1522

Abstract. – OBJECTIVE: This research at-tempts to identify the part the hippocampus plays in accelerated fracture-healing after trau-matic brain injury as well as to test functions of calcitonin gene-related peptide (CGRP) during this process.

MATERIALS AND METHODS: Experiments were carried out on Male Sprague-Dawley rats that were split into four groups at random: TBI-fracture group, fracture-only group, TBI-only group, and control group. In the first week, blood specimen would be drawn from rats among the groups except those of the control group at three-time points (24, 72 and 168 hours) post-damage. These rats would be assessed from the neurolog-ical perspective based on their grades of perfor-mance in a sequence of tests 24 hours before and 12 hours after brain injury. Blood samples were also taken from the control group 24 hours be-fore the injury, and whole brain tissues in the in-jured groups were harvested at 72 and 168 hours post-injury. We compared the serum CGRP con-centration, the distribution of CGRP, the CGRP expression, and the expression of CGRP in the hippocampus, the expression of CGRP in the hip-pocampus, the expression of CGRP in the hippo-campus, and the expression of CGRP in the brain by immunohistochemistry, Western blotting, RT- Of CGRP RNA expression levels.

RESULTS: Neurological examinations sug-gested that the functions of the cerebral cortex, cerebellum, and brain stem showed significant differences pre- and post-injury (p < 0.001). ELI-SA analysis indicated a great density of CGRP in TBI-fracture group at different time points. Fur-thermore, in the TBI-fracture group, CGRP in both hippocampus and the whole brain showed a noticeable augment in RT-PCR and western blot analysis at 72 and 168 h post-injury, and on-ly in this group, immunohistochemistry analysis indicated that CGRP was present in the hippo-campus at 168 hours post-injury.

European Review for Medical and Pharmacological Sciences 2017; 21: 1522-1531

Y. SONG, G.-X. HAN, L. CHEN, Y.-Z. ZHAI, J. DONG, W. CHEN, T.-S. LI, H.-Y. ZHU

Department of Emergency, General Hospital of Chinese PLA, Beijing, China

Corresponding Author: Haiyan Zhu, MD; e-mail: zhuhy301@aliyun.com

The role of the hippocampus and the functionof calcitonin gene-related peptide in the mechanism of traumatic brain injury accelerating fracture-healing

CONCLUSIONS: We observed that the hippo-campus and CGRP were responsible for quick bone-healing mechanisms. We suggest a role for the hippocampus in accelerated fracture healing. CGRP expression, as determined by IHC, cannot be observed in other groups, indi-cating that the hippocampus may be the specif-ic component of the brain that responds to “big stress”.

Key Words: Hippocampus, Fracture-healing, Traumatic brain in-

jury, Calcitonin gene-related peptide.

Introduction

Traumatic brain injury (TBI) combined with a fracture is common in road accidents, which is a dominant reason for death as well as disabi-lity. However, fracture-healing of patients with TBI is often accelerated in clinical practice. Ac-cording to Perkins and Skirving in 1987, spee-ded-up callus formation along with sooner union existed in patients who had long-bone fractures and TBI1. As a result, various researches have been conducted with the intention of identifying mechanisms causing this phenomenon over past several years2, which, however, remains uncle-ar today. The recovery for fracture united with peripheral nerve injury is likely to be tardy. On the contrary, fracture with damage to the cen-tral nervous system (CNS) accelerates the hea-ling process. Thus, the CNS should be the focal point of current research. The hippocampus is sensitive to outside injury information3, which is a primary part of the human brain, reposing on its medial temporal lobe. Despite the importance

CGRP in fracture-healing

1523

of the hippocampus, it has not been the focus of previous investigations.

Calcitonin gene-related peptide (CGRP) is a neuropeptide produced by cells of the central and peripheral nervous system4. The function of CGRP in bone metabolism has been well demon-strated. Osteoblasts possess receptors for CGRP which can stimulate osteoblast proliferation5, growth factor and cytokine synthesis, collagen synthesis and bone formation6. Following a su-barachnoid hemorrhage, CGRP is also released7. In addition, the survival of tissues is correlated with the greater density of CGRP immunoreacti-vity, and endogenous CGRP exerts a regulatory impact on ischemia8. Thus, CGRP is a candidate for connecting the CNS to the fracture site. The study attempts to identify the function of the hip-pocampus in accelerated fracture-healing as well as to test functions of CGRP in this process.

