|
|
||||||||
1 From the Department of Ophthalmology, Bascom Palmer Eye Institute, and the 2 Department of Cell Biology and Anatomy, University of Miami School of Medicine, Miami, Florida.
| Abstract |
|---|
|
|
|---|
METHODS. Expression of nerve growth factor (NGF); neurotrophins (NT)3 and NT4; brain-derived neurotrophic factor (BDNF); tyrosine kinase-transducing receptors TrkA, TrkB, and TrkC; and a pan-NT low-affinity receptor (p75NTR) was examined by immunostaining in the normal human corneolimbus, HLE grown on intact epithelially denuded AM, and stratified HLE, after subcutaneous implantation in NIH-bg-nu-xid BR mice. NGF protein level was assayed by an ELISA in extracts of intact and epithelially denuded AM. K252a, a specific inhibitor of TrkA autophosphorylation, was added to test whether it would inhibit HLE expansion on AM culture.
RESULTS. Strong positive TrkA staining was confined to the basal epithelial cell layer of normal corneal and limbal epithelia, with the highest intensity noted in the limbus. TrkA staining was also strongly positive in the basal layer of HLE cells cultured on intact and epithelially denuded AM and in basal and some suprabasal layers of stratified HLE transplanted in nude mice. Positive staining of p75NTR was noted in the full-thickness of the corneal epithelium but was limited to the superficial layers of the limbus and in HLE cells cultured on intact and epithelially denuded AM, but was weak in HLE transplanted to nude mice. Weak staining of NT3 and TrkC was noted in the suprabasal layers of corneal and limbal epithelia but was negative in the stratified HLE in nude mice. Negative staining of NGF, NT4, BDNF, and TrkB was noted in all specimens tested. The NGF protein level was readily measured as 35.6 ± 9.1 and 41 ± 12.5 pg/mg protein in the homogenate of the intact and epithelially denuded AM, respectively (P = 0.0256). K252a significantly inhibited the HLE outgrowth on intact AM culture (P = 0.024).
CONCLUSIONS. The strong expression of TrkA but not p75NTR in the limbal basal epithelial cells in vivo suggests that NGF signaling favors limbal epithelial stem cell survival. Such a phenotype is preserved in HLE cells on AM. Blocking NGF signaling significantly retarded HLE expansion on AM, supporting the notion that NGF is important in expansion of limbal epithelial progenitor cells. Furthermore, a high and therapeutic level of NGF was present in AM. Collectively, these findings indicate that denervated neurotrophic ulcers are associated with poor epithelial stem cell function at the limbus. Future studies are needed to determine whether AM transplantation to heal such ulcers may include the promotion of nerve regeneration and survival of epithelial progenitor cells.
| Introduction |
|---|
|
|
|---|
The biological effects of NTs are mediated by a common pan-NT low-affinity receptor (p75NTR), and one of the three high-affinity, tyrosine kinase-transducing receptors (TrkA, TrkB, and TrkC).13
14
15
16
That is, all NTs bind equally to p75NTR; NGF acts through TrkA,5
17
18
which also binds to NT3 and NT4/5 at a lower affinity,17
19
whereas BDNF and NT4 bind TrkB, and NT3 binds TrkC exclusively. It is believed that coexpression of p75NTR and Trk receptors leads to signaling through Trk receptors and promotion of cell survival by stimulation of NTs,20
21
whereas expression of p75NTR alone without Trk receptors promotes apoptosis by increased intracellular ceramide, activation of the nuclear transcription factor
B22, and modulation of growth factor production by target cells.23
24
In epithelial tissues, synthesis of biologically active NGF has already been demonstrated in the human epidermis in vivo25 and in vitro,26 27 and NGF plays an important role in facilitating the degree of re-epithelialization, the thickness of the granulation tissue, and the density of extracellular matrix, when topically applied to the full-thickness skin wound in normal and healing-impaired diabetic mice.28 29 30 It is believed that NGF accelerates the rate of epidermis wound healing through its high-affinity receptor TrkA, and acts synergistically with the other cytokines or growth factors that are released in injured tissues. NGF is also a mitogen for cultured rabbit corneal and limbal epithelial cells.31 When topically applied, NGF facilitates epithelial healing and helps recover the corneal sensitivity in patients with neurotrophic ulcers.32 Neurotrophic ulcers are characterized by persistent corneal epithelial defects and stromal ulceration and result from denervation of corneal sensory nerves.33 34 Because the human corneal basal epithelium expresses TrkA,35 these results collectively indicate that NGF has a direct action on the corneal epithelium. Its role in the stem cell-containing limbal epithelium remains undefined.
