Arntl deficiency in myeloid cells reduces neutrophil recruitment and delays skeletal muscle repair

After a muscle injury, a process comprising inflammation, repair, and regeneration must occur in a time-sensitive manner for skeletal muscle to be adequately repaired and regenerated. This complex process is assumed to be controlled by various myeloid cell types, including monocytes and macrophages, though the mechanism is not fully understood. Aryl hydrocarbon receptor nuclear translocator-like (Arntl or Bmal1) is a transcription factor that controls the circadian rhythm and has been implicated in regulating myeloid cell functions. In the present study, we generated myeloid cell-specific Arntl conditional knockout (cKO) mice to assess the role of Arntl expressed in myeloid cell populations during the repair process after muscle injury. Myeloid cell-specific Arntl deletion impaired muscle regeneration after cardiotoxin injection. Flow cytometric analyses revealed that, in cKO mice, the numbers of infiltrating neutrophils and Ly6Chi monocytes within the injured site were reduced on days 1 and 2, respectively, after muscle injury. Moreover, neutrophil migration and the numbers of circulating monocytes were significantly reduced in cKO mice, which suggests these effects may account, at least in part, for the impaired regeneration. These findings suggest that Arntl, expressed in the myeloid lineage regulates neutrophil and monocyte recruitment and is therefore required for skeletal muscle regeneration.


Results
Arntl expression in myeloid cells is essential for muscle regeneration. To test whether Arntl expression in macrophages is required for skeletal muscle regeneration, we crossed a mouse carrying floxed exons 6-8 in the Arntl gene locus with the Lyz2Cre line 24 , which expresses Cre-recombinase in their myeloid lineage 25 (Fig. 1a,b). Western blotting using bone marrow-derived macrophages and peritoneal exudate cells confirmed Arntl deletion in Lyz2Cre +/− Arntl flox/flox (cKO) mice (Fig. 1c-e). Although systemically Arntl-deficient mice show a decrease in skeletal muscle weight per total body weight despite a decrease in total body weight 19 , deletion of Arntl from myeloid cells did not affect body weight (Fig. S1a), TA muscle weight per body weight (Fig. S1b), or skeletal muscle fiber diameters (Fig. S1c-f).
We then induced skeletal muscle injury in Ctrl and cKO mice by injecting cardiotoxin (CTX) into the tibialis anterior (TA) muscle and performed histological analyses of the post-injury inflammatory and regenerative processes ( Fig. 2a-h). Intramuscular injection of CTX causes rapid muscle degeneration followed by recruitment of inflammatory cells and muscle regeneration at the injury site 5,26 . On days 1 and 2 after injury, swollen necrotic fibers 2,4 and infiltrating inflammatory cells were observed within the TA muscle in Ctrl mice (Fig. 2a,c). During the same period, cKO mice showed less inflammatory cell infiltration than Ctrl mice (Fig. 2b,d). By day 4 after injury in Ctrl mice, most parts of the necrotic fibers had been phagocytized by immune cells 27 , whereas a substantial numbers of necrotic myofibers remained in the cKO muscle 4 days after injury (Fig. 2e,f,i). By day 7, the phagocytized myofibers had been replaced by regenerating fibers with central nuclei in the Ctrl muscle (Fig. 2g). The number of regenerating muscle fibers with central nuclei tended to be higher in cKO fibers, but the increase did not reach the level of significance as compared to Ctrl (Fig. S2). The mean diameters of regenerating fibers were 9.3% smaller in cKO than Ctrl mice (Fig. 2h,j). Analysis of the distribution of myofiber diameters confirmed that fiber size was smaller in cKO than Ctrl mice (Fig. 2k). Immunofluorescent staining of TA muscles on day 7 after injury showed that the expression of MYH3, a marker of immature myofibers 28,29 , was higher in cKO than Ctrl mice (Fig. 2l-n). Additionally, expression levels of creatine kinase, muscle (Ckm) and myomesin 2 (Myom2), two markers of mature regenerating myofibers [30][31][32] , were lower in cKO than Ctrl mice on day 7 (Fig. 2o). These results suggest that regeneration is delayed in myeloid cell-specific Arntl-deficient mice.
