IOVS Blood
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

(Investigative Ophthalmology and Visual Science. 2008;49:2163-2171.)
© 2008 by The Association for Research in Vision and Ophthalmology, Inc.
DOI:  10.1167/iovs.07-1206
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Cai, J.
Right arrow Articles by Boulton, M. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Cai, J.
Right arrow Articles by Boulton, M. E.

The Angiopoietin/Tie-2 System Regulates Pericyte Survival and Recruitment in Diabetic Retinopathy

Jun Cai,1 Oksana Kehoe,2 Gill M. Smith,2 Philip Hykin,3 and Michael E. Boulton1

1From the Department of Ophthalmology and Visual Sciences, The University of Texas Medical Branch, Galveston, Texas; the 2School of Optometry and Vision Sciences, Cardiff University, Cardiff, Wales, United Kingdom; and 3Moorfields Eye Hospital, London, United Kingdom.


   Abstract
Top
 Abstract
Methods
 Results
 Discussion
 References
 
PURPOSE. The angiopoietin (Ang) system plays an important role in vascular stabilization and pathologic neovascularization. The hypothesis for the study was that, in addition to modulating endothelial cell behavior, the angiopoietin/Tie-2 system also regulates the pericyte apoptosis and/or the vessel maturation associated with diabetic retinopathy.

METHODS. Tie-2 expression in cultured retinal pericytes was analyzed by using ELISA, Western Blot analysis, and flow cytometry. CD13 (aminopeptidase N) expression in pericytes was determined by Western blot analysis and Ang effects verified with Tie-2 antisense treatment. Cell proliferation was monitored by crystal violet uptake, and pericyte migration was assessed in a scrape wound. Annexin V-FITC flow cytometry was used to quantify pericyte apoptosis.

RESULTS. Pericytes expressed a functionally active Tie-2 receptor upregulated by both Ang-1 and -2 (P < 0.05). In pericytes undergoing apoptosis induced by either TNF-{alpha} or high glucose Ang-1 increased survival (P < 0.05 for TNF-{alpha}; P < 0.01 for high glucose), whereas Ang-2 increased apoptosis. Ang-1 enhanced CD13 expression in a dose-dependent manner (P < 0.05) and stimulated pericyte migration in a synthetic matrix wound-healing assay that was associated with a change in F-actin organization. Addition of Tie-2 antisense confirmed that angiopoietins act through Tie-2.

CONCLUSIONS. These findings demonstrate that Tie-2 is functional in pericytes and may play an important role in the progression of diabetic retinopathy, by regulating pericyte loss and influencing the activation state and recruitment of pericytes.

A hallmark of diabetic retinopathy and a precursor of endothelial cell dysfunction is the loss of pericytes, so-called "pericyte drop-out," leading to acellular capillaries, which are no longer perfused.1 2 3 Although the loss of pericytes is believed to be linked to fluctuations in blood glucose, the mechanism is not well understood. In addition, the maturation of newly formed vessels during the late stages of proliferative diabetic retinopathy requires the incorporation of pericytes.4

The angiopoietins, a recently discovered family of vascular regulatory molecules binding to the Tie-2 tyrosine receptor, play an important role in retinal neovascularization.5 6 The TIE-2 gene encodes a protein of 1122 amino acids with its extracellular region containing three EGF-like repeats and three repeats with fibronectin type III homology located after the second Ig loop. The intracellular portion of Tie-2 contains two tyrosine kinase domains. Tie-2-deficient mice die at around embryonic day 10.5 and exhibit profound vascular defects indicating the importance of Tie-2 in vascular development.7 8 The activity of Tie-2 is differentially regulated by angiopoietin (Ang)-1 and -2, which share approximately 60% amino acid identity with similar affinity to Tie-2.9 10 Although primarily thought of as an endothelial cell protein, Tie-2 is also expressed by some hematopoietic stem cells,11 and its RNA has been reported in pericytes.12

Ang-1 binds to Tie-2 and induces phosphorylation of Tie-2. Ang-1/Tie-2 are proposed to mediate the mobilization of hematopoietic stem cells to the peripheral circulation13 and the formation of mature capillary networks by recruiting periendothelial cells such as pericytes.14 Both Ang-1 and Tie-2 are upregulated during the early stages of wound healing coinciding with wound angiogenesis15 and in various tumors and tumor cell lines.16 Furthermore, the constitutive secretion of Ang-1 by normal quiescent vessels is considered to stabilize vessels by maintaining contacts between endothelial cells and periendothelial cells.

Ang-2 has been suggested to block the constitutive stabilization or maturation function of Ang-1 by promoting smooth muscle cell–pericyte dropout therefore loosening contacts between endothelial cells and periendothelial cells.17 Ang-2 has been reported to be upregulated in mouse models of ischemia-induced retinal neovascularization, as well as during angiogenesis in retinal development.18

We investigated how the angiopoietin signaling system may regulate the course of diabetic retinopathy by modulating retinal pericyte survival and recruitment.


   Methods
Top
 Abstract
 Methods
Results
 Discussion
 References
 
Cell Culture
Retinal pericytes were isolated from bovine eyes and cultured by using a modification of previous methods.19 Isolated pericytes were resuspended in MEM with 20% fetal bovine serum at 37°C for 3 days. After cell attachment, the medium was changed every 3 to 4 days with MEM containing 10% fetal bovine serum.

Purification of bovine retinal pericytes was achieved with a kit (Cellection Pan Mouse IgG kit; Dynal Biotech, Bromborough, UK) according to the manufacturer’s instructions. Briefly, the cells collected from one primary culture T-25-cm2 flask by trypsinization were reacted with mouse anti-desmin monoclonal antibody (Chemicon Europe, Chandlers Ford, UK) for 10 minutes at 4°C. After washing, the cells were mixed with pan mouse IgG beads (Cellection Dynabeads; Dynal Biotech) for 20 minutes at 4°C with constant shaking. The samples were placed in a magnetic particle concentrator (MPC-S; Dynal Biotech) to separate immunoadsorbed cells for 2 minutes at room temperature. The bead-bound cells were then resuspended in release buffer for 15 minutes. After vigorous pipetting, the samples were place in the particle concentrator for 2 minutes, and the purified pericytes were collected and seeded into T25-cm2 flasks under the culture conditions just described. Cultures were subcultured at a ratio of 1:2 on reaching confluence, and the cells were used within three passages. For all experimental studies the fetal bovine serum concentration of the medium was reduced to 5%.

