Mass Spectrometry - an essential tool to understand and produce - - PowerPoint PPT Presentation

Mass Spectrometry

• an essential tool to

understand and produce proteins

Mark Abbott CEO Peak Proteins Ltd

Talk Outline

1) Brief introduction to Peak Proteins Ltd. 2) How we use mass spectrometry to solve problems and analyse the proteins we make 3) Examples of the use of mass spectrometry

Protein Structures; Art and Science!

G6b-B is dimerised around a ligand and bound to Fab fragments – more later

Protein production is easy!

QC and storage

• The right construct, cell, culture

conditions, purification method, analytical methods and storage conditions appropriate for end use.

• What does “right” mean?
• There is no one right way.
• Every protein is unique and requires

handling differently including for different end uses.

• How do you know/find the right

methods?

Peak Proteins

Proteins are very individual macromolecules

We aim to help you understand them better and make them work for you.

• Set up by ex-pharma employees, offer decades
• f expertise in protein reagent supply and

protein structure determination.

• Based at Alderley Park near Manchester.
• Understand the drug discovery process and the

need for high quality bespoke proteins and structural data.

• Proven ability to work with many protein classes

in small molecule and biologics projects.

• Our research-based, innovative approach to

solving protein requirements differentiates us from ‘off-the-shelf’ protein suppliers.

Protein mass spectrometry at Peak Proteins

• SCIEX X500B mass spectrometer
• SCIEX Exion LC
• SCIEX OS. BioToolKit and

BioPharmaView software

• Intact mass and peptide mapping
• Seamless transition between

methods, C4 and C18 column in

• ne oven so intact and peptide

mapping can be queued together

What do we get from intact mass analysis?

• Simple rpHPLC , ESI-MS analysis
• Is the Intact mass as expected?
• Are there post-translational

modifications?

• Glycosylation, disulphide bonds,

phosphorylation, proteolytic processing

• Has the product been

inadvertently “clipped”?

• Are there unexpected

modifications?

• Are there expected modifications

(biotinylation)?

Peptide mapping - process

• Band on SDS-PAGE gel
• Reduce, alkylate and digest
• Trypsin, chymotrypsin (Glu-

C)

• Separate by rpHPLC
• ESI-MSMS
• Peptide masses and MSMS

sequencing using CID

• BioPharmaView software to

search data against bespoke sequences

• Search data using Mascot

against UniProt

?

What do we get from peptide mapping?

• Particularly useful when the intact protein won’t fly on ESI-

MS or heterogeneity like N-glycosylation prevents clear identification or if the protein is very impure

• Confirmation of identity
• Post translational modifications
• Mutations
• Sequence confirmation
• Degradation
• Contaminant identification

Examples of how we use mass spectrometry

• Identification of products in the manufacture of a therapeutic protein
• Trouble shooting a purification
• Mapping post translational modifications to enable the engineering of a protein

suitable for crystallisation

Manufacture of a therapeutic protein

• Cytokine with 2 disulphide bonds
• Manufacturing process is;
• Express as inclusion bodies in

E.coli

• Solubilise in urea/reductant
• Refold and form disulphide bonds
• Further purification
• Need to maximise yield of correct

product in initial refold

• SDS-PAGE gel shows different

conditions

• Analyse via intact mass, peptide

mapping of disulphide bonds.

Intact mass – 19394 (NR); 19398 (Red.) – 2 disulphides Peptide mapping – 1 disulphide detected in D, work still ongoing as needs GluC

A B C D

Ladder Load 10 11 12 13 14 15 16 kDa 40 51 62 87 110 200 140 136 30 22 16 7 17 18 19

Trouble shooting a purification

A B C

Peptide mapping A – kinase X B – kinase X C – E.coli CRP (host protein) Intact mass 38190 – full length protein; 38189.5 measured 35612 – delete N-t 23 aa; 35611.5 measured

Recent client project – G6b-B

• Prof. Yotis Senis (University of Birmingham) – group studies regulation of platelets.
• Requested the X-ray structure of the extracellular domain of the Megakaryocyte

and platelet inhibitory receptor G6b (G6b-B).

• In complex with the Fab fragment of a potential therapeutic monoclonal in
• rder to help identify epitope for patent application.
• In complex with ligand to visualise & better understand binding/activation

mechanism.

Platelet function

• Platelets are highly reactive

anucleated cell fragments.

