Introduction

Adolescents from 10 to 19 years of age are in a transition from childhood to adulthood, and are at higher risk of non-intentional injury. According to the CDC childhood injury report, more than 9.2 million injured are treated in hospital emergency rooms, and more than 12,000 injured people under 19-year-old die each year. Burn is one of the leading causes of unintentional injury leading to higher morbidity and mortality among children in the US and worldwide. The proportion of pediatric admissions among total patients with burn injury increased by 63.9% from 2003 to 2012 at United States1. Meanwhile, the WHO investigated burn incidence in children and reported 96,000 lethal burn injuries globally in 20042. These death rates are the second-highest between 15 and 19 years. old children after infants. Worldwide, an increasing proportion of burn injuries was recorded in adolescent females2. Along with the improvement of care management, current mortality has been reduced to less than 3% in pediatric burn patients3. However, severe burn has significant impact and consequences in surviving children, including both physical disfiguration and mental in long-term effects. Burn impairs growth in children4, and has other effects after injury5.

The mechanisms of burn-induced muscle wasting are mostly described within the hypermetabolic response to severe burn, significantly contributing to sequelae after injury. Various treatments were developed to moderate the severity of this pathophysiological and metabolic response6. At the transcriptional level, Padfield et al. showed that groups of genes, including those regulating myogenesis and mitochondrial function, contributed to muscle wasting in a mouse model with a single limb burn7. Only a few studies further revealed gene expression changes in pediatric burns during a long-term period of recovery8.

As key gene regulators, microRNAs (miRNAs) inhibit gene expression at the translational level9. In burns, Liang et al. demonstrated miRNA profiles in cells within the burned dermis and found that 66 miRNAs were significantly altered10. We found miRNAs to mediate gene pathway changes in skeletal muscle after burn11. However, the data collected from most studies (including ours) were based on adult animal models, which may not reflect responses in adolescents and young subjects, and can lead to a bias in data interpretation when applied to pediatric patient populations.

The biological profile in adolescents (< 19 years old) is unique and different from adults12. Related to burns, children have more skin surface area in the head and less in the legs generally13. Also, energy metabolism differs among children and adults. Total energy utilization is 60–80 kcal/kg/day for a 10-year-old child and 30–40 kcal/kg/day for a 20-year-old adult14. At the cellular level, muscle progenitor satellite cells, a major resource for cell hemostasis, are depleted with age15. Genetic regulations also differ in children and adults. A study of 299 adults demonstrated dopamine receptor genes DRD2 are associated with weight gain in adults, with α-ketoglutarate dependent dioxygenase (FTO) gene expression closer to the young adolescence period16. We then wonder whether gene expression profiles respond differentially in severe burns in the young compared to adults in the muscle specifically. We then propose that miRNA-regulated gene expression contributes to muscle pathophysiological changes in response to severe burns that are distinct in the young.

To date, the sparse information on differential effects of age in response to injury may be related to the lack of young burn animal models. The life span of rodents differs from humans, reaching maturity in 3–6 months in mice. 5–8 week-old mice are then equivalent to human adolescents aged 12–18 years17. Comparing 5 week-old (adolescent) to 11 week-old mice (adult) in this study, we examined miRNA and mRNA age-related expression profiles in mouse skeletal muscle.

Results

General description of transcriptome alteration in young and adult burn mice

Among a total of 55,401 grand counts, 21,859 protein-coding genes and 2,202 miRNA counts were deferentially expressed in mouse muscle tissue, reflecting a proportional alteration of 39.45% genes and 3.97% miRNA in the total cohorts. These data reinforce previous findings suggesting a small number of miRNAs are present to regulate expression/function of a large number of target genes/mRNA.

