insect cell technology as a vaccine- producing platform · paula m alves insect cell technology as...

Paula M Alves

Insect cell technology as a vaccine-producing platform

VLPs as vaccines, vectors and adjuvants Les Pensières,

Fondation Mérieux Conference Center March 2014

Instituto de Biologia Experimental e Tecnológica Private not-for-profit Research Institute

Main

Activities

Health-

Pharma

Agro-

Industry

Almeirim, 29 Janeiro 2013

in Vitro Models

for pre-clinical

Research

Cell line

Development and

“omics” tools

Cell

Therapy

Pilot Plant iBET Infrastructure

—Bio-Process Development (e.g. Vaccines)

—Supply of purified proteins for structural studies in different biological systems (bacterial, yeasts, mammalian cells, insect cells / baculovirus)

—Supply of Pre-clinical grade Biopharmaceutical Products products up to gram scale (vaccines, Mabs, VLPs)

—Supply of Phase I/II Clinical grade material produced under cGMP can also be provided by collaboration with our spin-off GeniBET

—Supercritical Fluid Extraction

Analytical Services Unit iBET Infrastructure

Support to the implementation of quality systems

Quality Control and lot release of drug substance and drug product and experimental new drugs under GMP

– Including chromatography, mass spectrometry and cell based assays

Development, optimization and validation of analytical methods

– Chemical, biological, molecular biology and cell based assays

– Post development analytical method transfer

Support in biopharmaceutical development

– Protein identification by mass spectrometry including, intact mass, peptide mass fingerprinting, confirmation of the glycosylation pattern

– N-terminal sequencing by the Edman-reaction

– Detection of chemical, microbial, viral and Mycoplasma contamination

– Analysis of cell and viral banks

Almeirim, 29 Janeiro 2013

www.genibet.com

A cGMP Bridge to Personalized Medicine

Facilities

GMP – lots for hase I/II

Virus Unit

QC Labs

Cell Culture Fill and Finish

Bacterial Unit

http://www.genibet.com/

Number of chains deposited in the PDB, produced in mammalian cells and BEVS (March 2011). http://www.rcsb.org/pdb

Protein Expression Systems Entries in Protein Data Bank(PDB) from eukaryotic hosts

No

. Ch

ain

s

The Baculovirus Expression Vector System (BEVS)

Vijayachandran et al 2011

Powerful tool for the expression of proteins in short time frames

Widely used in drug discovery research target validation, Biochemical assays (high-throughput screening, …) structural biology campaigns


Vijayachandran et al 2011

Powerful tool for the expression of proteins in short time frames

Multi-protein complexes

Rasmussen et al 2011 Nature 477, 549–555

Production of complex proteins:

Membrane proteins (important drug targets)

Widely used in drug discovery research target validation, Biochemical assays (high-throughput screening, …) structural biology campaigns

Insect cells Recombinant Baculoviruses

50 m 300 nm 50 m

Infected cells

Co-infection with monocistronic rBac Single-infection with multicistronic rBac

Key advantages

1) Strong promoters (polh and p10) - good protein expression levels;

2) Similar protein folding and PTMs to mammalian cells;

3) The baculovirus genome (80-230 kb) allows expression of large transgenes and multiple genes.

Multi-genic products:

Infection of insect cells (Sf9 or Hi5 cells) with recombinant baculoviruses (Autographa californica multiple nucleopolyhedrovirus, AcMNPV)


Virus Recombinant Proteins Size (nm) Product Status

Adeno-associated virus VP1, VP2, VP3 20-25 Development

Bluetongue virus VP2, VP3, VP5, VP7 69 Preclinical

Ebola virus VP40 and glycoprotein 65-70 Preclinical

Enterovirus 71 VP0, VP1, VP2 25-27 Preclinical

H1N1 Influenza virus HA, NA, M1 80-120 Phase IIa (Novavax)

H3N2 Influenza virus HA, NA, M1 80-120 Phase IIa (Novavax)

H5N1 Influenza virus HA, NA, M1 80-120 Phase I/IIa (Novavax)

Hepatitis C virus Core, E1 and E2 proteins 40-60 Preclinical

Human immunod virus Pr55gag and env proteins 100-120 Preclinical

Human papiloma virus L1 capsid protein 30-50 Cervarix (GSK)

