: Identification of genes associated with the initiation of jerboa hindfoot muscle loss at P0 (A)Intersection of all genes that are differentially expressed between jerboa and mouse hindfoot muscle and between jerboa and mouse FDS.The orange sliver of 'disproportionately differentially expressed' genes lie outside of the 99% confidence interval of the regression of jerboa:mouse FDS versus hindfoot shown in (B).(C) The intersection of interspecies and intraspecies expression differences reveals genes that are differentially expressed in both comparisons.(D) A majority of differential expression correlates in the two comparisons; anti-correlated genes (gray dots) were removed.

Figure 3 :
Figure3: Identification of genes associated with the progression of jerboa hindfoot muscle loss at P3 (A) Intersection of all genes that are differentially expressed between jerboa and mouse hindfoot muscle and between jerboa and mouse FDS.The orange sliver of 'disproportionately differentially expressed' genes lie outside of the 99% confidence interval of the regression of jerboa:mouse FDS versus hindfoot shown in (B).(C) The intersection of interspecies and intraspecies expression differences reveals genes that are differentially expressed in both comparisons.(D) A majority of differential expression correlates in the two comparisons; anticorrelated genes (gray dots) were removed.

Figure 4 :
Figure 4: The top twenty most significantly enriched canonical pathways among genes associated with jerboa hindfoot muscle loss at P0 (A) and at P3 (B).Vertical dashed lines mark the threshold for significance [-log(0.05)=1.3].Asterisks mark pathways that reach significance for 'activation' (z-score >2) or inhibition (z-score<-2) reported within the Ingenuity Pathway Analysis.

Table 1 : Values for the Fisher's Exact Test of overlap between genes associated with jerboa hindfoot muscle loss at P0 or at P3 and datasets obtained from rodent disease models or human pathologies
. Significant values are bold and highlighted.