As illustrated in Table 1, 5.09% of muscle mRNA transcripts were differentially expressed between young sham and adult sham animals. Less than 1% of mRNAs were differentially expressed after burn in both age groups; 0.93% when comparing adult burn to sham and 0.83% when comparing young burn to sham. In all comparisons, most changes in mRNA transcripts had fold changes less than 1. Though 22 more mRNAs were expressed in adult burns than young burns, after comparing to each respective sham group, the number of differentially expressed mRNA with absolute fold changes > 1 was greater in the young burn animals. The overlap between differentially expressed genes (DEGs) across burn injury and ages was demonstrated in the Venn diagram (Fig. 1).

Table 1 The number of differently expressed mRNA and miRNAs in mouse gastrocnemius between young sham (YS), young burn (YB), adult sham (AS), and adult burn (AB) groups.
Full size table
Figure 1
figure 1

Venn diagram demonstrates the number of overlapped DEGs across burn injury and ages. (A) Comparisons between four groups, labeled as a young sham (YS), young burn (YB), adult sham (AS), and adult burn (AB). Four comparisons were colored in yellow (AB vs. AS), green (YB vs. YS), gray (AS vs. YS), and blue (AB vs. YB); (B) comparison between young burn and adult burn; (C) comparison between burn and sham.

Full size image

We found 154 miRNAs differentially expressed in young sham animals compared to adult sham mice while 147 miRNAs were differentially expressed when comparing young burns to adult burn animals (Table 1). Similar to mRNA, most miRNAs showed fold changes of less than 1. Mir-1930, mir-29a, mir-29b, and mir-29c were the most down-regulated (fold changes between − 1.7 and − 3.5), while mir-546, mir-3099, and mir-758 were the most up-regulated (fold changes between 3 and 5) in young sham when comparing to adult sham. Interestingly, only two miRNAs were differentially expressed following burn in both the young and adult mice cohorts at 24 h after injury; Mir-10a showed 0.33 fold downregulation in adult burned compared to adult sham mice, and mir-126a showed 0.30 fold upregulation in the young burn compared to young sham mice. All differentially expressed genes and miRNAs are listed in Supplement Table 1A,B.

Hierarchical clustering of the burn effect on gene expression in adult and young mice after exclusion of basal differences

Since the basal gene expression profile was different between young and adult sham groups, we evaluated the differential effects of burn on the young versus the adult groups using the likelihood ratio test implemented in DESeq2. Therefore, we compared a reduced model with age as a factor to the full model with the age and burn status. We found 415 genes significantly changed as presented in the heat map illustrating the hierarchical clustering (Fig. 2). A positive log2 fold change with the red pseudo color means the change of the individual gene due to the adult burn is higher than the young burn, and the blue color indicates the opposite. Generally, genes expressing muscle-specific protein (Myh2, Myh1, Myom3, Mybpc1) were downregulated in adult burn but not in the young burn group. Genes involved in proteolysis (Rnf115, Ubb) and metabolic processes (Gpt2, Pfkfb4) were relatively upregulated in adult burn animals.

Figure 2
figure 2

Using analysis of variance, the burn effect difference between young and adult was presented in the clustered heatmap. The heatmap was created with the ‘pheatmap’ R package (Pretty Heatmaps, version: 1.0.12, CRAN) showing the top 100 genes with the largest burn effect differences between the young and adult mice. Four groups were labeled as a young sham (YS), young burn (YB), adult sham (AS), and adult burn (AB). One animal presented as an outlier and thus been excluded in the young burn group. A positive log2 fold change means the change due to the adult burn is higher than the young burn with pseudo red color. The intensity of pseudo colors indicates the degree of an altered gene either upregulated with red color or downregulated with blue color. Two-way ANOVA with a post-hoc test was applied, and significance was accepted at p values < 0.05.

Full size image

GSEA analysis with Hallmark gene sets

Hallmark gene sets are coherently expressed signatures derived by aggregating many MSigDB gene sets to represent well-defined biological states or processes. We analyzed 50 Hallmark gene sets that included 40,071 genes, and found 1,106 differentially expressed genes with 38 overlapped gene sets in young sham compared to adult sham mice. This demonstrates an age effect on the expression of 2.76% of Hallmark gene sets. 21 and 20 gene sets (with 204 and 179 genes each) were affected by the burn in adult and young mice, respectively (Supplemental Table 2A).

