Current Eye Research
Volume I 1 number 12 1992, 1161-1172
Human recombinant bFGF stimulates corneal endothelial wound healing in rabbits
Peter Rieck’,*, Christian Hartmann3, Claudine Jacob4, Yves Pouliquen2 and Yves Courtois’
Curr Eye Res Downloaded from informahealthcare.com by University of Otago on 12/28/14 For personal use only.
‘INSERM U 118-CNRS UA 630, Assoc. Claude Bernard, Unite de Recherches Gerontologiques, 29 rue Wilhem, 75016 Paris, 21NSERM U 86, Laboratoire d’Ophtalmologie, Hotel Dieu de Paris, 1 place Parvis de Notre Dame, 75004 Paris, France, 3University Eye Clinic, Dept of Immunology, Joseph-Stelzmann Strasse 9, 5000 Koln 41, Germany and 4Farmitalia Carlo Erba, Via Carlo Imbonati 24, 20159 Milan, Italy
ABSTRACT We have previously shown that bovine, human placenta extracted and recombinant human basic Fibroblast Growth Factor (bFGF) are effective in enhancing corneal epithelial wound healing in vivo. In the present study, we investigated the effect of rh-bFGF on the regeneration of injured rabbit endothelium. A standardized wound was created by scraping of endothelial cells with a special device within the boundaries of a central epithelial trephine mark of 7 mm in diameter. A single dose of 1.5 pg rh-bFGF was injected into the anterior chamber immediately after wounding, while control eyes received the vehicle only (n = 27). Functional recovery and wound closure rates were assessed by means of ultrasonic pachymetry, corneal button wet weight, endothelial vital staining as well as direct computer assisted surface analysis of Janus green stained corneal buttons. Measurements were carried out 1, 2 , 4 , and 7 days after injury. Morphological evaluation and cell counts at D4 and D7 were also performed. Significant stimulation of endothelial regeneration in rh-bFGF treated eyes, was observed with all methodological approaches. These results demonstrate the effectiveness of rh-bFGF in enhancing experimental corneal endothelial wound healing and advocate for a possible clinical application of this growth factor in order to preserve endothelial cell function or to promote healing of this important monolayer in case of disease or injury
.
INTRODUCTION A functional endothelium is essential to corneal transparency. As the human corneal endothelium shows essentially no proliferative activity, the continuous loss of endothelial cells with age, trauma, intraocular surgery or disease as corneal dystrophies may lead to a reduction in the
tissue’s physiologic reserve and finally to endothelial decompensation with resulting corneal opacity. An exciting prospect is given by the possibility to pharmacologically stimulate the proliferation of this delicate monolayer. Promising agents in this context are mitogenic growth factors, and several pieces of evidence indicate that especially basic Fibroblast Growth Factor (bFGF) might play an important role to stimulate and maintain corneal endothelial cells. FGFs are polypeptides which form a family of heparin binding growth factors, actually consisting of seven members. The two major forms, acidic and basic FGF, share a 558 absolute homology in their primary structure (1) and have been isolated from various tissues (see (2) for review). In vitro, both a and bFGF act on a wide variety of cells from neuroectodermal and mesodermal origin and affect cell differentiation, morphology, regeneration and senescence (3). FGF interacts with specific cell surface receptors ( 4 ) and the genes encoding these receptors have recently been isolated (5,6). FGFs have a high affinity for heparin and heparin-like structures, and heparan sulfate proteoglycans have been characterized as the major component of basement membranes which bind both forms of the growth factor (7,8). Indeed, labelled acidic and basic FGF applied on wounded corneas in vivo or on frozen sections of the cornea in vitro
Received on March 30. 1992; accepted on November 6, 1992
19 Oxford University Press
1161
Current hYe
Curr Eye Res Downloaded from informahealthcare.com by University of Otago on 12/28/14 For personal use only.
