science webinar series...science /aaas business office. dr. john p. nolan. la jolla bioengineering...

Sponsored by:

Participating Experts:

Brought to you by the Science/AAAS Business Office

Dr. John P. NolanLa Jolla Bioengineering InstituteLa Jolla, CA

24 March, 201024 March, 2010

The Next Generation of Cell AnalysisThe Next Generation of Cell AnalysisGaining Insight With CytometryGaining Insight With Cytometry

Webinar SeriesWebinar SeriesScienceScience

Dr. Albert D. DonnenbergUniversity of PittsburghPittsburgh, PA

Dr. William G. TelfordNational Institutes of HealthBethesda, MD

Flow Cytometry: The State of the Art

Bill Telford, Ph.D.

Flow Cytometry Core LaboratoryNational Cancer Institute

National Institutes of Health

All of these components of flow cytometry have seen dramatic advances since the first commercial cytometers

in the 1970s.

First, you need a biological sample…

…usually labeled with a fluorescentmarker

You need to move

these cellsin a linear stream through afocused light source.

You need a light source

to excitethe fluorescent molecular on or inthe cell.

Usually a CW or quasi-CW laser, goodshort-

and long-term stability, low noise

PMT

You need a filter

that onlyadmits light from the fluorescent probe.

You need a sensitivelight detector

(usuallya photomultiplier tube).

You need electronics that can convertanalog fluorescent signals to digital ones.

Finally, you need acomputer to processthe digital signalsand display the data

What goes into a flow cytometer?

Immunophenotyping

The number and characteristics of fluorescent probes for flow cytometryhave improved dramatically.

Immunolabeling remains the dominant application, against both extracellularand intracellular targets. Intracellular targets now include phosphoproteinsand transcription factors, as well as cytokines and chemokines.

Heavily substituted low molecular weight fluorochromes. Probes like theAlexa

FluorTM

dyes are modifications of existing dyes like fluorescein

andrhodamine. They have improved overall brightness, are more optimal forphysiological pHs and show reduced photobleaching. They are available atexcitation and emmison

bandwidths covering the entire visible spectrum.

Phycobiliproteins

and their tandems.

Phycoerythrin

(PE) and allophycocyanin(APC) are mainstays of fluorescent labeling. Their FRET tandems

such as PE-Cy5, PE-Cy7 and APC-Cy7 provide a number of additional colors in the long red range. Newer acceptor fluorochromes

have made them brighter andmore stable.

Violet excited fluorochromes

such as Pacific Blue, Pacific Orange, V450 and V500 allow the use of violet laser diodes in multicolor immunolabeling.

Quantum nanoparticles

increasing diameter = increasing wavelength

655 nm705 nm

525 nm585 nm

semiconductor core (CdSe)

semiconductor shell (ZnS)

polymer coating

conjugated proteins

Encapsulated semiconductor nanocrystals

that fluoresce strongly at tightly defined wavelengths, depending on crystal diameter. QdotsTM (Invitrogen) and eFluorsTM (eBioscience) are commercially available.

Emission bandwidths are much more narrow than traditional fluorophores

Optimal excitation with a UV or violet laser source. Extremely resistant to photobleaching

605 nm

800 nm

500 600 700 800Wavelength (nm)

400 500 600 700 800 900

FITC PE PE-Cy5 PE-Cy7

APC APC-Cy5.5

APC-Cy7

PacificBlue

Emission wavelength (nm)

Blue-green(488 nm)laser

Red(633 nm)laser

Violet(405 nm)laser

5 colors

3 colors

Fourteen color analysis

UV(355 nm)laser

Qdot

605

Qdot

705

3 colors

PacificOrange

3 colors

Qdot

625 Qdot

800

PE-Cy5.5

Now achievable using commercial instruments and conjugates..

