Transcript West Nile Virus - Transfusion medicine

Recombinant Proteins:
What’s New in the Lab and the Clinic
Transfusion Medicine Residents
2008-09-25
Bill Sheffield
PubMed Search 2008-09-19
• Recombinant AND protein AND 2008
• >17,000 articles
• Add “AND review” …
Zhang AL, Luo JX, Zhang TY, Pan YW, Tan YH, Fu CY, Tu FZ.
Recent advances on the GAP promoter derived expression system of Pichia pastoris.
Mol Biol Rep. 2008 Sep 10. [Epub ahead of print]
PMID: 18781398 [PubMed - as supplied by publisher]
• Near-ubiquitous use in research, increasing
number of approved drugs
What’s a recombinant protein?
• A protein made using
recombinant DNA
• Recombinant DNA: DNA
molecules in which
sequences are assembled
in an order not found in
nature, usually for the
purposes of facilitating
protein production
Objectives
To explore general issues common to research
and pharmaceutical recombinant protein
production
To discuss (a) the recombinant protein drugs
currently distributed by CBS and (b) ongoing
development of these products
To present published and unpublished research
from the Sheffield laboratory relating to
recombinant plasma proteins
Molecular biology truths
• DNA makes RNA makes
protein
• Protein production from a
DNA template = protein
expression
• Most widely used means
to transfer DNA = plasmid
• Numerous expression
systems available
Source: www.artes-biotechnology.com
Plasmids
•
•
•
Small, double-stranded,
circular DNA elements
Usually propagated in bacteria
Used for recombinant protein
production in bacteria or
transferred to another host
(yeast, cultured mammalian or
insect cells)
Source: Becker WM, Kleinsmith LJ, Hardin, J. The World of the Cell, 4th
edition, 2000. Benjamin/Cummings Pub. Co., Don Mills ON.
Why make recombinant proteins?
• From a research point of
view…
• Test ideas about protein
structure and function by
making single amino acid
substitutions (or larger
changes)
• Has become increasingly
“easy” and “inexpensive”
• In some cases the natural
protein is not easy to obtain
in sufficient quantities
Why make recombinant proteins?
• From a clinical point of view…
• Safety (e.g. vs. purifying FVIII from pooled
plasma of 1000s of donors)
• Natural human material not available due
to abundance or source (e.g. G-CSF
[filgastrim/Neupogen], erythropoietin
[Epotin], interferons, fVIIa [Niastase],
tissue plasminogen activator [Activase])
• Cost?
Production choices…because
expression is “empirical”
• Choice of DNA (cDNA, genomic clone,
codon usage)
• Choice of vector/plasmid (promoter
strength, inducibility)
• Choice of expression system
– Post-translational modifications
– Purification strategies (e.g. tags)
– Efficacy/immunogenicity considerations
Expression systems
• Bacteria (usu. E. coli, others possible)
– Cheap but can mis-fold, do not glycosylate, inclusion bodies
• Yeast (Saccharomyces, Pichia, others)
– Fold well but acidify media, glycosylation may be high mannose
• Insect Cells (baculovirus system)
– Yields can be very high but glycans not well characterized
• Cultured Mammalian cells
–
–
–
–
Chinese Hamster Ovary Cells CHO (FDA happy)
Baby Hamster Kidney BHK (FDA happy)
Human cell lines (HEK 293 gaining acceptance)
Tumour cell lines (research use only)
• Weirder stuff
– Slime molds
– Plants
– Transgenic animals (e.g. goats expressing human proteins in their milk)
A word about protein folding
• Proteins: linear
chains of amino
acids
• Often shown in bar
graph form
• Actually must have
appropriate 3D
structure to
function, avoid
clearance, avoid
immunogenicity (?)
