dendritic cells copyright © 2018 direct reprogramming of ...€¦ · in both dc development and...

Rosa et al., Sci. Immunol. 3, eaau4292 (2018) 7 December 2018

S C I E N C E I M M U N O L O G Y | R E S E A R C H A R T I C L E

1 of 15

D E N D R I T I C C E L L S

Direct reprogramming of fibroblasts into antigen-presenting dendritic cellsFábio F. Rosa1,2,3*, Cristiana F. Pires1,2,3*, Ilia Kurochkin4, Alexandra G. Ferreira1,2,3, Andreia M. Gomes3, Luís G. Palma3, Kritika Shaiv3, Laura Solanas3, Cláudia Azenha3, Dmitri Papatsenko4, Oliver Schulz5, Caetano Reis e Sousa5, Carlos-Filipe Pereira1,2,3†

Ectopic expression of transcription factors has been used to reprogram differentiated somatic cells toward pluripo-tency or to directly reprogram them to other somatic cell lineages. This concept has been explored in the context of regenerative medicine. Here, we set out to generate dendritic cells (DCs) capable of presenting antigens from mouse and human fibroblasts. By screening combinations of 18 transcription factors that are expressed in DCs, we have identified PU.1, IRF8, and BATF3 transcription factors as being sufficient to reprogram both mouse and human fibroblasts to induced DCs (iDCs). iDCs acquire a conventional DC type 1–like transcriptional program, with features of interferon-induced maturation. iDCs secrete inflammatory cytokines and have the ability to engulf, process, and present antigens to T cells. Furthermore, we demonstrate that murine iDCs generated here were able to cross-present antigens to CD8+ T cells. Our reprogramming system should facilitate better understanding of DC specification programs and serve as a platform for the development of patient-specific DCs for immunotherapy.

INTRODUCTIONCell reprogramming refers to the concept of rewiring the transcrip-tional and epigenetic network of one cell state to that of a different cell type. Mouse and human somatic cells can be reprogrammed into induced pluripotent stem cells by the expression of transcription factor combinations (1). Alternative somatic cell fates can also be instructed by direct lineage reprogramming using transcription factors speci-fying target-cell identity including myocytes, neurons, and hepato-cytes (2). Hematopoietic progenitors and mature cells have also been generated using this strategy (3–5).

Among mononuclear phagocytes, dendritic cells (DCs) excel in their ability to sense, process, and present antigens to lymphocytes in secondary lymphoid organs orchestrating adaptive immune responses (6). DC subsets have been described based on receptor expression, ontogeny, and function. In the bone marrow (BM), monocyte DC progenitors (MDPs) give rise to common DC pro-genitors (CDPs), which differentiate only into cells from DC lineage, such as conventional DCs (cDCs) and plasmacytoid DCs (pDCs). cDCs can be further subdivided into functionally distinct cDC type 1 (cDC1) and cDC type 2 (cDC2) cell lineages. Whereas cDC1 cells are specialized in antigen cross-presentation, initiating cyto-toxic T cell responses (7), cDC2 cells prime CD4+ T cells through major histocompatibility complex class II (MHC-II)–restricted presentation. Several transcription factors have been implicated in both DC development and subset specification (6, 8, 9). Given the pivotal role of DCs in driving immune responses, we asked whether transcription factor combinations could instruct DC identity and endow fibroblasts with professional antigen presen-tation capacity.

RESULTSScreening for DC-inducing transcription factorsTo identify candidate transcription factors to induce DC identity and their function as antigen-presenting cells (APCs), we used three complementary approaches: (i) a developed predictive computa-tional tool, GPSMatch; (ii) literature mining; and (iii) analysis of available gene expression datasets, leading to the identification of 18 candidates (table S1). Gene ontology (GO) analysis highlighted the fundamental role in DC differentiation and activation, whereas genetic perturbations were associated with immune phenotypes (fig. S1A). We also verified that the 18 transcription factors are spe-cifically enriched in DCs (fig. S1B) when compared with other tis-sues and macrophages, which are less efficient APCs (fig. S1C, left) (10) and are expressed during DC ontogeny (fig. S1C, right) (9).