Protein Reference Excitation, nm

Emission, nm

Brightness, % of EGFP

EBFP Yang et al., J. Biol. Chem., 1998, 273, 8212 380 440 27

Azurite Mena et al., Nat. Biotechnol., 2006, 24, 1569 383 447 43

EBFP2 Ai et al., Biochemistry, 2007, 46, 5904 383 448 60

Cerulean Rizzo et al., Nat. Biotechnol., 2004, 22, 445 433 475 79

ECFP www.clontech.com 439 476 39

CyPet Nguyen et al., Nat. Biotechnol., 2005, 23, 355 435 477 53

TagCFP www.evrogen.com 458 480 84

AzamiGreen www.mblintl.com 492 505 121

TagGFP www.evrogen.com 482 505 100EGFP www.clontech.com 484 507 100

Emerald Cubitt et al., Methods Cell. Biol., 1999, 58, 19 487 509 116

T-Sapphire Zapata-Hommer et al., BMC Biotechnol., 2003, 3 399 511 78

TagYFP www.evrogen.com 508 524 137

EYFP www.clontech.com 514 527 151

Topaz Cubitt et al., Methods Cell. Biol., 1999, 58, 19 514 527 169

Venus Nagai et al., Nat. Biotechnol., 2002, 20, 87 515 528 156

Citrine Griesbeck et al., J. Biol. Chem., 2001, 276, 29188 516 529 174

YPet Nguyen et al., Nat. Biotechnol., 2005, 23, 355 517 530 238

Protein Reference Excitation, nm

Emission, nm

Brightness, % of EGFP

Kusabira Orange www.mblintl.com 548 559 92

mOrange Shaner et al., Nat. Biotechnol., 2004, 22, 1524 548 562 146

dTomato (dimer)

Shaner et al., Nat. Biotechnol., 2004, 22, 1524 554 581 142

DsRed (tetramer) www.clontech.com 558 583 176

DsRed-Express (tetramer) www.clontech.com 555 584 58

TagRFPMerzlyak et al., Nat. Methods,

2007, 4, 555www.evrogen.com

555 584 146

DsRed- Monomer www.clontech.com 556 586 10

mStrawberry Shaner et al., Nat. Biotechnol., 2004, 22, 1524 574 596 78

mCherry Shaner et al., Nat. Biotechnol., 2004, 22, 1524 587 610 47

mKeima Kogure et al., Nat. Biotechnol., 2006, 24, 577 440 620 12

mRaspberry Wang et al., PNAS, 2004, 101, 16745 598 625 37

Katushka (dimer)

Shcherbo et al., Nat. Methods, 2007, 4, 741 588 635 67

mKate (TagFP635)

Shcherbo et al., Nat. Methods, 2007, 4, 741

www.evrogen.com588 635 45

HcRed (tetramer)

Gurskaya et al., FEBS Lett., 2001, 507, 16 592 645 5

mPlum Wang et al., PNAS, 2004, 101, 16745 590 649 12

Fluorescent proteins for flow cytometry

Vladislav V. Verkhusha, Ph.D and William KingAlbert Einstein College of Medicine

Physiological probes for flow cytometry

An expanding number of fluorescent probes for morphological and physiological cell measurements are now available.

DNA binding dyes.

Cell impermeant

and permeant, with excitation/emission characteristics spanning the visible spectrum.

Viability dyes. Protein modifying dyes (i.e

Live/Dead) are now commonly used for viability measurements, and are included in immunolabelingprotocols.

Mitochondrial and cell membrane potential dyes. To measure Mitochondrial and cell membrane status, apoptosis.

Cell tracking dyes. Available in all colors, with a variety of

incorporation mechanisms for long and short term labeling.

Fluorogenic

enzyme substrates.

Membrane pump substrates.

Early cytometer

were typically equipped with a single water-cooled gaslaser emitting at 488 nm. Fluorochrome

analysis was limited to this wavelength. Lasers have become smaller and more efficient, but 488 nm remains as our base wavelength.