RSA
HLAH 6
HV3
RSA
His6
Mammalian cell expression:
Pluses and Minuses
• Pluses: Chaperonemediated folding and
“post-translational”
modifications (N- and Olinked glycans, S-S
bonds, gamma
carboxylation (VIIa, IX))
• Minuses: Quality control
mechanisms lower
yields
Annu. Rev. Cell Dev. Biol. 2000. 16:557-589
Recombinant CBS products
• Factor VIII (Kogenate [FS], Advate,
Recombinate)
• Factor IX (Benefix)
• Factor VIIa (Niastase)
• Palivizumab (Synagis)
– Humanized monoclonal antibody indicated for the
prevention of respiratory syncytial virus RSV LRT
infections in at-risk infants
– Produced in hybridoma cell line
– Some data suggests superiority over RSV-IVIG
Recombinant CBS coagulation proteins
TF-FVIIa
FX
FXa
FIXa
FVIIIa
TF-FVIIa
IIa
FXIa
FIX
FVIII
PROTHROMBIN
IIa
FXa FVa
IIa
IIa
FXI
FIBRINOGEN
FXIII
FV
IIa
FIBRIN
FXIIIa
CROSS-LINKED FIBRIN
FVIII
Huge (2332 amino acids, 165 – 280 kDa)
B domain is heavily glycosylated and dispensable
FVIII circulates as a heterodimer; HC varies in length
A1
A2
B
A3
C1
C2
Processed during synthesis
A1
A2
B
2
A3
C1
C2
Light Chain
Variable Heavy Chain
Recombinant FVIII products
• FVIII circulates complexed to von
Willebrand factor
• 50: 1 vWF: fVIII
A1
• Unstable without vWF
2
• Two recombinant expression
A3
strategies
•Co-express VIII & vWF in
CHO cells, remove vWF
before formulation (Genetics
Institute/Baxter)
•Express fVIII alone (Bayer)
• Initial products required
stabilization with albumin (HSA)
when formulated
A2
C1
B
C2
VWF
FVIII: New developments
• Alterations to reduce or eliminate exposure to HSA, FBS or other
human plasma proteins in culture or formulation (Kogenate FS,
Advate)
• Recombinate is still formulated with HSA
• Refacto (B-domain deleted rFVIII)
• Xyntha (Refacto made without HPP, purified using synthetic affinity
ligand, not murine MAb)
• In clinical development: PEGLip-FVIII-FS (Bayer), Kogenate FS
formulated with pegylated liposomes to increase biological half-life
(T&H Sept’08 16 patients)
• Pre-clinical: Baxter FVIII modified with “releasable” PEG conjugates
• Research: Mutant rFVIII molecules with mutated clearance motifs,
stabilized (S-S or fusion protein) A1-A3 linkages
Factor IX
• A vitamin K-dependent clotting factor
• 55.4 kDa plasma protein, 5g/mL (90 nM)
• Complex - 415 aa, N-, O-glycans, sulphotyrosines, -OH aspartate, g-COOH**
• Expressed in CHO cells modified to overexpress
processing peptidase that removes pro sequence
1
415
FIX
1
145 181
415
FIXa
S
S
FIXa
FVIIIa
FX
Recombinant factor IX and
potential development
• BeneFIX
• Only recombinant fIX product on market
• Product enhancements (2007) allowed more rapid,
needleless reconstitution
• Baxter (early 2008) announced intention to market rfIX,
develop pegylated rfIX with Nektar Therapeutics
• Pre-clinical work suggests that activity can be enhanced
(3X) by point mutation
• Clearance slowed by K5A mutation (Stafford, UNC) or
activation peptide deletion (Sheffield, CBS) (?)