Infectious bursal disease virus VP2, VP3, VP4 60-65 Development

Marburg virus VP40 and glycoprotein 50-100 Preclinical

Norwalk virus Capsid protein 38 Phase I (LigoCyte Pharma)

Poliovirus VP0, VP1, VP3 27 Development

Porcine parvovirus VP2 22 Preclinical

Rotavirus VP2, VP6, VP7 70-80 Preclinical

Simian immunod virus Pr56gag and env proteins 130 Preclinical

Baculovirus infection has been widely used for production of VLPs in insect cells

... especially multi-protein VLPs

BEV

S-d

eri

ved

VLP

s Virus-like particles

Market approval: 2010

Market approval: 2013 Market approval: 2007 in EU

Vaccines produced using BEVS

Human papilloma virus vaccine

Autologous cellular immunotherapy – prostate cancer

Influenza vaccine

Short production time required

Influenza VLP, Phase II

Alternative to the Baculovirus-insect cell system

Sf9 cell platform

Advantages:

Growth to high cell densities Eukariotic post-translational modifications Very efficient metabolism No lactate or ammonia accumulation

BEVS

Drawbacks:

Lytic infection, proteases Very late promoters – protein processing machinery compromised Need to separate viral and rec proteins Effort to maintain the virus stock rec virus unstable – problem scaling up

0

1

2

3

4

5

6

7

8

9

10

0 48 96 144 192 240 288

Gln

, Am

m (

mM

)

Age (h)

Nitrogen

source

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0

5

10

15

20

25

0 48 96 144 192 240 288

Lac

(mM

)

Ala

(m

M)

Age (h)

By-product

formation

0

10

20

30

40

50

0

2

4

6

8

10

12

0 48 96 144 192 240 288

Glc

(m

M)

Ce

ll d

en

sity

(1

06/m

l)

Age (h)

cells

Glc Gln

Amm

Ala

Lac

Recombinase-Mediated Cassette Exchange (RMCE) System

2 Steps of RMCE:

1) To tag and select for a high expression locus

2) To target the locus of interest

Flipase Recombination

5’- GAAGTTCCTATTC – TCTAGAAA – GTATAGGAACTT-3’

3’- CTTCAAGGATAAG– AGATCTTT – CATATCCTTGAAG -5’

Flp recognition target (FRT): spacer 13 bp repeat 13 bp repeat

Binding sites Binding sites Homology required

Establishing a Flexible Sf9 cell line: Promoter Selection

Which promoters?

OpIE1 OpIE2 Hsp70 Actin Metallotionin

Drosophila or baculovirus specific promoters

OpIE1

OpIE2

Hsp70

Mtn

Actin

eGFP

eGFP

eGFP

eGFP

eGFP

compared in terms of GFP expression Higher expression from OpIE2

Tagging and targeting cassette design

Recombination triggers activation of the neo resistant gene

eGFP Fw F5

Target ATG

Δneo dsRed

Fw

Hygro Tagging

F5

OpIE2

Flippase

OpIE2 OpIE1

OpIE2 OpIE1

OpIE1 has lower expression, but suitable for antibiotic resistance expression Fernandes et al. Biotechnol Bioeng 2012, 109, 2836-2844

Establishing a Flexible Sf9 cell line: Clone screening

Southern blot analysis

- Identification of clones with single-copy integration

• Single copy integration: one band of variable size, as restriction enzymes (EcoRI/SacI) cut only once inside the tagging cassette

All analyzed clones have single copy integration of the tagging cassette

#1

SacI EcoRI SacI+

EcoRI

#5

SacI+

EcoRI

Sf-9

SacI+

EcoRI

Clone

0

10

20

30

40

50

#1 #2 #3 #4 #5 #6 #7 #8 #9 #10 #11 #12 #13 #14

Mean

Flu

ore

scen

ce I

nte

nsit

y

Clones

Lipotransfection

Electroporation

**

*

*

*

*

*

*

Fernandes et al. Biotechnol Bioeng 2012, 109, 2836-2844

Integrated Tagging cassette

+ Flipase

Targeting plasmid:

eGFP Fw F5

ATG OpIE2 OpIE1

STEP 1 – Tagging = identification stable

high expression locus

•Selection w/ Hyg

•Screening w/ reporter gene

•Screening single copy integration

STEP 2 – Targetting = reuse the same

locus bt RMCE

•Selection w/ neomycin

•Predicatble expression of the

gene of interest (GOI)