Blue-green 488 nm

Fluorescein

(FITC)Phycoerythrin

(PE)PE-Cy5, PE-Cy7GFP, YFP

Red 633 or 635 nmAllophycocyanin

(APC)APC-Cy7Cy5, red Alexa

Fluor dyes

Violet 405 nm Pacific Blue, Pacific OrangeHoechst dyes, DAPI, Qdots

Red HeNe

lasers

and small red laser diodes

were subsequently added as a second laser source, giving us a variety of red-excited fluorochromes.

Inexpensive violet laser diodes appeared in cytometers, and provideexcitation for a number of fluorochromes, including quantum nanoparticles.

Other laser wavelengths (such as UV) could be generated by gas lasers, butwere difficult to produce and expensive. Not available on most instruments.

Until recently, our analysis capabilities were largely limited by the numberof laser wavelengths available for flow cytometry. Although very useful, theusual “triad” of blue-green, red and violet lasers only excites a fraction ofthe fluorochromes

available for cytometry.

405 nm

488 nm

633 nm

440 nm

532 nm543 nm546 nm550 nm555 nm561 nm570 nm580 nm592 nm

650 nm670 nm

Recent advances in laser technology

has given us a new generation of solid state lasers that cover virtually the entire visible spectrum (ultraviolet to infrared).

628 nm642 nm

505 nm515 nm

457 nm473 nm

Solid state lasers

are small, reliable, easy to integrate into existing instrumentation, and are rapidly decreasing in cost.

Most importantly, they are nowavailable in virtually any color,allowing excitation of almostany fluorescent molecule. NTT Photonics, Inc.

Multiple lasers is now the norm for flow cytometry…

Beckman-Coulter Astrios… 7 lasers

BD Biosciences LSR Fortessa… 5 lasers

BD LSR II SORP… up to 7 lasers

BD (Cytopeia) InFlux… 6 lasers

Stratedigm

S1400… 4 lasers

Partec

CyFlow… 4 lasers

DPSS 532 and 561 nm lasers for flow cytometry

50 mW

(Laser-Compact)

DPSS 532 nm green

DPSS 561 nm green-yellow

DPSS 532 and 561 nm lasers are becoming common fixtures on flow cytometers, supplementing existing 488 nm sources.

Phycoerythrin (PE)

532 nm488 nm

450 500 550 600 700 750400 650

Riboflavin

Rel

ativ

eflu

ores

cenc

e450 500 550 600 700 750400 650

Wavelength (nm)

450 500 550 600 700 750400 650

561 nm

Riboflavin

EXEM

532 and 561 nm lasers provide betterexcitation of PE and its tandems than488 nm sources, while reducing cellularautofluorescence.

Wavelength

Tsien

“fruit”

fluorescent proteins

The newest generation of fluorescent proteins (i.e. the Tsien

“fruit”

FPs) oftenexcite poorly at 488 nm, and require green or yellow light.

mHoney

dew

mOrange

dtTom

atomTan

gerin

emStra

wberry

All of the fruit FPs

except mHoneydewideally require a 500 to 600 nm excitation source.

excitation emissionmBa

nana

mCher

ry

Shaner, N.C. et al. Nat. Methods 2, 905-909 (2005).

Wavelength (nm) Wavelength (nm)

Fiber lasers

near-IR laserdiode pump

high reflector

544 nmlaser output

output couplerpump beam correction optics

808 nm

In a fiber laser, a specially doped fiberoptic is coupled to a IR pump laser.

The fiber is the lasing cavity.

fiber optic

Fiber lasers can generate virtually any wavelength.

MPBC 580 nm

MPBC 592 nmMPBC 628 nm

We can fine-tune our lasers to our fluorochromes.

Zecotek

550 nm

Supercontinuum

laser sources

Supercontinuum

or white-lightlaser sources emit over a broadrange of visible and infrared wavelengths. With one laser, weshould theoretically be able tochoose any wavelength we needfor analysis by selectively filteringthe laser light.

