`doi:l(). IDl6I'_i.ijum.2005.09.0l2, available unline at htq):/lwurw.acieneedinact.com
`
`r....nm....r14...: 9,:
`Or 1&.
`Majdlifllcial
`Surgegr
`
`Research Paper
`Wound Healing
`
`Histological evaluation of rabbit
`gingival wound healing
`transplanted with human
`amniotic membrane
`
`M. R|nastltE1,Harijadi’,
`A. L. S. Santnsoa, W. Sosruseno“
`‘Department oi Conservative Dentistry’,
`Faculty of Dentistry. Gadjah Mada University,
`Yogyakarla 55281, Indonesia; Zbepariment of
`Pathological Anatomy, Faculty of Medicine.
`Gadlah Mada University, Yogyakarla 55281,
`Indonesia; “Department oi Pedodontlc,
`Faculty oi Dentistry. Gadjah Mada University,
`Yogyakarta 55281, Indonesia; “Depanmem of
`Oral Biology, School oi Dental Sciences,
`Unlverslll Sains Malaysia, Kota Bharu 16150,
`Malaysia: “Department of Immunology.
`School 01 Medical Sciences, Unlversitl Sains
`Malaysia, Kata Bharu 16150, Malaysia
`
`M. Rirtastizi, Harijadi. A. L. S. Santoso, W. Sosroseno: Histological evaluation of
`rabbit gingival wound healing transplanted with human amniotic membrane. Int. J.
`0ralMaxi1[of2zc. Swfg. 2006; 35: 247-251. © 2005 International Association ofOral
`and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.
`
`Abstract. Human amniotic membrane has been used as a material to accelerate wound ’
`healing and reconstruct damaged organs. The aim ofthe present study was to assess
`histologically human amniotic membrane transplantation on 1-abbit’s gingival
`wound. Thrce— to 4-month-old male rabbits were divided into 2 groups, i.e., control
`(group I) and amniotic membrane-transplanted animals (group II). Buccal gingival
`wounds were created by a punch—biopsy instrument and covered by a 5-layered
`human amniotic membrane for group 11 or left uncovered for group I. Gingival
`biopsies were taken at days 1, 3, 5, 7 and 10, processed for parafiin sections and
`stained with haematoxylin—eosin or Von Gieson. Thickness of epithelial layer, the
`number of polymorphonuclear cells (PMN), fibroblasts and new blood vessels as
`well as density ofcollagen fibres were assessed. The results showed that the number
`of fibroblasts and new blood vessels, but not PMN, from group II was higher than
`that from group I (P < 0.05); Similarly, the epithelial thickness and density of
`collagen fibres-fmm group [I were significantly higher than those from group I
`{P < 0.05). The results of the present study indicate that amniotic membrane ~
`transplantation may induce rapid epithelialization and both granulation tissue and
`collagen formation but suppress inflammation, suggesting that amniotic membrane
`transplantation may promote rapid gingival wound healing in rabbits compared to
`secondary healing.
`
`Key words: amniotic membrane; ginglval; heal-
`ing; rabbit; wound.
`
`Accepted for publication 15 September 2005
`
`
`the innermost
`The amniotic membrane,
`layer ofthe foetal membrane, containsbase—
`mcnt membrane components and both
`growth factors and proteinasc inhibitors.
`The use of this material in surgical proce-
`dures such as a biological bandage for
`dressing burns’, andskin ulcers" and as
`
`an aid to physiological wound healing’
`has long been lfl'l0WI'l. The precise mechan-
`ism by which the amniotic membrane
`accelerates wound healing remains to be
`further investigated. Studies on corneal
`transplantation indicate that this membrane
`may enhance
`epitl-tclialization,
`inhibit
`
`both inflammation and scarring, regulate
`angiogenesis and posse anli—bac1ericidal
`activity3.
`The biological events of wound healing
`in gingival tissues are similar to those in
`the skin, but healing at the former site
`seems to be much faster and with less scar
`
`0901-5027103 0247 +05 $30.00/L)
`
`© 2005 International Association of Oral and Maxillofnciiil Surgeons. Published by Elsevicr Ltd. All rights reserved.
`
`MTF Ex. 1021, pg.1
`
`
`
`Histological assssment
`
`observers other than the authors were
`blinded to the treatment groups.
