Modulation of HOXA9 after skeletal muscle denervation and reinnervation

Background HOXA9 (Homeobox A9), whose expression is promoted by MLL1 (Mixed Lineage Leukemia 1) and WDR5 (WD-40 repeat protein 5), is a homeodomain-containing transcription factor which plays an essential role in regulating stem cell activity. HOXA9 inhibits regeneration of skeletal muscle and delays the recovery after muscle wound in aged mice, but is little known in denervated/reinnervated muscles. Methods we performed detailed time-process expression analysis on HOXA9 and its promotors, MLL1 and WDR5, in the rat gastrocnemius muscle after three types of sciatic nerve surgeries: nerve transection (denervation); end-to-end repairing (repairing); and the sham operation. Then the specific mechanisms of Hoxa9 were detected in vitro through primary satellite cells transfected respectively by pIRES2-DsRed2 empty plasmids, pIRES2-DsRed2-HOXA9 plasmids, pPLK/ GFP -Puro empty plasmids, and pPLK/GFP-Puro- HOXA9 shRNA plasmids. Results We found that HOXA9 expression was synchronous with the severity of muscle atrophy, as well as the upregulation of MLL1 and WDR5 associated with the denervation state to some extent. Indeed, experiments with primary satellite cells revealed that HOXA9 inhibited myogenic differentiation, but not destroy the differentiation potential, influenced the best-known atrophic pathways, and promoted apoptosis. Conclusion HOXA9 may play a pro-atrophic role in denervated muscle atrophy. Abcam, UK), rabbit anti-P-FoxO3a(1:500; ab47285; Abcam, UK), rabbit anti-NF-κB(1:500; ab16502; Abcam, UK), rabbit anti-P-NF-κB(1:1000; ab86299; Abcam, UK). After 3 washes, blots were incubated at 37◦C for 2h with appropriate secondary antibodies: HRP-labeled Goat Anti-Mouse IgG (H+L) (1:5000; Beyotime), or HRP-labeled Goat Anti-Rabbit IgG(H+L) (1:2000; Beyotime). Immunoreactive bands were developed by an enhanced chemiluminescence (ECL) kit (AR1171; Boster, China) and visualized through Chemidoc™ XRS+ (BIORAD, USA).

In addition, stem cells of muscle, the satellite cells are activated under denervation and try to differentiate into myofibers to repair muscle loss. Unfortunately, this restorative phenomenon remains shortly and inadequately until timely, sufficient reinnervation, which is often unlikely after serious nerve injuries, responsible for the continuous atrophy of skeletal muscle [12].
HOXA9, a member of Hox genes, is a key regulator of stem cells and plays a positive role in regulating the development of muscles and limbs during embryogenesis, while it can only be detected in a few stem cells after birth. Interestingly. HOXA9 is usually activated in several serious conditions such as cancer, leukemia, aging, and so on, showing a contrary effect: interference with the homeostasis of stem cells [13][14][15]. But its expressing and molecular basis in denervated muscles still remains unknown.
Here, we analyzed the express rules of HOXA9 and its promoters, MLL1and WDR5, in the rat gastrocnemius muscle after sciatic nerve operation, comparing denervation, reinnervation, and the general conditions. In addition, to furtherly investigate the effects of HOXA9 in skeletal muscles, experiments in vitro were performed in primary satellite cells.

Animal procedures
Male SD rats (200-250 g) were housed in a standard facility (23℃, 50% relative humidity, 12-h light/12-h dark cycle) and provided with adequate food and water. The rats were allocated randomly into the following groups (36 rats/group): denervation group; denervation + nerve repairing group (repairing); and sham operation group. Denervation and nerve repairing operations were performed surgically on the right hind legs as previously described [16] [4]. Surgeries were conducted of mice under anesthetization with pentobarbital (50 mg/kg ip.). The sciatic nerve was separated and exposed via a lateral incision in the mid-thigh. Afterwards, in denervation group we removed 0.5 cm of the sciatic nerve and buried both ends in the muscles; in the repairing group, the nerve was cut in the middle, and the epineurium was then immediately sutured with 9-0 nylon thread; in the sham group, no neurotomy was carried out after exposed. After that, we closed the incision with a 4-0 absorbable suture. 6 rats of each group were euthanized at the indicated time post-operation (0d, 3d, 7d, 2w, 3w, 4w), and gastrocnemii on both sides were removed to assess atrophy.

