- Email: [email protected]
Proliferation and apoptosis in the corneal stroma in longterm follow-up after photorefractive keratectomy
ARTICLE IN PRESS
Pathology – Research and Practice 201 (2005) 399–404 www.elsevier.de/prp
ANIMAL AND IN VITRO MODELS IN HUMAN DISEASES
Proliferation and apoptosis in the corneal stroma in longterm follow-up after photorefractive keratectomy No´ra Szentma´rya,, Zolta´n Zsolt Nagya, Miklo´s Rescha, Be´la Szendeb,c, Ildiko´ Su¨vegesa a
1st Department of Ophthalmology, Semmelweis University, Budapest 1st Department of Pathology and Experimental Cancer Research, Semmelweis University, Budapest c Research Group of Molecular Pathology, Hungarian Academy of Sciences and Semmelweis University, Budapest b
Received 14 October 2004; accepted 13 April 2005
Abstract This study aimed at investigating the proliferation and apoptosis of corneal cells following photorefractive keratectomy (PRK) treatment. PRK (6.0 D correction) was performed with the Asclepion-Meditec MEL70 G-scan excimer laser on the right eye of each of 33 rabbits under combined local and general anaesthesia. Animals were sacrificed at 4 h, 1, 4, 7, 14, 28, 56, and 112 days postoperatively, and corneal samples from these eight groups were examined histologically. Stromal cell proliferation was evaluated by immunocytochemical analysis of Ki67. Apoptosis was detected using the terminal deoxyribonucleotidyl transferase-mediated dUTP-digoxigenin nick-end labeling (TUNEL) assay method. The untreated left eyes served as controls. Ki67 positivity was detected in the upper stroma on day 1, 4, 7, and 14, and keratocyte apoptosis on day 1, 4, 7, and 14 after PRK, but not at an earlier or later time. Neither Ki67 positivity nor apoptotic activity was observed in the controls (untreated corneas). PRK was found to trigger proliferation and apoptosis of corneal keratocytes. The frequency and spatial distribution of keratocyte proliferation and apoptosis are likely to be important determinants of the corneal wound healing process, but the detailed regulatory mechanisms have not yet been characterized. r 2005 Elsevier GmbH. All rights reserved. Keywords: Proliferation; Apoptosis; Cornea; Photorefractive keratectomy
Introduction The exact regulatory mechanism of the tissue homeostasis of the cornea is still not known. Apoptosis (genetically programmed cell death) of the keratocytes plays an important role in the response to the various environmentally caused insults such as chemical or Corresponding author. Tel.: +36 30 248 3817; fax: +36 1 210 0309.
E-mail addresses: [email protected], [email protected] (N. Szentma´ry). 0344-0338/$ - see front matter r 2005 Elsevier GmbH. All rights reserved. doi:10.1016/j.prp.2005.04.003
mechanical injuries or exposure to ultraviolet light. The regulatory effect of numerous different proteins on the apoptotic process has been demonstrated; interleukin-1 [15,18], FAS ligand protein [6,17], bone morphogenic proteins 2 and 4 [4,7], bcl 2 protein, fibroblast-derived anti-apoptotic survival factor [18], bcl-2 anti-apoptotic protein [4], and type VI collagen [18] all modulate the process. The ability to express these proteins is determined genetically. Although photorefractive keratectomy (PRK) is very effective in treating refractive errors, the postoperative
ARTICLE IN PRESS 400
N. Szentma´ry et al. / Pathology – Research and Practice 201 (2005) 399–404
corneal wound healing processes and their regulation mechanisms are at present not known in detail [8–10,16,18]. According to the results of recent studies, keratocyte apoptosis and subsequent proliferation play a major role in the wound-healing and regeneration process of the cornea following laser treatment [3,4,8]. In this study, we investigated the apoptotic and proliferative activity in rabbit corneas in a series of different intervals up to 16 weeks following PRK.
Methods The investigations were performed using a total of 33 New Zealand colored rabbits; their age at the beginning of the investigation ranged from 3–6 months, and the weight range was 2.5–3.5 kgs. The animals were dealt with in accordance with the principles laid down in the ARVO Resolution on the Use of Animals in Research. Prior to PRK treatment, the rabbits were anesthetized systemically by intramuscular injection of ketamine hydrochloride (30 mg/kg). Local anesthesia using 2–3 drops of oxybuprocaine (Humacain 0.4%, Human Pharmaceuticals, Go¨do¨llo+ , Hungary) was also applied immediately before PRK. The treatment was performed on the right eye of each of the animals using the Asclepion-Meditec MEL70 G-scan laser (AsclepionMeditec GmbH, Jena, Germany); the laser wavelength was 193 nm, with a spot diameter of 1.8 mm, fluence of 250 mJ/cm2, and a pulse rate of 35 Hz. After epithelium removal using a hockey-knife scalpel, a 6 mm diameter ablation zone was treated, with tissue removal to a depth of 82 mm (6.0 diopter correction). Anti-inflammatory (indomethacin/Indosol) and antibiotic (neomycin) drops were instilled at the end of the intervention, and the cornea was allowed to reepithelize spontaneously. During the postoperative period, we closely followed and recorded the status of each cornea, especially focusing on epithelization, as well as on ulcer and haze formation. The eyes that were observed to heal with signs of stromal infiltration or ulcer formation were excluded from further detailed evaluation of postPRK apoptotic and proliferative activity. Animals were sacrificed, and the corneas were examined at eight different intervals after the corresponding PRK intervention, namely at 4 h, 1, 4, 7, 14, 28, 56, and 112 days postoperatively; the numbers of animals in each such group were 4, 4, 4, 5, 4, 4, 4, and 4, respectively. The untreated left eyes served as controls. Eyes that developed infiltration suggestive of infection were excluded from the study, and an additional animal was included in its place. The corneas were fixed in buffered (sodium acetate) formalin (4%; pH ¼ 6.7) and embedded in paraffin wax. We used hamatoxylin–eosin (HE) and periodic acid
