A Primary Study Of L-carnitine Protective Effect On Ocular Urface Of Mouse Dry Eye Model Induced By Hyperosmolar Saline

Background and purposeDry eye disease （DED）, also known as keratoconjunctivitis sicca or dysfunctional tearsyndrome is caused by abnormal quality and quantity or unusual kinetics of tear，which leadto tear film unstability causing ocular surface disease. Dry eye disease is the most commonocular surface disease. Normal human tear osmolarity is302±9.7mOsmol/L according tothe statistical analysis of literatures, tear osmotic pressure in patients with dry eye diseasewas326.9±22.1mOsmol/L. Taken316mOsmol/L as a diagnostic indicator of dry eyedisease, the overall accuracy is better than any other indicators of diagnosis of dry eyedisease.At present, the treatment of DED，mainly depending on artificial tear drops byincreasing the tear film stability, can alleviate the clinical signs and symptoms of dry eyepatients, however，for severe DED the effect of treatment is poor. Autologous serum as aneffective tear substitute, having no preservatives and no antigenicity, can promote theproliferation of conjunctival and corneal epithelial cell differentiation and was used inmoderate to severe dry eye disease treatment, but the source is limited. Therefore, there is apressing need to develop an effective drug for dry eye treatment and clinical application,based on the pathogenesis of DED.L-carnitine is a polar molecule. Most recently, the role of L-carnitine as an osmolytehas been demonstrated. In the ocular surface tissue of Mammalian, the highestconcentration of L-carnitine was in the lens, and lowest in the vitreous. Through localapplication of L-carnitine in rabbit eye, the content of L-carnitine in the cornea wassignificantly increased. L-carnitine and its derivatives were detected in the tear samples. InDED patients, concentration of L-carnitine was substantially lower than that in healthysubjects. The osmoprotectants L-carnitine were found to protect against stress activation of corneal epithelial cells cultured in hyperosmolar（400mOsmol/L）media. L-carnitine inophthalmology as a physiological supplement or pharmaceutical confirmed concentrationwas1%as its optimal dose. However, whether L-carnitine has osmoprotective effect on theocular surface eipthelial cells has not been reported.Our study is based on a mouse dry eye model induced by hyperosmolar saline, andvaluate the local application of L-carnitine may slow down the ocular surface damage of themouse dry eye model.Methods1. The selection and grouping of animalsSixty female BALB/c mice at the age of6-8weeks were randomly divided into threegroups （20in each）: Hyperosmolar saline group （HO）, Hyperosmolar and Isosmotic salinegroup （HO+IO）, as well as Hyperosmolar saline and1%L-carnitine group （HO+1%LCA）.The HO group was treated with500mOsmol/L sodium chloride solution; the HO+IO groupwith308mOsmol/L sodium chloride solution first and500mOsmol/L sodium chloridesolution30minutes later; and the HO+1%LCA group with1%of L-carnitine eye drop firstand500mOsmol/L sodium chloride solution30minutes later. Alternately,5times a day for28days.2. Preparation of L-carnitine eye drops1%of L-carnitine eye drops preparation: the L-carnitine powder soluble in water forinjection, sodium chloride powder to adjust the osmotic pressure to308mOsmol/L,hydrochloric acid to adjust pH to7.0.3. SchirmerⅠThe phenol red thread was placed in the eye canthus of BALB/c mice withmicro-tweezers under the slit lamp microscope. The phenol red thread was removed, thewetting length of the phenol red thread was measured after60s.4. Corneal fluorescein staining and scoringInstillation of1μl5%fluorescein sodium into the conjunctiva of BALB/c mice weremade, photographic records and scoring were made with cobalt blue light under the slitlamp microscope after3minutes.5. Tear fernsTear in fornix of BALB/c mice was collected with capillary tube, the tear was coated in a clean slide, dried at room temperature. Image analysis of fern crystal and photographicrecords were done under the microscope within48hours.6. Hematoxylin eosin staining and periodic acid-Schiff stainingThe morphology of corneal epithelial was observed by hematoxylin eosin stainingunder microscope. The thickness of the central corneal epithelial was measured by imageanalysis software.The top fornix of the conjunctival fornix was stained using periodicacid-Schiff staining.The number of fornix goblet cells was calculated under each visualfield.7. Scanning electron microscopeThe morphology of corneal surface was observed under scanning electron microscope.8. Determination of tear osmotic pressureFirst to establish tear osmotic pressure measurements conversion table in BALB/cmice. Collection of mouse tear and measurement of tear osmolarity on0,28thdays, basedon the establishment of the BALB/c mice, tear osmotic pressure was measured with theconversion table of tear osmolality.9. Stastistical methodsScores of corneal fluorescein were compared by Kruskal-Wallis H test. The thicknessof the corneal epithelium, tear secrection, the number of goblet cells and tear osmoticpressure were compared by ANOVA and LSD-t.Results1. Schirmer Ⅰ experimentalOn0,7^thdays, wetting length of phenol red thread of the mice in HO group, HO+IOgroup, HO+1%LCA group was not significantly statistically different （P>0.05）. On14^th,28^thday, wetting length of phenol red thread of the mice in three groups wassignificant difference（P＜0.05）. The wetting length of phenol red thread in HO group,HO+IO group, compared with that in HO+1%LCA group, was significantly decreased,however, no difference was shown in HO group and HO+IO group.2. Corneal fluorescein staining and scoringOn0day, the cornea of BALB/c mice in three groups was found a little fluoresceinstaining, and corneal score in each group showed no significant difference （P>0.05）. Withthe progress of the experience, punctate or sheet staining gradually emerged in the cornea of BALB/c mice in HO, HO+IO group. From the14^thday, in HO or HO+IO group,compared with HO+1%LCA group, the score of corneal fluorescein staining was elevated.On14^th,28^thday, corneal fluorescein staining score was significant difference in eachgroup.3. Tear fernsImage analysis of fern crystal and photographic records were done under themicroscope. Tear ferns is well formed in each groups. The fern in tear of BALB/c mice inHO, HO+IO group diminished from the7^thday, while at the same time, the tear fern isbetter in HO+1%LCA group than that in HO, and HO+IO group.4. Hematoxylin eosin stainingCorneal epithelial thickness was measured and corneal epithelium was observed. On0day, the corneal epithelial layer of BALB/c mice clearly showed4to5layers by HEstaining in each group, the basal cells arranged in columnar, cells became squamous whenclosed to the corneal surface, and corneal epithelial thickness in each group showed nosignificant difference （P>0.05）. With the experimental progress, corneal epithelial cellsappeared hyperplasia, the cell layers and corneal epithelium thickness were both increased.From the7^thday, the corneal epithelial thickness was significantly different in each group（P＜0.01）. On7^th,14^th,28thday, the corneal epithelial thickness in HO, HO+IO group,compared with HO+1%LCA group,was sharply increased.5. Periodic acid-Schiff stainingOn0day, clear goblet cells morphology and distribution in the epithelial cells, mainlyin the fornix, were observed in each group. The number of goblet cells was no difference inthree groups. From the7^thday, The number of goblet cells between the three groups wassignificant difference（P＜0.01）. On7^th,14^th,28thday, the number of goblet cells in HO,HO+IO group, compared with that in HO+1%LCA group, was significantly decreased, andthe difference in the each group was statistically significant.6. Scanning electron microscopyOn0day, epithelial cells were found intact, polygonal types and close connection.Microvilli were found on the cell surface. On the28thday, microvilli on corneal surfacewere reduced in HO and HO+IO group, compared with that in HO+1%LCA group.7. Determination of tear osmotic pressure Establishment of tear osmotic pressure measurements conversion table was done inBALB/c mice. On0day, tear osmotic pressure of the mice in HO group, HO+IO group,HO+1%LCA group was not significantly statistically different （P=0.712）. On the28thday, tear osmotic pressure of the mice in three groups was significant different（P=0.000）.The tear osmotic pressure in HO group, HO+IO group, compared with that inHO+1%LCA group, was significantly increased, however, no difference was shown in HOgroup and HO+IO group.Conclusion1. The local application of L-carnitine could slow down the pathological damage of theocular surface epithelial cells of mouse DED model.2. L-carnitine, through the role of osmoprotectants, had protective effect on ocularsurface epithelium of mouse dry eye model induced by hyperosmolar saline.3. The local application of eye drops containing L-carnitine and its derivatives mayhave the role of the clinical treatment of DED, and should be studied further.