`
`Animals were sacrificed at days 1, 3, 5, 7
`and 10. Three animals of each group
`were ‘sacrificed at the indicated day. The
`buceal gingival tissue was resocted, fixed
`in 10% neutral formalin buffer, embedded
`in paraffin and sectioned serially at 6 pm.
`The sections were stained with either
`haematoxylin—eosin (HE) or von Gieson
`staining and viewed under a light micro-
`scope. The number of polymorphonuclear
`(PMN) cells, Fibroblasts and new blood
`vessels were then accounted. Fibroblasts
`were morphologically characterized as
`pink-stained stellate or spindle-shaped
`cells with blue-stained nucleus. Blood
`vessels were characterized as endothelial
`cell-assembled
`capillary vessels
`that
`contained erythrocytes‘. The epithelial
`thickness was
`determined
`by
`using
`visopan (Merck, Reiehert) and the mea-
`surement was then converted to micro-
`meter
`(l cm =0.075 um). Density of
`collagen fibres was scored using scales
`modified from a study by Tnuis et a1.”
`as
`follows:
`1 = few collagen fibres;
`2 = few and partially spread collagen
`fibres; 3 = few and fully spread collagen
`fibres and 4- = dense collagen fibres. The
`
`Statistical analysis
`
`The differences between groupsl and II
`and within each group on the number of
`PMN cells, fibroblasts and blood vessels
`were analysed by ‘t’ test and mu1ti-analy-
`sis of variance, respectively. The density
`of collagen fibres between groups I and II
`and within each group were analysed by
`Mann—Whitney
`[Most
`and Kmskal—
`Waflis, respectively. Data were calculated
`by using a statistical package (SPSS Inc.,
`Chicago, IL, USA).
`
`Results
`
`For the sake of brevity and clarity, photo-
`micrographs are only shown from both the
`control and arnniotic membrane-treated
`animals at day 10, otherwise indicated.
`Histological sections
`taken at day 1
`showed that necrotic tissues with many
`infiltrated inflammatory cells, particularly
`PMN, in the wounded site of group I were
`much prominent as compared with those
`in group II (Fig. 1A and B). G'Lt1givalPMN
`from both groups at day 10 could hardly be
`
`
`
`Iili2l3ia iI i3
`
`Pnnrmni
`
`{0}
`
`\;._\:§§_\
`'
`O .......t........
`.....t...............i......._...__:
`-3
`5
`'
`1
`
`........._... \
`.......
`T
`"to
`
`Days
`
`Fig. 1'. Large necrotic tissue with polyrnorphonuelear cells (PMNJ infiltration in the gingival
`wounded site from group I (A, lcflpanel) at day 1 is shown as compared wit.i1 thztifrom amniotic
`membrane-uansplanted animals (group II) (B, right panel) (HE, 400x). Panel (C) represents
`mean and standard deviation of the number ofPMN. Arrows indicate PMN.
`
`MTF Ex. 1021, pg.2
`
`248
`
`Rinastiti at al.
`
`formation as compared with that at the
`latter sites‘. Attempts to produce materials
`for enhancement of periodontal wound
`healing and reconstruction of periodontal
`tissue damage have been made. For exam-
`ple, enamel matrix proteins '7 and acellular
`dermal matrix allografiz‘ were applied on
`the gingival wound in humans and the
`results showed that these materials pro-
`mote both epitheliaiization and tissue vas-
`cularization and reduce inflammation.
`Since amniotic membrane may induce
`rapid physiological wound healing,
`the
`aim of the present study was to determine
`whether gingival wound healing in rabbits
`may be enhanced by transplantation of this
`membrane.
`
`Materials and methods
`
`C-ilngival wound preparation and amniotic
`membrane transplantation
`Three-
`to 4-month-old male rabbits
`(weight 1-2 kg) were divided into 2
`groups, each consisted of 15 animals.
`Groups I and II were a control and an
`amniotic rnembmne-treated group, respec-
`tively. All animals were anaeslhetized
`intramuscularly with Ketamine
`PIC]
`(50 pg kg" body weight)
`1 h before
`surgical procedures. Immediately before
`surgical procedure, animals were anaes-
`thctized intramusculariy with Ketamin
`HCl (80 mg kg"' body weight). Gingival
`wound was generated by using a punch-
`biopsy instrument (diameter 4 mm) at the
`buccal site of gingiva between maxillary
`incisor and premolar.