neomycin resistance

gene inactive

Δneo dsRed Fw

Hygro

F5

OpIE2 OpIE2 OpIE1

Δneo Fw F5

OpIE2 OpIE2 OpIE1 eGFP ATG

neomycin resistance

gene active

Integrated Target cassette

Establishing a Flexible Sf9 cell line: Cassette exchange

Lanes 1, 3, 5 – Expression before cassette exchange Lanes 2, 4, 6 – Upon cassette exchange and 3 weeks in selection

mRNA levels of dsRed decreased and GFP increased

Establishing a Flexible Sf9 cell line: Cassette exchange

Successful flp-mediated recombination in Sf-9 cells

dsRed EGFP 18S

Clone #2

PCR of genomic DNA confirms cassette exchange

1600 bp

600 bp

Less 1000 bp confirms lack of hygromycin gene in tagged loci

- The same primers were used: OpIE1 promoter and neo marker.

Tagging

DNA Target

DNA

#2 #3 Clone

+flipase


Establishing a Flexible Sf9 cell line: Stability

Tagging clone #2

DsRed expression is stable along passages, without

selective pressure

0

100

200

300

400

0,0

0,4

0,8

2,9

6,6

12

,8

49

,4

85

,9

32

8,6

70

0,0

Co

un

ts

EGFP Fluorescence Intensity

Target Population 1

0

100

200

300

400

0,0

0,4

0,8

2,9

6,6

12

,8

49

,4

85

,9

32

8,6

70

0,0

Co

un

ts

EGFP Fluorescence Intensity

Target Population 2

Reproducible results in two independent cassette

exchange tests of clone #2

Two independent cassette exchange tests

0

1

2

3

4

5

6

7

0 24 48 72 96 120 144 168 192 216

Xv

(10

6ce

ll/m

l)

Culture time (h)

Cell line 2 Cell line 3

RMCE-based stable expression vs BEVS

doubling

time (h)

Xmax

(106cell/ml)

qEGFP

(µg/106cell/h)

eGFP titer

(mg/l)

Cell line 2 41.9 ± 4.2 6.2 ± 0.6 0.079 ± 0.004 47.9 ± 4.07

Cell line 3 41.7 ± 5.8 4.8 ± 0.5 0.078 ± 0.004 44.2 ± 3.69

Baculovirus infection

MOI=0.1, CCI=3 - 3.9 ± 0.4 0.140 ± 0.008 67.9 ± 5.67

96h 192h

#3 #3 #3

- Similar eGFP titers to those obtained with BEVS;

- More room for improvement

Cell line

Cell lines derived from “weak” tagged clones

Not all sites are equally amenable to recombination


Master Sf9 cell line: Expression of complex proteins

VP2

Rotavirus

VP7

• Rotavirus VP2-GFP

Master cell line

+ Flp

Targeting plasmid:

VP2-GFP

Fw F5

ATG OpIE2 OpIE1

Δneo dsRed Fw

Hygro

F5

OpIE2 OpIE2 OpIE1

Rotavirus VP2 core layer: 120 copies of VP2 (102 kDa) organized in 60 dimers

Fernandes et al. J Biotechnol 2014, 171: 34–38.

VP2

Rotavirus

VP7 Δneo OpIE2 OpIE2 OpIE1 VP2-GFP ATG

VP2-GFP

Tagg

ing

Targ

etG

FP

VP

2-G

FP

• VP2-GFP insertion in the tagged locus

Genomic PCR

Rotavirus VP2 core layer: 120 copies of VP2 (102 kDa) organized in 60 dimers

• Rotavirus VP2-GFP

Master Sf9 cell line: Expression of complex proteins


0

2

4

6

8

10

12

0 24 48 72 96 120 144 168 192

Via

ble

ce

ll co

nce

ntr

atio

n(1

06

cells

/mL)

Time (h)

Sf9-VP2GFP

BEVS

Time of infection

Shake flask cultures (50 ml)

VP2-GFP Expression: Stable Sf9-Flp vs Sf9-BEVS

Flow cytometry

Confocal microscopy

BEVS: higher fluorescence intensity at population level - stronger promoter (polh) and high transgene copy number