We used mouse embryonic fibroblasts (MEFs) harboring aDC- specific reporter system (Clec9a-Cre X R26-stop-tdT mouse,hereafter called Clec9a-tdT) to identify DC-inducing transcriptionfactors. In Clec9a-tdT reporter mice, tdT fluorescent protein is ex-pressed in DCs and committed precursors (11). We confirmed that, within the immune system, Clec9a gene expression is restricted toDCs (fig. S2A). Clec9a expression was undetectable in MDPs andstarts in CDPs (fig. S2B) (12), reaching high levels in the cDC1 sub-set (fig. S2C) (9, 13). In agreement, >98% of splenic cDC1s expressed tdT, whereas cDC2 and pDCs showed incomplete labeling (fig. S2,D and E). Clec9a-tdT MEFs were purified by fluorescence-activatedcell sorting (FACS) to remove rare contaminating hematopoietic(CD45+) and tdT+ cells (fig. S2F) and used to screen candidates cloned individually in doxycycline (Dox)–inducible lentiviral vectors (3).MEFs were transduced with 18 transcription factors or smaller com-binations, cultured in the presence of Dox, and evaluated for tdTexpression (Fig. 1A). After transduction with 18 candidates or oneof the pools of four transcription factors, we observed the emergenceof tdT+ cells (fig. S3A). The pool comprising PU.1, IRF4, IRF8, andBATF3 generated more tdT+ cells than 18 transcription factors(2.36 ± 0.75% and 0.61 ± 0.04%, respectively), suggesting that theminimal combination required to induce Clec9a-tdT activation iscontained within this pool (fig. S3B). TdT+ cells were not detectedafter transduction with the M2rtTA vector. As an additional control,

1Molecular Medicine and Gene Therapy, Lund Stem Cell Center, Lund University, BMC A12, 221 84, Lund, Sweden. 2Wallenberg Center for Molecular Medicine, Lund University, Lund, Sweden. 3Center for Neuroscience and Cell Biology, University of Coimbra, Largo Marquês do Pombal, 3004-517 Coimbra, Portugal. 4Skolkovo Insti-tute of Science and Technology, Nobel Street, Building 3, Moscow 143026, Russia. 5Immunobiology Laboratory, The Francis Crick Institute, 1 Midland Road, London NW1 1AT, UK.*These authors contributed equally to this work.†Corresponding author. Email: [email protected]

Copyright © 2018 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works

by guest on Novem

ber 23, 2020http://im

munology.sciencem

ag.org/D

ownloaded from

http://immunology.sciencemag.org/



2 of 15

10 µm

B

F

C

+ Dox

MEF

FUW-TetO-TFs

A

E13.5

I

FUW-M2rtTA

Clec9a Cre

tdTomatoStopR26

Off

Screen fortdTomato+ cells

H

M2rtTAPIB day 0 1 2

3 4 5 6

G

d0 d1 d2 d3 d5 d7 d9 d11

d15

0

1

2

3

4

Days

Clec9aCre/CreRosatdT/tdT

PU.1 +

tdT

omat

o+ c

ells

(%

)

0

2

4

6

M2r

tTA

PIB

PIB-P

U.1

PIB-IR

F8

PIB-B

ATF3PU.1

IRF8

BATF3

tdT

omat

o+ c

ells

(%

)M

2rtT

A d

ay 9

PIB

day

9

Clec9a Cre

tdTomatoR26

On

D

M2r

tTA

PIBIR

F4

STAT3TCF4

IKZF1

ID2

BCL11ARELB

ZBTB46

RUNX3GFI1

IRF2

NFIL3BCL6

L-M

YC

NR4A3

0

2

**

tdT

omat

o+ c

ells

(ra

tio)

+PIB

****

E

M2r

tTA

PIB

GFI1B

RUNX1

NFKB1

GATA2ETV6

MEIS

1NFE2

MYB

HOXA9KLF

4

HOXB4

NFAT1

BCL11B

STAT5IR

F10

2

+PIBtd

Tom

ato+

cel

ls (

ratio

)

*******

******

***** ***

***

* *

3

1

3

1

M2rtTA PU.1 + IRF8 + BATF3

tdTomato

Aut

o

3.9%0.0%0.0%

********

***

DAPI F-actin tdTomato Merged

9 yad ATtr2

M9 yad

BIP

1.5

d0 d1 d2 d3 d5 d7 d9d11d1

50

100

200

300

400

500

tdT

omat

o+ c

ells

(#)

