Serial
cytokine
levels
during
wound
healing
in
rabbit
maxillary
sinus
Acta Oto-Laryngologica,2010;130:607613ORIGINAL ARTICLESerial cytokine levels during wound healing in rabbit maxillary sinusmucosaXICAI SUN,DEHUI WANG,HUAPENG YU&LI HUDepartment of Otolaryngology,Eye,Ear,Nose and Throat Hospital,Shanghai Medical College,Fudan University,Shanghai,ChinaAbstractConclusion:Collagen deposition was increased in regenerative rabbit maxillary mucosa,and the elevated collagen depositionduring wound healing was associated with increased interleukin(IL)-6 and IL-8,and decreased IL-10.Objectives:To observethe natural progression of wound healing in rabbit maxillary sinus mucosa and to detect the expression of relevant cytokines.Methods:A rabbit wound-healing model was established in the maxillary sinus.At different times after injury,the sizes ofmaxillary ostia were recorded,and histological changes in the sinus mucosa were examined by means of hematoxylin and eosin(H&E)and Masson trichrome staining.Enzyme-linked immunosorbent assay(ELISA)was used to detect the concentrationsof IL-6,IL-8,IL-10,and collagen types I and III.Results:A large amount of immature spindled collagen deposition wasobserved in the basal lamina and lamina propria of regenerative mucosa.IL-6 was increased significantly in regenerativemucosa after injury,especially at 14 days post wounding when compared with normal mucosa.IL-8 was also increasedsignificantly in regenerative mucosa;however,IL-10 was decreased significantly when compared with normal mucosa.IL-6and IL-8 showed a significant correlation with the expression level of collagen type I.Keywords:Rabbit model,scarring,collagen deposition,inflammationIntroductionEndoscopic sinus surgery(ESS)has become the stan-dard surgical treatment for chronic sinusitis with orwithout nasal polyp that is resistant to medical man-agement.Although the functional results have beenextensively described,restenosis of surgical ostia andscarring of nasal cavity and sinuses after ESS are com-mon.Adhesion formation as one of form of scarring ismostly found in the anterior ethmoid region betweenthe middle turbinate and the lateral nasalwall,rangingfrom 1%to 36%in incidence 1.Another anatomicstructurethatisoftenaffectedbyrestenosisisthefrontalostium.Ithasbeenreportedthattheaveragesizeofthefrontal ostium decreased by 2833%after endoscopicmodified Lothrop procedure 2,3.These complica-tions can compromise the success of surgery and maypotentially lead to revision surgery.Current treatment for nasal cavity scarring andrestenosis of surgical ostia includes topical applicationof mitomycin C(MMC),stenting,and hyaluronan(HA)and its derivatives in animal models.MMC is anantineoplastic agent and has shown antifibroblasticeffect when applied topically.Despite this effect,restenosis is still common 1,4,5.Local stentingcan cause a foreign body reaction and the develop-ment of biofilm,which have limited its usefulness5,6.The topical application of HA has been suc-cessfully used to inhibit the scarring in a rabbit model,but its effect is significantly affected by the composi-tion or preparation in which the HA is delivered.Todate,none have been shown to be useful in the humanpatient with chronic post-ESS sinusitis 1,5.The wound healing process involves three phasesthat overlap in time and space:inflammation,tissueformation,and tissue remodeling.These phases areCorrespondence:Dehui Wang,Department of Otolaryngology,Eye,Ear,Nose and Throat Hospital,Shanghai Medical College,Fudan University,83 FenyangRoad,Shanghai,200031,China.Tel:+86 21 64 377 134 388.Fax:+86 21 64 377 151.E-mail:(Received 19 August 2009;accepted 17 September 2009)ISSN 0001-6489 print/ISSN 1651-2251 online?2010 Informa UK Ltd.(Informa Healthcare,Taylor&Francis AS)DOI:10.3109/00016480903352975Acta Otolaryngol Downloaded from by Shanhai Second Medical Univ on 03/23/12For personal use only.regulated by a wide variety of cytokines.Inflammationcan result in rapid repair of the wound,but it is alsoassociated with scar formation 7.Studies have sug-gested that a diminished inflammation response couldpromote regenerative healing and reduce scar forma-tion,which was related to the balance of pro-and anti-inflammatory cytokines 8.Interleukin(IL)-10 isknown to be a major regulator in suppressing theinflammatory response.In cutaneous wound healing,it has been found that IL-10 inhibited the expressionof proinflammatory cytokines,and was conducive toscarless wound healing 9.IL-6 is a multifunctionalcytokine that is up-regulated in a number of fibroticand inflammatory conditions 10.IL-8 is an impor-tant cytokine implicated in inflammation and postna-tal wound healing.Elevated levels of IL-8 have beenimplicated in the formation of scar in adult skin tissue11.Type III collagen is present in reticular fibers,which provide elasticity to tissue,while type I collagenfibrils often contribute to a tissues tensile strength12.In comparison to the extensive basic science andclinical experience with cutaneous wound