Myogenesis, epithelial-mesenchymal transition (EMT), and adipogenesis were the top 3 modified gene sets in young sham mice compared to their adult counterparts. mTORC1 signaling, MYC targets v1, and myogenesis were the top 3 gene-sets affected by burn in the adult mice. Complement, KRAS signaling, and epithelial to mesenchymal transition (EMT) were the top 3 gene sets affected in young mice after burn. The top 25 hallmark gene sets altered in each comparison group with the number of changed genes are listed in Table 2.

Table 2 Hallmark gene sets analysis in mouse gastrocnemius between young sham (YS), young burn (YB), adult sham (AS), and adult burn (AB) groups.
Full size table

Canonical Wiki-Pathways

Canonical Pathways gene sets analysis is derived from the WikiPathways pathway database. Using the WikiPathways database, we identified over 100 out of 587 overlapping gene sets in young sham compared to adult sham mice. Only 24 and 15 overlapping gene sets were identified when comparing adult burn and young burn mice to their sham counterparts, respectively (Supplemental Table 2B). VEGFA-VEGFR2, focal adhesion, and nuclear receptors meta-pathways were the top 3 altered gene sets when analyzing age effect by comparing young to adult sham animals. Proteasome degradation, parkin-ubiquitin-proteasomal system, and nuclear receptors meta-pathways are the top 3 gene sets altered by the burn in the adult burn while the proteasome degradation, focal adhesion, and PI3K-AKT pathways are the top 3 affected by burn in young animals. These data confirm the role of the ubiquitin–proteasome pathway in regulating the rapid muscle proteolysis starting at the 24–48 h ebb phase after burn. The top 25 canonical pathways gene sets with the number of genes altered in each pathway are listed in Table 3.

Table 3 Wiki gene sets analysis in mouse gastrocnemius between young sham (YS), young burn (YB), adult sham (AS), and adult burn (AB) groups. N: the number of altered genes.
Full size table

miRNA target prediction gene sets

To determine the interaction of miRNA and gene expression, we chose all miRNA target prediction gene sets from the combined superset of both the online database for miRNA target prediction and functional annotations (miRDB) and legacy sets in Gene Set Enrichment Analysis (GESA). Over 100 of 2,598 miRNAs were altered in the young sham compared to their adult counterparts (Supplemental Table 2C). Similar to mRNA data, a smaller number of miRNA were affected by burn; 30 and 14 miRNAs were identified in adult and young burned animals, respectively (Table 4). MIR-12316, MIR153-5P, and MIR8485 are the top 3 miRNAs related to the number of mRNAs differing in adult sham animals compared to the young ones. MIR124-3P, MIR506-3P, and MIR520D-5P are the top 3 miRNAs related to genes affected by burn in adult animals. MIR6867-5P, MIR3658, and MIR124-3P are the top 3 miRNAs regulating mRNA altered by the burn in young animals. These data confirm the specificity of miRNAs in regulating a large number of genes in response to both age and injury. All miRNA target prediction gene sets are listed in Table 4.

Table 4 miRNA interactions analysis in mouse gastrocnemius between young sham (YS), young burn (YB), adult sham (AS), and adult burn (AB) groups.
Full size table

Discussion

This study investigates the acute response of muscle-specific transcriptome to severe scald burn in a mouse model, and further deciphers potential interactions between miRNA and mRNA expression profiles. To assess transcriptomic differences between pediatric/adolescent and adult subjects, we chose to compare young (5-week-old) and adult (11-week-old) mice. We described the characteristic profiles of mRNA and miRNA expression in both young and adult mice and their changes in response to injury using an NGS analysis. These changes affected less than 1% genes and 0.05% miRNA covered by the 21,859 mouse GRCm38 genome library and 2202 miRNA database respectively, suggesting that miRNA possesses wide substrate specificity when regulating genes pathways involved in muscle homeostasis regardless of age.