This will provide the ultimate inwavelength flexibility.

Fianium

Ltd. (UK)NKT Photonics (Denmark)

Fianium, Ltd.SC-450

NKT PhotonicsSuperExtreme

Supercontinuum

laser sources

529/25 nm488/10 nm 575/25 nm 632/22 nm

400 800500 600 700Wavelength (nm)

Post-IR cold mirrorsRe

lati

ve p

ower

leve

l

Other tunable laser technologies are coming soon.

Data collection and analysis

Digital systems have largely supplanted analog systems for rapid, accurate data acqusition

and analysis.

Most mathematical operations (log conversion, compensation) are now doneby computer, or using digital signal processors, rather than hardwiredelectronics.

• More accurate calculations.

• Greater flexibility in data analysis.

• Rapid sorting decisions, resulting in laser sorting rates and fewer aborts.

Sophisticated offline data analysis programs (TreeStar

FlowJo, De Novo software FCS Express, Verity WinList) that can manage digitally acquired data, including complex spillover analysis.

Probability state analysis software (Gemstone).

Digital acquisition systems and software make polycolor

flow cytometry possible.

Cell sorting

Digital systems have greatly increased cell sorting throughput, eliminating“dead time” and allowing continuous sort decisions.

Electrostatic sort rates of 25,000 –

75,000 events/second can be achieved.

Six-stream sorters are becoming the benchmark (BD InFlux

and Beckman-

Coulter Astrios).

Parallel sorting for extremely high-throughput applications (iCyt

Reflection).

Biohazard containment is now standard on cell sorters, with some

instrumentshoused in biosafety

hoods.

Microfluidic

based sorting systems for clinical applications (Cytonome).

Up-and-coming cytometry technologies

Phase sensitive and spectral flow cytometry

Raman light scatter cytometry

Atomic mass cytometry

Microfluidics

systems (Lab on a Chip)

Scanning or image cytometry

Slide-based systemsStream-based systems (Amnis)Interventional scanning cytometry (Cyntellect)

Compucyte iCys

DVS Sciences CyTOF

Up to 50 parameters!

Small, inexpensive cytometers

Accuri

C6, Millipore GuavaCytometry4Life

http://www.dvssciences.com/images/cytof_31.jpg

NCI ETIB Flow CytometryCore Laboratory

Sponsored by:




24 March, 201024 March, 2010





Albert D. Donnenberg, Ph.D.UNIVERSITY of PITTSBURGH CANCER INSTITUTE

[email protected]

New Tools for Cytometry: Hardware and Software

AAAS March 2010

New Tools: Hardware

Imaging Flow Cytometry: ImageStream100

Debris & Monsters

MonstersDebris/RBC

SSc

BlueCD45 RedBrightfield

Gray

Lymphs

Monos

Grans Eos

K562 Target (CTO+)

CD3+ T cell

CD16/56+ NK cell

Gated on 7AAD+ Targets (50:1 E to T)

Cellular Cytotoxicity: High E:T

CD45-

HEA-

CD90+ Tumor Stem Cell in Primary NSC Lung Ca

Draq5 (Nucleus)

CD90+

Surface Mucin

NORMAL

Surface Mucin

TUMOR

CD45-

HEA-

CD90+ Tumor Stem Cell in Primary NSC Lung Ca

Draq5 (Nucleus)

Mucin+

CD90

CD90

New Tools: Software

VenturiOne: Efficient Use of 8 hyperthreads

during recalculation

Tesla graphics processing engine with 240 parallel CPU

Kaluza: Support of Graphics Processing Engine

Ve

ntu

riO

ne

: Pre

vie

w

Pa

in S

ho

ws

SP

LO

M

(Sc

att

erp

lot

Ma

trix

) o

n G

ate

d E

ven

ts

17 parameter dataset

CD34-/CD90- CD146+

CD34-/CD90+ CD146+

CD34+/CD90+ CD146-

CD90

CD34

CD146

CD34+/CD90+ CD146+

SA-ASC pericytesK

alu

za: n

-Dim

en

sio

na

l R

ad

ar

Plo

t

•Metafile graphics are composed of scalable objects•Use of play lists to organize data analysis /reanalysis•Exporting analysis results to CSV Files