healing,there have been few reports investigating mucosalwound healing.Therefore,the purpose of this studywas to determine the expression of IL-6,IL-8,andIL-10,and to correlate these cytokines with the levelsof collagen type I and III in rabbit maxillary sinusmucosa during wound healing,in an attempt tofurther clarify,mechanistically,the role of inflamma-tion in sinus mucosa wound healing.Material and methodsAnimal modelThe protocol was approved by the Institutional Ani-mal Care and Use Committee of Fudan University.Twenty-eight Pasteurella-free New Zealand whiterabbits of either sex,weighing between 2 and 3 kg,were used in this study.Animals were observed for1 week at the Animal Care Facility to exclude anyunderlying disease.Four rabbits did not undergosurgery and served as a control group.Twenty-fourrabbits were used to produce wound healing models.The surgical procedure was similar in technique tothat described by other authors.The animals wereanesthetized with intramuscular injections of keta-minehydrochloride(50mg/kg)andxylazine(5 mg/kg).The fur over the nasal dorsum was shavedand sterilized with iodine solution.Local anesthetic(3 ml of 1%lidocaine with 1:100 000 epinephrine)was infiltrated subcutaneously.Using a sterile surgicaltechnique,a 2.5 cm midline vertical incision wasperformed,and the soft tissue and periosteumoverlying the maxillary sinuses were elevated.Theanterior wall of the maxillary sinus was removedwith a rotating bur drill and rongeur.A 4 mm circularopening was created in the medial wall of the max-illary sinus into the nasal cavity.The skin incision wasthen closed with a running absorbable suture.Trans-dermal fentanyl was used for postoperative analgesia.There were no surgical complications,no postoper-ative fatalities,and no postoperative infections.Ostium measurementAt2-,4-,6-,8-,10-,and14-dayintervalsaftertheinitialsurgery,four rabbits were killed with intravenous pen-tobarbital,respectively.Immediately after death,themaxillary sinus was then reopened.The medial wall ofthe maxillary sinus,including the wound,the under-lying bone,and the nasal mucosa on the oppositesurface,were then removed under microscopic visu-alization.The diameter of each ostium was measuredfrom the mucosal edge directly using a tape withmillimeter markings by an investigator in a blindedfashion.The ostium area was estimated based on thearea(pr2)of a circle with the radius(r)taken as half ofthe measured diameter.The specimens were thendivided into two parts.One part was put into liquidnitrogenforELISAanalysisandanotherpartwasfixedin 4%paraformaldehyde buffered with phosphate-buffered saline(PBS,pH 7.4)for histologic analysis.HistologyWound samples were first fixed in 4%paraformalde-hydeforseveraldays.Thenthetissuewasdecalcifiedinaceticacid,andembeddedinparaffinperpendiculartothe ostium.Serial 5mm sections of paraffin-embeddedwound tissues were made using a microtome(Leica,Heerburgg,Switzerland)and stained routinely withhematoxylinandeosin(H&E),orMassontrichrometoassess collagen production and distribution.Thesesections were then examined for histological featuresincludingfibrosis,inflammation,andheterophilia byapathologist in a blinded fashion.Measurement of cytokines,collagen type I,and collagentype III by ELISAWound samples were homogenized in PBS,pH 7.4(adding 5 ml PBS/g of sample),and particulates wereremoved by centrifugation for 15 min at approxi-mately 1000 g,subpackaged,and stored at 80?C.Levels of IL-6,IL-8,IL-10,and collagen types I and608X.Sun et al.Acta Otolaryngol Downloaded from by Shanhai Second Medical Univ on 03/23/12For personal use only.III in wound extracts were quantified using commer-cial ELISA kits(Uscn Life Science&TechnologyCo.,USA)according to the manufacturers instruc-tions.The values obtained by ELISA were correctedby dilution factor and were finally expressed in pg/mlfor IL-6,IL-8,and IL-10,and in ng/ml for collagentype I and III proteins.Statistical analysisSAS 9.1 software was used for statistical analysis.Student-Newman-Keuls test was used to comparethe ostium area,collagen proteins,and expression ofcytokinesinthedifferentgroups.Spearmancorrelationanalysis was used to analyze the relation of cytokineexpressionandcollagendeposition.Avalueofp0.05was considered statistically significant.ResultsOstium areaThe average areas measured by a blinded evaluator forthe 2-,4-,6-,8-,10-,and 14-day specimens are shownin Table I.The average ostium area of the 2-day spec-imen and the initial wounding was 11.79 and 12.57mm2,respectively.This difference was not statisticallysignificant.Startingfromthe4-dayspecimen,theaver-agesizeofthewounddecreasedgradually(p0.0001).HistologyNormal control group.Histological analysis of the nor-mal maxillary mucosa demonstrated healthy signs(Figure 1A).Two or three layers of pseudostratifiedciliated epithelium were identified,with ciliated,gob-let,and basal cells on the basal lamina.Ciliated cellsoutnumbered other cell types.The lamina propriacontained numerous serous glands and vessels