Materials and Methods

AnimalsExperiments were carried out on 300-350g Male

Sprague-Dawley rats (Laboratory Animal Center of the Fourth Military Medical University, Xi’an, China). The animals were split into four groups at random: TBI-fracture group, fracture-only group, TBI-only group, and control group. Experiments on the rats were done in accordance with related Chinese regulations on Animal Welfare, and had acquired the permission from the Ethical Com-mittee of the Fourth Military Medical University, Xi’an, China.

Surgical TechniqueClosed transverse femoral shaft fractures were

created as stated by Bonnarens and Einhorn9. In short, after weighed by the balance, the rats would be anesthetized by injected 20% urethane solution (5 ml/kg body weight) intraperitoneally. Their left rear legs were de-haired, sterilized and draped. A median parapatellar incision had been conducted before the same operation into the joint capsule (stretching from the midline throughout the va-stus medialis muscle until the patellar tendon in-sertion). Tardy flexion of the knee while moving the patella resulted in lateral dislocation. An in-tercondylar entry point was made with a handheld drill so that a 1.2 mm diameter Kirschner wire could be inserted into the medullary space, which would stick in till the greater trochanter was met. Next, the wire was mildly retracted, cut, reinser-

ted, and buried under the knee cartilage surface. Surgical spots and the skin would be sutured. At the next step, an obtuse guillotine would be used to create the fracture, and radiographs would be made to locate the pin placement and the fracture configuration. Comminuted fractures or fractu-res not happening in the middle position of the diaphysis would not be included.

Traumatic Brain InjuryTBI was produced by a soundly-built strike ac-

celeration system10,11. This model was proven to reproducibly impart a diffuse axonal injury with a resemblance to what had been discovered in pa-tients with TBI. In case the rat exhibited traces of a normal healing of the closed femoral shaft fracture, a nickel-plated cap would be fixed abo-ve the rat’s skull. After that, the rat’s head would be placed on a foam board under stainless steel tubing with the smooth inner wall, the end of whi-ch was around the nickel plate straightly over the rat’s head. Then, a 250 g weight was fallen onto the rat’s head from as high as exactly 1 m above. Upon the contact, the foam board, as well as the rat, was taken away to avert a rebound hit so as to restrict the damage to a one-time strike. Rats were taken back to their cages and were permitted to heal in observation.

Neurological ExaminationThe rats have been assessed from the neurolo-

gical perspective based on their scores of beha-vior in a sequence of tests 24 hours before and 12 hours after brain injury according to the establi-shed protocols of Arrieta-Cruz et al12 and Ziporen et al13. See Table I for additional details.

Specimen CollectionBlood specimen have been drawn from ani-

mals in the TBI-fracture, fracture-only and TBI-only groups at three-time points in the first week (24, 72 and 168 hours) post-injury. Blood samples were also taken from the control group 24 hours before damage. All specimens were centrifuged at 1000 g for 15 min in 30 min after gathering, and the consequent supernatant was reserved at -80°C for ELISA analysis. Whole brain tissues in the injured groups were gathered at 72 and 168 hours post-damage and were formalin-fixed and climbed for Hematoxylin and Eosin (H&E) staining and immunohistochemistry (IHC). The hippocampus and the whole brain were also used for Western blot (at 72 hours) and reverse tran-scription-polymerase chain reaction (RT-PCR)

Y. Song, G.-X. Han, L. Chen, Y.-Z. Zhai, J. Dong, W. Chen, T.-S. Li, H.-Y. Zhu

1524

analysis (at 168 hours). Brain and blood speci-mens were also acquired from rats in the control group to serve as controls. All samples had been acquired at around 08:00 a.m. for the purpose of keeping resembling situation for all groups.

ELISA AnalysisEmploying a CGRP ELISA immunoassay kit

(Multi Science, Hangzhou, China), an ELISA analysis was performed for the quantitative de-termination of CGRP existing in the rat serum.

Table I. Neurological tests applied to brain-injured rats.