GESA analysis confirmed gene pathways that are differentially expressed in response to both age and burn. We used two different sets of GSEA to identify the pattern of gene pathways in response to age, and burn and found compatible data. We also sorted DEGs by either up-regulated or down-regulated and further clarified the activities of biological processing and signal pathway. BPs including skeletal system development and regulation of cell differentiation were activated in young mice compared to adult. Apoptotic process and regulation of cell death were activated in adult burn (comparing to adult sham), while interestingly, inflammatory and immune responses were activated in young burn (comparing to young sham) (Supplemental Table 3A). The results of WikiPathway analysis were also specified (Supplemental Table 3B) Focal adhesion was activated and oxidative stress pathway was deactivated in young sham when comparing adult sham. Proteasome degradation was activated in both young and adult following injury. Oxidative stress response was more presented in adult burn, while inflammatory response like complement activation and macrophage markers was more activated in young burn comparing to young sham. Although major catabolic gene responses to burn were observed in both young and adult animals, additional specific pathway alterations were distinguishable between young and adult mice. These findings may reflect differential effects observed in young adolescents compared to adult burn patients.

mRNA transcript profiles in the muscle, in response to severe injury, were age-specific. Transcriptome analysis of coagulation factors in a human closely related species such as the Chinese rhesus macaque showed that basal gene expression was affected by age19. Changes in the expression of muscle growth genes were also reported in bulls between 15 and 19 months of age20. In our study, a comparison of muscle-specific transcriptome profiles revealed an age effect on gene expression between 5-week-old and 11 week-old mice with paralleled burn effects on gene expression in both young and adult groups. The protein degradation pathway was the most affected gene pathway in both adult and young burned mice. Ubiquitin expression were shown to up-regulate in both human and rat skeletal muscles during ageing21. However, detailed information regarding specific genes, transcriptome fold changes, and the number of altered transcripts were different, which indicates a distinguished response to injury in young mice compared to adults. One example is gene expression regulated by polyubiquitination. Ubb gene encodes ubiquitin and was upregulated 0.64-fold in the adult burn group when comparing adult sham, but it was not differentially expressed in the young burn group. However, the Ubc gene encoding Polyubiquitin-C to maintain cellular ubiquitin level, was down-regulated (-1.72-fold) in the young burn group. A direct comparison between the young burned and the adult burned mice shows that Ubb gene level increases 0.81-fold and the Ubc gene level increases 1.13, further indicating a delicate tuning on the regulatory mechanism with different ages at the transcriptional level.

Myogenesis was another typical altered GO-BP pathway in the study. In this study, we observed the most number of DEGs enriched in myogenesis when comparing young sham to adult sham. The previous study showed that the progenitor cell activity in aged muscle can be restored when treated with young serum22. It is not surprise that the myogenic activity is greater in younger mice. Further stratification of up- and down-regulated genes, we confirmed that activated biological processing of skeletal muscle development and cell differentiation in young sham group. In this study we also noticed that myogenesis has also been affected by external stimulations. In a previous study of burn mouse model, we found that severe burn caused insufficient myogenic activation23. Corrick et al. set up an in vitro model with human serum stimulation from burn patients, and they explored that myogenesis impairment could be related to STAT pathway activation24. Further dissection of specific DEGs will help us understand better the role of myogenesis in the maintenance of muscle homeostasis in young burn victims. miRNAs serve as negative regulators of gene expression. Similar to mRNA, age is also prominent in miRNA expression changes after injury, and the miRNA profile presented with different characters of neurobiological development with age25. Using microarray and TaqMan-based expression analysis of miRBase 10.0, Moreau et al. observed that 312 miRNAs were differentially expressed in 48 post-mortem human tissue of fetal, young, and adult brains25. In our study, we observed increased expression of the miRNA 29s family with age. The miRNA 29s family includes three mature members a, b, c, that regulate different genes involved in cell growth, differentiation, apoptosis, and regulation of the immune response26. miR-29s family members target at least 16 extracellular matrix genes for focal adhesion, and are upregulated in aged mice aortas27. For muscle tissue, miR-29 impairs myoblast proliferation, to assist with cell arrest protein expression (p53, p16, and pRB) during aging28. We found that the endogenous levels of miR-29 are elevated with age, specifically in the C57BL/6 male mouse gastrocnemius 5-week-old compared to the 11-week-old. Though further work is indicated to investigate this link, age-related miRNAs could interfere with the genomic response to burn. Therefore, our data show that miR-29 could be an important therapeutic target of aging-related diseases, particularly after injury.