Research selectively binds to the basal laminas (9-11). However, little is actually known about this growth factor's effects in vivo. Recent progress in recombinant DNA technology with the possibility to produce growth factors on a large scale makes it possible to consider the therapeutic use of this protein, as it obviates the laborious and scarcely effective preparation of FGF from natural sources, such as bovine brain or human placenta. In the present study, we therefore investigated the ability of a recombinantderived human bFGF to stimulate endothelial regeneration in a rabbit in vivo model. MATERIAL AND METHODS Experimental animals 27 NZW rabbits, aged 6 months and weighing 4 - 5,5 kg were used for these experiments. Animal care practices were in compliance with the "Guiding Principles in the Care and the Use of Laboratory Animals" (NIH publication 86-23) and with the ARVO resolution on the Use of Animals in Research. Prior to surgery, the eyes were investigated for any pathology of the anterior segment under an operating microscope. At the end of each experimental series, the animals were sacrificed by an overdose of pentobarbital. Recombinant human bFGF (rh-bFGF) rh-bFGF was provided by Farmitalia Carlo Erba as a freeze-dried preparation containing 50 pg of active ingredient, 1 mg of mannitol, 1.5 mg of NaH,PO, and 0 . 4 mg of NaC1. Each vial was reconstituted with a solution containing mannitol (0.25 mg/ml), NaH,PO, (0.37 mg/ml) and NaCl (0.1 mg/ml) (mannitol containing medium, MCM) to obtain a final bFGF concentration of 10 pg/ml. In vitro biological activity of the treatment solution was assessed using the
1162
assay reported by Plouet et al. (12) measuring tritiated thymidine incorporation into cellular DNA of cultured bovine epithelial lens cells. The activity was confirmed to be in the range of that reported in the literature and of our laboratory experience (EC,, = 50-200 pg/ml). Endothelial wound model A special device was constructed f o r this purpose. It consists essentially of a 12-gauge Touhy needle normally used for epidural anesthesia, through which both ends of a thin metal wire were introduced, so as to form a loop at the tip of the needle (Fig. 1). 30 minutes before surgery, the animals were anesthetized by intramuscular injection of 1 mg flunitrazepam and ketaminehydrochloride (30mg/kg). Atropine sulfate drops (1%) as well as oxybuprocaine drops (0,4%)were topically applied immediately prior to surgery. The wounding procedure was performed under an operating microscope. A plastic surgical drape was applied to the eye and three 5-0 silk sutures were placed through the episclera for fixation of the globe. In order to outline wound margins, a superficial epithelial mark was produced by means of a conventional handtrephine (7mm in diameter). A 2-3 mm limbal incision was made with a Beaver blade (No. 65) and the instrument was introduced into the anterior chamber through this opening. Although a little amount of aqueous generally escaped, care was taken to maintain good depth of the anterior chamber. By carefully advancing the loop of the wire through the needle towards the inner corneal surface, endothelial cells were gently scratched within the bounderies of the trephine mark (Fig. 1). After complete endothelial abrasion
Curr Eye Res Downloaded from informahealthcare.com by University of Otago on 12/28/14 For personal use only.
Current Eye Research the instrument was withdrawn and the incision was closed by 2 separated 10-0 nylon sutures. Treatment regimen About 15 minutes after the wounding procedure, one eye of each rabbit received an intraocular injection of 1.5 pg rh-bFGF in 150 p1 using a sterile 30-gauge needle attached to a microsyringe. The contralateral eye received the same amount of the vehicle only. At the end of the operation, both eyes received ophthalmic gentamycin sulfate drops and atropin sulfate (1%) drops: this treatment was continued twice daily f o r two to three days postoperatively. Fo1low-up Post-operative clinical observations were performed every day in order to detect any pathological changes related to the treatment. Time points for the following assessments were 1, 2 , 4 and 7 days (n=6 per group) after operation, except for cell density determination, performed at D4 and D7. Three parameters of post-operative endothelial status were analyzed: 1. Functional recovery. a) Central corneal thickness. In vivo central corneal pachymetric measurements were performed using a DGH 1000 ultrasonic pachymeter (DGH Technology, Frazer, Pa), equipped with a hand-held transducer. b) Wet weight of corneal buttons. After dissection of enucleated eyes, the central 9 mm of the cornea was trephined and corneal buttons were immediately weighed using scales (Sartorius, Germany) with a precision range of 0.1 mg. 2 . Wound surface regression on time. The endothelial side of dissected corneoscleral buttons was stained for 1.5 minutes with a 0.1% solution of Janus green (Merck, Ger- many). This supravital dye stains denuded Descemet's membrane in a dark blue