NO Tests Evaluating method Method CNS localization

1 Skin color 0=blanched, 0.3=blue or violet, 0.6=bright, deep red flush, 1=pink 2 Urination and 0=absent, 1=present defecation3 Body position 0=incapacity, Noticed for consciousness, 0.5=limited movement, breeding, curiosity, 1=normal grooming, and autonomous motions.4 Front leg placing 0=absent, 1=present Grasping the dorsum Cerebral cortex of the front paws near a desk edge leads to laying of palms on the desk.5 Hind leg placing 0=absent, 1=present Letting one hind leg Cerebral cortex down from the table (along with other three legs on the table) causes uplifting of the leg to the table.6 Placing reflex 0=absent, 1=present The rat is captured at the Cerebral cortex tail above a table and is lowered till the beard is close to the table. The paws are then placed on the table.7 Gait 0=Inability to walk straight, Incapability of walking on Cerebellum 0.5=Inability to walk a 6-cm wide beam or walking on 6-cm-wide beam, in straight line 1=walk on 6-cm-wide beam 8 Equilibrium 0=absent, 1=present The rat is put head down on a platform tilted at 30°, and upward climbing is observed.9 Postural reflex 0=absent, 1=present Uplifting the rat by its tail results in posterior legs abduction, unfolding of toes, arching of the back, and uplifting of the head.10 Righting reflex 0=no response, When the rat is placed on 0.5=slow response, its back, it turns 1=active over its feet over immediately. and impossible to evaluate 11 Flexion reflex 0=slight withdrawal or none, The rat is picked up, and the toes are 0.5=moderate withdrawal pinched with forceps, leading to 1=brisk rapid withdrawal (not brisk), foot withdrawal.12 Grip 0=less than 14 seconds, A neurologically normal rat 0.5=from 15 to 30 seconds, should be able to hang suspended 1=more than 30 seconds on a stationary bar for at least 30 seconds.

CGRP in fracture-healing

1525

The CGRP assay was conducted based on the producer’s specifications. The optical density was surveyed at a wavelength of 450 nm by a microplate reader (Multiskan Ascent, Thermo Labsystems, Shanghai, China). Concentrations were worked out and presented in pg/ml.

Immunohistochemistry Assay A 4 µm-thick tissue section was set on a

slide. Having been deparaffinized and dehy-drated by xylene and ethanol, respectively, the slide was prepared for antigen retrieval by subjecting to microwave pretreatment for 10 min. Incubation with 3% H2O2 was adop-ted to prevent endogenous peroxidase in the tissue. After washes with phosphate-buffered saline, the slide was incubated for the who-le night with a rabbit anti-rat CGRP primary polyclonal antibody (Abcam, Cambridge, MA, USA) at 4°C. In the next step, the slide would be rinsed and treated for 15 min with a goat anti-rabbit secondary antibody as well as the Avidin-Biotin reagent for 30 min in succes-sion. The slide was then incubated in diami-nobenzidine for 10 min, counterstained with hematoxylin, dehydrated in a graded series of alcohols, cleared in xylene and mounted. Sections from the normal brain of the control group were included as controls.

Western Blot AnalysisWestern blotting was used in order to de-

tect CGRP expression in the hippocampus. The samples were homogenized in 10 mmol/L Tris homogenization buffer (pH 7.4) with pro-tease inhibitors (1 tablet in 50 m, Sigma-Al-drich, St. Louis, MO, USA) and centrifuged at 12,000 rpm for 20 min, after which the super-natant was gathered. Samples (50 μg protein) were filled into an 8% SDS gel; the proteins were divided by electrophoresis before moved to nitrocellulose membranes, which were then incubated with a rabbit anti-rat CGRP primary polyclonal antibody (Abcam, Cambridge, MA, USA). The membranes were rinsed 3 times for 10 min and incubated with goat anti-rabbit IgG-HRP (1:2000, Santa Cruz Biotechnology, San-ta Cruz, CA, USA). An ECL western blotting detection kit visualized the immunoreactive bands (Pierce, Rockford, IL, USA) on light-sen-sitive film. β-actin was used as an endogenous reference, and average light density analysis was employed to work out the ratio of CGRP expression to β-actin.

Reverse Transcription of mRNA and RT-PCR Analysis

Total RNA was separated from the whole brain by the Trizol reagent (Invitrogen, Carlsbad, CA, USA) in accordance with the producer’s in-structions. The primary sequences applied to amplify CGRP were 5’-AAGTTCTCCCCTT-TCCTG-3’ as well as 5’-GCCTCCTGTTCCTC-CTC-3’. The PCR protocol contained 1 cycle at 94°C for 30 seconds, 30 cycles at 52°C for 30 seconds, 1 cycle at 72°C for 30 seconds, and 1 cycle at 72°C for 8 minutes. β-actin was enlar-ged and used as an internal standard; the primary sequences were 5’-TGCCGCATCCTCTTCCT-3 and 5’-GGATGTCAACGTCACACTTC-3’. For the purpose of deciding the expression levels of CGRP RNA in the brain, average light density analysis worked out the ratio of CGRP mRNA to β-actin mRNA.