5
The sediment was then suspended in the growth medium consisting of DMEM/F12ham, 20% FBS, and 1% penicillin/streptomycin (Solarbio). We transferred the suspended cells to a collagen-coated petri dish, where they were pre-attached for two hours to purify satellite cells [17]. After this, the supernatant containing the non-adherent cells was transferred to a new collagen-coated petri dish. The growth medium was changed every two days. The cells were used in subsequent experiments upon reaching approximately 60%-70% confluence. To verify the purity of satellite cells, immunofluorescence staining with Desmin protein, a myogenic marker which expressed early in early stages of satellite cells' differentiation [18], was performed.

Plasmids and transfection
shRNA of HOXA9 (target: AAGTGTGAGTGTCAAGCGT) was inserted into the pPLK/GFP-Puro plasmids, and the coding region sequence of rat HOXA9 was cloned into the pires2-dsred2 vector to produce a recombinant HOXA9 construct (Public Protein/Plasmids Library, China).

Wet weight ratio
The gastrocnemius muscles of the surgical side and contralateral side were both taken, washed with normal saline, dried with filter paper, and weighed. The wet weight ratio was calculated by dividing the muscle weight of the operative side by that of the contralateral side.

Immunohistochemistry of paraffin sections
Mid-portions of the operational gastrocnemii were fixed through paraformaldehyde (4%), then dehydrated, and paraffin-embedded. 5-μm thick cross-sectional slices of the muscle wax samples were made on a slicing machine (Leica Biosystems, Germany). Following deparaffinage, hydration, heat mediated antigen retrieval (Tris/EDTA buffer PH 9.0, Solarbio), endogenous peroxidase activity blocking, and goat serum blocking, primary

Quantification of fiber diameter
Hematoxylin staining in HOXA9 immunohistochemical staining makes each part of the sections clearly visible. Based on this, Muscle fiber diameters were calculated and analyzed. The diameter was determined in Caseviewer 2.2 (3DHISTECH Ltd, Hungary), and the average diameter of every fiber was expressed as a mean of three measurements. At least 600 total muscle fibers from six fields of each section were assessed.

Western blot
Western blot analysis was performed as previously described [4]. Proteins were harvested from primary satellite cells or gastrocnemius muscles with Whole Cell Lysis Assay The gray value of bands was analyzed with Image Lab 6.0.0 (BIORAD, USA).

RT-PCR
Total RNA from gastrocnemius muscles was extracted using the RNA Extraction Kit (Code No. 9767; Takara, Japan). cDNA was synthesized using the primescript™ RT Master Mix

TUNEL staining TUNEL assay
In satellite cells 36h after transfection, the TUNEL assay was performed using the In Situ Cell Apoptosis Detection Kit I (POD) (Boster, China) as the specification requested.

Statistical analysis
Statistical analysis was carried out using the Graphpad Prism 7 Software (US) with all data shown as mean ± standard deviation calculated from at least six independent experiments. Each experimental measure was performed in triplicate. Results were analyzed using the One-way ANOVA when coming from three or more groups and using the t-test when from two experimental groups. P < 0.05 was considered statistically significant.

Results 9
Analysis of muscular atrophy after sciatic nerve operations

Wet weight and fiber diameter
Wet weight analysis (figure 1.A) revealed muscle atrophy in denervated and repaired muscles from 3d to 8w by the wet weight ratio (operational weight divided by specific weight). Consistent with previous research [4], gastrocnemius at the side of denervation operation atrophied rapidly; while in the repairing group, the mass loss was rescued after 2 weeks post-operation but still more serious than the sham group. The result showed the denervated atrophy was a gradual slowdown, with muscle wet weight dropping rapidly in the first 2 weeks (gastrocnemius muscle decreased by an average of 57%), then gradually in the second 2 weeks (average 17%) and only 11% over the following 4 weeks.
These data suggested that HOXA9 inhibited myogenic differentiation in satellite cells, but it didn't damage the differentiation capacity. Thus, it can be seen that HOXA9 could influence the typical atrophy pathways and lead to atrophy in satellite cells.