• Produced by megakaryocytes (MK’s) in

bone marrow, spleen & lungs

• On vascular injury platelets adhere to

exposed vascular extracellular matrix and become activated to form hemostatic plug & seal wound.

• Must be tightly regulated to avoid

hyper-reactivity and indiscriminate blockage e.g. acute coronary heart disease and stroke.

• Inhibition partly due to receptors

containing immunoreceptor tyrosine- based inhibition motifs (ITIM’s) e.g. G6b-B

SH2 P T P SH2 active Shp1 Shp2 PTP inactive SH2 SH2 Shp1 Shp2

G6b-B

ITIM ITSM

SFK

P P P

IgV

Senis et al. Mol Cell Prot 2007

The inhibitory ITIM receptor G6b-B

• G6b-B – an ITIM containing receptor highly

expressed in MK & platelets.

• KO mice have grossly distorted platelet

function – macrothrombocytopenia

• Binds heparin/heparan containing

saccharide ligands

• Type I transmembrane protein consists of

single IgV-like ECD, a transmembrane domain and cytoplasmic tail with ITIM and ITSM motifs.

• Upon ligand binding central tyrosines of

ITIM/ITSM are phosphorylated by Src family kinases to become docking site for phosphatases Shp1 & 2.

• Positions active Shp1/2 to dephosphorylate

key components of ITAM signaling pathway & attenuate activation signaling.

G6b-B ECD expression and purification

Derek Ogg CSO/Crystallographer Juli Warwicker Protein Scientist

signal extracellular membrane ITIMS Several ECD constructs generated

• Extracellular domain is single IgV-like domain of ~13kDa
• No published X-ray structure & has < 20% homology with IgV family structures in PDB.
• One potential N-linked glycosylation site (Asn32).
• 4 cysteines, at least one disulphide by homology.
• A number of G6b-B ECD constructs were expressed transiently in HEK293 cells.
• ECD construct encompassing residues 18-133 expressed well.
• Purification by cation exchange and size exclusion from culture medium.

N-term C-term

1 241 142 163 18

G6b-ECD: Initial purification

• Initial SDS-PAGE and LC-MS identified

protein consisted of 2 species

• Upper band with multiple masses

between 14-15kDa indicating N- glycosylation at the predicted site Asn32.

• Native G6b-B ECD protein did not

crystallise.

• Need to remove N-glycosylation
• The N-linked sugars could be partly

cleaved with PNGaseF - but difficult to get removal to go to completion.

• Therefore generated Asn32->Asp

mutant.

SP-Seph S75 SEC S75 chromatogram

Engineering out the glycosylation

• N32->D mutant eliminates upper band.
• G6b-ECD now appears as single species on SDS-

PAGE & LC-MS.

• Intact Mass LC-MS data from a Sciex X500B mass

spectrometer gives mass of N32->N G6b-ECD at 13410.2Da.

• This is +948Da from the predicted mass & consistent

with addition of a single common O-linked tetrasaccharide structure:

S75

NeuNAc Gal NeuNAc GalNAc Ser/Thr

Crystallisation of N32->D mutant

• No crystals were obtained of N32->D G6b ECD mutant

alone or in presence of DP12 (dodecasaccharide heparin fragment)

• Crystals of Fab-G6b-DP12 complex were obtained but

grew very slowly (2-3 months) and only diffracted to ≤4.0Å at Diamond Light Source

• At this resolution could clearly see the Fab and some

electron density near the CDRs for putatively bound G6b-B but unable to build model

• Improve resolution by also removing the O-

glycosylation?

• Considered sialidase and O-glycosidase but opted

against for cost reasons

Initial Fab-G6b ECD-DP12 crystals

O-glycosylation

• 13 Ser and 5 Thr residues in G6b-B ECD construct any of which in principle could be

O-glycosylated

• Bioinformatics with NETOGlyc 4.0 on UniProt identifies 4 residues with a “positive”

score

• All 4 are found close together in a predicted loop region containing 3 Ser & 2 Thr

residues

• LC-MSMS peptide mapping via chymotrypsin digest and analysis on Sciex X500B

instrument identified a 15aa peptide of this loop with + 948Da mass:

66 80

ASSSGTPTVPPLQPF

• Consistent with this loop being the site of O-glycosylation - but which residue?

LC-MSMS peptide mapping via chymotrypsin digest

• TOF-MS of digest shows peak at 11.5mins,

confirming O-glycosylation with 3+ ion ASSSGTPTVPPLQPF

• In source fragmentation to give non-

glycosylated 2+ ion and free

• ligosaccharide
• In source fragmentation confirms

individual saccharides

• Which Ser or Thr?