Statistical AnalysisThe data were noted down as the mean±SD and

subjected to a normality test to check the normal distribution. In the case of the failure of the nor-mality test, a non-parametric test would be used for the analysis of the data. In the case of the suc-cess of the normality test, one-way ANOVA and the succedent Student-Newman-Keuls test were employed for the assessment of statistical analysis with intention of comparing the variations between different groups. All statistical analyses were carri-ed out by the employment of the statistical program for social sciences (SPSS Inc., Chicago, IL, USA) 17.0. p

Y. Song, G.-X. Han, L. Chen, Y.-Z. Zhai, J. Dong, W. Chen, T.-S. Li, H.-Y. Zhu

1526

approximately 17% compared to the pre-injury level at 15 seconds since impact before gradual-ly dropping back to normal in 30 minutes of it. Heart rate reduced from a pre-injury level of 341±28 to 261±51 beats/min at 1 min post-injury before slowly returned to normal in 10 min post-injury. Generalized convulsions de-veloped immediately after injury and lasted approximately 15-30 seconds. Most animals suffered decortication flexion deformity of the forelimbs. No one group showed differences in capability of walking one day after the injury.

Pathological ChangesGross pathological observation. Subarach-

noid, intraventricular and brain stem petechial hemorrhages were discovered among brain-inju-red rats; however, the brains appeared otherwise normal without any contusions or additional focal lesions. Skull fractures did not develop. Micro-scopic changes. In the control group, the neurons distributed regularly and the structures were in-tact. In the brain-injured rats, the neurons in the

cortex and hippocampus exhibited a disordered distribution with a concomitant reduction in nor-mally distributed neurons. Shrunken neurons in-tegrated with perineuronal vacuolation were de-tected in these fields, and so were dark, contracted neurons with corkscrew-like processes (Figure 1).

Neurological Function AnalysisThere was a significant difference in the total

neurological scores among brain-injured animals before and after injury (p < 0.001). The cerebral cortex, cerebellum, and brain stem functions showed significant differences pre- and post-injury (p < 0.001); there were not prominent chan-ges in spinal cords (p > 0.05). Rats in the control group also presented no difference in function pre- and post-injury (p > 0.05; Table II).

ELISA AnalysisAt 24 hours after the injury, no prominent

changes in CGRP density were discovered in the fracture-only group or the TBI-only one in com-parison with the control group (p > 0.05); howe-ver, the TBI-fracture group presented more CGRP compared with the control group (p < 0.05). Then, at 72 hours post-fracture, all experimental groups expressed greater density of CGRP compared with the control one, while the TBI-fracture group had more CGRP compared with the fracture-only group (p < 0.05). At 168 hours post-fracture, CGRP expression in the TBI-fracture group along with the fracture-only group kept elevated (p > 0.05), and the TBI-fracture group had higher le-vels of CGRP compared with the fracture-only group (p < 0.05; Table III).

Immunohistochemistry Assay CGRP was discovered in the hippocampus

in the TBI-fracture group only at 168 hours post-injury; CGRP was not detected at 72 hours post-injury in either the TBI-fracture or the fracture-only group, nor was it discovered in the fracture-only group at 168 hours post-injury or the control group (Figure 2).

Figure 1. Photomicrographs of the hippocampus 3 days after head injury in rats. Neurons were distributed regular-ly, and structures were intact in Control group. The neu-rons were distributed in a disorderly manner, and thus the number of normally distributed neurons was fewer. Dark, contracted neurons were observed in Brain-injured group (Magnification×200).

Table II. Scores of neurological function analysis.

Injury time Total scores Cerebral cortex Cerebellum Brain stem Spinal cord Total scores (TBI group) (TBI group) (TBI group) (TBI group) (TBI group) (control group) Pre-injury 11.70±0.56 2.80±0.56 2.00±0.00 3.00±0.00 0.97±0.20 11.90±0.3212 h post-injury 6.20±0.99* 1.60±0.63* 0.26±0.50* 0.97±0.55* 0.80±0.32# 11.50±0.71 #

*p < 0.001 vs. scores before injury; #p > 0.05 vs. scores before injury

CGRP in fracture-healing

1527

Western Blot AnalysisCGRP expression was higher in the TBI-fractu-

re group compared with the fracture-only and control ones at 72 hours post-fracture (p < 0.05 and 0.001, respectively) and was increased further at 168 hours (p < 0.001). The TBI-fracture group expressed more CGRP contrasted with the fractu-re-only group (p < 0.01; Figure 3).

RT-PCR Analysis RT-PCR showed that the mRNA expression of

CGRP rose in the TBI-fracture group at 72 hours after brain injury (p < 0.05, contrasted to the con-trol group). At 168 hours after the fracture, CGRP

expression increased in the TBI-fracture and fracture-only groups compared with expression at 72 h (p < 0.001 and 0.05, respectively; Figure 4).