color. Stained corneas were trephined (9 mm diameter) and immediately subjected to computerized image analysis. The whole endothelial surface was digitized by means of a high resolution video camera (Panasonic, Japan). Multistep segmentation of the cornea, detection of the wound edges, rejection of artifacts and calculation of the wound area was performed semiautomatically using an Image Analysis software package (ImageniaR;Biocom, France). In three animals, this procedure was performed at time point 0, directly after wounding, to determine the accuracy and reproducibility of the initial wound size compared to the surface outlined by the epithelial trephine mark. 3. Endothelial morphology and cell density. The general morphological appearance of regenerating endothelium after FGF or placebo treatment was studied by staining endothelial cellular borders of freshly dissected corneal buttons with Alizarin Red S (ARS, 0.2%, 2 min). At D4 and D 7 , central endothelial cell density ( E C D ) was determined by computerized image analysis using the same equipment as described above. After staining with ARS, the central part of the cornea was photographed and the negatives were subjected to digitization. Ten square fields (100x100 pm each) were randomly selected from each micrograph and automatically traced by,the analysing system. Cell borders inside the square field were outlined following the precipitation of the dye and subsequent cell counting was processed by the computer. Mean ? S E were determined and the values multiplied by 100 to express results as cell number/mm2. Analysis of the data For all animals and all variables, the difference between the rh-bFGF treated eye
1163
Curr Eye Res Downloaded from informahealthcare.com by University of Otago on 12/28/14 For personal use only.
Current Eye Research
Fig. 1 Schematic representation of the endothelial wound model involved in this study including the needle used €or endothelial scraping. After incision at the limbus, the Thouhy needle was inserted into the anterior chamber and endothelial cells were
* v) v)
u
2 Y
scraped within the boundaries of a previously produced circular epithelial outline by means of a thin metal wire introduced through the needle. The detailed procedure is described in Materials and Methods.
**
*** mm
900
mfGF
800
0 I-
700
-I
22
600
K 0 500 0 400
300
0
1
2
4
7
TIME (DAYS)
Fig. 2 Corneal thickness after endothelial scraping as a function of time. Determinations were made by ultrasonic pachymetry. After injection of 1.5 pg rh-bFGF, corneal
thinning occured significantly faster than in control eyes (placebo injection). *: p=O.O49, **: p
Volume I 1 number 12 1992, 1161-1172
Human recombinant bFGF stimulates corneal endothelial wound healing in rabbits
Peter Rieck’,*, Christian Hartmann3, Claudine Jacob4, Yves Pouliquen2 and Yves Courtois’
Curr Eye Res Downloaded from informahealthcare.com by University of Otago on 12/28/14 For personal use only.
‘INSERM U 118-CNRS UA 630, Assoc. Claude Bernard, Unite de Recherches Gerontologiques, 29 rue Wilhem, 75016 Paris, 21NSERM U 86, Laboratoire d’Ophtalmologie, Hotel Dieu de Paris, 1 place Parvis de Notre Dame, 75004 Paris, France, 3University Eye Clinic, Dept of Immunology, Joseph-Stelzmann Strasse 9, 5000 Koln 41, Germany and 4Farmitalia Carlo Erba, Via Carlo Imbonati 24, 20159 Milan, Italy
ABSTRACT We have previously shown that bovine, human placenta extracted and recombinant human basic Fibroblast Growth Factor (bFGF) are effective in enhancing corneal epithelial wound healing in vivo. In the present study, we investigated the effect of rh-bFGF on the regeneration of injured rabbit endothelium. A standardized wound was created by scraping of endothelial cells with a special device within the boundaries of a central epithelial trephine mark of 7 mm in diameter. A single dose of 1.5 pg rh-bFGF was injected into the anterior chamber immediately after wounding, while control eyes received the vehicle only (n = 27). Functional recovery and wound closure rates were assessed by means of ultrasonic pachymetry, corneal button wet weight, endothelial vital staining as well as direct computer assisted surface analysis of Janus green stained corneal buttons. Measurements were carried out 1, 2 , 4 , and 7 days after injury. Morphological evaluation and cell counts at D4 and D7 were also performed. Significant stimulation of endothelial regeneration in rh-bFGF treated eyes, was observed with all methodological approaches. These results demonstrate the effectiveness of rh-bFGF in enhancing experimental corneal endothelial wound healing and advocate for a possible clinical application of this growth factor in order to preserve endothelial cell function or to promote healing of this important monolayer in case of disease or injury
.