• Blood, July 1, 2002 issue
• K5A, K5R, WT HufIX
made in 293 cells
• Injected into fIX ko mice to
restore physiological [IX]
• K5R 79% in liver in 2
minutes; K5A 59%
• K5A initially cleared
slower, later faster
Recombinant VIIa
• Produced by Novo Nordisk in BHK cells
• 406 amino acids; rfVII activates easily (autoactivation) by
cleavage after Arg152 (no AP)
• Activation encouraged by processing procedure to make
NovoSeven (Niastase)
• FVII also vitamin K dependent, some structural
similarities to fIX
New developments in FVIIa
• NovoSevenRT (Ready to Travel) heat-stable formulation
• NovoNordisk has “short-acting” and “long-acting” VIIa in clinical
development pipeline
• MaxyVII (multiple mutations to increase activity, decrease clearance
from circulation)
• NN1731 (V158D/E296V/M298Q) in clinical development
• Trial of NN1731 vs Niastase in hemophilia A patients with inhibitors
being treated for acute joint bleeds commenced recruitment in June
2007 (2 years)
• CSL Behring pursuing VIIa-FP, fusion protein to HSA, in preclinical
studies (Weimer T et al Thromb Haemostas April 2008; WHC)
An approach to watch (10 years+)
• Production of recombinant proteins by
genetically transformed mammalian
cells in culture very $$$
• Production by transgenic livestock
potentially less expensive, higher
capacity
• GTC Biotherapeutics (Framingham,
MA) produces recombinant
antithrombin (ATryn) in the milk of
transgenic goats
• ATryn approved in Europe, phase III in
USA – but indications for this protein
are limited
• GTC continues to pursue rfIX
production via this route and may be
able to avoid issues that caused
AmCross to drop this approach
Sheffield lab recombinant proteins
• Prelude:
• Published:
• Submitted:
Our simplest recombinant protein
project
Transforming alpha-1-proteinase
inhibitor into a thrombin inhibitor
The Trojan rabbit: Transforming
albumin into a cross-linked
component of a thrombus
Prelude: GST fusion proteins
• Schistosomal glutathione
sulfotransferase (GST) is
well expressed in E. coli
• Fused mouse fIX cDNA in
frame (made one mini-gene
with no stop codons between
GST and fIX DNA)
• Purified on glutathioneSepharose column
• Injected chickens
• Harvested antibodies from
eggs
• Adsorbed vs. GST beads
• Obtained specific polyclonal
antibody
Published: Serpin work
•
•
•
•
Serpins are a protein family
Are SERine Protease Inhibitors
Many serpins are plasma proteins
Antithrombin, alpha-1-proteinase inhibitor
(aka alpha-1-antitrypsin), heparin cofactor
II, PAI-1, alpha-2-antiplasmin, C1-esterase
inhibitor)…
• My lab has attempted protein engineering
to alter the specificity of serpins.
1-Proteinase Inhibitor (1-PI)
• 20 M in plasma
• 394 amino acids (52 kDa)
• Anti-neutrophil elastase
k2 = 4 X 109 M-1min-1
• Poor inhibitor of thrombin
k2 = 3 X 103 M-1min-1
http://www.expasy.ch/ch2dgifs/PLASMA_HUMAN/PLASMA_HUMAN.gif
• 1-PI M358R
• Good inhibitor of thrombin
k2 ~ 1 X 107 M-1min-1
• Antithrombotic drug candidate?
• Good inhibitor of Activated Protein C
k2 ~ 3 X 106 M-1min-1
• How to maximize anti-thrombin activity?
Huntington JA et al. Nature 2000
Harnessing the HCII “tail”
Baglin TP et al. PNAS 2002
• HCII 1-75 binds thrombin exosite 1
• Needs activation by GAGs
• We grafted HCII 1-75 to 1-PI M358R
394
1
1
Acidic
tail
- - 75
--
*
API (M358R)
*
HAPI (M358R)
HCII
1
480
k2 (M-1min-1)
Thrombin
APC
Selectivity
Ratio
1-PI M358R
1.0  0.5  107
1.4  0.1 106
7.9
HAPI M358R
2.30  22  108
1.8  0.1 106
128
Merging extra- and intra-loop approaches
1
Acidic
tail
- - 75
*
HAPI (M358R)
Filion ML et al.
Biochemistry 2004
EGTQATTVTTVGFMPRSTQPE
API / HAPI RCL 4
Hopkins PC et al.
J Mol Evol 2000
KGTEAAGAMAVVIAPRSLPPE
API / HAPI RCL 5
• Would making the RCL more HCII-like or more
AT-like increase selectivity for thrombin over APC?
Kinetic results
-thrombin
APC
-thrombin
/APC
k2
( 106
M-1min-1)
SI
k2
( 106
M-1min-1)
1-PI
M358R
11  0.5
3.2  0.3
1.4  0.1
7.9
HAPI
M358R
230  22b
2.9  0.1
1.8  0.1
128
HAPI
RCL4
22  4
26  1.6
0.01  0.002
2,200
HAPI
RCL5
140  5
2.1  0.2
0.013  0.002
10,769
TCT
• Prothrombindepleted plasma
+ thrombin +
inhibitor
• Measured time
to clot
• HAPI RCL5
more effective
than HAPI RCL4
In vivo stability in mice
Results were published (Sutherland et al.) in Thromb Haemostas November 2007)
Submitted: The Trojan “Bunny”
•
•
•
•
Albumin, the most abundant plasma protein
Not glycosylated but has many disulphide bonds
Pichia pastoris yeast best system for expression
Pp HSA shown to be biochemically identical to
HAS but scale defeated recombinant expression
except in Japan (Bipha Corp product approved
May 2007)
• We use the same system to make recombinant
albumin and albumin fusion proteins.
Albumin
• Most abundant plasma
protein protein (35-50
mg/ml)
• 584 amino acids, 67 kDa
• Slowly cleared (t1/2 = 4-
5 days in rabbits, 18
days in humans)
• Non-glycosylated
Genetic fusion to albumin
• Concept: Increase molecular volume and/or
negative charge
• Mechanism: Reduction in glomerular filtration,
other potential size-related effects, possible
stabilization, masking of labile bonds
RSA
HLAH 6
HV3
RSA
His6
• Geometry: Nterminal HV3, then
GLY6, then RSA,
then HIS6
Production of albumin fusion
proteins in yeast (P. pastoris)
• Yeast are
simple
eukaryotes
• Fold
complex
proteins
better than
E. coli
Interfering with blood clots
• Rationale: Speed clot lysis
of thrombi
• Alpha-2-antiplasmin is
cross-linked to fibrin by
fXIIIa & inhibits plasmin
• Concept: Fool the clot into
incorporating a modified
albumin that cannot inhibit
plasmin (kind of a Trojan
horse), make the clot
easier to dissolve
• We fused portions of
alpha-2-AP that
cannot inhibit
plasmin to HSA
• Only one, A2AP(1342)-HSA, was wellexpressed
• Amino acid
sequencing showed
that it was a mixture
of intact and cleaved
termini
• A2AP(13-42)HSA, became a
substrate for
fXIIIa, like native
A2AP
• Plasma clotted with
thrombin in the presence of
tPA lysed rapidly (3)
• Clot lysis was delayed by
inclusion of A2AP(13-42)
fusion protein (FP)
• Submitted to BMC
Biotechnology August 19,
2008
• Under review
• Not quite a Trojan Horse
(yet) but shows
promise…more like a Trojan
rabbit
Objectives achieved?
 General issues common to recombinant proteins on an
industrial and research scale
 CBS’ 4 recombinant drugs (FVIII, FIX, FVIIa, Synagis)
and the predicted next new product (PegLip FVIII)
 GST-mfIX, HAPI RCL5, and A2AP(13-42)-HSA, 3
examples of recombinant protein expression from my lab
 Acknowledgements to my lab
 Current members: Varsha Bhakta, Sharon Gataiance,
Louise Eltringham-Smith,
 Graduates Jason Sutherland (PhD), Marc Filion (MSc)