Schiff (PAS) staining for general histological examination. Immunocytochemical analysis (immunoperoxidase reaction) of Ki67 (DAKO, Glostrup, Denmark) was used for the evaluation of stromal cell proliferation and the terminal deoxyribonucleotidyl transferase-mediated dUTP-digoxigenin nick-end labeling (TUNEL) assay method (Apoptag, Q-Biogene, Strasbourg, France) for the detection of apoptosis. All antibodies were used in a dilution of 1:100 overnight at 37 1C. The sections were treated with proteinase K; to inactivate endogenous peroxidase, they were incubated in methanol and H2O2. The secondary anti-mouse IgG was used in a dilution of 1:100 the next day. Diaminobenzidine served as chromogen and methyl green as counterstain. Apoptotic cells were recognized on the basis of morphological changes and positive TUNEL reactions. The TUNEL method detects DNA fragmentation along nucleosomes and is considered to stain in particular the nuclei of apoptotic cells. However, there may be some instances where cells exhibiting necrotic morphology may also stain lightly. Therefore, it is important to evaluate TUNEL staining results in conjunction with morphological criteria (nuclear condensation or segmentation, cell shrinkage). The method was previously described and applied to human tumors [2,12] and animal models [13] by Szende et al. For each specimen and staining, the high-power field (hpf) (Olympus BH 10 ocular and 40 objective lens) with a maximum number of Ki67- or TUNEL-positive cells was identified and recorded as positive cell/hpf. The positive cells in the adjacent four contiguous hpf were then counted, added to the first, and recorded as positive cell/5 hpf. Statistical analysis of the number of Ki67-positive and TUNEL-positive cells was not performed because of the low number of corneas and Ki67-positive and apoptotic cells. Epithelial thickness was measured (ocular micrometer) in five nonoverlapping columns at the center of each cornea to achieve a normalized figure for comparison purposes. For statistical analysis, the software package SPSS (version 10.0 for Windows NT) was used. Comparisons between groups or variables were performed using a nonparametric test (Wilcoxon test for paired samples). A p-value less than 0.05 was considered to indicate statistical significance.
Results After PRK, the corneal epithelium regenerated spontaneously. Haze formation was detected in two cases (in samples taken 1 and 2 weeks after the intervention, respectively). Definite postoperative inflammation
ARTICLE IN PRESS N. Szentma´ry et al. / Pathology – Research and Practice 201 (2005) 399–404
keratocytes (two Ki67-positive cells/5 hpf in each positive section) in the upper third of the corneal stroma (Fig. 1). At this time point in one cornea, one TUNELpositive cell/5 hpf could be found; three corneas did not show TUNEL-positive reaction (0 TUNEL positive cells/hpf). Four days after the intervention, the epithelium in the treated area was found to be of nonconstant thickness. Polymorphonuclear cell infiltration of the outer onethird of the stroma was detected. There were 1–2 Ki67positive cells/5 hpf in the upper third of the corneal stroma. In two corneas at the zone with loss of epithelium, 2–3 TUNEL-positive cells/5 hpf were detected in the upper stroma (Fig. 2). One week after PRK, the epithelium showed irregularity of thickness in all the four cases. The stroma was infiltrated by eosinophilic and plasma cells and lymphocytes. In one case, in which haze formation was also observed, thinner collagen fibers with very irregular morphology were found. There were Ki67positive keratocytes in the upper third of the corneal stroma in three of the five corneas (2–3 positive cells/ 5 hpf ). In one cornea, 2 TUNEL-positive cells/5 hpf were detected in the upper one-third of the stroma (Fig. 3).
(infiltration, ulcer) was found to be present in three of the 33 treated corneas; these were excluded from further examination, and additional animals were included in their place. Table 1 shows the mean epithelial thickness at each time point. Four hours and 1 day after PRK, no central epithelium was present. The mean epithelial thickness was decreased significantly 4 and 14 days following PRK, and it was significantly increased 112 days after surgery. In the control (untreated corneas), proliferation or apoptosis was not detectable in any of the samples. The number of Ki67- or TUNEL-positive keratocytes/5 hpf at each time point after PRK-treatment is displayed in Table 2. Four hours after PRK, loss of epithelial cells and eosinophil and lymphocyte infiltration of the outer onethird of the stroma were noted in the PRK-treated area. No Ki67- or TUNEL-positive cells were detected in treated corneas (0 positive cells/5 hpf in each control section). One day after PRK, loss of epithelial cells and eosinophil and lymphocyte infiltration of the outer one-third of the stroma were still noted in the lasertreated area. In two corneas, there were Ki67-positive
Table 1.
401
Central epithelial thickness of the PRK-treated and control corneas at 4 h, 1, 4, 7, 14, 28, 56, and 112 days after surgery Central epithelial thickness (mm)
Time after PRK
4h
1 day
4 days
7 days
14 days
28 days
56 days
112 days
PRK-treated corneas (n ¼ 4)
0.070.0 (0–0)
0.070.0 (0–0)
30.0713.9 (15–56)
31.178.6 (8–45)
35.777.9 (22–48)
32.0711.9 (14–51)
42.7715.6 (23–66)
32.6711.7 (16–53
Control corneas (n ¼ 4)
44.6712.2 (22–56)
45.0712.8 (22–58)
51.072.5 (48–55)
34.172.6 (30–40)
42.579.0 (30–55)
30.274.6 (22–40)
41.579.4 (28–55)
26.472.3 (22–30)
p ¼ 0:00
p ¼ 0:21
p ¼ 0:02
p ¼ 0:61
p ¼ 0:73
p ¼ 0:02
p-values
(Mean7SD, minimum–maximum). A p-value less than 0.05 was considered to indicate statistical significance (bold). When there was no central corneal epithelium of the PRK-treated corneas, we did not make a statistical comparison.
