Doctoral Thesis
Refine
Year of publication
- 2014 (2) (remove)
Document Type
- Doctoral Thesis (2) (remove)
Has Fulltext
- yes (2)
Is part of the Bibliography
- no (2)
Keywords
- Thrombozytopenie (2) (remove)
Die Heparin-induzierte Thrombozytopenie (HIT) ist eine seltene, aber lebens-bedrohliche unerwünschte Wirkung einer Therapie mit Heparin. Ausgelöst wird sie durch die Bildung von Antikörpern, die gegen Komplexe aus dem positiv geladenem Thrombozytenprotein Plättchenfaktor 4 (PF4) und negativ geladenem Heparin gerichtet sind und Thrombozyten aktivieren können, welches sich klinisch in einer schweren prothrombotischen Diathese bei gleichzeitiger Thrombozytopenie äußert. Bemerkenswert ist, dass im Widerspruch zur klassischen Immunantwort bei der HIT auch von Heparin-naiven Patienten binnen weniger Tage anti-PF4/Heparin-Antikörper der Immunglobulinklasse G gebildet werden, welche eigentlich eine sekundäre Immunantwort repräsentieren. Darüber hinaus ist bekannt, dass PF4 auch mit verschiedenen polyanionischen Nichtheparinen Komplexe bildet, die von anti-PF4/Heparin-Antikörpern erkannt werden. In der vorliegenden Arbeit wurde untersucht, ob Nukleinsäuren, die aufgrund ihres Reichtums an Phosphatgruppen ebenfalls stark negativ geladenen sind, mit PF4 Komplexe bilden, die von anti-PF4/Heparin-Antikörpern erkannt werden. Da eine Bildung antigener Komplexe mit PF4 für therapeutisch eingesetzte Nukleinsäuren von hoher praktischer Relevanz wäre, wurde außerdem die Interaktion von PF4 mit verschiedenen Aptameren untersucht. Dazu wurde die Interaktion zwischen PF4 und Nukleinsäuren in systematischen Bindungsstudien mit verschieden strukturierten Nukleinsäuren charakterisiert. Mittels Circulardichroismus-Spektroskopie wurde der Einfluss eines Modellaptamers auf die Sekundärstruktur von PF4 mit dem Einfluss von Heparin auf die PF4-Struktur verglichen. Die Kreuzreaktivität von anti-PF4/Heparin-Antikörpern gegen PF4/Nukleinsäure- bzw. PF4/Aptamer-Komplexe wurde mittels Immunassay und Thrombozytenfunktionstest untersucht. Schließlich wurde im Mausmodell die in-vivo-Immunogenität eines PF4/Aptamer-Komplexes überprüft. Es konnte gezeigt werden, dass Nukleinsäuren längen- und strukturabhängig mit PF4 interagieren. Die Neigung zur Komplexbildung mit PF4 stieg mit der Länge der Nukleinsäure und dem Anteil doppelsträngiger Abschnitte. PF4/Nukleinsäure-Komplexe wurden von anti-PF4/Heparin-Antikörpern erkannt und konnten in Gegenwart dieser Antikörper eine Thrombozytenaktivierung vermitteln. Mit PF4/Aptamer-Komplexen immunisierte Mäuse produzierten anti-PF4/Aptamer-Antikörper, welche gegen PF4/Heparin-Komplexe kreuzreagierten. Diese Ergebnisse lassen vermuten, dass insbesondere Aptamere mit hohem Anteil an doppelsträngigen Abschnitten mit PF4 Komplexe bilden können, die ein ähnliches Epitop exprimieren wie PF4/Heparin-Komplexe und damit ein potentielles Risiko für die Erzeugung HIT-ähnlicher prothrombotischer Komplikationen bergen. Nukleinsäuren könnten einen endogenden Interaktionspartner von PF4 darstellen und das höhere HIT-Risiko von Patienten nach großen orthopädischen Eingriffen oder schweren Infektionen erklären, welche mit höheren Plasmaspiegeln zellfreier Nukleinsäuren einhergehen. Darüber hinaus könnte die Interaktion von PF4 mit endogenen Nukleinsäuren eine Erklärung für die frühe Produktion von anti-PF4/Heparin-Antikörper der Immunglobulinklasse G liefern.
The central aim of this thesis was the investigation of protein/polyanion interaction using circular dichroism (CD) spectroscopy, enzyme immune assay (EIA), isothermal titration calorimetry (ITC) and flow cytometry (FC). A further aim was to understand why an endogenous protein becomes immuno-genic when forming a complex. The focus was on the protein platelet factor (PF4), which gained wide interest in the clinical field, due to its role in the life-threatening, immune-driven, adverse drug effect heparin-induced thrombocytopenia (HIT). PF4 is a small homotetrameric chemokine with several basic amino acids on its surface, forming a positively charged ring. The antibodies that are formed during HIT recognize an epitope exposed on PF4, when it is in a complex with heparin at a certain molar ratio at which, PF4 tetramers are aligned on the heparin and forced into close approximation. The main results and conclusions of the thesis are summarized below: 5.1 Evolutionary Conservation of PF4 (Paper I – PF4/Evolution) By carrying out an amino acid sequence survey we found that the positively charged amino acids contributing to the heparin binding site on the surface of PF4 and related proteins are highly conserved in all vertebrates, including fish species. PF4 interacts with the phospholipid lipid A, the innermost part of the lipopolysaccharide (LPS) of Gram negative bacteria. We showed that the shorter the sugar chain of the O antigen, outer and inner core of the LPS were the more PF4 was binding. The interaction of PF4 with lipid A is inhibited by heparin, suggesting that the amino acids known to contribute to heparin binding are also involved in binding to lipid A. 5.2 PF4 Interaction with Polyanions (PA) of varying Length and Degree of Sulfation (Paper II – PF4/PA) CD spectroscopy was found to be a powerful technique to monitor structural changes of PF4 caused by binding to various clinically relevant polyanions. Therefore PF4 was titrated with different PA to investigate the dependencies: i. impact of the PF4:PA molar ratio, ii. degree of polymerization of the PA and iii. degree of sulfation of the PA. In all cases, exposure of HIT-relevant epitope(s) was only observed for PA that also induced changes in secondary structure of PF4. A comparison of results of an immune ¬assay with CD spectroscopic data showed that the extent of complex anti¬genicity correlates well with the magnitude of changes in PF4 secondary structure, and that the structural changes of PF4 have to exceed a certain threshold to achieve PF4/PA complex antigenicity. These findings allowed us to calculate expectation intervals for complex antigenicity solely using CD spectroscopic data. To our knowledge, this was the first demonstration that the capability of drugs to induce antigenicity of PF4 can be assessed without the necessity of in vivo studies or the use of antibodies obtained from immunized patients specific for the antigens. The antigenicity of PF4 in complex is not restricted to negative charges originating from sulfate groups, PA with phosphate groups are also capable (binding to phospholipids). We investigated inorganic polyphosphates (polyP) with a chain length of 75 Pi and showed that the induced secondary structural changes are even higher compared to the changes induced by the different heparins and that the PF4/P75 complexes are antigenic as well. 5.3 PF4 Interaction with defined oligomeric Heparins (Paper III – PF4/defined Heparins) We tested highly purified, monodisperse heparins. In contrast to the clinically relevant but relatively undefined (high polydispersity index) glycosamino glycans reported in paper II (PF4/PA). The defined heparins induced higher secondary structural changes. Here we showed for the first time that strong conformational changes during PF4/PA complex formation are necessary but not sufficient for to the expression of the anti-PF4/heparin antibody binding site. Also, the size of the complexes is not the only prerequisite for anti-PF4/heparin antibody binding (tested by atomic force microscopy). By ITC we found that antigenicity is only induced if the PF4/PA complex has a high binding enthalpy and the complex formation leads to a negative change in entropy. 5.4 PF4/Polyphosphates (polyP) Complex Antigenicity and Interaction with Escherichia coli (E. coli, Paper IV – PF4/polyP) PolyP with chain lengths of 45 Pi and 75 Pi induced remarkable secondary structural changes in the PF4 molecule, thereby exposing the epitope recognized by anti-PF4/heparin antibodies. The induced conformational changes were similar to the changes induced by the defined heparins. Again a high binding enthalpy was observed but here in connection with a positive change in entropy. Further we showed that polyP (≥45 Pi) enhance PF4 binding to the surface of Gram negative E. coli at intermediate concentration and disrupt the binding at elevated polyP concentrations. The increased amounts of PF4 on the bacterial surface also improved the binding of anti-PF4/heparin antibodies and thereby the phagocytosis of the bacteria by poly¬morpho¬nuclear leucocytes. 5.5 Nucleic acid based Aptamers induce structural Changes in the PF4 Molecule (Paper V – PF4/Aptamer) Nucleic acids are another class of molecules containing phosphate groups. Especially after cell damage their extra¬cellular concentration can be locally quite high (>2 mg/ml). We found that certain aptamers form complexes with PF4 and thereby inducing anti-PF4/aptamer antibodies which cross-react with PF4/heparin complexes. Moreover by CD spectroscopy we showed that the protein C-aptamer caused similar secondary structural changes of PF4 like heparin, but already at much lower concentration. The maximally induced changes by the protein-C aptamer were even higher and persisted over a broader concentration range. 5.6 Protamine Interaction with Heparin (Paper VI – PS/Heparin) After the intensive investigation of the complex formation between PF4 and many different classes of PA we assessed another protein for structural changes upon complex formation with heparin. Protamine (PS) a protein in routinely used in post-cardiac surgery to reverse the anticoagulant effects of heparin was found to unfold but not to refold with increasing concentration of PA in solution. 5.7 Conclusion and Outlook When starting this thesis, it was believed that repetitive structures formed by PF4 on a heparin chain mold the epitope recognized by antibodies inducing HIT. These repetitive structures might exhibit similarities with viral capsids and are therefore recognized by the immune system of some patients. We found that induced by the close approximation PF4 changes its conformation, thereby exposing a neoepitope. The conserved positively charged amino acids of the heparin binding site and the involvement of these amino acids in the binding to lipid A confirm our hypothesis of PF4 as part of an ancient immune-mediated host defense mechanism. As possible consequence of the “primitive mechanism of defense” the highly variable O-antigens of LPS might have significantly contributed to an efficient escape mechanism by hiding the structures that made the bacteria vulnerable. In turn polyP might be an adaption of the host improve pathogen recognition by PF4 and further by antibodies inducing phagocytosis of the PF4-marked objects. Although shown only for PF4 and PS, our findings might be applicable to other proteins that also express epitopes upon changes in their secondary structure. Our physicochemical methods may further be applied: i. to drug development for the prediction of antigenicity induced by polyanionic drugs, ii. to guide the development of synthetic heparins and other polyanion based drugs, e.g. aptamers, that do not lead to HIT and iii. to provide relevant aspects for other biological functions of heparins.