Expression of LRP1 in Retinal Pigment Epithelial Cells and Its Regulation by Growth Factors

Abstract

purpose. The retinal pigment epithelial (RPE) cells are mitotically inactive under normal conditions, but play a pivotal role in the pathogenesis of proliferative vitreoretinopathy (PVR). Triggered by changes in the concentrations of growth factors, RPE cells reenter the cell cycle, proliferate, and migrate onto the retinal surface, into the subretinal space, and into the vitreous. The receptor for α₂-macroglobulin (low-density lipoprotein receptor–related protein [LRP1], or CD91) is known to be involved in the processes of cell migration and invasion, as well as in the regulation of growth factor homeostasis. The purpose of this study was to investigate the expression of this receptor and its regulation, at the protein and mRNA levels, in human (h)RPE cells.

methods. The cell surface expression of the receptor was studied by immunocytochemistry and flow cytometry. The endocytosis-related activity of LRP1 in hRPE cells was examined by assessing the uptake of FITC-labeled, methylamine (MA)-treated α₂-M (α₂-M-MA). LRP1 mRNA expression was analyzed by means of the RNase protection assay (RPA) after the hRPE cells were stimulated with the growth factors TGF-β1, TGF-β2, PDGF, VEGF (each 10 ng/mL), or bFGF (5 ng/mL).

results. hRPE cells expressed LRP1 on their cell surface. The receptor mediated rapid binding and endocytosis of FITC-labeled α₂-M-MA. The expression of LRP1 mRNA strongly increased on stimulation of the cells with TGF-β1, TGF-β2, or VEGF, whereas PDGF or bFGF elicited only minor effects.

conclusions. The expression of functionally active LRP1 in hRPE cells suggests that the receptor may be involved in cell migration and invasion, as reported for other LRP1-expressing cells. Thus, certain growth factors may control RPE cell migration and invasion in vivo through a regulation of LRP1 expression. As LRP1 mediates the clearance of α₂-M, known to regulate the homeostasis of many cytokines and growth factors, this receptor may be a promising target for therapeutic intervention in PVR.

Proliferative vitreoretinopathy (PVR) is the main complication after retinal injury or retinal detachment surgery. This ocular disorder is characterized by a proliferation and migration of retinal cells onto the retinal surface, into the subretinal space, and/or into the vitreous. These processes are assumed to be induced by growth factors, such as transforming growth factor (TGF)-β, platelet derived growth factor (PDGF), basic fibroblast growth factor (bFGF), hepatocyte growth factor (HGF), and vascular endothelial growth factor (VEGF), which are released through a break in the blood–retinal barrier and are also secreted by stimulated retinal cells. ^{1 2 3 4 5 6 7}  

An important regulator of growth factor homeostasis is the protease inhibitor, α₂-macro-globulin (α₂-M). The binding of proteinases causes a change of its conformation (transformation) which promotes the binding of several growth factors and cytokines. ^{8 9 10} In fact, α₂-M has been found to enhance the clearance of these biologically active polypeptides from the circulation. The systemic application of proteases (known as enzyme or protease therapy) seems to ameliorate this process. ¹¹  

The clearance of α₂-M–bound cytokines from the circulation is mediated by the receptor for α₂-M, which is identical with the low-density lipoprotein receptor–related protein (LRP1 or CD91). LRP1 is a large, noncovalently associated heterodimeric membrane protein synthesized as a 600-kDa single polypeptide chain and cleaved in the trans-Golgi compartment into the 515-kDa α- and the 85-kDa β-subunit. ^{12 13 14} LRP1 is a glycoprotein in which the large α-subunit forms the extracellular domain and the smaller β-subunit forms the transmembrane domain and the intracellular tail. The receptor is expressed in many cell types, including hepatocytes, keratinocytes, fibroblasts, macrophages, Müller glial cells, mammary epithelial cells, and others. ^{15 16 17 18 19 20}  

In addition to α₂-M, other molecules such as lipoproteins, ²¹ lipoprotein lipase, ²² coagulation factor VIII, ²³ heat shock protein gp96, ²⁴ or the β-amyloid precursor protein ²⁵ are known to bind to the LRP1, as well. After binding to LRP1, the complexes are forwarded to the endosomal–cytosolic compartment of the cells by receptor-mediated endocytosis. 

The signal for LRP1 endocytosis is a YXXL motif located within the cytoplasmic domain. ²⁶ LRP1 displays an extremely fast endocytosis rate, consistent with its function as a clearance receptor. ²⁷ Therefore, LRP1 is suggested to play an important role in the clearance of α₂-M-associated growth factors and, thus, in the control of the pathologic events in PVR, where it may (partially) inhibit abnormal cell proliferation. 

In the present study, we investigated the expression of the LRP1, at both the mRNA and protein levels, in hRPE cells in vitro and in vivo. Furthermore, we examined the effects on LRP1 mRNA expression of various growth factors thought to be involved in the pathogenesis of PVR. 

Materials and Methods

Cell Cultures

Human RPE cells were obtained from several donors within 48 hours of death and were prepared as described previously. ²⁸ The use of human tissue was approved by the Ethics Committee of the University of Leipzig, and the study was performed according to the Declaration of Helsinki. After preparation, the hRPE cells were suspended in complete F-10 (Ham) medium with l-glutamine (GlutaMAX I; Invitrogen-Gibco, Paisley, UK) containing 10% fetal calf serum (FCS), GlutaMAX II (Invitrogen-Gibco), and gentamicin and were cultured in tissue culture flasks (Greiner, Nürtingen, Germany) in 5% CO₂ at 37°C. The epithelial nature of the RPE cells was routinely verified by immunohistochemical staining, using the monoclonal antibodies AE1 (recognizing most of the acidic [type I] keratins) and AE3 (recognizing most of the basic [type II] keratins). Both were obtained from Chemicon (Hofheim, Germany). The human fibroblasts were cultured in the same medium. All tissue culture components and solutions were purchased from Invitrogen-Gibco. 

Retinal Tissues

Samples of retinas were obtained from patients undergoing vitreoretinal surgery for the treatment of retinal detachment complicated by PVR. All patients gave their informed consent before inclusion in the study. The stage of PVR was estimated according to the updated classification of the Retina Society. ²⁹ The PVR samples were from patients with PVR grade CP4, CA4, or above, and they were mostly circumferential. During the vitreoretinal surgery, the PVR membranes were peeled away as much as possible and then the required retinectomy was performed. Retinal samples were taken from seven patients (four women, three men; mean age, 63.14 ± 9.77 years; range, 51–71 years). The tissue was promptly removed from the vitrectomy waste after surgery, immediately precipitated by centrifugation at 4°C, and washed twice with phosphate-buffered saline (PBS) to remove blood cells. The tissue was then used directly for total RNA preparation. 

Retinas obtained from the eyes of cornea donors prepared within 24 hours of death were used as normal control tissues. Control retinas were taken from seven donors (three women, four men; mean age, 66.28 ± 12.18 years; range, 46–86 years). 

Activation of Human RPE Cells with Different Cytokines

Cells were cultured in 250-mL culture flasks until 80% to 90% confluence was achieved. The cells were washed twice with PBS and growth arrested in fresh medium without FCS for approximately 24 hours. Thereafter, the medium was exchanged by fresh medium supplemented with recombinant human (rh)TGF-β1 (10 ng/mL), rhTGF-β2 (10 ng/mL), rhPDGF-BB (10 ng/mL), rhVEGF (10 ng/mL), or rhbFGF (5 ng/mL), respectively. The recombinant human cytokines were purchased from R&D Systems (Wiesbaden, Germany). One flask served as the unstimulated 0-hour control. Stimulation of the cells was continued for 1 and 24 hours, respectively. Thereafter, the cells were washed twice and used for subsequent analysis. 

Total RNA Preparation

Treated and untreated hRPE cells and human fibroblasts were trypsinized and collected by centrifugation, and the supernatants were aspirated completely. The cells and the tissue samples (normal and PVR retinas) were lysed, and total RNA was prepared (RNeasy-Mini Kit; Qiagen, Hilden, Germany). The eluted RNA was incubated for 10 minutes at 60°C; quantities were determined from OD₂₆₀ (Genequant_pro; Amersham Biosciences, Freiburg, Germany). 

Reverse Transcription–Polymerase Chain Reaction

The complementary (c)DNA was synthesized from 2 μg total RNA with a kit (First-Strand cDNA Synthesis for RT-PCR; Roche, Mannheim, Germany). A kit was used for PCR (Taq PCR Master Mix Kit; Qiagen). Two microliters of the first-strand mixture and 1 μm of each gene-specific sense and antisense primer were used for the amplification reaction in a final volume of 50 μL. Primer pairs were LRP1 (accession no. NM_002332) sense 5′-ACCACCCCTCCCGCCAGCCCA-3′ and antisense 5′-AGCCCGA-GCCGTCGCCTTGC-3′ producing a 876 bp amplicon; glyceraldehyde-3-phosphate dehydrogenase (GAPDH: accession no. M33197) sense 5′-GCAGGGGGGAGCCAAAAGGGT-3′ and antisense 5′-TGGGTGGCAGTGATGGCATGG-3′, producing a 215-bp amplicon. Thermocycling was then performed (PTC-200 Thermal Cycler; MJ Research, Watertown, MA). The following PCR cycle parameters were used: denaturation at 94°C for 3 minutes, followed by 32 cycles of denaturation at 94°C for 30 seconds, annealing at 64°C for 1 minute, and polymerization at 72°C for 2 minutes. A final extension step was performed at 72°C for 10 minutes. The LRP1 primer pair spanned an intron to ensure that the RT-PCR product was derived from RNA and not from genomic DNA. Oligonucleotide primers were obtained from MWG-Biotech (Ebersberg, Germany). The amplified products were separated on a 1.6% agarose gel including 10 ng/mL ethidium bromide and were visualized using the ultraviolet transilluminator of the imager (Fluor-S-Imager; BioRad, Munich, Germany). 

Immunocytochemistry

Human RPE cells grown on chamber slides and frozen tissue sections of human eyes were fixed with methanol (9 minutes, −20°C) and acetone (1 minute, −20°C). After air drying of the slides, blocking was performed with 5% normal goat serum in Tris-buffered saline (TBS) for 30 minutes. The slides were then incubated with the primary monoclonal antibody (mAb) directed against the α-chain of the LRP1 (1:10 dilution; clone 02-03; BioMac, Leipzig, Germany) at 4°C for 12 hours. The control was performed by omitting the primary mAb. After they were washed with TBS, the slides were incubated with a CY3-labeled polyclonal goat anti-mouse antibody (Dako, Hamburg, Germany) for 60 minutes. After they were washed with TBS (three times), the cell nuclei were counterstained with Hoechst 33258. The fluorescence was measured by microscope (Axiolab; Carl Zeiss Meditec, Jena, Germany). Tissue sections were analyzed with a laser scanning microscope (model LSM 510; Carl Zeiss Meditec). 

Flow Cytometry

Trypsinized hRPE cells were fixed with ice-cold methanol for 10 minutes at 4°C and washed twice with PBS. The cells were labeled with anti-LRP1 antibody (0.2 mg/mL in PBS+0.3% Tween-20) for 1 hour at 4°C. For the negative control, the mAb was substituted by PBS. After two washes with PBS, the cells were incubated with a secondary fluorescein isothiocyanate (FITC)-labeled antibody for 45 minutes. Flow cytometry was then performed (FACSCalibur; BD Biosciences, Franklin Lakes, NJ). 

Uptake of FITC-Labeled α₂-M-MA

Human RPE cells were cultured in eight-well chamber slides (Nunc Inc., Napierville, IL) in 5% CO₂ at 37°C until 80% to 90% confluence was achieved. Then, 100 nM of methylamine (MA)-treated FITC-labeled α₂-M (BioMac) was added to the cells after different time intervals. One well served as the untreated control. The uptake was documented by microscope (Axiolab; Carl Zeiss Meditec). 

RNase Protection Assay

The RNase protection assays (RPAs) were performed with a multiprobe kit (RiboQuant; BD-PharMingen, San Diego). The probe set contained templates for LRP1 and GAPDH. The GAPDH probe was used as the internal standard. The probe for LRP1 was performed by subcloning the PCR fragment described earlier into a plasmid (pGEM-T; Promega). The sequence of the inserted fragment was then checked by sequencing. After amplification and preparation, the plasmid was digested with restriction endonuclease ClaI (Roche). The fragments were separated by electrophoresis, and the band containing the LRP1 cDNA was extracted from the gel by gel purification (GFX PCR and Gel Band Purification Kit; Amersham Biosciences). The template set was used for the T7 RNA polymerase-directed synthesis of a specific ³²P-labeled antisense riboprobe set, and RPA was performed as described previously. ³⁰ The relative quantity of LRP1 mRNA was determined based on the signal intensity of the protected probe fragment bands using software from (Multi-Analyst; BioRad). Data were normalized to the GAPDH signal. All chemicals for gel electrophoresis were purchased from Roth (Karlsruhe, Germany). 

Statistical Analysis

Individual experiments were performed with three different RPE cell lines. The values were expressed as the mean ± SD and the results were compared using one-way analysis of variance (ANOVA) and were considered to be statistically significant if P ≤ 0.05. 

Results

Detection of Receptor mRNA by RT-PCR

Figure 1 demonstrates the presence of specific LRP1 amplicons of the calculated size of 876 bp, obtained by RT-PCR of hRPE-mRNA. In addition, cDNA synthesized from total RNA of human fibroblasts was used as the positive control because these cells are known to express abundant amounts of the LRP1. ¹⁵ The RT-PCR for GAPDH amplification was performed as the control for the amount and quality of cDNA. 

Immunochemistry

For immunochemical staining, both cultured hRPE cells and frozen tissue sections were incubated with the anti-LRP1 antibody directed against the α-subunit of the receptor. Intensive immunoreactivity was observed in cultured hRPE cells (Fig. 2) as well as in hRPE cells of the tissue samples (Fig. 3B) . The corresponding location of the RPE cells is clearly visible by their pigmentation in the accompanying light micrograph (Fig. 3A , arrow). Immunoreactivity was also observed in adjacent regions such as the choroid or sclera. This response is caused by cells (such as fibroblasts) located in this region. All controls (prepared by omitting the primary antibody) were devoid of fluorescence signals, indicating a specific binding of the monoclonal antibody. 

Flow Cytometry

To confirm these results, hRPE cells were analyzed by flow cytometry. A representative example is shown in Figures 4A and 4B . The binding of the receptor-specific mAb was followed up with an FITC-labeled goat anti-mouse Ig as the secondary antibody. Control experiments were performed with the primary antibody omitted. More than 96.84% ± 0.67% (n = 3) of the counted cells were found to express the receptor protein at the cell surface, with a mean fluorescence intensity of 51.67 ± 11.06 counts. 

For investigating the influence of different cytokines on the LRP1 protein expression, flow cytometry experiments were performed with growth-factor–stimulated hRPE cells. No influence on the protein-expression was obtained after a 1-hour stimulation (data not shown). A slightly but significantly elevated amount of LRP1 protein was present after 24 hours of modulation with the TGF-β isoforms (Fig. 4C) . VEGF and PDGF caused a minor but not significant effect on LRP1 protein synthesis. No change was detected in cells treated with bFGF. 

Uptake of FITC-Labeled α₂-M-MA by hRPE Cells

The hRPE cells incorporated labeled-transformed α₂-M (α₂-M-MA). The binding and uptake of the ligand proceeded very fast. An intensive intracellular fluorescence was observed in the cells in as early as 10 minutes of incubation with FITC-labeled α₂-M-MA (Fig. 5) . As expected, no fluorescence signal was detectable in the controls (i.e., in the absence of labeled ligand). 

Effects of Cytokines on LRP1 mRNA Expression of hRPE Cells

RPA was used to investigate the impact of specific growth factors on LRP1 mRNA expression (Fig. 6) . Occasionally, more than one protected band was observed after phosphorimaging. This is probably caused by a degradation of mRNA which naturally occurs in cells during processing. For the evaluation, radioactivity of all bands was used. The values were normalized to the expression level of the housekeeping gene GAPDH and were quantified as “fold increase” in mRNA-signal compared to the unstimulated 0-hour control. LRP1 mRNA was constitutively expressed, as shown in the nonstimulated 0-hour control sample (Fig. 6A) . Incubation of hRPE cells in the presence of TGF-β1 or -β2 caused the strongest upregulation of LRP1 mRNA expression. A significantly elevated amount of mRNA was observed as soon as 1 hour (1.48 ± 0.21- and 1.56 ± 0.31-fold, respectively) and was even more prominent after a 24-hour stimulation (2.01 ± 0.71- and 1.77 ± 0.3-fold, respectively). In addition to the TGF-β isoforms, VEGF also was found to stimulate LRP1 mRNA expression, resulting in a 1.30 ± 0.18-fold increase after 1 hour and a 1.62 ± 0.2-fold increase after 24 hours. PDGF was less effective, and no stimulation of expression was seen with bFGF. 

In addition, the expression of the LRP-1 mRNA in human retinal tissues was investigated. A slightly but significant upregulation (1.53 ± 0.20-fold) of LRP-1 mRNA was observed in the PVR retinas compared with normal retinas (Fig. 7) . 

Discussion

PVR is associated with dramatic changes of the cytokine milieu in both subretinal fluid and vitreous. ³¹ These changes are mainly caused by the breakdown of the blood–retinal barrier, which results in a uncontrolled cancerlike proliferation and migration of retinal cells, such as RPE and glial cells, and in an enhanced production of structure proteins including collagens, thrombospondin, or vitronectin. ^{1 32 33 34 35 36} It has been suggested that the migrating cells themselves may also constitute sources of these growth factors. ^{37 38 39}  

There is now good evidence for the assumption that hRPE cells play a central role in the development of PVR. ^{34 40} RPE cells are located on Bruch’s membrane between the choriocapillaris and the sensory retina. In adult eyes, these cells are well-differentiated and mitotically silent. However, if RPE cells become exposed to the vitreous under pathologic conditions, they may reenter the cell cycle and initiate proliferation, migration, and dedifferentiation. These responses include the increased formation of extracellular matrix proteins. ^{41 42}  

In the current study, we show that cultured hRPE cells express both LRP1 mRNA and LRP1 protein, similar to earlier findings in other cell types such as keratinocytes and fibroblasts. ¹⁵ The intense immunohistochemical demonstration of LRP1 in hRPE cells of frozen tissue and the detection of LRP1 mRNA in freshly isolated hRPE cells confirms that this receptor also is expressed by the cells in vivo. 

As cytokines are thought to be involved in the initiation of PVR, we studied the effect of various growth factors on LRP1 mRNA expression. The results demonstrated a significant upregulation of LRP1 mRNA after stimulation of hRPE cells with the TGF-β isoforms 1 and 2 or with VEGF as soon as after 1 hour. A further increase was obtained after 24 hours. A similar increase in the mRNA levels of LRP1 has been reported for nerve growth factor–stimulated neuronal cell lines and for dexamethasone-exposed HepG2 cells. ^{43 44} The increased LRP1 expression in these cells was accompanied by an enhanced binding and degradation of activated α₂-M. Treatment of vascular smooth muscle cells with PDGF-BB or EGF also induced an increased cellular binding of activated α₂-M; however, in this case no effect on the levels of LRP1 mRNA and protein was found. ⁴⁵ Similar results were observed if 3T3-L1 adipocytes were exposed to insulin. This effect was explained by a stimulated LRP-recycling on the cell surface. ⁴⁶ Thus, growth factors and cytokines may regulate LRP1 expression at different, transcriptional as well as posttranscriptional levels. Our results indicate that hRPE cells respond to at least three growth factors (TGF-β1, TGF-β2, and VEGF) by an upregulated LRP1 transcription. The corresponding changes of the LRP1 protein expression were rather small within 24 hours of stimulation, however. Thus, it could be that these growth factors regulate the endocytosis rate rather than the total protein level of LRP1, as is the case with NGF. ⁴³  

The link between LRP1 and cytokine clearance is constituted by α₂-M. We suggest that by binding to the receptor on hRPE cells, transformed α₂-M may act as a vehicle for the clearance of pathologically elevated growth factors, similar to that discussed for many other cell types and tissues. ^{8 47} In agreement with this idea, Schulz et al. ¹⁰ demonstrated the presence of α₂-M in bovine and human aqueous and vitreous and showed that the presence of this molecule (1) provides a protection of lens cells against the damaging effects of TGF-β and (2) inhibits cataractous alterations. Furthermore, it was demonstrated that TGF-β–induced collagen synthesis by human liver myofibroblasts is reduced by activated α₂-M. ⁴⁸ This suggests that triggering the clearance function of α₂-M in vivo may help to prevent the development of PVR. Complex formation of α₂-M with growth factors in vivo is mediated by the reaction of proteases with α₂-M. ⁴⁹ Based on this knowledge, proteases have been systemically administered for the treatment of cancer and fibrosis, ^{50 51} as well as to ameliorate immune-mediated diseases. ⁵² In analogy, protease treatment should be considered as a potential therapy in PVR, pushing a stimulated α₂-M–mediated clearance of growth factors ^{17 53 54 55} against an increased expression of LRP1 by RPE cells. In a more general view, protease treatment may become a novel strategy to prevent harmful systemic or local effects of excess cytokines in vivo. 

 Supported by Deutsche Forschungsgemeinschaft Grant Ko 1547/4-1.

 Submitted for publication June 26, 2003; revised November 12, 2003; accepted December 2, 2003.

 Disclosure: M. Hollborn, None; G. Birkenmeier, None; A. Saalbach, None; I. Iandiev, None; A. Reichenbach, None; P. Wiedemann, None; L. Kohen, 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.

 

Figure 1.

 

Detection of LRP1 mRNA by RT-PCR. RT-PCR was performed, and the resultant RT-PCR products were separated in 1.6% agarose gels and were analyzed under ultraviolet light. RT-PCR results using total RNA of cultured hRPE cells (left) and RT-PCR results using total RNA of freshly isolated hRPE cells (right). A 100-bp molecular weight marker was used for size estimation. The RT-PCR of GAPDH mRNA was performed as a control to check the amount and quality of cDNA. The negative control was performed by omitting the cDNA, and total RNA of human fibroblasts was used as the positive control.

Figure 2.

 

Immunocytochemical labeling of LRP1 in cultured hRPE cells. (A) Control experiments were performed without the first antibody. (B) Cell nuclei counterstained with Hoechst 33258. (C) Immunofluorescence labeling of LRP1 in hRPE cells. Magnification, ×400.

Figure 3.

 

Immunohistochemical labeling of LRP1 in human ocular tissue analyzed by laser-scanning microscopy. In this tissue sample, the sensory retina is not included, because only the region around the RPE cell layer was of interest. (A) Transmitted light microscopic image. Arrows: location of the pigmented layer within the tissue sample. (B) Immunofluorescence labeling of LRP1 protein in the sample. (C) Control (omitting the primary antibody). The calibration bar is valid for all panels.

Figure 4.

 

Flow cytometry analysis of hRPE cells using an anti-LRP1 antibody. (A) Negative control, without antibody. (B) Flow cytometry diagram using the anti-LRP1 antibody as the primary antibody and FITC-labeled anti-mouse Ig as the secondary one. (C) Illustration of the changes of LRP1 protein using hRPE cells stimulated with different cytokines for 24 hours. Data are expressed as a percentage of the untreated control (100%). Data are the mean ± SD of results in four independent experiments. *P < 0.05 versus untreated control.

Figure 5.

 

Uptake of FITC-labeled α₂-M-MA by hRPE cells. Cells were incubated in the presence of 100 nM FITC-labeled α₂-M-MA for 10 minutes and immediately subjected to fluorescence detection. (A) Untreated control cells; (B) counterstaining of the untreated control cells using Hoechst 33258; (C) incubation of hRPE cells with FITC-labeled α₂-M-MA. Magnification, ×400.

Figure 6.

 

Effects of growth factors on LRP1 mRNA expression in hRPE cells. (A) RPA using 5 μg total RNA for the hybridization with gene-specific anti-sense RNA probes. (B) Time- and stimulation-dependent modulation of LRP1 mRNA expression after normalization to the GAPDH signal. Each column represents the mean ± SD of results in three experiments using hRPE cells from different donors. Significant with *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001.

Figure 7.

 

LRP1 mRNA expression in normal and PVR retinas. (A) Representative RPA using 5 μg total RNA for the hybridization with gene-specific anti-sense RNA probes. (B) LRP1 mRNA expression signal after normalization to the GAPDH signal. Each column represents the mean ± SD. Significant at *P ≤ 0.05.

The authors thank Grit Müller for her excellent technical assistance and to Bernd Biedermann for the realization of the laser scanning microscopy.