Muscle Fibrosis,NF- κ B,and TGF- β Are Differentially Altered in Two Models of Paralysis (Botox Versus Neurectomy)

Animals and surgical procedures

Twenty male Sprague-Dawley rats (300–325 g) were obtained from Charles River Laboratories (Wilmington, MA). The rats were randomly assigned into the sciatic-femoral neurectomy group (SFN, n = 8 rats), Botox paralysis group (BTX, n = 8 rats), or control group (C, n = 4 rats). Botox is a well-established and well-researched neurotoxin, and prior research showed no differences between PBS-injected shams and baseline control animals. ²¹ Thus, a PBS injection control was not included. One animal from the SFN group was withdrawn because it met the criteria for early endpoint outlined in the Institutional Animal Care and Use Committee (IACUC) protocol due to self-injury and subsequent destress. In addition, one animal from the Botox study passed away due to unknown causes following secondary Botox injection and was removed from the study resulting in a final group size of n = 7 for both the SFN and BTX groups. Contralateral legs were used as internal controls for comparison with treatment limbs, whereas sham was used as a no-surgery control for both treatment groups.

All surgeries were performed humanely under a protocol approved by the IACUC at the Virginia Commonwealth University and followed the National Institutes of Health guide for the care and use of laboratory animals. All animals were single-housed in an individually ventilated, solid-bottomed polysulfone cage and were kept in an AAALAC-accredited animal facility in indoor housing capable of temperature and humidity control within limits appropriate for the species and a 12-h/12-h light/dark cycle. Food pellets and water were provided ad libitum. Rats were acclimated to their environment for five days before surgery.

Animal surgery

Before surgery, animals were anesthetized using aerosolized 4–5% isoflurane/400 mL/min O₂, and anesthesia was maintained during surgery with 2.5% isoflurane, 400 mL/min O₂. The hind limb surgery site was shaved with electric clippers and cleaned with 3× alternating swabs of isopropanol and chlorhexidine. Rats were transferred to the heated surgical table covered in a clean disposable underpad.

The sciatic neurectomy was performed with the rat in a prone position. A dorsolateral incision was made in the proximal thigh extending from the femoral head to mid-diaphysis. An intramuscular window was made in the bicep femoris to expose the sciatic nerve. Two to three centimeters of sciatic nerve was removed and the nerve stumps were knotted with 10-0 sutures to prevent nerve regeneration. The surgical site was closed with 4-0 vicryl sutures and wound clips. Next, to perform the femoral neurectomy, the rats were placed in a supine position before a dorsomedial incision was made extending from mid-diaphysis to the rat’s midline exposing the femoral nerve and artery. The femoral nerve was carefully separated from the surrounding tissue without damaging the femoral artery and a 5 mm segment was removed. The nerve stumps were sutured and the surgical site was closed. Control sham groups received the same surgery, but the nerves were left intact.

Paralysis was induced using Botulinum Toxin A (BOTOX^®, Allergan Inc., Irvine, CA [Botox]) acquired from the VCU Hospital Pharmacy (Richmond, VA) dissolved in 0.9% sterile saline at a concentration of 10U/mL. Each animal was injected with 2 units/muscle-group intramuscularly into the right hind limb muscles of the proximal femur, quadriceps femoris, hamstring, and calf muscles for a total of 8 units/limb/animal. Animals were prepared as outlined presurgery before day 32 injections. The secondary injection (day 32) was performed to maintain muscle paralysis for the full duration of the study.

For both procedures, rats were monitored during recovery to check for side effects. Animals were observed once daily at a minimum for health checks and monitoring Botox-treated animals for recovered mobility. No Botox-treated animal showed identifiable signs of recovered mobility by the end time point. Each rat was weighed weekly with free access to food and water while housed in a ventilated, temperature humidity-controlled, single rat cage contained in a room with a 12-h/12-h light/dark cycle. Wound clips were removed two weeks following neurectomy surgery and additional Botox injections were made on day 32 to maintain muscle paralysis. Fifty-two days following surgery, muscle force was measured and then animals were euthanized using 10% CO₂/volume/minute inhalation.