To our surprise, only one miRNA significantly changed in response to burn in each age group, with miR-10a downregulated and miR-126a upregulated in adult and young burned mice respectively. miR-10 is overexpressed following ischemic injury29. miR-126 has protective effects in an animal model of endotoxemia-induced vascular injury30,31. Using microarray analysis, we previously reported miRNA differential expression 14 days after burn in rats11. Hu et al., showed changes in the miRNA profiles over 1 and 7 days following traumatic brain injury, demonstrating one/multiple miRNA expression changes over time32. Changes in the genomic kinetics over time have been observed in critical trauma and burn patients33, which correspond with a systemic inflammatory34, and muscle pathophysiological changes35. In a previous study, we examined the muscle transcriptome profile from male adult rats at 14 days while the current study examined early changes in mouse muscle transcriptome profile at 1-day after-burn. Our results suggest that changes in mRNA and miRNA expression reflect a temporal protective feedback response starting 24 h after burn, and that further time course studies are needed to characterize this response.

Using miRNA and mRNA interactive pathway analysis, we found that the number of miRNAs interacting with genes is similar to the number of gene pathways affected in response to age and burn. The number of interactive mRNAs range from 90 to 34 in young mice comparing to adult mice without injury. The number ranges decreased from 15 to 4 in young burned mice or adult burned mice. The highly expressed gene-related miRNA in response to severe burn in both age groups were MIR-506_3p, MIR-124_3p, MIR-6867_5p, and MIR-3658 with the greatest number of related- altered genes in young burned mice. Interestingly, MIR-3658 has 74 altered genes connected in adult sham when compared to young sham. These novel findings provide a new avenue to study the roles in pediatric burns.

In summary, this study demonstrates the characteristic profile of gene expression from skeletal muscle in young and adult burns. Prominent age effects were presented in transcriptional levels with abundant alterations of genes, miRNAs, and interactions, leading to regulation pathways mainly on myogenesis, cell growth, and development. Our results further confirmed the effect of severe burn in mice muscle, with fewer gene sets and miRNAs affected in both age groups. Protein degradation, inflammation, and related pathways are affected by injury in both young and adult groups, however, more genes and related pathways were affected in the adults. The study descripted the different responses at the transcriptional level, including the altered DEGs and miRNA among the experimental groups, followed by an analysis of gene sets related pathways and interactions of miRNAs to mRNAs by GESA tool. Further work in correlation of muscle pathophysiological changes could be conducted from both animal model and clinical research.

This study also provides information on the miRNA and mRNA expression profiles in young burn mice and could lead to further studies on the mechanisms of gene regulation in pediatric burn patients. As for pediatric burn research, most metabolic parameters were studied broadly, however, genomic information reflecting those changes is missing. The current study opens a direction that emphasizes that age influences metabolic changes in with development and aging after burn. The pathophysiologic response evaluation combined with genomic profile examination will allow for a better picture of events, and explain variation in results greater than one single layer observation at the transcriptional level.

