scieee AI-readable full text Open interactive document viewer

Funktion und Einfluss der nicht-repetitiven, terminalen Domänen auf Speicherung und Assemblierung von Spinnenseidenproteinen

Eisoldt, Lukas

Full text

Funktion und Einfluss der nicht-repetitiven, terminalen Domänen auf Speicherung und Assemblierung von Spinnenseidenproteinen Dissertation Zur Erlangung des Grades - Doktor der Naturwissenschaften - im Promotionsprogramm „Molekulare Biowissenschaften“ der Bayreuther Graduiertenschule für Mathematik und Naturwissenschaften (BayNAT) Lukas Eisoldt Molekulare Biotechnologie (Master of Science) Bayreuth 2013 Die vorliegende Arbeit wurde in der Zeit von Februar 2008 bis Januar 2013 in Bayreuth am Lehrstuhl Biomaterialien unter Betreuung von Professor Dr. Thomas Scheibel angefertigt. Vollständiger Abdruck der von der Bayreuther Graduiertenschule für Mathematik und Naturwissenschaften (BayNAT) der Universität Bayreuth genehmigten Dissertation zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaften (Dr. rer. nat). Dissertation eingereicht am: 05.06.2013 Zulassung durch das Leitungsgremium: 10.06.2013 Wissenschaftliches Kolloquium: 24.09.2013 Amtierender Direktor: Prof. Dr. Franz Xaver Schmid Prüfungsausschuss: Prof. Thomas Scheibel (Erstgutachter) Prof. Franz Xaver Schmid (Zweitgutachter) Prof. Olaf Stemmann (Vorsitz) PD Dr. Stephan Schwarzinger Für Tina Inhaltsverzeichnis Inhaltsverzeichnis Tabellenund Abbildungsverzeichnis ................................................................ I Abkür zun gsverzeichnis .................................................................................... III Kurzfassung der Arbeit ...................................................................................... 1 Abstract ................................................................................................................ 3 1. Einleitung ......................................................................................................... 5 1.1 Seiden ................................................................................................................................ 5 1.2 Spinnenseiden ................................................................................................................... 5 1.3 Aufbau von Spinnenseidenproteinen ................................................................................ 8 1.4 Die terminalen Domänen ................................................................................................ 11 1.5 Major Ampullate Seide ................................................................................................... 13 1.5.1 Die Proteine der MA-Seide .......................................................................................... 13 1.5.2 Struktur der repetitiven Kerndomäne ....................................................................... 15 1.5.3 Die terminalen Domänen von MaSp-Proteinen ....................................................... 16 1.5.3.1 Die Struktur der aminoterminalen (NT) Domäne............................................. 17 1.5.4 Nanound Mikrostrukturierung der MA-Faser ........................................................ 19 1.5.5 Der natürliche Spinnprozess .................................................................................... 20 1.6 Rekombinante Spinnenseidenproteine ............................................................................ 22 1.7 Assemblierungseigenschaften von rekombinanten Spinnenseidenproteinen ................. 23 1.7.1 Fasern aus rekombinanter Seide ............................................................................... 23 1.8 Zielsetzung ...................................................................................................................... 25 2. Synopsis .......................................................................................................... 27 2.1 Übersicht der verwendeten Proteine ............................................................................... 27 2.2 Die NR3-Domäne ........................................................................................................... 28 2.2.1 Lösungsstruktur der NR3-Domäne .......................................................................... 28 2.2.2 Einfluss von Salz und Scherkräfte auf das NR3-Dimer ........................................... 30 2.2.3 Zusammenfassung: NR3-Domäne ........................................................................... 31 2.3 Einfluss von NR3 auf das Lösungsverhalten von eADF3((AQ)n) Proteinen ................. 31 2.3.1 Bildung von supramolekularen Assemblaten ........................................................... 31 Inhaltsverzeichnis 2.3.2 Einfluss der NR3-Domäne auf phosphatinduzierte Aggregation von Seidenproteinen ........................................................................................................................... 32 2.3.3 Einfluss von Scherkräften auf die Assemblierung von eADF3-Proteinen mit NR3-Domäne .................................................................................................................... 34 2.3.4 Modell zum Einfluss der NR3-Domäne während des Spinnprozesses .................... 37 2.4 Heterodimerisierung von NR3-NR4-Konstrukten .......................................................... 38 2.4.1 Bildung des Heterodimers ........................................................................................ 38 2.4.2 Assemblierungseigenschaften des Heterodimers ..................................................... 39 2.4.3 Assemblierungsmorphologien der Homodimere im Vergleich zum Heterodimer ...................................................................................................................... 39 2.4.4. Bedeutung des Heterodimers .................................................................................. 40 2.5 Der Einfluss der terminalen Domänen auf die Orientierung der repetitiven Kerndomäne von rekombinanten Spinnenseidenproteinen .................................................. 41 2.5.1 pH-abhängige Untersuchung der N1und N1L-Proteine mittels CDund Fluoreszenzspektroskopie ................................................................................................. 42 2.5.2 SEC-MALS-Analyse der N1und N1L-Proteine .................................................... 44 2.6 Modell des natürlichen Spinnprozesses .......................................................................... 45 3. Literaturverzeichnis ...................................................................................... 48 4. Publikationsliste............................................................................................. 56 5. Darstellung des Eigenanteils ........................................................................ 57 6. Teilarbeiten .................................................................................................... 58 Teilarbeit I ............................................................................................................................ 58 Teilarbeit II ........................................................................................................................... 73 Teilarbeit III .......................................................................................................................... 82 Teilarbeit IV ....................................................................................................................... 103 Teilarbeit V ......................................................................................................................... 111 Rechtliches ....................................................................................................... 120 Anhang ............................................................................................................. 121 Danksagung ...................................................................................................... 124 Abbildungssverzeichnis I Tabellenund Abbildungsverzeichnis Abbildung 1: Ordnungen, Unterordnungen, Familien und Arten von Webspinnen. .................. 6 Abbildung 2: Die verschiedenen Seidenarten eines Radnetzspinne ........................................... 7 Abbildung 3: Schematische Darstellung der zwei Spidroinklassen ........................................... 8 Abbildung 4: Vergleich der repetitiven Sequenzen der verschiedenen Spidroin Arten. .......... 10 Abbildung 5: Vergleich der Homologie terminaler Sequenzen von verschiedenen Spidroinarten und Spezies......................................................................................................... 12 Abbildung 6: Schematische Darstellung von MaSp-Proteinen und Sequenzvergleich repetitiver Blöcke von MaSp 1 und 2 Spidroinen ..................................................................... 14 Abbildung 7: Sequenzvergleich der aminound carboxyterminalen Domänen von MAProteinen ................................................................................................................................... 17 Abbildung 8: Struktur der aminoterminalen Domäne von MaSp1 aus E. australis ................. 18 Abbildung 9: Nanound Mikrostrukturierung der MA-Faser ................................................. 19 Abbildung 10: Schematische Darstellung des Spinnapparats ................................................... 21 Abbildung 11: Übersicht der in dieser Arbeit verwendeten rekombinanten Spinnenseidenproteine .............................................................................................................. 27 Abbildung 12: Lösungsstruktur von NR3. ................................................................................ 29 Abbildung 13: Phosphationen induzierte Aggregation von eADF3-Proteinen. ....................... 33 Abbildung 14: Untersuchung scherkraftinduzierter Assemblate von eADF3((AQ)24) und eADF3((AQ)12NR3) mittels Lichtmikroskopie. ....................................................................... 35 Abbildung 15: Polarisierte FTIR Messungen an scherkraftinduzierten Assemblaten von eADF3((AQ)24 und eADF3((AQ)12NR3) ............................................................................... 36 Abbildung 16: Einfluss der NR-Domäne auf die Assemblierung von Spinnenseidenproteinen ............................................................................................................ 37 Abbildung 17: AFM-Aufnahmen der drei dimeren Spezies ..................................................... 40 Abbildung 18: Molekulares Modell des Spinnprozesses innerhalb der MA-Drüse ................. 46 Abbildung 20: pH-abhängige Fluoreszensmessungen von N1-Proteinen. ............................... 43 Abbildung 21: SEC-MALS Analyse von N1-eADF3 und N1-eADF3-NR3 Proteinen in Abhängigkeit vom pH-Wert. .................................................................................................... 44 Tabellenverzeichnis II Tabelle 1: Vergleich der mechanischen Eigenschaften von ausgewählten Spinnenseidenfasern und synthetischen Fasermaterialien. ......................................................... 7 Tabelle2: Übersicht der Sequenzmotive der repetitiven Einheiten von MA-, MIund Flagelliform-spidroinen .............................................................................................................. 9 Tabelle 3: Übersicht der häufigsten Aminosäuren in den einzelnen Spidroinsequenzen ........... 9 Tabelle 4: Mechanische Werte von Fasern auf Basis von rekombinanten Spinnenseidenproteinen im vergleich mit natürlicher MA-Seide ............................................. 24 Abbildungssverzeichnis III Abkürzungsverzeichnis ADF Araneus diadematus fibroin AFM Atomic force microscopy CD Circular Dichroismus CTD Carboxyterminale Domäne DLS Dynamische Lichtstreuung eADF engineered ADF FPLC fast liquid chromatography FTIR Fourier Transform Infrarot GdmHCl Guanidinhydrochlorid HSQC heteronuclear single quantum coherence MaSp Major Ampullate Spidroin NMR Nuclear magnetic resonance NTD Aminoterminale Domäne ppm parts per million rmsd root mean square deviation SEC size exclusion chromatography MALDI-MS Matrix-assisted Laser Desorption/Ionization Massenspektrometrie MALS multi angle light scattering TEM Transmissionselektronenmikroskopie Einleitung [6] Abbildung 1: Ordnungen, Unterordnungen, Familien und Arten von Webspinnen. Die Ordnung der Webspinnen unterteilt sich in die drei Unterordungen VogelWolfsund Echte Webspinnen. Innerhalb der Echten Webspinnen (Araneomorphae) sind die netzbauenden Spinnen in der Superfamilie (SF) der Netzspinnen (Araneoidae) zusammengefasst. Baldachin- (Linyphiidae), Haubennetz- (Theridiiae) und Echte Radnetzspinnen (Araneidae) stellen die drei artenreichsten Netzspinnen dar und ergeben zusammen ca. ein Viertel der Ordnung der Webspinnen. Bekannte und gut untersuchte Arten der Netzspinnen sind die Gartenkreuzspinne (Araneus diadematus), die Goldene Radnetzspinne (Nephila spp.) sowie die Schwarze Witwe (Latrodectus hesperus) Weibliche Radnetzspinnen sind in der Lage bis zu sieben verschiedene Seidenarten, in dafür spezialisierten Drüsen zu produzieren. Allein fünf dieser Seiden werden benötigt um ein Radnetz zu bauen (Abbildung 2). Die Seiden aus den Großen Ampullen- (Major Ampullate, MA) und Flagelliformdrüsen (Flag) bilden mit Rahmen, Speichen und Fangspirale den Hauptteil des Netzes. Zur Verstärkung, z.B. an längeren Spannfäden, wird die Seide aus der kleinen Ampullendrüse (Minor Ampullate, MI) verwendet. Verankert wird das Netz mit Hilfe der Pyriformseide. Damit die Beute nach dem Einschlag durch das Zurückfedern des Netzes nicht herausgeschleudert wird, ist die Fangspirale in regelmäßigen Abständen mit der klebrigen Aggregatseide beschichtet. Die beiden nicht am Netzbau beteiligten Seiden werden für das Einwickeln der Beute (Aciniform) sowie zur Herstellung des Eischale (Tubiliformbzw. auch manchmal Cylindriformseide) verwendet. Netzspinnen (SF) Araneoidae Baldachinspinnen Linyphiidae Echte Radnetzspinnen Araneidae Goldene Radnetzspinne Nephila spp. Gartenkreuzspinne A raneus diadematus Haubennetzspinnen Theridiidae Schwarze Witwe Latrodectus hesperus Vogelspinnen Mygalomorphae W ebspinne n Araneae Echte Webspinnen Araneomorphae Gliederspinnen Mesothelae Einleitung [7] Abbildung 2: Die verschiedenen Seidenarten eines Radnetzspinne. Weibliche Radnetzspinnen können bis zu sechs verschiedene Seidenarten sowie eine Klebesubstanz in dafür spezialisierten Drüsen bilden. Die entsprechenden Seiden sind in der Grafik rot dargestellt. Modifiziert nach [3]. Ein wesentlicher Teil der Faszination für Spinnenseiden ist in den mechanischen Eigenschaften der Fasern begründet, die denen von synthetischen Hochleistungsfasern zum Teil überlegen sind (Tabelle 1). Tabelle 1: Vergleich der mechanischen Eigenschaften von ausgewählten Spinnenseidenfasern und synthetischen Fasermaterialien. Begriffserklärungen: Reisfestigkeit = Kraft, die benötigt wird bis die Faser reißt. Dehnbarkeit = Länge, bezogen auf die Ausgangslänge um die die Faser gedehnt werden kann, bis sie reißt. Zähigkeit = Energie, die benötigt wird um die Faser zu zerreißen. Der Wert errechnet sich aus Reißfestigkeit, Dehnbarkeit und dem Querschnitt der Faser. Material Reisfestigkeit (GPa) Dehnbarkeit (%) Zähigkeit (MJm-3) Quelle Major AmpullateSeide 1,1 27 160 [4] Minor AmpullateSeide 0,9 33 137 [5] Flagelliform-Seide 0,5 270 150 [4] Aciniform-Seide 1,1 40 230 [5] Hochleistungsstahl 1,5 0,8 6 [4] Kevlar 3,6 2,7 50 Nylon 0,9 18 80 Major Ampullate (MA) Seide Rahmen, Speichen, Abseilfaden Flagelliform (Flag) Seide Fangspirale Aggregat-Seide klebrige Beschichtung der Fangspirale Minor Ampullate (Mi) Seide Hilfsspirale und Netzverstärkung Pyriform-Seide Netzverankerung Aciniform-Seide Einwickeln von Beute Tubiliform-Seide Eischale Einleitung [8] Die Spinnenseidenfasern weisen eine besonders hohe Zähigkeit (Erklärung siehe Tabelle 1) auf, was hauptsächlich auf die hohe Dehnbarkeit zurückzuführen ist. So zeigt die Flagelliformseide der Fangspirale eine hohe Elastizität und kann bis zu 600% reversibel gedehnt werden, während die MA-Seide des Rahmens und der Speichen eine hohe Reißfestigkeit bei geringerer Elastizität aufweisen. Durch die unterschiedlichen Reisfestigkeiten und Dehnbarkeiten der Seiden in einem Radnetz kann der Aufprall von fliegender Beute, die größer und schwerer als die Spinne ist, ohne Schädigung des Netzes abgefedert werden [6]. 1.3 Aufbau von Spinnenseidenproteinen Sämtliche Spinnenseidenfasern bestehen größtenteils aus einem oder mehreren Proteinen, welche als Spidroine bezeichnet werden, die zwischen 250 und 650 kDa großen sind und alle den gleichen schematischen Aufbau besitzen. Sie bestehen aus einer großen repetitiven Kerndomäne, die ca. 90% des Proteins ausmacht sowie zwei nicht-repetitiven, terminalen Domänen. Die Spidroine lassen sich anhand der Länge der repetitiven Blöcke in zwei Klassen unterteilen (Abbildung 3). Abbildung 3: Schematische Darstellung der zwei Spidroinklassen. Sämtliche Spidroine bestehen aus einer repetitiven Kerndomäne (blau) und den nicht-repetitiven aminoterminalen (NT) und carboxyterminalen (CT) Domänen. Anhand der Länge der repetitiven Blöcke der Kerndomäne lassen sich die Spidroine in zwei Klassen einteilen. Die Blöcke in Klasse I (A) weisen eine durchschnittliche Länge von 30-60 Aminosäuren auf und werden wiederum von den Motiven (A)n, (GA)n, GGX, GPGXX aufgebaut. Die Längen der bis zu 100 mal wiederholten Blöcke unterscheiden sich innerhalb des Spidroins zum Teil signifikant. Die Größe der Blöcke in Klasse II (B) eine Länge beträgt 180-360 Aminosäuren. Sequenz und Länge der zwischen 16-25 mal vorkommenden Blöcke sind innerhalb eines Spidroins nahezu identisch. In der ersten Klassen weisen die die repetitiven Blöcke von Major, Minor Ampullate und Flagelliformseide eine durchschnittliche Länge von 30-60 Aminosäuren auf [7-12]. Die Blöcke bestehen wiederum aus der Aneinanderreihung der spezifischen Aminosäuremotive (Ala)4-14, (GlyAla)4-6 , GlyGlyX sowie GlyProGlyXX, wobei X für jede beliebige Aminosäure steht (Tabelle 2). Die Art und die Anzahl der Wiederholungen dieser vier Motive sind für jede Seidenart spezifisch. Die Länge und Sequenz der repetitiven Blöcke innerhalb ein und desselben Spidroins sind nicht homogen. A) Klasse I Major, Minor Ampullate und Flagelliform Spidroine B) Klasse II Aciniform,Tubiliform und Pyriform Spidroine NTD CTD 16-25 NTD CTD 50-100 30-60 AS (A) n , (GA) n , GGX, GPGXX 180-360 AS Einleitung [9] Tabelle2: Übersicht der Sequenzmotive der repetitiven Einheiten von MA-, MIund Flagelliformspidroinen. Die vier Aminosäuremotive sind bei den jeweiligen Spidroinen und damit bei den entsprechenden Seidenarten, die am häufigsten vorkommenden Sequenzmotive. Zu Klasse II gehören Aciniform-, Tubiliformund Pyriformseide, deren repetitive Blöcke eine Länge von ca. 180-360 Aminosäuren aufweisen, welche zwischen 16-25 Mal wiederholt werden. [13-20]. Die Sequenzen der einzelnen repetitiven Blöcke innerhalb eines Proteins sind untereinander sehr homogen und zum Teil bis auf wenige Basen auf genetischer Ebene nahezu identisch [15]. Die Blöcke der Aciniform und Tubiliform Spidroine besitzen eine komplexe, nicht-repetitive Primärsturktur, während die repetitiven Einheiten der Pyriformseide als eine Mischung zwischen beiden Klassen angesehen werden kann. Ein ca. 200 Aminosäuren langer repetitiver Block der Pyriformseide besteht aus mehreren 10-16 Resten langen, alaninreichen Blöcken sowie einer ca. 80 Aminosäuren langen komplexen Sequenz (Abbildung 3). Allen Spidroinen gemein ist, dass einige wenige Aminosäuren in ihren Sequenzen deutlich überrepräsentiert sind, wobei Alanin, Glycin, Glutamin, Prolin und Serin die am häufigsten vorkommenden Aminosäuren in Spinnenseidenproteinen darstellen (Tabelle 3). Tabelle 3: Übersicht der häufigsten Aminosäuren in den einzelnen Spidroinsequenzen. Die Anteile beziehen sich auf die jeweilig bekannte Sequenz wie sie in den Quellen angegeben sind. Aminosäuren, die mehr als 25% der bekannten Sequenz ausmachen, sind farblich (rot) hervorgehoben. Die prozentualen Häufigkeiten der einzelnen Aminosäuren wurden mit Protparam ermittelt. Spidroin Häufigste Aminosäuren (%) Quellen Major Ampullate Spidroin 1 Ala: 32,7-41,4 / Gly: 29,5-42,4 / Gln: 10,0-11,3 [7-9, 21-24] Major Ampullate Spidroin 2 Ala: 18,5-31,1 / Gly: 33,5-37,3 / Gln: 6,9-15,0 / Pro: 8,6-13,5% Minor Ampullate Spidroine Ala: 28,0-36,8 / Gly: 37,0-43,8 [10, 11, 2527] Flagelliform Spidroin Ala: 7,9-10,3 / Gly: 49,7-56,8 / Pro: 15,0-17,4 [11, 28, 29] Aciniform Spidroin Ala: 14,2-15,1 / Gly: 11,3-15,2 / Ser: 13,1-21,5 [13, 15] Tubiliform Spidroin Ala: 25,1-28,0 / Gln: 10,4-12,5 / Ser: 25,9-29,9 [30] Pyriform Spidroin Ala: 16,4-42,2 / Gln: 12,7-16,9 / Ser: 8,4-25,7 [19, 31] (A) 4-13 (GA) ~4-6 GGX GPGXX Protein Sequenzmotiv Seidenart Abseilfaden Hilfsfaden Fangspirale Major Ampullate Spidroin 1 (MaSp1) Major Ampullate Spidroin 2 (MaSp2) Minor Ampullate Spidroins (MiSp) Flagelliform Spidroin (Flag) Einleitung [10] In einigen Fällen, wie z. B. bei den MaSpund MiSp-Proteinen, können zwei Aminosäuren zusammen mehr als die Hälfte der Sequenz ausmachen. Die angegebenen Werte der am häufigsten vorkommenden Aminosäuren zeigen zudem, dass sich selbst innerhalb einer Spidroinart über verschiedene Spezies die Aminosäureanteile erheblich voneinander unterscheiden. Trotz Ähnlichkeiten im Aminosäuregehalt unterscheiden sich die Sequenzen der einzelnen Spidroinarten zum Teil signifikant voneinander (Abbildung 4). Abbildung 4: Vergleich der repetitiven Sequenzen der verschiedenen Spidroin Arten. Dargestellt sind 200 AS lange, repräsentative Abschnitte der repetitiven Kerndomäne der einzelnen Spidroine aus unterschiedlichen Spinnen. Major Ampullate Sp1 & 2 aus L. hesperus (UniProt: A6YIY1, A6YIY0): Dargestellt sind sechs repetitive Blöcke je bestehend aus Polyalanin- (rot), GGX- (blau) und GPGXX-Motiven (grün). Minor Ampullate Sp1 aus A. ventricosus (GeneBank: JX513956.1): Ausschnitt mit GA- (rot), GX/GGX- (grün), Polyalaninblöcken (blau), die den Großteil der Kerndomäne ausmachen sowie einer Spacersequenz (lila). Flagelliform Sp aus N. clavipes (UniProt: Q9NHW4): Ausschnitt mit GPGGX- (rot) und GGXBlöcken (blau), die mehr als 90% der rep. Domäne ausmachen, sowie der zweimal vorkommenden Spacersequenz (lila). Pyriform Sp1 aus L. hesperus (UniProt: C7T5D2): Gezeigt ist ein 200 AS langer repetitiver Block bestehend aus 10 – 16 AS langen, kürzeren Blöcken (rot & blau) sowie einer Spacer-Sequenz (lila). Aciniform Sp1 aus A. trifasciata (UniProt: Q64K55): Darstellung einer 200 AS langen Wiederholungseinheit. Tubuliform Sp1 aus L. hesperus (UniProt: Q4G1Y6): Darstellung einer 178 AS langen Wiederholungseinheit. Major Ampullate Sp1 AAAAAAAAGGAGQGGQGGYGQGGYGQGGYGQGGSGAAAAAAAAAGGAGQGGQGGYGQGGYGQGGAGQGGAGAAAAAAAAA GGAGQGGYGRGGAGQGGAAAAAAAAAGAGQGGYGGQGAGQGGSGAAAAAAAAGGAGQGGQGGYGQGGYGQGGSGAAAAAA AAGGAGQGGQGGYGQGGYGQGGAGQGGAGAAAAAAAAGGA 1 81 161 8 0 1 60 Major Ampullate Sp2 AAAAAASGAGPGRQLGYGPGGSGAAAAAAAGGPGYGGQQGYGPGGAGAAAAAAAGGAGPGRQQTYGPGGSGAAATAAGGS GPGGYGQGPSGYGPSGPGGQQGYGPGGSGAAAAAAAGEAGPGRQQGYGPRGSGAAAAAAAGGPGYGGQSGYGPGGAGAAA AAAAGGAGPGRQQEYGPGGSGAAAAAAAAAGSGPSGYGPG 8 0 1 60 1 81 161 Minor Ampullate Sp1 GAGAGARGADSAGAAAGYGGGVGTGTGSSAGYGRGAGAGAGAGAAAGSGAGAAGGYGGGYGAGAGAGAGAGGATGNRAGD AFAQVFSQNVINSGVITSTTVTKNSAQAAASSMVSTAAKSLGLDENTARSMANAMSSYAAAMAKSFRNSDEFIRNMSYQM GRMLSNAGAINESTASAAASSASSTVTETVRTYGPAAIFS 8 0 1 60 1 81 161 Flagelliform Sp GPGGSGPGGYGPGGSGPGGYGPGGYGPGGSGPGGSGPGGSGPGGYGPGGTGPGGSGPGGYGPGGSGPGGSGPGGYGPGGS GPGGFGPGGSGPGGYGPGGSGPGGAGPGGVGPGGFGPGGAGPGGAAPGGAGPGGAGPGGAGPGGAGPGGAGPGGAGPGGA GGAGGAGGSGGAGGSGGTTIIEDLDITIDGADGPITISEELPIS 8 0 1 60 1 81 161 AAARAQAQAEARAKAEAAARAQAQAEARAKAEATARAKAQAEAAARAQAQAEARAIAEAAARAQAQAEARAKAEAAARAQ AQAIARAEAAARAQAEAEARAYAEALARVQAEAAARAQAQTQSRTQAETHSQAHSASHASSQATSETHVESTAHTATETH EHTSSNSQAASHTQAASHSQAKAQSEAHTYSQAQSAAHTI Pyriform Sp1 8 0 1 60 1 81 161 GSAGPQGGFGATGGASAGLISRVANALANTSTLRTVLRTGVSQQIASSVVQRAAQSLASTLGVDGNNLARFAVQAVSRLP AGSDTSAYAQAFSSALFNAGVLNASNIDTLGSRVLSALLNGVSSAAQGLGINVDSGSVQSDISSSSSFLSTSSSSASYSQ ASASSTSGTGYTGPSGPSTGPSGYPGPLGGGAPFGQSGFG Aciniform Sp1 8 0 1 60 1 81 161 AQAASAAQSSAFAQSQAAAQAFSQAASRSASQSAAQAGSSSTSTTTTTSQAASQAASQSASSSSSAASQSAFSQASSSAL ASSSSFSSAFSSASSASAVGQVGYQIGLNAAQTLGISNAPALADAVSQAVRTVGVGASPFQYANAVSNAFGQLLGGQGIL TQENAAGLASSVSSAISS Tubuliform Sp1 8 0 1 60 1 81 161 Einleitung [11] 1.4 Die terminalen Domänen Laut UniProt-Datenbank (Stand Januar 2013) sind für ca. 90 Spidroine aller Art (außer Aggregatspidroin) die Sequenzen von mindestens einer der nicht-repetitiven, terminalen Domänen bekannt (detaillierte Übersicht s. Anhang Tabelle A1). Etwa 90% stammen von Arten aus der Unterordnung der Echten Webspinnen (Araneomorphae), wobei die Familie der Echten Radnetzspinnen (Araneidae) circa die Hälfte der bekannten Sequenzen ausmacht. Die restlichen 10% sind Familien aus der Unterordnung der Vogelspinnenartigen (Mygalomorphae) zugeordnet. Der überwiegende Teil der Sequenzen stammt aus cDNA-Datenbanken auf Basis von mRNA. Dabei wird häufig die mRNA der betreffenden Drüsen mittels PolythymidinPrimern (PolyT), die an den 3‘-Poly-Adeninschwanz der mRNA binden, isoliert und in cDNA überführt. Bei der späteren Sequenzierung sind daher eher Sequenzen zu finden, die vom ursprünglichen 3‘-Ende der mRNA stammen, welches nach Translation dem Carboxyterminus entspricht. Aufgrund dessen sowie den Schwierigkeiten bei der Sequenzierung von hochrepetitiven Bereichen sind etwa viermal so viele carboxyterminale wie aminoterminale Sequenzen bekannt (Anhang Tabelle A1). Nur für wenige Spidroinsequenzen sind daher beide terminale Domänen bekannt [8, 15, 27, 32-35]. Die Sequenzen von terminalen Domänen zeigen im Gegensatz zu den repetitiven Kerndomänen eine hohe speziesübergreifende Sequenzkonservierung (Abbildung 5) [30, 33, 36]. Die NTDomänen sind sogar über verschiedene Spinnenseidenarten von zum Teil weit entfernten Spezies konserviert, was für eine wichtige Funktion spricht [33, 37]. Ihre Größe liegt ohne Sekretionssignalsequenz mit 125-135 Aminosäuren zudem in einem engen Bereich. Mit 75 – 110 Aminosäuren Länge sind die CT-Domänen im Schnitt kleiner und variabler in der Länge als die NT-Domänen. Eine hohe Sequenzkonservierung ist nur innerhalb derselben Spidroinart von näher verwandten Spezies oder Familien (wie z.B. die Netzspinner) festzustellen [24, 36, 37]. Einleitung [12] A) NT-Sequenzen Nc_MaSp1 1 ---------------QNTPWSSTELADAFINAFMNEAGRTGAFTADQLDDMSTIGDTIKT Lh_MaSp2 1 -----------ALGQANTPWSSKENADAFIGAFMNAASQSGAFSSDQIDDMSVISNTLMA Uldiv_MiSp1 1 -FAVFCAQSYSALGQGASVWSSPQMAENFMNGFSMALSQAGAFSGQEMKDFDDVRDIMNS Nc_Flag 1 -----------RGIIANSPFSNPNTAEAFARSFVSNIVSSGEFGAQGAEDFDDIIQSLIQ Lh_AcSp1 1 VQVEGRKGHHHSSGSSKSPWANPAKANAFMKCLIQKISTSPVFPQQEKEDMEEIVETMMS Agap_TuSp 1 ---QGTSFASSATAGIRNIFGNPNTANNFVDCLKGGIQASPAFPRQEQADIQSIASSILS Bc_Fib1 1 -----------RTKVTNVPWTDEAKGKKFLSTFLDYALDHGLFPQQERDDLEAISQNLIP Konsensus 1 tpws a Fi fm tg F q dDm i tim Nc_MaSp1 46 AMDKMARSNKSSKGKLQALNMAFASSMAEIAAVEQG--GLSVDAKTNAIADSLNSAFYQT Lh_MaSp2 50 AMDNMGGR--ITQSKLQALDMAFASSVAEIAVAD----GQNVGAATNAISDALRSAFYQT Uldiv_MiSp1 60 AMDKMIRSGKSGRGAMRAMNAAFGSAIAEIVAANGG-----KEYQIGAVLDAVTNTLLQL Nc_Flag 50 AQ-SMGKGRHDTKAKAKAMQVALASSIAELVIAESS--GGDVQRKTNVISNALRNALMST Lh_AcSp1 61 AFSSMSTSGGSNAAKLQAMNMAFASSMAELVIAEDADNPDSISIKTEALAKSLQQCFKST Agap_TuSp 58 AGNT----ATKSKAIEQALSTALASSLAEIVITESG--GQDYSKQITDLNGILSNCFIQT Bc_Fib1 50 VFRKTMDSGGNAAAKMKALNMAFASSIAEIAVQEGG--AGSIEEKTQAVSEALAHAFLQT Konsensus 61 a m gkl AlnmAfaSsmAEi v e g g v t ai l f qt Nc_MaSp1 104 TGAANPQFVNEIRSLINMFAQSSANEVSY------ Lh_MaSp2 104 TGVVNNQFITGISSLIGMFAQVSGNEVSYS----- Uldiv_MiSp1 115 TGNADNGFLNEISRLITLFSSVEAND--------- Nc_Flag 107 TGSPNEEFVHEVQDLIQMLSQEQINEVDTSGP--- Lh_AcSp1 121 LGSVNRHFIAEIKDLIGMFAREAAAMEEAGD---- Agap_TuSp 112 TGVENKRFVNSIQNLIRLLAESAVSETTNSIQIGP Bc_Fib1 108 TGSVNIQFIKEIRALITLFAKEGQDNETEN----- Konsensus 121 tG n Fv ei LI mfa e B) CT-Sequenzen Nc_MaSp1 1 GSGASAASAAASRLSSPQASSRVSSAVSNLV---ASG-PTNSAALSSTISNVVSQIGASN Lh_MaSp2 1 -SVASSAASAASALSSPTTHARISSHASTLL---SSG-PTNAAALSNVISNAVSQVSASN Uldiv_MiSp1 1 -------SAASNRIVSAPAVNRMSAASSTLV---SNG-AFNVGALGSTISDMAAQIQAGS Nc_Flag 1 --------GSPGGAY--YPSSRVPDMVNGIMSAMQGS-GFNYQMFGNMLSQYSSGSGTCN Lh_AcSp1 1 --SLNNILDSPQGLKSPQASSRINRLSSSVVNALGPN-GLDINNFSDGLRTTLSQLSSSG Agap_TuSp 1 -----PVLSSETGLSSASASSRVNSLASSVASAIASGQALSADSFAKSLLIQASQIQSSA Bc_Fib1 1 -----RLLSSPSGLTSEAAKERISSIIASLLSAKSSQ-DFNALLLSNILPSLISKISQRA Konsensus 1 al s a Rv stlv n i sqi Nc_MaSp1 57 PGLSGCDVLIQALLEVVSALIQILGSSSIGQVNYGSAGQATQIVGQSVYQALG------ Lh_MaSp2 56 PGSSSCDVLVQALLEIITALISILDSSSVGQVNYGSSGQYAQIVGQSMQQAMG------ Uldiv_MiSp1 50 QGLSSAEATVQALLEVISVLTHMLSSANIGYVDFSRVGDSASAVSQSMAYAG------- Nc_Flag 50 P--NNVNVLMDALLA----ALHCLSNHGSSSFAPSPTPAAMSAYSNS------------ Lh_AcSp1 58 --LSKKEAAIETLMEAMVALLQVLNSAQVNQVDTSSTVVTSSSLAKALSSLF------- Agap_TuSp 56 PSFKADDVVHESLLEGISALIQVINSSYGSPLSLSNAQ----TVNAGLVNYFLV----- Bc_Fib1 55 SGLSPTEMVTEALLEVLAGCMEILSSFNVGAQSISSSRTSSNALVQSISNQFSGLNAAA Konsensus 61 s dvli aLle v ali il s i v s v qsv Abbildung 5: Vergleich der Homologie terminaler Sequenzen von verschiedenen Spidroinarten und Spezies. Um die Vergleichbarkeit der Sequenzkonservierung zu gewährleisten wurden nur Spidroinarten ausgewählt bei denen sowohl die aminoterminale (A) als auch die carboxyterminale (B) Domäne bekannt sind. Hochkonservierte Aminosäuren (>70%) sind schwarz hinterlegt und schwache (>50%) grau. In der Konsensussequenz sind die entsprechenden Aminosäuren nach Konservierungsgrad in Groß- (100%) und Kleinschreibung (>70%) dargestellt. Die Quellsequenzen der NT-Domänen wurden vor dem Vergleich mittels SignalP4.1 auf Signalsequenzen überprüft und ggfs. entfernt. Verwendete Sequenzen und Spezies: Nc_MaSp1: Nephila clavipes MaSp1(NT: B5SYS5; CT: P19837); Lh_MaSp2: Latrodectus hesperus MaSp2 (A6YIY0); Uldiv_MiSp: Uloborus diversus MiSp1 (NT: E1AHV5; CT: Q15G88) Nc_Flag: Nephila clavipes Falgelliform Spidroin (NT: O44358 – modifiziert nach [33]; CT: O44359); Lh_AcSp1: Latrodectus hesperus Aciniform Spidroin1 (GenBank: JX978171.1); Agap_TuSp: Agelenopsis aperta (NT: E1AHV9 CT: E1AHV2); Bc_Fib1: Bothriocyrtum californicum (NT: E1AHU2 CT: A8UAJ7). Die Alignments wurden mit ClustalOmega und Boxshade erstellt. Einleitung [13] 1.5 Major Ampullate Seide Trotz der Vielfalt an interessanten Spinnenseiden ist die Seide der Großen Ampullendrüse (Major Ampullate, MA) die mit Abstand am besten untersuchte. Sie wird für den Rahmen und die Speichen eines Radnetzes sowie als Abseiloder Fluchtfaden in Notsituationen verwendet. Da die Spinne deshalb immer kleines Stück MA-Seide hinter sich herzieht bezeichnet man sie häufig auch als dragline Seide (englisch to drag = ziehen). Der längste je entdeckte Spannfaden aus MA-Seide der Spinne Caerostris darwini ist 25 m lang und besitzt zudem die höchste je berichtete Reißfestigkeit aller Spinnenseidenarten von 1,7 GPa [38]. Neben den mechanischen Werten zeigt dragline-Seide einige weitere interessante Eigenschaften. Eine sich abseilende oder hängende Spinne dreht sich kaum um ihre eigene Achse, was bedeutet, dass die dragline-Seide eine torsionsdämpfende Funktion besitzt. Ähnliches kann man beobachten wenn man den Faden verdreht und anschließend loslässt. Er oszilliert nicht um seine Achse, sondern rotiert langsam in seine ursprüngliche Form zurück, was für eine Art Formgedächtnis spricht. [39]. Eine weitere Eigenschaft ist die sogenannte Superkontraktion, die auftritt sobald die Luftfeuchtigkeit über 60% liegt oder die Seide nass wird. Dabei nimmt der Fadendurchmesser zu und die Seidenfaser schrumpft dabei um bis zu 50% ihrer Länge [40-43]. 1.5.1 Die Proteine der MA-Seide Die MA-Seide besteht hauptsächlich aus zwei Proteinen, die als Major Ampullate Spidroin 1 und 2 (MaSp) bezeichnet werden [8, 9, 44]. Die repetitive Kerndomäne beider Proteine setzt sich im Wesentlichen aus alternierenden polyalaninund glycinreichen Blöcken zusammen (Abbildung 6 A). Ein Polyalaninblock ist zwischen 4 und 13 Resten lang, während die Länge der glycinreichen Region zwischen 12 und 50 Aminosäuren variiert. Die NTund CT-Domänen zwischen beiden Spidroinen sind jeweils hoch homolog [7, 8, 37]. Das Hauptunterscheidungsmerkmal zwischen MaSp1 und 2 ist die Sequenzzusammensetzung der glycinreichen Region. In MaSp1-Proteinen befinden sich in dieser Region hauptsächlich GGX-Motive, während in MaSp2 Proteinen stattdessen GPGXX-Motive zu finden sind. Die einzig bekannten Volllängen MaSp-Proteine stammen aus der schwarzen Witwe (Latrodectus hesperus) sowie der Wespenspinne (Argiope bruennichi) und sind 250 (MaSp1) bzw. 289 - 311 kDa (MaSp2) groß, wobei es diverse Hinweise gibt, dass andere MaSp-Proteine in derselben Größenordnung liegen [8, 45-47]. Einleitung [14] Abbildung 6: Schematische Darstellung von MaSp-Proteinen und Sequenzvergleich repetitiver Blöcke von MaSp 1 und 2 Spidroinen. (A) Schematische, nicht maßstabsgetreue Darstellung des Aufbaus von MaSp-Proteinen. Die repetitive Kerndomäne besteht im Wesentlichen aus alternierenden Polyalanin und glycinreichen Blöcken. Die glycinreiche Region enthält je nach MaSp-Variante entweder hauptsächlich GGX (MaSp1) oder GPGXX-Motive (MaSp2). Die Kerndomäne wird zudem von nicht-repetitiven aminoterminalen (NT) sowie carboxyterminalen (CT) Domänen flankiert. (B) & (C) Sequenzvergleich von repetitiven Blöcken von MaSp 1 (B) und MaSp2 (C) Proteinen verschiedener Spezies. Dargestellt sind repetitive Einheiten, die jeweils zwei Polyalaninblöcke enthalten. Aminosäuren mit einem Konservierungsgrad von >70% sind schwarz hervorgehoben. In der Konsensussequenz sind die entsprechenden Aminosäuren nach Konservierungsgrad in Groß- (100%) und Kleinschreibung (>70%) dargestellt. Die Abkürzungen der verwendete Spezies sowie die UniProt-Nummern der Sequenzen lauten: Lh = Latrodectus hesperus (MaSp1: A6YIY1; MaSp2: A6YIY0), Ea = Euprosthenops australis (MaSp1: A1INM2), Ab = Argiope bruennichi (MaSp1: I6XQ31; MaSp2: I6YNT3), Nc = Nephila clavipes (MaSp1: P19837; MaSp2: P46804), At = Argiope trifasciata (MaSp1: Q9BIU7), Nec = Nephilengys cruentata (MaSp1: A6YP78), Ad = Araneus diadematus (ADF3: Q16987; ADF4: Q16988) A) B) Polyalaninregion GGX-haltige Region Lh_MaSp1 1 -----AAAAAAAAGGAGQGGQGGYGQGGYGQ---------GGAGQGG Ea_MaSp1 1 AAAAAAAAAAAAAGGQGGQGQGRYGQ---GAGSSAAA---------- Ab_MaSp1 1 -----AAAAAAAAGGDGGSGLGGYGA---GRGHGVGLGGAG----GA Nc_MaSp1 1 -------AAAAAAGGAGQRGYGGLGNQGAGR---------GGLGGQG At_MaSp1 1 ------AAAAAAASGAGGAGRGGLGAGGAGQEYGAVSGGQGGAGQGG Nec_MaSp1 1 -----AAAAAASGAGQGGYGGPGAGQ---GAGA-------------- Konsensus aaAAAAaagG Gg G gg Gq G g g Lh_MaSp1 60 AGAAAAAAAAGGAGQGGQGGYGQGGYGQGGAGQGG Ea_MaSp1 61 AAAAAAAAAAAGRGQGGYGQGSGGN---------- Ab_MaSp1 62 GAASAAAAAGGQGGRGGFGGLGSQGAGGAGQGGAG Nc_MaSp1 56 AGAAA-AAAAGGAGQGGYGGLGNQGAGRGGQG--- At_MaSp1 68 EAAAAAAAAGGQGGQGGYGGLGSQGAGQGGYGQGG Nec_MaSp1 42 -----AAAAAGGAGQGGYGGLGGQGAG---QGAGA Konsensus agaaaaAAAaggaGqGGyGglg qgag gg g gg C) Polyalaninregion GPGXX-haltige Region Lh_MaSp2 1 ---AAAAAA-----AAAGSGPSGYG------PGAAAAA---A--AGS------- At_MaSp2 1 AAAAAAAAGPGYGPGAGQQGPGSGGQQGGPGSGQQGPGGAGQGGPRGQGPYGPG Ab_MaSp2 1 -AAAAAAAAGGYGPGAGQQGPGSQG------PGSGGQ-----QGPGGQGPYGPG Nc_MaSp2 1 AAAAAAAAS-----GPGQQGPGGYG------PGQQGPG---GYGPGQQGPSGPG Ad_ADF3 1 ---AAAAAAGGNGPGSGQQGAGQQG------PGQQGP----------------- Ad_ADF4 1 -----AAAA-----AAAASGPGGYG------PGSQGPS---G--PGVYGPGGPG Konsensus aaAAAa gagqqGpg G pG qgp pg gp gpg Lh_MaSp2 29 ------------------------AGPGTQQGYGPGGSG------ At_MaSp2 65 AAAAA-AAAAGGYGPGAGQQGPGSQGPGSGGQQGPGSQGPYGPSAb_MaSp2 55 AAAAAA AV--GGYGPGAGQQGPGSQGPGSGGQQGPGGQGPYGPSNc_MaSp2 53 SAAAAA AAASGPGQQGPGGYGPGQQG---PGGYGPGQQGLSGPGS Ad_ADF3 39 -GASAA AAAAGGYGPGSGQQGPGQQGPGGQGPYGPGAS------- Ad_ADF4 46 SSAAAAAA----AGSGPGGYGPENQGPSGPGGYGPGGSGSS---- Konsensus aaaaa aa g g g g gpg qGpg g yGPGg g (A) 4-13 (A) 4-13 (A) 4-13 glycinreiche Region MaSp1: GGX MaSp2: GPGXX NTD CTD Polyalaninregion n Einleitung [15] Die Sequenzkonservierung der repetitiven Kerndomäne von MaSp Proteinen ist vergleichsweise gering, da Reihenfolge und vor allem Anzahl der Wiederholungen der GGX und GPGXX-Motive innerhalb der glycinreichen Region zum Teil stark variieren (Abbildung 6 B, C). Beide MaSp-Proteine liegen im finalen Faden vor, jedoch variiert das Verhältnis der beiden Proteine speziesübergreifend zum Teil stark [44]. Die jeweilige Menge der beiden Spidroine im Faden ist zudem von vielerlei Umweltbedingungen wie der Nahrungsquellen oder dem Ernährungszustand abhängig [48-50]. Trotz variabler Anteile überwiegt in den meisten bisher untersuchten dragline Seiden der MaSp1-Anteil. Die einzige Ausnahme stellt die MA-Seide der Gartenkreuzspinne Araneus diadematus dar, die einen hohen Prolinanteil von 16% aufweist [7]. Dort besitzen beide bekannte MaSpProteine, genannt Araneus diadematus Fibroin 3 und 4 (ADF3 u. 4) in repetitiven Blöcken das GPGXX Motiv und sind demnach MaSp2-Analoga [7]. 1.5.2 Struktur der repetitiven Kerndomäne Die Struktur-Funktionsbeziehung der repetitiven Kerndomäne von MaSp-Proteinen wurde mit einer Vielzahl an Methoden, wie Festkörper-NMR, Röntgenstreuung oder IRSpektroskopie untersucht. Schon relativ früh konnte nachgewiesen werden, dass die Polyalaninmotive im Faden in einer β-Faltblattkonformation vorliegen [51]. Neuere Untersuchungen haben gezeigt, dass das Motiv antiparallele β-Faltblattkristalle mit einer Größe von 5x2x7 nm (Länge x Breite x Höhe) ausbildet [52-57]. Die Faltblätter im Kristall sind dabei parallel zur Faserachse angeordnet [58-60]. Die Strukturbestimmung der GGX-Motive innerhalb der Faser hingegen ist weniger eindeutig. Das Motiv kommt sowohl in β-Faltblatt als auch in weniger geordneten, αhelikalen Bereichen vor [56, 61-63]. Es gibt zudem Hinweise, dass das Motiv in Polyglycin-II-Helices vorliegt [56, 61, 64]. Zum Pentapeptid GPGXX liegen die wenigsten Strukturdaten vor. Dem Motiv wird aufgrund des Prolins eine spiralförmige Polyprolin-II-helikale Struktur zugeordnet, wie sie auch von ähnlichen repetitiven Sequenzen (z.B. VPGVG) in Elastin gebildet wird [65]. Diesem Motiv wird zudem eine wichtige Rolle bei der Superkontraktion der draglineFaser zugeschrieben, da Hinweise existieren, dass die Hydratisierung dieser prolinreichen Sequenz zu einer Konformationsumlagerung führt, die das Schrumpfen der Seide bewirkt [41-43, 66-71]. Einleitung [22] 1.6 Rekombinante Spinnenseidenproteine Spinnen weisen ein kannibalisches Revierverhalten auf, weshalb die industrielle Zucht, wie sie z. B. für den Seidenspinner Bombyx mori betrieben wird, als ökonomisch sinnvolle Quelle für Spinnenseiden nicht möglich ist. Um verwertbare Mengen an Spinnenseide zu gewinnen, können jedoch Spinnenseidenproteine rekombinant produziert werden, was in zwei unterschiedlichen Ansätzen verfolgt wird. Der erste Weg stellt die heterologe Produktion von natürlichen Sequenzen in verschiedenen Wirtsorganismen dar, wie z.B. eukarotischen Zellen, Tabakpflanzen und transgenen Seidenspinnern, Mäusen oder gar Ziegen [114-120]. Dabei werden häufig stark verkürzte 3‘-cDNA-Fragmente der Spinnenseiden verwendet. Die Ausbeuten bei der jeweiligen Produktion sind jedoch vergleichsweise gering, so dass keine dieser Ansätze über Machbarkeitsstudien hinausgekommen sind. Ein Grund für die niedrigen Ausbeuten liegt möglicherweise in der speziellen Codonverteilung der Spinnenseidengene [121]. Besonders deutlich wird dies bei den häufig vorkommenden Aminosäuren Glyzin und Alanin, die nur von einer kleinen Anzahl an möglichen Codons codiert werden. Da die Spidroine hohe Anteile von lediglich zwei bis drei Aminosäuren (z.B. Glycin und Alanin) aufweisen, wird vermutet, dass ein beschränkter aber dafür hochverfügbarer Pool an tRNA für die Spinne effizienter ist [8, 122]. Der zweite und wesentlich häufiger verwendete Ansatz rekombinante Spinnenseidenproteine zu gewinnen, ist die Verwendung von artifiziellen Spinnenseidengenen. Die repetitiven Blöcke dieser künstlichen Spinnenseiden bestehen häufig aus Konsensusmotiven der natürlichen Spidroinsequenzen, weshalb die künstlichen Spinnenseidenproteine eine hohe Ähnlichkeit zu ihren natürlichen Vorbildern aufweisen [121, 123, 124]. Die Verwendung von künstlichen Genen erlaubt zudem die Anpassung der Sequenz an die Codon-Verwendung des gewünschten Produktionsorganismus. In den meisten Fällen erfolgt die Herstellung der artifiziellen Spidroine in E. coli, vereinzelt auch in Hefen wie Pichia pastoris oder in Tabakpflanzen [114, 121, 123, 125, 126]. Die meisten produzierten Proteine haben ein Molekulargewicht von 25 bis 120 kDa und sind damit wesentlich kleiner als die natürlichen Spidroine. Die Produktion von rekombinanten Spidroinen mit einem Molekulargewicht von mehr als 200 kDa ist möglich, jedoch mit aufwändigen Eingriffen in den Metabolismus von E. coli verbunden [127]. Einleitung [23] Die in dieser Arbeit verwendeten rekombinanten Seiden basieren auf den beiden MaSp2Proteinen ADF3 und 4 der Gartenkreuzspinne Araneus diadematus und werden zu Beginn von Kapitel 2 im Detail beschrieben [7, 124]. 1.7 Assemblierungseigenschaften von rekombinanten Spinnenseidenproteinen Für diverse rekombinante Spinnenseideproteine existieren einer Vielzahl an nichtnatürlichen Assemblierungsformen wie Filme, Hydrogele, Kapseln oder Kugeln. Die Charakterisierung, Funktionen und vielfältigen Einsatzgebiete dieser Morphologien sind für diese Arbeit nicht relevant und anderer Stelle sehr ausführlich beschrieben, weswegen nun auf die Faserassemblierung eingegangen wird [128-138]. 1.7.1 Fasern aus rekombinanter Seide Die mechanischen Eigenschaften der natürlichen Spinnenseidenfasern sind besonders bezogen auf ihr Gewicht und die Dichte besonders attraktiv für verschiedene Anwendungsfelder. Die Herstellung einer Faser aus den rekombinanten Seiden stellt daher ein wichtiges Ziel dar. Mehrere Methoden zur Produktion einer Faser sind beschrieben, wobei die mit Abstand am häufigsten angewandte das sogenannte Nassspinnen (engl. wet spinning) darstellt [115, 139-142]. Bei diesem Prozess wird eine konzentrierte Spinnlösung (>10 % w/v) meist durch eine Injektionsnadel mit einem dünnen Querschnitt in ein flüssiges Fällbad injiziert. Das Fällbad enthält häufig hohe Anteile an Alkoholen wie Ethanol, Methanol oder Isopropnaol welche ein schnelles Assemblieren bzw. Aggregieren der Proteine bewirken. Die so gewonnenen Fasern sind den natürlichen Seiden hinsichtlich ihrer mechanischen Eigenschaften jedoch zum Teil weit unterlegen (Tabelle 4). Die Zähigkeit der Fasern aus rekombinanten Spidroinen liegt häufig eine Größenordnung unter den Natürlichen. Eine mögliche Erklärung dafür liegt in der Verwendung von Proteinen, die deutlich kürzer sind als die natürlichen. Zudem werden häufig rekombinante Spidroine verwendet die nur aus repetitiven Blöcken bestehen und nur vereinzelt die CTDomäne enthalten. Ein weiterer Grund für die schlechteren mechanischen Eigenschaften liegt womöglich im künstlichen Spinnprozess selbst. Um eine hochkonzentrierte Spinnlösung herzustellen werden die rekombinanten Seiden meist in organischen Lösungsmitteln, wie Hexafluoroisopropanol (HFIP) gelöst. Einleitung [24] Tabelle 4: Mechanische Werte von Fasern auf Basis von rekombinanten Spinnenseidenproteinen im Vergleich mit natürlicher MA-Seide. Die Werte wurden direkt aus den entsprechenden Quellen entnommen. Material Molekülmasse (kDa) Reisfestigkeit (GPa) Dehnbarkeit (%) Zähigkeit (MJm-3) Quelle Major AmpullateSeide 250 1,1 27 160 [4] 1) Rekomb. MaSp2 71 0,045 ± 0,008 3,9 ± 2,6 - [143] 2) Rekomb. MaSp1 285 0,5 ± 0,11 15 ± 5 - [127] 3) Rekomb. MaSp1 70 0,132 ± 0,049 22,78 ± 19.06 23,7 ± 18,5 [144] 4) Rekomb. MaSp2/Flag 61 0,049 ± 0,019 34,06 ± 25,30 10,6 ± 10,2 [141] 5) Rekomb. MaSp2 106 0,0646 ± 0,026 10,8 ± 3,1 - [145] 6) Rekomb. MaSp1 24 0,198 ± 0,019 - - [22] Durch das anschließende Injizieren in ein Fällbad mit mehr als 90 % Alkohol (z. Isopropanol) entstehen Fasern, die kaum Wasser enthalten und dadurch sehr spröde und brüchig werden. Hinzu kommt, dass die Proteine in HFIP eine unnatürliche hauptsächlich αhelikale Sekundärstruktur annehmen, und somit eine Vorstrukturierung, wie sie im natürlichen Spinntrakt vorliegt, nicht möglich ist. Für Fasern aus rekombinanten Seidenproteinen, die sich in wässrigem Milieu ausbilden, sind lediglich zwei Untersuchungen bekannt. Im ersten Fall bilden sich meterlange Fasern aus einem rekombinanten „Mini-Spidroin“ (ca. 24 kDa) in einem Selbstassemblierungsprozess aus einer relativ gering konzentrierten Lösung ohne äußere Einflüsse [22]. Die mechanischen Eigenschaften liegen jedoch ebenfalls eine Größenordnung unter den natürlichen Seiden (vgl. Tabelle 4, Material 6). Im zweiten Fall werden die Fasern auf Basis von rekombinanten ADF3-Proteinen mit Hilfe einer Mikrofluidikanlage gewonnen, die dem natürlichen Spinnkanal nachempfunden ist [134]. Dazu werden die in wässriger Lösung befindlichen Proteine durch ein modulares System aus dünnen Kanälen gepumpt. Das Verspinnen erfolgt, ähnlich wie natürlichen Spinntrakt, durch Absenken des pH-Wertes von 8 auf 6, der Zugabe von Phosphationen sowie einem Elongationsfluss initiiert. Die resultierenden Fasern sind wasserunlöslich und zeigen unter kreuz-polarisiertem Licht Doppelbrechung wie sie auch bei natürlichen MA-Seidenfäden beobachtet wurde, was für stark strukturierte bzw. kristalline Bereiche spricht [115, 146-148]. Einleitung [25] 1.8 Zielsetzung Zu Beginn dieser Arbeit waren hauptsächlich einzelne Aspekte des natürlichen Spinnprozesses, wie die strukturellen Umlagerungen der für MA-Spidroine typischen Aminosäuremotive (A)n, GGX und GPGXX, besser untersucht. In einer vorangegangenen Arbeit konnten bereits erste Erkenntnisse über das Assemblierungsverhalten von rekombinanten eADF3 und 4 Proteinen unter ähnlichen Bedingungen (pH-Absenkung, Zugabe von Phosphationen, Elongationskräfte) wie sie im natürlichen Spinnkanal auftreten, gewonnen werden [134]. Über die Funktionen der nicht-repetitiven terminalen Domänen während der Speicherung und des Verspinnens Domänen hingegen waren nicht viel bekannt. Um die Rolle der terminalen Domänen näher zu charakterisieren, wurden diese mit künstlichen repetitiven Domänen auf Basis der Proteine ADF3 und 4 gekoppelt, deren Löslichkeitsund Assemblierungseigenschaften zum Teil bereits charakterisiert waren [124, 132, 134, 135, 149]. Die Untersuchung der Rolle der beiden terminalen Domänen erfolgte in drei unterschiedlichen Teilprojekten. Der umfangreichste Teil stellt die Charakterisierung der nicht-repetitive carboxyterminale Domäne NR3 des Spinnenseidenproteins ADF3 der Gartenkreuzspinne Araneus diadematus dar. Als Basis dienten vorangegangene Arbeiten in denen bereits erste Erkenntnisse über den Einfluss von NR3 auf rekombinante Spidroine auf Basis von ADF3 gewonnen werden konnten [124, 134, 149]. Ein wesentlicher Teil der Arbeit mit der NR3-Domäne bestand darin die Struktur-Funktionsbeziehung besser zu charakterisieren. Die Struktur der Domäne sollte dabei mittels NMR in Kooperation mit Prof. Dr. Kessler und Dr. Franz Hagn (TU München) untersucht werden. Die Funktion der NR3-Domäne sollte mit Hilfe von rekombinanten eADF3 Proteinen analysiert werden, wobei der Schwerpunkt dabei auf der Analyse des Löslichkeitsund Assemblierungsverhalten, unter Bedingungen wie sie auch während des natürlichen Spinnprozesses auftreten, lag. So sollten die Proteine unter Scherstress oder in Anwesenheit von kosmotropen Salzen wie Kaliumphosphat untersucht werden. Da die Spidroine der Großen Ampullendrüse aller Wahrscheinlichkeit nach in den gleichen Zellen produziert werden [8, 35], sollten in einem zweiten Teilprojekt die Untersuchungen auf ein zweites rekombinantes Spinnenseidenprotein auf Basis von ADF4 ausgedehnt werden. Ein besonderer Fokus lag dabei auf möglichen Interaktionen, die durch die stark homologen CT-Domänen NR3 und NR4 vermittelt werden. Einleitung [26] Da natürliche MA-Spidroine zwei terminale Domänen aufweisen, sollten in einem dritten Teilprojekt rekombinante Seidenproteine mit NTund CT-Domäne untersucht werden. Die NT-Domäne bildet pH-abhängig ein antiparalleles Dimer [34, 79, 85], weshalb besonders das Löslichkeitsund Assemblierungsverhalten bei pH-Bedingungen wie sie in der Drüse vorkommen, untersucht wurden. Da die Sequenzen der NT-Domänen von ADF 3 und 4 nicht bekannt sind, wurden stattdessen diejenigen der MaSp-Proteine der schwarzen Witwe Latrodectus hesperus verwendet werden [8]. Synopsis [27] 2. Synopsis 2.1 Übersicht der verwendeten Proteine Die in dieser Arbeit verwendeten rekombinanten Seiden basieren auf den beiden MaSp2Proteinen ADF3 und 4 der Gartenkreuzspinne Araneus diadematus [7, 124]. Von den repetitiven Blöcken in ADF3 und 4 wurden entsprechende Konsensusmotive abgeleitet, die modulartig mittels eines speziell entwickelten Klonierungssystems nahtlos multimerisiert werden können. Aus den repetitiven Blöcken in ADF3 wurden die Module A (Alaninreich) und Q (Glutamin (Q)-reich) generiert, während aus ADF4 das Modul C (Carboxylgruppe des Glutamats) abgeleitet wurde (Abbildung 11). Abbildung 11: Übersicht der in dieser Arbeit verwendeten rekombinanten Spinnenseidenproteine. Aus den repetitiven Blöcken von (A) ADF3 wurden die Module A und Q, aus (B) ADF4 das Modul C abgeleitet. Die Module können nahtlos multimerisiert werden und mit den Sequenzen der nativen carboxyterminalen Domänen NR3/ NR4 aus ADF3/4 verknüpft werden. Neben den ADF-Termini wurden zudem die beiden terminalen Domänen von Masp1 der schwarzen Witwe (L. hesperus) in diverse eADF3-Proteine eingebaut. Alle N1-haltigen Proteine liegen auch in einer Variante mit einer 32 Aminosäure langen Linkersequenz vor, die sich am carboxyterminalen Ende der N1-Domäne befindet. GSSAAAAAAAASGPGGYGPENQGPSGPGGYGPGGP Modul A A) B) MaSp1 L . hesperus Modul C Modul Q CTD(NR3) CTD(C1)NTD(N1) Linker* * = optionale Linkersequenz: CTD(NR4) ADF4 ADF3 GPYGPGASAAAAAAGGYGPGSGQQ GPGQQGPGQQGPGQQGPGQQ eADF3((AQ)24) (98 kDa) eADF4(C16NR4) (58,2 kDa) eADF4(C16) (46 kDa) eADF3((AQ)12NR3) (59,8 kDa) eADF3(N1(AQ)C1) (29,6 kDa) eADF3(N1L(AQ)C1) (33,7 kDa) eADF3(N1(AQ)12NR3) (73,5 kDa) eADF3(N1L(AQ)12NR3) (76,4 kDa) eADF3(N1(AQ)12) (61,8 kDa) eADF3(N1L(AQ)12) (65,9 kDa) eADF3((AQ)24NR3) (110 kDa) SGGGGYGASSASAASASAAAPSGVAYQAPAQAQISFTLRGQQPVS Synopsis [28] Die repetitiven Einheiten können mit den jeweiligen nativen Sequenzen der carboxyterminalen Domänen NR3 und NR4 aus ADF3 respektive ADF4 kombiniert werden. Die rekombinanten Proteine werden als eADF3 und 4 (engineered ADF3 und 4) bezeichnet [124]. Da die aminoterminale Domäne von ADF3 und 4 nicht bekannt sind, wurde stattdessen diejenige von MaSp1 der schwarzen Witwe (L. hesperus), hier als N1 bezeichnet, verwendet. Die dazugehörige carboxyterminale C1 stand ebenfalls zur Verfügung. Die Sequenzen der beiden MaSp1-Termini wurden auf Basis der Veröffentlichung von Ayoub et al entworfen [8]. 2.2 Die NR3-Domäne 2.2.1 Lösungsstruktur der NR3-Domäne Die Struktur der carboxyterminalen Domäne NR3 des MA-Proteins ADF3 der Spinne Araneus diadematus konnte mittels NMR gelöst werden. Die Aufklärung der Struktur sowie Teile der biophysikalischen Charakterisierung der NR3-Domäne wurden hauptsächlich von Franz Hagn durchgeführt [150] (Abbildung 12). Die nicht-repetitive carboxyterminale Domäne NR3 des natürlichen Spinnenseidenproteins ADF3 bildet in Lösung ein durch eine Cysteinbrücke kovalent verknüpftes, parallel orientiertes Dimer (Abbildung 12). Jedes Monomer besteht aus fünf α-Helices, die die Form einer gekrümmten Hand annehmen. Helix 1 ragt dabei jeweils klammerartig an das andere Monomer heran und interagiert dort mit Helix 5 (Abbildung 12 A). Die Reste zwischen Helix 4 und 5 bilden zudem einen Bogen aus in dem der carboxyterminale Teil von Helix 4* liegt. Trotz paralleler Anordnung der Monomere zeigen die aminoterminalen Enden von Helix 1 und 1*, welche die Verbindung zu den repetitiven Bereichen des Seidenproteins darstellen, in einem Winkel von etwa 100° relativ zueinander von der NR3-Domäne weg. Das zentrale Dimerisierungsinterface um Helix 4 stellt mit mehreren Leucinen und Valinen den hydrophobsten Teil des Proteins dar (Abbildung 12 C, D). Die Interaktion der hydrophoben Seitenketten wird durch eine leichte Verdrehung der beiden zentralen Helices zueinander verstärkt (Abbildung 12 A). Am aminoterminalen Ende von Helix 4 befindet sich ein konserviertes Cystein, das beide Monomere über eine Disulfidbrücke kovalent verknüpft [24, 36, 151]. In thermischen Entfaltungsexperimenten unter reduzierenden Bedingungen oder mit carboxymethylierten Cysteinresten konnte gezeigt werden, dass die Disulfidbrücke das Protein signifikant stabilisiert. Jedes Monomer enthält zudem zwei Salzbrücken (R43-E101, R52-D92), die jeweils Helix 1, 2 und 4 verbinden (Abbildung 12 B). Synopsis [29] Abbildung 12: Lösungsstruktur von NR3. A) Aufund Seitenansicht des NR3-Dimers in vereinfachter Darstellung. Die Helices von Monomer 1 (blau) sind mit H1-H5 gekennzeichnet, die von Monomer 2 (rot) mit H1*-H5*. Die Seitenketten der Disulfidbrücke (Cys92) sowie der Salzbrücken R43-D93 und R52-E101 sind gezeigt. B) Vergrößerung des Bereiches um die Salzbrücken sowie des Cysteins in einem Monomer. C) Darstellung eines NR3-Monomers innerhalb der dimeren Struktur. Die Salzbrücken der vier geladenen Aminosäuren sind eingezeichnet. D) Moleküloberfläche des Monomers aus (C) eingefärbt nach der Hydrophobizitätsskala nach Kyte & Doolittle [152]. Grau: hydrophobe Reste; Blau/Dunkelblau: hydrophile Reste. Die Strukturen wurden mit PyMOL gezeichnet. Diese vier an der Ausbildung der ionischen Bindungen beteiligten Aminosäuren sind die einzigen geladenen in der gesamten bekannten natürlichen ADF3-Sequenz und ebenso wie das Cystein hochkonserviert innerhalb der carboxyterminalen Domänen der MaSp-Proteine [7]. Um die Bedeutung der Salzbrücken näher zu beleuchten wurden thermische Entfaltungsexperimente mit NR3-Mutanten durchgeführt, bei denen jeweils eine geladene Aminosäure gegen Alanin ausgetauscht wurde. Sämtliche Alanin-Mutanten weisen dabei zum Teil erheblich reduzierte Schmelzpunkte auf (Teilarbeit I, Supplement-Abbildung 2 B). Auch eine Umkehr der Polung der Salzbrücken z. B. von R43-D93 zu D43-R93 führte zu einer Reduktion des Schmelzpunktes um ca. 20°C, was zeigt, dass der Bereich um die Salzbrücken sehr sensitiv in Bezug auf die Seitenkettengröße ist. H1* S-S Brücke H1 H2* H3* H3 H2 H4* H4 H5* H5 S-S Brücke H2 R52 R43 D93 E101 H3 H3* H2* H4 H4* NH3 NH3* H5 H1 H2 H3 H4 R52 R43 D93 C92 E101 AufsichtA) B) C) D) Seitenansicht H1 H2 R52 R43 D93 C92 E101 H3 H4 H5 Synopsis [30] Mit Experimenten, die Änderungen der chemischen Verschiebungen einzelner Reste detektieren (chemical shift pertubations, CSP), konnte nachgewiesen werden, dass dieser Bereich des Moleküls besonders flexibel ist. In Anwesenheit von Harnstoff ist die Dynamik in diesem Teil des Moleküls am größten, was darauf schließen lässt, dass das Protein in diesem Bereich zu entfalten beginnt. In weiteren CSP-Experimenten konnte daher auch gezeigt werden, dass der für hydrophobe Bereiche sensitive Farbstoff AnilinonaphthalenSulfonsäure (ANS) bevorzugt in diesem Bereich bindet (Teilarbeit I, Abbildung 3). 2.2.2 Einfluss von Salz und Scherkräften auf das NR3-Dimer Da die Speicherung der hochkonzentrierten Spidroinlösung in einem Natriumchloridhaltigen Milieu erfolgt [107], wurde der Einfluss des Salzes auf die Struktur und insbesondere auf die Region um die Salzbrücken eingehender charakterisiert. Dazu wurde in chemischen Entfaltungsexperimenten die Stabilität der NR-Domäne bei steigender Proteinkonzentration in Anund Abwesenheit von Natriumchlorid untersucht. Mit steigender Proteinkonzentration von 5 auf 60 µM sinkt der Übergangspunkt bei Harnstoff-induzierter Entfaltung von ca. 2,8 M auf ca. 2 M Harnstoff. Die Steigung der Kurven und somit die Kooperativität des Entfaltungsvorganges nimmt dabei stetig ab, gezeigt durch den Verlust des typischen sigmoidalen Verlaufs bei 60 µM Proteinkonzentration [150]. Wird das Entfaltungsexperiment bei einer Proteinkonzentration von 60 µM in Anwesenheit von 250 mM Natriumchlorid durchgeführt, ist eine Stabilisierung des Proteins zu beobachten, wie an dem deutlich kooperativeren Übergang sowie einem leicht erhöhten Übergangspunkt zu erkennen. Durch Erhöhung der NaCl-Konzentration auf 500 mM steigt die Stabilität des Proteins sogar bei einer Konzentration von 60 µM wieder auf einem Niveau mit der 5 µM-Proteinlösung (Teilarbeit I, Abbildung 3 B). Ein ähnlicher NaCl-abhängiger Effekt ist bei thermischen Entfaltungsexperimenten mit NR3 in stark reduzierendem Milieu zu beobachten, in dem sich die Disulfidbrücke nicht ausbilden kann. Reduziertes NR3 zeigt einen ca. 20°C niedrigeren Schmelzpunkt als das oxidierte, kovalent verknüpfte NR3. Die Anwesenheit von NaCl eliminiert den destabilisierenden Einfluss der fehlenden Disulfidbrücke nahezu vollständig. Ein Grund für die beiden beobachteten NaCl-abhängigen Effekte kann eine Stabilisierung des hydrophoben Bereiches um das Cystein sein. Wie bereits erwähnt ist dieser Bereich des Moleküls am dynamischsten und zugleich durch hydrophobe Bereiche geprägt. Wie bei anderen Proteinen bereits beobachtet, können durch steigende Salzkonzentration die Synopsis [31] hydrophoben Interaktionen gestärkt werden, was die erhöhte Stabilität bei erhöhter Proteinkonzentration und unter reduzierenden Bedingungen erklärt [153-155]. Nach der Speicherung bei hohen Konzentrationen sind die natürlichen Spinnenseidenproteine unter anderem Scherkräften im Spinnkanal ausgesetzt, weshalb auch deren Einfluss auf die NR3-Domäne untersucht wurde. Scherkräfte wurden entweder durch „Vortexen“ oder durch Überkopf-Rotation bei 20-25 Umdrehungen pro Minute erzeugt. Die Experimente wurden zudem in Anwesenheit von 50 und 500 mM NaCl durchgeführt, um den Einfluss von Salz mit zu untersuchen. Es zeigte sich, dass ca. 50% von NR3 selbst bei langsamer Rotation über einen Zeitraum von 18-24 h aggregiert. Im Vergleich zu den thermischen Entfaltungsexperimenten ist durch die Anwesenheit von 500 mM NaCl kein ausgeprägter Stabilisierungseffekt zu erkennen. Durch Anfärben mit dem für hydrophobe Bereiche sensitiven Farbstoff ANS wird deutlich, dass die scherkraftinduzierte Aggregation mit der Exposition von hydrophoben Bereichen einhergeht (Teilarbeit I, Abbildung 3D) [150]. 2.2.3 Zusammenfassung: NR3-Domäne Während des Spinnprozesses steigt die Proteinkonzentration im Spinnkanal durch Wasserentzug stetig an, während die Natriumchloridkonzentration immer mehr abnimmt. Das während der Speicherung in der Drüse vorhandene Natriumchlorid [107] stabilisiert die NR3-Domäne bei steigenden Proteinkonzentrationen und kann somit unerwünschte Aggregation unterdrücken. Scherkräfte, wie sie während des Verspinnens auftreten, fördern hingegen die Aggregation der NR3-Domäne. Zusätzlich nimmt die NaCl-Konzentration während des Verspinnens ab, was eine gleichzeitige Destabilisierung der NR3-Domäne bewirkt und dadurch möglichweise die Assemblierung weiter beschleunigt. Sowohl bei der NaCl-abhängigen Stabilisierung wie auch der scherkraftinduzierten Aggregation sind Aminosäuren um das hochkonservierte Cystein sowie den Salzbrücken entscheidend involviert. 2.3 Einfluss von NR3 auf das Lösungsverhalten von eADF3((AQ)n) Proteinen 2.3.1 Bildung von supramolekularen Assemblaten Um die Funktion von NR3 besser zu verstehen, wurde die Domäne mit rekombianten Varianten der Kerndomäne des natürlichen Spinnenseidenproteins ADF3 kombiniert und deren Lösungsund Assemblierungsverhalten untersucht (vgl. Abb. 11). Die rekombinante Synopsis [38] 2.4 Heterodimerisierung von NR3-NR4-Konstrukten 2.4.1 Bildung des Heterodimers Die bisherigen Untersuchungen haben sich lediglich auf eines der beiden ADF-Proteine aus der dragline-Seide von A. diadematus konzentriert. Da die beiden Proteine ADF3 und ADF4 aller Wahrscheinlichkeit nach in der Großen Ampullendrüse in den gleichen Zellen produziert werden [8, 35], wurden die Untersuchungen auf das zweite bekannte Spinnenseidenprotein auf Basis von ADF4 ausgedehnt. Der Fokus lag hierbei auf möglichen Interaktionen vermittelt durch die homologen CT-Domänen NR3 und NR4. ADF3 und 4 unterscheiden sich hauptsächlich in ihrer repetitiven Kerndomäne. eADF4Proteine sind hydrophober als eADF3-Proteine und zeigen ein anderes Löslichkeitsund Assemblierungsverhalten [134, 135, 160]. So sind eADF3-Proteine in wässriger Lösung bis zu 30% (w/v) löslich, während eADF4-Proteine bis maximal 9% (w/v) löslich sind [124]. Die nicht-repetitiven carboxyterminalen Domänen (CTD) NR3 und NR4 hingegen sind zueinander, wie alle MaSp-CTD-Sequenzen, stark homolog. Daher stellte sich die Frage der natürlichen Interaktion von eADF3-NR3 und eADF4-NR4. Dazu wurden die beiden Proteine eADF3((AQ)12NR3) und eADF4(C16NR4) sowohl in vitro nach individueller Produktion als auch in vivo nach Coexpression der entsprechenden Gene untersucht. Beide Proteine bilden schon während der Produktion in E. coli über ihre carboxyterminalen Domänen kovalent verknüpfte Dimere aus, wobei der größte Teil der Proteine in der löslichen Fraktion nach Zellaufschluss zu finden ist. Die Coexpression hat auf die Löslichkeit der einzelnen Proteine keinen Einfluss. Mittels SDS-PAGE und anschließendem Western Blot konnte in den Coexpressionsproben eine Bande identifiziert werden, die eindeutig einem carboxyterminal vermitteltem Heterodimer zugeordnet werden konnte (Teilarbeit III, Abbildung 2b). Solch ein Heterodimer kann man auch durch in vitro Experimente erhalten, indem man zu einer 1:1 Mischung chemisch denaturiertem eADF3((AQ)12NR3) und eADF4(C16NR4) ein starkes Reduktionsmittel wie TCEP gibt und die Proteine unter reduzierenden Bedingungen zurückfaltet. Allerdings ist die Menge an gebildeten Heterodimer in vitro wesentlich geringer als in vivo, wo eine Verteilung der drei dimeren Spezies von nahezu 1:1:1 beobachtet wurde (Teilarbeit III, Abbildung 2c). Synopsis [39] 2.4.2 Assemblierungseigenschaften des Heterodimers Um das Heterodimer näher zu charakterisieren, wurde es mittels einer zweistufigen Reinigungsstrategie, bestehend aus einer Nickel-Affinitätschromatografie und einem Anionentauscherschritt gereinigt. Die strukturelle Integrität des Heterodimers wurde mittels Fern-UV-CD-Spektroskopie untersucht. Das Fern-UV-CD-Spektrum zeigt keine signifikanten Unterschiede zu denen der beiden Homodimeren (Teilarbeit III, Abbildung 2d). Die Struktur des Heterodimers wird reversibel, genau wie die beiden Homodimere, nach denaturierendem Erhitzen und Abkühlen quantitativ renaturiert. Die Schmelztemperatur des Heterodimers liegt mit 66°C zwei Grad über der des (AQ)12NR3-Dimers und 1.5 Grad unter der des C16NR4-Dimers. In einem weiteren Experiment wurde die in vitro Löslichkeit der drei dimeren Proteine bei Phosphat-induzierter Aggregation bzw. Assemblierung untersucht. Die Experimente wurden analog zu den bereits gezeigten aus Teilarbeit I und II durchgeführt. Dieser Assay zeigt, dass die beiden Homodimere (AQ)12NR3 und C16NR4 ca. im gleichen Phosphatkonzentrationsbereich von 150 – 300 mM am stärksten aggregieren. In Assemblierungskinetiken, die mittels Lichtstreuung in Anwesenheit von 100 mM Phosphat aufgenommen wurden, zeigte sich, dass C16NR4 wesentlich schneller assembliert als (AQ)12NR3. Das Heterodimer assembliert interessanterweise nur geringfügig langsamer als C16NR4. Werden die Assemblierungskinetiken in Anwesenheit des β-Faltblatt detektierenden Farbstoffs ThioflavinT (ThT) durchgeführt, zeigte (AQ)12NR3 signifikant weniger ThT-Bindung als die anderen beiden Dimere (Teilarbeit III, Abbildung 3c). 2.4.3 Assemblierungsmorphologien der Homodimere im Vergleich zum Heterodimer Ein wesentliches Unterscheidungsmerkmal zwischen eADF3 und eADF4-Proteinen ist ihr Selbstassemblierungsverhalten. Es konnte bereits gezeigt werden, dass eADF4(C16), ein Derivat ohne die NR4-Domäne, in Anwesenheit von bis zu 300 mM Kaliumphosphat Nanofibrillen ausbildet. Diese Fibrillen haben einen hohen β-Faltblattanteil [135]. eADF3Proteine hingegen bilden keine fibrillären Strukturen in dieser Größenordnung aus sondern neigen in Anwesenheit von Phosphat zur Ausbildung supramolekularer, mizellartiger Assemblate (Teilarbeit II). Um die Selbstassemblierungseigenschaften des AQ/CHeterodimers zu untersuchen, wurde es in Anwesenheit von 50 – 100 mM Kaliumphosphat für 1-2 Tage inkubiert und anschließend mittels Rasterkraft- (AFM) und Transmissions- Synopsis [40] elektronenmikroskopie (TEM) untersucht (Abbildung 17 & Teilarbeit III, Supplement Abbildung 1). Abbildung 17: AFM-Aufnahmen der drei dimeren Spezies. (A) C16NR4, (B) (AQ)12NR3 und (C) Heterodimer. Die Proteine wurden vor den Aufnahmen für 24 h in Tris/HCl-Puffer, pH 8,0 und 50 mM Kaliumphosphat bei Raumtemperatur inkubiert. Die Bilder wurden im tapping mode erstellt. C16NR4 bildet unter den hier getesteten Bedingungen Fibrillen von mehreren hundert Nanometern Länge, die denen der bereits beschriebenen C16_ΔNR4-Fibrillen sehr ähnlich sind [135]. (AQ)12NR3 bildet nicht-fibrilläre Morphologien aus, die eher an bereits beschriebenen sphärische, mizellartige Assemblate erinnern (Teilarbeit II). In Transmissionselektronenmikroskopischen Aufnahmen konnten diese Assemblate ebenfalls nachgewiesen werden (Teilarbeit III, Supplement Abbildung 1). Das Heterodimer hingegen bildet ebenfalls Nanofibrillen aus, die sich jedoch von C16NR4-Fibrillen signifikant unterscheiden. Sie sind wesentlich dünner und bilden ein hochverzweigtes Netzwerk, wobei die Abstände zwischen den einzelnen Verzweigungspunkten manchmal nur wenige Nanometer betragen. 2.4.4. Bedeutung des Heterodimers Beide MaSp-Proteine werden in der Drüse im selben Bereich sekretiert [8, 35, 161]. Genomische Untersuchungen an regulatorischen Elementen der Spidroingene deuten auf eine gemeinsame Regulierung der Gene/Proteine hin [8]. Über potentielle Interaktionen zwischen MaSp-Proteinen ist wenig bekannt, lediglich über die Verteilung im Faden [44, 46, 102]. ADF3 und 4 sind aufgrund ihres hohen Prolingehaltes MaSp2 Analoga, wobei ADF4 hydrophober ist als ADF3 [7]. Die beiden zugehörigen carboxyterminalen Domänen NR3 und NR4 weisen eine sehr hohe Homologie auf [81]. Die in diesem Abschnitt präsentierten Ergebnisse zeigen, dass beide ADF Proteine über ihre carboxyterminalen Domänen bereits 500 nm 500 nm500 nm C16NR4 (AQ)12NR3 Heterodimer ABC Synopsis [41] während der Produktion innerhalb der Zellen, z.B. in den Kompartimenten des sekretorischen Systems, interagieren können. Möglicherweise agiert das Heterodimer als zusätzlicher „Quervernetzer“ bzw. als Assemblierungsvermittler zwischen beiden dragline Proteinen. Aus biotechnologischer Sicht stellt die Möglichkeit unterschiedliche Seidenproteine mittels der carboxyterminalen Domänen zu verknüpfen eine einfache Methode dar um verschiedene Materialien mit einstellbaren Eigenschaften herzustellen. 2.5 Der Einfluss der terminalen Domänen auf die Orientierung der repetitiven Kerndomäne von rekombinanten Spinnenseidenproteinen In den bisherigen Arbeiten (Teilarbeit I-III) wurde ausschließlich der Einfluss der nichtrepetitiven carboxyterminalen Domäne auf eADF3 und 4 Proteine untersucht. Um weitere Erkenntnisse über den Einfluss der terminalen Domänen auf das Lösungsund Assemblierungsverhalten zu erhalten, wurden rekombinante eADF3-Proteine mit beiden terminalen Domänen untersucht (Teilarbeit VI). Da die native NT-Domäne der ADF-Proteine nicht bekannt ist, kam stattdessen die NT-Region von MaSp1 der schwarzen Witwe (L. hesperus) zur Verwendung (vgl. Abbildung 10). In einer anderen Arbeit wurde deren Struktur mittels NMR untersucht und gezeigt, dass diese eine sehr hohe strukturelle Homologie zur einzig bekannten Kristallstruktur einer MaSp-NT-Domäne einer anderen Spinnenart aufweist [79, 85]. Die NT-Domäne aus der schwarzen Witwe dimerisiert ebenfalls durch Absenken des pH-Wertes von 7,2 auf 6,0. Die Sequenz der NT-Domäne N1 wurde zunächst an eADF3((AQ)12NR3) gekoppelt, da dieses bei den vorangegangen Arbeiten intensiv charakterisiert wurde. Als Kontrolle wurde parallel dazu eADF3(N1-(AQ)12) ohne die CT-Domäne hergestellt. Um im Rahmen dieser Untersuchungen mögliche Interaktionen zwischen beiden terminalen Domänen besser zu charakterisieren, wurde zudem noch das kurze eADF3-Protein N1-AQ-C1 bestehend aus einem AQ-Block und den beiden terminalen Domänen von MaSp1 der schwarzen Witwe hergestellt (vgl. Abbildung 11). Die C1-Domäne ist homolog zur NR3-Domäne und bildet gekoppelt an eADF3-Proteine ebenso quantitativ Dimere aus (Daten nicht gezeigt). In der NT-Dimerstruktur ist am carboxyterminalen Ende eines Monomers eine Vertiefung zu erkennen, von der vermutet wird, dass dort Teile einer Linkerregion zur repetitiven Kerndomäne binden und so die Ausrichtung der Ketten arretieren könnten [79]. Da die bisher untersuchte N1-Sequenz diese Linkerregion nicht enthält, wurden zusätzlich Varian- Synopsis [42] ten mit der 32 Aminosäuren langen nativen Linkerregion des MaSp1-Proteins der schwarzen Witwe hergestellt (vgl. Abbildung 10) [79]. 2.5.1 pH-abhängige Untersuchung der N1und N1L-Proteine mittels CDund Fluoreszenzspektroskopie Die isolierte N1-Domäne zeigte mit fallendem pH-Wert eine erhöhte thermische Stabilität in CD-Experimenten, was als Indikator für die Dimerbildung gilt [85]. Die N1und N1L- (AQ)12 Konstrukte wurden daher ebenfalls mittels CD-Spektroskopie untersucht (Abbildung 18). Abbildung 18: pH-abhängiges Schmelzverhalten von N1-haltigen eADF3-Proteinen und der isolierten N1-Domäne. Es sind die Schmelzkurven von (A) N1-(AQ)12 & N1L-(AQ)12, (B) N1-(AQ)12NR3 & N1L- (AQ)12NR3, (C) N1(AQ)C1 & N1L(AQ)C1 sowie der isolierten (D) N1-Domäne jeweils bei pH 6,0 und 7,2 dargestellt. Die Proteinkonzentration betrug bei allen Messungen 4 µM. Die Heizund Kühlrate bei den thermischen Entfaltungsanalysen betrug 1°C/min. Alle Messungen wurden in 10 mM Kaliumphosphatpuffer durchgeführt. Bei allen Konstrukten sind die Schmelzkurven wie bei der isolierten N1-Domäne auch bei pH 6,0 im Vergleich zu pH 7,2 signifikant zu höheren Temperaturen verschoben (Abbil-  @ 222 nm norm. 1,0 0,8 0,6 0,4 0,2 0  @ 222 nm norm. 1,0 0,8 0,6 0,4 0,2 0  @ 222 nm norm. 1,0 0,8 0,6 0,4 0,2 0  @ 222 nm norm. 1,0 0,8 0,6 0,4 0,2 0 20 30 50 N1-(AQ) 12 NR3 pH 6,0 N1-(AQ) 12 NR3 pH 7,2 N1L-(AQ) 12 NR3 pH 6,0 N1L-(AQ) 12 NR3 pH 7,2 N1-(AQ) 12 pH 6,0 N1 pH 6,0 N1 pH 7,2 N1-(AQ) 12 pH 7,2 N1L-(AQ) 12 pH 6,0 N1L-(AQ) 12 pH 7,2 N1(AQ)C1 pH 6,0 N1-(AQ)C1 pH 7,2 N1L-(AQ)C1 pH 6,0 N1L-(AQ)C1 pH 7,2 60 70 8040 Temperatur (°C) 20 30 50 60 70 8040 Temperatur (°C) 20 30 50 60 70 8040 Temperatur (°C) 20 30 50 60 70 8040 Temperatur (°C) AB CD Synopsis [43] 400 440420400 350 380360340320 300 250 200 150 100 50 0 Fluoreszenz Wellenlänge (nm) N1-(AQ) 12 pH 6,0 N1-(AQ) 12 pH 7,2 N1-(AQ) 12 NR3 pH 6,0 N1-(AQ) 12 NR3 pH 7,2 N1-(AQ)C1 pH 6,0 N1-(AQ)C1 pH 7,2 N1L-(AQ) 12 pH 6,0 N1L-(AQ) 12 pH 7,2 N1L-(AQ)C1 pH 6,0 N1L-(AQ)C1 pH 7,2 NRN1 pH 6,0 NRN1 pH 7,2 dung 18). Die Linkerregion beeinflusst die thermische Stabilität bei beiden pH-Werten dabei nur marginal. Bei den vier Proteinen mit beiden terminalen Domänen sind bei pH 7,2 zwei Bereiche zu beobachten, die sehr wahrscheinlich den Entfaltungsvorgängen der beiden Termini entsprechen (Abbildung 18 B). Die Schmelztemperaturunterschiede zwischen den pH-Werten bei allen Konstrukten sind vergleichbar mit denen der isolierten N1Domäne. Die Spektren bei 20°C vor und nach dem Erhitzen der Proteine überlagern sich nahezu vollständig, was für ein quantitatives Rückfalten spricht. Die einzigen Ausnahmen stellen N1(AQ)C1 und die N1-Domäne bei pH 6,0 dar, die beim Erhitzen auf 85°C unter den gewählten Bedingungen vollständig aggregierten (Ergebnisse nicht gezeigt). Die N1-Domäne besitzt ein einzelnes Tryptophan welches abhängig vom pH-Wert unterschiedliche Fluoreszenzemissionen aufweist [79, 80]. Durch Absenken des pH-Wertes von 7,2 auf 6,0 sinkt bei der isolierten N1-Domäne die Fluoreszenzemission, und das Maximum verlagert sich von ca. 330 nach 345 nm. Dieser Fluoreszenzeffekt basiert auf einer konformationellen Umlagerung des Tryptophans, die während der Dimerisierung auftritt [34, 79, 80]. Um zu überprüfen ob dieser Effekt und damit eine mögliche Dimerisierung bei den N1und N1Linker-Kontrukten auftritt, wurden diese daher bei pH 7,2 und 6,0 mittels Fluoreszenzspektroskopie untersucht (Abbildung 19). Abbildung 19: pH-abhängige Fluoreszensmessungen von N1Proteinen. Dargestellt sind Emissionspektren bei pH 6,0 (geschlossene Symbole) und pH 7,2 (offene Symbole). Zum Vergleich sind die Spektren der isolierten N1-Domäne mitangegeben (von Joschka Bauer zur Verfügung gestellt). Die Lage der unterschiedlichen Maxima ist durch gestrichelte Linien hervorgehoben Die Proteine wurden vor der Messung in Puffern mit den entsprechenden pHWerten äquilibriert. Die Proteinkonzentration betrug bei den Messungen zwischen 3 und 6 µM. Die Anregungswellenlänge betrug 295 nm. Die Fluoreszenzmessungen zeigen, dass alle N1-(AQ)-Proteine bei den beiden pH-Werten das gleiche Verhalten aufweisen wie die isolierte N1-Domäne. Bei pH 6,0 ist die Fluoreszenzintensität geringer und das Maximum zu ca. 345 nm verschoben. Die Maxima bei pH 7,2 liegen alle bei ca. 330 nm. Die Spektren zeigen zudem, dass die Anwesenheit der Linkerregion keinen Einfluss auf den Fluoreszenzeffekt hat. Laut diesen Messungen soll- Synopsis [44] ten daher sämtliche N1-Domänen in den eADF3-Konstrukten durch pH-Absenkung dimerisieren. 2.5.2 SEC-MALS-Analyse der N1und N1L-Proteine Die pH-abhängige Dimerisierung der N1-Domäne wurde mittels an eine Größenausschlusschromatograpie gekoppelte statische Lichtstreuungsmessung (SEC-MALS) verfolgt [85]. Um zu überprüfen, ob die N1-haltigen eADF3-Konstrukte auch tatsächlich Dimerisieren oder evtl. Oligomerisieren wurden hier ebenfalls SEC-MALS Experimente durchgeführt (Abbildung 20). Abbildung 20: SEC-MALS Analyse von N1-eADF3 und N1-eADF3-NR3 Proteinen in Abhängigkeit vom pH-Wert. Gezeigt ist jeweils das UV-Chromatogramm (durchgezogene Linie) mit der normierten Absorption bei 280 nm sowie die mittels MALS ermittelten Molekulargewichte der Proteinspezies (Punkte). A) N1-(AQ)12 und N1L-(AQ)12. B) N1-(AQ)12NR3 und N1L-(AQ)12NR3. C) N1-(AQ)-C1 und N1L-(AQ)- C1. D) Tabellarische Darstellung der mittels MALS ermittelten Molekulargewichte der einzelnen Proteine bei pH 6,0 und 7,2. Die Proteinkonzentration in allen Läufen betrug ca. 12 µM. Die SEC-MALS Ergebnisse von N1-(AQ)12NR3 bzw. N1L-(AQ)12NR3 zeigen, dass die Absenkung des pH-Wertes keine Änderung des Assemblierungszustandes bewirkt. Bei beiden pH-Werten liegen die Proteine nahezu vollständig als NR3-vermittelte Dimere vor. Allerdings ist die Retentionszeit des N1L-Konstruktes bei pH 6,0 signifikant größer, was 8 10121416182022 Zeit (min) MW pH 6,0 MW pH 7,2 MW pH 6,0 MW pH 7,2 pH 6,0 pH 7,2 pH 6,0 pH 7,2 N1-(AQ) 12 N1-(AQ) 12 N1-(AQ) 12 NR3 N1L-(AQ) 12 NR3 N1L-(AQ) 12 MW pH 6,0 MW pH 7,2 MW pH 6,0 MW pH 7,2 pH 6,0 pH 7,2 pH 6,0 pH 7,2 N1-(AQ)-C1 N1L-(AQ)-C1 MW pH 6,0 MW pH 7,2 MWpH 6,0 MW pH 7,2 pH 6,0 pH 7,2 pH 6,0 pH 7,2 N1-(AQ) 12 NR3 N1L-(AQ) 12 NR3 N1L-(AQ) 12 N1-(AQ)-C1 N1L-(AQ)-C1 8 10121416182022 Zeit (min) 8 10121416182022 Zeit (min) Molekulargewicht (kDa) 40 60 20 0 80 100 120 140 160 Molekulargewicht (kDa) DMolekulargewichte (kDa) 100 150 50 0 200 250 300 350 400 Molekulargewicht (kDa) 50 75 25 0 100 125 150 norm. U V Absorption 0 0,2 0,4 0,6 0,8 1,0 norm. UV Absorption 0 0,2 0,4 0,6 0,8 1,0 norm. U V Absorption 0 0,2 0,4 0,6 0,8 1,0 AB C pH 6,0 pH 7,2 106 & 65,6 (75:25) 136,5 & 67,4 (81:19) 101,6 & 47,2 89:11 116,8 & 43,6 87:13 139,1137,7 156,5162,9 61,2 127,4 & 64,6 (6:94) 63,75 140,6 + 66,2 (10:90) Synopsis [45] auf eine Änderung des hydrodynamischen Radius und somit der Proteinpackung hinweist (Abbildung 20 B). Die wesentlich kleineren N1und N1L-(AQ)-C1-Konstrukte zeigen ebenfalls nur eine schwache Tendenz zur Ausbildung von intermolekularen Dimeren, wobei hier die Anwesenheit der Linkerregion sichtbar zu einem höheren Dimer-vom-Dimer Anteil beiträgt (Abbildung 20 C). Um zu überprüfen ob die NT-abhängige Dimerisierung tatsächlich in Verbindung mit repetitiven Blöcken funktioniert, wurde als Kontrolle N1- (AQ)12 bzw. N1L-(AQ)12 untersucht. Im Gegensatz zu den Proteinen mit beiden nichtrepetitiven terminalen Domänen zeigen liegen beide Proteine bei pH 6,0 hauptsächlich als N1-vermitteltes Dimer vor. Interessanterweise dissoziieren sie auch bei pH 7,2 nicht quantitativ. Auch hier zeigt sich, allerdings für das N1-Konstrukt, eine signifikant verschobene Retentionszeit bei pH 6,0. Da alle Proteine in etwa bei gleicher Konzentration auf die Säule appliziert wurden, deuten die Ergebnisse auf unterschiedliche Assoziations/Dissoziationskonstanten der verschiedenen N1/N1L-Proteine hin. Die Anwesenheit der carboxyterminalen Domäne scheint demnach einen Einfluss auf die N1-vermittelte Dimerisierung zu haben. Während Fluoreszenzund CD-Assays zeigen, dass alle N1-haltigen Proteine auf die Absenkung des pH-Wertes sehr ähnlich reagieren, konnte dies bei den SEC-MALS Experimenten nicht bestätigt werden. Die Ergebnisse zeigen, dass die carboxyterminale Domäne NR3 möglicherweise die Konformation und/oder Ausrichtung der repetitiven Bereiche mehr als bisher angenommen beeinflusst. Es sind jedoch noch weiterführende Experimente notwendig um diesen auch für das Verständnis des natürlichen Spinnprozesses wichtigen Sachverhalt weiter zu charakterisieren. Die AQProteine mit beiden terminalen Domänen stellen eine gute Basis für weiterführende Experimente dar. 2.6 Modell des natürlichen Spinnprozesses Im Rahmen dieser Arbeit ist es gelungen ein erstes übergreifendes Modell für die Initiierung der Fadenassemblierung zu erstellen (Teilarbeiten IV & V) (Abbildung 21). Die MA-Spidroine werden zunächst von dafür spezialisierten Zellen produziert und ins Lumen der Drüse sekretiert, welches einen pH-Wert von ca. 7,2 aufweist. Die Spidroingene werden aller Wahrscheinlichkeit nach coexprimiert [8, 35, 162]. Sie liegen aufgrund der CT-Domäne als parallel orientierte kovalent verknüpfte Dimere vor und werden bis zum Verspinnen in der Ampulla bei hohen Konzentrationen von bis zu 50% (w/v) gelagert. Synopsis [46] Abbildung 21: Molekulares Modell des Spinnprozesses innerhalb der MA-Drüse. Die Spidroine werden als kovalente Dimere sekretiert und anschließend bei hohen Konzentrationen in Form von mizellartigen Assemblaten mit flüssigkristallinen Eigenschaften gelagert (i, ii). Durch weiteres Absenken des pH-Wertes im Spinnkanal bilden die NT-Regionen antiparallele Dimere aus, die die Spidroine weiter vernetzen. Das Auftreten von Scherund Elongationskräften im Spinnkanal sowie der aktive Entzug von Wasser initiiert die Bildung der β-Faltblattkristalle (iii). Gleichzeitig entfalten die CT-Domänen und exponieren so hydrophobe Bereiche, die die Assemblierung weiter beschleunigen. Parallel dazu unterstützt die CT-domäne die Ausrichtung der strukturellen Elemente entlang der Faserachse ermöglichen (iii). Im finalen Faden sind die antiparallelen β-Faltblattkristalle entlang der Faserachse orientiert und tragen so zur mechanischen Stabilität bei. Modifiziert nach [3] Aufgrund des amphiphilen Charakters der repetitiven Kerndomänen lagern sich die Proteine zu supramolekularen, mizellartigen Assemblaten zusammen, die einen flüssigkristallinen Charakter aufweisen [105]. An der Bildung dieser Assemblate ist möglicherweise die CT-Domäne beteiligt. Das in der Ampulla vorhandene Natriumchlorid stabilisiert die CTDomäne gegen Entfaltung und damit gegen vorzeitige Assemblierung. Die nicht-repetitive aminoterminale (NT) Domäne liegt in der Ampulla aufgrund des pH-Wertes und der Anwesenheit von Natriumchlorid überwiegend als Monomer vor. Da der pH-Wert innerhalb der Ampulla von 7,2 bis auf ca. 6,5 kurz vor Beginn des S-förmigen Spinnkanal abnimmt, steigt die Tendenz der NT-Domäne antiparallele Dimere zu bilden [82, 83]. Mit Eintritt in den S-förmigen Spinnkanal durch eine Art Trichter verändern sich die physikochemischen Bedingungen in der Drüse. Es kommt zu einer weiteren Ansäuerung durch Protonenpumpen auf ca. pH 6,0 [82, 83] sowie zu Ionenaustauschprozessen bei denen Natriumchlorid durch das kosmotropere Kaliumphosphat ersetzt wird. Durch die Windungen sowie dem kovalent verknüpfter, dimerer C-Terminus monomerer N-Terminus Repetitive Block i) Fortsatz ii)Ampulla Die Spidroine sind teilw. vorstrukturiert Faser mit β-Faltblattkristallen, die entlang der Faserachse ausgerichtet sind Scherinduzierte Elongation sowie pHbedingte Dimerisierung der N-terminalen Domäne und zeigen flüssigkristallinen Charakter iii) Spinnkanal nicht-kovalentes, N-terminales Dimer . iv) Außen Wasserentzug und Scherkräfte verstärken die Phasenseparation und die damit verbundene Bildung der β-Faltblattstrukturen sowie die Ausrichtung der Proteine; durch Entfaltung der C-terminalen Domäne werden hydrophobe Bereiche freigesetzt, die die Assemblierung weiter beschleunigen. Synopsis [47] enger werdenden Spinnkanal treten zusätzlich Scherund Elongationskräfte auf. In Abwesenheit von Natriumchlorid und in Anwesenheit von Scherkräften entfaltet sich die CTDomäne und fördert durch die Exposition von hydrophoben Bereichen die Assemblierung der Spidroine. Die NT-Domäne dimerisiert unter den Bedingungen im Spinnkanal. Diese antiparallelen Dimere beschleunigen durch die mögliche Vernetzung der Spidroine die Assemblierung. Durch das Kaliumphosphat sowie die Scherund Elongationskräfte bilden sich intermolekulare β-Faltblattkristalle. Parallel dazu wird im Spinnkanal Wasser entzogen, was die Konzentration der Proteine erhöht und somit zur flüssig-fest Phasenseparation beiträgt [108]. Zusammen mit den zunehmenden Scherund Elongationskräften gegen Ende des Spinntraktes bewirkt die CT-Domäne eine Ausrichtung der strukturellen Elemente entlang der Faserachse. Durch aktiven Zug aus der Spinnwarze wird die strukturelle Ausrichtung finalisiert. Literaturverzeichnis [54] 134. Rammensee, S., et al., Assembly mechanism of recombinant spider silk proteins. Proceedings of the National Academy of Sciences of the United States of America, 2008. 105(18): p. 6590-6595. 135. Slotta, U.K., et al., An engineered spider silk protein forms microspheres. Angewandte Chemie-International Edition, 2008. 47(24): p. 4592-4594. 136. Lammel, A., et al., Processing conditions for the formation of spider silk microspheres. ChemSusChem, 2008. 1(5): p. 413-6. 137. Hermanson, K.D., et al., Engineered microcapsules fabricated from reconstituted spider silk. Advanced Materials, 2007. 19(14): p. 1810-+. 138. Schacht, K. and T. Scheibel, Controlled hydrogel formation of a recombinant spider silk protein. Biomacromolecules, 2011. 12(7): p. 2488-95. 139. Teule, F., et al., Combining flagelliform and dragline spider silk motifs to produce tunable synthetic biopolymer fibers. Biopolymers, 2012. 97(6): p. 418-31. 140. Teule, F., et al., A protocol for the production of recombinant spider silk-like proteins for artificial fiber spinning. Nature Protocols, 2009. 4(3): p. 341-355. 141. Teule, F., et al., Modifications of spider silk sequences in an attempt to control the mechanical properties of the synthetic fibers. Journal of Materials Science, 2007. 42(21): p. 8974-8985. 142. Fu, C.J., Z.Z. Shao, and F. Vollrath, Animal silks: their structures, properties and artificial production. Chemical Communications, 2009(43): p. 6515-6529. 143. Brooks, A.E., et al., Properties of synthetic spider silk fibers based on Argiope aurantia MaSp2. Biomacromolecules, 2008. 9(6): p. 1506-1510. 144. An, B., et al., Inducing β-Sheets Formation in Synthetic Spider Silk Fibers by Aqueous Post-Spin Stretching. Biomacromolecules, 2011. 12(6): p. 2375-2381. 145. Keerl, D. and T. Scheibel, Characterization of natural and biomimetic spider silk fibers. Bioinspired, Biomimetic and Nanobiomaterials, 2012. 1(2): p. 83-94. 146. Work, R.W., DIMENSIONS, BIREFRINGENCES, AND FORCE-ELONGATION BEHAVIOR OF MAJOR AND MINOR AMPULLATE SILK FIBERS FROM ORB-WEBSPINNING SPIDERS - EFFECTS OF WETTING ON THESE PROPERTIES. Textile Research Journal, 1977. 47(10): p. 650-662. 147. Carmichael, S. and C. Viney, Molecular order in spider major ampullate silk (dragline): Effects of spinning rate and post-spin drawing. Journal of Applied Polymer Science, 1999. 72(7): p. 895-903. 148. Carmichael, S., J.Y.J. Barghout, and C. Viney, The effect of post-spin drawing on spider silk microstructure: a birefringence model. International Journal of Biological Macromolecules, 1999. 24(2-3): p. 219-226. 149. Exler, J.H., D. Hummerich, and T. Scheibel, The amphiphilic properties of spider silks are important for spinning. Angewandte Chemie-International Edition, 2007. 46(19): p. 35593562. 150. Hagn, F.X., NMR Spectroscopy of Molecular Chaperones, the p53 Tumorsuppressor Network and Spider Silk Proteins. 2009. p. 227-258. 151. Sponner, A., et al., Conserved C-termini of spidroins are secreted by the major ampullate glands and retained in the silk thread. Biomacromolecules, 2004. 5(3): p. 840-845. 152. Kyte, J. and R.F. Doolittle, A simple method for displaying the hydropathic character of a protein. Journal of Molecular Biology, 1982. 157(1): p. 105-132. 153. Arakawa, T. and S.N. Timasheff, Preferential interactions of proteins with salts in concentrated solutions. Biochemistry, 1982. 21(25): p. 6545-52. 154. Mayr, L.M. and F.X. Schmid, Stabilization of a protein by guanidinium chloride. Biochemistry, 1993. 32(31): p. 7994-7998. 155. Baldwin, R.L., How Hofmeister ion interactions affect protein stability. Biophysical Journal, 1996. 71(4): p. 2056-2063. 156. Eisoldt, L., et al., The role of salt and shear on the storage and assembly of spider silk proteins. Journal of Structural Biology, 2010. 170(2): p. 413 - 419. Literaturverzeichnis [55] 157. Rammensee, S., et al., Assembly mechanism of recombinant spider silk proteins. Proc Natl Acad Sci U S A, 2008. 105(18): p. 6590-5. 158. Holland, C., J.S. Urbach, and D.L. Blair, Direct visualization of shear dependent silk fibrillogenesis. Soft Matter, 2012. 8(9): p. 2590-2594. 159. Lin, Z., et al., Solution structure of eggcase silk protein and its implications for silk fiber formation. Proceedings of the National Academy of Sciences of the United States of America, 2009. 106(22): p. 8906-11. 160. Huemmerich, D., et al., Novel assembly properties of recombinant spider dragline silk proteins. Current Biology, 2004. 14(22): p. 2070-2074. 161. Rising, A., et al., Spider silk proteins--mechanical property and gene sequence. Zoological Science, 2005. 22(3): p. 273-81. 162. Blamires, S.J., C.-L. Wu, and I.M. Tso, Variation in Protein Intake Induces Variation in Spider Silk Expression. Plos One, 2012. 7(2): p. e31626. 163. Tian, M.Z. and R.V. Lewis, Molecular characterization and evolutionary study of spider tubuliform (eggcase) silk protein. Biochemistry, 2005. 44(22): p. 8006-8012. 164. Lee, K.S., et al., Molecular cloning and expression of the C-terminus of spider flagelliform silk protein from Araneus ventricosus. Journal of biosciences, 2007. 32(4): p. 705-12. 165. Guinea, G.V., et al., Minor Ampullate Silks from Nephila and Argiope Spiders: Tensile Properties and Microstructural Characterization. Biomacromolecules, 2012. 166. Tai, P.-L., G.-Y. Hwang, and I.M. Tso, Inter-specific sequence conservation and intraindividual sequence variation in a spider silk gene. International Journal of Biological Macromolecules, 2004. 34(5): p. 237-243. 167. Ayoub, N.A. and C.Y. Hayashi, Multiple Recombining Loci Encode MaSp1, the Primary Constituent of Dragline Silk, in Widow Spiders (Latrodectus: Theridiidae). Molecular Biology and Evolution, 2008. 25(2): p. 277-286. 168. Garb, J.E., et al., Silk genes support the single origin of orb webs. Science, 2006. 312(5781): p. 1762-1762. 169. Bittencourt, D., et al., A MaSp2-like gene found in the Amazon mygalomorph spider Avicularia juruensis. Comparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology, 2010. 155(4): p. 419-426. 170. Garb, J.E., et al., Expansion and Intragenic Homogenization of Spider Silk Genes since the Triassic: Evidence from Mygalomorphae (Tarantulas and Their Kin) Spidroins. Molecular Biology and Evolution, 2007. 24(11): p. 2454-2464. 171. Starrett, J., et al., Early Events in the Evolution of Spider Silk Genes. Plos One, 2012. 7(6): p. e38084. Publikationsliste [56] 4. Publikationsliste I. Hagn, F., Eisoldt, L., Hardy, J.G., Vendrely, C., Coles, M.,Scheibel, T., Kessler, H. (2010) A highly conserved spider silk domain acts as a molecular switch that controls fibre assembly. Nature. 465, pp. 239-242 II. Eisoldt, L., Hardy, J.G., Heim, M., Scheibel, T. (2010) The role of salt and shear on the storage and assembly of spider silk proteins. J. Struct. Biol. 170, pp. 413419 III. Eisoldt, L., Döring, V., Schiemann, S., Scheibel, T. Counterplay of two MaSp2 spider silk proteins from A. diadematus: implications for silk fiber assembly. (2013) Manuscript in preparation. IV. Eisoldt, L., Smith, A., Scheibel, T. (2011) Decoding the secrets of spider silk. Materials Today. 40, pp. 80-86 V. Eisoldt, L., Thamm, C., Scheibel, T. (2012) The role of terminal domains during storage and assembly of spider silk proteins. Biopolymers. 97, pp. 355-361 Die Teilarbeiten I-V sind nachfolgend angehängt. Neben dieser Arbeit sind noch weitere Publikationen auf Basis meiner Daten geplant und z.T. in Vorbereitung. Diese sind daher nicht Teil dieser Dissertation: VI. Eisoldt, L., Bauer, J., Scheibel, T. (2013) Influence of the non-repetitive terminal domains on the spatial orientation of recombinant spider silk proteins in solution. Manuscript in preparation – Work in progress VII. Bauer, J., Eisoldt, L., Schmidpeter, P., Scheibel, T. (2013) Investigations on the dimerization mechanism of the MaSp1 NT-domain from the black widow. Manuscript in preparation – Work in progress VIII. Heidebrecht, A., Eisoldt, L., Schmidt, A., Diehl, J., Geffers, M., Scheibel, T. (2013) Synthetic spider silk fibers with adjustable mechanical properties. Manuscript in preparation - – Work in progress Darstellung des Eigenanteils [57] 5. Darstellung des Eigenanteils I. Hagn, F., Eisoldt, L., Hardy, J.G., Vendrely, C., Coles, M.,Scheibel, T., Kessler, H. (2010) A highly conserved spider silk domain acts as a molecular switch that controls fibre assembly. Nature. 465, pp. 239-242 Sämtliche NMR-Experimente sowie Entfaltungsstudien und ANS/SYPRO Orange Bindungen an der NR3-Domäne wurden von Franz Hagn durchgeführt. Die Klonierung, Reinigung und Herstellung der eADF3(AQ)-Konstrukte wurde von mir durchgeführt. Die Phosphatund Rotationsaggregationsassays wurden von mir durchgeführt. Die Lichtmikroskopischen Aufnahmen der Aggregate wurde von mir aufgenommen. Das Manuskript wurde von Franz Hagn, Dr. John Hardy, Horst Kessler, Thomas Schiebel und mir verfasst. II. Eisoldt, L., Hardy, J.G., Heim, M., Scheibel, T. (2010) The role of salt and shear on the storage and assembly of spider silk proteins. J. Struct. Biol. 170, pp. 413419 Sämtliche Proteine wurden von mir hergestellt und gereinigt. Alle Aggregationsassays sowie die lichtmikroskopischen Aufnahmen und Teile der FTIR-Messungen wurden von mir durchgeführt. Das Manuskript wurde von Dr. John Hardy, Thomas Scheibel und mir verfasst. III. Eisoldt, L., Döring, V., Schiemann, S., Scheibel, T. Counterplay of two MaSp2 spider silk proteins from A. diadematus: implications for silk fiber assembly. (2013) Manuscript in preparation. Die Reinigungsstrategie zum Heterodimer wurde von mir in Zusammenarbeit mit Svenja Schiemann und Volker Döring entwickelt. Die Proteine wurden von Volker Döring, Svenja Schiemann und mir gereinigt. Die CD-Messungen, AFMund TEM-Bilder wurden von mir angefertigt. Das Manuskript haben Thomas Scheibel und ich verfasst. IV. Eisoldt, L., Smith, A., Scheibel, T. (2011) Decoding the secrets of spider silk. Materials Today. 40, pp. 80-86 Die Konzeption des Artikels stammt von Thomas Scheibel und mir. Das Manuskript wurde von Thomas Scheibel und mir verfasst. Dr. Andrew Smith hat teilweise Literatur bereitgestellt. V. Eisoldt, L., Thamm, C., Scheibel, T. (2012) The role of terminal domains during storage and assembly of spider silk proteins. Biopolymers. 97, pp. 355-361 Die Konzeption des Artikels stammt von Thomas Scheibel und mir. Das Manuskript wurde von Thomas Scheibel und mir verfasst. Die Abbildungen wurden von Christopher Thamm erstellt. Teilarbeiten [58] 6. Teilarbeiten Teilarbeit I Hagn, F., Eisoldt, L., Hardy, J.G., Vendrely, C., Coles, M., Scheibel, T., Kessler, H. (2010) A highly conserved spider silk domain acts as a molecular switch that controls fibre assembly. Nature. 465, pp. 239-242 LETTERS A conserved spider silk domain acts as a molecular switch that controls fibre assembly Franz Hagn 1,2 , Lukas Eisoldt 3 , John G. Hardy 3 , Charlotte Vendrely 3 {, Murray Coles 4 , Thomas Scheibel 3 & Horst Kessler 1,2 A huge variety of proteins are able to form fibrillar structures 1 , especially at high protein concentrations. Hence, it is surprising that spider silk proteins can be stored in a soluble form at high concentrations and transformed into extremely stable fibres on demand 2,3 . Silk proteins are reminiscent of amphiphilic block copolymers containing stretches of polyalanine and glycine-rich polar elements forming a repetitive core flanked by highly conserved non-repetitive amino-terminal 4,5 and carboxy-terminal 6 domains. The N-terminal domain comprises a secretion signal, but further functions remain unassigned. The C-terminal domain was implicated in the control of solubility and fibre formation 7 initiated by changes in ionic composition 8,9 and mechanical stimuli known to align the repetitive sequence elements and promote b-sheet formation 10–14 . However, despite recent structural data 15 , little is known about this remarkable behaviour in molecular detail. Here we present the solution structure of the C-terminal domain of a spider dragline silk protein and provide evidence that the structural state of this domain is essential for controlled switching between the storage and assembly forms of silk proteins. In addition, the C-terminal domain also has a role in the alignment of secondary structural features formed by the repetitive elements in the backbone of spider silk proteins, which is known to be important for the mechanical properties of the fibre. Silk proteins are amphiphilic (Fig. 1a) and the C-terminal nonrepetitive (NR) domain of such proteins produced by orb-weaving spiders is one of the most conserved regions (Fig. 1b). To determine the structure of the C-terminal NR domain (NR3) of Araneus diadematus fibroin 3 (ADF-3), we used high-resolution NMR with an optimized protein construct (Fig. 1c and Supplementary Tables 1 and 2). The structure is a new protein fold composed of a parallel-oriented dimeric five-helix bundle in which the longest helix (helix 4) is the main dimerization site (Fig. 1c, Supplementary Fig. 1). The single cysteine residue of each monomer involved in intermolecular disulphide formation is located at the N-terminal end of helix 4. The dimerization interface consists predominantly of hydrophobic residues. Packing of these residues is facilitated by a slight right-handed twist between helices 4 in both monomers. Helix 1 of one monomer and helix 5 of the second monomer interact in a clamp-like manner. Two salt bridges (R43–D93 and R52–E101) are located within each monomer, which fix helices 1 and 2 at one side of helix 4 (Fig. 1c). These salt bridges use the only charged residues of the entire known sequence of ADF-3 and are located in the most conserved parts of the NR domain (Fig. 1b). Using mutagenesis experiments, our data show that the charged residues are essential for the structural integrity of the NR domain, as deletion of the salt bridges results in significantly reduced thermal stability of the entire domain (Supplementary Fig. 2). Of particular note is that although the NR domain is relatively hydrophobic in comparison to the repetitive backbone of the protein (Fig. 1a), in the folded state the hydrophobic residues are buried and the solventexposed surface displays hydrophilic amino acids such as glutamine and serine, assuring its hydration (Supplementary Fig. 1b). At low pH (pH 2, where the salt bridges are disrupted) NR3 shows markedly increased hydrophobicity and decreased secondary structure content as probed with 8-anilinonaphthalene-1-sulphonic acid (ANS) fluorescence, circular dichroism (CD) and NMR spectroscopy (Supplementary Fig. 3a–d). In addition, a destabilized NR3 salt bridge mutant (D93A) of an engineered ADF-3 analogue, (AQ) 12 NR3 containing 12 repeats of the ‘A’ and ‘Q’ repetitive sequence elements 16 (see Methods), also shows pronounced ANS binding (Fig. 2b) and molten globule-like thermal unfolding behaviour (a partially structured state that shows higher binding affinity to ANS than the native state), which can be stabilized by the addition of sodium chloride (Fig. 2c). This emphasizes the role of a correctly folded NR domain for silk protein storage and the inhibition of undesired aggregation. In vivo the transition from soluble protein to solid fibres involves a combination of chemical and mechanical stimuli (such as ion exchange, extraction of water, acidification and elongational flow 3,13,17 ). Hence, we were interested in determining the effect of such stimuli on spider silk-like proteins, and used dimeric (AQ) 12 NR3 and monomeric (AQ) 24 (an analogue of approximately the same molecular mass as the dimer without the NR carboxyterminal domain). In the absence of shear forces (as silk proteins are stored in the lumen in vivo) the extent of protein aggregation in vitro is determined by the concentration of salt, and even at extremely high concentrations of sodium chloride (up to 500 mM) the level of aggregation is low (less than 15%), whereas in the presence of equivalent concentrations of sodium phosphate the level of aggregation is markedly higher (up to 100%) owing to the more kosmotropic nature of the phosphate anion (Fig. 2a). Such a finding explains why the proteins are stored inside the lumen in the presence of the relatively chaotropic sodium chloride (approximately 100–150 mM) 9 ,and emphasizes the requirement for ion exchange within the spinning duct. Higher supramolecular assemblies are required to achieve efficient protein storage 8 and we have previously reported our proteins to be thermoresponsive, displaying fully reversible lower critical solution temperature (LCST, phase separation on increasing the temperature) behaviour owing to a self-assembly process 16 . This behaviour was displayed only for silk proteins incorporating the NR domain. Using differential scanning calorimetry (DSC) we show here that the LCST is decreased in the presence of higher concentrations of salt and/or protein (Fig. 2d, e), thereby stabilizing the proteins during storage. 1 Center for Integrated Protein Science (CIPS M )and 2 Institute for Advanced Study, Department Chemie at the Technische Universita ¨tMu¨nchen, 85747 Garching, Germany. 3 Lehrstuhl fu¨r Biomaterialien, Fakulta ¨tfu¨r Angewandte Naturwissenschaften, Universita ¨t Bayreuth, 95440 Bayreuth, Germany. 4 Department of Protein Evolution, Max-Planck-Institute for Developmental Biology, 72076 Tu¨bingen, Germany. {Present address: Universite ´de Cergy-Pontoise, 95302 Cergy-Pontoise cedex, France. Vol 465 | 13 May 2010 | doi:10.1038/nature08936 239 Macmillan Publishers Limited. All rights reserved ©2010 We have also found that the NR domain has an impact on the fibre formation process. On titration of chemical denaturants (urea or guanidine hydrochloride) into solutions of 15 N-labelled NR3, NMR chemical shift changes (indicating structural variations or binding) monitored with 15 N HSQC (heteronuclear single quantum coherence) experiments could be observed mainly within helix 1 and around the salt bridges (Fig. 3a, Supplementary Fig. 4), therefore this region seems to unfold first. Titrations with sodium chloride demonstrated chemical shift changes in the same region (where the two salt bridges are located, Fig. 3a), which also shows the highest hydrophobicity in the protein as monitored with chemical shift changes upon the addition of ANS (Fig. 3a and Supplementary Fig. 4). Interestingly, at higher protein concentrations the protein shows a greater tendency to unfold (Fig. 3b). Additionally, SYPRO Orange (an ANS-like fluorescent dye) binding experiments 18 demonstrate that the partially folded state of NR3 is more hydrophobic than the natively folded or the unfolded state (Fig. 3c). It is also noteworthy that the presence of sodium chloride (present during protein storage) stabilizes the NR domain against thermal or chemical denaturation (Fig. 3b, c and Supplementary Fig. 5a) and can even compensate for the loss in stability during reduction of the inter-molecular disulphide bridge (Supplementary Fig. 5b). Previous rheological studies emphasize the role of the NR domain for shear-induced aggregation of ADF-3, where solutions of proteins with the NR domain exhibit significant viscosity changes under shear, −2 −1 0 1 2 −3 −2 −1 0 1 2 3 −2 −1 0 1 2 3 Bombyx mori Hydrophobicity scale E. australis eADF3 Amino acid sequence NR NR Repetitive sequence elements NC ab 90° c H1 H2 H3 H4 H5 N H1′ H2′ H3′ H4′ H5′ S–S bridge C H4′H4 N C‘ C′ -------GPQ SSSAPVASAA ASRLSSPAAS SRVSSAVSSL GAGQGGYGGV GSGASAASAA ASRLSSPQAS SRVSSAVSNL --SSASASAA ASAASTVANS VSRLSSPSAV SRVSSAVSSL ADF-3 N.c. MaSp1 E.a. MaSp1 ADF-1 --------VA GASAGSYGGA VNRLSSAGAA SRVSSNVAAI VSSGPTNQAA LSNTISSVVS QVSASNPGLS GCDVLVQALL VASGPTNSAA LSSTISNVVS QIGASNPGLS GCDVLIQALL VSNGQVNMAA LPNIISNISS SVSASAPGAS GCEVIVQALL ASAG---AAA LPNVISNIYS GVLSS--GVS SSEALIQALL EVVSALVSIL GSSSIGQINY GASAQYTQMV GQSVAQALAG EVVSALIHIL GSSSIGQVNY GSAGQATQIV GQSVYQALGEVITALVQIV SSSSVGYINP SAVNQITNVV ANAMAQVMGEVISALIHVL GSASIGNVSS VGVNSALNAV QNAVGAYAGH1 H2 H2 H3 H4 H4 H5 ADF-3 N.c. MaSp1 E.a. MaSp1 ADF-1 ADF-3 N.c. MaSp1 E.a. MaSp1 ADF-1 N′ Salt bridges Salt bridges D93 R43 E101 R52 S–S bridge Salt bridges Q97 Figure 1 | Sequence analysis and structure of the non-repetitive (NR) domain of ADF-3. a, Hydrophobicity index of various silk proteins. b, Sequence alignment of the C-terminal NR domains of the major ampullate silk proteins (MaSp) of Araneus diadematus,Nephila clavipes and Euprosthenops australis, and minor ampullate silk fibroin 1 (ADF-1) of Araneus diadematus (top to bottom). Grey bars indicate the degree of conservation between these sequences. The conserved charged residues are coloured. c, Overlay of the 20 best-energy structures of the C-terminal NR domain of ADF3 having a root mean square deviation (r.m.s.d.) of 0.18 A ˚. 100 200 30050 150 250 500 100 200 30050 150 250 500 Aggregation (%) Salt concentration (mM) 0 20 40 60 80 100 450 500 550 600 2 × 105 6 × 105 1 × 106 Fluorescence (a.u.) (AQ)12NR3 D93A (AQ)12NR3 WT (AQ)24 Wavelength (nm) b 10 15 20 25 30 35 40 0.0 1.0 2.0 3.0 4.0 5.0 165 μM: 23.4 °C 50 µM: 25.4 °C 10 µM: 29.7 °C Temperature (°C) ΔC p(mcal °C−1) ΔCp(mcal °C−1) 10 15 20 25 30 35 40 0.0 0.5 1.0 1.5 2.0 2.5 No salt: 25.4 °C 100 mM: 25.0 °C 500 mM: 20.0 °C Temperature (°C) ed 20 40 60 80 100 4 × 103 8 × 103 1 × 104 Fluorescence (a.u.) Temperature (°C) (AQ)12NR3 D93A (AQ)12NR3 WT (AQ)24 c NaCl Na-phosphate a (AQ)12NR3 (AQ)12NR3 D93A (AQ)24 (AQ)24NR3 Figure 2 | Assembly and aggregation properties of our spider silk-like proteins. a, Salt-induced protein aggregation. Error bars indicate standard deviation. b, Binding of hydrophobic ANS by proteins. a.u., arbitrary units; WT, wild type. c, Thermal transition experiments indicate molten globulelike behaviour of the D93A variant (red symbols), whereas the wild-type protein binds to the dye only during unfolding (black symbols). A structural transition of the D93A variant can be induced by sodium chloride (open symbols: in presence of 300 mM sodium chloride). Without the NR domain, no transition can be observed (green symbols). d, The concentration dependence of the LCST of (AQ) 24 NR3. e, The sodium chloride dependence of the LCST of (AQ) 24 NR3. LETTERS NATURE | Vol 465 | 13 May 2010 240 Macmillan Publishers Limited. All rights reserved ©2010 whereas solutions of proteins without the NR domain exhibit no viscosity changes at various shear rates 14,19 . ANS binding also increases under shear, indicating that the NR domain partially unfolds, thereby exposing hydrophobic surfaces (Fig. 3d and Supplementary Fig. 6), leading to oligomerization and ultimately b-sheet-rich fibre formation. In further in vitro protein aggregation assays under shear stress, we find that aggregation is vastly increased in comparison to corresponding assays in the absence of shear, and that the protein without the NR domain is much more prone to unspecific aggregation (Figs 3e and 4). Exposure of solutions of proteins without the NR domain to shear stress resulted in the formation of ill-defined b-sheet-rich aggregates, typically with dimensions of microto millimetre scale (Fig. 4a). By contrast, proteins incorporating the NR domain formed relatively well-defined fibrous aggregates, which optical microscopy showed to be composed of ordered (aligned) bundles of long fibrillar aggregates, implying that the NR domain has a role in stabilizing the protein against undesirable aggregation. Polarized Fourier transform infrared spectra of the fibrous aggregates formed from our protein without the NR domain ((AQ) 24 ) show that the b-sheets are randomly aligned, whereas the b-sheets within fibres formed from our protein with the correctly folded NR domain (that is, wild type and not the D93A mutant) were predominantly aligned with the long axis of the fibres (in analogy to naturally spun fibres) owing to a controlled assembly process (Fig. 4b, c). To examine the influence of the length of the repetitive backbone on the behaviour of the proteins, we performed aggregation assays with (AQ) 24 NR3, finding that it was significantly more prone to aggregation than the equivalent (AQ) 12 NR3 analogue (Fig. 2a), which leads us to speculate that the a-helical N-terminal NR domain 4 (not included in our recombinant proteins) might also play an important function in protein storage and fibre assembly, especially in case of the natural system, where the proteins are larger. Our data indicate that the C-terminal NR domain has a role in both spidersilk protein storageand fibre assembly (Fig. 4d).During storage of spider silk proteins, the NR domain stabilizes a solution-competent state of the silk protein via the formation of higher supramolecular assemblies, and the C-terminal NR domain also acts as a trigger for the fast, controlled and efficient salting-out of the proteins in combination with the correct alignment of the b-sheet forming repetitive sequence elements during shear-induced assembly. The hydrophobic surface of the assembly state of the NR domain, in addition to the disulphide bridge, provides anchor points for the correct positioning of the repetitive sequence elements. This mechanism is reminiscent of other fibrillar proteins 20 like collagen 21 , where the highly conserved terminal pro-peptides form intermolecular disulphide bridges 22 and regulate pro-collagen solubility and lateral assembly 23 . METHODS SUMMARY Proteins samples were bacterially produced in Escherichia coli and purified as described 24 or by Nickel-NTA and size-exclusion chromatography. Twoand Urea NaCl ANS S–S bridge S–S bridge S–S bridge H1 H1 H1 H4 H4 H5′ H2 H3 H3 H2 H4 H5′ H5′ H2 H3 Urea (M) 5 μM 34 μM 60 μM 200 μM 01 23 4 60 μM NR3 + NaCl 0.0 0.5 1.0 Fraction UnfoldedFolded Urea (M) 0 5 10 15 20 SYPRO Orange fluorescence (a.u.) No NaCl + NaCl 01 234 UnfoldedFolded a bc NR3 (AQ)12NR3 (AQ)24 0 20 40 60 80 100 −+−+−+Shear 50 mM NaCl 500 mM NaCl Aggregation (%) Wavelength (nm) Fluorescence (a.u.) de 2 × 10 5 1 × 10 5 0 440 480 520 560 600 ANS binding Time Figure 3 | Stability and folding of spider dragline silk constructs. a, NMR chemical shift perturbation experiments with urea, sodium chloride and ANS indicate that the structural elements around the salt bridges represent the most labile and hydrophobic region within the protein. b, Urea-induced unfolding of NR3 as determined by CD spectroscopy. c, Binding of hydrophobic SYPRO orange to NR3 during urea-induced unfolding. d, Shear-induced ANS binding of NR3. e, Aggregation of our spider silk-like proteins with and without shear stress. Shear stress leads to increased aggregation and is the most likely trigger for the assembly process in nature. Error bars represent standard deviation. 100 µm 100 µm a b 1,0001,2001,4001,6001,800 0.00 0.05 0.10 0.15 0.20 Wavenumber (cm–1) Absorption (a.u.) 1,0001,2001,4001,6001,800 0.00 0.04 0.08 0.12 Absorption (a.u.) Wavenumber (cm–1) (AQ)12NR3 (AQ)12NR3 (AQ)12NR3(AQ)12NR3 (AQ)24 cβ-sheet content (%) Salt-induced aggregation Shear-induced aggregation pH Wild type D93A Wild type D93A NR NR Shear stress NR NR (h (h ydro ydro philic philic surface) surface) Repetitive sequence elements Hydrophobic element Hydrophilic element Higher oligomeric assemblies (turbid solutions) Reversible Hydrophilic surface NR Fibre d (AQ)24 (AQ)24 234 ± 3 35 ± 3 31/33/32 ± 2 30/33/30 ± 238 ± 1 35/33/36 ± 2 632 ± 4 32 ± 1 36/33/30 ± 2 28/29/29 ± 141 ± 2 30/32/31 ± 2 726 ± 3 35 ± 2 28/28/25 ± 1 35/34/35 ± 132 ± 6 32/32/32 ± 1 Partial unfolding of the NR domain exposes a hydrophobic surface which encourages aggregation and allows alignment of the repetitive elements under shear Figure 4 | Fibre assembly mechanism of dragline silk proteins. a, A welldefined fibre (dry state) of (AQ) 12 NR3 and ill-defined fibrous aggregate of (AQ) 24 formed under shear. b, Infrared absorption spectra of (AQ) 12 NR3 and (AQ) 24 fibres recorded at 0uand 90urelative to the long axis of the fibrous aggregates (black vs red line). c, Estimated b-sheet content of the aggregates. The values of b-sheet content for the aggregates formed under shear were recorded at 0u,45uand 90urelative to the long axis of the fibrous aggregates. d, The silk proteins are stored as higher oligomeric assemblies. Exposure of these assemblies to shear and salting out leads to partial unfolding of NR3 and controlled fibre assembly. NATURE | Vol 465 | 13 May 2010 LETTERS 241 Macmillan Publishers Limited. All rights reserved ©2010 three-dimensional NMR experiments were performed with 13 Cand 15 N isotopically enriched protein samples on NMR spectrometers (Bruker Biospin) operating at 600, 750 and 900MHz proton frequency, respectively. Out of 140 residues in each monomer, the NMR signals of 138 residues could be assigned using a set of conventional experiments 25 . Structural restraints were generated with nuclear Overhauser enhancement (NOE) spectroscopy experiments 26 and high-resolution structures were obtained with the program Xplor-NIH 27 . Protein folding studies were performed with circular dichroism (CD) and fluorescence spectroscopy and protein fibres and aggregates were investigated with optical microscopy and Fourier transform infrared (FTIR) spectroscopy. Full Methods and any associated references are available in the online version of the paper at www.nature.com/nature. Received 20 May 2009; accepted 11 February 2010. 1. Dobson, C. M. Protein folding and misfolding. Nature 426, 884 – 890 (2003). 2. Gosline, J. M., Guerette, P. A., Ortlepp, C. S. & Savage, K. N. The mechanical design of spider silks: from fibroin sequence to mechanical function. J. Exp. Biol. 202, 3295 – 3303 (1999). 3. Vollrath, F. & Knight, D. P. Liquid crystalline spinning of spider silk. Nature 410, 541 – 548 (2001). 4. Rising, A., Hjalm, G., Engstrom, W. & Johansson, J. N-terminal nonrepetitive domain common to dragline, flagelliform, and cylindriform spider silk proteins. Biomacromolecules 7, 3120 – 3124 (2006). 5. Motriuk-Smith, D., Smith, A., Hayashi, C. Y. & Lewis, R. V. Analysis of the conserved N-terminal domains in major ampullate spider silk proteins. Biomacromolecules 6, 3152 – 3159 (2005). 6. Sponner, A. et al. The conserved C-termini contribute to the properties of spider silk fibroins. Biochem. Biophys. Res. Commun. 338, 897 – 902 (2005). 7. Ittah, S., Cohen, S., Garty, S., Cohn, D. & Gat, U. An essential role for the C-terminal domain of a dragline spider silk protein in directing fiber formation. Biomacromolecules 7, 1790 – 1795 (2006). 8. Jin, H. J. & Kaplan, D. L. Mechanism of silk processing in insects and spiders. Nature 424, 1057 – 1061 (2003). 9. Knight, D. P. & Vollrath, F. Changes in element composition along the spinning duct in a Nephila spider. Naturwissenschaften 88, 179 – 182 (2001). 10. Knight, D. P. & Vollrath, F. Liquid crystals and flow elongation in a spider’s silk production line. Proc. Roy. Soc. Lond. B266 519 – 523 (1999). 11. Knight, D. P., Knight, M. M. & Vollrath, F. Beta transition and stress-induced phase separation in the spinning of spider dragline silk. Int. J. Biol. Macromol. 27, 205 – 210 (2000). 12. Lefe `vre, T., Boudreault, S., Cloutier, C. & Pe ´zolet, M. Conformational and orientational transformation of silk proteins in the major ampullate gland of Nephila clavipes spiders. Biomacromolecules 9, 2399 – 2407 (2008). 13. Heim, M., Keerl, D. & Scheibel, T. Spider silk: from soluble protein to extraordinary fiber. Angew. Chem. Int. Ed. 48, 3584 – 3596 (2009). 14. Rammensee, S., Slotta, U., Scheibel, T. & Bausch, A. R. Assembly mechanism of recombinant spider silk proteins. Proc. Natl Acad. Sci. USA 105, 6590 – 6595 (2008). 15. Lin, Z., Huang, W., Zhang, J., Fan, J. S. & Yang, D. Solution structure of eggcase silk protein and its implications for silk fiber formation. Proc. Natl Acad. Sci. USA 106, 8906 – 8911 (2009). 16. Exler, J. H., Hu¨mmerich, D. & Scheibel, T. The amphiphilic properties of spider silks are important for spinning. Angew. Chem. Int. Ed. 46, 3559 – 3562 (2007). 17. Hardy, J. G. & Scheibel, T. R. Silk-inspired polymers and proteins. Biochem. Soc. Trans. 37, 677 – 681 (2009). 18. Lavinder, J. J., Hari, S. B., Sullivan, B. J. & Magliery, T. J. High-throughput thermal scanning: a general, rapid dye-binding thermal shift screen for protein engineering. J. Am. Chem. Soc. 131, 3794 – 3795 (2009). 19. Ve ´zy, C., Hermanson, K. D., Scheibel, T. & Bausch, A. R. Interfacial rheological properties of recombinant spider-silk proteins. Biointerphases 4, 43 – 46 (2009). 20. Lupas, A. Coiled coils: new structures and new functions. Trends Biochem. Sci. 21, 375 – 382 (1996). 21. Brown, J. C. & Timpl, R. The collagen superfamily. Int. Arch. Allergy Immunol. 107, 484 – 490 (1995). 22. Sacca `, B., Renner, C. & Moroder, L. The chain register in heterotrimeric collagen peptides affects triple helix stability and folding kinetics. J. Mol. Biol. 324, 309 – 318 (2002). 23. Lees, J. F. & Bulleid, N. J. The role of cysteine residues in the folding and association of the COOH-terminal propeptide of types I and III procollagen. J. Biol. Chem. 269, 24354 – 24360 (1994). 24. Huemmerich, D. et al. Primary structure elements of spider dragline silks and their contribution to protein solubility. Biochemistry 43, 13604 – 13612 (2004). 25. Sattler, M., Schleucher, J. & Griesinger, C. Heteronuclear multidimensional NMR experiments for the structure determination of proteins in solution employing pulsed field gradients. Prog. Nucl. Magn. Reson. Spectr. 34, 93 – 158 (1999). 26. Diercks, T., Coles, M. & Kessler, H. An efficient strategy for assignment of crosspeaks in 3D heteronuclear NOESY experiments. J. Biomol. NMR 15, 177 – 180 (1999). 27. Schwieters, C. D., Kuszewski, J. J., Tjandra, N. & Clore, G. M. The Xplor-NIH NMR molecular structure determination package. J. Magn. Reson. 160, 65 – 73 (2003). Supplementary Information is linked to the online version of the paper at www.nature.com/nature. Acknowledgements This paper is in memoriam of R. Rudolph who triggered spider silk research in the lab of T.S. We want to thank D. Hu¨mmerich and J. Exler for initial work on the project, S. Quedzuweit for sample preparation in early stages of the project, M. Heim and M. Suhre for cloning work, and H. Krause for mass spectrometry. This work was supported by the Center for Integrated Protein Science Munich (CIPS M ) (to H.K. and T.S.) and the Deutsche Forschungsgemeinschaft SCHE 603/4-3 (to T.S.). F.H. was supported by the Elitenetzwerk Bayern, CompInt. J.G.H. was supported by the Alexander von Humboldt Foundation. Author Contributions F.H. designed research, performed cloning work, purified proteins, performed protein folding studies, recorded NMR experiments, performed structure calculations and wrote the manuscript; L.E. performed cloning work, purified proteins and performed aggregation assays. J.G.H. performed aggregation assays, characterized the fibres and wrote the manuscript; C.V. performed cloning work and purified proteins; M.C. was involved in structure calculation and provided software tools for structure calculation and analysis; T.S. designed research and wrote the manuscript; H.K. designed research and wrote the manuscript. All authors discussed the results and commented on the manuscript. Author Information The resonance assignment obtained was deposited at the BMRB data bank under accession code 16249 and the atomic coordinates of the best 20 structures plus a regularized average structure have been deposited at the Protein Data Bank under accession code 2khm. Reprints and permissions information is available at www.nature.com/reprints. The authors declare no competing financial interests. Correspondence and requests for materials should be addressed to H.K. ([email protected]) or T.S. ([email protected]). LETTERS NATURE | Vol 465 | 13 May 2010 242 Macmillan Publishers Limited. All rights reserved ©2010 METHODS Cloning, protein production and purification. The initial construct used for NMR resonance assignment (NR3-T7) was produced and purified as described previously 24 . The gene of a shorter construct (NR3D19) used for NOESY experiments was inserted into a pET28a expression system (Merck Biosciences)encoding an N-terminal His 6 -tag and expressed in E.coliBL21 (DE3) (Agilent Technologies) at 20 uC for 16–20 h. Purification was achieved by passage over a Ni-NTA column (GE healthcare) and a Superdex75 column. Between both columns the hexahistidine tag was removed by digestion with thrombin. Isotopically ( 13 C, 15 N) enriched protein samples were produced by growing bacteriain M9 medium supplemented with 1 g l 2115 NH 4 Cl and 2 g l 2113 C glucose (Eurisotope). For NMR measurements, the proteins were concentrated up to 1 mM in 10mM sodium phosphate pH 6.0. Addition of 1% (v/v) trifluoroethanol-d 3 (Eurisotope) was used for sample stabilization. The generation of a mixed 13 C, 15 N– 12 C, 14 N dimeric sample was achieved by mixing equal amounts of labelled and unlabelled His 6 -tagged protein samples and a subsequent addition of guanidine hydrochloride to a final concentration of 4 M and 50 mM DTT to achieve unfolding and reduction of the intermolecular disulphide bridge. For refolding, the mixture was dialyzed against 50 mM Tris/HCl pH 8.0 resulting in an almost quantitative refolding yield. By this procedure, the monomers are allowed to exchange resulting in a 1:2:1 population of the homo-unlabelled, the hetero-labelled and the homo13 C, 15 N-labelled species. Genes encoding proteins containing the repetitive elements (AQ) 12 and (AQ) 24 (A, hydrophobic polyalanine-rich motif: GPYGPGASAAAAAAGGYGPG SGQQ; Q, hydrophilic glutamineand glycine-rich motif: GPGQQGPGQQGP GQQGPGQQ) were cloned into a pET29 expression vector (Merck Biosciences) encoding an N-terminal T7-tag and were expressed in E. coli strain BLR (DE3) (Agilent Technologies) at 25 uC for 3 h. Purification was carried out as described previously 24 . Site-directed mutagenesis experiments with the gene of NR3D19were done using the QuikChange mutagenesis kit (Agilent Technologies). Circular dichroism (CD) spectroscopy. CD spectra and stability measurements were carried out on a Jasco J-715 spectropolarimeter (Jasco). For spectra and thermal transitions, a protein concentration of 10 mM was used, and thermal unfolding was monitored by the CD signal at 222 nm in a 1 mm path length cuvette using a heating rate of 60 K h 21 in 10 mM sodium phosphate pH 6.0. For chemical unfolding experiments, samples of 5, 34, 110, 200 and 230 mM concentration were incubated with increasing amounts of urea overnight at 4 uC, and the CD signal at 222 nm was recorded with a path length of 0.1 or 1 cm, respectively, at 20 uC. A bandwidth of 5 nm and a response of 2 s were used. Fitting of the thermally and urea-induced unfolding profiles was done as reported previously 28 assuming a two-state folding mechanism. NMR spectroscopy and structure calculation. Nuclear magnetic resonance (NMR) spectroscopic measurements were conducted at 298K on Bruker DMX600, DMX750 and Avance900 spectrometers (Bruker Biospin). Backbone resonance assignment was achieved by a set of standard triple resonance experiments 25 . Side-chain assignment was carried out using a combination of CCHCOSY (correlation spectroscopy via the nuclei C,C,H 29 ) and HCH-TOCSY ,(total correlation spectroscopy 30 ) on double-labelled samples. Stereospecific assignment of prochiral H Cb methylene and valine methyl groups and the resulting rotamer assignment were made using 3 J NHb couplings observed in the HNHB experiment and patterns of NOESY connectivities. Distance data were derived from a set of three-dimensional NOESY experiments including 15 N-HSQCNOESY, NNH-NOESY on a 15 N-labelled sample and 13 C-HSQC-NOESY and heteronuclear edited 3D-CCH NOESY and 3D-CNH NOESY experiments 26 . For obtaining inter-subunit NOE distance restraints, we used a 14 N, 12 Cfiltered/ 13 C edited two-dimensional-NOESY experiment on a mixed 13 C, 15 N– 12 C, 14 N labelled sample. NOESY cross-peaks in the three-dimensional spectra were converted into distance ranges after rescaling according to the corresponding HSQC intensities. Cross-peaks were divided into the following four classes: strong, medium, weak and very weak, which resulted in restraints on upper distance of 2.7, 3.2, 4.0 and 5.0 A ˚, respectively. Lower distance restraints were also included for very weak and absent sequential H N –H N cross-peaks using a minimum distance of 3.2 A ˚and for medium intensity or weaker sequential and intraresidual H N -H a cross-peaks using a minimum distance of 2.7A ˚. Allowance for the use of pseudo atoms (using r 26 averaging) were added for methyl groups (0.8 A ˚for one methyl and 1.5 A ˚for two methyls) and nonstereospecificallyassigned methylene groups (0.7 A ˚). Dihedral angle restraints were derived for backbone wand yangles based on C a ,C b ,C9and H a chemical shifts using the program TALOS 31 . Restraints were applied for predictions consistent with the input chemical shifts, 3 J HNHa coupling constants measured from an HNHA experiment and the observed NOE pattern within sequential residues. Accepted predictions were applied using the tolerance calculated by the program 65u. The 3 J HNHa couplings were also applied as direct coupling constant restraints on the backbone wangles. Hydrogen bond restraints were applied for 73 residues in secondary structural elements with low hydrogen exchange rates, as measured with MEXICO experiments 32,33 and where donor and acceptor were consistently identified in preliminary calculations. The restraints were applied with the inclusion of pseudo covalent bonds as described previously 34 . Structures were calculated with XPlor-NIH 27 using standard protocols. Experimental restraints were applied to only one monomer, with non-crystallographic symmetry restraints over the backbone of ordered residues (S34–L138) used to assure the symmetry of the dimer. Sets of 50 structures were calculated and a final set of 20 structures chosen on the basis of lowest restraints violations. An average structure was calculated and regularized to give a structure representative of the ensemble. The quality of the regularized average structure was assessed with the program PROCHECK-NMR 35 . Titration experiments were performed with 200–500 mM NR3 in 10mM sodium phosphate pH 6.0 at 293 K using standard 15 N-HSQC experiments. Chemical shift deviations were calculated as the mean of 1 H and 15 N chemical shift differences using following equation: Ddav~ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi DdH2z(DdN=5)2 q  =2. Fluorescence spectroscopy. Fluorescence spectra of 10mM ANS and SYPRO Orange in 10 mM sodium phosphate pH 6.0 were recorded with a Fluoromax4 spectrofluorometer (Horiba Jobin Yvon) using excitation at 350 nm (5 nm bandwidth) and by collecting emission spectra from 420 to 600 nm with a bandwidth of 5 nm. With SYPRO Orange, excitation was set to 485 nm and emission was recorded from 520 to 700 nm. Thermofluor experiments 36 were performed with a real-time PCR machine (Agilent Technologies) by detection of ANS fluorescence (10 mM ANS was used). Heating rate was 30 K h 21 , protein concentration was 50 mM and sample volume was 50 ml in a 96-well plate. Five individual measurements were done for each protein sample and solvent condition. Differential scanning calorimetry (DSC). DSC experiments were performed using a VP-DSC Micro-Calorimeter (MicroCal). Standard measurements were performed at a protein concentration of 50mM in 10 mM Tris/HCl, pH 8.0, with a heating rate of 20 K h 21 . Evaluation of the LCST behaviour was done using the ORIGIN software (MicroCal). Rotation-induced aggregation. Before the experiment the proteins were dissolved in guanidine thiocyanate solution (6 M) and then dialyzed against 10 mM Tris/HCl, pH 8.0. Then the proteins were diluted into buffer containing 10 mM Tris/HCl, pH 8 with 50 mM NaCl. Final protein concentrations were adjusted to 40 mM of NR3, 14 mMofAQ 12 NR3 and 8 mMofAQ 24 . The samples were incubated for 12 h at 25 uC rotating with 25 r.p.m. After rotation, the clearly visible aggregates were transferred onto glass slides for optical microscopy (DMI3000B, Leica). The samples were then centrifuged at 125,000gto remove non-visible aggregates, and the amount of remaining soluble protein in the supernatant was determined using an ultraviolet spectrometer (NanoDrop ND1000, Thermo Fischer). Fourier transform infrared (FTIR) spectroscopic studies. FTIR spectra of fibres were recorded on a liquid-nitrogen-cooled Bruker Tensor 27 FTIR spectrometer, fitted with a liquid-nitrogen-cooled Bruker Hyperion microscope with IR polarizers supplied by Optometrics Corporation. Spectroscopy was performed on fibres formed during the previously described aggregation assay that were washed with water to remove salts, air dried and placed on highly polished optical grade calcium fluoride disks (Crystal GmbH). Spectra of all aggregates were recorded in absorbance mode at 21 uC, with a 4 cm 21 resolution and 600 scans (corrected for background and atmosphere using OPUS software), and polarized FTIR spectra of aggregates of micrometre scale (and larger) were recorded at three different orientations (0u,45uand 90urelative to the long axis of the fibrous aggregates) to determine if any of the secondary structural elements were aligned with the fibres. 28. Privalov, P. L. Stability of proteins: small globular proteins. Adv. Protein Chem. 33, 167 – 241 (1979). 29. Ikura, M., Kay, L. E. & Bax, A. Improved three-dimensional 1 H13 C1 H correlation spectroscopy of a 13 C-labeled protein using constant-time evolution. J. Biomol. NMR 1, 299 – 304 (1991). 30. Kay, L. E., Ikura, M. & Bax, A. Proton – proton correlation via carbon – carbon couplings: a three-dimensional NMR approach for the assignment of aliphatic resonances in proteins labeled with carbon 13. J. Am. Chem. Soc. 112, 888 – 889 (1990). 31. Cornilescu, G., Delaglio, F. & Bax, A. Protein backbone angle restraints from searching a database for chemical shift and sequence homology. J. Biomol. NMR 13, 289 – 302 (1999). 32. Gemmecker, G., Jahnke, W. & Kessler, H. Measurement of fast proton-exchange rates in isotopically labeled compounds. J. Am. Chem. Soc. 115, 11620 – 11621 (1993). 33. Koide, S., Jahnke, W. & Wright, P. E. Measurement of intrinsic exchange-rates of amide protons in a 15 N-labeled peptide. J. Biomol. NMR 6, 306 – 312 (1995). 34. Truffault, V. et al. The solution structure of the N-terminal domain of riboflavin synthase. J. Mol. Biol. 309, 949 – 960 (2001). doi:10.1038/nature08936 Macmillan Publishers Limited. All rights reserved ©2010 6 www.nature.com/nature doi: 10.1038/nature08936 SUPPLEMENTARY INFORMATION  7 Supplementary Figure 6 Supplementary Figure 6: Structural changes within the NR domain under shear stress are independent of the presence of sodium chloride. 200 µM NR3 in 10 mM Na-phosphate pH6.0 was shaken at 900 rpm at 37°C with 250mM (red symbols) or without (grey symbols) NaCl for the indicated time. For each time point, 10µl of the NR3 solution was mixed with 490 µl buffer supplemented with 10 µM ANS for fluorescence measurements or 290 µl buffer for CD measurements. ANS (an environment sensitive dye) shows high fluorescence intensity in hydrophobic environment (NR domain unfolded) accompanied by a blue-shift of the fluorescence emission maximum. 7 www.nature.com/nature SUPPLEMENTARY INFORMATION doi: 10.1038/nature08936  8 Supplementary Tables Supplementary Table 1: Sequences of the NR3 constructs used for NMR structure determination. NR3 – T7-Tag# 10 20 30 40 50 60 MASMTGGQQM GRGSMGAASA AVSVGGYGPQ SSSAPVASAA ASRLSSPAAS SRVSSAVSSL 70 80 90 100 110 120 VSSGPTNQAA LSNTISSVVS QVSASNPGLS GCDVLVQALL EVVSALVSIL GSSSIGQINY 130 140 GASAQYTQMV GQSVAQALAG NR3Δ19# 10 20 30 40 50 60 GSHMAAVSVG GYGPQSSSAP VASAAASRLS SPAASSRVSS AVSSLVSSGP TNQAALSNTI 70 80 90 100 110 120 SSVVSQVSAS NPGLSGCDVL VQALLEVVSA LVSILGSSSI GQINYGASAQ YTQMVGQSVA QALAG # red letters indicate remaining peptide tags or remaining amino acids after tag removal, respectively. Numbering of residues in the structure is according to the first construct (NR3-T7-Tag). 8 www.nature.com/nature doi: 10.1038/nature08936 SUPPLEMENTARY INFORMATION  9 Supplementary Table 2: Structural statistics of the 20 lowest energy structures of the NR3 dimer. Rmsd from distance restraints (Å)a SAb <SA>r all (1313) 0.0191±0.0009 0.019 intraresidual (337) 0.0114±0.0005 0.011 sequential (398) 0.021±0.001 0.021 medium range (162) 0.015±0.001 0.015 long range (67) 0.026±0.003 0.023 intermonomer (112) 0.020±0.002 0.018 Rmsd from dihedral restraints (174) 0.071±0.005 0.078 Rmsd from J-coupling restraints (Hz)(58) 0.45±0.01 0.47 H-bonds restraints (Å /deg) (73)c 2.19±0.12 / 19.0±6.9 2.13±0.10 / 16.8±6.5 Deviations from ideal covalent geometry Bonds (Å 10-3) 3.637±0.023 3.742 Angles (Å) 0.479±0.003 0.487 Impropers (Å) 1.422±0.046 1.541 Ramachandran Map regions (%)d (residues 34-140) 96.8 / 3.2 / 0.0 / 0.0 96.8 / 3.2 / 0.0 / 0.0 SA versus <SA> SA versus <SA>r Atomic Rmsd (Å)e backbone all backbone all Ordered residues (34-138) 0.18±0.04Å 0.37±0.05Å 0.30±0.06Å 0.51±0.06Å a numbers in brackets indicate the number of restraints per monomer b structures are labeled as follows: SA, set of 20 final simulated annealing structures; <SA>, the mean structure calculated by averaging the coordinates of SA structures after fitting over secondary structure elements; <SA>r, the structure obtained by regularising the mean structure under experimental restraints. c H bonds were restrained by treating them as pseudocovalent bonds (see SI methods). Deviations are expressed as the average distance/average deviation from linearity for restrained H bonds. d Determined using the program PROCHECK-NMR. Percentages are for residues in allowed/additionally allowed/generously allowed/disallowed regions of the Ramachandran map. Values are calculated for the regularized structure. e Based on heavy atoms superimpositions.  Supplementary References 1 Privalov, P. L. Stability of proteins: small globular proteins. Adv Protein Chem 33, 167-241 (1979). Teilarbeiten [73] Teilarbeit II Eisoldt, L., Hardy, J.G., Heim, M., Scheibel, T. (2010) The role of salt and shear on the storage and assembly of spider silk proteins. J. Struct. Biol. 170, pp. 413-419 The role of salt and shear on the storage and assembly of spider silk proteins Lukas Eisoldt, John G. Hardy, Markus Heim, Thomas R. Scheibel * Fakultät für Angewandte Naturwissenschaften, Universität Bayreuth, Universitätsstr. 30, 95440 Bayreuth, Germany article info Article history: Received 12 November 2009 Received in revised form 22 December 2009 Accepted 27 December 2009 Available online 4 January 2010 Keywords: Spider silk Protein assembly Salt Shear abstract Major ampullate silk fibers of orb web-weaving spiders have impressive mechanical properties due to the fact that the underlying proteins partially fold into helical/amorphous structures, yielding relatively elastic matrices that are toughened by anisotropic nanoparticulate inclusions (formed from stacks of b-sheets of the same proteins). In vivo the transition from soluble protein to solid fibers involves a combination of chemical and mechanical stimuli (such as ion exchange, extraction of water and shear forces). Here we elucidate the effects of such stimuli on the in vitro aggregation of engineered and recombinantly produced major ampullate silk-like proteins (focusing on structure–function relationships with respect to their primary structures), and discuss their relevance to the storage and assembly of spider silk proteins in vivo. Ó2009 Elsevier Inc. All rights reserved. 1. Introduction Orb web-weaving spiders produce a number of task-specific silk protein-based fibers. One of these fibers is assembled from proteins produced by the major ampullate (MA) silk gland. MA silk fibers are used both in the construction of the web as the frame upon which to attach the capture spiral fibers (flagelliform silk fibers coated with aggregate silk) and as a lifeline to escape from predators (Aprhisiart and Vollrath, 1994; Lin et al., 1995). MA silk fibers of orb web-weaving spiders have impressive mechanical properties due to the fact that the underlying proteins partially fold into helical/amorphous structures yielding relatively elastic matrices that are toughened by anisotropic nanoparticulate inclusions (formed from stacks of b-sheets of the same proteins) (Gosline et al., 1999). The primary structures of MA proteins are reminiscent of amphiphilic multiblock copolymers (Exler et al., 2007; Geisler et al., 2008; Zbilut et al., 2005, 2006), in which the repetitive sequence elements (typically composed of multiple repeats of A n , (GA) n , (GGX) n , and (GPGXX) n ) are flanked by highly conserved non-repetitive (NR) aminoand carboxy-terminal domains (Bini et al., 2004; Foo et al., 2006; Hedhammar et al., 2008; Lewis, 2006; Rising et al., 2006). In common with other silk proteins, the MA proteins of Araneus diadematus spiders (ADF-3 and ADF-4) are stored at very high concentrations (>30 wt.%) without the onset of undesirable aggregation within the lumen of the spider. However, our in vitro studies of the repetitive sequence elements of recombinantly produced proteins based upon the consensus sequences of the MA proteins of Araneus diadematus spiders (Guerette et al., 1996) without the NR terminal domains clearly demonstrate that these proteins are not soluble at such high concentrations as found in vivo (Huemmerich et al., 2004a, 2004b). Our findings suggest that the NR terminal domains control both the storage of the proteins (by increasing their solubility and/or deterring undesirable aggregation events) and fiber production (allowing/facilitating controlled protein assembly into fibers upon demand) (Hagn et al., in press). In vivo Raman spectromicroscopy (Lefevre et al., 2008, 2007b) and in vitro 13 C NMR (Hijirida et al., 1996) studies of the spinning dope indicate that the repetitive sequence elements adopt highly hydrated random coil and 3 1 -helix (polyproline II-like) secondary structures which can be easily transformed into b-sheets. We (Huemmerich et al., 2004b) and subsequently others (Huang et al., 2006; Ittah et al., 2007, 2006; Lin et al., 2009; Stark et al., 2007) have shown the C-terminal NR domains to adopt a predominantly a -helical conformation in vitro. These domains have been postulated to form the outer layer of droplet-like structures observed in the spinning dope in vivo (Jin and Kaplan, 2003; Lin et al., 2009; Vollrath and Knight, 1999). We have also shown that the C-terminal NR domain is a prerequisite of certain supramolecular self-assembly processes (manifested in the form of fully reversible lower critical solubilisation temperature behavior) of our recombinantly produced proteins (Exler et al., 2007). In vivo the transition from soluble protein to solid fibers involves a combination of chemical and mechanical stimuli (such as ion exchange, acidification, extraction of water and shear forces (Bini et al., 2004; Foo et al., 2006; Heim et al., 2009; Vollrath and Knight, 2001) that have also been demonstrated to promote protein assembly into fibers in vitro (Hardy and Scheibel, 2009a; Hardy et al., 2008; Rammensee et al., 2008). We were consequently 1047-8477/$ - see front matter Ó2009 Elsevier Inc. All rights reserved. doi:10.1016/j.jsb.2009.12.027 *Corresponding author. E-mail address: [email protected] (T.R. Scheibel). Journal of Structural Biology 170 (2010) 413–419 Contents lists available at ScienceDirect Journal of Structural Biology journal homepage: www.elsevier.com/locate/yjsbi interested in determining the effect of such stimuli upon the stability of our MA silk-like proteins (Hardy and Scheibel, 2009b). A protein’s solubility in aqueous solution is governed by many factors, of which, one of the most important is its primary structure, which also determines it’s secondary, tertiary and quaternary structures (Levy et al., 2006; Rossmann and Argos, 1981; SenGupta and Scheibel, 2007). The presence of other solutes (such as ions) also has an effect upon a protein’s solubility in water, as they interfere with the highly ordered layer of water (known as the hydration layer) on the protein’s surface (Gerstein and Chothia, 1996; Kim and Cremer, 2001; Lesk et al., 1980; Zhang and Cremer, 2006). Low concentrations of salt tend to improve the solubility of proteins (known as ‘salting in’), due to the formation of ion-rich hydration layers in the vicinity of charged and polar amino acid residues (as described by the Debye–Hückel theory), whereas high concentrations of salt tend to have the opposite effect, causing the protein to precipitate (known as ‘salting out’). The magnitude of this effect is dependent upon the particular ions and usually follows the Hofmeister series, in which ‘chaotropic’ ions favor salting-in of proteins and ‘kosmotropic’ ions favor salting-out of proteins, and anions are well-known to have a much greater effect than cations (Baldwin, 1996; Geisler et al., 2008; Gurau et al., 2004; Horinek et al., 2008; Pegram and Record, 2008; Pirzer et al., 2009; Zhang and Cremer, 2006; Zhang et al., 2005). Knight and Vollrath have investigated the MA gland of Nephila edulis spiders, finding that the physiological concentration of sodium chloride in the spinning dope in the lumen to be of the order of 150 mM, and the total concentration of salts was observed to decrease during the spinning process. They also found that as the dope flowed through the spinning duct, sodium cations were replaced by potassium cations, yet other metal cations (e.g. calcium, copper or magnesium) were almost undetectable in the lumen, duct and the naturally spun fibers. Furthermore, they observed that weakly chaotropic chloride anions are replaced with strongly kosmotropic phosphate anions (significantly increased concentrations) and sulfate anions (small increase with a relatively high level of experimental error) during the passage of the spinning dope along the spinning duct (Knight and Vollrath, 2001). Here we investigate the effects of salts (sodium chloride and sodium phosphate) and shear on the in vitro aggregation of our recombinantly produced MA silk-like proteins, and discuss their relevance to the storage and assembly of spider silk proteins in vivo. In this study we use recombinantly produced silk proteins based on the consensus motifs of the MA silk Aranaeus diadematus fibroin 3 (ADF3) from the garden cross spider. The repetitive core domains of our proteins are composed of two different sequence modules, denoted A and Q, where module A is a hydrophobic polyalanine-rich motif, and module Q is a more hydrophilic glutamineand glycine-rich motif (see materials and methods). In this study we use proteins containing 12 or 24 repeats of (AQ). The C-terminal NR3 domain is derived from adf3 (gi|1263286, obtained from J. Gosline, Vancouver, BC) by using PCR. This domain is mainly a - helical and dimerises via disulfide bond formation between the single cysteine residues contained in two individual NR3 domains (Huemmerich et al., 2004b)(Fig. 1); therefore proteins bearing the NR3 domain are dimeric under non-reducing conditions. 2. Materials and methods 2.1. Cloning, protein expression and purification The proteins (AQ) 12 NR3 (dimer of 116 kDa), (AQ) 24 NR3 (dimer of 212 kDa) and (AQ) 24 (monomer of 95 kDa), containing the repetitive elements (A, hydrophobic polyalanine-rich motif: GPYGPGASAAAAAAGGYGPGSGQQ; Q, hydrophilic glutamineand glycine-rich motif: GPGQQGPGQQGPGQQGPGQQ) were cloned, expressed and purified as previously reported. A D93A mutant (numbering according to the residues of the NR domain; see Fig. 1C) of (AQ) 12 NR3 was obtained according to the QuikChange protocol from Stratagene (USA). 2.2. Salt-induced aggregation assay Experiments were carried out in accordance with our previously described protocol (Huemmerich et al., 2004b). Prior to the experiment the proteins were dissolved in guanidinium thiocyanate solution (6 M) and then dialyzed against 10 mM Tris/HCl, pH 8.0, 50 mM NaCl. The assay was started by the addition of buffered (10 mM Tris/HCl, pH 8.0) solutions of NaCl and NaH 2 PO 4 (pH 8.0). Protein concentrations were, respectively, 8.8 l M, 5.3 l M and 4.7 l M for AQ 12 NR3 (wt and D93A), AQ 24 and AQ 24 NR3. After 1 h of incubation at 25 °C the samples were centrifuged at 17700 g to remove any visible aggregates, and the amount of soluble protein remaining in the supernatant was determined using a UV spectrometer (NanoDrop ND1000, ThermoFischer, USA). 2.3. Shear-induced aggregation assay Prior to the experiment the proteins were dissolved in guanidinium thiocyanate solution (6 M) and then dialyzed against 10 mM Tris/HCl, pH 8.0. Then the proteins were diluted into buffer containing 10 mM Tris/HCl, pH 8.0 with 50 mM NaCl. Final protein concentrations were adjusted to 17.6 l MofAQ 12 NR3 and 10.6 l M of AQ 24 . The samples were incubated for 16 h at 25 °C without/ with rotation at 25 rpm (Intellimixer RM-2, NeoLab, Germany). After rotation, the clearly visible aggregates were transferred onto A B C Fig. 1. Employed recombinantly produced proteins based on the consensus sequences of the major ampullate protein (ADF-3) of Araneus diadematus spiders. (A) A schematic representation of the repetitive sequence elements (A, hydrophobic polyalanine-rich motif and Q, hydrophilic glutamineand glycine-rich motif) and the C-terminal non-repetitive domain (NR3). (B) A schematic representation of the engineered and recombinantly produced major ampullate silk-like proteins composed of the elements shown above. For each protein the molecular weight in kDa is given in brackets. Note that proteins bearing the NR domain are dimeric due to interdomain disulfide bond formation. (C) The primary sequence of the C-terminal non-repetitive domain (NR3) of the major ampullate protein (ADF-3) of Araneus diadematus spiders, highlighting the residues involved in salt bridges and the cysteine residue responsible for interprotein dimerization. In the (AQ) 12 NR3 (D93A) mutant the single aspartic acid residue is replaced by an alanine residue, inhibiting the formation of one of the two possible salt bridges. 414 L. Eisoldt et al. / Journal of Structural Biology 170 (2010) 413–419 glass slides for optical microscopy (DMI3000B, Leica, Switzerland). The samples were then centrifuged at 125000 g to remove non-visible aggregates, and the amount of remaining soluble protein in the supernatant was determined using a UV spectrometer (NanoDrop ND1000, ThermoFischer, USA). The percentage of aggregated protein was calculated as described in section 2.2. 2.4. Fourier transform infrared (FTIR) spectroscopic studies FTIR spectra of fibers were recorded on a liquid nitrogen cooled Bruker Tensor 27 FTIR spectrometer, fitted with a liquid nitrogen cooled Bruker Hyperion microscope with IR polarizers supplied by Optometrics Corporation (Ayer, MA, USA). Spectroscopy was performed on fibers formed during the previously described aggregation assay that were washed with water to remove salts, air dried and placed on highly polished optical grade calcium fluoride disks (Crystal GmbH, Berlin, Germany). Spectra of all aggregates were recorded in absorbance mode at 21 °C, with a 4 cm 1 resolution and 60 scans (corrected for background and atmosphere using OPUS software), and polarized FTIR spectra of aggregates of l m scale (and larger) were recorded at three different orientations (0°,45°and 90°relative to the long axis of the fibrous aggregates) to determine if any of the secondary structural elements were aligned with the fibers. 2.5. Transmission electron microscopy Prior to the experiment the proteins were dissolved in guanidinium thiocyanate solution (6 M) and then dialyzed against 10 mM Tris/HCl, pH 8.0. A droplet (3 l L) was placed on pioloform-coated copper grids and the excess liquid was removed by wicking using a Kimwipe (Kimberly-Clark Corporation, USA). The grid was then 100 80 60 40 20 0 100 200 300 400 500 ]%[ ytilibulos nietorp 100 200 300 400 500 100 80 60 40 20 0 ]%[ ytilibulos nietorp salt concentration [mM] A C 100 200 300 400 500 B 100 200 300 400 500 salt concentration [mM] D (AQ)24 (AQ)12NR3 no shear no shear shear shear solubility [%] 0 20 40 60 80 100 E Fig. 2. Salt-induced aggregation of engineered and recombinantly produced major ampullate silk-like proteins after incubation for 1 h at RT: induced by the presence of sodium chloride (grey circles) and by the presence of sodium phosphate (black circles), the concentration regime required to aggregate the proteins is indicated by striped boxes. (A) (AQ) 12 NR3 (wild type). (B) (AQ) 12 NR3 (D93A). (C) (AQ) 24 NR3 (wild type). (D) (AQ) 24 . (E) A comparison of the degree of aggregation of our spider silk proteins with and without shear; induced in the presence of 50 mM sodium chloride (grey bars), or 500 mM sodium chloride (black bars). L. Eisoldt et al. / Journal of Structural Biology 170 (2010) 413–419 415 stained with uranyl acetate (1% in water, pH 4.5) and the excess liquid was removed by wicking using a Kimwipe, and allowed to dry in a dust free dessicator. The grids were viewed with a JEOL JEM2100 transmission electron microscope (Jeol, Eching, Germany). 3. Results and discussion 3.1. The effect of salt on protein aggregation As noted in the introduction, MA spidroins are stored in the lumen of the gland in the presence of sodium and chloride ions, and during fiber spinning these ions are exchanged for potassium and phosphate ions (Knight and Vollrath, 2001). Here we investigate the effects of chloride and phosphate anions on the in vitro aggregation of our recombinantly produced MA silk-like proteins (Fig. 1). Our solubility/aggregation assays in the absence of shear forces (similar to the situation of the silk proteins in the lumen of the gland) show that the extent of protein aggregation is determined by the concentration of salt. Our assays with sodium chloride (found in the lumen) indicate that the level of aggregation is low (less than 20%) for all of the tested engineered proteins, even at extremely high salt concentrations (up to 500 mM); whereas in the presence of equivalent concentrations of sodium phosphate (found in the spinning duct) the level of aggregation for each protein is 100 µm 100 µm 20 µm AB DC 1.0 0.8 0.6 0.4 0.2 0 a.u. wavenumber [cm-1 mc[rebmunevaw] -1] 0.30 0.25 0.20 0.15 0.10 0.05 0 a.u. 3500 3000 1500 1000 3500 3000 1500 1000 1100 1050 1000 950 900 1100 1050 1000 950 900 20 µm EF HG Fig. 3. Shear-induced aggregation of our recombinantly produced major ampullate silk-like proteins. (A) Optical microscope image of aggregates produced by shear-induced aggregation of aqueous solutions of (AQ) 24 . (B) Magnified image of the (AQ) 24 aggregates shown in A) (white frame). (C) Optical microscope image of well-defined fibrous aggregates produced by shear-induced aggregation of aqueous solutions of (AQ) 12 NR3. (D) Magnified image of one part of the (AQ) 12 NR3 fiber (black frame). (E) Polarized FTIR spectra from (AQ) 24 aggregates from A/B recorded at 0°(black circles) and 90°(grey squares) relative to the long fiber axis. (F) Polarized FTIR spectra from AQ 12 NR3 fibrils from C/D recorded at 0°(black circles) and 90°(grey squares) relative to the long fiber axis. (G) Magnification of the region between 1100 and 900 cm 1 from the AQ 24 spectra (E). The arrow indicates the characteristic polyalanine b-sheet band at around 960 cm 1 (Papadopoulos et al., 2009). (H) Magnification of the region between 1100 and 900 cm 1 from the (AQ) 12 NR3 spectra (F). The arrow indicates the characteristic polyalanine b-sheet band at around 960 cm 1 . 416 L. Eisoldt et al. / Journal of Structural Biology 170 (2010) 413–419 notably higher (up to 100%) due to the more kosmotropic nature of the phosphate anion (Fig. 2). Therefore, weakly chaotropic chloride anions aid the long term storage of MA spidroins in the lumen (salting-in) and are substituted for strongly kosmotropic phosphate anions (salting out) during the natural fiber spinning process. 3.2. The effect of shear on protein aggregation During the passage of the spinning dope from the lumen through the spinning duct to the spigot (the point at which the fiber exits the animal), the dope is exposed to shear forces which are known to encourage aggregation in a variety of globular and fibrillar proteins (Bekard and Dunstan, 2009; Cromwell et al., 2006; Hamilton-Brown et al., 2008; Hill et al., 2006; Vezy et al., 2009; Yamaura et al., 1982, 1985); we were therefore keen to find out if shear forces have an effect upon the aggregation of our recombinantly produced MA silk-like proteins in vitro. Our aggregation assays in the presence of shear forces (as the silk proteins are exposed to in the duct during fiber spinning) show that the extent of protein aggregation is dramatically increased in relation to corresponding assays in the absence of shear, in the presence of both low and high concentrations of salt; moreover, the aggregates formed after longer periods of exposure to shear were fibrous with dimensions of l m to mm scale (Fig. 3). Particularly noteworthy is that water is absorbed from the dope during its passage along the spinning duct (Kojic et al., 2004; Tillinghast et al., 1984), which increases the concentration of proteins and consequently increases the shear forces between the proteins in the dope due to strengthened hydrophobic interactions. This leads us to suggest that shear forces are likely to play a greater role than salt effects during the fiber spinning process in vivo (Casem et al., 2002; Knight and Vollrath, 2002; Perez-Rigueiro et al., 2001a, 2001b). 3.3. The role of the C-terminal NR domain in fiber storage and assembly It is widely accepted that higher order supramolecular assemblies are required to achieve silk protein storage at high concentrations within the lumen, and as noted in the introduction, we have previously shown the highly conserved C-terminal NR domain to be a prerequisite of certain supramolecular self-assembly processes (manifested in the form of fully reversible lower critical solubilisation temperature behavior) of our recombinantly produced proteins (Exler et al., 2007). Interestingly, our transmission electron microscope studies of freshly dialyzed solutions of (AQ) 12 NR3 show evidence of vesicle-like supramolecular assemblies (of nm to l m scale), which is not the case for (AQ) 24 , a protein of similar molecular weight without the C-terminal NR domain (Fig. 4), which has also been observed by light microscopy in a previous study (Rammensee et al., 2008). Our salt-induced aggregation studies revealed no significant differences in the behavior of (AQ) 12 NR3 and (AQ) 24 in the presence of sodium chloride, whereas in the presence of sodium phosphate (>200 mM) (AQ) 24 is much more prone to aggregation (Fig. 2A–D) suggesting that the C-terminal NR domain dependent supramolecular assemblies are stabilized to a certain degree against aggregation. Shear stress had a even pronounced effect on protein aggregation (Fig. 2E). Furthermore, exposure of solutions of (AQ) 24 to shear stress resulted in the formation of ill-defined b-sheet rich aggregates (Fig. 3A and B and Supplementary Table 1), whereas exposure of solutions of (AQ) 12 NR3 to shear stress yielded relatively well-defined fibrous aggregates which optical microscopy revealed to be composed of ordered (aligned) bundles of long fibrous aggregates (Fig. 3C and D). Interestingly, polarized Fourier Transform Infrared Spectra recorded at two different orientations (0°and 90°relative to the long axis of the fibrous aggregates) showed that the b-sheets within the fibrous aggregates of (AQ) 24 are randomly aligned, whereas the b-sheets within the fibrous aggregates of (AQ) 12 NR3 were predominantly aligned with the long axis of the fibers (in analogy to naturally spun fibers (Boulet-Audet et al., 2008; Lefevre et al., 2007a; Rousseau et al., 2009, 2004)) due to a controlled assembly process (Fig. 3E and F) (Hagn et al., in press). This finding was further confirmed by FTIR in the region between 1100 and 900 cm 1 where specific signals for b-polyalanine crystals (964 cm 1 ) can be observed, as reported recently (Papadopoulos et al., 2009). Interestingly, a clear difference at the two employed angles using the polarizer can be seen for (AQ) 12 NR3 aggregates (Fig. 3H), while there is no spectral difference for the (AQ) 24 aggregates, pointing out that the polyalanine crystals are oriented in the fibers formed vesicle 200 nm 100 nm 100 nm A BC buffer ctrl (AQ) 24 (AQ) 12 NR3 Fig. 4. Images obtained via transmission electron microscopy. (A) Dialysis buffer. (B) Freshly dialyzed solutions of (AQ) 24 . (C) Freshly dialyzed solutions of (AQ) 12 NR3, showing evidence of higher order supramolecular assemblies. L. Eisoldt et al. / Journal of Structural Biology 170 (2010) 413–419 417 from the protein with the NR domain, but not in the fibers formed from the protein without the NR domain. These findings suggest that the C-terminal NR domain controls both the storage of the proteins (by increasing their solubility and/or deterring undesirable aggregation events) and fiber formation (by facilitating controlled protein assembly into fibers upon demand). 3.4. The role of the N-terminal NR domain in fiber storage and assembly The N-terminal NR domains of silk proteins have been shown to be highly conserved between different spider species (Hedhammar et al., 2008; Lin et al., 2009; Motriuk-Smith et al., 2005; Rising et al., 2006). However, due to a lack of any published sequence data of the N-terminal NR domains of MA silk proteins of Araneus diadematus, this domain was not included in our engineered protein constructs. Despite this fact, our in vitro salt-induced aggregation assays with proteins with higher molecular weights and with longer repetitive sequence elements ((AQ) 24 NR3 as opposed to (AQ) 12 NR3) showed them to be more prone to aggregation (Fig. 2), which leads us to speculate that the N-terminal NR domain might also have an effect on higher order supramolecular assembly of silk proteins and therefore play an important role in protein storage and fiber assembly in vivo, as the natural proteins are larger than our recombinantly produced analogues (Guerette et al., 1996; Huemmerich et al., 2004b). 4. Conclusions Our salt-induced aggregation assays clearly demonstrate that chloride (which is present in the lumen during silk protein storage) does not induce significant levels of protein aggregation, whereas phosphate (present in the spinning duct) does induce significant levels of protein aggregation, due to the strongly kosmotropic nature of the phosphate anion. Our shear-induced aggregation assays clearly show that shear forces are likely to play a greater role than salt effects during the fiber spinning process in vivo, and moreover, a key role in the alignment of the secondary structural elements within the fibers (which controls their mechanical properties). Our transmission electron microscope studies show that the C-terminal NR domain is a prerequisite for higher order supramolecular assembly of the proteins, which is important for both protein storage and the controlled assembly of fibers. Finally, we speculate that the N-terminal NR domains also play a role in protein storage (and potentially fiber formation). Acknowledgments We gratefully acknowledge the help of Stefan Geimer (Universität Bayreuth) for assistance with the transmission electron microscopy studies. We also acknowledge Franz Hagn and Horst Kessler (Technische Universität München) for supplying us with primers for the D93A mutants. L.E. acknowledges financial support from the German Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung, BMBF 13N9736). J.G.H. acknowledges financial support from the Alexander von Humboldt Foundation. M.H acknowledges financial support from the ‘Graduiertenförderung nach dem bayerischen Eliteförderungsgesetz‘ (Universitäten Bayern, e.V.). T.R.S. acknowledges financial support from the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG SCHE 603/4-3) and the German Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung, BMBF 13N9736). This manuscript is in memoriam of Rainer Rudolph. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.jsb.2009.12.027. References Aprhisiart, A., Vollrath, F., 1994. Design-features of the orb web of the spider, Araneus diadematus. Behavioral Ecology 5, 280–287. Baldwin, R.L., 1996. How Hofmeister ion interactions affect protein stability. Biophysical Journal 71, 2056–2063. Bekard, I.B., Dunstan, D.E., 2009. Shear-induced deformation of bovine insulin in couette flow. Journal of Physical Chemistry B 113, 8453–8457. Bini, E., Knight, D.P., Kaplan, D.L., 2004. Mapping domain structures in silks from insects and spiders related to protein assembly. Journal of Molecular Biology 335, 27–40. Boulet-Audet, M., Lefevre, T., Buffeteau, T., Pezolet, M., 2008. Attenuated total reflection infrared spectroscopy: an efficient technique to quantitatively determine the orientation and conformation of proteins in single silk fibers. Applied Spectroscopy 62, 956–962. Casem, M.L., Tran, L.P.P., Moore, A.M.F., 2002. Ultrastructure of the major ampullate gland of the black widow spider, Latrodectus hesperus. Tissue & Cell 34, 427– 436. Cromwell, M.E.M., Hilario, E., Jacobson, F., 2006. Protein aggregation and bioprocessing. Aaps Journal 8, E572–E579. Exler, J.H., Hummerich, D., Scheibel, T., 2007. The amphiphilic properties of spider silks are important for spinning. Angewandte Chemie-International Edition 46, 3559–3562. Foo, C.W.P., Bini, E., Hensman, J., Knight, D.P., Lewis, R.V., Kaplan, D.L., 2006. Role of pH and charge on silk protein assembly in insects and spiders. Applied Physics a-Materials Science & Processing 82, 223–233. Geisler, M., Pirzer, T., Ackerschott, C., Lud, S., Garrido, J., Scheibel, T., Hugel, T., 2008. Hydrophobic and Hofmeister effects on the adhesion of spider silk proteins onto solid substrates: an AFM-based single-molecule study. Langmuir 24, 1350–1355. Gerstein, M., Chothia, C., 1996. Packing at the protein–water interface. Proceedings of the National Academy of Sciences of the United States of America 93, 10167– 10172. Gosline, J.M., Guerette, P.A., Ortlepp, C.S., Savage, K.N., 1999. The mechanical design of spider silks: from fibroin sequence to mechanical function. Journal of Experimental Biology 202, 3295–3303. Guerette, P.A., Ginzinger, D.G., Weber, B.H.F., Gosline, J.M., 1996. Silk properties determined by gland-specific expression of a spider fibroin gene family. Science 272, 112–115. Gurau, M.C., Lim, S.M., Castellana, E.T., Albertorio, F., Kataoka, S., Cremer, P.S., 2004. On the mechanism of the Hofmeister effect. Journal of the American Chemical Society 126, 10522–10523. Hagn, F., Eisoldt, L., Hardy, J.G., Vendrely, C., Murray, C., Scheibel, T., Kessler, H., in press. A highly conserved spider silk domain acts as a molecular switch that controls fibre assembly. Nature. Hamilton-Brown, P., Bekard, I., Ducker, W.A., Dunstan, D.E., 2008. How does shear affect a beta fibrillogenesis? Journal of Physical Chemistry B 112, 16249–16252. Hardy, J.G., Scheibel, T.R., 2009a. Production and processing of spider silk proteins. Journal of Polymer Science Part A-Polymer Chemistry 47, 3957–3963. Hardy, J.G., Scheibel, T.R., 2009b. Silk-inspired polymers and proteins. Biochemical Society Transactions 37, 677–681. Hardy, J.G., Romer, L.M., Scheibel, T.R., 2008. Polymeric materials based on silk proteins. Polymer 49, 4309–4327. Hedhammar, M., Rising, A., Grip, S., Martinez, A.S., Nordling, K., Casals, C., Stark, M., Johansson, J., 2008. Structural properties of recombinant nonrepetitive and repetitive parts of major ampullate spidroin 1 from Euprosthenops australis: implications for fiber formation. Biochemistry 47, 3407–3417. Heim, M., Keerl, D., Scheibel, T., 2009. Spider silk: from soluble protein to extraordinary fiber. Angewandte Chemie-International Edition 48, 3584–3596. Hijirida, D.H., Do, K.G., Michal, C., Wong, S., Zax, D., Jelinski, L.W., 1996. C-13 NMR of Nephila clavipes major ampullate silk gland. Biophysical Journal 71, 3442–3447. Hill, E.K., Krebs, B., Goodall, D.G., Howlett, G.J., Dunstan, D.E., 2006. Shear flow induces amyloid fibril formation. Biomacromolecules 7, 10–13. Horinek, D., Serr, A., Geisler, M., Pirzer, T., Slotta, U., Lud, S.Q., Garrido, J.A., Scheibel, T., Hugel, T., Netz, R.R., 2008. Peptide adsorption on a hydrophobic surface results from an interplay of solvation, surface, and intrapeptide forces. Proceedings of the National Academy of Sciences of the United States of America 105, 2842–2847. Huang, W., Lin, Z., Sin, Y.M., Li, D., Gong, Z., Yang, D., 2006. Characterization and expression of a cDNA encoding a tubuliform silk protein of the golden web spider Nephila antipodiana. Biochimie 88, 849–858. Huemmerich, D., Scheibel, T., Vollrath, F., Cohen, S., Gat, U., Ittah, S., 2004a. Novel assembly properties of recombinant spider dragline silk proteins. Current Biology 14, 2070–2074. Huemmerich, D., Helsen, C.W., Quedzuweit, S., Oschmann, J., Rudolph, R., Scheibel, T., 2004b. Primary structure elements of spider dragline silks and their contribution to protein solubility. Biochemistry 43, 13604–13612. Ittah, S., Michaeli, A., Goldblum, A., Gat, U., 2007. A model for the structure of the Cterminal domain of dragline spider silk and the role of its conserved cysteine. Biomacromolecules 8, 2768–2773. 418 L. Eisoldt et al. / Journal of Structural Biology 170 (2010) 413–419 4  as recently described20. To investigate the possible interactions of the two recombinant spider silk proteins eADF3((AQ)12NR3) and eADF4(C16NR4) the encoding genes were co-expressed in E. coli from a plasmid where each gene is under the control of its own T7-promoter (Figure 1c).To distinguish between both similar sized proteins in various assays eADF3((AQ)12NR3) was fused with an aminoterminal T7-tag and eADF4(C16NR4) with a hexahistidine (His6)-tag. When eADF3((AQ)12NR3) and eADF4(C16NR4) are produced individually they form disulfide linked dimers even in the reducing cytoplasm of E.coli, since the redoxpotential of the involved cysteins is significantly lower than that of the cytosol of E. coli2. When produced individually the majority (>66%) can be found in the soluble fractions as judged by western blot analysis (Figure 2a). Coexpression has no influence on the solubility of the proteins under the conditions tested (Figure 2a). Strikingly a carboxyterminal domain mediated heterodimerisation of both proteins could be detected (Figure 2b). Next, we tested the formation of such heterodimer in vitro. Therefore, a strong reducing agent (TCEP) was added to a 1:1 mixture of fully chemically denatured (AQ)12NR3 and C16NR4, and the proteins were refold under reducing conditions. After refolding, both homodimers as well as the heterodimer could be detected (Fig 2c). However, the amount of heterodimer formed in vitro was much lower in vivo with a nearly statistical distribution of 1:1:1 (Figure 2b). To characterize the heterodimer, a purification strategy was established (see Materials and Method section) yielding pure heterodimer (Figure 2c). Far-UV CD spectra of the three proteins showed no substantial differences (Figure 2d). The broad minimum at 205 nm and a plateau at 219 nm indicate a mainly random-coil/PPI-II dominated structure with α-helical portions. The random coil signals arise from the repetitive core domain20, whereas the α-helical contributions are solely from the carboxyterminal regions NR3 and NR42, 20. Melting experiments showed a melting 5  point of 66°C for the heterodimer which is 2°C above the one of the (AQ)12NR3 homodimer and 1.5 °C below that of the C16NR4 dimer (Figure 2e). Upon cooling, after thermal denaturation the heterodimer refolds quantitatively like the homodimers (Figure 2d)20. These experiments indicate that the NR3-NR4-heterodimer has a similar structural integrity as the respective homodimers. As shown previously, eADF3 and eADF4 proteins have different assembly properties when exposed phosphate ions20, 21. Here, we tested the assembly propensity of the three dimeric species in the presence of increasing phosphate ion concentrations. (AQ)12NR3 as well as the heterodimer aggregated at phosphate concentrations above 50 mM, whereas C16NR4 startet to aggregate above 150 mM phosphate (Figure 3a). With increasing phosphate concentration the aggregation kinetics of C16NR4 were accelerated (Figure 3b). Interestingly, the kinetics of the heterodimer were only slightly slower than that of C16NR4. Next, binding of the β-sheet sensitive dye ThioflavinT to the different assemblies were tested (Figure 3c). (AQ)12NR3 showed significantly (5-10 times) less ThT binding than the other two dimeric species. To characterize the self-assemblies of the heterodimer in comparison to that of the homodimers they were investigated by atomic force microscopy (AFM). AFM images of the C16NR4 samples showed fibrils that are mainly unbranched with an average height of 3-3.5 nm similar to previously reported C16_∆CTD fibrils 22 (Figure 3d). In case of (AQ)12NR3 only amorphous morphologies rather than distinct fibrils could be detected (Figure 3e), with a spherical micelle-like assembly form as recently reported (Suppl. Figure 1)16. Interestingly the heterodimer also formed nano fibrils which are clearly distinctive in form and shape from the C16NR4 fibrils (Figure 3C and F), with a network of short, sometimes branched fibrils with an average height of 1 nm. The 6  distances between the branching points is sometimes only a few nanometers in length. In summary, we present evidence that the two MaSp2-like silk proteins ADF3 and 4 from A. diadematus can form covalently linked heterodimers through their native nonrepetitive carboxyterminal domains (CTD), with different assembly properties than the individual homodimers (Figure 3d-h), since the heterodimer is able to form fibrillar networks (Figure 3f/i). The findings have impact on understanding the natural silk fiber formation. Both proteins are secreted in the same compartment of the spinning gland 23, 24. Recent findings from genomic analysis suggest that both MA proteins are produced within the same rather than in separated cells which leads to the assumption that both proteins can already interact shortly after translation within the cells (e.g. in the secretory pathway compartments) 25, 26 Further, our results highlight the possibility to conveniently combine different recombinant silk proteins via the highly similar CTDs to obtain new silk materials with adjustable properties. Acknowledgments We thank Dr. Martin Humenik and Dr. Andrew Smith for their valuable input regarding the fiber assembly experiments. We also thank Kristin Schacht for helping to obtain the TEM images. This work was funded by the Bundesministerium für Bildung und Forschung (BMBF) and the Deutsche Forschungsgemeinschaft (DFG). Conflict of interests The authors declare no conflict of interest. 7  References 1. Vollrath, F., Strength and structure of spiders' silks. J Biotechnol 2000, 74, (2), 67-83. 2. Hagn, F.; Eisoldt, L.; Hardy, J. G.; Vendrely, C.; Coles, M.; Scheibel, T.; Kessler, H., A highly conserved spider silk domain acts as a molecular switch that controls fibre assembly. Nature 2010, 465, (7295), 239-42. 3. Hedhammar, M.; Rising, A.; Grip, S.; Martinez, A. S.; Nordling, K.; Casals, C.; Stark, M.; Johansson, J., Structural properties of recombinant nonrepetitive and repetitive parts of major ampullate spidroin 1 from Euprosthenops australis: Implications for fiber formation. Biochemistry 2008, 47, (11), 3407-3417. 4. Hinman, M. B.; Lewis, R. V., Isolation of a clone encoding a second dragline silk fibroin. Nephila clavipes dragline silk is a two-protein fiber. J Biol Chem 1992, 267, (27), 19320-4. 5. Guerette, P. A.; Ginzinger, D. G.; Weber, B. H. F.; Gosline, J. M., Silk properties determined by gland-specific expression of a spider fibroin gene family. Science 1996, 272, (5258), 112-115. 6. Xu, M.; Lewis, R. V., Structure of a Protein Superfiber - Spider Dragline Silk. Proc. Natl. Acad. Sci. U. S. A. 1990, 87, (18), 7120-7124. 7. Rising, A.; Johansson, J.; Larson, G.; Bongcam-Rudloff, E.; Engstroem, W.; Hjalmt, G., Major ampullate spidroins from Euprosthenops australis: multiplicity at protein, mRNA and gene levels. Insect Mol Biol 2007, 16, (5), 551-561. 8. Hijirida, D. H.; Do, K. G.; Michal, C.; Wong, S.; Zax, D.; Jelinski, L. W., 13C NMR of Nephila clavipes major ampullate silk gland. Biophys. J. 1996, 71, (6), 3442-7. 9. Simmons, A.; Ray, E.; Jelinski, L. W., Solid-State 13C NMR of Nephila clavipes Dragline Silk Establishes Structure and Identity of Crystalline Regions. Macromolecules 1994, 27, (18), 5235-5237. 10. Parkhe, A. D.; Seeley, S. K.; Gardner, K.; Thompson, L.; Lewis, R. V., Structural studies of spider silk proteins in the fiber. J Mol Recognit 1997, 10, (1), 1-6. 11. van Beek, J. D.; Hess, S.; Vollrath, F.; Meier, B. H., The molecular structure of spider dragline silk: folding and orientation of the protein backbone. Proc. Natl. Acad. Sci. U. S. A. 2002, 99, (16), 10266-71. 12. Lefevre, T.; Leclerc, J.; Rioux-Dube, J. F.; Buffeteau, T.; Paquin, M. C.; Rousseau, M. E.; Cloutier, I.; Auger, M.; Gagne, S. M.; Boudreault, S.; Cloutier, C.; Pezolet, M., Conformation of spider silk proteins in situ in the intact major ampullate gland and in solution. Biomacromolecules 2007, 8, (8), 2342-2344. 13. Simmons, A. H.; Michal, C. A.; Jelinski, L. W., Molecular orientation and twocomponent nature of the crystalline fraction of spider dragline silk. Science 1996, 271, (5245), 84-7. 14. Sponner, A.; Unger, E.; Grosse, F.; Klaus, W., Differential polymerization of the two main protein components of dragline silk during fibre spinning. Nat. Mater. 2005, 4, (10), 772775. 15. Li, S. F.; McGhie, A. J.; Tang, S. L., New internal structure of spider dragline silk revealed by atomic force microscopy. Biophys. J. 1994, 66, (4), 1209-1212. 16. Eisoldt, L.; Hardy, J. G.; Heim, M.; Scheibel, T. R., The role of salt and shear on the storage and assembly of spider silk proteins. J. Struct. Biol. 2010, 170, (2), 413 - 419. 17. Stark, M.; Grip, S.; Rising, A.; Hedhammar, M.; Engstrom, W.; Hjalm, G.; Johansson, J., Macroscopic fibers self-assembled from recombinant miniature spider silk proteins. Biomacromolecules 2007, 8, (5), 1695-1701. 18. Sponner, A.; Vater, W.; Rommerskirch, W.; Vollrath, F.; Unger, E.; Grosse, F.; Weisshart, K., The conserved C-termini contribute to the properties of spider silk fibroins. Biochemical and Biophysical Research Communications 2005, 338, (2), 897-902. 19. Exler, J. H.; Hummerich, D.; Scheibel, T., The amphiphilic properties of spider silks are important for spinning. Angew. Chem.-Int. Edit. 2007, 46, (19), 3559-3562. 8  20. Huemmerich, D.; Helsen, C. W.; Quedzuweit, S.; Oschmann, J.; Rudolph, R.; Scheibel, T., Primary structure elements of spider dragline silks and their contribution to protein solubility. Biochemistry 2004, 43, (42), 13604-13612. 21. Rammensee, S.; Slotta, U.; Scheibel, T.; Bausch, A. R., Assembly mechanism of recombinant spider silk proteins. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, (18), 6590-6595. 22. Slotta, U. K.; Rammensee, S.; Gorb, S.; Scheibel, T., An engineered spider silk protein forms microspheres. Angew. Chem.-Int. Edit. 2008, 47, (24), 4592-4594. 23. Vollrath, F.; Knight, D. P., Liquid crystalline spinning of spider silk. Nature 2001, 410, (6828), 541-8. 24. Vollrath, F.; Knight, D. P., Structure and function of the silk production pathway in the spider Nephila edulis. Int. J. Biol. Macromol. 1999, 24, (2-3), 243-9. 25. Ayoub, N. A.; Garb, J. E.; Tinghitella, R. M.; Collin, M. A.; Hayashi, C. Y., Blueprint for a High-Performance Biomaterial: Full-Length Spider Dragline Silk Genes. Plos One 2007, 2, (6), -. 26. Gaines, W. A. t.; Marcotte, W. R., Jr., Identification and characterization of multiple Spidroin 1 genes encoding major ampullate silk proteins in Nephila clavipes. Insect Mol Biol 2008, 17, (5), 465-74. 27. Heukeshoven, J.; Dernick, R., Improved silver staining procedure for fast staining in PhastSystem Development Unit. I. Staining of sodium dodecyl sulfate gels. ELECTROPHORESIS 1988, 9, (1), 28-32. 28. Korz, D. J.; Rinas, U.; Hellmuth, K.; Sanders, E. A.; Deckwer, W. D., Simple fed-batch technique for high cell density cultivation of Escherichia coli. J Biotechnol 1995, 39, (1), 5965. 29. Eisoldt, L.; Thamm, C.; Scheibel, T., The role of terminal domains during storage and assembly of spider silk proteins. Biopolymers 2012, 97, (6), 355-361. 9  Figures and Legends Figure 1: Co-Production of recombinant variants of ADF3 and ADF4. (A) Simplified scheme of the dragline silk from A. diadematus. The two MaSp2-like proteins ADF3 and ADF4 are colacilzed within the fiber. (B) Structure of the parallel dimer of the nonrepetitive carboxyterminal domain NR3. Each monomer, one depicted in blue and one in cyan, comprises a five helix bundle each with one conserved cysteine in helix 4 able to form a disulfide bond covalently linking the two monomers. The aminoterminal regions (NH 3 and NH 3* ) are highlighted where the CTD is connected to the repetitive domain. (C) Illustration of the plasmid used for the coexpression of the genes encoding eADF3((AQ) 12 NR3) and eADF4(C 16 NR4). Each silk gene is under the control of its own T7 promoter (T7-1 and T7-2). C 16 NR4 is produced with an aminoterminal hexahistidine tag, whereas (AQ) 12 NR3 has T7-tag. 10  Figure 2: Analysis of heterodimer formation. (a) Comparison of the soluble and insoluble fraction after production and cell disruption of individually produced (AQ) 12 NR3 and C 16 NR4 and the respective co-expressed (dual produced) samples. P = pellet (insoluble) fraction, SN = supernatant (soluble) fraction. (b) Westernblot analysis of (AQ) 12 NR3 and C 16 NR4 heterodimer formation. Lysate: E. coli lysate containing both eADF variants. heterodimer: purified (AQ) 12 NR3-C 16 NR4 heterodimer. (AQ) 12 NR3 and C 16 NR4 : purified homodimers, respectively. M: monomeric (AQ) 12 NR3. (c) SDS-PAGE analysis (silver stained) of (AQ) 12 NR3/C 16 NR4 heterodimers formed in vitro and in vivo. In vitro formation was achieved by refolding a mixture of both proteins under reducing conditions. The black arrows depict the heterodimer. The last lane shows the purified heterodimer used for all assays in this study. (d) Far UV spectra of homodimers of eADF3((AQ) 12 NR3) (black) and eADF4(C 16 NR4) (blue) as well as the heterodimer before and after heating to 90°C (red). (e) Thermal melting points. Protein concentrations in all experiments were 3.4 µM. 11  Figure 3: Assembly properties of the three dimeric species. (a) Silk protein assembly in dependence of phosphate concentration after 1 hour of incubation at RT. Data points were fitted with a boltzman function for better visualization. (b) Assembly kinetics of the three dimer species in presence of 100 mM sodium phosphate (pH 8.0) at 20 °C. Selfassembly was determined by measuring turbidity at 340 nm. (c) Binding of ThioflavinT in presence of 100 mM potassium phosphate after 24 h of incubation clearly indicated the presence of β-sheet rich structures in samples made of the heterodimer or C 16 NR4. Morphologies of the assemblies were investigated by AFM. Prior to analysis the three dimeric species were incubated for 24h in Tris-buffer 12  pH 8.0 with 50 mM of potassium phosphate at room temperature. (d) C16NR4, (e) (AQ)12NR3, (f) heterodimer. AFM pictures were made in tapping mode. (g) Model of the different in presence of phosphate ions. eADF3 preferentially forms large, globular shaped assemblies. eADF4 in contrast forms nanosized unbranched fibrils that are β-sheet rich. The heterodimer forms a highly interconnected network of nanofilaments that also have a high β-sheets content. 13  Materials & Methods Cloning. Engineered silk genes based on Aranaeus diadematus fibroin (ADF) 3 and 4 as described previously20 were cloned into a pRSFduet-1mut plasmid (a modified pRSFduet-1 plasmid (Novagen – Merk, Germany) with one additional base pair prior to the NdeI restriction site in MCS2) containing two multiple cloning sites (MCS) each with a preceding T7-promoter. Sequences encoding for eADF4(C16NR4) were cut out from pCS_C16NR420 using BamHI/HindIII. The gene encoding for eADF3((AQ)12NR3 with an aminoterminal T7 tag was obtained by digesting pET29_(AQ)12NR3 with NdeI/XhoI. Both genes were cloned stepwise into pRSFduet1mut digested with the respective restriction enzymes. The final construct encodes for eADF3((AQ)12NR3) bearing an aminoterminal T7 tag and eADF4(C16NR4) with an aminoterminal hexa-histidine tag. To isolate the desired genes after restriction digest the samples were separated on a 0.8% agarose gel. The desired bands were sliced out and purified using a kit (Promega, Germany). T4-Ligase and restriction enzymes were purchased from New England Biolabs (USA). Expression analysis with SDS-PAGE and Western Blot. Freshly prepared HMS174 (DE3) E. coli cells were transformed with the plasmid containing the two recombinant silk genes and plated on agar plates with kanamycin. 4 ml precultures from single colonies were grown overnight at 30-37°C overnight. Shaking flasks with 100 ml LB media were inoculated 1:500 (v/v) with the precultures and grown at 3037°C. The cultures were allowed to grow until they reached an optical density at 600 nm (OD600) of 0.6 – 0.8 until protein production was started with the addition of 0.1 mM IPTG. During expression the cells were incubated at 30-37°C. After 4 h of expression the cells were harvested by centrifugation at 6000xg and 6°C in a Sorvall RC6 Plus centrifuge (Thermo Scientific, USA). For cell disruption 1 g of cell pellet 20  Supplemental Information Supplemental Figure 1: Transmission electron microscopy images of the three dimeric species. (A) C16NR4 forming fibrils with length of several hundred nanometers. (B) (AQ)12NR3 was observed forming mainly spherical, micelle-like assemblies (B-1) and sporadic as fibrillar structures with a low electron density (B-2). (C) The heterodimer forms a network of thin and short (<10 nm), sometimes stick-like fibrils. Prior to analysis the three dimeric species were incubated for 18-24h in Trisbuffer pH 8.0 with 50 mM of potassium phosphate at room temperature. Droplets containing the samples were placed on pioloform-coated copper grids and the excess liquid was removed using paper towels (Kimberly-ClarK Corp, USA). The grid was stained with uranyl acetate (1% in water) and the excess liquid was again removed with papertowels. The samples were viewed with a JEOL JEM-2100 transmission electron microscope (Jeol, Germany). Teilarbeiten [103] Teilarbeit IV Eisoldt, L., Smith, A., Scheibel, T. (2011) Decoding the secrets of spider silk. Materials Today. 40, pp. 80-86 ISSN:1369 7021 © Elsevier Ltd 2011 MARCH 2011 | VOLUME 14 | NUMBER 3 80 Decoding the secrets of spider silk Silk has been produced by a large variety of arthropods for hundreds of millions of years. Humans have used silk materials for several thousands of years for all kinds of applications, ranging from textiles to wound dressings to uses in the military. The main source of silk is the mulberry silkworm Bombyx mori, since it is easy to farm (farming started about 5000 years ago in China). The silk of spiders cannot be collected as easily, but the nets of orb weaving spiders (Araneidae) have also been historically employed for applications such as fishing or wound coverage, due to their outstanding mechanical and biomedical properties. Orb webs can be found all over the world, with web diameters ranging from a few centimeters to several meters. Very recently spiders in Madagascar have been discovered to spin single threads which span across rivers, with a length of about 25 m1. Orb webs have been adapted over millions of years to withstand and compensate for the kinetic impacts of the large (in comparison to the fiber diameter) prey of the spider. When examined by the naked eye all threads in an orb web appear to be the same, but in fact a web is built from up to five different silk types. A female orb weaving spider is even able to produce up to seven different silks, including one special silk for egg cases2 (Fig. 1). Spider silks are tailored for specific purposes and exhibit a great variation in mechanical properties. Between the different silk types their strength (the stress needed to break the fiber) ranges from 0.02 to a remarkable 1.7 GPa, which exceeds steel (1.5 GPa), while the extensibility varies between 10 and 500 %3,4. Interestingly, most spider silks have a sophisticated combination of strength and extensibility, which yields a very high toughness (the amount of energy absorbed per volume before breakage) that exceeds most natural or man-made fibers. Also common for all silks is their viscoelastic behavior, since upon stretching energy is dissipated in the form of heat, diminishing any elastic recoil5. The most investigated spider silk is dragline silk, the lifeline a spider always drags behind, which shows the most remarkable combination of strength and extensibility. It is produced in the major ampullate gland (individual glands exist for each silk type) of a spider, and therefore its Spider silks have been employed by man for several thousands of years. Spider silks possess extraordinary mechanical properties due to a combination of strength and extensibility that are superior to most man-made fibers. Spider silk fibers are a protein-based material produced in a highly sophisticated hierarchical process under mild conditions. Here, we review the current understanding of spider silk and its assembly process, as well as discuss the application of silk-based materials to the fields of biomedicine and materials engineering. Lukas Eisoldt, Andrew Smith, Thomas Scheibel* Universität Bayreuth, Fakultät für Angewandte Naturwissenschaften, Lehrstuhl für Biomaterialien, Universitätsstraße 30, 95447 Bayreuth, Germany *E-mail: [email protected] MT14_3p80-87.indd 80 2/24/2011 11:25:12 AM Decoding the secrets of spider silk REVIEW MARCH 2011 | VOLUME 14 | NUMBER 3 81 correct biological name is major ampullate silk (MAS). MAS is used for the frame and radii of an orb web, as well as the aforementioned lifeline by most spiders. The mechanical properties of dragline silk of several different species have been extensively analyzed4,6-8. In comparison to man-made fibers, the maximum strength of up to 1.7 GPa (dragline of Caerostris darwini) is in the range of high-tech materials. Although, Kevlar or carbon fibers have a higher strength and stiffness, spider silk fibers have a much higher toughness due to their greater extensibility9. However, dragline silk offers even more interesting features. For example, a hanging spider hardly ever twists, indicating an interesting torsional dampening behavior of the dragline thread. When a dragline thread is twisted it does not oscillate around the new position, like a Kevlar fiber would. After a while, the fiber slowly returns to its initial position, indicating that there is a sort of a shape memory within the fiber10. Another interesting feature of spider silk is its ability to undergo supercontraction. When dragline silk is wetted, or when the relative humidity is above 60 %, a silk thread swells in diameter and shrinks in length by about 50 %11,12. Other spider silk types show this behavior as well, but in these cases this behavior is much less pronounced13. Besides dragline silk the other silk types also show interesting properties like an extraordinary extensibility of up to 500 % (flagelliform silk) or a glue-like behavior (aggregate silk), but in this review we will continue to focus on dragline silks. Primary structure of spider silks – less (complexity) is more All spider silk threads are mainly composed of one or more proteins, called spidroins, which tend to be large (up to 350 kDa per monomer)14. Despite their different properties, all spidroins share a common primary structure pattern comprising a large central core of repeated modular units, accounting for approximately 90 % of the amino acids of the protein, flanked by non-repetitive domains (Fig. 2a). These non-repetitive terminal domains have a folded globular structure and are comprised of approximately 100 – 140 amino acids. Their sequences are highly conserved throughout different spider species and silk types15-17. The terminal domains are essential for storage of spidroins in the silk glands and fiber formation in the spinning duct, since it is here that they trigger crucial steps in the complex assembly of the spidroins upon environmental changes, like pH or ionic composition and strength18,19. The sequence of the repetitive core is tailored for the individual mechanical functions of the different silk types. In general, it consists of modular units each with about 40 – 200 amino acids7,14,20,21, and the modular units are repeated up to approximately 100 times within the core domain14. In major ampullate, minor ampullate and flagelliform silk, the modular units are mainly comprised of a subset of the sequence motifs (Ala)4-14, (GlyAla)n, GlyGlyX and GlyProGlyXX, Fig. 1 Schematic overview of different silk types produced by female orb weaving spiders (Araneae). Each si lk type (highlighted in red) is tailored for a specific purpose. MT14_3p80-87.indd 81 2/24/2011 11:25:14 AM REVIEW Decoding the secrets of spider silk MARCH 2011 | VOLUME 14 | NUMBER 3 82 with X representing a variable amino acid (Fig. 2b). The subset, the sequential order and the number of these motifs in each module are important for the mechanical properties of the final fiber. In contrast to terminal domains, the core region is intrinsically unfolded, as long as the spidroins are stored in the gland22-25. From sequence to structure The dragline fiber is not only the best characterized spider silk in terms of mechanical properties, but also concerning its structurefunction relationship. Unlike other spider silks, dragline fibers can be collected from orb webs or by “milking” (forced silking) of spiders. Electron, atomic force and light microscopy have revealed a coreshell structure for the dragline fiber26,27. The shell is quite thin and contains different biomolecules like lipids, glycoproteins and other silk proteins28. The core consists mainly of two proteins produced in the major ampullate gland named MaSp (Major ampullate Spidroin) 1 and 2. A single modular unit of MaSp1 usually comprises a polyalanine block and several GGX motifs. In modules of MaSp2 the GGX motif is replaced by the motif GPGXX, increasing the proline content of MaSp2 dramatically. Several dozens of repeats of these modular units build the complete core region of these spidroins. The potential structures of the individual motifs and their contribution to the properties of the final fiber have been intensively investigated over the last two decades (Fig. 2b). Using different NMR techniques and x-ray diffraction it has been shown that the polyalanine motif forms defined nanosized crystals (2 x 5 x 7 nm) based on tightly packed anti-parallel β-sheets22,29-34, in which it is very likely that polyalanine stretches from different silk molecules constituting very strong non-covalent cross-links. The structure of the GGX motif in dragline silk is less understood. Recent NMR studies provided evidence that these motifs might form β-sheets as well as less ordered helical structures33,35-37. Meanwhile, the structure of the GPGXX motif from MaSp2 seems to be clear. Due to the proline residue, this pentapeptide likely forms β-turns which, when repeated, yield a spiral structure as suggested for elastin38. It should be noted that in the case of Araneus diadematus both MAS components, named ADF 3 and 4 (Araneus diadematus fibroin) due to historical biological reasons, have similar proline contents (~ 13 %)21. Compared to other MAS fibers Araneus silk shows no significant differences concerning its mechanical behavior indicating that the presence of low-proline-content spidroins is not necessary for the mechanical performance. It is worth noting that a macroscopic effect of the proline content is only seen when dragline silk is wetted. The amount of proline influences the degree of supercontraction and the mechanical properties of wetted silk39-41. However, it seems to be more important to have a pair of hydrophobic/hydrophilic spidroins, since comparing the hydrophobicity of MA spidroins between different species revealed that MaSp1/ADF4 are more hydrophobic and MaSp2/ADF3 are more hydrophilic. Fig. 2 The hierarchical setup of spider silk proteins. (a) Illustration of a spidroin comprised of up to 100 repetitive modules (blue) and two terminal domains (red and green). The components are not drawn to scale. (b) The repetitive modules of the most prominent silk types comprise a distinct subset of amino acid motifs. (c) Each motif is thought to fulfill a distinct structural role and contributes to the mechanical properties of the final fiber. (b) (a) (c) MT14_3p80-87.indd 82 2/24/2011 11:25:16 AM Decoding the secrets of spider silk REVIEW MARCH 2011 | VOLUME 14 | NUMBER 3 83 The natural spinning process The primary structure of silk proteins alone does not explain the outstanding properties of a dragline silk fiber. Fibers technically spun from reconstituted spidroins (i.e. spidroins obtained from chemically dissolved spider silk fibers) or from spidroins from the original spinning dope show completely different mechanical properties compared to fibers spun by spiders, indicating that the spinning process is also crucial42,43. The spiders spinning apparatus can be divided into four parts (Fig. 3a). In the first part, the tail, the spidroins are secreted from specialized cells. In the second part, the sac or ampulla, the spidroins are stored, and further compounds are added later forming the skin of the fiber. The protein concentration during storage in the ampulla can reach up to 50 % (w/v)22,44. This is a remarkably high concentration, since most proteins tend to aggregate irreversibly under such conditions. To prevent unspecific aggregation, spidroins form micellar-like structures controlled by their amphiphilicity19,45-48 and by the non-repetitive termini which induce or support the formation of such supramolecular assemblies18,19,48 (Fig. 3b). The stored spidroins in the ampulla, constitute the so-called spinning dope which has characteristics of a nematic liquid crystalline phase47,49. In the third part of the spinning apparatus, an S-shaped duct, the liquid crystalline fluid is exposed to a constant elongational flow. The liquid crystallinity allows the molecules to flow in a pre-aligned manner and to further align along the flow axis during the passage through the duct. Due to its tapering, the shear forces increase along the duct leading to the formation of β-sheet crystals19,40-50. In the last part of the spinning apparatus the shear stress further increases, due to pulling of the fiber out of the spigot (Fig. 3b). The pulling forces lead to some structural changes in the spidroins (especially their terminal regions) accompanied by exposure of hydrophobic areas followed by a phase separation between the solvent (i.e. water which is actively removed by epithelial cells51-53) and the spidroins54,55 (Fig. 3b). The structural changes in the terminal domains re-arrange the position of the core regions within the micellar-like structures, and this together with mechanical forces supports phase separation and spidroin assembly. As well as mechanical influences, chemical changes trigger a structural arrangement, especially in the S-shaped duct. There the exchange of sodium and chloride ions for the relatively kosmotropic potassium and phosphate ions is responsible for the exposure of hydrophobic areas within the C-terminal non-repetitive domain19,56. Upon lowering the Fig. 3 The natural spinning process. (a) Illustration of a spider’s spinning gland divided into four parts. (b) Schematic model of the silk fiber assembly mechanism occurring along the spinning apparatus. (b) (a) MT14_3p80-87.indd 83 2/24/2011 11:25:16 AM REVIEW Decoding the secrets of spider silk MARCH 2011 | VOLUME 14 | NUMBER 3 84 pH value from roughly 7 to 6, the amino-terminal non-repetitive domains dimerize, likely acting as additional physical cross-links between the spidroins48,57,58. Before the fiber exits the spigot it passes the so-called “valve”, an organ that assists in restarting the spinning process after internal rupture52. It twists the duct back and forth, thereby