INTRODUCTION A functional endothelium is essential to corneal transparency. As the human corneal endothelium shows essentially no proliferative activity, the continuous loss of endothelial cells with age, trauma, intraocular surgery or disease as corneal dystrophies may lead to a reduction in the
tissue’s physiologic reserve and finally to endothelial decompensation with resulting corneal opacity. An exciting prospect is given by the possibility to pharmacologically stimulate the proliferation of this delicate monolayer. Promising agents in this context are mitogenic growth factors, and several pieces of evidence indicate that especially basic Fibroblast Growth Factor (bFGF) might play an important role to stimulate and maintain corneal endothelial cells. FGFs are polypeptides which form a family of heparin binding growth factors, actually consisting of seven members. The two major forms, acidic and basic FGF, share a 558 absolute homology in their primary structure (1) and have been isolated from various tissues (see (2) for review). In vitro, both a and bFGF act on a wide variety of cells from neuroectodermal and mesodermal origin and affect cell differentiation, morphology, regeneration and senescence (3). FGF interacts with specific cell surface receptors ( 4 ) and the genes encoding these receptors have recently been isolated (5,6). FGFs have a high affinity for heparin and heparin-like structures, and heparan sulfate proteoglycans have been characterized as the major component of basement membranes which bind both forms of the growth factor (7,8). Indeed, labelled acidic and basic FGF applied on wounded corneas in vivo or on frozen sections of the cornea in vitro
Received on March 30. 1992; accepted on November 6, 1992
19 Oxford University Press
1161
Current hYe
Curr Eye Res Downloaded from informahealthcare.com by University of Otago on 12/28/14 For personal use only.
Research selectively binds to the basal laminas (9-11). However, little is actually known about this growth factor's effects in vivo. Recent progress in recombinant DNA technology with the possibility to produce growth factors on a large scale makes it possible to consider the therapeutic use of this protein, as it obviates the laborious and scarcely effective preparation of FGF from natural sources, such as bovine brain or human placenta. In the present study, we therefore investigated the ability of a recombinantderived human bFGF to stimulate endothelial regeneration in a rabbit in vivo model. MATERIAL AND METHODS Experimental animals 27 NZW rabbits, aged 6 months and weighing 4 - 5,5 kg were used for these experiments. Animal care practices were in compliance with the "Guiding Principles in the Care and the Use of Laboratory Animals" (NIH publication 86-23) and with the ARVO resolution on the Use of Animals in Research. Prior to surgery, the eyes were investigated for any pathology of the anterior segment under an operating microscope. At the end of each experimental series, the animals were sacrificed by an overdose of pentobarbital. Recombinant human bFGF (rh-bFGF) rh-bFGF was provided by Farmitalia Carlo Erba as a freeze-dried preparation containing 50 pg of active ingredient, 1 mg of mannitol, 1.5 mg of NaH,PO, and 0 . 4 mg of NaC1. Each vial was reconstituted with a solution containing mannitol (0.25 mg/ml), NaH,PO, (0.37 mg/ml) and NaCl (0.1 mg/ml) (mannitol containing medium, MCM) to obtain a final bFGF concentration of 10 pg/ml. In vitro biological activity of the treatment solution was assessed using the
1162
assay reported by Plouet et al. (12) measuring tritiated thymidine incorporation into cellular DNA of cultured bovine epithelial lens cells. The activity was confirmed to be in the range of that reported in the literature and of our laboratory experience (EC,, = 50-200 pg/ml). Endothelial wound model A special device was constructed f o r this purpose. It consists essentially of a 12-gauge Touhy needle normally used for epidural anesthesia, through which both ends of a thin metal wire were introduced, so as to form a loop at the tip of the needle (Fig. 1). 30 minutes before surgery, the animals were anesthetized by intramuscular injection of 1 mg flunitrazepam and ketaminehydrochloride (30mg/kg). Atropine sulfate drops (1%) as well as oxybuprocaine drops (0,4%)were topically applied immediately prior to surgery. The wounding procedure was performed under an operating microscope. A plastic surgical drape was applied to the eye and three 5-0 silk sutures were placed through the episclera for fixation of the globe. In order to outline wound margins, a superficial epithelial mark was produced by means of a conventional handtrephine (7mm in diameter). A 2-3 mm limbal incision was made with a Beaver blade (No. 65) and the instrument was introduced into the anterior chamber through this opening. Although a little amount of aqueous generally escaped, care was taken to maintain good depth of the anterior chamber. By carefully advancing the loop of the wire through the needle towards the inner corneal surface, endothelial cells were gently scratched within the bounderies of the trephine mark (Fig. 1). After complete endothelial abrasion
Curr Eye Res Downloaded from informahealthcare.com by University of Otago on 12/28/14 For personal use only.
Current Eye Research the instrument was withdrawn and the incision was closed by 2 separated 10-0 nylon sutures. Treatment regimen About 15 minutes after the wounding procedure, one eye of each rabbit received an intraocular injection of 1.5 pg rh-bFGF in 150 p1 using a sterile 30-gauge needle attached to a microsyringe. The contralateral eye received the same amount of the vehicle only. At the end of the operation, both eyes received ophthalmic gentamycin sulfate drops and atropin sulfate (1%) drops: this treatment was continued twice daily f o r two to three days postoperatively. Fo1low-up Post-operative clinical observations were performed every day in order to detect any pathological changes related to the treatment. Time points for the following assessments were 1, 2 , 4 and 7 days (n=6 per group) after operation, except for cell density determination, performed at D4 and D7. Three parameters of post-operative endothelial status were analyzed: 1. Functional recovery. a) Central corneal thickness. In vivo central corneal pachymetric measurements were performed using a DGH 1000 ultrasonic pachymeter (DGH Technology, Frazer, Pa), equipped with a hand-held transducer. b) Wet weight of corneal buttons. After dissection of enucleated eyes, the central 9 mm of the cornea was trephined and corneal buttons were immediately weighed using scales (Sartorius, Germany) with a precision range of 0.1 mg. 2 . Wound surface regression on time. The endothelial side of dissected corneoscleral buttons was stained for 1.5 minutes with a 0.1% solution of Janus green (Merck, Ger- many). This supravital dye stains denuded Descemet's membrane in a dark blue
color. Stained corneas were trephined (9 mm diameter) and immediately subjected to computerized image analysis. The whole endothelial surface was digitized by means of a high resolution video camera (Panasonic, Japan). Multistep segmentation of the cornea, detection of the wound edges, rejection of artifacts and calculation of the wound area was performed semiautomatically using an Image Analysis software package (ImageniaR;Biocom, France). In three animals, this procedure was performed at time point 0, directly after wounding, to determine the accuracy and reproducibility of the initial wound size compared to the surface outlined by the epithelial trephine mark. 3. Endothelial morphology and cell density. The general morphological appearance of regenerating endothelium after FGF or placebo treatment was studied by staining endothelial cellular borders of freshly dissected corneal buttons with Alizarin Red S (ARS, 0.2%, 2 min). At D4 and D 7 , central endothelial cell density ( E C D ) was determined by computerized image analysis using the same equipment as described above. After staining with ARS, the central part of the cornea was photographed and the negatives were subjected to digitization. Ten square fields (100x100 pm each) were randomly selected from each micrograph and automatically traced by,the analysing system. Cell borders inside the square field were outlined following the precipitation of the dye and subsequent cell counting was processed by the computer. Mean ? S E were determined and the values multiplied by 100 to express results as cell number/mm2. Analysis of the data For all animals and all variables, the difference between the rh-bFGF treated eye
1163
Curr Eye Res Downloaded from informahealthcare.com by University of Otago on 12/28/14 For personal use only.
Current Eye Research
Fig. 1 Schematic representation of the endothelial wound model involved in this study including the needle used €or endothelial scraping. After incision at the limbus, the Thouhy needle was inserted into the anterior chamber and endothelial cells were
* v) v)
u
2 Y
scraped within the boundaries of a previously produced circular epithelial outline by means of a thin metal wire introduced through the needle. The detailed procedure is described in Materials and Methods.
**
*** mm
900
mfGF
800
0 I-
700
-I
22
600
K 0 500 0 400
300
0
1
2
4
7
TIME (DAYS)
Fig. 2 Corneal thickness after endothelial scraping as a function of time. Determinations were made by ultrasonic pachymetry. After injection of 1.5 pg rh-bFGF, corneal
thinning occured significantly faster than in control eyes (placebo injection). *: p=O.O49, **: p