Discussion

Clinical observations of long-bone fractures in patients who also had TBI have shown an in-crease in fracture healing speed. In 1987, patients with long-bone fractures and TBI manifested ad-ded callus formation and sooner union, according to Perkins and Skirving14. Over past few years, a number of studies have been carried out to reveal the mechanisms responsible for this phenomenon. However, until now, mechanisms behind the si-tuation were unclear. Many orthopedic resear-

Figure 3. Expression of CGRP in the hippocampus at diffe-rent time points post-injury (0, 72 and 168 h). Data were calcula-ted as the ratio of expression to β-actin expression. (A) Control (0.0259±0.0253). (B) TBI-fracture group at 72 h (0.2851±0.0868). (C) Fracture-only group at 72 h (0.1625±0.0631). (D) TBI-fractu-re group at 168 h (0.5851±0.0995). (E) Fracture-only group at 168 h (0.3282±0.0732). *p < 0.001 vs. control, #p < 0.05 vs. fractu-re-only group at 72 h, **p

Y. Song, G.-X. Han, L. Chen, Y.-Z. Zhai, J. Dong, W. Chen, T.-S. Li, H.-Y. Zhu

1528

chers focused directly on humoral factors or local fracture sites, and many cytokines and growth factors have been shown to participate in the me-chanisms. Nevertheless, our studies mainly focu-sed in another direction. When a fracture occurs at the same time as peripheral nerve injury, the recovery may be tardy. On the contrary, fractures associated with spinal injury and cerebral cortex injury speed up the recovery course2. Differences mentioned above imply that the CNS should be the focus of current research, and humoral factors may be used to carry out such research.

Using the neurological tests of Arrieta-Cruz et al12 and Ziporen et al13 as a guide, we developed a new assay in which the difficult-to-implement and strong subjective items had been removed. In this assay, the neurological state of animals was assessed based on a sequence of neurological tests that were based on neuroanatomy and neurophy-siology. This assay is a gross technique for reve-aling main neurological dysfunction, resembling a clinical neurological examination. The assay is not perfect, and not all of the measures are sta-tistically independent of each other. However, in the present experimental TBI model, the assay was able to reveal deficits. In our work, changes in neurological scores indicated that functions of the cerebral cortex, cerebellum, and brain stem were altered.

As a 37-amino-acid neuropeptide, CGRP is en-coded by the Calca gene and is expressed and al-ternatively spliced in cells of central and periphe-ral nervous system4. The role of CGRP in bone metabolism has been well demonstrated. Osteo-blastic cells possess receptors for CGRP, which can stimulate osteoblast proliferation5, growth factor and cytokine synthesis, collagen synthesis as well as bone formation6. CGRP also adds the quantity of bone colonies formed by bone marrow stromal cells in vitro15 and inhibits bone resorp-tion in vitro16. In our previous report, CGRP was indicated to play a large part in accelerating bone healing. CGRP exists in a large amount in the central and peripheral nervous system. CGRP-re-ceptor component protein (RCP) is a coupling protein needed for the CGRP-receptor signaling. RCP-immunoreactive perikarya extensively and alternatively exist in the cerebral cortex, hip-pocampus and other parts of the brain17. CGRP significantly ameliorates vasospasm, improves cerebral blood flow, and reduces cortical and en-dothelial cell death. Moreover, CGRP can stimu-late angiogenesis18 and increase the formation of insulin-like growth factor-I in the brain, which

imposes advantageous influence on the cognitive function through improving synaptic transmis-sion and through stimulating neurogenesis in the hippocampus19. The greater density of CGRP is correlated with tissue survival, while endogenous CGRP exerts a regulatory impact on ischemia8. Following subarachnoid hemorrhage, CGRP is released7. In our study, serum CGRP increased 24 hours post-injury in the TBI-fracture group and kept raising with time. However, in the fractu-re-only group, the values did not exceed those in the TBI-fracture group at any time point. This finding indicated that the increased CGRP in the TBI-fracture group can only be attributed to TBI. Furthermore, the accelerated healing indirectly suggested that CGRP has a marked effect on both TBI and fracture healing.