Methods

Twelve C57BL6 male mice (six 4-weeks-old, young; and six 10-weeks-old, adult) were obtained from Charles River Laboratory (Wilmington, MA). All procedures followed the NIH and ARRIVE guidelines and were approved by the local IACUC at the University of Texas Medical Branch at Galveston (UTMB). Animals were housed in a temperature-controlled room with a 12-h light/ dark cycle with free access to food and water.

After one-week acclimation, half of the young and adult animals received a 25% body surface area (TBSA) scald burn23. Briefly, mice were anesthetized with 2% isoflurane inhalation and shaved on the back and belly, 30 min after receiving a subcutaneous injection of 0.05 mg/kg buprenorphine. Mice were placed in a mold with an opening to expose about 12.5% body surface after receiving 1 mL of 0.9% saline injection subcutaneously along the spinal column. Mice were then immersed in 98 °C water for 10 s on the dorsal and 2 s on the ventral side to receive a total 25% TBSA full-thickness scald burn. The animals received 1.5 mL of intraperitoneal lactated Ringer’s solution for resuscitation during the burn procedure. Sham animals received the same procedure with the anesthetics and analgesics but did not receive a scald burn or resuscitation. Animals all survived after burn procedure, and which be consistently observed in our previous studies23,36. Animals were euthanized 24 h after burn with a CO2 inhalation overdose. Muscle tissues were harvested and stored at − 80 °C for further analysis.

Severe burn caused metabolic changes include the early ebb stress phase and later flow phase. The transition time point is usually 24-48 h after injury37. Therefore, we chose the initial time point 24 h to explore the regulatory mechanism of post-burn muscle wasting in transcriptional level. Gastrocnemius tissue contains mixed slow-and fast- myofiber types, and its anatomic location is far away from burn wound site in mouse trunk. We and others analyzed the gastrocnemius tissue to represent the muscle response to systemic influence in mouse and rat models38,39. Gastrocnemius tissues (~ 25 mg) were processed for total RNA extraction using Qiagen RNeasy Mini Kit (Germantown, MD). After measurement of concentration with Nanodrop 2000/2000c Spectrophotometers (ThermoFisher, Waltham, MA), 2 µg of total RNA were analyzed at the Next Generation Sequence (NGS) core laboratory at UTMB. (Supplemental file 4) RNA sample quality was confirmed using an Agilent Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA). Poly-adenylated RNA (mRNA) and microRNA (miRNA) sequence libraries were prepared with NEBNext Ultra II RNA and Small RNA kits, respectively (New England BioLabs, Ipswich, MA). Libraries were sequenced on an Illumina NextSeq 550, using the mid-output kit and paired-end 75 base parameters (Fig. 3).

Figure 3
figure 3

Experiment flow chart outlines the study design and sample processing. The Image was created in BioRender.

Full size image

Reads from the mRNA samples were mapped to the mouse GRCm38 genome reference using the STAR read aligner, with the quantMode GeneCounts option to quantify read counts based on the Gencode version M25 mouse annotation. Reads from the miRNA samples were trimmed of adapters using the mirDeep2.0 software package, then mapped to the mouse genome reference with STAR. The featureCounts program in the subread software package was used to count reads mapping to miRNAs. Differential gene expression for mRNAs and miRNAs was estimated with the Differential gene expression analysis based on the negative binomial distribution (DESeq2) package following the authors’ vignette. The significance limit was set to an adjusted p value of < 0.05. Descriptive statistics was applied for transcript profile characterization. Data replicability was demonstrated in the PCA plot (Supplemental Fig. 1) Gene expression pathway and miRNA interaction were further analyzed using an online program, Gene Set Enrichment Analysis (GSEA) (www.gsea-msigdb.org). Two-way ANOVA with a post-hoc test was applied, and significance was accepted at p values < 0.05.

Conference presentation

Data was presented at the 44th Shock annual meeting, virtual on October 10–14, 2021.