A

C

Contro

l RASLE

0

20

40

60

80

100

% C

D4+

T ce

lls /

PBM

C

Contro

l RASLE

0

20

40

60

80

100

%C

D8+

T ce

lls /

PBM

C

Contro

l RASLE

0

20

40

60

80

100

CD

45R

O+

/ CD

4+ T

cells

B

DN

A1

DNA2

CD

4

CD8

Bea

ds

Cisplatin

CD

14

CD3 CD

45R

O

CD45RA

DN

A1

DNA2

CD

4

CD8B

eads

Cisplatin

CD

45

CD3

CD

45R

O

CD4

Supplementary Figure 1: Identification of major T cell subsets in mass cytometry data.A) Example of biaxial gating for T cell populations in the AMP mass cytometry data. B) Example of biaxial gating for T cell populations in the BWH validation cohort mass cytometry data. C) Quantification of CD4+ T cells, CD8+ T cells, and memory CD4+ T cells in the AMP cohort. Error bars show median±interquartile range.

96.9

98.8 82.4

Memory CD4+ T cells 39.8

Memory CD4+ T Cells 59.9

CD4+ Cells64.6

CD8+ Cells25.759.8

99.8

97.1

CD4+ cells 65.9

B

HLA

-DR

ICOS

PD-1hi CXCR5- CD4+ T cellsD

12 1 13 11 8 2 14 7 10 15 5 6 9 3 4C

lust

er A

Clu

ster

B

CD57TbetGranzyme BHLA-DRCD64PU.1CCR2FoxP3HeliosCD25CD27PD-1KLRG1CD69GATA3TCRgdRORgTCCR7CD94AHRCD161CXCR3CD127CTLA4CD28ICOS

A

−2 0 2Row Z−Score

Metacluster 4

Metacluster

C p = 0.21 **

Contro

lSLE

0.00

0.01

0.02

0.03

0.04

Contro

lSLE

0.00

0.01

0.02

0.03

0.04

Met

aclu

ster

4 -

Clu

ster

B

Met

aclu

ster

4 -

Clu

ster

AMetacluster 4

HLA

-DR

exp

ress

ion

Cluster A

Cluster B

Supplementary Figure 2: Phenotypic features of cells in metacluster 4.A) Heatmap of marker expression in FlowSOM metaclusters as in Figure 1D including the 2 clusters (cluster A, cluster B) that comprise metacluster 4. B) FlowSOM map demonstrating level of expression of HLA-DR in individual clusters. C) Abundance of the clusters that comprise metacluster 4, cluster A and cluster B, in SLE patients and controls as in Figure 1C. Error bars show median±interquartile range. **p<0.01 by Mann-Whitney test. D) Example contour plot of expression of HLA-DR and ICOS on gated PD-1hi CXCR5- CD4+ T cells in mass cytometry data.

0 10 20 300

20

40

60

SLEDAI

0 10 20 300

5

10

15

SLEDAI

0 10 20 300

5

10

15

SLEDAI

0 10 20 300

1

2

3

SLEDAI

0 10 20 300

3

6

9

SLEDAI

0 10 20 300.0

0.5

1.0

1.5

SLEDAI

0 10 20 300

20

40

60

SLEDAI

0 10 20 300

5

10

15

SLEDAI

r=0.46p=0.029

% P

D-1

very

hig

h CXC

R5ne

g / m

CD

4%

PD

-1ve

ry h

igh C

XCR

5+ / m

CD

4

% P

D-1

hi C

XCR

5neg /

mC

D4

% P

D-1

hi C

XCR

5+ / m

CD

4

% P

D-1

inte

rmed

iate

CXC

R5ne

g / m

CD

4%

PD

-1in

term

edia

te C

XCR

5+ / m

CD

4

(Tph

cel

l gat

e)(T

fh c

ell g

ate)

% P

D-1

tota

l pos

itive

CXC

R5ne

g / m

CD

4%

PD

-1to

tal p

ositi

ve C

XCR

5+ / m

CD

4

r=0.49p=0.018

r=0.10p=0.66

r=0.41p=0.052

r=-0.10p=0.66

r=0.38p=0.094

r=-0.23p=0.28

r=0.12p=0.58

Supplementary Figure 3: Correlation of PD-1+ T cell populations with lupus disease activity.Correlation between lupus disease activity by SELENA-SLEDAI and the frequency of different CD4+ T cell populations defined based on expression of CXCR5 and level of expression of PD-1 (n=21 patients). Spearman correlation statistics shown.

A B

0 20 40 600

5

10

15

Age

% T

ph c

ells

/ m

emC

D4

Black/A

frican

America

nW

hite

Asian

Unkno

wn or n

ot rep

orted

0

5

10

15%

Tph

cel

ls /

mem

CD

4r=-0.32p=0.005

Supplementary Figure 4: Tph cell frequencies in SLE patients according to race, ethnicity, and age.A) Tph cell frequency in SLE patients grouped by self-reported race or ethnicity. Error bars show mean±SD. B) Correlation between Tph cell frequency and age in SLE patients. Spearman correlation statistics shown.

Hispan

ic

Not Hisp

anic

Unkno

wn or n

ot rep

orted

0

5

10

15

% T

ph c

ells

/ m

emC

D4

ATph cells

CXC

R3

mea

n ex

pres

sion

PD-1negative memCD4+ Tfh cells

CXC

R5

mea

n ex

pres

sion

B

CXC

R3

mea

n ex

pres

sion

CXC

R3

mea

n ex

pres

sion

CC

R2

mea

n ex

pres

sion

CC

R2

mea

n ex

pres

sion

D E

0

40

80

120

160

CC

R2

mea

n ex

pres

sion

0

20

40

60

80

100

% C

CR

2+

0

20

40

60

80

100

% C

XCR

5+

0

10

20

30

40

CXC

R5

mea

n ex

pres

sion

Con

trol

RA

SLE

Con

trol

RA

SLE

PD-1neg Tph

Con

trol

RA

SLE

Con

trol

RA

SLE

PD-1neg Tph

Con

trol

RA

SLE

Con

trol

RA

SLE

PD-1neg Tfh

Con

trol

RA

SLE

Con

trol

RA

SLE

PD-1neg Tfh

Cnsns

* ****nsns

ns **** nsns ********

0

20

40

60

0

20

40

60

80

0

20

40

60

80 ns**** * ns**** ns * ns*

0

20

40

60

0

1

2

3

4

5

Tph cells

Contro

l

New-on

set S

LE

Establi

shed

SLE

0

1

2

3

4

5

Contro

l

New-on

set S

LE

Establi

shed

SLE

Contro

l

New-on

set S

LE

Establi

shed

SLE

Contro

l

New-on

set S

LE

Establi

shed

SLE

Contro

l

New-on

set S

LE

Establi

shed

SLE

Contro

l

New-on

set S

LE

Establi

shed

SLE

ns ***** ns * ns ns **** **

Supplementary Figure 5: Chemokine receptor expression on Tph cells in SLE.A) Mean expression of CXCR3 on Tph cells, Tfh cells, and PD-1neg memory CD4+ T cells in mass cytometry analysis of samples from the BWH validation cohort. B) Mean expression of CCR2 and frequency of CCR2+ cells in PD-1neg memory CD4+ T cells and Tph cells from control (n=25), RA (n=25), and lupus nephritis (n=27) donors in the AMP cohort. C) Mean expression of CXCR5 and frequency of CXCR5+ cells in PD-1neg memory CD4+ T cells and Tfh cells as in (B). Tfh cells are CXCR5+ by definition. D) Mean expression of CCR2 on Tph cells and PD-1neg memory CD4+ T cells as in (A). E) Mean expression of CXCR5 on Tfh cells as in (A). Error bars show mean±SD. * p<0.05,** p<0.01, *** p<0.001, **** p<0.0001 by Kruskal–Wallis with Dunn’s multiple comparisons test (A-E) statistics shown.

PD-1negative memCD4+

B

-

Supplementary Figure 6: RNA-seq sorting and interferon scores in T cell populations.A) Example of gating for sorting T cell populations for RNA-seq. B) The interferon score was calculated for each T cell population from each donor using the RNA-seq transcriptomes. Plot shows the IFN score for controls (blue, n=4-5), RA patients (green, n=4-5), and SLE patients (red, n=6) for each T cell population. Error bars show mean±SD.

Naive cells Tregs Tfh cells Tph cells-1.0

-0.5

0.0

0.5

1.0

IFN

sco

re

Control

RA

SLE

A

Single cells

Live CD3+cells

Memory CD4+ Naive CD4+ Treg

Non Treg TfhTph

Lymphocytes

nsns

0.0

0.1

0.2

0.3

0.4

0.5

IL10

rela

tive

expr

essi

on

PD1 - - hi hi - + - + CXCR5

Naive c

ontro

l

Naive S

LE

Treg c

ontro

l

Treg S

LE

Tfh co

ntrol

Tfh SLE

Tph co

ntrol

Tph S

LE0

5

10

15

20IL10

FPK

M

Supplementary Figure 7: Expression of IL10 and IFNG in SLE Tph cells A) Expression of IL10 and IFNG in T cell subsets in RNA-seq data. B) IL10 expression by qPCR in PMA+ionomycin stimulated memory CD4+ T cell subsets from SLE patients (n=4). Error bars show mean±SD (G). Mann Whitney test comparing Tph cells or Tfh cells from controls andSLE patients (A), or Kruskal–Wallis with Dunn’s multiple comparisons test (B).

A B

Naive c

ontro

l

Naive S

LE

Treg c

ontro

l

Treg S

LE

Tfh co

ntrol

Tfh SLE

Tph co

ntrol

Tph S

LE

0

50

100

150

IFNG

FPK

M

nsns

A

DN

A1

DNA2

CD

45

CD19

Bea

ds

Cisplatin

CD

14

CD3

CD

27

CD38

CD

20

CD38

DN

A1

DNA2

CD

3

CD19

Bea

ds

Cisplatin

CD

45

CD19

CD

11c

CD19

CD

16

CD14

CD

21

CXCR5

B

C

0 20 40 60 800

5

10

15

% CD21- CXCR5- / B cells0 20 40 60 80

0

5

10

15

% CD21- CXCR5- / B cells

% T

ph c

ells

/ m

emC

D4+ T

cel

ls

% T

fh c

ells

/ m

emC

D+ T

cel

ls

r = 0.60p = 0.0009

r = 0.67p < 0.0001r = 0.60p = 0.0016

r = 0.15p = 0.46

r = 0.32p = 0.0047r = 0.05p = 0.81

E

0 10 20 30 400

20

40

60

80

CD11c+ / B cells

CD

21- C

XCR

5- / B

cells

Dr = 0.78p < 0.0001

0

10

20

30

% C

D11

c+ C

D21

- / B c

ells

G

Plasmablasts

B Cells

B Cells

F

CD

21

CD11c

CD11c+ CD21-

Supplementary Figure 8: Tph cells correlate with CD11c+ B cells.A) Gating strategy for B cells and plasmablasts in the AMP mass cytometry data using the B cell panel. B) Gating strategy for CD11c+ B cells in the AMP mass cytometry data using the myeloid panel. C) Example visualization of CD21 and CXCR5 on B cells gated as in (A) in the AMP mass cytometry data. D) Correlation between frequency of CD11c+ B cells and CD21- CXCR5- B cells in all patients in the AMP cohort. E) Correla-tion between the frequency of Tph cells or Tfh cells and CD21- CXCR5- B cells in the AMP cohort (black = controls, green = RA patients, red = SLE patients). Spearman correlation statistics show (D,E). F) Example of detection of CD11c+ CD21- B cells in co-cultures of memory B cells from controls co-cultured with indicated CD4+ T cell subsets from SLE patients G) Quantification of CD11c+ B cells among B cells (excluding plasmab-lasts) in co-cultures as in (F). Pooled data from 6 donors in 6 different experiments. Error bars show mean±SD.

Tph cells Tfh cells

0

50

100

150

PRDM1

Naive c

ontro

l

Naive R

A

Naive S

LE

Treg c

ontro

l

Treg R

A

Treg S

LE

Tfh co

ntrol

Tfh RA

Tfh SLE

Tph co

ntrol

Tph R

A

Tph S

LE0

5

10

SOX4

Naive c

ontro

l

Naive R

A

Naive S

LE

Treg c

ontro

l

Treg R

A

Treg S

LE

Tfh co

ntrol

Tfh RA

Tfh SLE

Tph co

ntrol

Tph R

A

Tph S

LE

Naive c

ontro

l

Naive R

A

Naive S

LE

Treg c

ontro

l

Treg R

A

Treg S

LE

Tfh co

ntrol

Tfh RA

Tfh SLE

Tph co

ntrol

Tph R

A

Tph S

LE0

10

20

30

40

50BCL6

ns ns ns ** ns ns

Supplementary Figure 9: Expression of BCL6, PRDM1, and SOX4 in circulating Tph and Tfh cells. Expression of BCL6, PRDM1, and SOX4 in T cell subsets in RNA-seq data. Error bars show mean±SD. * p<0.05,** p<0.01, *** p<0.001, **** p<0.0001 by Mann-Whitney test comparing Tph cells or Tfh cells from controls and SLE patients.

We recognize participants in Phase 1 of the Accelerating Medicines Partnership RA/SLE

Network, which includes: Jennifer H. Anolik1, William Apruzzese2, Arnon Arazi3, Ami Ben

Artzi4, John P. Atkinson5, Joan M. Bathon6, Celine C. Berthier7, David L. Boyle8, Michael

B. Brenner1, S. Louis Bridges9, Jill P. Buyon10, Vivian P. Bykerk11, John A. Carrino11,

Arnold Ceponis8, Adam Chicoine2, Robert Clancy10, Sean Connery12, Andrew C.

Cordle13, Melissa Cunningham14, Maria Dall’Era15, Robert B. Darnell16, Anne Davidson17,

Evan Der18, Betty Diamond17, Laura T. Donlin11, Andrea Fava19, Andrew Filer20, Gary S.

Firestein8, Chamith Y. Fonseka2,3, Lindsy Forbess4, Richard A. Furie21, Anqi Gao2,

Daniel Goldman19, Susan M. Goodman11, Ellen M. Gravallese2, Peter K. Gregersen17,

Rohit Gupta22, Joel M. Guthridge23, Nir Hacohen3, David A. Hildeman24, Michael

Holers25, Paul J. Hoover2, Diane Horowitz21, Raymond Hsu15, Laura B. Hughes9, Mariko

L. Ishimori4, Lionel B. Ivashkiv11, Judith A. James23, A. Helena Jonsson2, Ruba Kado7,

Ken C. Kalunian8, Diane L. Kamen14, Joshua Keegan2, Stephen Kelly26, Gregory Keras2,

Mattias Kretzler7, James A. Lederer2, Peter Lowry25, Holden Maecker22, Arthur M.

Mandelin27, Elena M. Massarotti2, Eric L. Matteson28, William J. McCune7, Mandy J.

McGeachy13, Maureen McMahon29, Fumitaka Mizoguchi2, Larry W. Moreland13, Oganna

Nwawka11, Dana E. Orange11, Meyeon Park15, Deborah Parks5, Fernanda Payan-

Schober12, William F. Pendergraft30, Harris R. Perlman27, Alessandra B. Pernis11,

Michelle Petri19, Mina Pichavant22, Costantino Pitzalis31, Chaim Putterman18, Deepak A.

Rao2, Soumya Raychaudhuri2,3, Christopher Ritchlin1, William H. Robinson22, Karen

Saloman-Escoto32, Daimon Simmons2, Kamil Slowikowski2,3, Hermant Suryawanshi33,

Darren A. Tabechian1, Ralf G. Thiele1, Thomas Tuschl33, Paul J. Utz22, Dia

Waguespack34, Gerald Watts2, Kevin Wei2, Michael Weisman4, Aaron Wise13, E. Steve

Woodle35, David Wofsy15, Fan Zhang2,3.

1University of Rochester Medical Center, Rochester, New York. 2Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts. 3Broad Institute of Massachusetts Institute and Technology and Harvard University, Cambridge, Massachusetts. 4Cedars-Sinai Medical Center, Los Angeles, California. 5Washington University School of Medicine, Saint Louis, Missouri. 6Columbia University, New York, New York. 7University of Michigan, Ann Arbor, Michigan. 8University of California, San Diego, La Jolla, California. 9University of Alabama at Birmingham, Birmingham, Alabama. 10New York University School of Medicine, New York, New York. 11Hospital for Special Surgery, New York, New York. 12Paul L. Foster School of Medicine, Texas Tech University Health Sciences Center, El Paso, Texas.

13University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania. 14Medical University of South Carolina, Charleston, South Carolina.

15University of California San Francisco, San Francisco, California.

16New York Genome Center and the Rockefeller University, New York, New York. 17The Feinstein Institute for Medical Research, Northwell Health, Manhasset, New York.

18Albert Einstein College of Medicine and Montefiore Medical Center, Bronx, New York.

19Johns Hopkins University, Baltimore, Maryland.

20NIHR Birmingham Biomedical Research Centre, University Hospitals Birmingham NHS Foundation Trust and University of Birmingham, Birmingham, UK

21Northwell Health, Great Neck, New York.

22Stanford University School of Medicine, Stanford, California.

23Oklahoma Medical Research Foundation, Oklahoma City, Oklahoma, USA. 24University of Cincinnati and Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio.

25University of Colorado School of Medicine, Aurora, Colorado.

26Barts Health NHS Trust, London, UK

27Northwestern University Feinberg School of Medicine, Chicago, Illinois.

28Mayo Clinic College of Medicine, Rochester, Minnesota.

29University of California Los Angeles, Los Angeles, California.

30The Integrative Medical Clinic of North Carolina, PLLC, Chapel Hill, North Carolina.

31William Harvey Research Institute, Queen Mary University of London, London, UK

32University of Massachusetts Medical School, Worcester, Massachusetts.

33Rockefeller University, New York, New York.

34University of Texas, Houston, Texas.

35University of Cincinnati College of Medicine, Cincinnati, Ohio.