Muscle function and physiology

Muscle force measurements were performed in vivo using the 1300A Whole Animal System (Aurora Scientific, ON, Canada). Rats were anesthetized following the same anesthetization methodology as performed for surgery. Following general anesthesia, each rat was placed in a supine position on the sterile device platform. Each leg was individually fixed at the knee with the paw pressed firmly onto the footpad. Two electrodes (Grass Technologies subdermal electrodes, Fisher Scientific) were inserted into the lateral subdermal space adjacent to the distal crural muscle compartment. Muscle length and stimulating voltage were optimized for each limb before twitch and tetany measurements. Muscle length was optimized by gently stretching the limb to full extension while the knee was fixed. With electrodes in place, the stimulation voltage was optimized by slowly increasing the stimulation voltage until full motor recruitment was achieved denoted by a sustained plateau in muscle force output. Twitch measurements were performed three times in a row by sequentially stimulating with a 1 ms electrical pulse with a 1-s waiting period between pulses. After a 120-s waiting period after the twitch tests, tetany measurements were performed by stimulating the muscle in 500-ms intervals at 150 Hz followed by 120-s rest periods between stimulations to allow for muscle recovery. Maximum force, total time, contraction rates, total energy, minimum force, and relaxation rates were recorded for each test.

Muscle and nerve morphology and weight

Immediately after euthanasia, each limb’s muscles were exposed and the surgical sites were checked for signs of abnormal fibrosis or irregularities. Following the preliminary quality control check, the gastrocnemius and the available intact sciatic nerves along with the remaining distal nerve stumps were isolated and removed. Both tissues were individually weighed. Images of respective treatment and contralateral muscles were taken using a Nikon Rebel T3 camera after placing the muscles next to a surgical ruler, and then weighed. Muscles were then measured for width and length. Following measurements, muscles were cut in half. One half of the muscle was transferred to individual tubes containing 4% paraformaldehyde stored at room temperature. The other half muscle was put in empty 50 mL conical tubes stored at −80°C.

RNA isolation and Real Time quantitative polymerase chain reaction (RT-qPCR)

Total RNA was isolated from each gastrocnemius muscle in each group. While on ice, a 30–40 mg sample was taken from the muscle belly, minced, and transferred to a Bead Bug Zirconium 6 mm Bulk Bead tube (Benchmark Scientific, Sayreville, NJ) containing 1 mL TRIzol (Thermo Fisher Scientific, Waltham, MA). Each tube was run three times through the device at a maximum speed for 60 s with a 5-min rest period on ice between runs. Sciatic nerve samples were homogenized using the same method as muscle except with QIAzol, but due to the nerve’s high lipid content, the RNeasy Lipid Tissue Mini Kit (Qiagen, Hilden, Germany) was used to isolate RNA. Isolated RNA concentration and purity were measured using a microplate spectrophotometer (BioTek Take3, BioTek Synergy H1, Winooski, VT). Primers used in RT-qPCR to measure myogenesis and synaptogenesis are listed in Table 1. Primers used to measure markers of fibrogenesis and inflammation are listed in Table 2. All PCR data are presented as fold change (2 ^-ΔΔCT) against sham.

Histology and morphometrics

Gastrocnemius muscles were stored in 4% paraformaldehyde at room temperature, dehydrated, and embedded in paraffin wax. Five-micron sections were placed on Histobond slides (VWR, Radnor, PA) and then deparaffinized and rehydrated before staining with Weigert’s hematoxylin Masson’s trichrome (Sigma-Aldrich, St. Louis, MO), Biebrich scarlet-acid fuchsin (Sigma-Aldrich), and aniline blue (Sigma-Aldrich). Coverslips were mounted with xylene-based mounting media and dried before imaging at 10× magnification. Each region of interest (ROI) was measured for both fibrosis and Feret diameter. Fibrosis was measured by thresholding the image for blue pixels detecting the aniline blue-stained collagen ECM compared with the total number of pixels of the ROI as a percentage using MATLAB. Feret diameter and fiber counts in each ROI were manually measured in ImageJ.