120 KDa

100 KDa

Similar volumetric yields

Master Sf9 cell line: Bioprocess Optimization

Feeding Strategy A Feeding Strategy B

Feed 1

(96 h) 10 mM Ser, 1 mM Cys 10 mM Ser, 1 mM Cys

Feed 2

(144 h)

20 mM Glucose, 2 mM Gln and

insect medium supplement (1x)

20 mM Glucose, 2 mM Gln, 4g/L

peptones, lipidic cocktail (1x) and

vitamin solution (1x)

Feed 3

(192 h)

15 mM Glucose, 2 mM Gln and

insect medium supplement (1x)

15 mM Glucose, 2 mM Gln, lipidic

cocktail (1x) and vitamin solution (1x)

0

4

8

12

16

20

0 48 96 144 192 240 288

Via

ble

ce

ll co

nce

ntr

atio

n

(106

ce

lls/m

l)

Time (h)

Control

Feeding A

Feeding B

Biorector culture w/ feeding strategy A

Increase in maximum cell density: 20x106cell/ml

168 h

50 nm

50 nm

0

5

10

15

20

0 48 96 144 192 240

Via

ble

ce

ll co

nce

ntr

atio

n(1

06ce

lls/m

L)

Time (h)

well assembled ~50 nm core-like particles



Aim: to increase cell density and/or extend culture time

0

2

4

6

8

10

12

0,0

0,3

0,6

0,9

1,2

1,5

1,8

2,1

2,4

2,7

3,0

0 24 48 72 96 120 144 168 192 216

Ce

lls

, G

ln

His

, L

eu

, T

hr,

Tyr

(mM

)

Culture time (h)

His

Leu

Tyr

Gln

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0 24 48 72 96 120 144 168 192 216

Ser

(mM

), C

ell

s (1

06

cells

/ml)

Cys

tin

e (m

M)

Culture Time (h)

Cystine

Viable cells

Serine

Analysis of nutrient consumption

Serine and cystine depletion at 120h of culture

Culture medium: SF900II

At 120h the cells consumption of aminoacids stops

0.0

0.3

0.5

0.8

1.0

1.3

1.5

1.8

2.0

2.3

2.5

0

3

6

9

12

15

18

21

0 24 48 72 96 120 144 168 192 216 240

Succ

inat

e (m

M)

Ala

, (m

M),

ce

lls

Culture time (h)

Cells

SuccinateAlanine

Exometabolome analysis of bioreactor culture by 1H-NMR

By-product accumulation in supernatant: Alanine: steady accumulation along culture time Succinate: consumed in the begining and then produced Lactate: accumulation starts at 192h culture time Glycerol: accumulation starts at 192h culture time

Coincides with increase in glucose consumption

0

5

10

15

20

25

30

35

40

45

50

0

3

6

9

12

15

18

21

0 24 48 72 96 120 144 168 192 216 240

Glu

cose

(mM

)

Lac,

Gly

cero

l (m

M),

ce

llsCulture time (h)

Glycerol

Lactate

Cells

Glucose

Aim: Disclose strategies to extend production phase


• 4 weeks to have Sf9-Flp expressing a new protein

• Proof-of-concept with GFP and VP2 rotavirus core-like-particles

• Bioprocess optimization allowed similar protein titers to those obtained with BEVS

But we were working with “weak” clones...

... because the strongest clones were not amenable to recombination

Sf9 tagging clones

Status of 1st generation Sf9-Flp expression platform

250 µm

Gre

en

co

lon

ies

2 w

ee

ks

po

st-t

ran

sfe

ctio

n

Tagged clone #2 Flpe iFlp

250 µm

Flpe iFlp

2 representative areas

Tagged clone #3

EGFP

Recombination efficiency

iFlp >> Flpe 2 weeks instead of 3 to select a

population with exchanged cassettes

• iFlp – new flipase codon optimized for insect cells

2 representative areas

Improving recombination efficiency

Cell Sorter Protocol

Vidigal et al. J Biotechnol, 2013, 168(4), 436–439.

• Standard mammalian sorting conditions yield low recovery viabilities (cells to be sorted re-suspended in PBS with 2% FBS)

• Addition of pluronic F68 to cell preparation improved post-sorting viabilities

• Hi5 cells are more resistant to shear stress of the sorting process (less PF68 needed to maintain high viability)

Stable expression: High Five vs. Sf9 cells

2 weeks in selection

• Transfection of Sf9/Hi5 with the same tagging cassette

DsRedintensity

Sf9 (3-fold) >> Hi5 cells Hi5 Sf9

Sorting the 10% top expressers

Populations successfully enriched!

• Sorted Hi5 and Sf9 cell populations are stable along passages

Hi5

With Hygro Without Hygro

Sf9


Production of HCMV gB Protein in insect cells

Cytomegalovirus

Full-length gB (fusion competent)

gB ectodomain (post-fusion state)

Substantially different

Aim: Full-length gB structure

gB - transmembrane glycoprotein, required for virion-mediated membrane fusion

Production Platform: BEVS - HighFive cells

• Intensive process optimisation allowed 0.35 mg/L

uncl. FL-gB

gB TM chain (furin processed)

Protein degradation

Patrone et al. Plos ONE. in press

HT crystal screening (MPL-Diamond, UK)

FL-gB HSV-1 gBecto

Sf9 / Hi5 populations expressing FL-gB

Higher specific productivity in Hi5 cells

gB standard(ng)

12.55

Sf9-gB sorted

10th9th8th

Hi5-gB sorted

6th5th4th

Hi5

-eG

FP

Sf9

-eG

FP

120kDa

55kDa

gB precursor

gB TM chain (Furin processed)

day

• Cassette exchange; G418 selection; Sorting of DsRed- cells

ii) FL-gB

0

2

4

6

8

10

12

14

0 1 2 3 4 5 6 7 8 9 10 11

Via

ble

ce

ll d

en

sity

(1

06ce

ll/m

l)

Time (days)

gB-expressing cells

Hi5 Sf9

0.00

0.10

0.20

0.30

0.40

0.50

Hi5-BEVS Hi5-gB Sf9-gBP

rote

in y

ield

(mg

/L)

• Comparison of expression from single copy integration


Final titers in Hi5-Flp similar to those obtained with optimized Hi5-BEVS

All lanes loaded with protein extracts from same number of cells

120 KDa

55 KDa

iii) Enveloped VLPs Gag-eGFP

• Comparison of expression from single copy integration


Gag-eGFP

Sf9 cells

eGFP

Hi5 120h Sf9 216h

0.0

2.0

4.0

6.0

8.0

10.0

0 48 96 144 192 240

Ce

ll C

on

cen

trat

ion

(x1

06 /

ml)

Time (h)

Sf9 GAGeGFPHi5 GAGeGFP

• Higher specific productivity in Hi5 cells

83 KDa

40 KDa

HIV-1 Gag polyprotein fused to eGFP

eGFP

All lanes loaded with equal supernatant volume

Gag analysis in supernatant

Summary

Ongoing work

2nd generation platform:

• iFlp allows shorter development timelines (3 weeks to have a new protein being expressed)

• Suitable sorting conditions for insect cell lines – powerful tool to assist RMCE implementation

• Expression of gB and gag-VLPs from single copy integration much higher in Hi5 cells

• Limiting dilution and clone screening to isolate best clones of Sf9 and Hi5 cell populations

• Development of stable platforms to express enveloped VLPs

Re-usable insect cell platforms to produce multiple rec proteins and VLps:

skipping extensive clone screening alternative to baculovirus infection single or multiple protein expression

METABOLIC PROFILING OF INSECT CELL LINES

Disclosing cell line determinants behind system’s productivity

Temperature

pH

Shear

stress

osmolality

Toxic

metabolites

Nutrients

limitation

ENVIROME

O2

Large number of independent metabolic pathways

Only a subset is active for a given envirome

Glc

Lac

Ala Gln

Glu

Amm

CO2

osmolality

Temperature

THE ROLE OF THE ENVIROME IN METABOLIC ACTIVITY

Algorithm outputs: Active metabolic pathways Correlation coefficients between envirome factors and pathway fluxes

PROJECTION TO LATENT PATHWAY (PLP) ALGORITHM

Teixeira et al. (2011) BMC Syst Biol 5:92 Ferreira et al. (2012) BMC Syst Biol 5:181

The stoichiometry of the metabolic network is built in the PLS structure to bridge metabolic activity and the extracellular state.