Fig. 1. Screening for DC-inducing factors identifies PU.1, IRF8, and BATF3 (PIB) combination. (A) Experimental design to screen DC-inducing transcription factors (TFs). Clec9a-tdTomato (Clec9a-tdT) double transgenic mouse embryonic fibroblasts (MEFs) were cotransduced with lentiviral particles encoding candidate TFs and M2rtTA and cultured in the presence of doxycycline (Dox) and monitored for tdT (red) expression. (B) MEFs transduced with M2rtTA, PU.1 + C/EBP, or PU.1 + IRF8 + BATF3 (PIB) were analyzed by fluorescent microscopy and flow cytometry 5 days after the addition of Dox. Scale bar, 200 m. (C) Quantification of tdT+ cells after removal of in-dividual TFs from the pool or their individual expression at day 5 (n = 3, mean ± SD). (D and E) Quantification of tdT+ cells after transduction with PIB combined with (D) individual TFs from the 18 candidates or (E) hematopoietic TFs at day 8 (n = 2 to 6, mean ± SD). (F and G) Kinetics of Clec9a-tdT reporter activation analyzed by (F) flow cytometry or (G) time-lapse microscopy. White arrowheads indicate an emerging tdT+ cell. Scale bar, 200 m. (H) Immunofluorescence for F-actin (green) and DAPI (blue) highlighting tdT+ cell morphology. Scale bars, 50 m. (I) Scanning electron microscopy analysis of tdT+ and M2rtTA-transduced cells. Scale bars, 10 m. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, one-way ANOVA with Bonferroni’s test.

by guest on Novem


munology.sciencem

ag.org/D

ownloaded from




3 of 15

we tested C/EBP and PU.1 described to induce macrophage-like cells (4). As expected, no Clec9a-tdT reporter activation was observed de-spite the induction of 5.5% CD45+ cells (Fig. 1B and fig. S3C).

Inducing DC fate with a minimal transcription factor networkWe next removed each of the factors from the pool of four tran-scription factors (fig. S3D). PU.1, IRF8, and BATF3 (PIB) removal reduced reporter activation, whereas removal of IRF4 did not have an impact, suggesting that PIB are essential for DC reprogram-ming. Indeed, removal of PU.1, IRF8, and BATF3 from the PIB combination completely abolished reporter activation, whereas their individual expression was not sufficient to generate tdT+ cells (Fig. 1C). The reprogramming efficiency estimated by reporter ac-tivation was 4.26 ± 0.44% and was not increased by the presence of cytokines involved in DC specification (6), stimulation with lipo-polysaccharide (LPS), different media compositions (fig. S3E), or coculture with stromal cells (fig. S3F) (14). In addition, survival of tdT+ cells in culture was not improved in the presence of Flt3l (fig. S3G). Then, we assessed the impact of adding transcription factors from the initial candidate pool (Fig. 1D) and other hematopoietic regulators (Fig. 1E). Addition of TCF4 increased reporter activa-tion (2.2-fold) (Fig. 1D) but could not replace the action of PU.1, IRF8, or BATF3 (fig. S3H). Thus, we focused on the PIB combina-tion sufficient to induce reporter activation. TdT+ cells started to be detected between days 1 and 2 and peaked between days 5 and 7 (Fig. 1F). Using time-lapse microscopy, we observed that reporter activation occurs after ~30 hours and that tdT+ cells exhibited morphology changes, migration capacity, and dendrites that were gradually established within 6 days (Fig. 1G and movie S1). We used these data to quantify cell division, which represents a rare event (3 of 155 and 4 of 159 before and after reporter activation, respectively) and therefore may not be strictly required for the re-programming process. TdT+ cells displayed increased side scatter (fig. S3I), consistent with the observed stellate morphology and the establishment of dendrites, while losing the initial fibroblast mor-phology and F-actin fibers (Fig. 1, H and I, and fig. S4). Thus, we identified PIB as the minimal transcription factor network neces-sary to induce Clec9a-tdT reporter activation and acquisition of DC morphology.