The amniotic membrane (AM) is the innermost layer of the fetal membranes and consists of a simple epithelium, a thick basement membrane, and an avascular stroma. When appropriately procured, processed, and preserved, AM has been successfully used as a substrate replacement for ocular surface reconstruction36 37 38 39 and is effective in treating severe and progressive neurotrophic ulcers.40 41 42 43 44 To further elucidate how AM may help epithelialization in neurotrophic ulcers, we investigated the expression of NTs and their cognate receptors, with special attention to NGF and its receptors, TrkA and p75NTR, in the stem cell-containing limbal epithelium in vivo or ex vivo, expanded by either intact or epithelially denuded human AM culture. Furthermore, we analyzed the concentration of NGF in intact and epithelially denuded AM and examined how a specific inhibitor of TrkA autophosphorylation in the NGF signaling pathway may affect the outgrowth of limbal explants on AM cultures.
| Materials and Methods |
|---|
|
|
|---|
Chemical Reagents and Cell Culture Media
Dulbeccos modified eagle medium (DMEM), F-12 nutrient mixture, fetal bovine serum (FBS), HEPES-buffer, amphotericin B, gentamicin, and dispase II, were purchased from Gibco-BRL (Grand Island, NY). All other reagents and chemicals including mouse-derived epidermal growth factor, cholera-toxin (subunit A), dimethyl sulfoxide (DMSO), hydrocortisone, and insulin-transferrin-sodium selenite media supplement were purchased from Sigma Chemical Co. (St. Louis, MO). Plastic cell culture dishes (35 and 60 mm) and 15- and 50-mL sterile centrifuge tubes were obtained from BD Biosciences (Lincoln Park, NJ). Thirty 0.4-µm culture plate inserts (Millicell-PC) and polycarbonate filters were from Millipore (Bedford, MA). Optimal cutting temperature (OCT) compound and cryomolds were from Sakura Finetek (Torrance, CA). Affinity-purified rabbit polyclonal antibody against TrkA full-length form (1:200), TrkB (1:200), and TrkC (1:200) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Affinity-purified rabbit polyclonal antibody against p75NTR (1:100), rat monoclonal antibody against NGF (1:100), and chicken polyclonal antibody against BDNF (1:100), NT3 (1:100), and NT4 (1:100) were all purchased from Promega (Madison, WI). The specificity of each of these antibodies has been verified by the manufacturers and other reports. The secondary biotin-conjugated antibody goat anti-rabbit (1:100) for TrkA, TrkB, TrkC, and p75NTR was from Santa Cruz Biotechnology. The biotin-conjugated mouse anti-rat antibody (1:600) was from Sigma Chemical Co., and the biotin-conjugated rabbit anti-chicken antibody was from Promega. An avidin-biotin complex immunoperoxidase staining kit (Elite ABC) was purchased from Vector Laboratories (Burlingame, CA), a bicinchoninic acid (BCA) protein assay kit from Pierce (Rockford, IL), and an NGF immunoassay system (Emax) from Promega. The tyrosine kinase inhibitor K252a was purchased from Calbiochem-Novabiochem (San Diego, CA).