Immunoblotting

All immunoblotting was based on our previously published protocol, ²² the STAR protocol, ²³ an improvement on the standard Bio-Rad protocol, and with protein loading conditions optimized for skeletal muscle using previously published literature. ²⁴ A biopsy punch from the muscle belly of the gastrocnemius was taken for protein samples. The 40–50 mg of tissue was homogenized in 1 mL NP-40 (BP-119, Boston BioProducts, Ashland, MA) with protease inhibitor cocktail and 25 mM sodium fluoride with a 6.0 mm zirconium bead in a beadbug homogenizer (BeadBug^TM Cat #: 31–212, Genesee Scientific, San Diego, CA) at 4000 rpm for 60 s 5 times. The tubes were kept on ice for at least 5 min between runs. The homogenate was centrifuged at 9703 rpm (10,000× g) (Centrifuge 5427R, Eppendorf, Hamburg, Germany) for 5 min to remove insoluble cellular material, and the supernatant was used for dot blotting according to the Bio-Rad manufacturer’s protocol, which uses a 0.45 μM nitrocellulose membrane. Following sample application, the blots were stained with the Revert 700 Total Protein stain for 5 min followed by two washes with Revert Wash buffer for 30 s each. The blots were briefly rinsed with 1×TBS twice and then rehydrated in 1×TBS for 10 min before blocking. Membranes were then blocked for an hour at RT with StartingBlock blocking buffer (Thermo Fisher Scientific). Blots were incubated overnight with primary antibodies (Table 3) at 4°C on a 3D rocker. Next, the membranes were washed 3× with 1×TBS, 0.1% Tween20 solution and then incubated with secondary antibodies (926–32210, LI-COR Biosciences, Lincoln, NE) for 60 min at room temperature. Blots were imaged on the LI-COR Odyssey CLx (LI-COR Biosciences) at medium scan quality, 85 µm resolution, and 0.0 mm offset. A silicone roller was used to remove bubbles before scanning. The entire dot, including rings, was considered for quantification and background noise accounted for in each dot in the LI-COR Image Studio v5.5 software. Signal intensity of targets was normalized to their respective individual sample taken as a percentage of the total protein intensity. All blots were quantified using the following equation. Normalized Signal = Target Intensity ( 800 nm ) Total Protein Intensity ( 700 nm ) × 100 %

Statistical analysis

Surgical group size was based on a power analysis with an α = 0.05 and power = 0.8 using the a dz = 2.43 for the SFN group and dz = 2.7 for the BTX. These data were based on prior gene expression experiments. The analysis led to a final group size of n = 8/group for treatment groups. (sham, n = 4; BTX C, n = 8; BTX T, n = 8; SFN C, n = 8; SFN T, n = 8). Statistical tests were performed using GraphPad Prism 10.0 (GraphPad, LA Holla, CA) and are represented as mean ± SEM. A one-way analysis of variance test with a Tukey post hoc test was used for each data set, with different letters representing statistical significance. Bar graphs with different letters (e.g., A vs. B, B vs. C, or A vs. C) from each other are deemed significant (p value < 0.05). This also includes bars with multiple letters (e.g., AB vs. C) and are significant between these comparisons (p value < 0.05). Bars sharing a letter (e.g., A vs. A, AB vs. B, or BC vs. C) are not significant compared with each other. If the bar contains the same letter, then they are not significant. Outliers from assays were determined with a modified Z-score test removing outliers, three scaled absolute deviations from the median. ²⁵

Muscle function

SFN and BTX treatments caused similar functional deficits in twitch and tetanic muscle force results. In twitch testing, both treatments resulted in decreased maximum torque, a slower contraction rate, and a slower relaxation rate (Fig. 2A–C). Tetanic results showed a similar trend with maximum torque and maximum contraction and relaxation rates decreased in treatment limbs compared with contralateral limbs (Fig. 2D–F), Supplementary Fig. 2.

OPEN IN VIEWER

Figure 2.

Limb functional force testing demonstrates deficits in both Botox paralysis and SFN treatments compared with their respective contralateral group after 52 days. Representative force curve diagrams included for reference. Twitch testing showed maximum torque (A), maximum contraction rate (B), and maximum relaxation rate (C) decreased in both treatment limb groups compared with contralateral and sham. The same metrics were affected as measured by tetanic testing. (D–F). Bars with different letters indicate a significant difference (e.g., A vs. B) (p < 0.05) while bars containing the same letter are not significant (e.g., AB vs. B) using a one-way ANOVA with a Tukey’s post hoc test. Data are shown as mean ± SEM (n = 7/limb). Botox, botulinum toxin A; SFN, Sciatic-femoral neurectomy.

Gross morphometrics and histology

Gross images of the muscle show decreased muscle size in the treated gastrocnemius muscles (Fig. 3A). Gastrocnemius muscle weight and anatomical cross-sectional area decreased compared with contralateral groups and sham both qualitatively and quantitatively (Fig. 3B). Histological sections were taken from the midsection of the gastrocnemius in each limb and stained with Masson’s trichrome. Sham limbs and contralateral control legs displayed typical morphology consistent with healthy muscle (Fig. 3C). Treated limbs showed decreased fiber diameter and increased levels of aniline blue stain compared with contralateral groups and sham (Fig. 3D). These results were quantified morphometrically, demonstrating increased fibrosis in SFN compared with controls, while BTX-treated limbs were not significant (Fig. 3E). In addition, Feret diameter (Fig. 3F) and fiber count (Fig. 3G) showed decreases in both SFN- and BTX-treated limbs compared with the contralateral groups and sham.

OPEN IN VIEWER

Figure 3.

Gross morphology and histomorphometrics of gastrocnemius tissue receiving either a BTX or SFN treatment. Qualitative and quantitative assessments in gastrocnemius size showed decreases in tissue size, wet weight, and anatomical CSA (A, B). Representative cross sections taken from mid-gastrocnemius stained with Masson’s trichrome stain. Scale: 1 pixel: 100 μm (C, D) Histomorphometrics demonstrated that increased amounts of fiber (blue) stain indicate increased collagen only in SFN-treated limbs while both treatment limbs decreased muscle fiber size (E). Fiber diameters were decreased in both treatment groups. (F) Bars with different letters indicate a significant difference (e.g., A vs. B) (p < 0.05), while bars containing the same letter are not significant (e.g., AB vs. B) using a one-way ANOVA with a Tukey’s post hoc test. Data are shown as mean ± SEM (n = 7/limb). SFN, Sciatic-femoral neurectomy.

Gastrocnemius RNA expression

Both treatment groups demonstrated increased levels of mRNA expression in markers for remodeling and new fiber formation as indicated by Myh1, Myh3, Pax7, Myog, Myf5, Myf6, and Itga7. Myh1 and Myh3 expression increased in the SFN and BTX treatment groups, with Myh3 increasing over 40-fold than sham (Fig. 4A, B). Myh7 was not significantly different in treatment limbs compared with contralateral or sham (Fig. 4C). Myh2 was not different between each group (Supplementary Fig. S1). Pax7 levels increased in both the SFN and BTX treatment groups compared with the contralateral groups and sham (Fig. 4D). Myog, Itga7, Myf5, and Myf6 increased compared with controls for both SFN and BTX but were lower in BTX compared with SFN (Fig. 4E–H).

OPEN IN VIEWER

Figure 4.

RT-qPCR measuring markers of satellite cell activation and myogenic regulatory genes. Myh1 (A) and Myh3 (B) increased in both treatment types, while Myh7 (C) was unaffected. Pax7, an indicator of satellite cell activation, increased in both treatment types, (D) while Myf5, Myf6, Myog, and Itga7 (E–H) increased in SFN treatment over BTX treatment, suggesting commitment to myogenesis and new fiber formation was lower in BTX-treated limbs. Bars with different letters indicate a significant difference (e.g., A vs. B) (p < 0.05), while bars containing the same letter are not significant (e.g., AB vs. B) using a one-way ANOVA with a Tukey’s post hoc test. Data are shown as mean ± SEM (n = 7/limb). SFN, Sciatic-femoral neurectomy.

NMJ mRNA and protein level

While only SFN-treated limbs increased mRNA production of Acetylcholine receptor gamma (Chrng) compared with its contralateral group and sham (Fig. 5A), both treatments upregulated mRNAs for Chrne, Musk, Lrp4, and Rapsn (Fig. 5B–E). Furthermore, mRNA levels for Chrne and Musk were highest in SFN muscles compared with BTX, while mRNA levels for Lrp4 and Rapsn were unchanged between treatments. Similarly, both treatments caused increases in the associated proteins AChR-γ, AChR-ε, MuSK, LRP4, and Rapsyn (Fig. 5F–J), Supplementary Fig. 3–5. Differences between SFN and BTX treatments were also observed at the protein level where AChR-γ, AChR-ε, and Rapsyn were higher in SFN-treated limbs compared with BTX-treated limbs.

OPEN IN VIEWER

Figure 5.

RT-qPCR measuring neuromuscular junction end plate genes. (A) Chrng, the immature γ subunit of the acetylcholine receptor, increased in SFN. (B) Similarly, the mature ε subunit of the acetylcholine receptor encoded by Chrne increased in SFN and was expressed at higher levels than BTX-treated limbs. (C) Musk, (D) Lrp4, and (E) Rapsn all increased in SFN- and BTX-treated limbs over their contralateral group and the sham group. The protein levels of AChR- γ, AChR-ε, MuSK, LRP4, and Rapsyn all increased in both treated limbs over their respective contralateral group and sham. (F–J) Bars with different letters indicate a significant difference (e.g., A vs. B) (p < 0.05), while bars containing the same letter are not significant (e.g., AB vs. B) using a one-way ANOVA with a Tukey’s post hoc test. Data are shown as mean ± SEM (n = 7/limb). LRP4, lipoprotein receptor-related protein 4; AChRs, acetylcholine receptors; MuSK, muscle-specific kinase.

ECM protein level

After 52 days of paralysis, collagen types I (Fig. 6A, Supplementary Fig. 6) and IV (Fig. 6C, Supplementary Fig. 7) increased in SFN-treated limbs twofold than in contralateral and sham controls. Collagen type III levels were not different between any group (Fig. 6B, Supplementary Fig. 6). In contrast, collagen type VI (Fig. 6D, Supplementary Fig. 7) increased in both SFN- and BTX-treated limbs compared with its contralateral and sham groups. In SFN-treated limbs, laminin and fibronectin levels increased compared with all other groups (Fig. 6E, F, Supplementary Fig. 8), while chondroitin sulfate proteoglycan was not statistically different in treated limbs compared with the respective contralateral and sham controls (Fig. 6G, Supplementary Fig. 9). Heparan sulfate proteoglycan was upregulated in BTX-treated limbs compared to its contralateral but was not different compared with sham (Fig. 6H, Supplementary Fig. 9).

OPEN IN VIEWER

Figure 6.

Protein blotting analysis probing for key ECM proteins confirmed histomorphometrics and showed distinct changes in the composition of fibrosis between BTX and SFN treatments. Collagen types were differentially produced between SFN and BTX treatments, with collagen type 1 increasing only in SFN (A), collagen type III levels being unaffected (B), collagen type IV also only increasing in SFN (C), but collagen type VI increased in limbs in both treatments (D). Furthermore, laminin (E) and fibronectin (F) only increased in SFN-treated limbs. Chondroitin sulfate (G) and heparan (H) sulfate proteoglycan deposition was varied with only heparan sulfate increasing in BTX over the BTX contralateral limb. Bars with different letters indicate a significant difference (e.g., A vs. B) (p < 0.05), while bars containing the same letter are not significant (e.g., AB vs. B) using a one-way ANOVA with a Tukey’s post hoc test. Data are shown as mean ± SEM (n = 7/limb). ECM, extracellular matrix; SFN, Sciatic-femoral neurectomy.

Fibrosis and inflammatory mRNA level

In healthy tissues, fibrotic and inflammatory-related markers such as interleukin-, TGF-β-, and NF-κB-related mRNAs were maintained at low basal levels. Interleukin IL1A was not significantly different between groups, while IL1B and IL12B were increased in SFN and BTX (Fig. 7A–C). IL6 was not different between groups, while IL33 was twofold higher in SFN compared with BTX (Fig. 7E). Tgfb1 expressed 10-fold higher in SFN compared with sham, while BTX was fourfold higher (Fig. 7F). Tgfb2 and Tgfbr1 were not different between treatment limbs but were increased compared with sham (Fig. 7G, H). Akt1 and Ctgf increased in both treatment groups, and SFN was highest (Fig. 7I, J). Nfkb1 only increased in SFN compared with sham and BTX (Fig. 7K). Ikbkg, Nfkbia, and Ikbkb increased in both treatment limbs compared with sham (Fig. 7L–N). Traf6 increased in both treatment groups, but was higher in SFN compared with BTX (Fig. 7O).

OPEN IN VIEWER

Figure 7.

Overview of the inflammatory and fibrosis RT-qPCR displaying genes associated with interleukin, TGF-β, and NF-κB pathways. Groups shown are sham control, Botox-treated limbs, and SFN-treated limbs. Contralateral limbs were removed from the analysis as they showed no difference from sham. Bars with different letters indicate a significant difference (e.g., A vs. B) (p < 0.05), while bars containing the same letter are not significant (e.g., AB vs. B) using a one-way ANOVA with a Tukey’s post-hoc test. Data are shown as mean ± SEM (n = 7/limb). Botox, botulinum toxin A; SFN, Sciatic-femoral neurectomy.

Of the 71 total genes, 66 were detected and 48 were significantly different compared with sham between BTX and SFN. Of the 48 significantly expressed, 19 were shared between both treatment limbs, 28 uniquely significant in SFN, and only 1 was unique to BTX. The heatmap shows qualitatively the differences in expression compared with sham (Fig. 8A), while the Venn diagram breaks down the differences in significant gene expression (Fig. 8B). Edn1 was increased twofold higher above Sham and SFN and was unique to BTX (Fig. 8C). The full gene panel results are shown in Supplementary Figures S10–S12.

OPEN IN VIEWER

Figure 8.

Theoretical signaling pathway of endothelin-1 (EDN1) causing activation of β-arrestin may be responsible in mitigating inflammation and fibrogenesis in Botox-treated limbs compared with the neurectomy model, despite the two models having similar atrophy and functional deficits. Twenty-eight genes are significantly expressed higher in SFN-treated limbs, and 19 genes are increased in both treatments, while only EDN1 was significantly increased in BTX-treated limbs (A–C). A general pathway overview of the hypothesized role of ET-1 (EDN1) to β-arrestin activation and cytokine binding to the toll-like receptor (D). Botox, botulinum toxin A.