Expression of cDC1 and APC surface markersWe then asked whether activation of the Clec9a-tdT reporter was reflected in the expression of DC markers at the cell surface. The cDC1 markers XCR1 and CD103 are expressed in 31.6 and 27.3% of tdT+ cells, respectively, suggesting the specification of a cDC1-like program (Fig. 2A). Moreover, we could detect only residual expres-sion of cDC2, pDC, macrophage, or monocyte-derived DC markers (Fig. 2B and fig. S5A). The in vivo combined expression of PIB is mostly enriched in cDC1 cells (fig. S5B). Spi1, which encodes PU.1, is equally expressed in MDPs, CDPs, and pre-DCs, whereas Irf8 ex-pression markedly increases in CDPs and Batf3 is up-regulated later at the pre-DC stage (fig. S5C). Spi1 levels are higher in both pre-DC subsets, whereas Irf8 and Batf3 are mostly enriched in pre-cDC1 and cDC1 cells (fig. S5D). When compared with other mononuclear phagocytes, Clec9a, Spi1, Irf8, and Batf3 display combined enrich-ment in the cDC1 subset (fig. S5E). Thus, these data suggest that PIB promotes reprogramming of fibroblasts toward DC-like cells that express bona fide cDC1 markers.

Then, we evaluated the expression of MHC molecules, key for the establishment of APC functionality; 56.7 and 71.4% of tdT+ cells at day 7 expressed MHC-I and MHC-II at the surface, respectively, whereas the tdT− compartment contained a lower percentage of MHC-I+ and MHC-II+ cells (11.2 and 14.2%, respectively) (Fig. 2C). This population of cells may represent partially reprogrammed cells or inefficient Cre-mediated deletion. The expression of MHC-II in the tdT+ compartment was gradually acquired starting at day 2 and peaked between days 7 and 11 (Fig. 2D). By excluding each of the PIB factors individually, we observed that PU.1, but not IRF8 or BATF3, is essential for MHC-II expression (Fig. 2E), consistent with the de-scribed role in controlling the master regulator of MHC class II genes transactivator (CIITA) (15, 16). As IRF4 induces MHC-II expression in B cells but not in DCs (16), we asked whether IRF4 would com-pensate for the loss of PU.1 during reprogramming. Inclusion of IRF4 did not increase MHC-II expression or compensate for the loss of PU.1 (Fig. 2F). We then evaluated expression of the costimulatory molecules required for efficient antigen presentation. CD80 and CD86 were coexpressed in 35.2% of tdT+ MHC-II+ cells (Fig. 2G); 8.6% of tdT+ also coexpressed CD40 and MHC-II, a population that increased 3.8-fold (32.4%) after Toll-like receptor 4 (TLR4) stimu-lation with LPS (Fig. 2, H and I); 22.8% of tdT+ MHC-II+ CD40+ cells coexpressed CD86 (Fig. 2I), a response that resembles BM- derived DCs (BM-DCs) upon stimulation (fig. S5F). These data strongly support that PIB reprogram fibroblasts into induced DCs (iDCs).

Transcriptional reprogramming of iDCsTo define the transcriptional changes during iDC reprogramming, we generated full-length single-cell transcriptomes (Fig. 3A). One hundred sixty-three individual cells were profiled from nontrans-duced MEFs; sorted iDCs at days 3, 7 and 9; and freshly isolated splenic cDC1 cells. t-Distributed stochastic neighbor embedding (t-SNE) analysis of genome-wide transcriptomes revealed two main clusters of cells (Fig. 3B). iDCs at day 3 and the majority of day 7 cells mapped close to MEFs, whereas a small percentage of day 7 and all day 9 iDCs clustered together with cDC1. TdT+ cells showed progressive transcriptional changes starting at day 3. At day 9, iDCs are notably similar to bona fide DCs as confirmed by principal components analysis (PCA) and hierarchical clustering (fig. S6, A and B). We next extracted the most variable genes across the dataset, which could be organized in four clusters (Fig. 3C). Cluster I com-prised highly expressed genes in MEFs that are silenced during DC reprogramming (Fig. 3D) (3). Cluster II included transcripts acti-vated early during reprogramming (days 3 to 7) and associated with intracellular trafficking and type I interferon (IFN) signaling (Eea1 and Ifit3, respectively) (Fig. 3E). Cluster III encompassed genes en-riched at day 9 iDCs (Fig. 3E). MHC-II and cross-presentation genes [H2-Pb, Ctsc, and Cd74 (17)] are enriched in this cluster (Fig. 3E). Accordingly, top biological process GO and pathway analysis showed enrichment for antigen processing and presentation (Fig. 3F and fig. S6C). Top cellular component GO categories included lysosome and lytic vacuole, in agreement with the described role of lysosome signaling in coordinating antigen processing and migration of DCs (18). MicroRNA target prediction showed enrichment of miR-155 and miR-124 targets (fig. S6C), previously implicated in DC function and specification. Last, cluster IV included genes enriched in cDC1s and iDCs (Fig. 3E), such as Ptprc and Itgax, encoding CD45 and CD11c, respectively, and Ccr2 chemokine involved in DC migration (Fig. 3E).