CD45 APC‐Cy7

ABC NOT DEF

B: 45.2%

CD45 APC‐Cy7

ABC NOT DEF

CLEAN CD45‐

CD3 FITC

FL1‐

FITC

VenturiOne: Metafile Graphics and other analysis enhancements

SPICE: Data mining tool for visualization of complex multiparameter

data

Gemstone: New paradigm for modeling modulation of markers during differentiation

Thanks to

• AAAS

• Accuri

• AVDLab

• UPCI-Cytometry Facility

Sponsored by:




24 March, 201024 March, 2010





Screening and Systems Cytometry More Samples, More Parameters

John P. Nolan, Ph.D.

La Jolla Bioengineering Institute

[email protected]

Needs for Large Scale Cytometry Proteomics, Systems Biology, Drug Discovery, Clinical

Diagnostics

• Quantitative measurement of many molecular features under many conditions

– Dozens to hundreds of proteins/epitopes– Dose‐response, time course

– Replicates, controls• Solutions:

– More Samples: High Throughput FC

– More Information: Highly Multiparameter

FC

More Samples: High Throughput FC Bruce Edwards, Larry Sklar

–

U. New Mexico

Edwards et al (2006) Nature Protocols

As little as 2 ul/sampleA 96‐well plate can be sampled in as little as 2.5 minutes. A 384‐well plate can be sampled in as little as 10 minutes.

Multiplexing with Beads

Nolan and Mandy 2001 Cell Mol Biol

47

Multiplexing with Cells: Cell Barcoding Peter Krutzik, Garry Nolan –

Stanford

Krutzik

et al (2008) Nature Protocols

Barcode Deconvolution

Krutzik

et al (2008) Nature Protocols

Challenges of Multiparameter

FCHighly multi‐color FC requires

multiple lasers, detector assemblies – expensive and

complicated

Software, reagent developments are making

multicolor FC more accessible

Usable spectral range filled, further increases will be

incremental

635 nm Excitation - 3 colors

Wavelength (nm)400 500 600 700 800 900

Inte

nsity

0

20

40

60

80

100 APCAPC-A680

APC-A750


Wavelength (nm)400 500 600 700 800 900

Inte

nsity

0

20

40

60

80

100

FITCPEPE-Texas Red

PE-Cy5PE-Cy7

405 nm Excitation with Quantum Dot Reagents - 9 colors

Wavelength (nm)400 500 600 700 800 900

Inte

nsity

0

20

40

60

80

100

CascadeBlue

QD525 QD565QD545 QD585

QD605QD655

QD705QD800

Approaches to Higher Parameter FC

• More lasers, detectors – but spectral space is limited, complexity and cost are high

• More efficient use of optical spectrum – Single cell Raman spectroscopy using nanoparticle

SERS tags

• Non‐optical approaches – Single cell mass spectrometry using mass‐tags

Conventional Flow Cytometry

Sample Inlet

SheathInlet

Laser Beam

SampleStream

SheathStream

Filters

Dichroic

mirrors

Detectors

Sample Inlet

SheathInlet

Laser Beam

SampleStream

SheathStream

Optical Fiber CCD

Spectrograph

Raman Flow Cytometry Nolan Lab – La Jolla Bioengineering Institute

0

10000

20000

30000

40000

50000

60000

20 40 60 80 100 120 140 1600

100

200300400500

Inte

nsity

P ixel Number

Even

t Num

ber


Wavelength (nm)400 500 600 700 800 900

Inte

nsity

0

20

40

60

80

100 APCAPC-A680

APC-A750

633 nm Excitation: SERS Tags

Wavelength (nm)640 650 660 670 680 690 700

Inte

nsity

Surface Enhanced Raman Scattering (SERS)