Preparation of Human Amniotic Membrane
Human AM was kindly provided by Bio-Tissue (Miami, FL) and stored at -80°C. AM was devitalized by freezing and thawing and washed three times with Hanks balanced saline solution (HBSS) before being fastened onto a 30-mm culture insert (Millicel-CM; Millipore), to be placed in a six-well plate, as previously described.45
Fifty percent of the membranes used for limbal cultures were treated with 0.1% sterile EDTA solution for 30 minutes and then gently scrubbed with an epithelial scrubber (Amoils Epithelial Scrubber; Innova, Inc., Toronto, Ontario, Canada) to remove the amniotic epithelium without breaking the underlying basement membrane, as previously described.46
With this method, 90% to 100% of the epithelium could be removed.
Preparation of Human Corneolimbal Tissue
Human tissue was handled according to the tenets of the Declaration of Helsinki. Human corneas obtained from the Florida Lions Eye Bank (Miami, FL) were of transplant quality but had been excluded from clinical use for nonocular reasons. Each cornea was cut into two halves through the 12 oclock to 6 oclock meridian. One half was embedded in OCT and snap frozen in liquid nitrogen. The other half was fixed in 4% paraformaldehyde in 0.1 M sodium phosphate buffer (pH 7.4) for 4 hours and embedded in paraffin. The limbal area was defined by the end of the Bowman layer, where vascularized soft connective tissue was noted. The peripheral cornea was defined as the region located 2 mm from this limbal landmark toward the cornea, with the underlying stroma being compact and avascular. The central cornea was defined as 4 mm from the limbal landmark toward the center of the button.
Human Limbal Explant Culture on Amniotic Membrane
The corneolimbal tissue was rinsed three times with DMEM containing 50 µg/mL gentamicin and 1.25 µg/mL amphotericin B. After removal of excessive sclera, iris, corneal endothelium, conjunctiva, and Tenon capsule, the remaining tissue was placed in a culture dish and exposed for 5 to 10 minutes to 1.2 U/mL dispase II in Mg2+-and Ca2+-free solution (DMEM) at 37°C under 95% humidity and 5% CO2. After one rinse with DMEM containing 10% FBS, the limbus was separated from the rest of the tissue with a trephine and was cut into cubes of approximately 1 x 1.5 x 2.5 mm with a scalpel. On the center of either intact or EDTA-treated AM, an explant was placed and cultured in SHEM medium made of an equal volume of HEPES-buffered DMEM containing bicarbonate and Hams F12 supplemented with 10% FBS, 0.5% DMSO, 50 µg/mL gentamicin, 1.25 µg/mL amphotericin B, 2 ng/mL mouse epidermal growth factor (EGF), 5 µg/mL insulin, 5 µg/mL transferrin, 5 ng/mL selenium, 0.5 mg/mL hydrocortisone, and 30 ng/mL cholera toxin. Human limbal epithelial (HLE) cells thus obtained were maintained at 37°C under 95% humidity and 5% CO2, the medium was changed three times weekly, and cell outgrowth was monitored daily for 3 weeks with an inverted phase microscope (Nikon, Tokyo, Japan).
Subcutaneous Transplantation to NIH-bg-nu-xid BR Mice
At confluence, HLE cells on intact AM were transplanted to the subcutaneous plane of NIH-bg-nu-xid BR mice after the skin covering the rectus abdominis was undermined to expose an area measuring approximately 1.5 x 1.5 cm. The skin flap was then closed with a running coated 7-0 suture (Vicryl, Ethicon, Somerville, NJ). A firm subcutaneous nodule that formed during a 5-day period was excised together with the surrounding skin and the muscle, embedded in OCT, snap frozen in liquid nitrogen, and processed for histology and immunostaining.
