Chemical & Biological Engineering and Clinical Translational

Research Center

State University of New York, Buffalo, NY, USA

Demystifying the Glycosciences• Glycobiology: Study of the function of sugars

attached to proteins and membranes, including protein- and lipid- linked sugars.

• Glycoconjugates: Formed when mono-, oligo- or poly-saccharides attach to proteins and lipids. This occurs in ER and Golgi (mostly).

• Glycans: Carbohydrate entity attached to proteins and lipids. Found outside cells, in cytoplasm (complex) and nucleus decorating transcription factors (simple).

• Lectins: Glycan binding proteins

Function of glycans• Structural component of cell wall and extracellular

matrix proteins [Cancer and stem cell biomarkers]• Intra- and extra-cellular trafficking of

glycoconjugates [Protein therapeutics half-life]• Cell adhesion: during cell-cell and cell-matrix

interaction. [Inflammation, human-virus and human-microbiome interactions]

• Cell signaling: Intracellular and extracellular [transcription factor regulation. O-GlcNAcformation on Ser/Thr competes with phosphorylation]

Why is glycosylation so complicated?

A. It is not part of the central dogma – not part of regular course work.

B. Because the biochemists made it look more complicated than it really is:

– Database lists 700 different monosaccharides. In humans there are only 9!

– They said that there are 1012 possible carbohydrates. In reality it is closer to 102-103 classical structures.

– The names are so complicated. Why doesn’t everyone just use IUPAC nomenclature

C. Because it is complicated

Common monosaccharides

• Most glycans are hexose sugars: 4 chiral carbons resulting in 16 different molecules (epimers and enantiomers).

• But many are of relative low abundance. Humans are made of 9 major monosaccharides.

All sugars are similar in structure

D-Glc (Glucose)D-GlcA

[oxidation of C6]

Xyl (Xylose)[remove C6]

D-Man (Mannose)[C2-epimerase]

D-GlcNAc[C2-acetylatedamino gp.]

D-Gal (Galactose)[C4-epimerase]

D-ManNAc

Sialic acid

+ pyruvate

L-fucose[6-deoxy-D-Gal]

D-GalNAc

O

OH

OH

OH

OH

OOH

OH

OH

OH

OH

OOH

OH

OH

OH

CH3

OOH

OH

NHAc

OH

OH

O

H

HH

H

OH

OH

H NHAc

OH

OH

O

OH

OHOH

OH

OH

O

OHOH

OH

OH

OOH

D-ManNAc

O

H

HH

H

OH

OH

H OH

OH

OH

1

2

34

5

9

7

6

8

Different families of glycans

Wiley Interdiscip Rev Syst Biol Med. 2015 Jul-Aug;7(4):163-81

Of course there are more sugars in plants and the microbiome…

http://www.ncbi.nlm.nih.gov/books/NBK310273/

… and more on bacteria that regulate immune function

A. TEM of T. forsythia showing glycans. B. Schematic O-glycan core to protein. Terminal trisaccharide is circled. C. Immune signaling by C-type lectin-like receptor (CLR) and TLR2 activated by O-glycans and TLR2 ligands (e.g., BspA) orchestrates.

Front. Microbiol., 17 October 2013

Relevance to human diseases

• Human glycans bind bacterial lectin-like adhesins

• Bacterial carbohydrates bind mammalian lectins

• Bacterial carbohydrates mimic human glycans, and condition the microbiome

Fusobacterium nucleatum• Plays a role in periodontal disease.• Associated with colon cancer

wikipedia

Plasma Membrane

Roseman, S. J. Biol. Chem. 2001;276:41527-41542

Glycocalyx – Physically big

Translated protein

Post-translational modification

Functionalprotein

Ribosome

ER/Golgi

Cell-surface

Glycosyltransferases

Neelamegham and Mahal, Current Opinion of Structural Biology, 2016

Lectins: Glycan binding proteinsC-type lectins (selectins)

Cell. 2000 Oct 27;103(3):467-79.

Siglecs• Contains

shallow Argthat binds sialicacid

Schnaar, J. Leuk. Biol, June, 2016

Galectins

Nature Reviews Microbiology 7, 424-438 (June 2009)

Systems Glycobiology

?

Input

Decorated GlycoProbes

GlycoProbes

Output

WT or edited/knock-out cell

Cell rolling or adhesion phenotype

GNAT(Glycosylation Network

Analysis Toolbox)

GlycoPAT(GlycoProteomicsAnalysis Toolbox)

http://sourceforge.net/projects/gnatmatlab/

Smallmolecules

ST3Gal-3

ST3Gal-4

ST3Gal-6 α(1,3)FUT

β3

β6

α3

β4α

β4 α3

β3

α3

αβ6 β4 α3

α3

β3

α3

α β6

Type II α(2,3) Sialyllactosamine

sialyl LewisX

Which α(2,3)sialyltransferase, ST3Gal-3, -4 or -6, contributes to human selectin-ligand

biosynthesis?

Sialic acid

Galactose

Mannose

GlcNAc

GalNAc

Fucose

Mono.Sugar

Alexander Buffone Nandini Mondal

Mondal et al. Blood. 125:687-96, 2015.

Wild-Type HL-60

ST3Gal-3‾HL-60

ST3Gal-4‾HL-60

ST3Gal-6‾HL-60

ST3Gal-3‾4‾HL-60

ST3Gal-4‾6‾HL-60

ST

3Gal

-3

ST

3Gal

-4

ST

3Gal

-6

β-a

ctin

Making ST3Gal ‾ HL60 cells

Single knock down cells

ST3Gal-3 ͞

ST3Gal-4 ͞

ST3Gal-6 ͞

Dual knock down cells

ST3Gal-3 ͞ 4 ͞

ST3Gal-4 ͞ 6 ͞

5’LTR U6ST3GalshRNA

DsRed 3’LTRPGKVector-3

Buffone et al. J Biol Chem. 288(3):1620-33, 2013.

HL60 cell rolling on E-Selectin

A B

C D

Genome Editing with CRISPR-Cas9

The Scientist, 2014

HL-60 ST3Gal-4 knock out generated using CRISPR/Cas9

WT 51400 5’-GGGTAG…ATCTTCCTGCGGCTTGAGGATTATTTCTGGGTC–3’51622

KO 51400 5’-GGGTAG…ATCTT-----------------ATTTCTGGGTC–3’51622

KO 51400 5’-GGGT------238bp-------GAGGATTATTTCTGGGTC-3’51622

Guide

RNA NLS NLShSpCas9 bGHpAU6 CBh

β4 α3

α3

β3

α3

α β6

β4

β4

α3

100 101 102 103 104100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

Co

un

t

100

75

50

25

0

LeXsLeXHECA-452 ECL lectin

Abrogation of leukocyte rolling on all selectins

0

50

100

150

WT

P2

H3

ST

3G

al-4

-KOIn

tera

cti

ng

cell

s/m

m2

†‡†‡

WT

ST

3G

al-4

-KO

*

**

WT

ST

3G

al-4

-KO

AdherentRolling

*

**

E-selectin L-selectinP-selectin

0

25

50

75

100

0 10 20 30

Ro

llin

g c

ell

s (

%)

Rolling velocity (μm/sec)

WT (9/160)ST3Gal-4-KO (10/85)

Profiling glycans

N-glycans: ST3Gal-4 controls sLeX biosynthesis

Sialic acid

Galactose

Mannose

GlcNAc

GalNAc

Fucose

Mono.Sugar

WT HL-60

ST3Gal-4 KO HL-60

O-glycans: ST3Gal-4 deletion abrogates sLeX biosynthesis

Sialic acid

Galactose

Mannose

GlcNAc

GalNAc

Fucose

Mono.Sugar

Human neutrophils derived from hematopoietic stem cells (HSCs)

57% 43%

ST3Gal-4¯hHSc derived

neutrophils

DsRed Expression

0

20

40

60

80

100

120

0 2 4 6 8 10 12

Recep

tor

exp

ressio

n (

MF

I)

Time (Day)

CD11bCD34

100 101 102 103 104

1024

754

512

252

0

Fo

rward

scatt

er

5’LTR U6ST3GalshRNA

DsRed 3’LTRPGKVector

Human neutrophil rolling on selectins

0

100

200

300

400

500

DsR

ed

ST

3G

al-4‾In

tera

cti

ng

cell

s/m

m2

*

**

0

100

200

300

400

500

600

700

DsR

ed

ST

3G

al-4‾

*

0

50

100

150

200

250

300

DsR

ed

ST

3G

al-4‾

AdherentRolling

*

L-selectinP-selectinE-selectin

DsRed+ cells(stimulated HUVECs)

The multistep cell adhesion cascade

Mac-1 (αM:β2)Chemokine

receptor

LFA-1 (αL:β2)

E-selectin

Chemokine (IL-8)

ICAM-1

CD31

P-selectin

Blood flow

Endothelium

1. Tethering/Capture 2. Rolling 3. Activation 4. Firm adhesion & diapedesis

Rapid slow

Glycan dependent Glycan independent

L-selectin

Blood. 125(4):687-96, 2015ST3Gal-4 dependent

Rapid glycan profiling and chemical synthesis

Nature Methods 13, 81–86 (2016)

The multistep cell adhesion cascade

Mac-1 (αM:β2)Chemokine

receptor

LFA-1 (αL:β2)

E-selectin

Chemokine (IL-8)

ICAM-1

CD31

P-selectin

Blood flow

Endothelium

1. Tethering/Capture 2. Rolling 3. Activation 4. Firm adhesion & diapedesis

Rapid slow

L-selectin

ST3Gal-4 dependentGlycolipids Glycolipids/N-glycans

Glycan dependent

O-glycans

N-glycans

GlycoPAT: High-throughput glycoproteomics analysis

Tandem mass spectrometry in Orbitrap

Protease

Digestion

(Trypsin,

Glu-C etc.) nanoLC

MS1

Separation based on

precursor m/z

MS2

fragmentationm/z

Inte

nsi

ty

Sialic acid

Galactose

Mannose

GlcNAc

GalNAc

Fucose

Mono.Sugar

With Jun Qu

Digest in silicoCandidate glycopeptide list

In SmallGlyPep format

Protein list

Non-glycan PTM list

Digesting protease

3

4

5

Overall algorithm

List of potential glycans

Enzymes

in system

Some of the

expected glycans

Connection Inference

using

GNAT

1 2

GlycoPAT: High-throughput glycoproteomics analysis

Overall algorithm

List of potential glycans

Enzymes

in system

Some of the

expected glycans

Connection Inference

using

GNAT

12

Digest in silicoCandidate glycopeptide list

In SmallGlyPep format

Protein list

Non-glycan PTM list

Digesting protease

34

5

MS1 match

within tolerance?

6 Tandem MS

expt. data

Experimental

[M+H]Theoretical

[M+H]

MSn Spectra scoring

Xcorr

P-value

% Ion match

Top10 peaks

Ensemble & False Discovery

Score (ES) Rate (FDR)

MSn frag. spectra

Theoretical

MSn

Spectra

yes

GlycoPAT: High-throughput glycoproteomics analysis

• Xcorr• P-value• Top10• % ion-match

Ensemble score

Scoring parameters

GlycoPAT: High-throughput glycoproteomics analysis

Header

Scoring Table

FDR

options

Characterization of:

• Simple standard proteins

• Mixtures of standards

• Plasma cryoprecipitate from 5cc of blood

Cell Adhesion GlycoEngineering

• MSCs (Mesenchymal stem

cells) and CDCs (Cardiosphere

derived stem cells) are

promising cellular therapuetics.

• However, promising results

from animal studies are not

translating to clinical benefits

• Local infusion to damaged

tissue is not beneficial

• Systemic infusion proximal

to the therapeutic site

– minimally invasive

– allows repeated treatmentStrauer B E , and Kornowski R Circulation. 2003;107:929-934

Goals• Make stem cells home to sites of

inflammation, much like white blood cells

• Create methods that can be used in a clinical setting with minimal effort

• Move studies to clinically relevant large animals

Chi Lo

Experimental plan

Test modified cell binding to selectins

in vitro

Check whether treatment increases

homing in vivo

Increase α(1,3) fucose

Coupling rPSGL-1 –19Fc[FUT7+]

Adding scaffold and

sugar

Adding sugar

Thr-57}

Tyrosine

sulfationTransmembrane

domain

Decamer

Repeats

O-glycanN-glycan

Cytoplasmic

domain42 118 279 321 341 412

PSGL-1

Sialyl Lewis-x

NeuAcGal GlcNAc

Fuc

Recombinant PSGL-1

19 aa

PSGL-1

Human

IgG1

19Fc [FUT7+]

Characterization of 19Fc

19Fc

Mass (m/z)

% I

nte

ns

ity

19Fc[FUT7+]

Mass (m/z)%

In

ten

sit

y

Lo, C.Y., et al. Biomaterials, 2013. 34(33): p. 8213-22.

Coupling recombinant PSGL-1

Palmitate

Protein G (PPG)

+ +

19Fc or

19Fc[FUT7+]

MSC or CDC

CDC MSC

30 min. 30 min.

19Fc coupling with PPG increases

interactions with P-selectin

*

**

**

*

Experimental plan

Test modified cell binding to selectins

in vitro

Check whether treatment increases

homing in vivo

Increase α(1,3) fucose

Coupling rPSGL-1 –19Fc[FUT7+]

Adding scaffold and

sugar

Adding sugar

Over-expression of FUT7

Lentiviral constructs

CMV R U5

5’ LTR 3’ LTR

U3EGFPCMV FUT7 R U5

Fucose

Sialyl Lewis-x

NeuAc Gal GlcNAc

Fuc

FUT7 increases CDC

interactions with

stimulated HUVECs

CDC CDC FUT7

CDC FUT7 +

Esel blocking

Large animal studies: Swine

Test modified cell binding to selectins

in vitro

Check whether treatment increases

homing in vivo

Increase α(1,3) fucose

Coupling rPSGL-1 –19Fc[FUT7+]

Adding scaffold and

sugar

Adding sugar

Pig experiments: 30 min. brief ischemia-

reperfusion of LAD

Catheter placement

for cell injection

echocardiogram

CD

C-F

UT7

CD

C

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

Co

pie

s o

f F

UT

7/S

RY

+ c

ell

0

10

20

30

40

50

60

70

80

90

100 ×103

Left anterior

descending (LAD)

region

A B C

To

tal

fun

cti

on

ali

ze

d c

ell

s r

eta

ine

d

/g t

iss

ue

0

5

10

15

20

25

1 2 3

Avg

# F

UT

7 g

en

e c

op

ies

/10

0 n

g f

em

ale

ge

no

mic

DN

A

Regions:

*

Avg

# F

UT

7 g

en

e c

op

ies

/10

0 n

g f

em

ale

ge

no

mic

DN

A

Co

pie

s o

f F

UT

7/S

RY

+ c

ell

Pig experiments: 30 min. brief

ischemia-reperfusion of LAD

CD

C-F

UT7

CD

C

Lo, C.Y. et al. Biomaterials (2016), 74: 19-30

Conclusion• Glycosylation is a poorly communicated field--

its easier than it looks

• There are tremendous opportunities for exploration:

– New basic science

– New applications

– New drugs

The final frontier--- boldly go where no (wo)man has gone before

AcknowledgementsCurrent lab members:

Anju Kelkar, Ph.D. Arezoo Momeni

Kai Cheng Changjie Zhang

Hannah Wang Yuqi Zhu

Xinheng Yu Yusen Zou

Collaborators:

Aristotelis Antonopoulos, Anne Dell and Stuart Haslam: Imperial College-London

S. G. Sampthkumar: National Institute of Immunology (India)

John Canty: Cardiology, Univ. at Buffalo

Joseph Lau, Khushi L. Matta: Roswell Park Cancer Institute, Buffalo, NY

Funding support:

NIH Heart, Lung and Blood Institute: Systems Biology program

American Heart Association

NY State Stem Cell Program