To rule out endothelial cell contamination, equal amounts of lysates from pericyte cultures and microvascular endothelial cells (as a positive control) were subjected to Western blot analysis with anti-PECAM-1 antibody (Santa Cruz Biotechnology Inc., Santa Cruz, CA).

To confirm that pericytes are able to retain their phenotype during progressive passage, the cells were propagated for 5 days in the second passage and then incubated with mouse anti-desmin monoclonal antibody (Chemicon Europe) at room temperature for 30 minutes (mouse IgG1 antibody acted as a control). After incubation with an FITC-conjugated secondary antibody the number of desmin-positive cells was determined by flow cytometry.

To study the pericyte response to exogenous Ang-1 or -2 cells were, unless otherwise stated, treated with Ang-1 or -2 (Regeneron Pharmaceuticals, Inc., Tarrytown, NY) at a concentration of 100 ng/mL, based on preliminary experiments and the functional response achieved in Figure 4 . For the unstimulated control, Ang was substituted with 0.1% BSA.


Figure 4
View larger version (25K):
[in this window]
[in a new window]

  		FIGURE 4. Angiopoietins mediate retinal pericyte proliferation and migration. (A) Ang-2 modestly increased proliferation of retinal pericytes in a dose-dependent manner, whereas Ang-1 had no effect. The number of cells is expressed as the mean ± SEM of counts in triplicate wells. *P < 0.05 versus the unstimulated control group in which Ang was substituted with 0.1% BSA; (B) Ang-1 had no effect on the migration of retinal pericytes cultured on plastic. In contrast, Ang-1 treatment significantly enhanced retinal pericyte migration on a synthetic-matrix–coated surface. Tie-2 antisense treatment abolished the enhancing effect of Ang-1 on retinal pericyte migration on the matrix. However, there was no significant increase in retinal pericyte migration after treatment with Ang-2 in the presence or absence of matrix. The data are represented as the mean ± SEM (n = 3).

 
ELISA Determination of Tie-2
After treatment with 100 ng/mL Ang-1 or -2 at 37°C for 48 hours, the medium was collected, and the pericytes were lysed by adding 1 mL per flask of RIPA buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1% NP-40, 0.25% sodium deoxycholate, 1 mM NaF, 1 mM Na3VO4 and 1 mM EDTA containing the protease inhibitors phenylmethylsulfonyl fluoride, aprotinin, leupeptin, and pepstatin) at 4°C for 30 minutes. The amount of protein was determined by the BCA protein assay before detection of Tie-2 by ELISA (Quantikine; R&D Systems, Europe, Abingdon, UK) according to the manufacturer’s instructions and absorbance was read at 450 nm. The detection limit of this assay was 0.02 ng/mL. Tie-2 levels were normalized for total protein concentrations in the cell lysates and expressed as ng Tie-2 per milligram of total protein.

Tie-2 Phosphorylation
To assess the phosphorylation status of Tie-2 after angiopoietin treatment for 48 hours, phosphorylated Tie-2 was immunoprecipitated from the cell lysate containing protein at 500 µg/mL by incubation with 10 µL mouse anti-tyrosine phosphorylation monoclonal antibody (PY 20; Santa Cruz Biotechnology, Inc.) for 1.5 hours at 4°C, followed by addition of 20 µL protein A/G agarose (Santa Cruz Biotechnology, Inc.) overnight at 4°C. After a wash in RIPA buffer, the mixture was centrifuged at 12,000g for 20 minutes. The pellet was resuspended with 1 mL of RIPA buffer containing protease inhibitors at 65°C for 5 minutes, to disrupt protein–protein interactions, and was then subjected to ELISA for Tie-2.

Flow Cytometric Analysis of Tie-2
Subconfluent pericytes treated with 100 ng/mL Ang-1 or -2 for 48 hours were harvested by centrifugation at 200g for 5 minutes. After three washes in PBS containing 0.5% FCS, the cells were resuspended in the same buffer to a final concentration of 4 x 107/mL and 25 µL of the cells (1 x 106) were added to a tube. A 10-µL aliquot of the PE-conjugated anti-Tie-2 antibody (R&D Systems Europe) was added to the cells at 4°C for 45 minutes, followed by two washes in the same buffer before flow cytometric analysis (FACSCalibur; BD Biosciences, Oxford, UK). As a control, the cells in a separate tube were treated with PE-labeled mouse IgG1 antibody.

Antisense Treatment
Antisense 5'-GCTAAAGAATCCATGCTTCCCC-3' and scrambled oligos 5'-GGGGGAAGCATGGATTCTTTAGC-3' that corresponded to base pairs 318 to 337 of bovine Tie-2 mRNA were constructed by MWG Biotech (Milton Keynes, UK). In oligo transfections, 6 µM of oligo, predetermined as the optimal concentration, in MEM was incubated with DMRIE-C for 45 minutes at 25°C. After the pericytes were washed, liposome-oligo complexes were incubated with the cells for 5 hours at 37°C followed by addition of MEM containing 10% FCS for 19 hours. Based on the titration result, pericytes (Fig. 2C) were treated with 6 µM of the Tie-2 antisense and scrambled oligo daily over a 3-day period.


