IL-1β-mediated macrophage inltration into the brain after peripheral tissue injury

Background: Inltration of macrophages into the central nervous system (CNS) is involved in many neurological disorders, such as Alzheimer’s disease (AD), amyotrophic lateral sclerosis (ALS) and autism. Despite extensive studies into neuroinammation associated macrophage inltration into the CNS, its underlying mechanisms and pathological roles remain unclear, especially when triggered by peripheral inammation. Methods: To further elucidate the role and mechanism of peripheral inammation in neurological disorders, we exploited interleukin 1 beta (il1b) mutant transgenic zebrash (Danio rerio) with uorescent protein expression restricted to macrophages to track the macrophage migration under peripheral inammation following tail amputation. Results: We found that macrophage inltration into the brain of zebrash embryo following peripheral tissue injury can be alleviated via genetically targeting il1b. In addition, through circulation-independent migration, macrophages inltrate brains with evidence of increased apoptosis. We further identied the expression of camk2g1 in the brains of zebrash with hyperactive behavior following peripheral tissue injury. This il1b-regulated protein is associated with neuropsychiatry disorders. Conclusion: These ndings demonstrated that peripheral tissue injury induces il1b-mediated macrophage inltration into the brain and a hyperactive behavior.

neurological disorders. For instance, these invading macrophages mediate cognitive dysfunction as well as seizures [11][12][13] but can also contribute to the removal of beta-amyloid plaques, neuroprotection and tissue repair in the CNS [14][15][16]. Despite these far-reaching roles of in ltrating macrophages in in ammatory neurological disorders, mechanisms regulating their recruitment and pathogenesis remain elusive.
Systemic in ammation is a pathological process characterized by activation of the innate immune system in response to peripheral mechanical trauma, infection/sepsis, or surgery [17]. Systemic in ammation elicits neuroin ammation and serves as a contributing factor to multiple neurological disorders, including post-traumatic stress disorder (PTSD), post-operative cognitive dysfunction (POCD) and other neurodegenerative diseases [18][19][20]. Different experimental models of systemic in ammation, such as surgery (laparotomy), lipopolysaccharides (LPS) or live bacteria injection have been used to elicit neuroin ammation and cognitive dysfunction [21][22][23]. Furthermore, in ltration of peripheral macrophages into the brain was reported previously in rodent models of in ammatory liver injury and AD with arthritis [24,25]. However, systemic and neuroimmune responses stimulated by sterile in ammation such as mechanical trauma is not well understood.
Here, we used live cells tracking and genetic manipulation in zebra sh (Danio rerio) and demonstrated that il1b mediated peripheral macrophage in ltration into the brain following peripheral tissue injury and consequently contributed to a hyperactive behavior.
Generation of il1b mutant zebra sh by TALEN il1b mutant zebra sh was generated via TALEN resulting in a loss of function as previously described [28]. Brie y, plasmid containing mRNA sequence of TALEN left or right arm were constructed separately using the FusX assembly system according to the design (Additional File 1: Fig. S1A). Equal volume of in vitro transcribed mRNAs of TALEN left and right arms were mixed and co-injected into the one-cell-stage zebra sh embryos. F1, F2, and F3 zebra sh embryos were obtained by outcrossing founder (F0) with WT zebra sh, incrossing F1 zebra sh and incrossing F2 zebra sh, respectively. Moreover, il1b Mut reporter lines, such as Tg(kdrl:GFP;coro1a:DsRed), was generated by crossing homozygous il1b Mut with reporter lines. All il1b mutation was con rmed by restriction fragment length polymorphism (RFLP) assay using il1b genotyping primer (listed in Additional File 1: Table S1) and NsiI-HF® restriction enzymes (NEB) as well as sanger sequencing (Additional File 1: Fig. S1B and S1C).
Modeling Peripheral Tissue Injury in Zebra sh Embryo WT or transgenic zebra sh embryos at 3 dpf were anaesthetized using 0.16 mg/ml tricaine (Sigma-Aldrich) and then a tail amputation assay was applied in the position posterior to blood circulation of the caudal n to model the peripheral tissue injury as previously described [29]. After amputation, zebra sh embryos were immediately transferred back to sterilized E3 medium (5 mM NaCl, 0.17 mM KCL, 0.33 mM CaCl, and 0.33 mM MgSO4, pH 7.4) containing methylene blue.

RNA Extraction and Quantitative PCR
Total RNA was extracted from the tail region or the head region of zebra sh embryos using RNAiso Plus (Takara). cDNA was then synthesized using RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher) in accordance with manufacturer's protocol. Quantitative PCR was nally performed on ABI 7300 Real-Time PCR System with FastStart Universal SYBR Green Master (Roche) reagents (primers listed in Additional File 1: Table S1).

Light-sheet Imaging and Flow Cytometry
Zebra sh embryos at 4-6 dpf were anesthetized using 0.16 mg/ml tricaine (Sigma-Aldrich) and then mounted in 1.5% low-melting agarose using glass capillary for uorescence imaging by Zeiss Lightsheet Z.1 Selective Plane Illumination Microscope with or without a time-lapse mode. For ow cytometry, the head region of WT or Tg(mepg1:GFP) zebra sh embryos was isolated and then homogenized and ltered to get the cells suspension. Beckman Coulter FC 500 was utilized to quantify the number and percentage of GFP + cells under various conditions. Acridine Orange, TUNEL and PH3 Staining Acridine orange at 10 µg/ml (Sigma-Aldrich) and ApopTag® Fluorescein In Situ Apoptosis Detection Kit (Millipore) were used to detect apoptotic cells in live and xed zebra sh embryos, respectively, following the protocols described previously [32]. In addition, immunostaining together with phospho-Histone H3 (PH3) primary antibody (Cell Signaling Technology) was applied to detect the cells undergoing mitosis according to the previously described protocol [33].

Mass Spectrometry-based Proteomics
Total protein was extracted using cell lysis buffer (Sigma-Aldrich) from the head region of zebra sh embryos. After puri cation and trypsin (Promega) treatment, the peptides were desalted using Pierce C18 Spin Columns (Thermo sher). Proteomics was then performed on Thermo Fisher Orbitrap Fusion Lumos Mass Spectrometer coupled with Dionex UltiMate 3000 RSLCnano. Label-free relative quanti cation was processed with Progenesis QI software (http://www.nonlinear.com/progenesis/qi/) and the abundance of various proteins was quanti ed based on three independent experiments and normalized according to the housekeeping protein (gapdh).
Behavioral Assay of Locomotor Activity il1b Mut Zebra sh larvae at 7 dpf/4 dpa and their siblings with or without amputation were transferred into the 35 mm Petri dish containing 2 mL E3 medium. The protocol of locomotor activity was adapted from the Locomotion Assay with slight modi cation [34]. Zebra sh larvae from four groups was recorded for a 30-min period using a camera at the same time. A 10-min period after 10-min acclimation was analyzed using Fiji -Image J (https://imagej.net/Fiji) with tracking plugin and the total travel distance, active swim time and mean velocity were calculated.

Statistical analysis
Data are presented as mean (M) ± standard deviation (S.D.). Unpaired t-test, one-way analysis of variance (ANOVA), and two-way ANOVA with Tukey's HSD with were performed where appropriate using Statistical Package for the Social Sciences (SPSS) Version 14.0 and a p-value less than 0.05 was considered statistically signi cant.

Results
Peripheral tissue injury induced neuroin ammation, macrophage in ltration and apoptosis in the brain Peripheral tissue injury from tail amputation in zebra sh embryos is widely utilized to investigate acute and sterile innate immune responses to the mechanical trauma-induced in ammation [26]. Tail amputation was performed at 72 hours post-fertilization (hpf), and at 24 hours post-amputation (hpa), elevated in ammatory cytokines/chemokines, including il1b, il34, il6, tnfa and ccl2 but not il8 and il10 mRNA, were detected in the tail region, indicating a peripheral/systemic in ammation (Fig. 1A). Increased mRNA expression of il1b, il34, il6, il8, il10 but not tnfa, and decreased mRNA expression of ccl2 were detected in the head region at 24 hpa, suggesting neuroin ammation ( Fig. 1B and 1C). The number of leukocytes labeled with coro1a (coro1a+), largely macrophages/microglia and neutrophils [35], were markedly increased in the brain at 4 days post-fertilization (dpf)/24 hpa, 5 dpf/48 hpa and 6 dpf/72 hpa ( Fig. 1D-1E). Speci cally, the number of mpeg1-expressing (mpeg1+) macrophages increased in the brain, while no difference in the number of mpx-expressing (mpx+) neutrophils was found in the brain after amputation ( Fig. 1F-1H). Similarly elevated number of macrophages in the brain was also identi ed in acute response to amputation at 3 hpa and 6 hpa due to a redistribution, where mpeg1 + macrophages were recruited from caudal hematopoietic tissue (CHT) and aorta-gonad-mesonephros (AGM) to the brain and injury site while the total number of macrophages remained unchanged (Additional File 1: Fig. S2A-S2C). To investigate the potential mechanism underlying recruitment of macrophages to the brain, apoptosis was also examined in the brain as circulatory macrophages are more susceptible than microglia to cell death and apoptosis during CNS injury [36]. Acridine orange-labeled (AO+) apoptotic cell phagocytosed by coro1a + macrophages/microglia increased in the brain at 24 hpa ( Fig. 1I and 1J). Collectively, peripheral tissue injury elicited neuroin ammation, in ltration of macrophages and apoptosis in the brain under systemic in ammation.
il1b mediated in ltration of macrophages into the brain in response to peripheral tissue injury To de ne the role of systemic in ammation in peripheral tissue injury, a transcription activator-like effector nucleases (TALEN)-mediated mutant line of il1b [28] was established (Additional File 1: Fig. S1A-S1C). While somatic and stable (il1b +/− or il1b −/− ) i1lb mutants displayed normal gross development, an il1b knockout mutation (il1b Mut ) signi cantly decreased mRNA expression of cytokines regulated by il1b, including il6, il8, and tnfa but not il34, il10 and ccl2 in il1b −/− zebra sh embryos (Additional File 1: Fig.   S1D). Basal levels of macrophages/microglia were measured in il1b +/− zebra sh before tail amputation at 72 hpf and the number of pu.1-labled myeloid progenitors in CHT remained unchanged (Additional File 1: Fig. S1E and S1F). In addition, the number of neutral red-labeled (neutral red+; lysosome marker) microglia mildly decreased in both il1b +/− and il1b −/− zebra sh embryos compared with the il1b +/+ siblings (Additional File 1: Fig. S1G-S1H). However, the number of mpeg1 + macrophages/microglia in the brain and CHT as well as the mRNA expression of mpeg1 and mpx in the head remained unchanged in il1b +/− zebra sh embryo compared with wild-type siblings (Additional File 1: Fig. S1I-S1K). While il1b Mut showed no impact on early development of macrophages/microglia in the brain and CHT at 72 hpf, the number of macrophages in CHT at 4 dpf and the number of macrophages/microglia in brain at 5 dpf of heterozygous il1b Mut declined ( Fig. 2A-2E). Upon tail amputation, peripheral tissue injury-induced in ltration of macrophages alleviated in il1b Mut at 5 dpf/2 dpa ( Fig. 2A-2C). Moreover, the number of macrophages in CHT was also reduced after amputation in both heterozygous il1b Mut and wild-type siblings at 1 dpa, possibly due to the redistribution of macrophages ( Fig. 2D and 2E). To separate in ltrating macrophages from microglia in the brain, macrophage-speci c (mfap4) and microglia-speci c (apoeb) markers were used [37]. The number of both mfap4 + and apoeb + cells were elevated in the wildtype siblings but not in homozygous il1b Mut zebra sh after amputation (Fig. 2F-2I). With all these results, it is most likely that il1b mediates in ltration of both macrophages and microglia into the brain of zebra sh embryos after peripheral tissue injury.
Macrophages in ltrated the brain through circulation independent migration after peripheral tissue injury A time-lapse imaging was applied to investigate the migration routes of il1b-mediated macrophages in ltration into the brain at 4 dpf when the microglia ceased to invade the brain during development [32].
Using Tg(kdrl:GFP;coro1a:DsRed) double-transgenic zebra sh line to visualize cerebral blood vessels (endothelial cells) and pan-leukocytes (Fig. 3A), circulation-independent migration of macrophage in ltration into the brain were observed at 4 dpf/1 dpa while no macrophage was found to move out from the blood stream. A schematic diagram was shown to present the distribution of active macrophages around the brain (Fig. 3B). Speci cally, peripheral macrophages located in the neighboring tissues in ltrate the brain through the lateral periphery of the hindbrain after tail amputation. In contrast, such in ltration could not be seen after amputation in CTRL, il1b Mut (both il1b +/− and il1b −/− showing similar results) and il1b Mut + (Additional File 1: Video S1-S4). Macrophages distributed around mid-hindbrain boundary (MHB) failed to in ltrate the brains of zebra sh with or without tail amputation (Fig. 3C-3F, Additional File 1: Video S5). In addition, compared to CTRL (Fig. 3C), more active macrophages/microglia were found within the brain of amputated zebra sh ( Fig. 3D; Additional File 1: Video S6), which might be due to the in ltration of macrophages through the lateral periphery of the hindbrain (Fig. 3D; Additional File 1: Video S7). Therefore, this indicated that peripheral tissue injury induced il1b-regulated in ltration of macrophages from neighboring tissues through the lateral periphery of the hindbrain. Besides macrophages in ltration into the brain through circulation-independent migration, il1b-mediated cerebral blood vessel narrowing was observed. The diameter of cerebral branch vessels reduced while the main vessels increased after tail amputation, which were ameliorated in heterozygous il1b Mut (Additional File 1: Fig. S3A and S3B). As expected, a declined blood ow was detected in the cerebral branch vessels by using Tg(kdrl:GFP;gata1:DsRed) with gata1-labeled (gata1+) erythrocytes (Additional File 1: Fig. S3C). Surprisingly, circulatory macrophages were found to be trapped within the narrow cerebral branch vessels, which might also contribute to an increased number of macrophages in the brain following amputation (Additional File 1: Fig. S3D). While in ltration of macrophages into the brain was accomplished via circulation-independent migration, the response of cerebral blood vessels and blood ow to peripheral tissue injury might participate in other pathological processes, such as apoptosis, that affected in ltration of macrophages to the brain.
Apoptosis but not proliferation contributed to the il1bmediated macrophage number increase in the brain Recruitment of macrophages/microglial is an essential step in the clearance of apoptotic neurons through phagocytosis, where the dying cells release various " nd me" signals [38]. Thus, elevated apoptotic signaling molecules in the brain may contribute to the in ltration of macrophages/microglia. We further examined whether apoptosis is the mechanism underlying il1b-mediated phagocytes recruitment as il1b is a direct triggering factor for neuronal apoptosis [39]. Unlike the increase in apoptotic cells in the brain after tail amputation in wild-type embryos, tail amputation in homozygous il1b Mut zebra sh did not trigger any signi cant increase in the number of AO + and TUNEL-labeled (TUNEL+) apoptotic cells. The results were consistent with the unchanged number of macrophages/microglia in the brain, in which apoptotic cells in the il1b Mut remained the same as in CTRL group (Fig. 4A-4D). Moreover, a reduced number of TUNEL + apoptotic cells were also found in the CHT of zebra sh with tail amputation, probably due to the redistribution of leukocytes, while il1b Mut also resulted in a reduced number of apoptotic cells in CHT (Fig. 4E and 4F). In addition to il1b-induced apoptosis, we observed that neutral red + and LysoTracker + lysosome, an organelle for phagocytosis, decreased in the macrophages/microglia in the brain following tail amputation. Mutation of il1b (il1b −/− ) also led to a reduction of mpeg1+ & LysoTracker + colocalized cells (Additional File 1: Fig. S4A-S4D). Accumulation of lysosomes and macrophages at the injury site was also reduced in heterozygous and homozygous il1b Mut (Additional File 1: Fig. S4E-S4G). Importantly, the LysoSensor + acidi ed lysosomes were reduced only after tail amputation indicating an impaired phagocytosis, and this could contribute to the increased apoptotic cells in the brain (Additional File 1: Fig. S4H-S4I). To eliminate the potential in uence of glial cell proliferation on the increased number of macrophages/microglia in the brain, phospho-Histone H3 (PH3) was utilized to label proliferating cells. PH3 + cells in the midbrain remained unchanged after tail amputation, which indicates that increased number of macrophages/microglia was not due to the proliferation ( Fig. 4G and 4H). On the contrary, the decreased number of macrophages/microglia in homozygous il1b Mut might attribute to the reduced proliferation as shown by the decreased number of PH3 + cells ( Fig. 4G and 4H). In CHT, the number of PH3 + cells decreased after tail amputation in both wild-type siblings and il1b Mut (Fig. 4I and 4J). Collectively, increased apoptosis but not cell proliferation contributed to il1b-meidated in ltration of macrophages/microglia into the brain.

il1b -mediated proteomic changes and disordered behavior in response to peripheral tissue injury
To further investigate the potential effect of peripheral tissue injury on the brain and mechanisms underlying il1b-meidated in ltration of macrophages, proteomic analysis was performed with the head of zebra sh embryo (Fig. 5A). Despite the limited number of proteins identi ed, 29 out of 803 (3.6%) proteins were found in the head showing signi cant changes after tail amputation (AMPU) vs control (CTRL), which included 27 increased and 2 decreased proteins (Fig. 5B). We tracked these 29 proteins in various proteomic analysis including il1b Mut vs CTRL, il1b Mut + AMPU vs il1b Mut , il1b Mut + AMPU vs CTRL (data not shown) and il1b Mut + AMPU vs AMPU (data not shown), and excluded proteins that did not respond to il1b Mut at different conditions (Fig. 5C). Finally, 21 out of the 29 (72%) proteins were found to be regulated by il1b in AMPU, with the expression increased after amputation and alleviated in il1b Mut (Fig. 5D). The function of these proteins was classi ed to explore the effect of peripheral tissue injuryinduced il1b on the brain. In accordance with previous ndings, apoptosis inducing factor mitochondria associated 4 (aifm4) and apolipoprotein Eb (apoeb), the speci c marker of pro-apoptosis and microglia were found. Proteins involved in neuroprotection and other biological processes, such as development and metabolism, in response to il1b-mediated damage to the brain were also identi ed. Importantly, camk2g1, orthologous to human CAMK2G that is involved in multiple psychiatry disorders [40] is also on the list (Fig. 5D). Since the il1b-mediated increased expression of camk2g1 may result in psychiatry disorders, locomotor behavior of zebra sh was recorded at 7 dpf/4 dpa. A hyperactive behavior characterized by increased travel distances and active swim time without affecting average velocity (data not shown) was observed in amputated wild-type siblings but not in homozygous il1b Mut zebra sh embryos. This indicates that peripheral tissue injury induced an il1b-mediated hyperactive behavior ( Fig. 5E and 5F). In summary, peripheral tissue injury induced il1b-mediated induction of camk2g1 in the brain, which together with the other 20 proteins may contribute to the observed hyperactive behavior.

Discussion
In ltration of peripheral macrophages into the CNS has been well documented in various central neurological disorders, such as neurodegenerative diseases, psychiatry disorders and traumatic brain injuries [3][4][5]. However, little is known about the mechanisms responsible for the in ltration and its pathological roles in systemic in ammation-mediated neurological disorders. Here we demonstrated a dynamic interplay of peripheral tissue injury, systemic in ammation, neuroin ammation, apoptosis and macrophage in ltration using a zebra sh model. Most importantly, we demonstrated that il1b is a key regulator in mediating the in ltration of peripheral macrophages into the brain from neighboring tissues through circulation-independent migration, a process which is probably triggered by the accumulation of apoptotic cells in the brain [38]. The proteomic analysis also identi ed an elevation in an il1b-regulated protein, camk2g1 following peripheral tissue injury. This protein is elevated in the brain affected by neuropsychiatry disorders, and its increase accompanied with neuroin ammation, in ltrating macrophages and neuronal apoptosis might contribute to the hyperactive behavior of zebra sh observed in this study [3,40].
Many studies have reported that circulatory macrophages in ltrate the brain following peripheral in ammation and further contribute to the CNS diseases. However, they were limited to the several relatively rare disease states including arthritis and liver injury [24,25,41]. Our work provided experimental evidence of systemic in ammation induced in ltration of macrophages into the brain in response to peripheral tissue injury, and suggests a potential role of in ltrated macrophages in the development of peripheral trauma-induced neurological disorders such as PTSD and POCD [3,42]. Furthermore, we con rmed that il1b but not tnfa, the latter being generally recognized as a cytokine in systemic in ammation-induced in ltration of macrophages and neurological disorders, served as the key proin ammatory regulator under peripheral tissue injury [24,25]. While microglia-secreted CCL2, a chemokine attracting macrophages into the CNS [24,43], remained unchanged in this model, our results demonstrated that the accumulation of apoptotic cells in the brain triggered by il1b impaired phagocytosis, as characterized by reduced lysosomes in macrophage/microglia, and may contribute to the recruitment of macrophages/microglia following peripheral tissue injury [36,39,44]. These observations suggest that the mechanisms of the systemic in ammation mediated neurological disorders may vary between pathological conditions, which warrant further studies. It is widely believed that breakdown of BBB promotes circulatory macrophage in ltration, though in ltrating macrophages can be found in the CNS with intact BBB and migrating through the cerebrospinal uid (CSF) [9,45]. In this current study, circulation-independent in ltration of macrophages was observed, where macrophages invaded the brain from the neighboring tissues through the lateral periphery of the hindbrain, speci cally between metencephalon (or cerebellum) and myelencephalon, differs from the microglial invasion to the zebra sh midbrain during development [46]. This suggests that in addition to circulation, other less studied migration routes might be involved in the in ltration of macrophages into the brain under speci c disease states.
Involved in the pathogenesis of multiple neurological disorders, in ltration of macrophages into the brain might result in seizure or sickness behaviour or anxiety in response to peripheral in ammation and neuroin ammation, respectively [3,24,47]. We demonstrated that peripheral tissue injury results in hyperactive behavior marked by elevated expression of camk2g1 in the context of both systemic in ammation and neuroin ammation. More importantly, il1b mutation mitigated the increased levels of camk2g1 and the hyperactive behavior. CaMK2G, the human ortholog of camk2g1, is associated with major neuropsychiatry disorders. Correlation between polymorphism in CaMK2G and human memory performance was reported previously [48]. In addition, abnormal levels of alternative CaMK2G splicing was found in the brain of patients with autism spectrum disorder (ASD) [49]. More importantly, the elevated CaMK2G-encoding protein (γCaMKII) level in the brain has been detected in rodent models of anxiety, schizophrenia and major depressive disorder (MDD) [50][51][52]. Thus, hyperactive behavior observed in our peripheral tissue injury model might represent the camk2g1-related neuropsychiatry disorder. While the roles of CaMK2G in macrophages in ltration and subsequent impact on neurological functions require further investigations [53], the il1b-mediated elevated expression of camk2g1 together with in ltration of macrophages, neuronal apoptosis and neuroin ammation might serve as the potential mechanism underlying peripheral tissue injury-induced neuropsychiatry disorders.

Conclusion
Although peripheral injury/in ammation has long been linked to neurological disorders, the underlying mechanisms remain unclear. The present study demonstrated that peripheral tissue injury induces il1bmediated macrophage in ltration into the brain and a hyperactive behavior, which may contribute to development of therapeutic strategies against macrophage in ltration in neurological disorders following peripheral injury. Abbreviations AD (Alzheimer&#39;s disease); AGM (aorta-gonad-mesonephros); aifm4 (factor mitochondria associated 4); ALS (amyotrophic lateral sclerosis); AMPU (amputation); AO (acridine orange); apoeb (apolipoprotein Eb); ASD (autism spectrum disorder); BBB (blood-brain barrier); CHT (caudal hematopoietic tissue); CSF B. Availability of data and materials All data supporting this study are available from the corresponding author upon reasonable request.

C. Competing interests
The authors declare no competing interests.