`‘The wound was
`left uncovered for group I or was covered
`by a sterile 5-layered human amniotic
`membrane (5mm x 5 mm in size)
`for
`group H. The freeze-dried human amniotic
`membranes were kindly donated by The
`Biomaterial Center and Tissue Bank, Dr
`Soetorno’s Hospital, School of Medicine,
`Airlangga University, Surabaya, Indone-
`sia. The basement membrane of the
`first layer was attached by using sterile
`saline onto the epithelial surface of the
`second
`layer. This
`procedure was
`repeated until all 5 layers were bound.
`The amniotic membrane was transferred
`to the gingiva with the basement mem-
`brane attaching the gingival wound and
`fitted to cover the whole wound by #5 -0
`vicryl suturing, followed by trimming off
`the excess portion. The analgesic medica-
`tion was orally given 3 times a day for 2
`days after surgical procedure. The experi-
`mental protocol was approved by the ethi-
`cal
`cornmittee, School of Medicine,
`Gadjah Mada University, Yogyakarta,
`lndoncsia.
`
`
`
`Human amniotic membrane enhances‘ rabbit gingival wound healing
`
`249
`
`observed (see Fig. 2A and B). Statistically,
`the number of gingival FMN from arnnio-
`tic membrane-treated animals (group II)
`was significantly lower than that from
`group I
`throughout
`the _ experiments
`(P < 0.05) (Fig. 1C).
`taken at day 10
`Photornicrographs
`showed that Stratified squamous epithe-
`lium from group II was thicker than that
`from group I and rate peg formation was
`seen only in group 11 (Fig. 2A and B). The
`thielcrress ofgingival epithelial layer in the
`wound site from both groups was steadily
`increased until day 10 (Fig. 2C). Statistical
`analysis revealed that gingival epithelial
`layer from group II was thicker than that
`from group I at days 5-10 (P < 0.05)
`(Fig. 2C).
`Histological appearance of wounded
`gingival tissues from both groups of ani-
`mals at day 10 showed that granulation
`tissues, which consisted of new blood
`vessels and large number of fibroblasts,
`from group H were much denser than those
`from group I (Fig. 3A and B). Blood
`vessels from group II were observed as
`early as day 3, whereas those from group I
`were seen at day 5 (P < 0.05) (Fig. 3C).
`The number of new blood vessels from
`group II at days 7 and 10 was significantly
`higher than that from group I (P'< 0.05)
`and fibroblasts at the wounded gingiva
`from both groups started to appear at
`day 3 (Fig. 3D). Further analysis revealed
`that the number of fibroblasts from group
`II at days 3-10 was higher than that from
`group I (P < 0.05) (Fig. 3D).
`Histological sections taken at day 10
`showed that
`immature and unorganized
`gingival collagen fibres from group I at
`day 10 were seen (Fig. 4A). However,
`dense and well-organized collagen fibres
`bcncath gingival epithelial layer of amnio-
`tic membrane-treated animals (group II)
`could be observed at the same period of
`experiment (Fig. 4B). Collagen fibre den-
`sity at the wounded gingival site of both
`groups gradually increased from days 3 to
`10. The density of collagen fibres from
`group I] at days 3 to lll were much higher
`than that from groupl (P < 0.05) (Fig. 4C).
`
`Dlscusslon
`
`The results of present study showed that
`amniotic membrane may suppress the
`migration of PMN at the wounded gingi-
`val site. Similar results have also been
`previously
`documented
`that
`human
`arrmiotic membrane transplantation in
`acute corneal alkali bum” and experimen-
`tal herpetic keratitiss may reduce PMN
`infiltratioll, thereby inhibiting inflamma-
`tion. Thc exact mechanism by which
`
`MTF Ex. 1021, pg.3
`
` 5
`
`o.o'o ‘
`
`1
`
`3
`
`5
`
`7
`
`to
`
`Fig. 2. The epithelial thickness at the gingival wounded site from the control (group I) (A. left
`panel) at day 10 is less as compared with that from amniotic membrane-transplantedanimals (group
`II) (B, right panel) (HE, 100x). Panel (C) represents mean and standard deviation ofthe epithelial
`tlriclmess.u.
`
`E.........................................................ICE
`:
`we
`B-'
`s
`
`§
`_E
`2
`l
`-3
`§ 6?
`3
`an
`5-
`>'
`2
`z
`-
`1.
`‘
`E 4;
`
`'
`-
`to
`.