Table 2. Number of Ki67 or TUNEL-positive keratocytes/5 hpf in PRK-treated corneas at different time points after laser treatment (4, 4, 4, 5, 4, 4, 4, 4 corneas in each group) Number of positive cells/5 hpf 4h
1 day
4 days
7 days
14 days
28 days
56 days
112 days
Ki67 TUNEL Ki67 TUNEL Ki67 TUNEL Ki67 TUNEL Ki67 TUNEL Ki67 TUNEL Ki67 TUNEL Ki67 TUNEL 1 2 3 4 5
0 0 0 0 —
0 0 0 0 —
2 2 0 0 —
1 0 0 0 —
2 1 1 2 —
3 2 0 0 —
* indicates cornea regenerated with haze formation.
3* 2 2 0 0
0* 2 0 0 0
2* 2 1 1 —
1* 1 1 2 —
0 0 0 0 —
0 0 0 0 —
0 0 0 0 —
0 0 0 0 —
0 0 0 0 —
0 0 0 0 —
ARTICLE IN PRESS 402
N. Szentma´ry et al. / Pathology – Research and Practice 201 (2005) 399–404
Fig. 1. One day after PRK: Ki67-positive cells (arrows) are detected in the stroma (Ki67 method, original magnification 20). The arrows indicate Ki67-positive cells.
Fig. 3. Seven days after PRK: besides numerous inflammatory cells, some apoptotic cells are detectable in the stroma (arrow). (TUNEL assay, original magnification 40). The arrow indicates TUNEL-positive cell.
cornea, regenerated with haze formation. The TUNEL assay verified the presence of one-two TUNEL-positive cells/5 hpf in the four corneas examined, with no notable difference between the transparent corneas and the one which regenerated with haze formation (Fig. 4). Regarding the corneas examined 4, 8, and 12 weeks after PRK, the stromal fibers and the basal membrane showed irregularity, but Ki67-positive or apoptotic cells were not detected (0 positive cells/5 hpf for each immunohistochemical reaction).
Discussion
Fig. 2. Four days after PRK: PMN cell infiltration and apoptotic cells can be detected in the stroma (TUNEL assay, original magnification 40). The arrows indicate TUNELpositive cells.
Two weeks after the intervention, HE staining revealed a limited level of eosinophilic cell infiltration of the stroma. There were Ki67-positive keratocytes (two positive cells/5 hpf) in the upper stroma of the
Wound healing is critical to the outcomes of all corneal surgical procedures. Apoptosis (programmed cell death) plays an active role in corneal homeostasis, in the corneal response to injury and infection, and in the processes of certain hereditary corneal diseases [15,16]. It is a controlled form of cell death used by an organism to rid itself of unwanted cells, without the release of enzymes and other cellular components that otherwise would damage surrounding tissue and cells. Either epithelial or stromal injury can induce apoptosis in the stroma. During the wound-healing process, cytokine mediators (interleukin-1, bone morphogenic proteins 2 and 4) are released from the damaged tissues and activate the apoptotic program [16]. Interleukin-1 is known to be the master regulator of the corneal woundhealing response [18]. This cytokine also participates in the normal regeneration of epithelial cells. After epithelial injury, interleukin-1 activates apoptosis of keratocytes and corneal fibroblasts via the FAS-ligand system.
ARTICLE IN PRESS N. Szentma´ry et al. / Pathology – Research and Practice 201 (2005) 399–404
Fig. 4. Fourteen days after PRK: under the epithelium active keratocytes, inflammatory cells and apoptotic cells are detectable in the outer one-third of the stroma. (TUNEL assay, original magnification 40). The arrows indicate TUNEL-positive cells.
However, the wound-healing process in the cornea is different to that for tissues containing blood vessels. In the normal healthy cornea, inflammatory cells are found only rarely when using light microscopy after HE staining (perhaps 1–2 of such cells are seen in the typical field of view, at 40 magnification). After injury, the inflammatory cells migrate to the corneal stroma as a result of the chemotactic effect of the elevated levels of cytokines released from the limbal vessels and in the tear film. Immunohistochemical examinations suggest that the majority of these cells belong to the macrophagemonocyte cell type. The increase in their number is detectable 12–24 h after injury [18], and they persist for over 1 week [8]. During healing, the destroyed cells are replaced by the proliferation and migration of the remaining cytokineinduced keratocytes [8,16,18]. Some of the dividing keratocytes and inflammatory cells are caused to transform into myofibroblasts and to participate in the wound-healing response by producing collagen, hyaluronic acid, and growth factors [14,18]. Meanwhile,
403
some of the newly formed keratocytes and inflammatory cells themselves undergo apoptosis and necrosis [18]. Therefore, the quantity and distribution of apoptotic and proliferating keratocytes must be of key significance in the corneal wound healing response. Several studies were conducted recently [3,4,6,8], but the exact regulatory mechanisms of apoptosis and the stromal–epithelial interactions are still topics for further research. The wound healing of the cornea after refractive surgery is similar to that occurring after an abrasion injury. As a consequence of the removal of the epithelium and the photoablation, apoptosis of keratocytes [1] and the presence of inflammatory cells [11] are detected. The polymorphonuclear and apoptotic cells also participated in the wound-healing response in our present study. In agreement with previous studies, inflammatory-cell infiltration of the PRK-treated corneas was detectable until the second week after PRK, when only a limited level of eosinophilic cell infiltration of the stroma was found. Similarly to previous reports [5], the central epithelial thickness was found to be significantly decreased 4 and 14 days after PRK and to be significantly increased 112 days after laser treatment. The thickened epithelium may reduce the refractive correction by compensating the bare stroma ablated by PRK. We compared our findings with the results