Figure 2
View larger version (36K):
[in this window]
[in a new window]

  		FIGURE 2. Ang-1 upregulated CD13 (aminopeptidase N) expression in retinal pericytes. Western blots are from a representative experiment. The densitometric analysis incorporates the mean results of at least three separate experiments. (A) Whole-cell lysates from the retinal pericytes treated with 100 ng/mL Ang-1 for 48 hours were probed with an antibody for CD13. Ang-1 treatment led to a dose-dependent increase in CD13 (150 kDa) expression in retinal pericytes. Densitometric analyses are presented as the relative ratio of CD13 to {alpha}-tubulin (55 kDa) and the ratio relative to control is arbitrarily presented as 1. Data are expressed as the mean ± SEM (n = 4). *P < 0.05, **P < 0.01 versus the control. (B) The pericytes exposed to Tie-2 antisense eliminated the Ang-1-induced expression of CD13, whereas Tie-2 sense had no effect. Densitometric analyses are presented as the relative ratio of CD13 to {alpha}-tubulin, and the ratio relative to control is arbitrarily presented as 1. Data are expressed as the mean ± SEM (n = 4).**P < 0.001 versus Ang-1 only. (C) Oligo at 6 µM was the optimal concentration for antisense Tie-2, as determined by Western blot analysis of Tie-2 (140 kDa) expression by pericytes after transfection with different concentrations of the antisense Tie-2 oligos. Densitometric analyses are presented as the relative ratio of Tie-2 to {alpha}-tubulin, and the ratio relative to control is arbitrarily presented as 1. Data are expressed as the mean ± SEM (n = 4). **P < 0.01 versus 0 µM Tie-2 antisense.

 
Western Blot Analysis for CD13
CD13 (aminopeptidase N) is a marker of hematopoietic cells20 and cell mobility and is a putative indicator of pericyte maturation in the cerebral vascular bed.21 It is a type II membrane-bound metalloproteinase expressed on various cell types and a Zn2+-dependent ectopeptidase that degrades preferentially proteins with an NH2-terminal neutral amino acids.22 23 CD13 expression in pericytes was analyzed by Western blot analysis. Subconfluent pericytes were cultured in the presence or the absence of 100 ng/mL Ang-1 or -2 at 37°C for 48 hours. The proteins were extracted with RIPA buffer at 4°C, and the protein content was determined by the BCA protein assay. Equal amounts of protein from each sample were resolved by SDS polyacrylamide gel and transferred onto nitrocellulose membrane. The membranes were incubated with an antibody for CD13 (Santa Cruz Biotechnology, Inc.) at room temperature for 2 hours. The membranes were then washed with 3% milk/TBS containing 0.05% Tween-20 followed by HRP-conjugated secondary antibody (Santa Cruz Biotechnology, Inc.) at room temperature for 1 hour. After they were washed, the membranes were incubated with enhanced chemiluminescent substrate (ECL; Santa Cruz Biotechnology Inc.) and exposed to film (Biomax MR; Sigma-Aldrich, Poole, Dorset, UK). The blots were stripped and reprobed with goat polyclonal anti-{alpha} tubulin antibody (1:250; Santa Cruz Biotechnology, Inc.).

Detection of Apoptosis
Apoptosis was evaluated by using a FITC-conjugated annexin V/propidium iodide (PI) assay kit (R&D Systems Europe), based on annexin-V binding to phosphatidylserine exposed on the outer leaflet of the plasma membrane lipid bilayer of cells entering the apoptotic pathway. Apoptosis was induced using two approaches, TNF-{alpha} and high glucose. Pericytes were exposed to (1) 100 ng/mL Ang-1 or -2 in the presence or absence of 100 ng/mL TNF-{alpha} at 37°C for 48 hours or (2) high glucose, 15 and 25 mM, or control medium containing 5.6 mM glucose. Glucose levels were replenished daily. Ang-1 or -2 (100 ng/mL) was added to the culture medium 48 hours before cell harvest.

For apoptosis detection, the cells were collected by centrifugation (200g for 5 minutes), washed in PBS, and resuspended in the annexin V incubation reagent in the dark for 15 minutes before flow cytometric analysis.

Proliferation Assay
Cells were plated in 96-well plates at 750 cells per well in MEM with 10% FCS for 24 hours. They were transferred to serum-free MEM for 45 minutes and then treated with various concentrations of Ang-1 or -2 in MEM containing 1% FCS at 37°C for 48 hours. The relative cell number was determined by crystal violet uptake.24 To assess the degree of cell cycle synchronicity within the population of pericytes, the cultures were stimulated with Ang-1 or -2 (100 ng/mL) for 48 hours and then fixed with 70% ethanol. After digestion with RNase, DNA was stained with propidium iodide and DNA density determined by flow cytometry, to obtain a measure of cell cycle progression.

Migration Assay
In vitro wound healing was used to evaluate cell migration. The pericytes were cultured to near confluence in 24-well plates with or without precoating with 12.5% extracellular matrix (Matrigel; BD Biosciences) in MEM containing 10% FCS at 37°C. After being maintained at quiescence in serum-free MEM for 45 minutes, the cell monolayers were wounded by a 1-cm3 tip pipette in one direction. The wounded cells were washed with PBS to remove cellular debris and were incubated with various concentrations Ang-1 or -2 at 37°C for 7 hours. Cell migration was monitored at initial wounding and at 7 hours under a phase-contrast microscope and calculated as migration distance (in micrometers) = (distance at time 0 – distance at time 7 hours)/2.

Visualization of F-Actin
As assembly of actin polymers appears to be an absolute requirement for pseudopod extension and cell migration, visualization of F-actin with phalloidin was used as a molecular marker for cell migration.25 Retinal pericytes grown on slides precoated with extracellular matrix (Matrigel; BD Biosciences) were treated with 100 ng/mL Ang-1 or -2 at 37°C for 48 hours. The cells were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100. Changes in F-actin structures were detected by incubating the cells for 30 minutes at RT with 0.33 µM TRITC-labeled phalloidin, and the cells were examined by fluorescence microscopy. A minimum of five cells in five random microscope fields for each treatment (at least 25 cells) were analyzed.

Statistical Analysis
All experiments were conducted at least three times. Unless otherwise indicated, all data are expressed as the mean ± SEM. An unpaired, two-tailed Student’s t-test was used to determine the significance of the results of Tie-2 expression, proliferation, migration, and apoptosis assays by analysis of the two group means. The Mann-Whitney test was used to determine statistical significance in the laser densitometry of Western blot analysis. Statistical analysis was performed (Minitab, ver. 14; Minitab, Inc., State College, PA), with P < 0.05 considered statistically significant.