by guest on Novem


munology.sciencem

ag.org/D

ownloaded from




4 of 15

In addition, Ccr1, Ccr5, and Ccr7 were also up-regulated in repro-grammed cells at days 7 and 9 (fig. S6D). Cluster IV also included up-regulated cDC1-restricted genes, such as Clec9a, Tlr3 (7), and Nlrc5 (19). We detected a robust increase in splenic cDC1 signature genes during reprogramming (Fig. 3G) (12) as well as cDC1-like Flt3l and Flt3l/Notch BM-DC genes (fig. S6E) (20). Collectively, these data suggest that PIB-mediated DC reprogramming favors the cDC1 subset.

This analysis also revealed activation of known DC transcriptional regulators, including Zbtb46, Bcl11a, Ikzf1, and Bcl6, included in the candidate transcription factor list (Fig. 3H). ZBTB46 was proposed as a selective marker for distinguishing cDCs from other tissue phago-cytes (21). Endogenous Spi1, Irf8, and Batf3 expression started to be detected at day 3 (Fig. 3I), reaching comparable levels to cDC1 cells at day 9, which suggests that a stable DC fate has been acquired. Then, we performed gene set enrichment analysis (GSEA) to compare the

dynamic pathway transitions between days 0, 3, 7, and 9 (fig. S7A). At day 3, 29 immune pathways were enriched, with interleukin-4 (IL-4) ranking on top, whereas at day 9, 24 gene sets were enriched, including multiple IL pathways and Oncostatin M associated with DC maturation. We then mapped the transcription factor networks involved in these transitions (fig. S7B). A dense network of 56 transcription factors highly connected to PIB at the early phases of reprogram-ming evolved to a network of 104 transcription factors that includes Zbtb46. These networks are enriched in mature DCs but not in DC progenitors, irrespective of reprogramming stage (fig. S7C). Con-cordantly, no colonies were observed in methylcellulose assays with PIB-transduced MEFs between days 3 and 25 and sorted cDC1 cells (fig. S7D). Then, we investigated whether the reprogramming pro-cess induces genomic instability and tumorigenic events. Day 9 iDC cultures displayed a normal karyotype (fig. S8A). Moreover, expres-sion of Myc, Kras, and Mdm2 oncogenes was down-regulated during

PIB tdT– PIB tdT+

MH

C-I

MH

C-I

I

tdTomato

M2rtTA

1.5% 14.2% 71.4%

FED

46.6%

53.4%

MH

C-I

I

tdTomato

CD80

CD

86

10.0% 12.9%

20.0%57.1%

21.5% 35.4%

19.0%24.1%

PIB tdT+

MHC-II+

MHC-II–

M2r

tTA

PIB

5

10

15

20

MH

C-I

I+ c

ells

(%

)

PIB tdT–

0

20

40

60

0

MH

C-I

I+ c

ells

gat

ed td

T+ (

%)

**

PIB tdT+

G

PIB + IRF4M

2rtT

A PIB

PIB+I

RF4

H

0.1% 11.2% 56.7%

1.6%

CD40

8.6%

2.8%MH

C-I

I

PIB tdT+ LPS

13.0%

CD40

32.4%

MH

C-I

I

4.8%

CD

86

MHC-II+ CD40+

22.8%MHC-II– CD40+

CD40

*

XCR1

Mod

al

tdT+ CA

I

27.3%

0 103

104

105

0

20

40

60

80

100tdT–

31.6%

0 103

104

105

0

20

40

60

80

100CD103

B

1.3%

CD4

CD11b

B220

F4/80

CD640

5

10tdT+

tdT–

d0 d1 d2 d3 d5 d7 d9d11d1

3d1

50

20

40

60

80

100 PIB tdT+

PIB tdTM2rtTA

MH

C-I

I+ c

ells

(%

)

Sur

face

mar

ker+

cel

ls (

%)