of the largest and most detailed investigation into stromal cellular response to corneal surgery, conducted by Mohan et al. [8], also investigating rabbits. The study of Mohan et al., using the TUNEL assay method, has shown apoptosis of keratocytes 4, 24, and 72 h after PRK [8]. However, 24 h after surgery, transmission electron microscopy revealed that most of the dying cells had the appearance of necrosis. Helena et al. found that the TUNEL assay was more strongly positive at 4 h than it was earlier [3]. In our present study, 4 h after PRK treatment, genetically programmed cell death could not be detected. The PRK did induce a limited level of apoptosis of the stromal keratocytes, detected by the TUNEL assay 1 day, 4 days, 7 days, and 14 days after the intervention. The study of Mohan et al. [8] detected Ki67-positive cells in the anterior portion of the stromal wound 4 h, 1, 3, and 7 days after photorefractive laser treatment. In our study, proliferation of keratocytes could be detected 1, 4, 7, and 14 days after PRK using the Ki67 method, but not at an earlier or later time. Therefore, the proliferating and apoptotic keratocytes were detected at time points later than in the above study. In addition, the appearance of proliferating keratocytes in one cornea preceded that of apoptotic keratocytes in our study, which is in contradiction to previous findings. The differences in the time points, when apoptotic and proliferating keratocytes were detected, can in part
ARTICLE IN PRESS 404
N. Szentma´ry et al. / Pathology – Research and Practice 201 (2005) 399–404
be explained by the extremely high genetic diversity. The largest and most detailed investigation into stromal cellular response to corneal surgery, conducted by Mohan et al. [8], included New Zealand white rabbits. The use of New Zealand colored rabbits in our study could result in a distinct wound healing response. The different TUNEL technique used in our study, compared to the study of Mohan et al. [8], could also alter the results. Mohan et al. have found that some cells that appeared to be necrotic by transmission electron microscopy may also have shown TUNEL-positive staining using the fluorescence-based TUNEL assay. In summary, PRK was found to trigger proliferation and apoptosis of corneal keratocytes. We conclude that the keratocyte apoptosis and proliferation detected after PRK plays an important role in postoperative woundhealing; but the detailed regulatory mechanisms and their molecular genetic background remain to be clarified.
Acknowledgment This work was supported by funds from the Hungarian Ministry of Health (Grant no. ETT 239/01).
References [1] M. Campos, S. Raman, M. Lee, P.J. McDonnell, Keratocyte loss after different methods of de-epithelialization, Ophthalmology 101 (5) (1994) 890–894. [2] E. Farczadi, I. Kaszas, M. Baki, et al., Changes in apoptosis, mitosis, Her-2, p53 and Bcl2 expression in breast carcinomas after short term tamoxifen treatment, Neoplasma 49 (2002) 101–103. [3] M.C. Helena, F. Baerveldt, W.J. Kim, et al., Keratocyte apoptosis after corneal surgery, Invest. Ophthalmol. Vis. Sci. 39 (1998) 276–283. [4] G. Jianping, T.A. Gelber-Schwab, J.V. Addeo, et al., Apoptosis in the rabbit cornea after photorefractive keratectomy, Cornea 16 (2) (1997) 200–208. [5] T. Miyamoto, S. Saika, A. Yamanaka, et al., Wound healing in rabbit corneas after photorefractive keratectomy and laser in situ keratomileusis, J. Cataract. Refract. Surg. 29 (2003) 153–158.
[6] R. Mohan Rajiv, Q. Liang, W.J. Kim, et al., Apoptosis in the cornea: further characterization of FAS/FAS ligand system, Exp. Eye Res. 65 (1997) 575–589. [7] R. Mohan Rajiv, W.J. Kim, R. Mohan Rahul, et al., Bone morphogenic proteins 2 and 4 and their receptors in the adult human cornea, Invest. Ophthalmol. Vis. Sci. 39 (1998) 2626–2636. [8] R.R. Mohan, A. Hutcheon, R. Choi, et al., Apoptosis, necrosis, proliferation, and myofibroblast generation in the stroma following LASIK and PRK, Exp. Eye Res. 76 (2003) 71–87. [9] Z.Z. Nagy, P. Hiscott, B. Seitz, et al., Morhological response to UV-B irradiation after excimer-laser photorefractive keratectomy, German J. Ophthalmol. 5 (1997) 352–361. [10] Z.Z. Nagy, J. Ne´meth, I. Su¨veges, et al., Die Untersuchung der subepithelialen Narbenbildung nach photorefraktiver Keratectomie mit dem Ultraschallbiomikroskop, Klin. Monatsbl. Augenhlk. 209 (1996) 283–285. [11] T.P. O’Brien, Q. Li, M.F. Ashraf, et al., Inflammatory response in the early stages of wound healing after excimer laser keratectomy, Arch. Ophthalmol. 116 (1998) 1470–1474. [12] B. Szende, I. Romics, K. Minik, et al., Repeated biopsies in evaluation of therapeutic effects in prostate carcinoma, The Prostate 49 (2001) 93–100. [13] E. Toronyi, J. Hamar, K. Magyar, et al., Antiapoptotic effect of ()-deprenyl in rat kidney after ischaemiareperfusion, Med. Sci. Monit. 8 (2002) BR65–BR68. [14] S.E. Wilson, Everett Kinsey lecture. Keratocyte apoptosis in refractive surgery, Contact Lens Assoc. Ophthalmol. J. 24 (1998) 181–1851. [15] S.E. Wilson, Y.Y. He, W. Jian, et al., Epithelial injury induces keratocyte apoptosis: hypothesized role for the interleukin-1 system in the modulation of corneal tissue organization and wound healing, Exp. Eye Res. 62 (1996) 325–337. [16] S.E. Wilson, W.J. Kim, Keratocyte apoptosis: implications on corneal wound healing, tissue organization, and disease, Invest. Ophthalmol. Vis. Sci. 39 (1998) 220–226. [17] S.E. Wilson, Q. Li, J. Weng, et al., The FAS–FAS ligand system and other modulators of apoptosis in the cornea, Invest. Ophthalmol. Vis. Sci. 37 (1996) 1582–1592. [18] S.E. Wilson, R. Mohan Rahul, R. Mohan Rajiv, et al., The corneal wound healing response: cytokine-mediated interaction of the epithelium,stroma and inflammatory cells, Prog. Retin. Eye Res. 20 (2001) 625–637.