Immunohistochemical Staining
Immunoreactivity of NGF, BDNF, NT3, NT4, TrkA, TrkB, TrkC, and p75NTR in normal corneolimbal epithelia, in the confluent HLE cells on either the intact or epithelially denuded AM, and in the resultant specimen from NIH-bg-nu-xid BR mice was studied in acetone-fixed cryostat sections (6 µm). Briefly, after endogenous peroxidase and nonspecific avidin-biotin binding were blocked, sections were incubated at room temperature using primary antisera against the corresponding antigens for 2 hours washed in PBS (three times, each for 5 minutes) followed by incubation with the appropriate biotinylated secondary antibody (1:100) for 30 minutes and washing in PBS (three times, each for 5 minutes). Immunoreactive proteins were detected using an avidin-biotin-horseradish peroxidase reagent for 30 minutes and 3-amino-9-ethylcarbazole (AEC) as a chromogen. Finally, sections were counterstained with hematoxylin and photographed. For all antisera, incubation without primary antisera was used as a negative control, and the optic nerve paraffin-embedded sections were used as a positive control (data not shown). The sensitivity and specificity of each primary antibody were confirmed.
Amniotic Membrane Extracts for Total Protein Assay and NGF ELISA
Fresh intact or epithelially denuded AM was weighed and frozen at -80°C and homogenized at a 1:30 (wt/vol) dilution in a high salt extraction buffer comprising 100 mM Tris-HCl (pH 7.6), 1 M NaCl, 2% BSA, 4 mM EDTA, and 0.5% Triton X-100 and supplemented with protease inhibitors, as previously described.47
After centrifugation at 6.500g for 30 minutes at room temperature, the supernatants were stored at -80°C until assayed. Total protein assay was performed using the BCA protein assay kit, according to the Bradford method provided by the manufacturer. The NGF content was measured using the NGF immunoassay system, according to the manufacturers instructions. This assay is a two-site sandwich ELISA that offers optimal sensitivity as low as 15.6 pg/mL.
Inhibition of NGF Signaling by K252a
As previously demonstrated,48
49
the alkaloid K252a, a tyrosine kinase inhibitor isolated from the culture broth of Nocardiopsis sp, totally inhibits NGF-induced differentiation of PC12 pheochromocytoma cells at nanomolar concentrations (100% inhibition at 200 nM). We sought to explore whether NGF signaling was involved in HLE cell expansion on intact AM by adding 200 nM K252a to the culture medium at day 1 and every other day thereafter for 3 weeks. Untreated cultures were added with an equal amount of DMSO as the control. Their outgrowth was monitored daily and terminated at the end of 3 weeks by fixing in methanol and staining with hematoxylin.
Statistical Analysis
Data were subject to nonparametric Mann-Whitney test and expressed as the mean ± SEM. P < 0.05 was considered significant.
| Results |
|---|
|
|
|---|
|
|
|
K252a Inhibition of HLE Cell Outgrowth on AM
The presence of TrkA and p75NTR receptors in HLE cells suggests that NGF signaling may effect the outgrowth of HLE cells on AM. To test this hypothesis, we added the TrkA-specific tyrosine kinase inhibitor K252a in HLE cells cultures on intact AM. Compared with the untreated control, addition of K252a significantly reduced the outgrowth rate measured daily for a period of 3 weeks. At the end of 3 weeks, of six control culture wells, there was marked outgrowth to confluence in three, moderate growth to partial confluence in two, and no growth in one, because of a necrotic limbal explant (Fig. 4
, upper panel, the control plate). In contrast, five of six wells treated with K252a did not show any growth, whereas the remaining one showed only a limited growth (Fig. 4
, lower panel, K252a-treated plate). The number of wells in each six-well plate showing positive HLE cell outgrowth at days 7, 14, and 21 was as follows: K252a-treated plate: 0, 1, and 1, respectively: control plate: 5 wells at each time point. The differences between the average sizes of the control and the K252a-treated outgrowths were analyzed by the nonparametric Mann-Whitney test and were significant at each time point (P = 0.015 at day 7, 0.026 at day 14, and 0.026 at day 21). The mean ± SEM outgrowth size is presented in Figure 5
. This result strongly supports that the TrkA signaling pathway mediated by NGF plays an important role in the expansion of limbal epithelial cells on AM.