Background

rNe

u A

ctiv

ity

(U/1

06

cel

ls)

0

1x10-1

2x10-1

3x10-1

Sf9 Hi5 0

4x10-2

8x10-2

Ext Int Ext Int

Sf9 Hi5

GFP

(µ

g/1

06

cel

ls)

1.2x10-1

rNe

u A

ctiv

ity

(U/1

06

cel

ls)

0

1x10-1

2x10-1

3x10-1

Sf9 Hi5 0

4x10-2

8x10-2

Ext Int Ext Int

Sf9 Hi5

GFP

(µ

g/1

06

cel

ls)

1.2x10-1

Distinct productivity phenotypes: cell line dependent

Why?

I. Why? Contributing to Fundamental knowledge

II. Why? Bioprocess development and optimization – designing “rational” feeding strategies

Multi-level analysis of insect cells physiology in the biotechnological context:

Study of the impact of baculovirus infection on the host cells

Disclose cell-specific determinants behind highly productive phenotypes

Identification of new targets for engineering purposes

Rational design of an improved bioprocess for vaccine production:

Target oriented optimization of system bottlenecks

WHY do Hi5 cells produce more than Sf9 cells ????

IC-BEVS metabolic profile

II. Rationale

Sf9 cells

Hi5 cells

Growth

Infection

Growth

Infection

Multivariate Data Analysis

Pathway analysis

Paving the way to productivity tracer’s identification

Results

I. IC-BEVS metabolic profile

23

extracellular

metabolites

62

intracellular

metabolites

Results

II. Multivariate Data Analysis

Component 1 (58.9%)

Co

mp

on

ent

3 (

24

%)

PLS-DA (Projection to latent structures – discriminant analysis (PLS-DA)

Score plot

High

Low

CMP

Taurine

CDP

Biotin

GSSG

L-Cystine

SAM

GSH

Thymine

VIP scores

Monteiro, F. et al (2013)

Results

III. Pathway analysis - Pathway Impact

Ala, asp & glu metabolism

Phe metabolism

Val, leu & ile biosynthesis

Gln & glu metabolism

His metabolism

Phe, tyr & trp biosynthesis

a - tRNA biosynthesis

Tyr metabolism

Cys & met metabolism

Riboflavin metabolism

Gly, ser & thr metabolism

Metabolic pathways regulated by Bv infection

PI=1

PI=0.7

PI=0.6

PI=0.5

PI=0.4

PI=0.3

PI=0.2

PI=0.1

Sf9

Monteiro, F. et al (2013)

Results

III. Pathway analysis - Pathway Impact

Metabolic pathways regulated by Bv infection

PI=1

PI=0.7

PI=0.6

PI=0.5

PI=0.4

PI=0.3

PI=0.2

PI=0.1

Hi5

Val, leu & ile biosynthesis

Phe, tyr & trp biosynthesis

His metabolism

Ala, asp & glu metabolism

Phe metabolism

Gly, ser & thr metabolism

tRNA biosynthesis

Glycolysis, Gluconeog.

Tyr metabolism

FA elongation

Cys & met metabolism

Nicotinate & NAM metab.

TCA cycle

a - GSH metabolism

Taurine & hypotau. metab.

Pyr metabolism

Trp metabolism

Biotin metabolism

Pyrimidine metabolism

Glyoxylate & dicarbox. metab. Monteiro, F. et al (2013)

Conclusions

Multivariate analysis pin-pointed specific metabolite families related to hyper productive states;

Hi5 cells re-orient their metabolic activity towards feeding the increased biosynthetic activity, a key factor to hyper-productivity, whilst Sf9 do not; The “metabolomic decomposition” of the IC-BEVS system identified traits leveraging productivity, and specific pathways of utmost importance to support it.

Summary and Future Work Acknowledgements

Kristala Prather (MIT) Jopp Van den Heuvel (HZI) Hansjorg Hauser (HZI)

Funding:

Ana Sofia Coroadinha Manuel Carrondo Cristina Peixoto Marcos Sousa

Ana Teixeira

Fabiana Fernandes

João Vidigal

Mafalda Dias

Marco Patrone

Francisca Monteiro

http://www.mitportugal.org/

insect cell technology as a vaccine- producing platform · paula m alves insect cell technology as...

Documents