HexNAc NeuAc Hex NeuAc

NeuNAc Gal NeuNAc GalNAc Ser/Thr

Mass spectrometry analysis of G6b-B mutants

Mutation (All have N32D) Predicted mass +O-glycosylation Observed mass S67A 13398 13394 S68A 13398 13394 S69A 13398 13394 T71A 13384 13380 T73A 13384 1=13380 2=12432 4M(AAAAT) 13336 13332 5M(AAAAA) 13306 12354

1 2 Cation exchange purification of T73A mutant

• Suggests that O-glycosylation on Thr73 is preferred

site but can also occur elsewhere on loop

• This heterogeneity may hinder ordered crystal

formation

G6b-B crystals - 5M & 4M

• Both 5M & 4M mutants were screened for

crystallization with and without Fab & DP12 (heparin fragment)

• Crystals of 5M alone (no DP12) were obtained but

diffracted only to 10Å resolution

• 4M G6b-B alone however crystallised within 2

weeks and diffracted to 2Å

• 4M G6b-B in complex with Fab + DP12 also

crystallised in a similar timescale and diffracted to 3.0Å

• Allowed G6b ECD-Fab-DP12 complex structure to

be solved.

Vis UV

G6b(4M) + Fab G6b(4M)

G6b-B ECD-Fab-DP12 - 3.0A X-ray structure

• 3.0 Ång data collected at Diamond Light Source.
• Crystal structure solved by Molecular Replacement using a Fab model.
• Structure reveals a dimer of two G6b-B ECD-Fab complexes
• Deposited in PDB (6R0X)

G6b-B ECD epitope identified

• X-ray structure revealed that

the Fab epitope largely formed by N-terminal strand of the G6b-B ECD

• All CDR regions except CDR 2 of

VL chain involved in binding interactions

• Particularly important

interaction is formed by salt bridge between Asp24 of G6b- B and Arg69 in VH (CDR2)

VL VH

G6b-B ECD

G6b-B ECD-DP12 interaction

• The G6b-B ECD dimer has heparin chain

(DP12) bound tightly in groove formed at dimer interface

• Spatially separated from Fab binding site
• Anti-parallel arrangement of 2 Ig-like

domains is unique among known heparin/HS binding structures

• Electron density for only 8 of the 12

saccharide units of DP12 can be observed in structure

• Consistent with binding data that at least 8

heparin units need for high affinity binding

• Also ~90x higher heparin affinity for G6b-

B dimer over monomer constructs

Heparin-mediated G6b-B dimerization

• Heparin mediated G6b-ECD dimerization is supported by

size exclusion chromatography Elution volume (ml) Absorbance (280 nm)

20 – 30 – 40 – 60 – 5 10 15 20 25 0 – 10 – 50 –

G6b-B ECD Ribonuclease A (13.7 kDa) Carbonic anhydrase (29 kDa) G6b-B ECD + heparin (DP12)

NPGASLDGRPGDRVDLSCGGVSHPIRWVWAPSFPACKGLSKGRRPILWAAAAGAPTVP PLQPFVGRLRSLDSGIRRLELLLSAGDSGTFFCKGRHEDESRTVLHVLGDRTYCKAPG Sequence Charge WVWAPSFPACK TYCK 3 WVWAPSFPACK TYCK 2 VDLSCGGVSHPIR LELLLSAGDSGTFFCK 5 VDLSCGGVSHPIR RLELLLSAGDSGTFFCK 5

• Disulphide bonding pattern could not be

unambiguously defined from structure

• Tryptic peptide mapping
• A – Suggested pattern
• B – Tryptic map under reducing and non-

reducing conditions

• C – Full MS/MS analysis without

reduction and table of linked peptides

Peptide mapping of disulphide bonds in G6b

A B C

Impact of mass spectrometry on G6b crystallography

• The “light shone” by mass spectrometry has been essential to understand the

post-translational modifications of G6b

• This has enabled the engineering of a construct much more suitable for

crystallisation

• Mass spectrometry has also confirmed both disulphides, one of which was not

visible in the structure

Acknowledgements (G6b work)

Helen McMiken Rachel Rowlinson Juli Warwicker Catherine Geh Derek Ogg Tina Howard Mark Abbott

• Prof. Yotis Senis

Timo Vögle

Thank you!