|
|
| Discussion |
|---|
|
|
|---|
Lambiase et al.35 have shown that the human corneal basal epithelium expresses TrkA. Herein, we noted that human limbal basal and some suprabasal epithelial layers also expressed disproportionally more TrkA than p75NTR, suggesting that survival, but not apoptosis, in limbal epithelial progenitor cells may be maintained preferentially by NGF signaling. In contrast, both TrkA and p75NTR were similarly expressed by corneal basal cells, suggesting a significant role of NGF, as a cofactor, in regulating corneal transient amplifying cell survival and apoptosis. The finding that p75NTR, but not TrkA, was expressed by limbal and corneal suprabasal and superficial epithelial cells, suggests that apoptosis dominates in these postmitotic and terminally differentiated epithelial cells. Because NT3 and TrkC were not expressed by the basal cell layer but instead were weakly expressed by suprabasal cell layers of both corneal and limbal epithelia, we believe signaling through NGF, but not NT3, is involved in the control of limbal epithelial stem cells. Knowing that the limbal epithelial progenitor cells are predominantly equipped with TrkA, but not p75NTR, for NGF signaling, the first natural question that arises is where NGF is coming from. Besides the natural source of sensory neurons, You et al.61 have detected the NGF transcript in cultured corneal epithelium and stroma. Our immunohistochemical technique may not be sensitive enough to detect NGF in the corneal or limbal epithelia. If NGF is indeed produced by these epithelial cells in vivo, it will function as an autocrine factor.
Without knowing how NGF signaling may be involved in regulating limbal epithelial progenitor cells in vivo, we turned to an in vitro model of AM culture, a system that has been used to expand limbal epithelial stem cells for transplantation in patients with limbal stem cell deficiency.62 63 64 Our results showed that the immunoreactivity of TrkA and p75NTR was indeed found in ex vivo expanded HLE cells on both intact and epithelially denuded AM cultures, as was the expression of NT3 and its receptor TrkC. This result indicates that the NGF signaling system was preserved in HLE cells by AM cultures. Furthermore, using an ELISA we detected a high amount of NGF (35.6 ± 9.1 and 41 ± 12.5 pg/mg protein) in the extract obtained from both intact and epithelially denuded AM, respectively. We attribute these relatively high levels of NGF, in part, to the use of a high-salt extraction buffer, which according to Zettler et al.47 increases the detectable levels of NGF protein up to 10 times more than those detected by previous procedures. In addition, freezing and thawing of our samples three times before ELISA may have promoted the release of NGF from receptors, as has been reported.65 NGF has been found in a number of tissues in the body, with the highest concentration noted in the submandibular salivary gland of the adult male mouse, the venoms of poisonous snakes, human placenta, guinea-pig prostate, and bovine seminal plasma and vesicles.66 One report pointed out that AM of the placenta is one of the human tissues possessing a high specific NGF bioactivity.67 Because the difference in these two levels was statistically significant (P = 0.026), we believe that the most NGF in AM resides in the stroma. Therefore, the reason NGF was not detected by immunohistochemistry but was detected at high levels by ELISA may be that NGF was sequestered in the matrix of the AM, which masked the antigenicity of NGF. Although we cannot rule out the possibility that NGF may be produced elsewhere and then sequestered in AM, there is evidence to show that NGF, BDNF, and NT-3 are synthesized by human amniotic epithelial cells.68 It remains unclear whether NGF was actually released during culturing of HLE cells. Further experiments to resolve these questions are important in delineating the role of NGF in ex vivo expansion of HLE cells.
To determine whether NGF signaling is involved in ex vivo expansion of limbal epithelial progenitor cells on AM, we added K252a, which has been reported as a specific inhibitor of NGF-mediated signal transduction in PC12 pheochromocytoma cells,49 69 a rat myogenic cell line L6,70 a catecholaminergic central nervous system (CNS) cell line,71 a Y-79 retinoblastoma-derived cell line,72 and a prostate adenocarcinoma cell line.73 74 The inhibitory action has been determined at the level of tyrosine phosphorylation and kinase activity of TrkA.49 At 200 nM, K252a is a selective inhibitor for all three forms of Trk and not serine-threonine kinases,75 does not affect signal transduction of bFGF69 and EGF,49 69 and does not penetrate cell membranes or affect thymidine incorporation directly.73 The dramatic antiproliferative effect of K252a on HLE cell outgrowth on AM cultures (Figs. 4 5) supports the notion that NGF signal transduction through TrkA is essential to promote HLE progenitor cell proliferation in this AM culture system.
After xenotransplantation to nude mice, the basal cell layer of the resultant stratified epithelium expressed strong TrkA but weak p75NTR, a phenotype resembling that of the normal limbal epithelium in vivo. Therefore, it can be envisioned that AM may provide NGF in a manner similar to topical application of NGF32 for the treatment of neurotrophic corneas. In addition, AM may support HLE cells expression of an NGF signaling system equipped with disproportionally more TrkA than p75NTR, favoring survival over apoptosis in HLE progenitor cells. Collectively, these data help explain why poor epithelial healing ensues in neurotrophic ulcers when the trigeminal sensory nerve is damaged and why AM transplantation is effective in healing these ulcers in humans.40 41 42 43 44 Furthermore, these data also support the hypothesis that the AM culture system preferentially preserves and expands limbal epithelial progenitor cells including stem cells and explains why it is useful for treating patients with limbal stem cell deficiency.62 63 64
| Footnotes |
|---|
Submitted for publication August 13, 2001; revised November 28, 2001; accepted December 21, 2001.
Commercial relationships policy: P (SCGT); N (AT, MG).
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked "advertisement" in accordance with 18 U.S.C.
1734 solely to indicate this fact.
Corresponding author: Scheffer C. G. Tseng, Ocular Surface Center and Ocular Surface Research and Education Foundation, 8780 SW 92nd Street, Suite 203, Miami, FL 33176; stseng{at}ocularsurface.com
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T. Nakamura, K.-i. Endo, and S. Kinoshita Identification of Human Oral Keratinocyte Stem/Progenitor Cells by Neurotrophin Receptor p75 and the Role of Neurotrophin/p75 Signaling Stem Cells, March 1, 2007; 25(3): 628 - 638. [Abstract] [Full Text] [PDF] |
||||
![]() |
C.-C. Sun, J.-H. Su Pang, C.-Y. Cheng, H.-F. Cheng, Y.-S. Lee, W.-C. Ku, C.-H. Hsiao, J.-K. Chen, and C.-M. Yang Interleukin-1 Receptor Antagonist (IL-1RA) Prevents Apoptosis in Ex Vivo Expansion of Human Limbal Epithelial Cells Cultivated on Human Amniotic Membrane Stem Cells, September 1, 2006; 24(9): 2130 - 2139. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. H. Lee, I. H. Ryu, E. K. Kim, J. E. Lee, S. Hong, and H. K. Lee Induced Expression of Insulin-like Growth Factor-1 by Amniotic Membrane-Conditioned Medium in Cultured Human Corneal Epithelial Cells. Invest. Ophthalmol. Vis. Sci., March 1, 2006; 47(3): 864 - 872. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. He, H.-T. Cho, W. Li, T. Kawakita, L. Jong, and S. C. G. Tseng Signaling-Transduction Pathways Required for Ex Vivo Expansion of Human Limbal Explants on Intact Amniotic Membrane Invest. Ophthalmol. Vis. Sci., January 1, 2006; 47(1): 151 - 157. [Abstract] [Full Text] [PDF] |
||||
![]() |
M Rihl, E Kruithof, C Barthel, F De Keyser, E M Veys, H Zeidler, D T Y Yu, J G Kuipers, and D Baeten Involvement of neurotrophins and their receptors in spondyloarthritis synovitis: relation to inflammation and response to treatment Ann Rheum Dis, November 1, 2005; 64(11): 1542 - 1549. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |