ELITE NETZWERK BAYERN

English  Sprachen Icon  |  Gebärdensprache  |  Leichte Sprache  |  Kontakt


Forschungsarbeit

Amyloid aggregates interact with essential metastable proteins causing cytotoxicity

By Dr. Heidi Olzscha (12.10.2011)

Aberrant protein folding is often the result of destabilizing mutations and may cause the loss or impairment of critical functions. However, in number of diseases, aberrant protein folding and aggregation results in a toxic gain of function. In these disorders, specific proteins, differing substantially in size and sequence, typically self-assemble into amyloid-like fibrils with cross-beta structure. This phenomenon underlies some of the most debilitating neurodegenerative disorders, including Alzheimer's, Parkinson's and Huntington's disease.

To distinguish between the loss-of-function/gain-of-function aspect and elucidate to what degree aberrant protein interactions and the protein quality control system is involved in the cytotoxicity of amyloid proteins, we have established a novel cellular model with amyloid-like proteins. The sequences of these proteins were designed to contain b-strands with an alternating pattern of polar and non-polar residues, while the exact identities of the side chains were varied combinatorially. As a control, we used a designed α-helical protein α-S824 (Fig. 1).

Fig. 1: Structures of the de novo generated amyloid-like proteins β4, β17, β23, and the de novo α-helical protein α-S824.[Bildunterschrift / Subline]: Fig. 1: Structures of the de novo generated amyloid-like proteins β4, β17, β23, and the de novo α-helical protein α-S824. β4, β17, and β23 are designed to form 6 β-strands, indicated by red arrows, connected by 5 turns. The β-strands consist of 7 amino acids with an alternating hydrophilic/hydrophobic pattern. Hydrophilic amino acids are depicted in blue, hydrophobic amino acids are shown in yellow. The α-helical proteins α-S824 occurs as a 4-helix-bundle protein (green). The N-terminal Myc-epitope is shown in orange.

Upon expression in human HEK293T cells, the β-proteins, but not the α-helical protein α-S824, formed aggregates. Fig. 2 also illustrates that the cell shape is altered in cells expressing the de novo amyloid proteins in comparison to cells which are transfected with an empty vector only (control) or with the de novo α-helical protein α-S824. Furthermore, the actin cytoskeleton is disrupted and filopodia are diminished or missing. Taken together, these morphological alterations of the cells gave an initial hint that the expression of the amyloid-like proteins is cytotoxic.

Specific tests for cytotoxicity caused by amyloidogenic proteins demonstrated that the three de novo amyloid-like proteins β4, β17, and β23 were cytotoxic in the following gradation: β23 > β17 > β4 >> α-S824 (Fig. 3).

In addition, expression of the model β-sheet proteins leads to a deficiency of the normal cytosolic stress response, thereby limiting the capacity of cells to mount an effective defense.

Fig. 2: De novo amyloid-like proteins form aggregates in human cells (β 23) and alter the morphology of the cell and the actin cytoskeleton. Fig. 3: Cell viability of HEK293T cells expressing β-sheet proteins.[Bildunterschrift / Subline]: Fig. 2: De novo amyloid-like proteins form aggregates in human cells (β 23) and alter the morphology of the cell and the actin cytoskeleton. Control cells were transfected with an empty vector. The de novo proteins were detected by an α-Myc-antibody conjugated to a fluorescent dye, the actin cytoskeleton was visualized by phalloidin conjugated to rhodamine. Fig. 3: Cell viability of HEK293T cells expressing β-sheet proteins. Cells were transfected with the proteins and 3 days later the cell viability was analyzed by means of MTT assay. β 23 is the most toxic protein, followed by β 17 and β 4. The soluble α-helical protein α-S824 displays nearly the same viability as control cells.

A comparison of the proteins that interact preferentially with β4, β17 and β23, as defined by the SILAC experiments (Fig. 4), revealed that a gradual increase in molecular weight and decrease in hydrophobicity of the interactors, along with a slight increase in their disorder, correlated with the differential cytotoxicity of the three β-proteins (Fig. 5).

Essential proteins often occupy critical ‘hub’ positions in the network. The interactors of β4, β17 and β23 are more frequently linked than lysate proteins, through direct interactions, with proteins that have been found in association with neurodegenerative disease proteins. Therefore, the capacity of the β-protein aggregates to interact with and sequester highly connected cellular proteins correlates strongly with their relative cytotoxicity. It is thus likely that the aggregates compete for binding to disordered regions with a protein’s normal interactors and the more toxic forms may be able to compete more effectively. Certain proteins may misfold upon interaction with the aggregates but remain in solution. As a result, dependent on the interaction strength of the aggregates, an increasing number of key functions may be affected, eventually resulting in fatal network collapse.

Moreover, a selective overexpression of a detected and highly enriched interactor, Hsp110, can partly suppress the cytotoxicity of β4 and β17, prevent the formation of aggregates and keep the cell in their normal shape.

 

Fig. 4: β 23 interactome. Fig. 5: Physico-chemical properties and functional context of β-protein interactors.[Bildunterschrift / Subline]: Fig. 4: β 23 interactome. The 103 enriched proteins with β 23 are grouped here according to cellular location and function (cell scheme: © wissenmedia in der inmediaONE] GmbH, Gütersloh). Fig. 5: Physico-chemical properties and functional context of β-protein interactors. The molecular weight, number and length of IURs correlate with the cytotoxicity of the β-sheet proteins (fig. 3). The hydophobicity of the interactors correlate in an inverse manner with the cytotoxicity (left bar diagram). In the diagram on the right, the functional context of the β-protein interactors within the protein interaction network is shown. The number of total interactors, the essential interactors and with neurodegenerative diseases associated interactors correlate with the observed cytotoxicity.

Taken together, these results give an explanation why amyloidogenic proteins with different sequences and lengths lead to similar cytotoxic events which cannot be compensated for by the cellular defense mechanisms. Overexpression of enriched interacting chaperones and eventually other enriched interacting proteins could be a useful tool to suppress cytotoxicity and the formation of amyloid aggregates (Fig. 6).

Fig. 6: Model for the sequestration of essential metastable proteins.[Bildunterschrift / Subline]: Fig. 6: Model for the sequestration of essential metastable proteins. Pre-existent proteins are structurally flexible in their functional state and are involved in multiple protein-protein interactions, which may be disrupted by their association with the β-aggregates. Newly-synthesized proteins are structurally vulnerable to co-aggregation during folding and assembly. Interaction of the metastable proteins with the β-aggregates is facilitated by the limited capacity of chaperones to shield aggregation-prone surfaces and by the failure of the cells to induce an efficient stress response. Overexpression of selected interactors and especially molecular chaperones can protect the cell against cytotoxicity and the formation of aggregates.

References

Heidi Olzscha, Sonya M. Schermann, Andreas C. Woerner, Stefan Pinkert, Michael H. Hecht, Gian G. Tartaglia, Michele Vendruscolo, Manajit Hayer-Hartl, F. Ulrich Hartl, and R. Martin Vabulas. (2011). Amyloid-like Aggregates Sequester Numerous Metastable Proteins with Essential Cellular Functions. Cell. 144, 67-78.


Heidi Olzscha
Heidi Olzscha
* 1977

Education
  • May 2011 – present
  • University of Oxford, Department of Oncology, Postdoctoral researcher (EMBO fellow), Prof. Dr. Nicholas B. La Thangue
  • January 2005 – September 2010
  • Ph.D. student at Max Planck Institute of Biochemistry, Department of Cellular Biochemistry, Martinsried, Germany, Prof. Dr. F.-Ulrich Hartl (Ph.D. supervisor)
  • Doctoral degree Dr. rer. nat. (equivalent to Ph.D.) in biochemistry, Project: Analysis of structural determinants underlying the toxic effects of amyloid proteins
  • October 1999 – August 2004
  • Studies of Biochemistry/molecular biology, University of Hamburg, Germany
  • Diploma thesis on the Characterisation of phosphotyrosine binding sites of the human FLT3 receptor with modular domains

Publications
  • Heidi Olzscha, Sonya M. Schermann, Andreas C. Woerner, Stefan Pinkert, Michael H. Hecht, Gian G. Tartaglia, Michele Vendruscolo, Manajit Hayer-Hartl, F. Ulrich Hartl, and R. Martin Vabulas. (2011).
  • Amyloid-like Aggregates Sequester Numerous Metastable Proteins with Essential Cellular Functions. Cell. 144, 67-78

Scientific prizes and fellowships
  • EMBO long-term postdoctoral fellowship for 2 years, Oxford, UK, 2011
  • Max Planck Institute of Biochemistry Junior Research Award 2011 for outstanding research in biochemistry, Martinsried, Germany, 2011
  • Poster prize winner (1st place), EMBO Conference Series 'Cellular protein homeostasis in disease and ageing', Dubrovnik, Croatia, 2009
  • Selected to participate in the 56th meeting of Nobel Laureates (Chemistry) in Lindau, Germany, 2006
  • Member of the elite network Bavaria, international graduate program 'Protein dynamics in health and disease', Munich, Germany, 2005-2010
  • PhD Kekulé Fellowship from Fonds of Chemical Industrie Germany (FCI, Germany), Martinsried, Germany, 2005-2007