The blood-brain barrier (BBB) contains non-fenestrated endothelial cells associated with tight junctions and conforms to a lasting base-ment membrane, which is a structural and fun-ctional barrier between cerebral capillaries and brain parenchyma, and restricts the free pass of large molecules from blood into the CNS20. The BBB reflects an important barrier to various CNS-active agents21,22. The novel approach of re-

Figure 4. Expression of CGRP mRNA in the whole brain at different time points post-injury (0, 72 and 168 h). Data were calculated as the ratio of expression to β-actin expres-sion. (A) Control group (0.3046±0.0810). (B) TBI-fractu-re group at 72 h (0.3995±0.0314). (C) Fracture-only group at 72 h (0.2661±0.0499). (D) TBI-fracture group at 168 h (0.4935±0.0445). (E) Fracture-only group at 168 h (0.3869±0.0721). *p < 0.05 vs. control, #p < 0.001 vs. fractu-re-only group at 72 h, **p < 0.01 vs. TBI-fracture group at 72 h.

CGRP in fracture-healing

1529

ducing available CGRP is limited by the BBB23. Cortez et al3 suggested that the BBB was broken after experimental brain injury. Transient ische-mia is indicated to injure the basal lamina of the cerebral microvasculature and increase BBB per-meability24,25. CNS-specific markers were mar-ked in both cerebrospinal fluid (CSF) and serum from brain-injured patients26. In our recent study, CGRP staining was found in the choroid plexus in the TBI-fracture group, other than those in other groups. In a recent study27, the osteoinductive po-tential of CSF from TBI patients was described. Our previous study indicated that CGRP induced bone formation. The dynamic changes in the se-rum levels of CGRP in the TBI-fracture group may be connected with the leakage and gradual healing of the BBB after brain injury. This idea conforms to the hypothesis that the osteoinducti-ve effect could show the disruption of the BBB as a result of trauma, thus permitting the leakage of osteoinductive CSF components into the systemic circulation28. Therefore, the putative speeded up osteogenic impacts in TBI combined with fractu-re are considered to be caused by the release of CGRP from injured neural tissue or as part of the CNS response to brain injury. However, the area of the brain responsible for this mechanism re-mains unclear.

The hippocampus is the main part of the brain of humans and other mammals and exists in the medial temporal lobe of the brain29. It involved in memory forming, organizing, and storing. The hippocampus is sensitive to outside injury infor-mation. Experiments by Cortez et al3 discovered specific necrosis of neurons in CA1 after expe-rimental TBI. Both CGRP and CGRP receptors were found in the hippocampus30. The co-locali-zation of CGRP and CGRP-receptor component proteins in neurons suggested that CGRP im-poses an autocrine or paracrine impact on these neurons. Bulloch et al31 indicated that injury in-duced CGRP-like immunoreactivity in neurons in the area of damage. CGRP expression in sur-viving neurons within damage-related regions of the hippocampus was shown a likely significant, and likely offensive part of the reaction of the nervous system to damage31. The transportation of newly synthesized CGRP into regenerating axons was observed after injury32. In our stu-dy, CGRP expression in the hippocampus at 72 hours and 168 hours post-injury was highest in the TBI-fracture group. Additionally, there were CGRP staining in the choroid plexus and hip-pocampus in the TBI-fracture group. Taken to-

gether, we suggested that TBI-induced CGRP in the brain, and CGRP acted in an autocrine or pa-racrine manner, protecting the injured brain tis-sue. CGRP that was expressed in the hippocam-pus in response to both fracture and brain injury flowed into the serum through the damaged BBB and accelerated fracture healing in conjunction with CGRP secreted from the peripheral nerves. Our results suggested that the hippocampus and CGRP were responsible for quick bone-healing mechanisms. We are the first to propose a role for the hippocampus in accelerated fracture healing. CGRP expression, as determined by IHC, cannot be observed in other groups, indicating that the hippocampus may be the specific component of the brain that responds to “big stress”.

Combined with neuroscience and orthopedics, this topic is complicated and challenging. Clinical observations of more rapid healing are only evi-dent at the surface, but the reason for the increased healing speed may be TBI. For years, orthopedic researchers focused only on humoral or local fac-tors that influenced fracture healing; neurosurgical researchers seldom studied this field. The under-standing of the sophisticated mechanisms related to stepped-up fracture-recovery remains at the ini-tial steps, and a lot more is waiting for research.

Conclusions

Taken above, dynamic changes in CGRP expression in the hippocampus were not in ac-cordance with those of the whole brain. Thus, the increase in CGRP mRNA in the brain in the TBI-fracture group may also be released from another part of the brain. Determining the speci-fic area of the hippocampus that is responsible for these mechanisms and whether other parts of the brain also will be the topic of a future study. As the mechanism of rapid fracture-healing becomes further understood, patients will hopefully soon benefit from faster healing after the fracture.