   Results
Top
 Abstract
 Methods
 Results
Discussion
 References
 
Regulation of Tie-2 Expression in Retinal Pericytes
We determined the levels of both the cellular and soluble forms of Tie-2. (A soluble form of Tie-2 has been found to be released into the culture supernatant of other cell types.26 ) Both membrane and soluble Tie-2 expression was observed in pericytes grown under standard culture conditions (Fig. 1A) . With angiopoietin treatment at 100 ng/mL for 48 hours, Tie-2 in the cell lysate significantly increased by 29.5% (4.03 ng Tie-2/mg of protein for unstimulated vs. 5.21 ng/mg Ang-1 stimulated) and 11.5% (4.03 ng/mg for unstimulated vs. 4.49 ng/mg Ang-2 stimulated) when stimulated with Ang-1 and -2, respectively. In the supernatant, however, the soluble Tie-2 significantly decreased (by ~70%) on angiopoietin stimulation (1.62 ng Tie-2/mg of protein without stimulation vs. 0.54 ng/mg and 0.44 ng/mg Tie-2 when stimulated with Ang-1 and -2, respectively). Ang-1 induced tyrosine phosphorylation of Tie-2 in pericytes within 5 minutes, reaching a maximum level at 1 hour, followed by a slight downregulation (Fig. 1B) . In contrast, the kinetics of Tie2 phosphorylation induced by Ang-2 was a gradual upregulation, with the maximum phosphorylation occurring at 48 hours (Fig. 1B) .


Figure 1
View larger version (28K):
[in this window]
[in a new window]

  		FIGURE 1. Tie-2 expression in retinal pericytes. (A) Lysates and supernatants from retinal pericyte cultures were analyzed for Tie-2 by ELISA. Tie-2 levels were normalized for total protein concentrations in the cell lysates. Tie-2 was present in the lysate of unstimulated pericytes and increased on addition of Ang-1 and -2 for 48 hours. The levels of soluble Tie-2 in the medium decreased for pericytes treated with Ang-1 and -2. (B) The cell lysates were immunoprecipitated with anti-tyrosine phosphorylation antibody (PY20). The immunoprecipitated proteins were subject to ELISA analysis of Tie-2. The results are represented as the mean ± SEM *P < 0.05 versus the control group. To confirm the specificity of PY20 immunoprecipitation, cell lysates immunoprecipitated with PY20 and subjected to Western blot analysis demonstrated a single band for Tie-2 (inset). (C) The pericytes were analyzed by flow cytometry for expression of Tie-2. Top left: percentages of pericytes bearing Tie-2 immune reactivity at the pericyte surface and mean fluorescence intensity (MFI) of Tie-2-positive cells. (D) The percentage of cells within the culture staining desmin-positive is shown, to confirm the purity of the pericyte cultures with increasing passages. Data represent the mean results of three experiments. Vertical bars: upper limit of the negative isotype control. (E) Microvascular endothelial cell contamination of pericyte cell lysates was ruled out by Western blot analysis with an anti-PECAM-1 antibody (microvascular endothelial cell lysates were used as the positive control). Data are expressed as the mean ± SEM (n = 3).

 
Flow cytometry demonstrated that more than 98% of cultured pericytes expressed Tie-2 (normal control: 98.4% ± 0.5%; 100 ng/mL Ang-1 or -2 for 48 hours: 99.5% ± 0.3% or 99.7% ± 0.2%, respectively; Fig 1C ). Compared with normal control cells, the angiopoietin-treated pericytes exhibited an increase in levels of Tie-2 shown as MFI (arbitrary units; normal control: 99.8 ± 17.6; Ang-1: 178.7 ± 9.8; Ang-2: 138.3 ± 14.4; Fig 1C ). Furthermore, the cells exhibited typical pericyte morphology when assessed by microscopy, and greater than 94% of the cells expressed desmin (see representative flow cytometric analysis in Fig. 1D ). Consistent with the desmin expression, no PECAM-1 (a marker for endothelial cells but not pericytes) was detected in pericyte cultures by Western blot analysis, whereas microvascular endothelial cells strongly expressed PECAM-1 (Fig. 1E) .

Ang-1 Regulation of CD13 Expression in Retinal Pericytes
CD13 protein was expressed by retinal pericytes and was upregulated by Ang-1 in a dose-dependent manner (Fig. 2A) . Tie-2 antisense (but not Tie-2 sense) treatment significantly reduced CD13 expression (Fig. 2B) and abolished Ang-1-induced CD13 expression in pericytes (Fig. 2C) . Ang-2 had no effect on CD13 expression in retinal pericytes (data not shown).

Differential Effect of Ang-1 and -2 on Pericyte Survival
TNF-{alpha} (100 ng/mL) induced maximum apoptosis of pericytes (25%) compared with baseline (7%) by 24 hours (Fig. 3) . Ang-1 at 100 ng/mL significantly inhibited TNF-{alpha}-induced apoptosis with the number of apoptotic cells decreasing by approximately 50%. By contrast 100 ng/mL Ang-2 enhanced TNF-{alpha}-induced apoptosis reaching up to 30% by 48 hours (Fig. 3B) . A similar response was observed with high glucose, Ang-1 decreased glucose-induced apoptosis by 60%, and Ang-2 enhanced apoptosis by 47% at 25 mM glucose (Fig 3C) .


Figure 3
View larger version (25K):
[in this window]
[in a new window]

  		FIGURE 3. The effect of angiopoietins on TNF-{alpha}-induced apoptosis of retinal pericytes. Apoptosis was induced in TNF-{alpha}-treated retinal pericytes. Pericytes, exposed to Ang-1 and -2 (100 ng/mL), either in the presence or absence of TNF-{alpha} (100 ng/mL) for the indicated times, were stained with annexin-V-FITC and propidium iodide and analyzed by flow cytometry. (A) Representative experiment from flow cytometric analysis of annexin-V FITC/propidium iodide stained pericytes. (B) Cell apoptosis was expressed as the percentage of apoptotic cells in the total cell population. TNF-{alpha}-induced maximum apoptosis of pericytes by 24 hours. Ang-2 enhanced TNF-{alpha}-induced apoptosis with maximum effect at 48 hours, whereas Ang-1 significantly reduced apoptosis. Data are expressed as the mean ± SEM of results in three separate experiments. (C) Apoptosis in pericytes exposed to high glucose (15 and 25 mM) and the effects of Ang-1 and -2 as a percentage of apoptotic cells in the total cell population. Data are expressed as the mean ± SEM of results in three separate experiments.

 
Effects of Ang-1 and -2 on Retinal Pericyte Behavior
Proliferation.
Retinal pericytes exposed to Ang-2 (between 1 and 200 ng/mL) exhibited a small but significant dose-dependent increase in cell proliferation (Fig. 4A) , whereas Ang-1 had no mitogenic effect on the cells.

The effects of Ang-1 and -2 on cell cycle progression were evaluated by flow cytometric analysis of the DNA content. As shown in Table 1 most of the cells under control conditions were in the G0/G1 phase. Exposure to Ang-2 resulted in a significant (P < 0.05) shift toward the S-phase and G2/M phase, whereas Ang-1 had no effect compared with control.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Ang-1- and -2-Mediated Pericyte Cell Cycle Progression

 
Migration.
Addition of Ang-1 and -2 had no obvious effect on retinal pericyte migration on standard cell culture plastic (Fig. 4B) . A recent study27 showed that an antibody against CD13 inhibits cell migration on basement membrane matrix (Matrigel; BD Biosciences). Because in the present study Ang-1 upregulated CD13 expression in the pericytes, we assessed pericyte migration on basement membrane matrix–coated slides. As shown in Figure 4B , retinal pericyte migration was significantly stimulated by Ang-1 in a dose-dependent manner on basement membrane matrix-coated slides, and the effect was almost completely inhibited by Tie-2 antisense. There was no significant increase in pericyte migration after treatment with Ang-2 (Fig. 4B) .

We next determined whether the changes in cell migration were associated with alterations in the F-actin cytoskeletal organization. The control cells were flat with thin, uniform, parallel actin filaments (stress fibers) departing from single foci and extending throughout the length of the cell, with the presence of lamellipodia (Figs. 5A 5D) . There were a few dominant, elongated pseudopodia and a few actin-containing fine cell extensions. Ang-1 treatment caused dramatic reorganization of the actin cytoskeleton that resulted in a reduction in the number of stress fibers and lamellipodia, with an uneven thickening of the remaining stress fibers and an increase in both dominant leading-edge pseudopodia observed at the ends of the cells and in invadopodia (Fig. 5B) . By contrast, Ang-2-treated retinal pericytes remained flat with the partial disassociation of actin into aggregates (Fig. 5E) . Tie-2 antisense treatment abolished the effects of Ang-1 and -2 on changes in actin cytoskeleton in the retinal pericytes (Figs. 5C 5F) .


Figure 5
View larger version (21K):
[in this window]
[in a new window]

  		FIGURE 5. Cytoskeletal changes in retinal pericytes exposed to Ang-1 and -2. Retinal pericytes were fixed and stained with phalloidin-TRITC. A minimum of five cells in five random microscope fields for each treatment (a total of at least 25 cells) were analyzed. Control cells showed that thin, uniform, parallel actin filaments with few elongated pseudopodia (arrowhead) (A, D). Treatment with 100 ng/mL Ang-1 resulted in a reduction in the number of stress fibers, with an uneven thickening of the remaining fibers in the dominant leading-edge pseudopodia observed at the ends of the cells (arrow) (B). Ang-2 (100 ng/mL)-treated pericytes remained flat with partial dissociation of actin into aggregates (chevron) (E). Tie-2 antisense treatment abolished the effects of both Ang-1 and -2 on changes in actin cytoskeleton in the retinal pericytes (C, F). Bar, 50 µm.

 

   Discussion
Top
 Abstract
 Methods
 Results
 Discussion
References
 
Retinal pericytes are intimately associated with the vascular endothelium and have a critical role in maintaining the functional integrity of the capillary unit.28 Acting as contractile cells with a well-developed actin microfilamentous network, pericytes alter capillary luminal diameter and regulate retinal blood flow.22 Retinal pericytes also appear to have a critical trophic function in promoting vessel maturation and endothelial survival. Selective degeneration of pericytes in the retinal capillary vessels is a distinguishing feature of early retinal vascular damage in diabetes and can result in decreased capillary tonicity, formation of microaneurysms, and vessel dilation.23

It has been proposed that hyperglycemia causes pericyte dysfunction, apoptosis, and ultimately pericyte loss.29 30 Our finding that Ang-2 protein is upregulated in the diabetic retina is consistent with that observed in a chronic hyperglycemia animal model of diabetic retinopathy.2 Ang-2 has different effects on Tie-2 phosphorylation dependent on the cell types on which it acts. For example, Ang-2 binds to endothelial cell Tie-2 without inducing its phosphorylation but can activate Tie-2 in genetically modified mesenchymal cells.31 Although Tie-2 has been presumed to be restricted to endothelial cells and hematopoietic stem cells, we provide evidence that Tie-2 receptors are expressed by retinal pericytes, confirming other reports that Tie-2 expression can be found in nonendothelial mesenchymal cell types.10 11 32 Tie-2 immunolocalization to nonvascular cells in both the inner and outer retina is in agreement with that reported by Ohashi et al.33

Both Ang-1 and -2 enhanced Tie-2 expression in cultured retinal pericytes, indicating that upregulation rather than constitutive expression of Tie-2 occurs during the progression of diabetic retinopathy. Consistent with these findings, both Ang-1 and -2 induced Tie-2 autophosphorylation, although robust Tie-2 activation by Ang-2 required prolonged exposure to Ang-2. Furthermore, the decrease in sTie-2 levels in pericyte-conditioned media after Ang-1 or -2 treatment suggests that Tie-2 may not be susceptible to proteolysis after ligand binding.34

Retinal capillary pericytes extend long cytoplasmic processes to form interdigitating contacts with endothelial cells that facilitate the maturation, remodeling, and maintenance of the vascular system. During retinal blood vessel development, Ang-2 acts as a factor that primes endothelial cells for angiogenesis by destabilizing interactions between endothelial cells and perivascular cells.18 Increased Ang-2 is thought to lead to persistent disruption of the cellular cross-talk between pericytes and endothelial cells in early diabetic retinopathy, culminating in pericyte loss and vessel destabilization. However, our data that Ang-2 promotes retinal pericyte proliferation in a dose-dependent manner add a new aspect to the complexity of the Ang-2/Tie-2 system. A recent report35 suggests that mitotically active pericytes can form angiogenic sprouts or tubes during the early phase of neovascularization in tumors and in developing retinal tissues. Although Ang-1 has been identified as the primary activating ligand for Tie-2, Ang-2 possesses similar receptor affinity to Tie-2 and it has been proposed that in response to Ang-1, Tie-2 is rapidly internalized and targeted for degradation.36 By contrast, Ang-2 only weakly activates Tie-2 without significantly stimulating Tie-2 internalization, leading to amplification and a delay before full activation of Tie-2 occurs.

CD13 was initially identified as an important marker of subpopulations of hematopoietic cells.20 CD13 is identical with a predominant metalloproteinase (MMP), a Zn2+-dependent ectopeptidase37 that activates or inactivates bioactive peptides on the cell surface by preferential cleaving of proteins with NH2-terminal neutral amino acids38 39 and that regulates the availability of peptides to adjacent cells. A recent study showed that CD13 is expressed by vascular cells and may play a role in angiogenesis.40 In our study, retinal pericytes also expressed CD13 and Ang-1 enhanced retinal pericyte CD13 expression in a dose-dependent manner. Recently, MMPs have been reported to be involved in tumor pericyte recruitment.41 The strong enhancement of a migratory cell phenotype (an increase in actin polymerization and pseudopod extension) in Ang-1-treated pericytes on synthetic matrix (Matrigel; BD Biosciences) in this study, accompanied with increased expression CD13, indicates that Ang-1 indeed increased the motility of pericytes, at least in part, via CD13 mediated-MMP activity. Moreover, that Ang-1 stimulated pericyte migration but not proliferation reinforces our hypothesis that interaction of Ang-1 with Tie-2 in pericytes leads to vessel maturation via recruitment of pericytes and smooth muscle cells. Of interest, CD13 has also been reported to be a putative marker for cerebral pericyte maturation.20

It has been proposed that overexpression of Ang-1 in vivo results in a dramatic increase in microvessel number and vessel branching42 indicating that Ang-1 plays an important role in retinal neovascularization, including maturation and remodeling, rather than being a effector of diabetic retinopathy. Furthermore, it has been shown that Ang-1 can rescue the vessels from leakiness caused by VEGF-A without interfering with induction of angiogenesis.43 Our observation that retinal pericytes express Tie-2 raises the possibility that Ang-1 helps to maintain and stabilize mature vessels by stimulating pericyte viability. Because Ang-1 also leads to increased Akt activation in endothelial cells, thus enhancing survival signals,44 we subsequently tested whether Ang-1 can block retinal pericyte apoptosis induced by high glucose29 30 or TNF-{alpha} (a factor implicated in the pathogenesis of diabetic retinopathy45 ). Retinal pericytes treated with Ang-1 demonstrated significantly decreased apoptosis compared with those treated with TNF-{alpha} or high glucose alone, indicating that Ang-1 can improve pericyte survival.46 As Ang-1 has been shown to be highly expressed in adult tissue, our results further emphasize its role in maintaining previously developed and mature blood vessels.

It is now becoming evident that the effect of angiopoietins is highly context dependent and contingent on the nature of the local environment, the cellular phenotype, and the stage of the disease. Fiedler and Reiss47 elegantly demonstrated that Ang-2 functions are dependent on the presence of other cytokines, and their findings are supported by our observation that Ang-2 enhanced the apoptotic effect of TNF-{alpha} but had no effect alone. Pericyte phenotype differs as to whether it is in its quiescent state encased within the vascular basement membrane or proliferating at the vanguard of newly forming blood vessels. Although we would not expect this phenotype to proliferate, we would expect it to be susceptible to apoptosis. However, the angiopoietins would be expected to promote a different response in the activated pericytes associated with angiogenesis, and our data support this. Diabetic retinopathy is a dynamic process and at stages during the disease both pericyte phenotypes will be present. Thus, it is not surprising that the angiopoietins can have multiple effects dependent on the context of endogenous and exogenous factors.

In conclusion, our results suggest (1) a critical role for Ang-2 in pericyte loss in early diabetic retinopathy, (2) a possible role for Ang-2 in pericyte proliferation at a later stage in the progression of diabetic retinopathy, and (3) Ang-1-mediated improvement of survival, activation, and migration of retinal pericytes during the establishment of retinal new vessels. Targeted regulation of angiopoietin isoforms may offer a therapeutic approach as an adjunct treatment for the microvascular complications associated with diabetic retinopathy. Das et al.48 have reported that a Tie-2 antagonist, muTek delta Fc, can inhibit revascularization in an oxygen model of retinopathy.


   Footnotes
 
Supported by Diabetes UK, the Wellcome Trust, UK, and Research to Prevent Blindness; and Regeneron Pharmaceuticals, Inc. (Tarrytown, NY), which provided Ang-1 and -2.

Submitted for publication September 17, 2007; revised November 14, 2007, and January 4, 2008; accepted March 6, 2008.

Disclosure: J. Cai, Regeneron Pharmaceuticals, Inc. (F); O. Kehoe, None; G.M. Smith, None; P. Hykin, None; M.E. Boulton, Regeneron Pharmaceuticals, Inc. (F)

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: Michael E. Boulton, Department of Ophthalmology and Visual Sciences, The University of Texas Medical Branch, 301 University Boulevard, Galveston, TX 77550; boultonm{at}utmb.edu.


   References
Top
 Abstract
 Methods
 Results
 Discussion
 References
 

  1. Miller J, D'Amico D. Proliferative diabetic retinopathy. Albert D Jakobiec F eds. Principles and Practice of Ophthalmology. 1994;760–782. WB Saunders Philadelphia.
  2. Hammes HP, Lin J, Wagner P, et al. Angiopoietin-2 causes pericyte dropout in the normal retina: evidence for involvement in diabetic retinopathy. Diabetes. 2004;53:1104–1110.[Abstract/Free Full Text]
  3. Cai J, Boulton M. The pathogenesis of diabetic retinopathy: old concepts and new questions. Eye. 2002;16:242–260.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  4. Ishibashi T, Inomata H, Sakamoto T, Ryan SJ. Pericytes of newly formed vessels in experimental subretinal neovascularization. Arch Ophthalmol. 1995;113:227–231.[Abstract/Free Full Text]
  5. Takagi H, Koyama S, Seike H, et al. Potential role of the angiopoietin/tie2 system in ischemia-induced retinal neovascularization. Invest Ophthalmol Vis Sci. 2003;44:393–402.[Abstract/Free Full Text]
  6. Jones N, Master Z, Jones J, et al. Identification of Tek/Tie2 binding partners: binding to a multifunctional docking site mediates cell survival and migration. J Biol Chem. 1999;274:30896–30905.[Abstract/Free Full Text]
  7. Jones N, Voskas D, Master Z, Sarao R, Jones J, Dumont DJ. Rescue of the early vascular defects in Tek/Tie2 null mice reveals an essential survival function. EMBO Rep. 2001;2:438–445.[Web of Science][Medline][Order article via Infotrieve]
  8. Fadel BM, Boutet SC, Quertermous T. Octamer-dependent in vivo expression of the endothelial cell-specific Tie2 gene. J Biol Chem. 1999;274:20376–20383.[Abstract/Free Full Text]
  9. Maisonpierre PC, Suri C, Jones PF, et al. Angiopoietin-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis. Science. 1997;277:55–60.[Abstract/Free Full Text]
  10. Davis S, Aldrich TH, Jones PF, et al. Isolation of angiopoietin-1, a ligand for the Tie2 receptor, by secretion-trap expression cloning. Cell. 1996;87:1161–1169.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  11. Puri MC, Bernstein A. Requirement for the TIE family of receptor tyrosine kinases in adult but not fetal hematopoiesis. Proc Natl Acad Sci USA. 2003;100:12753–12758.[Abstract/Free Full Text]
  12. Park YS, Kim NH, Jo I. Hypoxia and vascular endothelial growth factor acutely up-regulate angiopoietin-1 and Tie2 mRNA in bovine retinal pericytes. Microvasc Res. 2003;65:125–131.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  13. Suri C, Jones PF, Patan S, et al. Requisite role of angiopoietin-1, a ligand for the TIE2 receptor, during embryonic angiogenesis. Cell. 1996;87:1171–1180.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  14. Hawighorst T, Skobe M, Streit M, et al. Activation of the tie2 receptor by angiopoietin-1 enhances tumor vessel maturation and impairs squamous cell carcinoma growth. Am J Pathol. 2002;160:1381–1392.[Abstract/Free Full Text]
  15. Kampfer H, Pfeilschifter J, Frank S. Expressional regulation of angiopoietin-1 and -2 and the Tie-1 and -2 receptor tyrosine kinases during cutaneous wound healing: a comparative study of normal and impaired repair. Lab Invest. 2001;81:361–373.[Web of Science][Medline][Order article via Infotrieve]
  16. Yamakawa M, Liu LX, Belanger AJ, et al. Expression of angiopoietins in renal epithelial and clear cell carcinoma cells: regulation by hypoxia and participation in angiogenesis. Am J Physiol. 2004;287:F649–F657.[Web of Science]
  17. Hanahan D. Signaling, vascular morphogenesis and maintenance. Science. 1997;277:48–50.[Free Full Text]
  18. Hackett SF, Ozaki H, Strauss RW, et al. Angiopoietin 2 expression in the retina: upregulation during physiologic and pathologic neovascularization. J Cell Physiol. 2000;184:275–284.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  19. Cai J, Ahmad S, Jiang WG, et al. Activation of vascular endothelial growth factor receptor-1 sustains angiogenesis and Bcl-2 expression via the phosphatidylinositol 3-kinase pathway in endothelial cells. Diabetes. 2003;52:2959–2968.[Abstract/Free Full Text]
  20. Griffin JD, Ritz J, Nadler LM, Schlossman SF. Expression of myeloid differentiation antigens on normal and malignant myeloid cells. J Lab Invest. 1981;68:932–941.
  21. Ramsauer M, Krause D, Dermietzel R. Angiogenesis of the blood-brain barrier in vitro and the function of cerebral pericytes. FASEB J. 2002;16:1274–1276.[Abstract/Free Full Text]
  22. Cai J, Jiang WG, Grant MB, Boulton M. Pigment epithelium-derived factor inhibits angiogenesis via regulated intracellular proteolysis of vascular endothelial growth factor receptor 1. J Biol Chem. 2006;281:3604–3613.[Abstract/Free Full Text]
  23. Lauffenburger DA, Horwitz AF. Cell migration: a physically integrated molecular process. Cell. 1996;84:359–369.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  24. Reusch P, Barleon B, Weindel K, et al. Identification of a soluble form of the angiopoietin receptor TIE-2 released from endothelial cells and present in human blood. Angiogenesis. 2001;4:123–131.[CrossRef][Medline][Order article via Infotrieve]
  25. Ikeda N, Nakajima Y, Tokuhara T, et al. Clinical significance of aminopeptidase N/CD13 expression in human pancreatic carcinoma. Clin Cancer Res. 2003;9:1503–1508.[Abstract/Free Full Text]
  26. Kern TS, Engerman RL. Pharmacological inhibition of diabetic retinopathy: aminoguanidine and aspirin. Diabetes. 2001;50:1636–1642.[Abstract/Free Full Text]
