Müller Cell Production of Insulin-like Growth Factor–Binding Proteins In Vitro: Modulation with Phenotype and Growth Factor Stimulation

Abstract

purpose. Müller cells are present in diabetic fibrocontractive ocular tissues and generate tractional forces in response to insulin-like growth factors. Recent studies indicate that diabetes-associated increases in vitreous insulin-like growth factor activity are, in part, attributable to changes in insulin-like growth factor binding proteins (IGFBPs). The objectives of this study were to evaluate Müller cells as a source of IGFBPs and characterize changes associated with cell phenotype and growth factor stimuli known to be present in diabetic vitreous.

methods. Müller cells isolated from normal porcine retina were maintained in culture for 1 and 5 weeks, yielding phenotypes described as proliferative and myofibroblastic. RNA preparations from porcine liver, retina, and Müller cell cultures were evaluated by RT-PCR and Northern blot analysis. IGFBP production was verified by Western ligand and Western blot analysis of Müller-cell–conditioned media and detergent-extracted proteins.

results. Molecular biological analyses of RNA from normal retina and from proliferative and myofibroblastic Müller cells did not detect message for IGFBP-1, but revealed progressive increases in message abundance for IGFBP-2, -3, -4, and -6. IGFBP-5 message was detected in all samples, but was least abundant in myofibroblastic Müller cells. Stimulation of myofibroblastic Müller cells by IGF-I and -II, but not PDGF, further increased message abundance and production of IGFBP-2, -4, -5, and -6, but not IGFBP-3.

conclusions. Müller cell production of IGFBPs changes with phenotype and, in most cases, is highest in the cells most likely to participate in fibrocontractive retinal disease. IGFBP production by these cells is further increased by IGF-I and -II, growth factors known to be present and active in proliferative vitreoretinal disorders, suggesting that Müller cells represent a potential source of vitreous IGFBPs in disorders involving this cell type.

Diabetes remains the leading cause of vision loss in adults between the second and seventh decades of life, and the risk of the development of proliferative diabetic retinopathy (PDR) increases with disease duration. ^{1 2 3} A substantial body of information suggests that Müller cells, the principal retinal glia, play causal roles in this disorder. Studies of diabetic retina and animal models of hyperglycemia have revealed changes in Müller cell morphology, protein expression, and physiology well in advance of detectable retinopathy. ^{4 5 6 7 8} In immunohistochemical studies of diabetic epiretinal membranes, Müller cells were identified as a component of these proliferative tissues, ^{9 10 11 12} and findings in studies of Müller cell behavior in an animal model have demonstrated the capacity of this cell type to cause traction retinal detachment. ¹³  

Much of what we know about Müller cell tractional force generation has been derived from studies of cell behavior in vitro. After isolation and introduction into culture, Müller cells adopt a proliferative phenotype, with increased expression of glial fibrillary acid protein (GFAP) and reduced expression of several constitutively expressed enzymes ^{14 15} —changes that generally mimic Müller cell reactive changes to retinal insult, such as injury or detachment. ¹⁶ After several weeks in culture, Müller cells adopt a myofibroblastic phenotype characterized by progressive loss of GFAP expression, de novo expression of the myoid marker α-smooth muscle actin (αSMA), and the capacity to generate tractional forces through extracellular matrix contraction. ^{15 17} These studies also revealed that Müller cell tractional force generation is not constitutive, but is stimulated by several exogenous growth factors, including members of the platelet-derived growth factor (PDGF) and insulin-like growth factor (IGF) families. In addition, studies of Müller cell responses to vitreous fluids from patients with nondiabetic and diabetes-associated retinal disorders revealed significant increases in extracellular matrix contraction-promoting activity, most of which was attributable to IGF rather than PDGF. ^{18 19}  

The origins of vitreous IGF activity in these disorders are uncertain at present. Although there are reports of diabetes-associated increases in vitreous IGF-I and -II, ^{20 21 22} the concentrations of these same growth factors in normal vitreous are also well above the threshold of Müller cell sensitivity. ^{15 17} As IGF biological activity in normal vitreous is low or undetectable, these observations suggest that IGF is in some way controlled or attenuated. ^{18 19} The IGF system is composed of two ligands (IGF-I and -II) and at least six high-affinity IGF binding proteins (IGFBPs). Each is the product of a separate gene and has the capacity to inhibit and potentiate growth factor activities, depending on the experimental system used for study. ^{23 24 25 26} IGFBPs are present in normal human vitreous and in altered concentrations in diabetic vitreous and, as a result, there is now considerable interest in understanding the disease-related changes that may contribute to loss of growth factor control. ^{20 27}  

The liver is the major source of plasma IGFBPs, ^{24 28 29} and it is easily accepted that hemorrhage or increased effusion arising from diabetes-associated changes in the blood retinal barrier permeability could alter vitreous IGFBP concentrations. ^{21 30 31} Changes in IGFBP production by local ocular cells is another plausible mechanism, as RNA for IGFBP-2, -3, and -5 have been detected in ocular tissues. ^{23 32 33 34 35} There is also the potential for IGFBP contribution from fibroproliferative processes, as the cells involved are in direct contact with vitreous fluids. In light of the probable involvement of Müller cells in the pathogenesis of PDR and regulatory effects of vitreous IGFBPs, is it important that binding protein production by local cells be considered a mechanism that contributes to vitreous change. As no information is available about the capacity of Müller cells to produce IGFBPs, this unexplored and relatively broad topic was the focus of this study. Increased emphasis was placed on IGFBP-1, -2, and -3, because these are the species known to be present in normal and/or diabetic vitreous. However, in this study, we evaluated the potential production of all six high-affinity IGFBPs as, to our knowledge, the presence of IGFBP-4, -5, -6 in diabetic vitreous had not been considered in any published studies. Müller cell phenotype was yet another mechanistic consideration as IGFBP biosynthesis could change in concert with cell function between the normal, proliferative, and myofibroblastic states. Finally, the potential modulating influences of exogenous growth factors were also considered, as we now know that the vitreous humor in normal and diabetic eyes elicits different Müller cell responses. ¹⁹  

Methods

Isolation and Culture of Porcine Müller Cells

Müller cells were isolated from normal porcine retina and maintained in culture, using previously published methods. ¹⁷ The methods used to secure animal tissue complied with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Institutional Review Board at the University of Alabama at Birmingham. Using polyclonal rabbit anti-GFAP and mouse monoclonal anti-αSMA as described previously, ¹⁷ we evaluated the Müller cell phenotype in culture by indirect immunofluorescence of fixed, permeabilized cells attached to glass coverslips. Proliferative Müller cells at 1 week in culture were 96% positive for GFAP and 0% positive for αSMA, whereas myofibroblastic Müller cells at 5 weeks in culture were 0% positive for GFAP and 98% positive for αSMA. 

Primer Design

Porcine and human sequences for IGFBP-1, -2, -3, -4, -5, and -6 were obtained from the National Center for Biotechnology Information (NCBI; Bethesda, MD). Specific primers were designed to yield PCR amplification products between 220 and 500 bp, by using the GeneFisher or Primer3 programs provided by NCBI (www.ncbi.nlm.nih.gov/). Potential primer sequences were then searched for homology against all known sequences in the genetic databases by using the Basic Local Alignment Search Tool (BLAST) available at NCBI. (Table 1) . Successful primer design was confirmed with RNA harvested from the livers of normal donor animals. 

Reverse Transcription–Polymerase Chain Reaction

Total RNA was isolated from porcine liver, porcine retina, and Müller cell cultures with extraction reagent used according to the manufacturer’s instructions (TRIzol; Invitrogen-Gibco, Grand Island, NY). RT-PCR reactions were performed with 1 μg total RNA, 20 pM each primer, and commercial RT-PCR reagents (Ready-To-Go; Amersham Pharmacia Biotech, Piscataway, NJ). Reaction programs were performed on a thermocycler (MiniCycler model PTC-150; MJ Research, Watertown, MA) and included (1) a reverse transcription program of 20 minutes at 42°C and 5 minutes at 95°C; (2) 35 cycles of 1 minute at 95°C, 45 seconds at the appropriate annealing temperature, and 45 seconds at 72°C; and (3) 5 minutes at 72°C. For negative control reactions, reverse transcriptase was inactivated at 95°C for 10 minutes before addition of the primer and template. PCR products were separated on 2% agarose gels, visualized with ethidium bromide, and documented with photography (Polaroid camera, Polaroid Corp., Cambridge, MA, equipped with a Tiffen 40.5-mm Deep Yellow 15 filter; Fisher Biotech, Pittsburgh, PA). 

Real-Time RT-PCR

Reactions were performed with a thermocycler and commercial kit (DNA Engine Opticon 2; MJ Research; and the Quantitect SYBR green RT-PCR kit; Qiagen, Valencia, CA), using total RNA. To verify the accuracy of spectrophotometric quantification of the templates, we measured levels of glyceraldehyde-3-phosphate dehydrogenase (G3PDH) and 18S rRNA with commercially available primers (G3PDH; BD Biosciences-Clontech, Palo Alto, CA; 18S, Ambion, Austin, TX). Each template and primer set combination was optimized by using a temperature gradient, and the templates were evaluated in serial dilutions to ensure that evaluations were within the linear range of RNA concentrations. Reaction products were separated on 2% agarose/Tris/borate/EDTA (TBE) gels and stained with ethidium bromide to confirm product size. The optimized annealing temperature and template amounts were used in a single run, along with a no-template control to identify primer-dimers and a no-reverse-transcriptase control for DNA contamination. The final runs of the templates and primer sets were performed in triplicate with G3PDH as the control. Analysis points or cycle thresholds (C _T ) were selected as 10% of maximum fluorescence normalized to G3PDH (ΔC _T ) and then to the experimental variable (ΔΔC _T ). Increases (x-fold) were calculated using the formula 2^ΔΔCT. 

Northern Blot Analysis

Total RNA preparations were suspended in MOPS (3-(N-morpholino)-propanesulfonic acid)/formamide loading buffer, heated at 65°C for 5 minutes, and then placed on ice. The 1.2% agarose/formaldehyde gels were loaded with 5 or 10 μg of RNA from porcine liver and retina or 5 μg each of RNA from Müller cell cultures. Gels were run in MOPS buffer at 100 V for 2 hours, rinsed in diethyl pyrocarbonate (DEPC)–treated water three times for 10 minutes each, and neutrally transferred for 6 hours in 20× sodium chloride/sodium citrate (SSC) to a nylon membrane (Nytran Supercharge Membrane; Schleicher & Schuell, Keene, NH) using a commercial transfer system (Turboblotter Rapid Downward Transfer System; Schleicher & Schuell). Membranes were rinsed briefly in 2× SSC and the RNA covalently bonded by UV irradiation (UV Stratalinker 1800; Stratagene, La Jolla, CA). Radioactive probes were prepared (Strip-EZ PCR; Ambion) per the supplier’s instructions. Template cDNA for each IGFBP probe was prepared as previously described for RT-PCR, but with an additional gel purification step (Qiaquick; Qiagen). Radio-labeled dATP was incorporated into the PCR product during a 35-cycle amplification followed by purification with the purification kit. Blots were prehybridized in hybridization buffer (UltraHyb; Ambion) for 1 hour at 42°C and hybridized with 1 × 10⁶ cpm/mL of the radio-labeled probe overnight at the same temperature. The membrane was subjected to washes of increasing stringency with SSC, and background reduction was checked with a Geiger counter. Blots were placed in a film cassette with the film (Biomax MS; Eastman Kodak, Rochester, NY) and two intensifying screens, exposed at −80°C, and developed at different times. 

Extracellular Matrix Contraction Assays

Müller cell responses to IGF-I and -II, and R³IGF-I and R⁶IGF-II were evaluated with a previously described assay of Müller cell extracellular matrix contraction. ¹⁵ Recombinant growth factors and analogues were obtained from GroPep, Ltd. (Adelaide, Australia). 

Western Ligand Blot Analysis

Experiments to evaluate IGFBP production by Müller cells were performed by using the same procedures and reagents as for routine tissue culture except that after trypsin release, the cells were washed twice in serum-free DMEM containing 1 mg/mL crystalline BSA, before seeding into collagen-coated 50-cm² tissue culture dishes, and the cells were preincubated in serum-free medium for 24 hours before exchange for media with and without growth factors. At the conclusion of the experiment, the media were removed and stored at −70°C, and the cell layers were either released with trypsin and counted electronically or lysed with Laemmli sample buffer without reducing agents. ³⁶ Culture media volumes evaluated were normalized to cell counts. Secreted proteins were concentrated by using a centrifugal filter device (Microcon YM-3; Millipore, Bedford, MA), combined with nonreducing Laemmli sample buffer, ³⁶ heated to 100°C for 3 minutes, and separated on 10% SDS-polyacrylamide gels. After electrophoresis at 20 mA, proteins were transferred to nitrocellulose membranes (Hybond ECL; Amersham Pharmacia Biotech) for 2 hours at 100 V and blocked with 3% BSA in Tris-buffered saline (TBS; 0.15 M NaCl and 0.02 M Tris-HCl [pH 7.8]) for 60 minutes at room temperature. IGFBPs were detected with 20 ng/mL biotinylated IGF-II (GroPep, Ltd.) in TBS with 0.05% Tween-20 (TBST) for 3 hours at room temperature. Bound IGF-II was detected with 10 ng/mL horseradish peroxidase–conjugated biotin-binding protein (NeutraAvidin; Pierce, Rockford, IL) in TBST for 60 minutes at room temperature. Chemiluminescence was developed (Super Signal West Femto; Pierce) according to the manufacturer’s instructions, and the membranes were exposed to autoradiograph film (Hyperfilm; Amersham). 

Western Blot Analysis

Culture media and cell lysates were prepared as for the ligand analysis but were reduced by the addition of 1.25% 2-mercaptoethanol in Laemmli sample buffer. After transfer, nitrocellulose membranes were blocked for 1 hour at room temperature with 3% nonfat dry milk in TBST. IGFBPs were detected with the following antibodies diluted in blocking solution: rabbit polyclonal anti-IGFBP-1 (1:2000) and anti-IGFBP-2 (1:2000) from Upstate Biotechnology, Inc. (Lake Placid, NY), rabbit polyclonal anti-IGFBP-3 (1:1000) and anti-IGFBP-6 (1:1000) from GroPep, Ltd., and mouse monoclonal anti-IGFBP-4 (1:500) and anti-IGFBP-5 (1:2000) from Austral Biologicals (San Ramon, CA). After a 2-hour incubation in primary antibody, blots were incubated for 1 hour with horseradish-peroxidase–conjugated donkey anti-rabbit IgG or goat anti-mouse IgG (Jackson ImmunoResearch, Inc., West Grove, PA) diluted in blocking buffer at 1:200,000. Chemiluminescent development was as described for Western ligand blots. Recombinant IGFBP-1, -3, and -5 were obtained from Upstate Biotechnology, Inc., IGFBP-2 from GroPep Ltd., and IGFBP-4 and -6 from Austral Biologicals. 

Results

IGFBP Biosynthesis and Müller Cell Phenotype

Müller cells were isolated, characterized, and maintained in continuous culture for 1 and 5 weeks, yielding phenotypes referred to as proliferative (GFAP positive, αSMA negative) and myofibroblastic (GFAP negative, αSMA positive). To evaluate changes in IGFBP biosynthesis associated with phenotype, total RNA harvested from these cultures and normal porcine retina were examined by RT-PCR with IGFBP-specific primers. Normal porcine liver, a source of all six high-affinity IGFBPs, served as a positive control for primer function. Amplimers of the predicted sizes were present in each reaction with liver RNA (Fig. 1) , confirming successful primer design. Although there was no evidence of IGFBP-1–specific message in normal retina or Müller cells of either phenotype, amplimers for IGFBP-2 through -6 were detected in all three RNA preparations. Northern blot analyses were performed on the same RNA preparation to evaluate transcript size and relative abundance. IGFBP-specific message abundance varied considerably in normal liver, but was consistent with production profiles described for fasting animals (Fig. 2) . ²⁴ Message levels for the five IGFBP species examined were undetectable or low in retina relative to G3PDH. There was little change between normal retina and proliferative Müller cells, except for modest increases in message levels for IGFBP-4, -5, and -6. However, the abundance of mRNA for IGFBP-2, -3, -4, and -6 was substantially increased in the myofibroblastic Müller cell phenotype, whereas expression of IGFBP-5 was reduced. Changes in IGFBP-specific message were confirmed in real-time RT-PCR reactions, using the same primer sets as in Figure 1 to evaluate, in this case, mRNA abundance between the culture phenotypes. IGFBP-2, -3, -4, and -6 message levels in myofibroblastic Müller cells increased from 6- to 70-fold compared with the proliferative phenotype, whereas IGFBP-5 message was reduced by approximately one half (Table 2) . 

IGFBP Biosynthesis and Growth Factor Stimuli

To evaluate the influence of Müller cell stimuli associated with fibrocontractive disorders, the effects of PDGF and IGF system ligands were examined in myofibroblastic Müller cells. To account for the potentially confounding effects of endogenous IGFBP production by Müller cells on IGF-I and -II activities, two analogues with reduced IGFBP affinity (R³IGF-I and R⁶IGF-II) were also examined. Control experiments performed using Müller cell extracellular matrix contraction as a biologically relevant assay indicated that growth factor analogue activities were comparable to that of the native ligands (Fig. 3) . Statistical analyses of the dose–response curves confirmed this observation (Table 3 , % contraction per mole), but also revealed that the analogues were more active than the native ligand at lower concentrations (Table 3 , C₅₀), which is consistent with modest IGFBP production by Müller cells under these conditions. 

Growth factor effects on IGFBP production were examined by incubating myofibroblastic Müller cell cultures in serum-free media with and without these ligands with the culture supernatants evaluated by Western ligand blot analysis. To account for the potentially confounding effects of growth factor-induced cell proliferation, cell densities were also evaluated at the end of the experiment, and the media volumes analyzed were normalized to these values. As a practical matter, cell densities under the different experimental conditions varied by <10%, indicating that little or no cell proliferation occurred in response to mitogens in serum-free medium (not shown). Müller cells incubated without exogenous growth factors produced at least three functional IGFBPs that migrated to positions of approximately 45, 35, and 22 kDa in nonreducing conditions (Fig. 4) . PDGF stimulation induced little change in this profile except for perhaps modest increases in the abundance of the 35- and 22-kDa species. In contrast, stimulation by either IGF-I or R³IGF-I induced substantial increases in all three binding proteins. IGF-II and R⁶IGF-II induced an intermediate response, with comparable increases in the 45- and 22-kDa species, but decreased production of the 35-kDa binding protein. 

To examine the differential effects of the IGF system ligands on production of specific IGFBPs, myofibroblastic Müller cells were incubated with R³IGF-I or R⁶IGF-II or without growth factor stimuli and Müller-cell–conditioned media, and lysates were evaluated in Western blot analysis under reducing conditions. The potential for antibody cross-reactivity between IGFBP species was evaluated in Western blot analysis of recombinant IGFBP mixtures and, in all cases, the primary antibodies detected proteins at relative positions consistent with the reported sizes of IGFBP monomers and/or dimers (Fig. 5) . Western blot analysis of conditioned media revealed that R³IGF-I induced increases in Müller-cell–secreted IGFBP-2, -4, -5, and -6, whereas IGFBP-3 levels appeared to be unchanged from the unstimulated cells (Fig. 6) . Similar results were obtained with R⁶IGF-II–stimulated cells, except that IGFBP-2 levels appeared to decrease, rather than increase. However, Western blot analysis of comparable cell equivalents of lysate from the same experiment indicated that R⁶IGF-II induced increased IGFBP-2 retention by the cells, and this effect, rather than decreased production, most likely accounts for the reduced levels in conditioned media. RNA preparations from these cultures were also evaluated by Northern blot and growth-factor–induced increases in the abundance of full-length transcripts were evident for IGFBP-2, -4, -5, and -6 (Fig. 7) . In contrast, there appeared to be a slight decrease in the abundance of the IGFBP-3 relative to the control G3PDH. Real-time RT-PCR evaluations of these samples confirmed these trends, indicating two- to ninefold increases in specific message for IGFBP-2, -4, -5, and -6, with a modest reduction in IGFBP-3 (Table 4) . 

Discussion

The primary goal of this study was to evaluate Müller cell potential to synthesize and secrete growth-factor–modulating IGFBPs and, in this, the results were relatively straightforward. Notwithstanding changes associated with phenotype and the influences of growth factors, this study provided evidence that Müller cells produce five of the six high-affinity IGFBPs. On the surface, these data provide impetus for studies to explore the potential of Müller cell to modulate vitreous IGFBP content in fibroproliferative disorders involving this cell type. Absent evidence of IGFBP-1 production by other relevant ocular tissue such as the ciliary body, these results also support the premise that vitreous IGFBP-1 detected in advanced diabetic retinopathy originates from plasma and that vitreous IGFBP content is either influenced by local production or is, at the least, under local control. ^{31 37}  

This study also provided evidence that the relationship between Müller cell phenotype and IGFBP production can be profound. RT-PCR evaluations of normal retina served as an indicator of basal IGFBP biosynthesis by normal Müller cells in addition to other retinal cells. Although the positive results obtained cannot be ascribed to a particular cell type, conclusions regarding the absence of detectable message by this highly sensitive approach are applicable to Müller cells in addition to the other cells in the retina. These data, combined with the Northern blot analyses revealed significant message levels for IGFBP-3 and -5. The findings with IGFBP-3 are consistent with previous studies in which similar approaches and retinas from other species were used. ²³ Similarly, at least one study reported IGFBP-5 expression in the developing retina, ³⁴ an observation that can now be extended to adult tissues. Although the IGFBP biosynthetic profile of proliferative Müller cells was similar to normal retina with respect to IGFBP species-specific message, the most interesting result was the rather dramatic change in myofibroblastic Müller cells, with significant increases in message levels for IGFBP-2, -3, -4, and -6 and a modest reduction in IGFBP-5 message. Similarly compelling was evidence of increased IGFBP production by myofibroblastic Müller cells in response to growth factors known to be active in diabetic vitreous. Although still somewhat speculative, the most direct interpretation of these results is that myofibroblastic Müller cells produce IGFBPs and potentially modulate vitreous IGFBP content in disease. 

As described in the introduction, the presence of Müller cells in diabetic fibrocontractive tissues has been reported by several laboratories. ^{9 10 11 12} Identification, in most cases, was based on immunocytochemical detected of the intermediate filament protein GFAP which, based on our observations of Müller cell phenotype plasticity in culture, suggests that these cells would represent the proliferative rather than myofibroblastic phenotype. However, recent examinations of epiretinal tissues from patients with PDR, by using dual-label indirect immunofluorescence to localize GFAP and αSMA, revealed cells consistent with proliferative and myofibroblastic phenotypes and, most important, a transitional region in which both antigens could be localized to the same cells. ³⁸  

In a recent study from this laboratory, the effects of the six high-affinity IGFBPs on Müller cell responses to IGF-I and -II were systematically evaluated, and the most significant finding was that IGFBP-3, a predominate binding protein in normal vitreous, is an extremely potent inhibitor of IGF-I and -II with effects measurable at concentrations equimolar to the growth factors. ³⁹ IGFBP-2, another abundant binding protein in vitreous, are three to four times less active than IGFBP-3, but is nonetheless a potent inhibitor of the effects of IGF-I and -II. The inhibitory activities of the remaining IGFBPs were an order of magnitude or more less effective against IGF-I, but more effectively attenuated IGF-II activity. Assuming that myofibroblastic Müller cells supplement the vitreous environment, the predominant effect should be to increase vitreous growth-factor–binding capacity and increase attenuation of free growth factor, preferentially sequestering IGF-II over -I. A biosynthetic mechanism to self-limit IGF-induced effects would also be consistent with the observation that free IGFs and not PDGF enhance production of four of the five Müller-cell–synthesized IGFBPs. These data also suggest that IGF-I, rather than -II, is most likely the predominant free, and thus biologically active, growth factor in vitreous and places even greater mechanistic relevance on the role of the two IGFBPs (-2 and -3) with the highest capacity to control the actions of this growth factor. 

 Supported by the Juvenile Diabetes Research Foundation International, the International Retinal Research Foundation (Birmingham, AL), National Eye Institute Grant EY13258, and Research to Prevent Blindness.

 Submitted for publication April 19, 2004; revised July 13, 2004; accepted August 13, 2004.

 Disclosure: J.L. King, None; C. Guidry, None

 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.

 

Table 1.

 

Primer Pairs Developed for Analysis of IGFBP Expression

IGFBP	Product Size (bp)	Accession Number	Primer Sequence
1	502	AB053605	fwd-acagactagccagggagcag
			rev-gcaggacctctgaagagcaa
2	453	AF120326	fwd-cgagcaggttgcagacaatgg
			rev-tggatcagcttcccggtgttg
3	467	AF085482	fwd-ccaggaaacggcagtgagtcc
			rev-tccatgctgtagcagtgcacg
4	258	NM001552	fwd-ctgacaaggacgagggtga
			rev-ggatggggatgatgtaaagg
5	428	U41340	fwd-ggagacctactcgcccaaga
			rev-tcaacgttgctgctgtcgaagc
6	221	AF327653	fwd-caaggagagtaagccccaag
			rev-acaattgggcacgtaga

 fwd, forward; rev, reverse.

Figure 1.

 

Evaluation of IGFBP expression in retina and Müller cell cultures by RT-PCR. RNA preparations from porcine liver (lanes 1 and 5), retina (lanes 2 and 6), and proliferative (lanes 3 and 7) and myofibroblastic Müller cells were evaluated by RT-PCR with IGFBP-specific primers (lanes 1–4) and, as a reaction control, G3PDH (lanes 5–8). DNA standards are in the left lane of each panel.

Figure 2.

 

Evaluation of IGFBP expression by Northern blot analyses. RNA preparations from liver (lane 1), retina (lane 2) and proliferative (lane 3) and myofibroblastic (lane 4) Müller cells were separated on agarose gels, transferred to membranous supports and probed for the presence of IGFBP-specific mRNA, using ³²P-labeled cDNA probes. The positions of the 28S (4.7 kb) and 18S (1.9 kb) ribosomal RNA subunits are shown at the left.

Table 2.

 

Summary of IGFBP Message Abundance in Proliferative and Myofibroblastic Müller Cells

IGFBP	Proliferative vs. Myofibroblastic ΔΔCT ± SD (x-fold increase)
2	6.30 ± 0.33 (78.8)
3	4.66 ± 0.20 (25.3)
4	2.61 ± 0.39 (6.1)
5	−1.11 ± 0.31 (0.5)
6	3.56 ± 0.27 (11.8)

Figure 3.

 

Müller cell responses to IGF analogues. Myofibroblastic Müller cells attached to collagen gels were incubated in the indicated concentrations of IGF-I, IGF-II, and analogues with reduced affinities for IGFBPs, including R³IGF-I and R⁶IGF-II. Presented are the dose–response profiles obtained after 24 hours of incubations, representing the means and standard deviations obtained from triplicate cultures under each condition.

Table 3.

 

Summary of IGF Ligand Activities

	% Contraction per Mole	r corr	C 50
IGF-I	3.07 × 1011	0.99	6.01 × 10−11
R3-IGF-I	2.94 × 1011	0.91	3.20 × 10−11
IGF-II	3.31 × 1010	0.92	5.01 × 10−10
R6-IGF-II	3.11 × 1010	0.94	3.39 × 10−10

 % Contraction per Mole is calculated from regression analysis of 24-hour dose–response curves normalized in internal positive and negative controls, using a minimum of three concentrations, each in triplicate. C ₅₀ is the calculated concentration of ligand necessary to induce half-maximum response.

Figure 4.

 

Müller cell secretion of functional IGFBPs. Myofibroblastic Müller cells in serum-free medium were incubated for 72 hours with IGF-I (lane 1), R³IGF-I (lane 2), IGF-II (lane 3), R⁶IGF-II (lane 4), no additions (lane 5) or PDGF (lane 6) after which the media were harvested and analyzed by Western ligand blot analysis using biotinylated IGF-II. The positions of molecular weight standards are shown at the left.

Figure 5.

 

Evaluation of anti-IGFBP specificity. Mixtures of the six recombinant high-affinity IGFBPs were separated by electrophoresis, transferred to membranous supports, and probed in Western blot analysis with polyclonal antibodies raised against IGFBP types 1 through 6 (lanes 1–6, respectively). The positions of prestained molecular weight standards are shown at the left.

Figure 6.

 

Evaluation of Müller cell IGFBP production by Western ligand blot analysis. Myofibroblastic Müller cells in culture were incubated for 72 hours without additions (lanes 1 and 4) or with the addition of R³IGF-I (lanes 2 and 5) or R⁶IGF-II (lanes 3 and 6). Control media (lane C), Müller-cell–conditioned media (lanes 1–3) and detergent-extracted proteins from the cell layer (lanes 4–6) were probed in Western blot analysis, using the IGFBP-specific antibodies shown at left. The positions of prestained molecular weight standards are indicated between the two panels.

Figure 7.

 

Evaluation of Müller cell IGFBP expression by Northern blot analysis. RNA preparations from normal porcine liver (lane 1) and myofibroblastic Müller cells exposed to media without additions (lane 2) or with the addition of R³IGF-I (lane 3) or R⁶IGF-II (lane 4) were separated on agarose gels, transferred to membranous supports and probed for the IGFBP-specific mRNA indicated using ³²P-labeled cDNA probes. The positions of the 28S or 18S ribosomal RNA subunits are shown at the left.

Table 4.

 

Summary of IGFBP Message Abundance in Response to Growth Factor Stimulation

IGFBP	Unstimulated vs. R3IGF-I ΔΔCT ± SD (x-fold increase)	Unstimulated vs. R6IGF-II ΔΔCT ± SD (x-fold increase)
2	3.25 ± 0.53 (9.5)	4.07 ± 0.44 (16.8)
3	0.16 ± 0.40 (0.9)	0.35 ± 0.06 (0.8)
4	1.70 ± 0.18 (3.3)	2.09 ± 0.34 (4.3)
5	1.81 ± 0.18 (3.5)	1.09 ± 0.19 (2.1)
6	3.20 ± 0.97 (9.2)	3.32 ± 1.30 (9.9)