• Near metal (Au, Ag) surfaces, Raman signals enhanced 1012‐ fold

• Intensities rival fluorescence: single molecule detection

• Functionalize with Abs, other reagents for cell staining

Raman Shift (cm-1)0 500 1000 1500 2000

Inte

nsity

Wavelength (nm)640 660 680 700 720

Spectral Unmixing

of SERS Tags

Raman Shift (cm-1)

0 200 400 600 800 1000120014001600

Inte

nsity

0

5000

10000

15000

20000

25000

Raman Shift (cm-1)

0 200 400 600 800 1000120014001600

Inte

nsity

0

5000

10000

15000

20000

25000

0

10000

20000

30000

40000

50000

60000

20 40 60 80 100 120 140 1600

100

200300400500

Inte

nsity

Pixel Number

Even

t Num

ber

Three Tag Mixture

Raman Shift (cm-1)

0 200 400 600 800 1000120014001600

Inte

nsity

0

5000

10000

15000

20000

25000

Spectra from Individual Cells Reference spectra Individual Tag Intensities

DMB

Cou

nt

100 101 102 103 1040

6

12

18

24

NBA

Cou

nt

100 101 102 103 1040

6

12

17

23

R6G

Cou

nt

100 101 102 103 1040

14

28

41

55

DMB

NBA

R6G

Raman Shift (cm-1)

0 200 400 600 800 1000120014001600

Inte

nsity

0

5000

10000

15000

20000

25000

Bkgd

Average Spectra

Pixel Number

0 200 400 600 800 1000 1200 1400 1600

Inte

nsity

0

5000

10000

15000

20000

25000

Blank beadsDMB-COINSNBA-COINSR6G-COINSMixture

E

Mass‐Tagged Antibodies and Mass Spec Scott Tanner, U. Toronto

http://www.stemspec.ca/Project/History/UofT.html

Antibodies (or other reagents) are tagged

with chelating polymers that are labeled

with different metal atoms

Inductively coupled plasma time of

flight mass spectrometry (ICP‐TOF‐MS)

can detect and measure the different

metal atoms

Sample Inlet

Nebulizer

Plasma

Time of Flight–Mass Spectrometry of Individual Cells


Drift Tube

Detector

Multiparameter

TOF Cytometry


Summary• The demand for more quantitative information

about cells and cell systems are being addressed by a combination of new instruments, new

reagents, and new analysis methods• Increasing sample throughput

– Hardware: Increases throughput, automation– Assay design: Cell barcoding

reduces time and cost

• Increasing the number of parameters measured– Conventional Approach: Multilaser

systems

– New Optical Approach: Raman cytometry– Non‐optical Approach: Mass cytometry– High Content Analysis: Image cytometry

Acknowledgements

• Bruce Edwards, Larry Sklar

–

University of New Mexico– High throughput flow cytometry, molecular library screening

• Peter Krutzik, Garry Nolan – Stanford University– Multiparameter

analysis of cell signaling, assays and

software

• Scott Tanner – University of Toronto– Mass cytometry for highly multiparameter

measurements

• Nolan Lab – La Jolla Bioengineering Institute– Raman Cytometry for Diagnostics and Drug Discovery– Bioengineering Research Partnership (NIH/NIBIB)

Look out for more webinars in the series at:

www.sciencemag.org/webinar

For related information on this webinar topic, go to:

www.AccuriCytometers.com

To provide feedback on this webinar, please e‐mail

your comments to [email protected]

Sponsored by:


24 March, 201024 March, 2010



science webinar series...science /aaas business office. dr. john p. nolan. la jolla bioengineering...

Documents