PIB

-PU.1

-IRF8

-BATF3

-PU.1

-IRF8

-BATF3

–

Fig. 2. Induced DCs express classical APC surface molecules and cDC1 receptors. (A) Flow cytometry analysis of CD103 and XCR1 in tdT+ and tdT− populations of PIB-transduced MEFs 8 days after the addition of Dox. (B) Quantification of CD8, CD4, CD11b, B220, F4/80, CD64, and SIRP expression in tdT+ and tdT− populations (n = 2, line indicates mean). (C) MHC-I and MHC-II expression in MEFs transduced with M2rtTA and PIB at day 7. (D) Kinetics of MHC-II surface expression. (E) Quantification of MHC-II+ cells after removal of individual transcription factors from PIB at day 5 (n = 3 to 4, mean ± SD). (F) Quantification of MHC-II+ cells at day 5 within tdT+ population after transduction with PIB, PIB and IRF4, or upon their individual removal (n = 2 to 3, mean ± SD). (G) CD80 and CD86 expression in tdT+ MHC-II+ and tdT+ MHC-II− populations at day 5. (H) CD40 and MHC-II expression at day 7 before or (I) after overnight LPS stimulation. CD80 expression in MHC-II+ CD40+ (red) and MHC-II− CD40+ (blue) populations. *P < 0.05, **P < 0.01, one-way ANOVA with Bonferroni’s test.

by guest on Novem


munology.sciencem

ag.org/D

ownloaded from




5 of 15

BA

DC

–10

–5

0

5

10

t-SNE 1

t-S

NE

2

MEFd3 iDCd7 iDCd9 iDCcDC1

Genome-wide t-SNE

F

TranslationGeneration of metabolites and energyElectron transport chainIntracellular transportEstablishment of protein localization

Chromatin modificationProtein transportEstablishment of protein localizationMacromolecule catabolic processProtein catabolic process

Acta2

010

020

030

0 Fbn1

01

2

05

1015

Lox

20

Grem1

00.

20.

40.

60.

8

Exp

ress

ion

02

46

8

Col5a2 Thy1

020

4060

80

ME

F d3 d7

cDC

1 d

9

ME

F d3 d7

cDC

1d9

ME

F d3 d7

cDC

1d9

ME

F d3 d7

cDC

1d9

ME

F d3 d7

cDC

1d9

ME

F d3 d7

cDC

1d9

Cluster I

0246

Eea1

0

2

4

6 Aldh1a2

02

4

6Ifit3

0.02.55.0

7.5

Ctsc

0

5

10

Cd74

0

24

6

III retsulCII retsulC

0

1

2

3Itgax

0

2

4

6Ccr2

02468

Tlr3

0

2

4

6Clec9a

0

2

4

6Nlrc5

ME

F d3 d7

cDC

1d9

ME

F d3 d7

cDC

1d9

ME

F d3 d7

cDC

1d9

ME

F d3 d7

cDC

1d9

ME

F d3 d7

cDC

1d9

02468

ME

F d3 d7

cDC

1d9

Cluster IV

210–1

MEF d3 d7 cDC1d9

ostnGrem1Col11a1

Fibin

Col6a2

Vim

Acta2

Hoxa10

Thy1Fbn1Col5a2Tnc

Lox

–2

IV

DC reprogramming

Ccr2

Il16

Fcgr2b

Tlr3

Nlrc5

Clec2d

Sema3a

Tyro3

H2-m3H2-dmb2

Ctsc

Eea1Aldh1a2Ifit3C

lust

er

Zbtb46

02468

Bcl11a

0246

Irf7

0246

0246

Stat28Exp

ress

ion

(log 2)

G I

Antigen presentation of peptide antigenActin cytoskeleton organizationAntigen presentation of exo. peptideAntigen presentation of exo. antigen

Cell cycleChromosome organization

Reg. GTPase-mediated signal transd.DNA metabolic processReg. Ras protein signal transduction

IVIII

III

0

5

10

15

0

369

Endo

5

10

15Irf8

2468

Endo Irf8

0

5

10

Batf3

0.02.5

5.07.5

Endo Batf3

0 0

Clu

ster

Exp

ress

ion

(log 2)

Exp

ress

ion

(log 2)