andless inflammatory cells.Tissue section stained withtrichrome showed less and healthy collagen deposi-tion in the lamina propria(Figure 1B).The wound groups.The ciliary density of mucosa in thewound groups was markedly diminished,with a fewtufts of cilia surrounded predominantly by denudedmucosal segments.The lamina propria largely con-sisted of acute and chronically inflamed edematouscongestedgranulationtissueandhadlessserousgland.Tissue section stained with trichrome showed a largeamount of immature spindled collagen deposition inthe basal lamina and lamina propria(Figure 2).Concentration of IL-6,IL-8,IL-10,and collagen types Iand III by ELISAThe levels of expression of IL-6,IL-8,and IL-10 inthe normal maxillary mucosa and the woundedmucosa are shown in Figure 3.IL-6 was increasedsignificantly in regenerative mucosa of the 14-daypostoperative group when compared with normalmucosa(p 0.0001).There were no significant dif-ferences in IL-6 expression level between the 2-day,4-day,6-day,8-day,and 10-day postoperative groups.During the postoperative period,IL-8 showed animpressive increase in concentrations in regenerativemucosa in each postoperative group(p=0.0011),andIL-10 was decreased when compared with normalcontrol(p=0.0003).However,IL-8 and IL-10showed no significant difference in concentration inthe six postoperative groups.The concentrations of collagen type I and III of thenormal maxillary mucosa and the wounded mucosaare shown in Figure 4.Collagen type I was signifi-cantly increased in regenerative mucosa when com-pared with the normal maxillary mucosa(p 0.0001).In the 14-day postoperative group,the concentrationof collagen I was significantly higher when comparedwith the other postoperative groups(p 0.0001).Allthe postoperative groups presented significantly lesscollagen type III in regenerative mucosa than thenormal control(p 0.0001).Spearman correlation analysis showed an indepen-dent and significant relation between IL-6 and colla-gen type I expression levels(p 0.0001),IL-8 andcollagen type I expression levels(p=0.0184)in thenormal control group and the wound.However,nocorrelation was found between IL-10 and collagentype I concentration(p=0.9625).The concentrationof IL-10 was associated with the expression level ofcollagen type III(p=0.0004),but the amount of IL-6and IL-8 showed no significant correlation with theexpression level of collagen type III(p=0.8878 and0.9438,respectively).Table I.Measurement of maxillary ostium area.Days after surgerynMean area(mm2)SDNormal control412.570.002411.790.304410.180.46645.991.25842.020.371040.570.081440.000.00Wound healing in rabbit maxillary sinus mucosa609Acta Otolaryngol Downloaded from by Shanhai Second Medical Univ on 03/23/12For personal use only.DiscussionESS is currently considered a standard procedure insinus surgery,but a certain proportion of patients donotshowgoodpostoperativehealing13,14.Restenosis of surgical ostia and scarring of nasalcavity and sinuses are frequent manifestations ofpoor healing,so it is necessary to further analyzefactors associated with the postoperative scar for-mation.As tissue biopsy in postoperative patientscan interfere with normal healing of sinus mucosa,in this study we successfully established a rabbitmodel to quantitatively evaluate the expression levelsof proinflammatory cytokines(IL-6 and IL-8),anti-A.B.Figure 1.Histological observations in normal control group.(A)Representative photomicrographs of H&E staining,and(B)representativephotomicrographs of Masson trichrome staining.Bar=20 mm.A.B.C.D.Figure 2.Histological observations of the 10-day(A,B)and 6-day(C,D)postoperative groups.A representative wound after H&E(leftcolumn)and Masson trichrome(right column)staining is depicted,highlighting the abnormal collagen deposition,cellular infiltrate,andreduced serous gland.Bar=20 mm.610X.Sun et al.Acta Otolaryngol Downloaded from by Shanhai Second Medical Univ on 03/23/12For personal use only.inflammatory cytokine(IL-10),and collagen(types Iand III)at different time points after operation.Onthis basis,an association was found between proin-flammatory cytokines and collagen deposition inrabbit maxillary mucosa.Scar formation is the physiological and inevitableend point of wound repair and there is substantialevidence that inflammation is an essential prerequisitefor scarring 8.The majority of the reports regardingwound repair are based on skin or gingiva tissue,butlittle is known about healing of the respiratorymucosa.Forsgren et al.15 made microscopic obser-vations on the maxillary sinus mucosa of rabbits afteroperation.They concludedthat the mucosaisre-epithelialized within 2 weeks,but the regenerationof the lamina propria is incomplete,and reactivecellular processes such as bone remodeling,fibroblastproliferation,and formation of polyps are character-istic of regenerating mucosa.In another study,theyinvestigated the wound healing of the infected max-illary sinus mucosa following surgical procedures in arabbit model,and they concluded that persi