Pathology – Research and Practice 201 (2005) 399–404 www.elsevier.de/prp
ANIMAL AND IN VITRO MODELS IN HUMAN DISEASES
Proliferation and apoptosis in the corneal stroma in longterm follow-up after photorefractive keratectomy No´ra Szentma´rya,, Zolta´n Zsolt Nagya, Miklo´s Rescha, Be´la Szendeb,c, Ildiko´ Su¨vegesa a
1st Department of Ophthalmology, Semmelweis University, Budapest 1st Department of Pathology and Experimental Cancer Research, Semmelweis University, Budapest c Research Group of Molecular Pathology, Hungarian Academy of Sciences and Semmelweis University, Budapest b
Received 14 October 2004; accepted 13 April 2005
Abstract This study aimed at investigating the proliferation and apoptosis of corneal cells following photorefractive keratectomy (PRK) treatment. PRK (6.0 D correction) was performed with the Asclepion-Meditec MEL70 G-scan excimer laser on the right eye of each of 33 rabbits under combined local and general anaesthesia. Animals were sacrificed at 4 h, 1, 4, 7, 14, 28, 56, and 112 days postoperatively, and corneal samples from these eight groups were examined histologically. Stromal cell proliferation was evaluated by immunocytochemical analysis of Ki67. Apoptosis was detected using the terminal deoxyribonucleotidyl transferase-mediated dUTP-digoxigenin nick-end labeling (TUNEL) assay method. The untreated left eyes served as controls. Ki67 positivity was detected in the upper stroma on day 1, 4, 7, and 14, and keratocyte apoptosis on day 1, 4, 7, and 14 after PRK, but not at an earlier or later time. Neither Ki67 positivity nor apoptotic activity was observed in the controls (untreated corneas). PRK was found to trigger proliferation and apoptosis of corneal keratocytes. The frequency and spatial distribution of keratocyte proliferation and apoptosis are likely to be important determinants of the corneal wound healing process, but the detailed regulatory mechanisms have not yet been characterized. r 2005 Elsevier GmbH. All rights reserved. Keywords: Proliferation; Apoptosis; Cornea; Photorefractive keratectomy
Introduction The exact regulatory mechanism of the tissue homeostasis of the cornea is still not known. Apoptosis (genetically programmed cell death) of the keratocytes plays an important role in the response to the various environmentally caused insults such as chemical or Corresponding author. Tel.: +36 30 248 3817; fax: +36 1 210 0309.
E-mail addresses: [email protected], [email protected] (N. Szentma´ry). 0344-0338/$ - see front matter r 2005 Elsevier GmbH. All rights reserved. doi:10.1016/j.prp.2005.04.003
mechanical injuries or exposure to ultraviolet light. The regulatory effect of numerous different proteins on the apoptotic process has been demonstrated; interleukin-1 [15,18], FAS ligand protein [6,17], bone morphogenic proteins 2 and 4 [4,7], bcl 2 protein, fibroblast-derived anti-apoptotic survival factor [18], bcl-2 anti-apoptotic protein [4], and type VI collagen [18] all modulate the process. The ability to express these proteins is determined genetically. Although photorefractive keratectomy (PRK) is very effective in treating refractive errors, the postoperative
ARTICLE IN PRESS 400
N. Szentma´ry et al. / Pathology – Research and Practice 201 (2005) 399–404
corneal wound healing processes and their regulation mechanisms are at present not known in detail [8–10,16,18]. According to the results of recent studies, keratocyte apoptosis and subsequent proliferation play a major role in the wound-healing and regeneration process of the cornea following laser treatment [3,4,8]. In this study, we investigated the apoptotic and proliferative activity in rabbit corneas in a series of different intervals up to 16 weeks following PRK.
Methods The investigations were performed using a total of 33 New Zealand colored rabbits; their age at the beginning of the investigation ranged from 3–6 months, and the weight range was 2.5–3.5 kgs. The animals were dealt with in accordance with the principles laid down in the ARVO Resolution on the Use of Animals in Research. Prior to PRK treatment, the rabbits were anesthetized systemically by intramuscular injection of ketamine hydrochloride (30 mg/kg). Local anesthesia using 2–3 drops of oxybuprocaine (Humacain 0.4%, Human Pharmaceuticals, Go¨do¨llo+ , Hungary) was also applied immediately before PRK. The treatment was performed on the right eye of each of the animals using the Asclepion-Meditec MEL70 G-scan laser (AsclepionMeditec GmbH, Jena, Germany); the laser wavelength was 193 nm, with a spot diameter of 1.8 mm, fluence of 250 mJ/cm2, and a pulse rate of 35 Hz. After epithelium removal using a hockey-knife scalpel, a 6 mm diameter ablation zone was treated, with tissue removal to a depth of 82 mm (6.0 diopter correction). Anti-inflammatory (indomethacin/Indosol) and antibiotic (neomycin) drops were instilled at the end of the intervention, and the cornea was allowed to reepithelize spontaneously. During the postoperative period, we closely followed and recorded the status of each cornea, especially focusing on epithelization, as well as on ulcer and haze formation. The eyes that were observed to heal with signs of stromal infiltration or ulcer formation were excluded from further detailed evaluation of postPRK apoptotic and proliferative activity. Animals were sacrificed, and the corneas were examined at eight different intervals after the corresponding PRK intervention, namely at 4 h, 1, 4, 7, 14, 28, 56, and 112 days postoperatively; the numbers of animals in each such group were 4, 4, 4, 5, 4, 4, 4, and 4, respectively. The untreated left eyes served as controls. Eyes that developed infiltration suggestive of infection were excluded from the study, and an additional animal was included in its place. The corneas were fixed in buffered (sodium acetate) formalin (4%; pH ¼ 6.7) and embedded in paraffin wax. We used hamatoxylin–eosin (HE) and periodic acid