  27. Chakravarthy U, Gardiner TA. Endothelium-derived agents in pericyte function/dysfunction. Prog Retin Eye Res. 1999;18:511–527.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  28. Lindahl P, Johansson BR, Leveen P, Betsholtz C. Pericyte loss and microaneurysm formation in PDGF-B-deficient mice. Science. 1997;277:242–245.[Abstract/Free Full Text]
  29. Miller AG, Smith DG, Bhat M, Nagaraj RH. Glyoxalase is critical for human retinal capillary pericyte survival under hyperglycaemic conditions. J Biol Chem. 2006;281:11864–11871.[Abstract/Free Full Text]
  30. Romeo G, Liu WH, Asnaghi V, Kern TS, Lorenz M. Activation of nuclear factor-{kappa}B induced by diabetes and high glucose regulates a proapoptotic program in retinal pericytes. Diabetes. 2002;51:2241–2248.[Abstract/Free Full Text]
  31. Teichert-Kuliszewska K, Maisonpierre PC, Jones N, et al. Biological action of angiopoietin-2 in a fibrin matrix model of angiogenesis is associated with activation of Tie2. Cardiovasc Res. 2001;49:659–670.[Abstract/Free Full Text]
  32. Iurlaro M, Scatena M, Zhu WH, et al. Rat aorta-derived mural precursor cells express the Tie2 receptor and respond directly to stimulation by angiopoietins. J Cell Sci. 2003;116:3635–3643.[Abstract/Free Full Text]
  33. Ohashi H, Takagi H, Koyama S, et al. Alterations in expression of angiopoietins and the Tie-2 receptor in the retina of streptozotocin induced diabetic rats. Mol Vis. 2004;10:608–617.[Web of Science][Medline][Order article via Infotrieve]
  34. Yabkowitz R, Meyer S, Black T, Elliott G, Merewether LA, Yamane HK. Inflammatory cytokines and vascular endothelial growth factor stimulate the release of soluble tie receptor from human endothelial cells via metalloprotease activation. Blood. 1999;93:1969–1979.[Abstract/Free Full Text]
  35. Ozerdem U, Stallcup WB. Early contribution of pericytes to angiogenic sprouting and tube formation. Angiogenesis. 2003;6:241–249.[CrossRef][Medline][Order article via Infotrieve]
  36. Bogdanovic E, Nguyen VPKH, Dumont DJ. Activation of Tie2 by angiopoietin-1 and angiopoietin-2 results in their release and receptor internalization. J Cell Sci. 2006;119:3551–3560.[Abstract/Free Full Text]
  37. Shimamura M, Hazato T, Iwaguchi T. Enkephalin-degrading aminopeptidase in the longitudinal muscle layer of guinea pig small intestine: its properties and action on neuropeptides. J Biochem. 1991;109:492–497.[Abstract/Free Full Text]
  38. Olsen J, Cowell GM, Konigshofer E, et al. Complete amino acid sequence of human intestinal aminopeptidase N as deduced from cloned cDNA. FEBS Lett. 1988;238:307–314.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  39. Ashmun RA, Look AT. Metalloprotease activity of CD13/aminopeptidase N on the surface of human myeloid cells. Blood. 1990;75:462–469.[Abstract/Free Full Text]
  40. Pasqualini R, Koivunen E, Kain R, et al. Aminopeptidase N is a receptor for tumor-homing peptides and a target for inhibiting angiogenesis. Cancer Res. 2000;60:722–727.[Abstract/Free Full Text]
  41. Chantrain CF, Henriet P, Jodele S, et al. Mechanisms of pericyte recruitment in tumour angiogenesis: a new role for metalloproteinases. Eur J Cancer. 2006;42:310–318.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  42. Machein MR, Knedla A, Knoth R, Wagner S, Neuschl E, Plate KH. Angiopoietin-1 promotes tumor angiogenesis in a rat glioma model. Am J Pathol. 2004;165:1557–1570.[Abstract/Free Full Text]
  43. Gamble JR, Drew J, Trezise L, et al. Angiopoietin-1 is an antipermeability and anti-inflammatory agent in vitro and targets cell junctions. Circ Res. 2000;87:603–607.[Abstract/Free Full Text]
  44. Kim I, Kim HG, So JN, Kim JH, Kwak HJ, Koh GY. Angiopoietin-1 regulates endothelial cell survival through the phosphatidylinositol 3'-kinase/Akt signal transduction pathway. Circ Res. 2000;86:24–29.[Abstract/Free Full Text]
  45. Gardiner TA, Gibson DS, de Gooyer TE, de la Cruz VF, McDonald DM, Stitt AW. Inhibition of tumor necrosis factor-alpha improves physiological angiogenesis and reduces pathological neovascularization in ischemic retinopathy. Am J Pathol. 2005;166:637–644.[Abstract/Free Full Text]
  46. Joussen AM, Poulaki V, Mitsiades N. Nonsteroidal anti-inflammatory drugs prevent early diabetic retinopathy via TNF-alpha suppression. FASEB J. 2002;16:438–440.[Free Full Text]
  47. Fiedler U, Reiss Y, Scharpfenecker M, et al. Angiopoietin-2 sensitizes endothelial cells to TNF-alpha and has a crucial role in the induction of inflammation. Nat Med. 2006;12:235–239.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  48. Das A, Fanslow W, Cerretti D, et al. Angiopoietin/Tek interactions regulate MMP-9 expression and retinal neovascularization. Lab Invest. 2003;83:1637–1645.[CrossRef][Web of Science][Medline][Order article via Infotrieve]



This article has been cited by other articles:


Home page
 Arterioscler. Thromb. Vasc. Bio.Home page
K. Gaengel, G. Genove, A. Armulik, and C. Betsholtz
Endothelial-Mural Cell Signaling in Vascular Development and Angiogenesis
Arterioscler. Thromb. Vasc. Biol., May 1, 2009; 29(5): 630 - 638.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Cai, J.
Right arrow Articles by Boulton, M. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Cai, J.
Right arrow Articles by Boulton, M. E.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS