京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第19561号 / 医博第4068号 / 新制||医||1013(附属図書館) / 32597 / 京都大学大学院医学研究科医学専攻 / (主査)教授 吉村 長久, 教授 戸口田 淳也, 教授 高橋 淳 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第21002号 / 医博第4348号 / 新制||医||1027(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 大森 孝一, 教授 山田 亮, 教授 Shohab YOUSSEFIAN / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

The retinal pigment epithelium (RPE) is a monolayer of cells located directly behind the photoreceptor cells in the retina. These cells are involved in a variety of functions that support the visual process in the eye, namely 1) they form a blood-retina barrier which separates the neural retina from the choroid's blood supply, 2) the apical processes of RPE cells diurnally phagocytose the outer segments of photoreceptor cells, and 3) they participate in the renewal of the photopigment 11-cis retinal. Age-related macular degneration (AMD) is the leading cause of blindness in people over the age of 50 years in North America and other developed countries. AMD involves the death of retinal pigment epithelial (RPE) cells in the macula early in the progress of the disease. Like some other postmitotic cells, the RPE accumulates autofluorescent lysosomal storage bodies (lipofuscin) during senescence. Lipofuscin is reported to begin accumulating in the human RPE around age 20 and continues to accumulate throughout an individual's life. This progressive accumulation of lipofuscin can eventually occupy a substantial fraction of the RPE cytoplasmic volume and may lead to impairment of normal RPE functions, resulting in retinal degeneration and loss of visual function as in AMD. Another autofluorescent granule that accumulates in RPE cells and may contribute to the etiology of AMD is a complex granule exhibiting properties of both melanosomes and lipofuscin granules called melanolipofuscin (MLF). In contrast with the accumulation of LF in the RPE, MLF accumulation has been reported by Feeney-Burns to more closely reflect the onset of AMD. Although there have been significant advances in our understanding of AMD, knowledge of the mechanisms responsible for its progression remain unclear. This dissertation details experiments that were designed to better understand the factors that may play a causal role in AMD as well as the development of methods to assist in AMD research. Specifically, the protein composition of retinal LF was assessed to elucidate its origin. These findings are reported in chapter 2. The accumulation, composition and phototoxicity of MLF were analyzed to assess MLF's origin and possible contribution to AMD. These results are reported in chapter 3. Because protein oxidation is possibly a common posttranslational modification to proteins which accumulate in lipofuscin and melanolipofuscin granules, a method for the detection and analysis of oxidized proteins was developed and is reported in chapter 4. Chapter 5 details the proteomic differences between ARPE-19 cells - the only human RPE cell line available for research - in their differentiated and undifferentiated states and compares these to the proteome of human RPE cells. These results are also compared to the phenotypic difference of these cells as observed by transmission electron microscopy.

This thesis is divided into two parts.Part I describes the organic extraction, separation, and liquid chromatographic-mass spectrometric analysis of chromophores from human and bovine retinal pigment epithelium. Flurorophores in the retinal pigment epithelium have been implicated in age related macular degeneration. In addition, the synthesis and characterization of a number of bis-retinoid type compounds that may potentially be found in such extracts, or that may be used for insight into pyridinium bis-retinoid reactivity, was accomplished.Part II describes a study of pyridinium bis-retinoid-folic acid coupling with respect to linker type, linker length, and nature of the linkage. Folic acid has been used as a targeting compound for a variety of cancer types. Development of HPLC and UV-Vis conditions suitable for the analysis of this new type of macromolecule was performed.

Dedifferentiation and proliferation of retinal pigment epithelial (RPE) cells are characteristics of retinal diseases. Dedifferentiation is likely associated with changes of inwardly rectifying potassium (Kir) channels. The roles of Kir4.2 channels in viability, and proliferation of cultured RPE cells were investigated. Gene expression levels were determined using qRT-PCR. RPE cells expressed Kir2.1, 2.2, 2.4, 3.2, 4.1, 4.2, 6.1, and 7.1 mRNA. Kir4.2 protein was verified by immunocytochemistry and Western blotting. Kir4.2 mRNA in cultured cells was upregulated by hypoxia (hypoxia mimetic CoCl2 or 0.2% O2) and extracellular hyperosmolarity (addition of high NaCl or sucrose). Kir4.2 mRNA was suppressed by vascular endothelial growth factor (VEGF), blood serum, and thrombin whereas platelet-derived growth factor (PDGF), basic fibroblast growth factor (bFGF), and transforming growth factor-1 (TGF-1) increased it. Hyperosmotic Kir4.2 gene expression was mediated by TGF-1 receptor signaling while hypoxic gene transcription was dependent on PDGF receptor signaling. VEGF receptor-2 blockade increased Kir4.2 mRNA level under control, hyperosmotic, and hypoxic conditions. SiRNA-mediated knockdown of Kir4.2 decreased the cell viability and proliferation under control and hyperosmotic conditions. Kir4.2 channels play functional roles in maintaining the viability and proliferation of RPE cells. Downregulation of Kir4.2 by VEGF, via activation of VEGF receptor-2 and induction of blood-retinal barrier breakdown, may contribute to decreased viability of RPE cells under pathological conditions.

It is well known that ultraviolet (UV) B (280-315 nm) and blue light (400-500 nm) radiation can produce phototoxic lesions in the neural retina and the retinal pigment epithelium (RPE). In the first section of this thesis, bovine lens cells (epithelium and superficial cortical fibre cell) and human retinal pigment epithelial (ARPE-19) cells were used to characterize in vitro changes following oxidative stress with UVB radiation in ocular lens optics and cellular function in terms of mitochondrial dynamics. In the second part, human retinal pigment epithelial (ARPE-19) cells and in vitro bioassays were used together to develop an in vitro approach for UV radiation-induced retinal toxicology research. In the third chapter, the in vitro approach developed above was used with intraocular lens (IOL) materials to evaluate the UV radiation blocking efficiency of commercially available IOL’s. Lastly, narrowband blue light irradiation and in vitro assays were used to determine more precisely the wavelengths of blue light responsible for photochemical lesions of the retina as an effort to contribute to future IOL designs. The results from mitochondrial dynamics of lens cells and RPE cells show significant decreases in mitochondrial movement after UVB irradiation in a dose dependent manner. Results obtained from four in vitro assays (Alamar blue assay, confocal microscopy for mitochondrial distribution and nucleic acids damage, phagocytotic activity assay) for evaluating the UVB-induced damage in ARPE-19 show significant decreases in cell viability as well as phagocytotic activity of RPE cells after UVB radiation. In addition, the results show that UV radiation can also induce the degradation of DNA/RNA and mitochondria of RPE cells in a dose dependent manner. The results of the UV blocking efficiency test of commercially available IOL materials show very effective UV blocking ability, allowing no cellular damage at all, in comparison to an IOL uncovered control cell. The results of three different wavelengths of blue light exposure show that only 400 nm blue light radiation can cause significant damage to RPE cells, while 420 and 435.8 nm blue light radiation cause no cellular damage at all. In conclusion, UVB and blue light radiation can cause phototoxic damage to the retinal pigment epithelium as a result of oxidative stress, and in vitro bioassays used for this research may offer a sensitive, and meaningful biomarker approach, not only for evaluating RPE function after oxidative and chemical stress, but also for evaluating IOL effectiveness.

It is well known that ultraviolet (UV) B (280-315 nm) and blue light (400-500 nm) radiation can produce phototoxic lesions in the neural retina and the retinal pigment epithelium (RPE). In the first section of this thesis, bovine lens cells (epithelium and superficial cortical fibre cell) and human retinal pigment epithelial (ARPE-19) cells were used to characterize in vitro changes following oxidative stress with UVB radiation in ocular lens optics and cellular function in terms of mitochondrial dynamics. In the second part, human retinal pigment epithelial (ARPE-19) cells and in vitro bioassays were used together to develop an in vitro approach for UV radiation-induced retinal toxicology research. In the third chapter, the in vitro approach developed above was used with intraocular lens (IOL) materials to evaluate the UV radiation blocking efficiency of commercially available IOL’s. Lastly, narrowband blue light irradiation and in vitro assays were used to determine more precisely the wavelengths of blue light responsible for photochemical lesions of the retina as an effort to contribute to future IOL designs. The results from mitochondrial dynamics of lens cells and RPE cells show significant decreases in mitochondrial movement after UVB irradiation in a dose dependent manner. Results obtained from four in vitro assays (Alamar blue assay, confocal microscopy for mitochondrial distribution and nucleic acids damage, phagocytotic activity assay) for evaluating the UVB-induced damage in ARPE-19 show significant decreases in cell viability as well as phagocytotic activity of RPE cells after UVB radiation. In addition, the results show that UV radiation can also induce the degradation of DNA/RNA and mitochondria of RPE cells in a dose dependent manner. The results of the UV blocking efficiency test of commercially available IOL materials show very effective UV blocking ability, allowing no cellular damage at all, in comparison to an IOL uncovered control cell. The results of three different wavelengths of blue light exposure show that only 400 nm blue light radiation can cause significant damage to RPE cells, while 420 and 435.8 nm blue light radiation cause no cellular damage at all. In conclusion, UVB and blue light radiation can cause phototoxic damage to the retinal pigment epithelium as a result of oxidative stress, and in vitro bioassays used for this research may offer a sensitive, and meaningful biomarker approach, not only for evaluating RPE function after oxidative and chemical stress, but also for evaluating IOL effectiveness.

Influence of stress factors on tunneling nanotubes in cultivated human retinal pigment epithelial cells (ARPE-19). The eye as one of the most important sense organs of the human body is exposed to visible light radiation and other stress factors every day. Especially the retina (of the eye) is a sensible tissue for oxidative damage (Wu et al., 2006). The retinal pigment epithelium (RPE) is an important layer of the retina, which forms the outer layer and phagocytises the shed disc membranes of the photoreceptor outer segments. Furthermore, the RPE is involved in the maintenance of the visual cycle and regulates the retinal balance (Bok, 1993). To maintain those functions, a steady communication between the RPE-cells and the adjacent neighbour cells is necessary. Tunneling nanotubes (TNTs) build a newly discovered variety of cell communication and thus establish intercellular signal transduction and transport different cell components including pathogens (Rustom et al., 2004; Onfelt et al., 2006; Sherer und Mothes, 2008; Veranic et al., 2008). The formation of TNTs in the neuron-like pheochromocytoma cell line PC12 was first reported by Rustom et al in 2004. In the following years a growing number of cell types containing TNTs were described. For example a lot of TNT-reports were found between immune cells (Onfelt et al., 2004; Sowinski et al., 2008). Chinnery et al. first described TNTs in vivo in 2008. Here they found TNTs between dendritic cells in the cornea of the mouse. An important characteristic of TNTs is that they do not attach to the substratum. They contain F-actin as a characteristic feature of there structure (Rustom et al., 2004). Our study group detected the formation of TNTs between ARPE-19-cells, a human retinal pigment epithelial cell line. They contain F-actin, but no microtubules. Further it was observed an exchange of electrical signals, small molecules and even the transfer of organelles between cells via TNTs (see publication Wittig et al., 2012). It is often described in the literature, that TNTs are very sensitive against stress factors, like prolonged light excitation, mechanical and chemical stress, which then can result in rupture of the TNTs (Rustom et al., 2004; Koyanagi et al., 2005; Gurke et al., 2008a; Pontes et al., 2008; Sowinski et al., 2008; Domhan et al., 2011; Wang und Gerdes, 2012). Up to now it is widely unclear how pathological conditions influences TNTs. There are several studies, which report an induction but also an inhibition of TNT-formation by different factors. The reaction of cell-cell-interactions between RPE cells on stress factors is not jet analysed. So our motivation was, to analyse the influence of different stress factors on the number, the morphology and formation of TNTs. ARPE-19-cells were treated with blue light, with a wavelength of 470 and 405 nm, with 3000 μM glyoxal, with 200 μM H2O2, with medium without serum as well as with cytochalasin-D and latrunculin-B. With the help of differential interference contrast (DIC) microscopy the formed TNTs were counted and the morphology was evaluated. A 24 hours cultivation of untreated ARPE-19 cells resulted in 15 TNTs per 100 cells on average. After excitation of the ARPE-19-cells with blue light 470 and 405 nm the number of TNTs decreased 50 % and 28,5 % accordingly in comparison to untreated cells (100 %). Furthermore, the cell culture, which was treated with glyoxal and H2O2 resulted in a reduction of 17,5 % and 53 % TNTs in comparison to the untreated cell culture. Cells which were cultured with serum free medium had an decreased TNT-number of 56.8 % in comparison with serum containing medium. TNTs of untreated ARPE-19-cells have a diameter from 50 to 300 nm (Wittig et al., 2012). Every TNTs, which were formed under named stress factors had the same diameter like untreated cells. In this study an average TNT length of 23 +/- 16 μm was measured between cells without treatment. This correlated with the TNT-lengths of cells which excitated with blue light 405 and 470 nm with 26 +/- 13 μm and 24 +/- 14 μm. In contrast the TNT-lenghts of cells treated with glyoxal and H2O2 with 16 +/- 11 μm and 15 +/- 13 μm were less and from cells cultured without serum with 34 +/- 20 μm were above the average length of TNTs of untreated cells. TNTs of ARPE-19-cells without treatment and TNTs which were treated with stress factors contained F-actin but no microtubules. Depolymerisation of F-actin, induced by addition of cytochalasin-D or latrunculin-B, led to disappearance of TNTs. This is an evidence for the importance of F-actin as an essential component of TNTs between ARPE-19-cells. Under the influence of blue light excitation the TNTs formed as good as untreated cells after contact of migrating cells. Reason for the reduced TNT-formation under stress factors could be explained by the generation of oxidative stress due to reactive oxygen species (ROS). ROS induced under blue light- or glyoxal-treatment as well as H2O2 could influence cell function by inactivation of cell-mediated proteins or induction of F-actin oxidation with subsequent destruction of the actin-network and inhibition of the actin-polymerisation (Chen, 1993; Ballinger et al., 1999; Thornalley et al., 1999; Valen et al., 1999; Dalle-Donne et al., 2002; Nilsson et al., 2003; Shangari und O'Brien, 2004; Zhu et al., 2005; Knels et al. 2008; Roehlecke et al., 2009). The reduced actin-polymerisation as well as the disruption of the TNTs due to changes at the actin-cytoskeleton and at the membranes could explain the reduced TNT-formation (Valen et al., 1999; Dalle-Donne et al., 2002; Reber et al., 2002; Zhu et al., 2005; Knels et al., 2008). The inhibition of the cell growth under oxidative stress conditions and under nutritional deficiency by serum free medium could lead to a reduced TNT-formation too. In this study we found a reduction of TNT-number between ARPE-19-cells under different stress conditions. It is possible, that TNTs are formed between RPE- and photoreceptor-cells in vivo, where they can exchange useful or recyclable materials between cells (Wang et al., 2011; Wittig et al., 2012). Disruption of TNTs by reactive oxygen species could cause a decreased exchange of informations. It is possible, that the cells, RPE- as well as photoreceptor-cells, die due to a deficiency of nutrients. This could be another reason in the formation of age related macular degeneration, which shows a destruction of RPE-cells and secondary of the photoreceptorcells. / Das Auge ist als eines der wichtigsten Sinnesorgane des Menschen täglich sichtbarer Lichtstrahlung und weiteren Stressfaktoren ausgesetzt. Die Netzhaut des Auges ist besonders empfindlich für oxidative Schäden (Wu et al., 2006). Eine bedeutende Schicht der Netzhaut im Auge stellt das retinale Pigmentepithel (RPE) dar, welches die äußere Schicht der Retina bildet und täglich die abgeworfenen Photorezeptoraußensegmentscheiben phagozytiert. Zudem ist das RPE wesentlich am visuellen Prozess sowie der Aufrechterhaltung des retinalen Gleichgewichts beteiligt (Bok, 1993). Um diese Funktionen zu gewährleisten, ist eine ständige Kommunikation zwischen den RPEZellen sowie zu angrenzenden Nachbarzellen innerhalb der Netzhaut notwendig. So ist über Tunneling Nanotubes (TNTs), als neu entdeckte Kommunikationsform, ein interzellulärer Transport von Signalen und verschiedensten Zellkomponenten, aber auch von Pathogenen, möglich (Rustom et al., 2004; Onfelt et al., 2006; Sherer und Mothes, 2008; Veranic et al., 2008). Erstmals 2004 beschrieben Rustom et al. die Bildung von TNTs zwischen Rattennierenzellen in vitro. In den folgenden Jahren kam es zu einer Vielzahl weiterer TNT-Entdeckungen zwischen verschiedensten Zellen in vitro. So findet man zum Beispiel vermehrt TNTBeschreibungen zwischen Immunzellen (Onfelt et al., 2004; Sowinski et al., 2008). Ein erster Nachweis an TNTs in vivo erfolgte 2008 durch die Arbeitsgruppe Chinnery et al.. Hierbei fand man TNTs zwischen dendritischen Zellen in der Mauscornea. Ein wichtiges Merkmal von TNTs ist, dass sie sich als frei im Medium schwebende interzelluläre Verbindungen darstellen, ohne Kontakt zum Substrat zu haben. TNTs sind im Wesentlichen als stabilisierendes Hauptstrukturmerkmal aus Aktin aufgebaut (Rustom et al., 2004). In unserer Arbeitsgruppe wurde die Bildung von TNTs zwischen ARPE-19-Zellen, einer humanen Pigmentepithelzelllinie, entdeckt. Neben dem strukturellen Aufbau aus Aktin, konnte ein Austausch von elektrischen Signalen sowie molekularen Stoffen und der Transport von Organellen (Mitochondrien) durch TNTs zwischen ARPE-19-Zellen nachgewiesen werden (siehe Publikation Wittig et al., 2012). Wie schon mehrfach in der Literatur beschrieben, reagieren TNTs sehr sensibel auf Stressfaktoren, so zum Beispiel auf längere Lichtreizung, mechanischen und chemischen Stress, was jeweils zur Ruptur der Strukturen führen kann (Rustom et al., 2004; Koyanagi et al., 2005; Gurke et al., 2008; Pontes et al., 2008; Sowinski et al., 2008; Domhan et al., 2011; Wang und Gerdes, 2012). Weitgehend unklar ist bisher der Einfluss von pathologischen Bedingungen auf die TNTs. Es gibt mehrere Studien, in denen durch verschiedenste Faktoren über eine Induktion, aber auch über eine Hemmung der TNT-Bildung berichtet wurde. Die Reaktion von Zell-Zell-Interaktionen zwischen RPE-Zellen auf Stressfaktoren wurde bisher in wissenschaftlichen Arbeiten nicht untersucht. Dies nahmen wir zum Anlass, den Einfluss von unterschiedlichen Stressfaktoren auf die Anzahl von TNTs, ihre Morphologie und Bildung zu untersuchen. Es erfolgte eine Behandlung der ARPE-19-Zellen mit Blaulicht in den Wellenlängen 405 und 470 nm, mit 3000 μM Glyoxal, mit 200 μM H2O2, mit serumfreiem Medium sowie mit Cytochalasin D und Latrunculin B. Die gebildeten TNTs wurden anschließend mit Hilfe der Lichtmikroskopie ausgezählt sowie deren Morphologie beurteilt. So bildeten unbehandelte ARPE-19-Zellen nach 24 Stunden Kultivierung im Durchschnitt 15 TNTs pro 100 Zellen aus. Nach 24stündiger Bestrahlung der ARPE-19-Zellen mit Blaulicht 470 nm und 405 nm fiel die TNT-Anzahl auf 50 % und 28,5 % im Vergleich zu unbehandelten Zellen (100 %). Weiterhin fanden sich in den Glyoxal- und H2O2-behandelten Kulturschalen 17,5 % und 53 % TNTs verglichen mit der unbehandelten Zellkultur. In der serumfreien Kulturschale verringerten sich die TNTs 24 Stunden nach Ausplattierung der Zellen auf 56,8 % im Vergleich zu in Medium mit Serum kultivierten Zellen. TNTs unbehandelter ARPE-19-Zellen besitzen einen Durchmesser von 50 bis 300 nm (Wittig et al., 2012). Alle unter oben genannten Stressfaktoren gebildeten TNTs befanden sich in Hinblick auf ihren Durchmesser im Bereich der TNTs unbehandelter Zellen. Bei TNTs unbehandelter Zellen wurde in dieser Arbeit eine durchschnittliche Länge von 23 +/- 16 μm gemessen. Dies entsprach dem TNT-Längendurchschnitt von mit Blaulicht 405 nm und 470 nm bestrahlter ARPE-19-Zellen mit 26 +/- 13 μm und mit 24 +/- 14 μm. Unter Glyoxal und H2O2 gebildete TNTs lagen im Gegensatz dazu mit 16 +/- 11 μm und 15 +/- 13 μm unterhalb und unter serumfreier Kultivierung mit 34 +/- 20 μm über dem TNTLängendurchschnitt unbehandelter Zellen. Alle TNTs, sowohl unbehandelter als auch mit Stressfaktoren behandelter ARPE-19-Zellen, sind aus Aktin aufgebaut. Jedoch ließ sich kein Tubulin nachweisen. Nach Zugabe von Aktinpolymerisationshemmern waren keine TNTs nachweisbar, was beweist, dass F-Aktin essentieller Bestandteil von TNTs zwischen ARPE-19-Zellen ist. Unter dem Einfluss von Blaulicht 470 und 405 nm bildeten sich die TNTs, wie auch bei unbehandelten Zellen, durch ein Zusammentreffen der Zellen mit anschließendem Auseinandergleiten. Die Ursache für die verminderte Bildung an TNTs unter verschiedenen Stressfaktoren könnte in der Entstehung von oxidativem Stress durch die Ausbildung von reaktiven Sauerstoffspezies (ROS) begründet sein. So können zum Beispiel die unter Blaulicht- und Glyoxalexposition entstehenden ROS sowie H2O2, als eine Hauptform der ROS, die Zellfunktion durch Inaktivierung zellulärer Proteine beeinflussen sowie eine direkte Oxidation an Aktin hervorrufen mit folglicher Aktinnetzwerkzerstörung und Hemmung der Aktinpolymerisation (Chen, 1993; Ballinger et al., 1999; Thornalley et al., 1999; Valen et al., 1999; Dalle-Donne et al., 2002; Nilsson et al., 2003; Shangari und O'Brien, 2004; Zhu et al., 2005; Knels, Worm et al. 2008; Roehlecke et al., 2009). Die verminderte Aktinpolymerisation, aber auch die Zerreißungen der TNTs durch Veränderungen am Aktinzytoskelett sowie an den Membranen könnten zu einer verringerten TNT-Bildung führen (Valen et al., 1999; Dalle-Donne et al., 2002; Reber et al., 2002; Zhu et al., 2005; Knels et al., 2008). Auch eine Hemmung des Zellwachstums unter oxidativen Stressbedingungen sowie unter Nährstoffmangel durch Serumentzug könnte mit einer verminderten TNT-Bildung einhergehen. Wir haben in unserer Untersuchung gezeigt, dass es durch verschiedene Stresseinflüsse zu einer Reduktion der TNTs zwischen ARPE-19-Zellen kommt. Es ist denkbar, dass solche TNTs in vivo zwischen RPE- und Photorezeptorzellen ausgebildet werden, wo sie nützliches oder recycelbares Material zwischen Zellen austauschen (Wang et al., 2011; Wittig et al., 2012). Bei Zerstörung der TNTs durch zum Beispiel oxidative Faktoren könnte es zu einer Verringerung des Informationsaustausches kommen. Es ist möglich, dass durch die Minderversorgung die Zellen absterben, sowohl RPE- als auch Photorezeptorzellen. Dies könnte ein weiterer möglicher Ursachenansatz in der Entstehung der altersabhängigen Makuladegeneration sein, welche als Erkrankungserscheinung den Untergang der RPEZellen und damit sekundär der Photorezeptorzellen aufweist.

We investigated attachment and migration of human retinal pigment epithelial cells (primary, SV40-transfected and ARPE-19) on nanoscopically defined, two-dimensional matrices composed of parallel-aligned collagen type I fibrils. These matrices were used non-cross-linked (native) or after riboflavin/UV-A cross-linking to study cell attachment and migration by time-lapse video microscopy. Expression of collagen type I and IV, MMP-2 and of the collagen-binding integrin subunit α2 were examined by immunofluorescence and Western blotting. SV40-RPE cells quickly attached to the nanostructured collagen matrices and aligned along the collagen fibrils. However, they disrupted both native and cross-linked collagen matrices within 5 h. Primary RPE cells aligned more slowly without destroying either native or cross-linked substrates. Compared to primary RPE cells, ARPE-19 cells showed reduced alignment but partially disrupted the matrices within 20 h after seeding. Expression of the collagen type I-binding integrin subunit α2 was highest in SV40-RPE cells, lower in primary RPE cells and almost undetectable in ARPE-19 cells. Thus, integrin α2 expression levels directly correlated with the degree of cell alignment in all examined RPE cell types. Specific integrin subunit α2-mediated matrix binding was verified by preincubation with an α2-function-blocking antibody, which impaired cell adhesion and alignment to varying degrees in primary and SV40-RPE cells. Since native matrices supported extended and directed primary RPE cell growth, optimizing the matrix production procedure may in the future yield nanostructured collagen matrices serving as transferable cell sheet carriers. / Dieser Beitrag ist mit Zustimmung des Rechteinhabers aufgrund einer (DFG-geförderten) Allianz- bzw. Nationallizenz frei zugänglich.

Thesis (Ph.D.)--University of Alabama at Birmingham, 2008. / Title from PDF title page (viewed on Sept. 17, 2009). Includes bibliographical references.