FundingThis study was supported by the National Natural Science Research Foundation of China (No: 81670467), Beijing Sci-entific & Technologic Supernova Supportive Project (No: 15111000030000/XXJH2015B100), Youth Project of Na-tional Natural fund (No: 81400648), the Beijing Natural Sci-ence Fund (No: 7152136), Youth Culture Project of Chinese PLA (No: 13QNP171), Surface Project of Hainan Province Natural Fund (No: 20158315), and Clinical Science Fund of PLA General Hospital (No: 2015FC-ZHCG-2007, 2012FC-TSYS-4028).

Y. Song, G.-X. Han, L. Chen, Y.-Z. Zhai, J. Dong, W. Chen, T.-S. Li, H.-Y. Zhu

1530

AcknowledgmentsWe thank Rong Lv, Jie Wu and Jing Li for help with the photomicrographs. We also thank Xin Bu for help with the data analysis and Shusen Wang and Xiaorui Cao for their thoughtful discussions.

Authors’ contributionsSong Y carried out all experiment in this study, Zhu HY participated in the design of the study and performed the statistical analysis. Han GX, Chen L, Zhai YZ, Dong J, Chen W and Li TS conceived of the study, and participated in its design and coordination and helped to draft the man-uscript. All authors read and approved the final manuscript.

Conflict of interestThe authors declare no conflicts of interest.

References

1) Perkins r, skirving AP. Callus formation and the rate of healing of femoral fractures in patients with head injuries. J Bone Joint Surg Br 1987; 69: 521-524.

2) vrettos t, Poimenidi e, AthAnAsoPoulos P, BAlAsis s, kArAgiorgos n, siklis t, gAtzounis g, Fligkou F. The effect of permissive hypotension in combined trau-matic brain injury and blunt abdominal trauma: an experimental study in swines. Eur Rev Med Phar-macol Sci 2016; 20: 620-630.

3) Cortez sC, mCintosh tk, noBle lJ. Experimental flu-id percussion brain injury: vascular disruption and neuronal and glial alterations. Brain Res 1989; 482: 271-282.

4) AmArA sg, JonAs v, rosenFeld mg, ong es, evAns rm. Alternative RNA processing in calcitonin gene expression generates mRNAs encoding different polypeptide products. Nature 1982; 298: 240-244.

5) Cornish J, CAllon ke, lin CQ, XiAo Cl, gAmBle gd, CooPer gJ, reid ir. Comparison of the effects of calcitonin gene-related peptide and amylin on osteoblasts. J Bone Miner Res 1999; 14: 1302-1309.

6) vignery A, mCCArthy tl. The neuropeptide calci-tonin gene-related peptide stimulates insulin-like growth factor I production by primary fetal rat oste-oblasts. Bone 1996; 18: 331-335.

7) edvinsson l. Calcitonin gene-related peptide (CGRP) in cerebrovascular disease. Scientifi-cWorldJournal 2002; 2: 1484-1490.

8) BuCinskAite v, BroddA-JAnsen g, stenFors C, theodors-son e, lundeBerg t. Increased concentrations of calcitonin gene-related peptide-like immunoreacti-vity in rat brain and peripheral tissue after ischa-emia: correlation to flap survival. Neuropeptides 1998; 32: 179-183.

9) BonnArens F, einhorn tA. Production of a standard closed fracture in laboratory animal bone. J Or-thop Res 1984; 2: 97-101.

10) FodA mA, mArmArou A. A new model of diffuse brain injury in rats. Part II: Morphological characteriza-tion. J Neurosurg 1994; 80: 301-313.

11) mArmArou A, FodA mA, vAn den Brink W, CAmPBell J, kitA h, demetriAdou k. A new model of diffuse brain injury in rats. Part I: Pathophysiology and biome-chanics. J Neurosurg 1994; 80: 291-300.

12) ArrietA-Cruz i, PFAFF dW, shelley dn. Mouse model of diffuse brain damage following anoxia, evalua-ted by a new assay of generalized arousal. Exp Neurol 2007; 205: 449-460.

13) ziPoren l, shoenFeld y, levy y, korCzyn Ad. Neurolo-gical dysfunction and hyperactive behavior asso-ciated with antiphospholipid antibodies. A mouse model. J Clin Invest 1997; 100: 613-619.

14) Perkins r, skirving AP. Callus formation and the rate of healing of femoral fractures in patients with head injuries. J Bone Joint Surg Br 1987; 69: 521-524.

15) shih C, BernArd gW. Calcitonin gene related pep-tide enhances bone colony development in vitro. Clin Orthop Relat Res 1997; 335-344.

16) AkoPiAn A, demulder A, ouriAghli F, CorAzzA F, Fondu P, BergmAnn P. Effects of CGRP on human osteo-clast-like cell formation: a possible connection with the bone loss in neurological disorders? Peptides 2000; 21: 559-564.

17) mA W, ChABot Jg, PoWell kJ, JhAmAndAs k, diCkerson im, Quirion r. Localization and modulation of cal-citonin gene-related peptide-receptor component protein-immunoreactive cells in the rat central and peripheral nervous systems. Neurosci 2003; 120: 677-694.

18) sun Bl, shen FP, Wu QJ, Chi sm, yAng mF, yuAn h, Xie Fm, zhAng yB, Chen J, zhAng F. Intranasal delivery of calcitonin gene-related peptide reduces cerebral vasospasm in rats. Front Biosci (Elite Ed) 2010; 2: 1502-1513.

19) hArAdA n, nArimAtsu n, kurihArA h, nAkAgAtA n, okAJimA k. Stimulation of sensory neurons impro-ves cognitive function by promoting the hippo-campal production of insulin-like growth factor-I in mice. Transl Res 2009; 154: 90-102.

20) SobańSka k, karbownik a, Szałek e, Płotek w, Gra-bowSki t, Szewczyk a, MarcinkowSka D, PołoM w, mAtuszeWski m, grześkoWiAk e. The influence of the time-of-day administration of sunitinib on the penetration through the blood-brain and blooda-queous humour barriers in rabbits. Eur Rev Med Pharmacol Sci 2016; 20: 166.

21) yAng Cs, ChAng Ch, tsAi PJ, Chen Wy, tseng Fg, lo lW. Nanoparticle-based in vivo investigation on blood-brain barrier permeability following ischemia and reperfusion. Anal Chem 2004; 76: 4465-4471.

22) Beggs s, liu XJ, kWAn C, sAlter mW. Peripheral ner-ve injury and TRPV1-expressing primary afferent C-fibers cause opening of the blood-brain barrier. Mol Pain 2010; 6: 74.

23) edvinsson l. CGRP blockers in migraine therapy: where do they act? Br J Pharmacol 2008; 155: 967-969.

CGRP in fracture-healing

1531

24) somogyvAri-vigh A, PAn W, reglodi d, kAstin AJ, Ari-murA A. Effect of middle cerebral artery occlusion on the passage of pituitary adenylate cyclase acti-vating polypeptide across the blood-brain barrier in the rat. Regul Pept 2000; 91: 89-95.

25) trinkl A, vosko mr, WunderliCh n, diChgAns m, hAmAnn gF. Pravastatin reduces microvascular basal lamina damage following focal cerebral ischemia and reperfusion. Eur J Neurosci 2006; 24: 520-526.

26) Petzold A, keir g, lim d, smith m, thomPson eJ. Ce-rebrospinal fluid (CSF) and serum S100B: release and wash-out pattern. Brain Res Bull 2003; 61: 281-285.

27) gAutsChi oP, toFFoli Am, JoesBury kA, skirving AP, FilgueirA l, zellWeger r. Osteoinductive effect of cerebrospinal fluid from brain-injured patients. J Neurotrauma 2007; 24: 154-162.

28) klein By, shohAmi e, reikhinshtein y, Ben-BAssAt h, lie-BergAll m. Serum-mediated osteogenic effects of head injury on cultured rat marrow stromal cells. Calcif Tissue Int 1999; 65: 217-222.

29) zhi Xl, li Cy, Xue m, hu y, Ji y. Changes in cognitive function due to combined propofol and remifenta-nil treatment are associated with phosphorylation of Tau in the hippocampus,abnormal total water and calcium contents of the brain, and elevated serum S100β levels. Eur Rev Med Pharmacol Sci 2016; 20: 2156.

30) liu Xd, zhAng JJ, WAng y, yu lC. Inhibitory effects of calcitonin gene-related peptide on long-term potentiation induced in hippocampal slices of rats. Neurosci Lett 2011; 494: 10-13.

31) BulloCh k, milner tA, PrAsAd A, hsu m, BuzsAki g, mCeWen Bs. Induction of calcitonin gene-related peptide-like immunoreactivity in hippocampal neu-rons following ischemia: a putative regional mo-dulator of the CNS injury/immune response. Exp Neurol 1998; 150: 195-205.

32) toth CC, Willis d, tWiss Jl, WAlsh s, mArtinez JA, liu WQ, midhA r, zoChodne dW. Locally synthesized calcitonin gene-related Peptide has a critical role in peripheral nerve regeneration. J Neuropathol Exp Neurol 2009; 68: 326-337.