Schiff (PAS) staining for general histological examination. Immunocytochemical analysis (immunoperoxidase reaction) of Ki67 (DAKO, Glostrup, Denmark) was used for the evaluation of stromal cell proliferation and the terminal deoxyribonucleotidyl transferase-mediated dUTP-digoxigenin nick-end labeling (TUNEL) assay method (Apoptag, Q-Biogene, Strasbourg, France) for the detection of apoptosis. All antibodies were used in a dilution of 1:100 overnight at 37 1C. The sections were treated with proteinase K; to inactivate endogenous peroxidase, they were incubated in methanol and H2O2. The secondary anti-mouse IgG was used in a dilution of 1:100 the next day. Diaminobenzidine served as chromogen and methyl green as counterstain. Apoptotic cells were recognized on the basis of morphological changes and positive TUNEL reactions. The TUNEL method detects DNA fragmentation along nucleosomes and is considered to stain in particular the nuclei of apoptotic cells. However, there may be some instances where cells exhibiting necrotic morphology may also stain lightly. Therefore, it is important to evaluate TUNEL staining results in conjunction with morphological criteria (nuclear condensation or segmentation, cell shrinkage). The method was previously described and applied to human tumors [2,12] and animal models [13] by Szende et al. For each specimen and staining, the high-power field (hpf) (Olympus BH 10 ocular and 40 objective lens) with a maximum number of Ki67- or TUNEL-positive cells was identified and recorded as positive cell/hpf. The positive cells in the adjacent four contiguous hpf were then counted, added to the first, and recorded as positive cell/5 hpf. Statistical analysis of the number of Ki67-positive and TUNEL-positive cells was not performed because of the low number of corneas and Ki67-positive and apoptotic cells. Epithelial thickness was measured (ocular micrometer) in five nonoverlapping columns at the center of each cornea to achieve a normalized figure for comparison purposes. For statistical analysis, the software package SPSS (version 10.0 for Windows NT) was used. Comparisons between groups or variables were performed using a nonparametric test (Wilcoxon test for paired samples). A p-value less than 0.05 was considered to indicate statistical significance.
Results After PRK, the corneal epithelium regenerated spontaneously. Haze formation was detected in two cases (in samples taken 1 and 2 weeks after the intervention, respectively). Definite postoperative inflammation
ARTICLE IN PRESS N. Szentma´ry et al. / Pathology – Research and Practice 201 (2005) 399–404
keratocytes (two Ki67-positive cells/5 hpf in each positive section) in the upper third of the corneal stroma (Fig. 1). At this time point in one cornea, one TUNELpositive cell/5 hpf could be found; three corneas did not show TUNEL-positive reaction (0 TUNEL positive cells/hpf). Four days after the intervention, the epithelium in the treated area was found to be of nonconstant thickness. Polymorphonuclear cell infiltration of the outer onethird of the stroma was detected. There were 1–2 Ki67positive cells/5 hpf in the upper third of the corneal stroma. In two corneas at the zone with loss of epithelium, 2–3 TUNEL-positive cells/5 hpf were detected in the upper stroma (Fig. 2). One week after PRK, the epithelium showed irregularity of thickness in all the four cases. The stroma was infiltrated by eosinophilic and plasma cells and lymphocytes. In one case, in which haze formation was also observed, thinner collagen fibers with very irregular morphology were found. There were Ki67positive keratocytes in the upper third of the corneal stroma in three of the five corneas (2–3 positive cells/ 5 hpf ). In one cornea, 2 TUNEL-positive cells/5 hpf were detected in the upper one-third of the stroma (Fig. 3).
(infiltration, ulcer) was found to be present in three of the 33 treated corneas; these were excluded from further examination, and additional animals were included in their place. Table 1 shows the mean epithelial thickness at each time point. Four hours and 1 day after PRK, no central epithelium was present. The mean epithelial thickness was decreased significantly 4 and 14 days following PRK, and it was significantly increased 112 days after surgery. In the control (untreated corneas), proliferation or apoptosis was not detectable in any of the samples. The number of Ki67- or TUNEL-positive keratocytes/5 hpf at each time point after PRK-treatment is displayed in Table 2. Four hours after PRK, loss of epithelial cells and eosinophil and lymphocyte infiltration of the outer onethird of the stroma were noted in the PRK-treated area. No Ki67- or TUNEL-positive cells were detected in treated corneas (0 positive cells/5 hpf in each control section). One day after PRK, loss of epithelial cells and eosinophil and lymphocyte infiltration of the outer one-third of the stroma were still noted in the lasertreated area. In two corneas, there were Ki67-positive
Table 1.
401
Central epithelial thickness of the PRK-treated and control corneas at 4 h, 1, 4, 7, 14, 28, 56, and 112 days after surgery Central epithelial thickness (mm)
Time after PRK
4h
1 day
4 days
7 days
14 days
28 days
56 days
112 days
PRK-treated corneas (n ¼ 4)
0.070.0 (0–0)
0.070.0 (0–0)
30.0713.9 (15–56)
31.178.6 (8–45)
35.777.9 (22–48)
32.0711.9 (14–51)
42.7715.6 (23–66)
32.6711.7 (16–53
Control corneas (n ¼ 4)
44.6712.2 (22–56)
45.0712.8 (22–58)
51.072.5 (48–55)
34.172.6 (30–40)
42.579.0 (30–55)
30.274.6 (22–40)
41.579.4 (28–55)
26.472.3 (22–30)
p ¼ 0:00
p ¼ 0:21
p ¼ 0:02
p ¼ 0:61
p ¼ 0:73
p ¼ 0:02
p-values
(Mean7SD, minimum–maximum). A p-value less than 0.05 was considered to indicate statistical significance (bold). When there was no central corneal epithelium of the PRK-treated corneas, we did not make a statistical comparison.
