What are the serpins?

What are the serpins?

• It is a family of proteins characterised by a common molecular architecture

• Most of the serpins are serine protease inhibitors, but some of them have other functions

• Today, more than 500 serpins have been identified in animals, plants, bacteria and viruses

• Serpin structure

• Inhibitory mechanism of serpins

• Serpin polymerisation

• inhibits the plasminogen activators, uPA and tPA

• regulates fibrinolysis (dissolving of blood clots) and cell migration

• the specificity and stability of PAI-1 is regulated by cofactors such as heparin and vitronectin

• lacks cysteine residues

(from Sharp et al. 1999)

Why PAI-1 spontaneously converts to latent form?

PAI-1 (plasminogen activator inhibitor type 1),the only serpin which spontaneously converts to the

latent form

Distance measurement using donor-donor energy migration (DDEM)

Time

Localisation of the RCL in PAI-1 by intramolecular distance measurements

Distances measured

P1’ - 313 P3 - 313

Distances in stable PAI-1 mutant (X-ray ) (Å)

Distances determinedby the DDEM method (Å)

6968

55 2 55 2

Conclusion: formation of disulfide bonds between the cysteines in RCL and cysteines in the A--sheet suggests that the RCL in active PAI-1 can be preinserted.

Complex

Oxidized

IntactCleaved

Preinsertion of the RCL studied by the ability to form intramolecular disulfide bonds

Conclusion

In contrast to other serpins, active PAI-1 has RCL located close to the core and preinserted. This may be a reason why PAI-1 spontaneously converts to latent form.

P. Hägglöf et al., J. Mol. Biol. 2003.

Inhibitory mechanism of serpins

What was known: •serpins form very stable/irreversible complexes with their target proteases• when the complexes were analysed by SDS-PAGE or amino acid sequencing, the serpins were cleaved

Major questions:• Are serpins cleaved in the native complexes or the cleavage is an artifact of the analyses?• How look the serpin/protease complex?

Quantification of free N-terminals in native serpin/protease complexes

PCF Result: in native serpin-protease complexes the N-terminus of PCF is blocked to the same extent as the other N-termini

Conclusion: in the native serpin/protease complex the reactive centre of serpin is cleaved and the protease covalently bound to the serpin

M. Wilczynska, et al., J. Biol. Chem. 1995.

What is the conformation of serpin/protease complex?

Hypothetical conformations of stable serpin/protease complex

Distance measurement in thePAI-1/uPA complex

Conclusion: the distance data exclude the “docking conformation” of the PAI-1/uPA complex but does not distinguish between full and partial-insertion models X

M. Wilczynska, et al., Nat. Struct. Biol. 1997.

Structural analysis of PAI-1/uPA complex by distance measurement and triangulation

Conclusion: the distances measured are compatible with full-insertion model

Full-insertion model

Model of the complex Distances (Å)

P3-266 P3-185 P3-P1’ P3-313

Partial-insertion model

Distances measured by DDEM

43,6 34,2 60,3 39,2

49,8 52,1 60,3 8,6

52 52 60 <30

M. Fa, et al., Struct. Fold. and Des. 2000.

Serpin inhibitory mechanism is driven by serpin metastability

Serpin inhibition involves reactive center cleavage and full loop insertion, so the covalently linked protease is translocated from the initial docking site to distal end of serpin.

Loop-sheet polymerisation of serpins

Wild-type serpins polymerise only under mild denaturing conditions.

Some of natural serpin mutants spontaneously polymerise in vivo. This results in diseases like cirrhosis and emphysemia (polymerisation of 1-antitrypsin), angioedema (polymerisation of C1-inhibitor), and dementia (polymerisation of neuroserpin).

The polymerisation is accompanied by loss of inhibitory activity.

Plasminogen activator inhibitor type 2, PAI-2, the only serpin which polymerises as wild-type protein

• PAI-2 exists as: * extracellular glycosylated form * intracellular non-glycosylated form• PAI-2 has the largest CD-loop in the serpin family

What are the molecular determinants of PAI-2 polymerisation?

Comparison between PAI-2 and 1-AT

Breachregion

Conclusion: the breach region does not determine the polymerisation ability of PAI-2

M. Wilczynska et al., Febs Lett. 2003

Polymerisation of native and DTT-reduced PAI-2

Conclusion: reduction of PAI-2 makes the protein resistant to polymerisation.

Non-denaturing PAGE

Native ReducedPAI-2 PAI-2

Identification of a cysteine which is important for polymerisation ability of PAI-2

Conclusion: Substitution of C79 or C161 to serine makes PAI-2 resistant to polymerisation.

Non-denaturing PAGE

Cysteines 79 and 161 form disulfide bond

Analysis of trypsin-degraded wt PAI-2 by Maldi-tof mass spectrometry

Polymerisation of PAI-2 mutant with two cysteines only (C79 and C161) under different redox conditions

Conclusions: The polymerogenic form PAI-2 is stabilised by the C79/C161 disulfide bond. The polymerogenic and stable monomeric forms of PAI-2 are interconvertible.?Oxidation

2

Polymerogenic form Stable monomerogenic form

Triangulation of the C79 in stable monomeric PAI-2 by intramolecular distance measurements using DDEM

Conclusion: Stable monomeric form of PAI-2 has the CD-loop folded on a side of the molecule

Is the translocation of CD-loop in PAI-2 linked to conformational changes in the A-β-sheet of the

inhibitor?

Conclusion: the A--sheet of PAI-2 is more open in the polymerogenic form than in the stable monomeric form of the inhibitor.

Annealing of synthetic RCL-peptide into wt PAI-2 and its mutants to compare the opening of the A--sheet

+R C Lp ep tid ePA I-2+u PAN o n -an n ealed

A n n ealed

C o m p lex

C leav ed =an n ealed

* S D S /PA G E* W estern b lo t* Q u an tificatio n o f cleav ed PA I-2 b y ph o sp h o rim ag er

W t P A I-2

C79S P A I-2C145S P A I-2C161S P A I-2N o n s e c ific p e p tid e

C5S P A I-2

0 25 50 75 1000

10

20

30

Peptide excess

Annealing [%]

Polymerogenic form of PAI-2

Stable monomeric form of PAI-2

50 Å

Conversion of PAI-2 from the polymerogenic form to the stable monomeric form is accompanied by closing of the sheet A and by translocation of the CD-loop from the bottom to the side of the molecule.

Reduction

Oxidation

Conclusion

Polymerisation of PAI-2 from the cytosol (wt PAI-2) and from the secretory pathway (SP-PAI-2) of CHO cells

Conclusions: • in the cytosol, PAI-2 exists mainly in the stable monomeric form• in secretory pathway, PAI-2 is in the polymerogenic form.

Do the polymerogenic and stable monomeric forms of PAI-2 exist in nature?

Interaction of PAI-2 with vitronectin

Conclusion: PAI-2 can form disulfide-bond to vitronectin via the C79 in CD-loop

Conclusions• PAI-2 is a unique serpin with two mobile loops: the RCL and the CD-loop• The CD-loop of PAI-2 is a redox-sensitive molecular switch that regulates conversion between the polymerogenic and the stable monomeric

forms of PAI-2.

• Polymerisation of PAI-2 in vivo may be regulated by redox status of the cell. • Disulfide-binding of vitronectin to the C79 in the CD-loop of PAI-2 may stabilise the active PAI-2 in extracellular compartments.

Polymerogenic form of PAI-2

Stable monomeric form of PAI-2

50 Å

Reduction

Oxidationpolymerisation

M. Wilczynska et al., EMBO J. 2003;S. Lobov et al., J. Mol. Biol. 2004.