moving the ruptured fiber end into the opened valve. Once the fiber has left the spider, the spider pulls and stretches the fiber leading to further water evaporation and additional alignment of the molecules inside the fiber. All the mechanisms of this highly sophisticated spinning process enable the spider to efficiently produce a very tough fiber under mild conditions; a process yet unrivaled by any man-made fiber spinning process. The structural organization of a dragline fiber The structure of dragline silk fibers reveals three hierarchical levels59. On the macroscopic level the fiber shows a core-shell structure (Fig. 4a). On the mesoscopic level, fibrils oriented along the fiber axis form the core of the fiber, resembling the structure of a rope. On the molecular level small crystallites (2 x 5 x 7 nm) comprising tightly packed alanine β-sheets and larger crystalline regions (> 100 nm) based on glycine-alanine sequences with high β-sheet content can be found. Both crystalline regions are interconnected by a so-called amorphous matrix which contains loosely arranged helical structures oriented along the fiber axis (Fig. 4b). In this two-phase-model the crystalline regions mediate the strength of the fiber, whereas the matrix is responsible for the elasticity of the fiber. However, results obtained from various x-ray and NMR methods indicate that parts of the amorphous regions are much more ordered and oriented, leading to a model with a third phase which interconnects the crystalline regions with the amorphous phase60-62. For further detailed information we recommend consulting additional literature63-66. Biotechnological production of spider silk proteins In contrast to silkworms (Bombyx mori), most spiders cannot be farmed due to their cannibalistic behavior. Collecting silks from webs is also very time-consuming and therefore not profitable, especially if a specific silk type (other than dragline silk) is desired. In order to achieve spider silk on a large scale, biotechnological production of spidroins is the only feasible solution. For recombinant protein production, the DNA sequence from a donor organism (here the spider) has to be recombined with that from an acceptor (a so-called host) which is used for the production. Most recombinant silk proteins are produced in the bacterium Escherichia coli which is a well-established host for the industrial scale production of proteins. It allows a fast and scalable production within 3 – 4 days from the initiation of production to the finally purified protein. Attempts to produce recombinant MA silk proteins based on fragments of native silk genes from spiders in bacteria resulted in low yields, mainly due to the different codon usage of spiders and E. coli67. However, in gene engineering approaches the codon usage of the genes can be adapted to the host organism. In the case of silk genes, the repetitive character allows a straight forward approach. As depicted above, most spider silk proteins show a simple primary structure composed of repeated modular units which contain 100 – 200 amino acids based on smaller sequence motifs (e.g. polyalanine). The arrangement and the exact sequence of the motifs differ from silk type to silk type and from species to species, but within an individual spidroin they are quite regular. Thus, by analyzing a spidroin’s amino acid sequence, one can identify the modular units and design artificial gene encoding for silk proteins using host-specific codons. With this method it is possible to create recombinant silk proteins with a high sequence similarity (on a protein level) to the natural blueprints. Additionally, completely novel silk-proteins can be created by hybridizing motifs of different silk types. This approach has been used by several groups to produce various engineered silk proteins. Most of these engineered proteins are based on the sequences of MaSp1/268,69 from various species or ADF3/470, but some also engineered recombinant flagelliform spidroins71-74. The majority of the recombinantly produced silk proteins are smaller (30 – 110 kDa) than natural spidroins (300 – 350 kDa). This is due to the biological limitation of E. coli, which lowers the protein yields dramatically with increasing protein size75,76. However, this problem can be partly circumvented by modifying the production host77. Strikingly, all artificial spidroins, except one protein based on the sequence of ADF346, show much lower solubilities in aqueous solution than the natural spidroins (see68,69 and references therein). A lot of these proteins tend to self-assemble or aggregate into small fibrillar structures with sizes ranging from a few micrometers to centimeters78. Although these spontaneously formed fibrils show β-sheet structures, they are only remotely reminiscent of natural silk fibers, since in this selfassembly process some crucial structural alignment steps do not take place. Fig. 4 Illustration of the hierarchical spider silk structure. (a) The fiber shows a skin-core structure with fibrils forming the core. (b) On the nanostructural level the fibrils comprise small tightly packed β -sheet crystals (red arrow) and larger crystalline regions (black arrow) interconnected by an amorphous matrix . (b) (a) MT14_3p80-87.indd 84 2/24/2011 11:25:18 AM Decoding the secrets of spider silk REVIEW MARCH 2011 | VOLUME 14 | NUMBER 3 85 Exploiting the application potential of recombinant spider silks So far, no biomimetic spinning process exists that would allow the formation of native-like spider silk fibers from recombinant silk proteins. However, despite the nonexistent spinning process most recombinant spider silks can self-assemble into non-natural shapes such as spheres, capsules, films, non-wovens or hydrogels (Fig. 5), which have a high application potential79-83. Depending on the processing conditions the assembly forms have different material properties linked to the molecular structure of the employed silk protein. Morphologies with high β-sheet contents (spheres, hydrogels) that are formed by fast salting-out or slow selfassembly processes are water-insoluble55,81,84,85. In contrast, forced assembly (e.g. into films or non-wovens) out of fast evaporating organic solvents such as hexafluoroisopropanol yields α-helical rich compositions rendering the structures water soluble. Interestingly, such structures can be post-treated with substances that induce β-sheet formation leading to the gain of water-insolubility80,82,86. Not only water solubility, but the mechanical properties of the silk materials are also linked to the β-sheet content. Since recombinant spider silk proteins are monodisperse (in comparison to synthetic polymers), high degrees of assembly control and, therefore, the final material’s properties are possible. Concerning applications, the aforementioned silk shapes can be used as carriers for drugs or as scaffolds in tissue engineering80, since natural silks show a good biocompatibility (cytocompatibility, low immunogenicity, biodegradeability, low toxicity) in vivo and in vitro69,87-92. Recombinant spider silk proteins have a similar cytocompatibility to natural silks making them suitable for biomedical applications93-95. Besides its biocompatibility recombinant spider silk structures share some other interesting features with natural spider silk fibers, like its smooth surface, making them suitable for technical applications as well. These features can be used in technical applications, e.g. non-wovens as air filtering devices or spheres, as additives for cosmetics to give the product a “silken” smooth feeling. Future perspective Spider silks are impressive biopolymers that have evolved over millions of years. Over the last several decades a lot of progress has been made in unraveling some of its secrets. Nevertheless, certain questions, especially concerning solubility, storage and assembly of the underlying spider silk proteins, which are crucial for the technical production of e.g. silk fibers, are still unanswered. The biotechnological production of recombinant spider silk in larger scales is a landmark in spider silk research, since now investigations are enabled to answer such questions. Furthermore, recombinant spider silk proteins can be Fig. 5 Different non-natural assembly forms (bold text) of recombinant spider silk proteins and potential applications thereof. MT14_3p80-87.indd 85 2/24/2011 11:25:18 AM REVIEW Decoding the secrets of spider silk MARCH 2011 | VOLUME 14 | NUMBER 3 86 REFERENCES 1. Agnarsson, I., et al., PLoS ONE (2010) 5, e11234. 2. Andersen, S. O., Comp Biochem Physiol (1970) 35 , 705. 3. Denny, M., J Exp Biol (1976) 65, 483. 4. Blackledge, T. A., and Hayashi, C. Y., J Exp Biol (2006) 209, 2452. 5. Gosline, J. M., et al., Endeavour (1986) 10, 37. 6. Vollrath, F., J Biotechnol (2000) 74, 67. 7. Hayashi, C. Y., et al., Mol Biol Evol (2004) 21, 1950. 8. Madsen, B., et al., Int J Biol Macromol (1999) 24, 301. 9. Heim, M., et al., Angew Chem-Int Edit (2009) 48, 3584. 10. Emile, O., et al., Nature (2006) 440, 621. 11. Work, R. W., and Morosoff, N., Text Res J (1982) 52, 349. 12. Jelinski, L. W., et al., Int J Biol Macromol (1999) 24, 197. 13. Gosline, J. M., et al., Elastomeric Network Models for the Frame and Viscid Silks from the Orb Web of the Spider Araneus diadematus. In Silk Polymers, American Chemical Society, (1993), 544, 328. 14. Ayoub, N. A., et al., Plos One (2007) 2, e514. 15. Motriuk-Smith, D., et al., Biomacromolecules (2005) 6, 3152. 16. Rising, A., et al., Biomacromolecules (2006) 7, 3120. 17. Challis, R. J., et al., Insect Mol Biol (2006) 15, 45. 18. Askarieh, G., et al., Nature (2010) 465, 236. 19. Hagn, F., et al., Nature (2010) 465, 239. 20. Hinman, M. B., and Lewis, R. V., J Biol Chem (1992) 267, 19320. 21. Guerette, P. A., et al., Science (1996) 272, 112. 22. Hijirida, D. H., et al., Biophys J (1996) 71, 3442. 23. Hronska, M., et al., Biomacromolecules (2004) 5, 834. 24. Lefevre, T., et al., Biomacromolecules (2008) 9, 2399. 25. Lefevre, T., et al., Biomacromolecules (2007) 8, 2342. 26. Li, S. F., et al., Biophys. J. (1994) 66, 1209. 27. Sponner, A., et al., PLoS ONE (2007) 2, e998. 28. Augsten, K., et al., Scanning (2000) 22, 12. 29. Warwicker, J. O., J Mol Biol (1960) 2, 350. 30. Grubb, D. T., and Jelinski, L. W., Macromolecules (1997) 30, 2860. 31. Simmons, A., et al., Macromolecules (1994) 27, 5235. 32. Parkhe, A. D., et al., J Mol Recognit (1997) 10, 1. 33. van Beek, J. D., et al., PNAS (2002) 99, 10266. 34. Creager, M. S., et al., Biomacromolecules (2010) 11, 2039. 35. Jenkins, J. E., et al., Biomacromolecules (2010) 11, 192. 36. Simmons, A. H., et al., Science (1996) 271, 84. 37. Lefevre, T., et al., Biophys J (2007) 92, 2885. 38. Hayashi, C. Y., et al., Int J Biol Macromol (1999) 24, 271. 39. Savage, K. N., and Gosline, J. M., J Exp Biol (2008) 211, 1937. 40. Eles, P. T., and Michal, C. A., Biomacromolecules (2004) 5, 661. 41. Boutry, C., and Blackledge, T. A., J Exp Biol (2010) 213, 3505. 42. Shao, Z., et al., Macromolecules (2003) 36, 1157. 43. Seidel, A., et al., Macromolecules (2000) 33, 775. 44. Vollrath, F., and Knight, D. P., Nature (2001) 410, 541. 45. Jin, H. J., and Kaplan, D. L., Nature (2003) 424, 1057. 46. Exler, J. H., et al., Angew Chem-Int Edit (2007) 46, 3559. 47. Knight, D. P., and Vollrath, F., Proc R Soc Lond Ser B-Biol Sci (1999) 266, 519. 48. Hagn, F., et al., Angew Chem-Int Edit (2010) 50, 310. 49. Willcox, P. J., et al., Macromolecules (1996) 29, 5106. 50. Eisoldt, L., et al., J. Struct. Biol. (2010), 51. Bonthrone, K. M., et al., Proc R Soc Lond Ser B-Biol Sci (1992) 248, 141. 52. Vollrath, F., and Knight, D. P., Int J Biol Macromol (1999) 24, 243. 53. Vollrath, F., et al., Proc R Soc Lond Ser B-Biol Sci (1998) 265, 817. 54. Knight, D. P., et al., Int J Biol Macromol (2000) 27, 205. 55. Rammensee, S., et al., PNAS (2008) 105, 6590. 56. Knight, D. P., and Vollrath, F., Naturwissenschaften (2001) 88, 179. 57. Landreh, M., et al., J Mol Biol (2010) 404, 328. 58. Gaines, W. A., et al., J Biol Chem (2010) 285, 40745. 59. Heim, M., et al., Chem Soc Rev (2010) 39, 156. 60. Thiel, B. L., et al., Biopolymers (1997) 41, 703. 61. Trancik, J. E., et al., Polymer (2006) 47, 5633. 62. Fossey, S. A., and Tripathy, S., Int J Biol Macromol (1999) 24, 119. 63. Porter, D., and Vollrath, F., Adv Mater (2009) 21, 487. 64. Vollrath, F., and Porter, D., Soft Matter (2006) 2, 377. 65. Ene, R., et al., Soft Matter (2009) 5, 4568. 66. Papadopoulos, P., et al., Colloid Polym Sci (2009) 287, 231. 67. Arcidiacono, S., et al., Appl Microbiol Biotechnol (1998) 49, 31. 68. Vendrely, C., and Scheibel, T., Macromol Biosci (2007) 7, 401. 69. Rising, A., et al., Cell Mol Life Sci (2010) 68, 169. 70. Huemmerich, D., et al., Biochemistry (2004) 43, 13604. 71. Teule, F., et al., J Mater Sci (2007) 42, 8974. 72. Miao, Y. G., et al., Appl Microbiol Biotechnol (2006) 71, 192. 73. Zhou, Y. T., et al., Biomacromolecules (2001) 2, 111. 74. Heim, M., et al., J Struct Biol (2010) 170, 420. 75. Fahnestock, S. R., et al., J Biotechnol (2000) 74, 105. 76. Fahnestock, S. R., and Irwin, S. L., Appl Microbiol Biotechnol (1997) 47, 23. 77. Xia, X.-X., et al., PNAS (2010) 107, 14059. 78. Huemmerich, D., et al., Curr Biol (2004) 14, 2070. 79. Hermanson, K. D., et al., Advanced Materials (2007) 19, 1810. 80. Spiess, K., et al., Macromol Biosci (2010) 10, 998. 81. Slotta, U. K., et al., Angew Chem-Int Edit (2008) 47, 4592. 82. Spiess, K., et al., Soft Matter (2010) 6, 4168. 83. Hardy, J. G., and Scheibel, T. R., Prog Polym Sci (2010) 35, 1093. 84. Slotta, U., et al., Supramol Chem (2006) 18, 465. 85. Lammel, A., et al., ChemSusChem (2008) 1, 413. 86. Huemmerich, D., et al., Appl Phys A-Mater Sci Process (2006) 82, 219. 87. Vollrath, F., et al., In Vivo (2002) 16, 229. 88. Allmeling, C., et al., J Cell Mol Med (2006) 10, 770. 89. Allmeling, C., et al., Cell Proliferat (2008) 41, 408. 90. Gellynck, K., et al., J Mater Sci-Mater M (2008) 19, 2963. 91. Gellynck, K., et al., J Mater Sci-Mater M (2008) 19, 3399. 92. Hakimi, O., et al., J Biomed Mater Res A (2010) 92A, 1366. 93. Agapov, I., et al., Dokl Biochem Biophys (2009) 426, 127. 94. Widhe, M., et al., Biomaterials (2010) 31, 9575. 95. Fredriksson, C., et al., Materials (2009) 2, 1908. processed into many different morphologies and shapes which have great potential in various technical and biomedical applications. With recombinant production technologies it is possible to create tailormade silk-based biopolymers for many different purposes on a large scale within a relatively short amount of time. Acknowledgements: This work was supported by the Bayrische Staatsministerium für Umwelt und Gesundheit, project U8793-2008/11-2. We would to thank Dr. Martin Humenik for his creative input and technical support in creating the artwork. MT14_3p80-87.indd 86 2/24/2011 11:25:19 AM Teilarbeiten [111] Teilarbeit V Eisoldt, L., Thamm, C., Scheibel, T. (2012) The role of terminal domains during storage and assembly of spider silk proteins. Biopolymers. 97, pp. 355-36 3. Waite, J. H.; Qin, X. X.; Coyne, K. J. Matrix Biol 1998, 17, 93– 106. 4. Winkler, S.; Kaplan, D. L. Rev Mol Biotechnol 2000, 74, 85–93. 5. Hardy, J. G.; Romer, L. M.; Scheibel, T. R. Polymer 2008, 49, 4309–4327. 6. Eisoldt, L.; Smith, A.; Scheibel, T. Mater Today 2011, 14, 80–86. 7. Heim, M.; Keerl, D.; Scheibel, T. Angew Chem Int Ed 2009, 48, 3584–3596. 8. Rising, A.; Widhe, M.; Johansson, J.; Hedhammar, M. Cell Mol Life Sci 2010, 1–16. 9. Omenetto, F. G.; Kaplan, D. L. Science 2010, 329, 528–531. 10. Sponner, A.; Unger, E.; Grosse, F.; Weisshart, K. Biomacromolecules 2004, 5, 840–845. 11. Denny, M. J Exp Biol 1976, 65, 483–506. 12. Blackledge, T. A.; Hayashi, C. Y. J Exp Biol 2006, 209, 2452– 2461. 13. Vollrath, F. J Biotechnol 2000, 74, 67–83. 14. Madsen, B.; Shao, Z. Z.; Vollrath, F. Int J Biol Macromol 1999, 24, 301–306. 15. Hayashi, C. Y.; Blackledge, T. A.; Lewis, R. V. Mol Biol Evol 2004, 21, 1950–1959. 16. Ayoub, N. A.; Garb, J. E.; Tinghitella, R. M.; Collin, M. A.; Hayashi, C. Y. Plos One 2007, 2. 17. Hinman, M. B.; Lewis, R. V. J Biol Chem 1992, 267, 19320– 19324. 18. Guerette, P. A.; Ginzinger, D. G.; Weber, B. H. F.; Gosline, J. M. Science 1996, 272, 112–115. 19. Motriuk-Smith, D.; Smith, A.; Hayashi, C. Y.; Lewis, R. V. Biomacromolecules 2005, 6, 3152–3159. 20. Rising, A.; Hjalm, G.; Engstrom, W.; Johansson, J. Biomacromolecules 2006, 7, 3120–3124. 21. Challis, R. J.; Goodacre, S. L.; Hewitt, G. M. Insect Mol Biol 2006, 15, 45–56. 22. Garb, J.; Ayoub, N.; Hayashi, C. BMC Evolutionary Biology 2010, 10, 243. 23. Hu, X. Y.; Kohler, K.; Falick, A. M.; Moore, A. M. F.; Jones, P. R.; Vierra, C. Biochemistry 2006, 45, 3506–3516. 24. La Mattina, C.; Reza, R.; Hu, X.; Falick, A. M.; Vasanthavada, K.; McNary, S.; Yee, R.; Vierra, C. A. Biochemistry 2008, 47, 4692–4700. 25. Sponner, A.; Unger, E.; Grosse, F.; Klaus, W. Nat Mater 2005, 4, 772–775. 26. Askarieh, G.; Hedhammar, M.; Nordling, K.; Saenz, A.; Casals, C.; Rising, A.; Johansson, J.; Knight, S. D. Nature 2010, 465, 236–U125. 27. Hagn, F.; Eisoldt, L.; Hardy, J. G.; Vendrely, C.; Coles, M.; Scheibel, T.; Kessler, H. Nature 2010, 465, 239–242. 28. Hagn, F.; Thamm, C.; Scheibel, T.; Kessler, H. Angew Chem Int Ed 2010, 50, 310–313. 29. Hijirida, D. H.; Do, K. G.; Michal, C.; Wong, S.; Zax, D.; Jelinski, L. W. Biophys J 1996, 71, 3442–3447. 30. Vollrath, F.; Knight, D. P. Nature 2001, 410, 541–548. 31. Knight, D. P.; Vollrath, F. Naturwissenschaften 2001, 88, 179– 182. 32. Jin, H. J.; Kaplan, D. L. Nature 2003, 424, 1057–1061. 33. Hu, X. Y.; Yuan, J.; Wang, X. D.; Vasanthavada, K.; Falick, A. M.; Jones, P. R.; La Mattina, C.; Vierra, C. A. Biochemistry 2007, 46, 3294–3303. 34. Knight, D. P.; Vollrath, F. Proc R Soc Lond Ser B Biol Sci 1999, 266, 519–523. 35. Landreh, M.; Askarieh, G.; Nordling, K.; Hedhammar, M.; Rising, A.; Casals, C.; Astorga-Wells, J.; Alvelius, G.; Knight, S. D.; Johansson, J.; Jo ¨rnvall, H.; Bergman, T. J Mol Biol 2010, 404, 328–336. 36. Gaines, W. A.; Sehorn, M. G.; Marcotte, W. R., Jr. J Biol Chem 2010, 285, 40745–40753. 37. Huemmerich, D.; Helsen, C. W.; Quedzuweit, S.; Oschmann, J.; Rudolph, R.; Scheibel, T. Biochemistry 2004, 43, 13604–13612. 38. Exler, J. H.; Hummerich, D.; Scheibel, T. Angew Chem Int Ed 2007, 46, 3559–3562. 39. Eisoldt, L.; Hardy, J. G.; Heim, M.; Scheibel, T. R. J Struct Biol 2010, 170, 413–419. 40. Stark, M.; Grip, S.; Rising, A.; Hedhammar, M.; Engstrom, W.; Hjalm, G.; Johansson, J. Biomacromolecules 2007, 8, 1695– 1701. 41. Sponner, A.; Vater, W.; Rommerskirch, W.; Vollrath, F.; Unger, E.; Grosse, F.; Weisshart, K. Biochem Biophys Res Commun 2005, 338, 897–902. 42. Hedhammar, M.; Rising, A.; Grip, S.; Martinez, A. S.; Nordling, K.; Casals, C.; Stark, M.; Johansson, J. Biochemistry 2008, 47, 3407–3417. 43. Rammensee, S.; Slotta, U.; Scheibel, T.; Bausch, A. R. Proc Natl Acad Sci U S A 2008, 105, 6590–6595. 44. Papadopoulos, P.; Ene, R.; Weidner, I.; Kremer, F. Macromol Rapid Commun 2009, 30, 851–857. 45. Papadopoulos, P.; Solter, J.; Kremer, F. Colloid Polym Sci 2009, 287, 231–236. Reviewing Editor: David Nate Breslauer Storage and Assembly of Spider Silk Proteins 361 Biopolymers Rechtliches [120] Rechtliches Für die Publikationen in den Teilarbeiten II, IV und V wurden die Rechte zur elektronischen und gedruckten Vervielfältigung des jeweiligen vollständigen Artikels im Rahmen dieser Dissertation bei den Verlagen eingeholt. Für Teilarbeit I gewährt der Verlag (Nature Publishing Group) jedem Autor das Recht den vollständigen Artikel im Rahmen einer Thesis/Dissertation in elektronischer und gedruckter Form zu vervielfältigen (Punkt a. unter „Author Requests“ auf http://www.nature.com/reprints/permission-requests.html). Anhang [121] Anhang Tabelle A1: Übersicht bekannter carboxyterminaler (CT) und aminoterminaler (NT) Sequenzen von verschiedenen Spidroinen und Arten. Die Sequenzen wurden der UniProt-Datenbank entnommen (http://www.uniprot.org/uniprot/, Stand Januar 2013) und soweit möglich publizierte Quellen mit angegeben. Der Übersicht halber wurden redundante Spidroinsequenzen, die auf unterschiedlichen cDNA-Klonen basieren nicht mit aufgenommen. Die Sequenzen sind nach Unterordnung und jeweilige Familie sortiert. Die Arten innerhalb der Familien sind alphabetisch sortiert. Bei Spidrointypen, die mit „Fibroin“ gekennzeichnet sind, konnten die jeweiligen repetitiven Bereiche nicht den bekannten Spidroinklassen zugeordnet werden. Die terminalen Domänen zeigen jedoch eine signifikante Verwandtschaft zu Sequenzen von terminalen Spidroindomänen. Die publizierte Nucleotidsequenz aus der die mit * gekennzeichneten NT-Sequenz des Flagelliform Spidroins von N. clavipes (O44358) hervorgeht enthält an Position 110 ein Adenosin zu viel, welches die Proteinsequenz ab Position 36 verfälscht. Die korrigierte Sequenz ist in [33] angegeben, Verwendete Abkürzungen: MaSp: Major Ampullate Spidroin; MiSp: Minor Ampullate Spidroin; Flag: Flagelliform Spidroin; AcSp:Aciniform Spidroin; TuSp: Tubuliform Spidroin; Pyriform Spidroin Spezies Spidroin Bekannter Terminus UniProt-Nummer Quelle Unterordnung Echte Webspinnen (Araneomorphae) Echte Radnetzspinnen (Araneidae) Araneus bicentenarius MaSp1 CT P46802 [24] MaSp2 CT Q9TY62 [24] Araneus diadematus MaSp1 CT Q16986 [7] MaSp2 CT Q16987 [7] Araneus genmoides TuSp1 CT Q3BCG3 [163] Araneus ventricosus MaSp1 CT G9JKN1 - MiSp NT + CT K4MTL7 [27] Flag CT A0FK72 [164] AcSp1 CT E1AW39 - Argiope amonea MaSp1 CT Q2TV51 - MaSp2 CT Q3ZTQ8 - AcSp1 CT E1AW40 - Argiope argentata MiSp1 CT I4CJN7 [165] TuSp1 NT + CT NT: E1AHW1 CT: Q4G1Y5 [30, 37] Argiope aurantia MaSp2 CT Q9BIV0 [29] TuSp1 CT Q4G1X5 [30] Argiope bruennichi MaSp1 CT I6XQ31 [47] MaSp2 NT + CT I6YNT3 [47] Argiope trifasciata MaSp1 CT Q9BIU7 [29] MaSp2 NT + CT NT: Q2VLH1 CT: Q9BIU6 [21] Flag CT Q9BIU9 [29] AcSp1 CT Q64K55 [13] PySp1 CT E2DEK3 - Cyrtophora moluccensis MaSp1 CT Q5UNI2 [166] TuSp1 CT Q4G1X3 [30] Gasteracantha cancriformis MaSp2 CT Q9BIU1 [29] Gea haptagon TuSp1 CT Q4G1X4 [30] Metepeira grandiosa MiSp1 NT + CT NT: E1AHV7 CT: E1AHU9 [37] Nephila clavipes MaSp1 NT + CT NT: B5SYS5 CT: P19837 [23, 35] MaSp2 NT + CT NT: B5SYS7 CT: P46804 [9, 35] Flag NT + CT *NT: O44358 CT: O44359 [12] TuSp1 CT Q3BCG2 [163] PySp1 CT E2DEK5 [20] PySp2 CT G0WJI5 [31] Anhang [122] Nephila inaurata madagascariensis MaSp1 CT Q9BIT6 [29] MaSp2 NT + CT NT: Q2VLH2 CT: Q9BIT5 [21, 29] Flag NT Q9NHW2 [28] Nephila pilipes MaSp1 CT Q5UNI6 [166] Nephila senegalensis MaSp1 CT Q9BIT4 [28] Masp2 CT Q9BIT3 [28] Nephilengyys cruentata MaSp1 CT A6YP78 [11] MiSp1 CT A6YP79 [11] Flag CT A6YP76 [11] TuSp1 CT A6YP77 [11] PySp1 CT E3SH49 - Parawixia bistriata MaSp1 CT I6LKZ4 - MaSp2 CT I6LKZ5 - MiSp1 CT E9JEC3 - AcSp1 CT I6LKZ2 - Tetragnatha kauaiensis MaSp1 CT Q9BIS8 [29] Tetragnatha versicolor MaSp1 CT Q9BIS7 [29] Haubennetzspinnen (Theridiidae) Latrodectus geometricus MaSp1 NT+ CT NT: B0F665 CT: Q9BIU0 [29, 167] MaSp2 NT + CT NT: B0F656 CT: Q9BIT8 [29, 167] TuSp1 CT Q4G1X7 [30] Latrodectus hasseltii TuSp1 CT Q4G1X6 [30] Latrodectus hesperus MaSp1 NT + CT A6YIY1 [8] Masp2 NT + CT A6YIY0 [8] MiSp NT + CT NT: E1AHV0, CT: E1AHV1 [37] AcSp1 NT + CT GenBank: JX978171.1 [15] TuSp NT + CT NT: Q1KXY8 CT: Q4G1Y6 [30] PySp CT C7T5D2 [19] Latrodectus mactans MaSp1 NT E5G5R4 [34] Tusp1 CT Q4G1X9 [30] Netzwerfende Spinnen (Deinopidae) Deinopis spinosa MaSp2 CT Q15G92 [168] MiSp1 CT Q15G96 [168] Flag CT Q15G95 [168] TuSp1 CT Q4G1Y3 [168] Kräuselradnetzspinnen (Uloboridae) Uloborus diversus MaSp1 CT Q15G89 [168] MaSp2 CT Q15G85 [168] MiSp1 NT + CT NT: E1AHV5 CT: Q15G88 [37] AcSp1 CT Q15G87 [168] TuSp1 CT Q4G1Y4 [168] Octonoba varians MaSp1 CT Q5UNK3 [166] Trichterspinnen (Agelenidae) Agelenopsis aperta MaSp1 NT + CT NT: E1AHV3 CT: Q6Q294 [37, 169] TuSp1 NT + CT NT: E1AHV9 CT: E1AHV2 [37] Raubspinnen (Pisauridae) Euprosthenops australis MaSp1 NT + CT NT: Q05H60 CT: A7M8L1 [32, 33] MaSp2 CT A7M8K6 [32] Psechridae Psechrus sinensis MaSp1 CT Q5UNJ3 [166] Luchsspinnen (Oxyopidae) Peucetia viridans MaSp1 CT D5JG83 [47] Diguetidae Diguetia canities MaSp1 NT + CT NT: E1AHU4 CT: E1AHU5 [37] Lochröhrenspinnen (Filistatidae) Anhang [123] Kukulcania hibernalis MaSp1 NT E1AHU3 [37] Unterordnung Vogelspinnartige (Mygalomorphae) Cyrtaucheniidae Aptostichus spp. Fibroin1 CT A8UAJ5 [170] Fibroin2 CT A8UAJ6 [170] Antrodiaetidae Atypoides riversi Fibroin1 CT I6ULS5 [171] Eigentliche Falltürspinnen (Ctenizidae) Bothriocyrtum californicum Fibroin1 NT + CT NT: E1AHU2 CT: A8UAJ7 [37, 170] Doppelschwanzspinnen (Dipluridae) Euagrus chisoseus Fibroin1 CT Q9BIU2 [29] Mecicobothriidae Megahexura fulva Fibroin1 CT I6ULT0 [171] Vogelspinnen (Theraphosidae) Avicularia juruensis MaSp1 CT D4HL83 [169] MaSp2 CT D4HL84 [169] Trichternetzspinnen (Hexathelidae) Macrothele holsti MaSp1 CT Q5UNJ2 [166] Danksagung [124] Danksagung Ein großer Dank geht an meinen Betreuer und wissenschaftlichen Mentor Prof. Thomas Scheibel, der stets für mich da war und mir die Gelegenheit wie auch die Freiheiten gab auf diesem spannenden Projekt zu promovieren. Den Mitgliedern des ursprünglichen „Fiberlabs“, namentlich Anja, Christian, Kristina, Simone und Ute, möchte ich für all die Hilfen, schon vor Beginn der Doktorarbeit, und den schönen Stunden in und vor allem auch außerhalb des Labors danken. All meinen Kollegen und Freunden des Lehrstuhls Biomaterialien gebührt ebenfalls mein Dank. Hier sei natürlich besonders die Gummibärchenverpflegungsstation um Claudia Mädels-Büro, die unzählige Male für den nötigen Zuckerschub erwähnt. Das hat so manchen Forschertag oder –abend gerettet. Ein Dank auch an die regelmäßigen Gäste des TripleB, Aniela, Chris, Elena, Felix, Kristin, Martina und Steffi für die sehr gemütliche Zeit sowie an den tapferen Mitstreiter Markus für gewisse FAN-Feten-Bars. Eileen, danke für die kurzweiligen Fahrten von Bayreuth nach München und zurück. Vielen Dank auch an unsere Post-Docs Martin, Andrew und John für all die kleinen und nützlichen Tipps. John gebührt mein aufrichtiger Dank, besonders in den ersten 2,5 Jahren der Arbeit! Ein spezieller Gruß geht an dieser Stelle an den Herz Achter bzw. die Bayreuther Schafkopfrunde um Andreas, Gregor, Markus, Johannes und Michi. Anderl und Johannes, sowie hier natürlich auch Jasmin, Claudia S. und Dani, danke ich natürlich auch für die tatkräftige Unterstützung im Labor. Ein besonderer Dank geht an meine Kollegen und Freunde im Büro Martin und Michael. Ohne die vielen wissenschaftlichen Diskussionen und Anregungen wäre vieles nicht so gekommen wie es heute hier steht. Auch die weniger wissenschaftlichen Gespräche waren wichtig und werden in guter Erinnerung bleiben. Meinem langjährigen Mitstreiter Michael gebührt ein besonderer Dank, nicht nur für all die gemeinsamen und mühevollen Stunden auf der Suche nach dem einen Haken, der Betreuung von OSI-Layer 8 oder sonst einer der unzähligen Administrationstätigkeiten, sondern auch für all die Hilfen, Gespräche und das gesamte Drumherum. Natürlich möchte ich mich bei meiner Familie bedanken, ohne die das alles, beginnend mit dem Studium, nicht möglich gewesen wäre. Ohne eure Unterstützung wäre das nicht gegangen. Zuletzt geht das größte Dankeschön an meine Frau Tina, die während der gesamten Zeit an meiner Seite war und vieles ermöglicht und auch ertragen musste. Vielen, vielen lieben Dank, mein Schatz. Erklärung [125] Hiermit versichere ich an Eides statt, die vorliegende Arbeit selbständig angefertigt und keine anderen als die angegebenen Quellen und Hilfsmittel verwendet zu haben. Des Weiteren erkläre ich, dass ich nicht anderweitig mit oder ohne Erfolg versucht habe, diese Dissertation einzureichen. Ich habe keine gleichartige Doktorprüfung an einer anderen Hochschule endgültig nicht bestanden. Bayreuth, den 05.06.2013 ……………………………… Lukas Eisoldt