`/-
`A
`
`
`‘r
`i
`—-‘
`
`T /’
`
`.........................................................50
`‘-'
`E.
`3
`<-
`3'.‘
`60 g
`‘.
`‘I
`was
`N
`-:
`» 13
`;'£'.
`g-
`1.
`igzcg
`7»:
`E to
`
`Q ...'2 ‘-7:
`
`\
`
`(Cl.
`
`Days
`
`(D)
`
`Fig. 3 . Increased number ofnew blood vessels and fibroblasts in the gingival wounded site from
`anuriotic membrane-transplanted animals (group II) (B, right panel) is soon as compared with
`that from group I(A, left panel) at day 10 (HE. 400x). Panel (C) represents mean and standard
`deviation of the ntunber of blood vessels and fibroblasts. Thick and thin arrows indicate blood
`vessels and fibroblasts, respectively.
`'
`
`‘r’
`
`i
`i
`.=
`
`5
`
`3
`
`.
`wc‘-—
`‘
`‘
`“‘~" fl
`
`.
`.r
`E
`
`it
`!"
`
`92::
`
`5.
`3'
`.
`0.10 yIk
`3*r
`
`r I
`
`’
`
`-‘E
`"u:
`3
`fl.E
`r~
`
`
`
`250
`
`Rinastiti et a!.
`
`r l' r_:5;3;u:u:i:xu- uuuuizmz:
`
`. Eli?
`
`(C)
`
`Ufiilffi‘
`
`Fig. 4. Collagen fibres at the gingival wounded site from amniotic membrane—Lransp1a.nted
`animals (group II) (B, right panel) at day 10 are much denser than those from the control (group
`I) (A, left panel) (Von Gieson, 400x). Panel (C) represents mean and standard deviation of
`collagen density. Arrow indicates collagen fibres,
`
`aumiotic membrane reduced the number
`of gingival PMN seen in the present study
`is unclear, but may be via the induction of
`cell apoptosis", inhibition of cell migra-
`tion due to the suppression of chemolrine
`and anti-inflammatory cytokinc (c.g., IL-
`10) synthesis” and thc prcvcnfiort of
`microbial contamination due to its ability
`to tightly attach an the wound”.
`The induction of rapid epitheliallzatlon
`of the gingival wound by amniotic mem-
`brane was demonstrated in the present
`study. These results support the previous
`studies showing increased epitltelia1iza-
`tion in alkali-burned corneal
`surfaceg
`and experimental hcrpetic lceratitis‘ trans-
`planted with amniotic membrane. Anu1io-
`tic membrane contains growth factors.
`necessary for cpithclializationl 1, and pro-
`vides substrates such as laminins for rapid
`epithelial cell attachment which, in turn,
`up-regulates the expression of growth fac-
`tor receptors such as EGF receptor on
`epithelial cells”'”. Therefore, one may
`assume that these growth factors might
`induce rapid gingival epithelial cell migra-
`tion and attachment to the wound area,
`thereby stimulating rapid cell proliferation
`
`and differentiation. This contention is,
`however, speculative and further studies
`are needed to clarify.
`Formation of granulation tissue marks
`the proliferative stage of wound healing
`and is cbaractcrizcd by dense fibroblast-
`derived cellular matrix and neovascular-
`ization. Indeed, the present study showed
`that amniotic membrane may stimulate
`rapid formation of granulation tissue i.n
`the gingival wound by inducing rapidly
`increased number of fibroblasts and vas-
`cularization. Since amniotic membrane
`contains bFGF, EGF, TGF-B and inter-
`leukin-l (IL-1 )1 1"“, one may assume that
`these growth factors may stimulate fibro-
`blast growth and ncovascularization in the
`gingival wound. In contrast. previous stu-
`dies demonstrated that amniotic mem-
`brane transplanted on the cornea in vitro
`and in vivo suppressed neovasculariza—
`tion‘°"3. The exact reason to explain this
`discrepancy is not clear. FAULK et al.‘
`showed that neovascularization in human
`leg ulcers could be induced by human
`amnion membrane transplantation. There-
`fore, onc may assume that neovascu]ariza-
`tion of the wound healing following
`
`amniotic
`transplantation with human
`membrane is dependent on the anatomical
`site of the wound, being that amniotic
`membrane-induced suppression of neo-
`vascularization may solely occur in the
`wounded cornea. Further
`studies
`are
`needed to delineate this contention.
`The ability of amniotic membrane to
`stimulate rapid production of collagen
`fibres in the gingival wound was also
`suggested in the present study. This mem-
`brane contains TGF—B and tissue inhibi-
`tors of metalloproteinase polypeptides
`(TIMPs), which up-regulate the produc-
`tion of collagen fibres by fibroblasts in the
`wound hcaling7'22. If so,
`that
`increased
`density of collagen fibres in the amniotic
`membrane-treated gingival wound seen in
`the present study may be augmented by
`TJMPS and 'I‘GF—B carurot be ruled out and
`requires fiirther studies.
`The histological findings of rabbit gin-
`gival wound healing transplanted with
`amniotic membrane seen in the present
`study seem to he in consistence with those
`of human periodontal wound healing
`guided with enamel matrix proteins” or
`accllular dermal matrix allograft“. More-
`over,
`cryopreserved human amniotic
`membrane transplantation stimulates very
`minimal allo and xenografl rejection due
`to the fact that this membrane expresses
`low major histocompatibility complex
`class II molecules but high immunoregu-
`latory molecules such as Fas ligand and
`IILA-G”. Therefore, the use of amniotic
`membrane transplantation to accelerate
`gingival wound healing in humans,
`regardless of the genetic background,
`is
`promising.
`
`Acknowledgements. 'I'his work was part of
`the thesis for Masters degree (to M.R.) in
`Gadjah Mada University and supported by
`the Postgraduate Scholarship (to MR.)
`from the Ministry of National Education,
`the Indonesian Govcmmcnt The authors
`gratefiilly-thank Dr Abdurmhman (School
`of Medicine, Airlnngga University) for
`providing human amniotic membranes
`and Mrs Ika Rahutarni for statistical ana-
`lysis.
`
`References
`
`1. ALTAVILLA D, Sarrrn A, Cuomorrn D,
`GALEANO M, Dnomrro B, COLONNA M,
`Totuua V, Russo G, SARDELLA A, URNA
`G, Carvrro GM, CAVALLAR1 V, SQUA-
`nturo G, Sounnmro F. Inhibition oflipid
`pcroxidation restores impaired vascular
`endothelial growth factor expression
`and stimulates wound healing and angio-
`
`_
`
`MTF Ex. 1021, pg.4
`
`
`
`Human arrmiotic membrane enhances rizbbit gingival wound healing
`
`251
`
`genesis in the genetically diabetic mouse.
`Diabetes 2001: 50: 667-674.
`.Btit:rM:tN 8, You I... Srmoau U,
`ROHRBCHNELDER K,
`\7ot.cKt—:1z HE,
`Ktzusa FE. Amniotic membrane down-
`regulates chcmoltine expression in human
`keratinooytes. Invest Opthalrnol Vis Sci
`1999: 40(Suppl.):S578.
`. DUA HS, Gowns JAP, KING AJ, Manna-
`AJAN VS. The amniotic membrane in
`ophthalmology. Surv Ophthalmol 2004:
`49: 51-77.
`. FAULK WP, iV]ATTHEW'S R, S'l‘E\’EN8 PI,
`BENNETT JP, BURGDS H, Hts BL. Human
`amnion as an adjunct in wound healing.
`Lancet 1980: 1: ll56—ll58.
`. Gomns MF, Dos Amos MJ, NOGUELRA
`TO, GUIMARAES SA. I-Iistologic evalua-
`tion of the osteoinductivc property of
`autogenous dernineralized dentin matrix
`on surgical bone defects in rabbit skull
`using human amniotic membrane for
`guidod bone regeneration. IntJ Oral Max-
`illofac Implant 2001: 16: 563-571.
`. H.RKI<n~u=.N L, Un'"ro V-J, LARJAVA H.
`Cell biology of gingival wound healing.
`Periodontol 2000 2000: 24: 127-152.
`. HAD Y, MA DI-I, HWANG DG. Identifica-
`tion of anti-angiogenic and anti-inflam-
`matory proteins
`in human amniotic
`membrane. Cornea 2000: 19: 348-352.
`. HEILIGIDIHAUS A, BAUER D, MELLER D,
`STEUBL K-P, Tsaruo SCG. Improvement
`of HSV-l nocrotizing lceratitis with
`amniotic membrane
`transplantation.
`Invest Optltalmol Vis Sci 2001: 42:
`I969-1974.
`. KIM IS, KIM JC,N4tBI-C,Jr:.oNo IM, SONG
`CY. Amniotic membrane patching pm-
`motes healing and inhibits proteinase
`activity on wound healing following
`
`10.
`
`ll.
`
`12.
`
`13.
`
`14.
`
`15.
`
`l6.
`
`l7.
`
`acute corneal alkali burn. Exp Eye Res
`2000: 70: 329-337.
`Kosmmsnt N, KABUYAMA Y, SASAKI S.
`Karo K, I-Iomut Y. Suppression of cor-
`neal neovascularization by culture super-
`natant of human amniotic cells. Cornea
`2002: 21: 62-67.
`Korztmt N, Imtmoro T, SoTozoNo T,
`Futnwooo NJ, QUANTOCK AJ, KINOSITA
`S. Growth factors mRNA and proteins in
`preserved human amniotic membrane.
`Curr Eye Res 1997: 20: 173-177.
`Kuno M, SONODA Y, Mumuatnisu R,
`Usur M.
`Immunogenicity of human
`amniotic membrane
`in
`experimental
`xenotransplantation.
`Invest Opthalmol
`Vis Sci 2001: 42: 1539-1546.
`Kunmucus MA, DANE-BHVAR C, DAVEN-
`PORT .1, Kim A. Human comm] epithelial
`cell adhesion to laminins. Curr Eye Res
`1999: 19: 106-114.
`MARTINEZ-PARDO N[E,R1'.vr=.s FRIAB ML,
`Ramos DURON LE, Gurtsnssz SALGADD
`E, Gomez IC, MARIN MA, LUNA ZAR-
`AGOZA D. Clinical application of amnio- -
`tic membranes
`on
`a
`patient with
`epidcrrnolysis bullosa. Ann Transplant
`1999: 4: 68-73.
`PAXK WC, Tsnrto SCG. Modulation of
`acute
`inflammation and keratinocyte
`death by suturing, blood and amniotic
`membrane in PRK.
`Invest Opthalmol
`Vis Sci 2000: 41: 2906-2914.
`Rom NS, Manon R, Swan KF, LYNDEN
`TW. Expression of IL-I and IL-6 protein
`and mRNA in amniochorionic mem-
`brane. Placenta 1993: 14: A63.
`SALLUM EA, Cnsan MZ, CAFFESSE. RG,
`Fums LP, Nocm Ir FH. SALLUM AW.
`Coronally positioned flap with or without -
`enamel matrix protein derivative for the
`
`18.
`
`19.
`
`20.
`
`21.
`
`22.
`
`treatment of gingival recessions. Am J
`Dent 2003: 16: 287-291.
`SHAO C, SIMA I, ZHANG SX, Jun J,
`REINACH P, WANG Z, MA IX. Suppres-
`sion of corneal ncovascularization by
`PEDF release from human amniotic
`membranes. Invest Opthalmol Vis Sci
`2004: 45: 1758-1762.
`TALAS DU, NAYCI A, Ans S, POLAT A,
`CDMELEKOGLU U, BAGDATOGLU C,
`RENDA N. The effects of corticosteroids
`on the healing of tracheal anastomosos in
`a rat model. Pharrnacol Res 2002: 45:
`299-304.
`V
`TALMI YP, SIGLEE L, mot: E. Antibac-
`tcrial properties ofhuman amniotic mem-
`brane. Placenla 1991: 12: 235-288.
`WAGS!-IALL E, LEWIS Z, Bnnren SB,
`SINENSKY MC, HOCHBERG M. Accllular
`dcrrnal matrix allograft in the treatment of
`mucogingival defects in children:
`illus-
`trative case report ASDC J Dent Child
`2002: 69: 39-43.
`Ym H-K, FANG C-P, FAN DS, FAN 13-],
`Yu EY-W, LAM D5-C. Growth factor
`changes in ex vivo human expansion of
`human limbal epithelial cells on human
`amniotic membrane. Cornea 2002: 21:
`101-105.
`
`Address:
`Wihas Sormserto
`Department of Oral Biology
`School ofDen1al Sciences
`Universiti Sam: Malaysia
`Kata Bharu 16150
`Malryria
`Fax: +60 9 7642026.
`E-mail: wihaskora@kb.u.rm.my
`
`MTF Ex. 1021, pg.5