Discussion
The purpose of this study was to investigate the mechanism of HOXA9 in skeletal muscles after peripheral nerve injury and reinnervation. To explore the molecular changes occurring in denervated and reinnervated muscles, we made use of three models of median nerve surgeries models to simulate the most common clinical conditions, namely denervation (peripheral nerve injury), repairing (peripheral nerve injury followed by surgical reconstructive surgery) and sham (the general population).
We analyzed the muscle wet weight ratio and fiber diameter to follow histological changes in these three conditions. In this study, we found that the muscle atrophy worsened over time after denervation, while the shrinking trend was restored from 3w post repairing. The muscle fiber size distribution confirmed these data: at 3w, 4w, 8w post-operation, the distribution was gradually similar between the repairing and the sham groups; in contrast, in denervated muscles, the number of smaller fibers was higher, consistent with literature data [24].
Denervation has been reported to increase the sensitivity of satellite cells to apoptosis [25,26], which may result in the inability of muscle regeneration after a long period of time without nerve input. And the activation of satellite cells is a significant pathway of 13 muscle recovery after denervated [27]. HOXA9 belonging to the HOX gene family is a key regulator of stem cells and tissue patterning during embryogenesis, playing roles either as an oncogene or as a tumor suppressor in various tumors [28]. Its involvement in cancers is well studied [29][30][31]. Since Simon Schwoerer's discovery that HOXA9 was particularly activated in aged skeletal muscle and limited satellite cells' function and muscle regeneration by activating senescence-related signaling pathways [15], the mechanism of HOXA9 in skeletal muscle has attracted extensive attention, but it is still little known after nerve injury.
We have demonstrated for the first time that HOXA9 is regulated in denervated and reinnervated skeletal muscle. We found the HOXA9 protein, not its upstream regulator MLL1 or WDR5, kept pace with the extent of muscle atrophy in a sense, continuing to increase after denervation, reaching a peak at 3w post repairing and then decreasing gradually to baseline, changed insignificantly in the sham group. Moreover, HOXA9 protein expressions seemed to be less after repairing than the denervation group in the case of a similar degree of muscular atrophy (3d, 1w, 2w), though the statistical difference was only detected at 2w. In particular, HOXA9 protein in the repairing group decreased to the control group level at 8w post-operation, while the wet weight ratio still lower than the sham muscle, showing asynchronous to atrophy and relative irreversibility of muscle injury after serious nerve injury in the long-term analysis. These data suggested there must be some mechanism between HOXA9 protein and denervated muscle atrophy.
Considering the regulatory effect of HOXA9 on stem cells, is it through the regulation of satellite cells? The vitro experiments confirmed this assumption. HOXA9 inhibited myogenic differentiation, affected the classical signaling pathway of muscular atrophy, and induced apoptosis of satellite cells, but the Pax7 showed no significant change, meant the differentiation potential of the remaining satellite cells had not weakened nor 14 disappeared. In summary, HOXA9 seemed to inhibit the development process of satellite cells towards myotubes without damaging its activity of differentiation. Coincidentally, satellite cells from denervated muscles were able to differentiate in culture even after up to seven months of denervation [32], but it is unattainable in vivo [33]. There must be special mechanisms in vivo which regulate satellite cell activity after peripheral nerve injury and determines the fate of the muscle. Is it HOXA9? Maybe, but this will require further verifications.
There are still some questions here. First of all, the spatial distribution of HOXA9 protein after denervation, especially at 4w and 8w, a large part of which was located in the cytoplasm rather than the nuclear, was significantly different from the other groups in which HOXA9 protein was mainly located in the nucleus. At the same time, HOXA9 protein in the nuclear also increases with the aggravation of muscular atrophy in both denervation and repairing groups. It may be that HOXA9 in the cytoplasm is an activated form or some metabolites. However, no further research was carried out in this study. Besides, in vitro experiments, HOXA9 mRNA expression level in the pIRES2-DsRed2-HOXA9 group increased abnormally to as much as 2750.83 times, while the HOXA9 protein increased to only 4.16 times. We suspect that there are mechanisms in cells that we have not found that inhibit HOXA9 activity and maintain the stability of the intracellular environment. This may be related to the silence of HOXA9 after birth in vivo. But we have no evidence for that at the moment.

Conclusion
To summarize, this study revealed the modulation of HOXA9 after skeletal muscle

Consent for publication
Not applicable.

Availability of data and materials
The datasets during and/or analyzed during the current study available from the corresponding author on reasonable request.   Wet weight and fiber diameter analysis of denervated and reinnervated muscles. Denervation 3d n=940, 1w n=952, 2w n=843, 3w n=828, 4w n=843 ,8w n=749;
Scale bar:100μm. B+C Western blot and relative grey value analysis of HOXA9,   Scale bar: 100 μm. B+D Quantification of apoptosis based on TUNEL staining as in A or C. A total of 6 fields per group were analyzed. *P < 0.05.