It was previously suggested that Arntl deficiency leads to increased induction of inflammatory cytokines via TLR4 activation in macrophages [11][12][13][14] . To determine whether that effect was reproduced in our Arntl deficient macrophages, we analyzed cultured peritoneal macrophages stimulated with 100 ng/mL KLA. The observed higher mRNA and protein expression of Il6 in cKO than Ctrl macrophages recapitulated the phenotype reported in earlier studies 13,14 (Fig. S3a,b). We then isolated myeloid cells (CD45 + CD11b + ) from the injured muscles of cKO and Ctrl mice on day 3 after injury and compared the Il6 mRNA expression, and the expression was increased in the cKO myeloid cells (Fig. S3c). Previous studies found that IL6 produced by inflammatory monocytes and macrophages at an early phase of injury is essential for activating muscle stem cells and initiating the regenerative process 8,10,33 . In Arntl-cKO mice, therefore, the sustained IL6 production may have inhibited muscle regeneration. To test that hypothesis, we further investigated the crosstalk between macrophages and muscle stem cells in vitro. Peritoneal macrophages were obtained from cKO and Ctrl mice and cultured, after which the conditioned media were collected. Muscle stem cells collected from wild-type mice were then cultured with the conditioned media for 48 h. The results showed that muscle stem cell differentiation was decreased in the medium conditioned by cKO macrophages (Fig. S4), which suggests Arntl deficiency may alter the phenotype and function of macrophages.
Arntl deletion in myeloid cells reduces neutrophil and monocyte recruitment into injured muscles. Given the histological result showing reduced inflammatory cell infiltration of the injured area in cKO mice, we next addressed which subpopulations of inflammatory cells were affected by the deletion of Arntl from myeloid cells. Flow cytometric analysis of the intramuscular myeloid cells within injured TA muscles revealed that, in Ctrl mice, the numbers of neutrophils (CD45 + CD11b + Ly6G + ) were markedly increased after injury, peaking 1 day post-injury, and that the numbers of infiltrating neutrophils were lower in cKO than Ctrl mice on day 1 (Fig. 3a). In Ctrl mice, numbers of monocytes (CD45 + CD11b + Ly6G -Ly6C hi ) continued to increase on days 2 and 4, but had decreased significantly by day 7 (Fig. 3b). As with neutrophils, the numbers of monocytes infiltrating the injured TA muscle were lower in cKO than Ctrl mice (Fig. 3b). The number of Ly6C lo macrophages (CD45 + CD11b + Ly6G -Ly6C lo ) increased on day 4 in both Ctrl and cKO, but there were no significant differences between genotypes (Fig. 3c). Thus, Arntl deficiency in myeloid cells limited muscle recruitment of neutrophils and monocytes in response to acute injury.
Arntl is needed for neutrophil chemotaxis during skeletal muscle injury. Within the bone marrow, neutrophils undergo differentiation from granulocyte-monocyte progenitors, after which the fully differen-  34 .While myeloid Arntl deficiency is known to abolish rhythmic migration during systemic inflammation 35 , the involvement of Arntl in neutrophils mobilization from bone marrow in response to acute skeletal muscle inflammation is not clear from previous reports. In the present study, flow cytometry revealed that neutrophil numbers in bone marrow and the peripheral blood did not differ significantly between cKO and Ctrl mice (Fig. S5a-c). Neutrophils are recruited to sites of injury through interactions between chemokines (C-X-C motif) ligand 1 (Cxcl1) 36 and ligand 2 (Cxcl2) 37 and their receptor, chemokine (C-X-C motif) receptor 2 (Cxcr2). When we analyzed whole TA muscle tissue to test reduced expression of any of these mediators could account for the reduced neutrophil recruitment seen cKO mice, we detected no significant difference in Cxcl1 or Cxcl2 expression between cKO and Ctrl muscle (Fig. S6a, p > 0.05). Using neutrophils isolated from bone marrow, we also tested whether Arntl deletion affected neutrophil migration by altering expression of Cxcr2 38 . We found that Cxcr2 expression of was significantly lower in cKO than Ctrl neutrophils (Fig. 4a,b), which would be expected to decrease neutrophil recruitment to the chemokines 39,40 . To assess changes in neutrophil migration due to Arntl deficiency, we performed chemotaxis assays in vitro, which confirmed that chemotaxis to CXCL2 was reduced in cKO neutrophils (Fig. 4c). It thus Arntl is needed to increase the number of circulatory monocytes. Monocytes are produced in the bone marrow from hematopoietic stem cells through multiple processes of development and commitment as follows: common myeloid progenitors, granulocyte and macrophage progenitors, macrophage-dendric cell progenitors, common monocyte progenitors and mature monocytes 41 . Expression of Lyz2 is also observed in macrophage-dendritic cell progenitors as well as common monocyte progenitors 42 . This suggests, Lyz2-Credependent Arntl deficiency could potentially inhibit normal development from macrophage-dendritic cell and common monocyte progenitors to mature monocytes in the bone marrow. Our flow cytometric analysis showed  . Arntl deletion from myeloid cells reduces the numbers of neutrophils and monocytes infiltrating injured muscle. (a) Quantification of neutrophils (CD45 + CD11b + Ly6G + ) in untreated Ctrl (n = 4) and cKO (n = 4) TA muscles and in Ctrl (n = 4) and cKO (n = 4) muscles on day 1, Ctrl (n = 10) and cKO (n = 8) muscles on day 2, Ctrl (n = 7) and cKO (n = 8) muscles on day 4, and Ctrl (n = 6) and cKO (n = 4) muscles on day 7 after CTX injury. (b) Quantification of monocytes (CD45 + CD11b + Ly6G -Ly6C hi ) in untreated Ctrl (n = 4) and cKO (n = 4) TA muscles and in Ctrl (n = 4) and cKO (n = 4) muscles on day 1, Ctrl (n = 10) and cKO (n = 8) muscles on day 2, Ctrl (n = 7) and cKO (n = 8) muscles on day 4, and Ctrl (n = 6) and cKO (n = 4) muscles on day 7 after CTX injury. (c) Quantification of macrophages (CD45 + CD11b + Ly6G -Ly6C lo ) in untreated Ctrl (n = 4) and cKO (n = 4) TA muscles and in Ctrl (n = 4) and cKO (n = 4) muscles on day 1, Ctrl (n = 10) and cKO (n = 8) muscles on day 2, Ctrl (n = 7) and cKO (n = 8) muscles on day 4, and Ctrl (n = 6) and cKO (n = 4) muscles on day 7 after CTX injury. Data are expressed as the means ± SEM. * p < 0.05 vs. untreated Ctrl muscles, # p < 0.05 vs. Ctrl muscles per day with two-way ANOVA with Bonferroni's multiple comparisons test. www.nature.com/scientificreports/ that, as previously reported 3,6 , numbers of monocytes in the bone marrow and blood were increased after muscle injury in Ctrl mice (Fig. 4d,e). On the other hand, the numbers of Ly6C hi monocytes in the bone marrow and blood were lower in cKO than Ctrl mice both before and 2 days after injury (Fig. 4d,e). Within injured tissue, chemokine (C-C motif) ligand 2 (Ccl2) 43 and chemokine (C-X3-C motif) ligand 1 (Cx3cl1) 44 are important mediators of monocyte/macrophage recruitment and infiltration into injury sites 3,45 . Levels of their mRNA expression were similar between cKO and Ctrl muscles (Fig. S6b, p > 0.05). Likewise, monocyte 46 migration assays performed with monocytes isolated from bone marrow revealed no significant difference in the expression levels of Ccr2 and Cx3cr1 between Ctrl and cKO monocytes (Fig. S5d). This suggests reductions in the numbers of bone marrow and circulating monocytes, not changes in the expression of chemokines or their receptors, is the primary mechanism contributing to the reduced recruitment of monocytes to regenerating muscle in cKO mice.  -test (a, b, d, e)

Discussion
Time-dependent infiltration of myeloid cells plays a critical role in skeletal muscle regeneration. To the best of our knowledge, the present study is the first investigation to examine skeletal muscle regeneration in myeloidcell-specific Arntl-deficient mice. Our findings show that Arntl expressed in myeloid cells is essential for normal skeletal muscle regeneration and that Arntl knockout in myeloid cells (1) impairs infiltration by neutrophils and monocytes into the muscle injury site; (2) suppresses expression of Cxcr2 in neutrophils in the bone marrow, thereby reducing their chemotactic migration; and (3) leads to a reduction in the numbers of Ly6C hi monocytes in the bone marrow. These results suggest that Arntl in myeloid cells plays an essential role in neutrophil and monocyte infiltration for skeletal muscle regeneration.
Cxcr2 is a major chemokine receptor expressed by neutrophils 38 . Reduced neutrophil recruitment in Cxcr2deficient models has been linked to delays in bacterial clearance and tissue repair 47,48 . We detected decreased numbers of infiltrating cKO neutrophils at sites of muscle injury, decreased expression of Cxcr2 in neutrophils isolated from the bone marrow of cKO mice, and decreased chemotaxis of cKO neutrophils. These results are consistent with the reduction in neutrophil infiltration seen in Cxcr2-deficient mice and suggest that decreased expression of Cxcr2 in cKO neutrophils contributed to the reduction in neutrophil infiltration into damaged muscle tissue we observed in cKO mice.
Ccr2 and Cx3cr1 are major chemokine receptors mediating tissue infiltration by monocytes/macrophages 46 . We detected no significant difference in levels of Ccr2 and Cx3cr1 expression in bone marrow monocytes between cKO and Ctrl mice. This may indicate that factors other than monocyte Ccr2 and Cx3cr1 expression contribute to the suppression of monocyte infiltration into skeletal muscle after injury in myeloid Arntl-deficient mice. The numbers of monocytes in bone marrow increases after muscle injury 6,49 , and a lack of bone marrow monocytes leads to a decrease in monocytes infiltrating muscle tissue 49 . In cKO mice, we observed a decrease in the numbers of monocytes in the bone marrow. It follows then that this reduction in the numbers of bone marrow monocytes likely led to reduced recruitment of monocytes to muscle tissue in cKO mice. Monocytes are produced in bone marrow from hematopoietic stem cells through multiple processes of development and commitment 41 . Expression of Lyz2 is also observed in macrophage-dendritic cell and common monocyte progenitors, which are middle stages of monocyte differentiation 42 . Lyz2-Cre-dependent Arntl deficiency may therefore inhibit normal monocyte production in the bone marrow. However, the present study provides no direct evidence of the critically affected process steps.
Depleting circulating monocytes in mouse models using clodronate 50,51 or CD11b diphtheria toxin receptor 52,53 has been shown to reduce inflammatory cell infiltration during the early stages of muscle regeneration, which causes necrotic fibers to persist and impairs muscle regeneration 8,49,54 . In the present study, cKO mice exhibited decreased infiltration by monocytes into injured muscle on day 2 after injury and the persistent presence of necrotic fibers on day 4 after injury. These results are consistent with muscle regeneration in a circulating monocyte depletion model and suggest that the impaired muscle regeneration seen in cKO mice reflects the reduction in monocyte infiltration. In the present study, however, Arntl deletion in myeloid cells did not lead to retention of necrotic fibers to the late stage after injury and did not significantly delay in muscle regeneration, though it decreased the number of circulating monocytes. In contrast to a general monocyte depletion model, in which circulating monocytes are reduced by 80-90% 8,54 , Arntl deficiency induced only a 20% reduction in the circulating monocyte in the present study (Fig. 4d). That numbers of circulating monocytes were not dramatically reduced may account the lack of a dramatic delay in cKO muscle regeneration.
Previous studies demonstrated that Arntl deletion from macrophages increases production of pro-inflammatory cytokines via TLR4 activation [11][12][13][14] . Our earlier report using cultured macrophages confirmed that loss of Arntl causes a prolonged inflammatory response in macrophages and delays its convergence 15 . Similarly, we observed that in vitro and in vivo Il6 expression was increased in our cKO macrophages. Many of the cytokines expressed in cultured macrophages, including IL6, are also reportedly expressed in macrophages in muscle tissue during regeneration and are involved in regenerative regulation 8,10 . These findings suggest there are similarities between cultured Arntl-cKO macrophages and regenerating skeletal muscle macrophages. IL6 acts mainly on the activation and proliferation of muscle stem cells during muscle regeneration 33,55 . The result of experiment in which muscle stem cell cultured in conditioned medium from Ctrl and cKO peritoneal macrophages confirmed that the Arntl-cKO macrophage-conditioned medium decreases muscle differentiation (Fig. S4). These results suggest that, apart from suppressing myeloid cell infiltration, the lack of convergence of macrophage-derived IL6 production may inhibit eventual skeletal muscle regeneration.
Because granulocytes, monocytes, and macrophages are affected in the Lyz2cre-deficient model, it is difficult to determine which cell type is primarily responsible for the observed inflammatory phenotype 25,56 . Neutrophils are known to contribute to the phagocytic removal of tissue debris and recruitment of monocytes [57][58][59] . Whether they are necessary for monocyte infiltration after muscle injury remains controversial, however 60,61 . The present study does not clarify whether the observed reduction in monocyte infiltration into skeletal muscle reflects a reduction in neutrophil infiltration. More detailed studies of specific cell types will be required to assess the contribution of Arntl to the recruitment of each myeloid cell to damaged skeletal muscle.
In conclusion, we have discovered that Arntl plays an essential role in skeletal muscle tissue regeneration after injury by regulating the appropriate timing of the gradual infiltration of neutrophils and monocytes into the injured muscle early during regeneration. The results of this study, in which the peripheral clocks of myeloid cells failed to function appropriately and inhibited muscle regeneration suggests that this mechanism may also contribute to the muscle atrophy that occurs in the space where circadian rhythms is disrupted. Animals. Eight-to 12-week-old mice, housed at 22ºC under a 12-h light:dark cycle and fed ad libitum, were used in these experiments. Arntl flox/flox mice were generated on a C57BL/6 background as described previously 24 .

Muscle regeneration.
To induce muscle injury, the mice anesthetized with isoflurane (1.5-2%) (Pfizer, NY, USA) and 100 μl of 10 mM CTX (Sigma-Aldrich, St. Louis, MO, USA) were injected into the TA muscle of anesthetized mice using a 29 G syringe at 6 pm (ZT 10). TA muscles were then collected at 8 am (ZT 0) 1, 2, 4, or 7 days after CTX injection, fixed by immersion in Tissue-Tek Ufix (Sakura finetek, Tokyo, Japan), embedded in paraffin blocks, and cut into 10-μm-thick sections. The sections were then deparaffinized, rehydrated, stained with hematoxylin for 3 min, washed with running water for 15 min, stained in eosin for 15 min, and quickly washed in a 70, 80, 90, 95, and 100% ethanol series before finally washing twice in xylene. Images of hematoxylin/eosin-stained sections were then acquired using a microscope (BZ-X810, Keyence, Osaka, Japan) and analyzed to assess the minor fiber axis and to quantify regenerating and necrotic fibers. The numbers of regenerating fibers that had centrally located myonuclei were counted, and the minor axis was measured in 2000-3000 regenerating fibers in each mouse. Necrotic fibers were identified as round myofibers lacking the centrally-located nuclei prevalent in regenerated myofibers 62,63 . The necrotic fiber area was measured as the sum of the areas of the necrotic fibers in each section.
Migration assays. Neutrophil migration was assessed using 24-well microchambers and polycarbonate filters (5 μm pore size) (Corning, NY, USA) as described previously 40,64 . In brief, CD45 + CD11b + Ly6G + neutrophils sorted from mouse bone marrow cells were placed in the upper wells (1 × 10 5 cells/well) of Transwell chambers, and 600 µL of RPMI 1640 medium with or without 100 ng/ml CXCL2 (BioLegend) were added to the lower wells. For migration assays, cells were incubated for 60 min at 37 °C. After the incubation, cells that migrated to the bottom part of the membrane and the lower wells was stained with Hoechst 33,342. The numbers of cells per field were counted in 10 randomly selected visual fields under a microscope (BZ-X810, Keyence, Osaka, Japan), and the mean estimate for individual samples was calculated.
Quantitative RT-PCR. Total RNA was isolated from homogenized TA muscles or cultured cells using ISO-GEN (Nippon Gene, Tokyo, Japan). The RNA was isolated using the phenol-chloroform extraction and isopropanol precipitation protocol according to the manufacturer's instructions. Total RNA was extracted from sorted cells using a RNeasy Mini Kit (Qiagen, Valencia, CA, USA). Complementary DNA (cDNA) was synthesized using ReverTra Ace qPCR RT Master Mix with genomic DNA Remover (TOYOBO CO., LTD., Osaka, Japan). cDNA was analyzed with real-time PCR in a QuantStudio 5 Real-time PCR system (Applied Biosystems, Foster City, CA, USA) using PowerUp SYBR Green Master Mix (Applied Biosystems). The primers used for qPCR are listed in Table 1. Glyceraldehyde-3-phosphate dehydrogenase (Gapdh) expression was used as an internal control. For preparation of macrophage-conditioned medium, peritoneal macrophages were cultured first in RPMI-1640 medium containing 10% FBS at 37 °C overnight and then RPMI-1640 without FBS for an additional 24 h. Muscle stem cells were cultured for 2 days in new conditioned differentiation medium composed of macrophageconditioned and differentiation medium as described previously 67 .
For myotube differentiation assays, myotubes were fixed in 4% paraformaldehyde solution and then inoculated with anti-skeletal muscle myosin antibody (MYH, clone F59, Santa Cruz) and stained with Hoechst 33342. Images of myotubes were acquired from randomly selected visual fields under a microscope (BZ-X810, Keyence). Ratio of nuclei within myotubes in individual samples were counted as a fusion index 68 . Immunoblotting. Protein was extracted from cells using RIPA buffer and quantified using a BCA Protein assay kit (Pierce, Rockford, IL) following the manufacturer's protocol. Sample were then mixed with SDS loading buffer, electrophoresed on 10% (vol/vol) acrylamide gels, transferred onto PVDF membranes, and blocked with 5% milk for 1 h before incubation with a primary antibody. The antibodies used for western blotting were rabbit anti-BMAL1 (ab93806, Abcam, Cambridge, UK) and anti-β-tubulin (clone 10G10, Wako, Osaka, Japan). The blots were then developed using ECL Prime detection reagent (GE, Waukesha, WI, USA). Raw images are in Fig. S8. Statistical analysis. Data are presented as means ± SEM, except where otherwise indicated. Sample sizes were not based on power calculations. Statistical significance is determined using the two-tailed Student's t-test. Two-way ANOVA with post-hoc Bonferroni's multiple comparison test was used for experiments involving two  www.nature.com/scientificreports/