Table 2. Number of Ki67 or TUNEL-positive keratocytes/5 hpf in PRK-treated corneas at different time points after laser treatment (4, 4, 4, 5, 4, 4, 4, 4 corneas in each group) Number of positive cells/5 hpf 4h
1 day
4 days
7 days
14 days
28 days
56 days
112 days
Ki67 TUNEL Ki67 TUNEL Ki67 TUNEL Ki67 TUNEL Ki67 TUNEL Ki67 TUNEL Ki67 TUNEL Ki67 TUNEL 1 2 3 4 5
0 0 0 0 —
0 0 0 0 —
2 2 0 0 —
1 0 0 0 —
2 1 1 2 —
3 2 0 0 —
* indicates cornea regenerated with haze formation.
3* 2 2 0 0
0* 2 0 0 0
2* 2 1 1 —
1* 1 1 2 —
0 0 0 0 —
0 0 0 0 —
0 0 0 0 —
0 0 0 0 —
0 0 0 0 —
0 0 0 0 —
ARTICLE IN PRESS 402
N. Szentma´ry et al. / Pathology – Research and Practice 201 (2005) 399–404
Fig. 1. One day after PRK: Ki67-positive cells (arrows) are detected in the stroma (Ki67 method, original magnification 20). The arrows indicate Ki67-positive cells.
Fig. 3. Seven days after PRK: besides numerous inflammatory cells, some apoptotic cells are detectable in the stroma (arrow). (TUNEL assay, original magnification 40). The arrow indicates TUNEL-positive cell.
cornea, regenerated with haze formation. The TUNEL assay verified the presence of one-two TUNEL-positive cells/5 hpf in the four corneas examined, with no notable difference between the transparent corneas and the one which regenerated with haze formation (Fig. 4). Regarding the corneas examined 4, 8, and 12 weeks after PRK, the stromal fibers and the basal membrane showed irregularity, but Ki67-positive or apoptotic cells were not detected (0 positive cells/5 hpf for each immunohistochemical reaction).
Discussion
Fig. 2. Four days after PRK: PMN cell infiltration and apoptotic cells can be detected in the stroma (TUNEL assay, original magnification 40). The arrows indicate TUNELpositive cells.
Two weeks after the intervention, HE staining revealed a limited level of eosinophilic cell infiltration of the stroma. There were Ki67-positive keratocytes (two positive cells/5 hpf) in the upper stroma of the
Wound healing is critical to the outcomes of all corneal surgical procedures. Apoptosis (programmed cell death) plays an active role in corneal homeostasis, in the corneal response to injury and infection, and in the processes of certain hereditary corneal diseases [15,16]. It is a controlled form of cell death used by an organism to rid itself of unwanted cells, without the release of enzymes and other cellular components that otherwise would damage surrounding tissue and cells. Either epithelial or stromal injury can induce apoptosis in the stroma. During the wound-healing process, cytokine mediators (interleukin-1, bone morphogenic proteins 2 and 4) are released from the damaged tissues and activate the apoptotic program [16]. Interleukin-1 is known to be the master regulator of the corneal woundhealing response [18]. This cytokine also participates in the normal regeneration of epithelial cells. After epithelial injury, interleukin-1 activates apoptosis of keratocytes and corneal fibroblasts via the FAS-ligand system.
ARTICLE IN PRESS N. Szentma´ry et al. / Pathology – Research and Practice 201 (2005) 399–404
Fig. 4. Fourteen days after PRK: under the epithelium active keratocytes, inflammatory cells and apoptotic cells are detectable in the outer one-third of the stroma. (TUNEL assay, original magnification 40). The arrows indicate TUNEL-positive cells.
However, the wound-healing process in the cornea is different to that for tissues containing blood vessels. In the normal healthy cornea, inflammatory cells are found only rarely when using light microscopy after HE staining (perhaps 1–2 of such cells are seen in the typical field of view, at 40 magnification). After injury, the inflammatory cells migrate to the corneal stroma as a result of the chemotactic effect of the elevated levels of cytokines released from the limbal vessels and in the tear film. Immunohistochemical examinations suggest that the majority of these cells belong to the macrophagemonocyte cell type. The increase in their number is detectable 12–24 h after injury [18], and they persist for over 1 week [8]. During healing, the destroyed cells are replaced by the proliferation and migration of the remaining cytokineinduced keratocytes [8,16,18]. Some of the dividing keratocytes and inflammatory cells are caused to transform into myofibroblasts and to participate in the wound-healing response by producing collagen, hyaluronic acid, and growth factors [14,18]. Meanwhile,
403
some of the newly formed keratocytes and inflammatory cells themselves undergo apoptosis and necrosis [18]. Therefore, the quantity and distribution of apoptotic and proliferating keratocytes must be of key significance in the corneal wound healing response. Several studies were conducted recently [3,4,6,8], but the exact regulatory mechanisms of apoptosis and the stromal–epithelial interactions are still topics for further research. The wound healing of the cornea after refractive surgery is similar to that occurring after an abrasion injury. As a consequence of the removal of the epithelium and the photoablation, apoptosis of keratocytes [1] and the presence of inflammatory cells [11] are detected. The polymorphonuclear and apoptotic cells also participated in the wound-healing response in our present study. In agreement with previous studies, inflammatory-cell infiltration of the PRK-treated corneas was detectable until the second week after PRK, when only a limited level of eosinophilic cell infiltration of the stroma was found. Similarly to previous reports [5], the central epithelial thickness was found to be significantly decreased 4 and 14 days after PRK and to be significantly increased 112 days after laser treatment. The thickened epithelium may reduce the refractive correction by compensating the bare stroma ablated by PRK. We compared our findings with the results of the largest and most detailed investigation into stromal cellular response to corneal surgery, conducted by Mohan et al. [8], also investigating rabbits. The study of Mohan et al., using the TUNEL assay method, has shown apoptosis of keratocytes 4, 24, and 72 h after PRK [8]. However, 24 h after surgery, transmission electron microscopy revealed that most of the dying cells had the appearance of necrosis. Helena et al. found that the TUNEL assay was more strongly positive at 4 h than it was earlier [3]. In our present study, 4 h after PRK treatment, genetically programmed cell death could not be detected. The PRK did induce a limited level of apoptosis of the stromal keratocytes, detected by the TUNEL assay 1 day, 4 days, 7 days, and 14 days after the intervention. The study of Mohan et al. [8] detected Ki67-positive cells in the anterior portion of the stromal wound 4 h, 1, 3, and 7 days after photorefractive laser treatment. In our study, proliferation of keratocytes could be detected 1, 4, 7, and 14 days after PRK using the Ki67 method, but not at an earlier or later time. Therefore, the proliferating and apoptotic keratocytes were detected at time points later than in the above study. In addition, the appearance of proliferating keratocytes in one cornea preceded that of apoptotic keratocytes in our study, which is in contradiction to previous findings. The differences in the time points, when apoptotic and proliferating keratocytes were detected, can in part
ARTICLE IN PRESS 404
N. Szentma´ry et al. / Pathology – Research and Practice 201 (2005) 399–404
be explained by the extremely high genetic diversity. The largest and most detailed investigation into stromal cellular response to corneal surgery, conducted by Mohan et al. [8], included New Zealand white rabbits. The use of New Zealand colored rabbits in our study could result in a distinct wound healing response. The different TUNEL technique used in our study, compared to the study of Mohan et al. [8], could also alter the results. Mohan et al. have found that some cells that appeared to be necrotic by transmission electron microscopy may also have shown TUNEL-positive staining using the fluorescence-based TUNEL assay. In summary, PRK was found to trigger proliferation and apoptosis of corneal keratocytes. We conclude that the keratocyte apoptosis and proliferation detected after PRK plays an important role in postoperative woundhealing; but the detailed regulatory mechanisms and their molecular genetic background remain to be clarified.
Acknowledgment This work was supported by funds from the Hungarian Ministry of Health (Grant no. ETT 239/01).
References [1] M. Campos, S. Raman, M. Lee, P.J. McDonnell, Keratocyte loss after different methods of de-epithelialization, Ophthalmology 101 (5) (1994) 890–894. [2] E. Farczadi, I. Kaszas, M. Baki, et al., Changes in apoptosis, mitosis, Her-2, p53 and Bcl2 expression in breast carcinomas after short term tamoxifen treatment, Neoplasma 49 (2002) 101–103. [3] M.C. Helena, F. Baerveldt, W.J. Kim, et al., Keratocyte apoptosis after corneal surgery, Invest. Ophthalmol. Vis. Sci. 39 (1998) 276–283. [4] G. Jianping, T.A. Gelber-Schwab, J.V. Addeo, et al., Apoptosis in the rabbit cornea after photorefractive keratectomy, Cornea 16 (2) (1997) 200–208. [5] T. Miyamoto, S. Saika, A. Yamanaka, et al., Wound healing in rabbit corneas after photorefractive keratectomy and laser in situ keratomileusis, J. Cataract. Refract. Surg. 29 (2003) 153–158.
[6] R. Mohan Rajiv, Q. Liang, W.J. Kim, et al., Apoptosis in the cornea: further characterization of FAS/FAS ligand system, Exp. Eye Res. 65 (1997) 575–589. [7] R. Mohan Rajiv, W.J. Kim, R. Mohan Rahul, et al., Bone morphogenic proteins 2 and 4 and their receptors in the adult human cornea, Invest. Ophthalmol. Vis. Sci. 39 (1998) 2626–2636. [8] R.R. Mohan, A. Hutcheon, R. Choi, et al., Apoptosis, necrosis, proliferation, and myofibroblast generation in the stroma following LASIK and PRK, Exp. Eye Res. 76 (2003) 71–87. [9] Z.Z. Nagy, P. Hiscott, B. Seitz, et al., Morhological response to UV-B irradiation after excimer-laser photorefractive keratectomy, German J. Ophthalmol. 5 (1997) 352–361. [10] Z.Z. Nagy, J. Ne´meth, I. Su¨veges, et al., Die Untersuchung der subepithelialen Narbenbildung nach photorefraktiver Keratectomie mit dem Ultraschallbiomikroskop, Klin. Monatsbl. Augenhlk. 209 (1996) 283–285. [11] T.P. O’Brien, Q. Li, M.F. Ashraf, et al., Inflammatory response in the early stages of wound healing after excimer laser keratectomy, Arch. Ophthalmol. 116 (1998) 1470–1474. [12] B. Szende, I. Romics, K. Minik, et al., Repeated biopsies in evaluation of therapeutic effects in prostate carcinoma, The Prostate 49 (2001) 93–100. [13] E. Toronyi, J. Hamar, K. Magyar, et al., Antiapoptotic effect of ()-deprenyl in rat kidney after ischaemiareperfusion, Med. Sci. Monit. 8 (2002) BR65–BR68. [14] S.E. Wilson, Everett Kinsey lecture. Keratocyte apoptosis in refractive surgery, Contact Lens Assoc. Ophthalmol. J. 24 (1998) 181–1851. [15] S.E. Wilson, Y.Y. He, W. Jian, et al., Epithelial injury induces keratocyte apoptosis: hypothesized role for the interleukin-1 system in the modulation of corneal tissue organization and wound healing, Exp. Eye Res. 62 (1996) 325–337. [16] S.E. Wilson, W.J. Kim, Keratocyte apoptosis: implications on corneal wound healing, tissue organization, and disease, Invest. Ophthalmol. Vis. Sci. 39 (1998) 220–226. [17] S.E. Wilson, Q. Li, J. Weng, et al., The FAS–FAS ligand system and other modulators of apoptosis in the cornea, Invest. Ophthalmol. Vis. Sci. 37 (1996) 1582–1592. [18] S.E. Wilson, R. Mohan Rahul, R. Mohan Rajiv, et al., The corneal wound healing response: cytokine-mediated interaction of the epithelium,stroma and inflammatory cells, Prog. Retin. Eye Res. 20 (2001) 625–637.