0 5–5–10

d3d0

PI B

MEF

d7 d9

Full transcript single-cell RNA-seq

iDCs

C57BL/6 Spleen cDC1

Cell sorting

MHC-II+

Clec9a-tdT+

Cell sorting

CD11c+

MHC-II+

CD8a+

E

ME

F d3 d7

cDC

1d9

0246 Ikzf1 Bcl6

02468

ME

F d3 d7

cDC

1d9

III

III

Ribonucleoprotein complexRibosomeMitochondrionMitochondrial partRibosomal subunit

Membrane-enclosed lumenIntracellular organelle lumenOrganelle lumenEndoplasmatic reticulumNon-memb. bound organelle

VacuoleLysosomeLytic vacuoleCytosolEndoplasmatic reticulum

CytoskeletonNon-mem. bound organelleIntracellellular organelleEndomembrane systemMicrotubule cytoskeleton M

EF d3 d7

cDC

1d9

ME

F d3 d7

cDC

1d9

Biological processes Cellular components

Antigen processing and presentation

0.00.10.20.30.4 cDC1

cDC2

HMEF d3 d7 d9

Exp

ress

ion

Fig. 3. PIB induce global cDC1-like gene expression program. (A) PIB-transduced MEFs were FACS sorted and profiled using full-length transcript single-cell mRNA-seq at day 3 (tdT+), day 7 (tdT+), and day 9 (tdT+ MHC-II+). MEF and splenic cDC1 cells (CD11c+ MHC-II+ CD8+) were used as controls. (B) t-SNE analysis of genome-wide tran-scriptomes showing clustering of 163 single cells. Each dot represents an individual cell (27 MEFs; 16 day 3, 30 day 7, and 29 day 9 iDCs; and 61 cDC1s). (C) Heat map showing expression of the 6525 most variable genes and four gene clusters. (D) Expression levels of fibroblast-associated genes shown as Census counts median values ± 95% confidence interval. (E) Violin plots showing expression distribution of genes in clusters II, III, and IV. Log values of Census counts are shown; horizontal lines corre-spond to median values. (F) Top five GO biological processes (left) and cellular components (right). (G) Cumulative median expression levels of cDC1 and cDC2 gene signatures. (H) Violin plots showing expression distribution of DC transcriptional regulators. (I) Total (left) and endogenous (right) expression of Spi1, Irf8, and Batf3 are shown as log counts presented as box plots.

by guest on Novem


munology.sciencem

ag.org/D

ownloaded from




6 of 15

the reprogramming process, while cell cycle arrest genes Cdkn1a, Trp63, and Tsc1 were induced (fig. S8, B and C). Together, these data support the idea that the DC reprogramming process is direct, does not transit through an intermediate progenitor state, and is not asso-ciated with tumorigenesis.

We next reconstructed the DC reprogramming path by estab-lishing a pseudotemporal order based on the gradual transitions of single-cell transcriptomes (Fig. 4A and fig. S9A). We observed three branches where MEFs, all day 3 iDCs, and 26 day 7 iDCs were placed along the first branch (state 1) and ordered consistently with the re-programming timeline captured by component 1 (Fig. 4B). Single- cell trajectory reconstitution identified a branching point, and samples were divided into states 2 and 3, captured by component 2. State 2 includes 82% of cDC1s, 3 day 7 and 13 day 9 iDCs. Conversely, 55% of day 9 iDCs and 11 cDC1s are placed within state 3 (Fig. 4B). GO and pathway analysis revealed enrichment in state 3 for type I and type II (IFN-) IFN signaling (Fig. 4C), inflammatory mediators of DC activation and maturation (22, 23). Branched expression analysis modeling (BEAM) also identified two kinetic clusters of branch- dependent genes up-regulated in state 3 (clusters II and IV) function-ally enriched in antigen presentation and immune- related processes (fig. S9, B and C), suggesting that the two states are indicative of different maturation profiles. Accordingly, branch kinetic curves re-flect a continuous up-regulation of Ciita and H2-Eb1 toward state 3 (Fig. 4D), as well as antigen processing (Lgmn) and initiation of T cell response (Tnfrs1a) related genes. Because state 3 contains most of the day 9 iDCs, we sought to confirm that similar matura-tion traits were observed when comparing cDC1s (naïve) with day 9 iDCs. GSEA for immunologic signatures showed 774 gene sets en-riched on day 9 iDCs, including IFN-– and IFN-–stimulated DC gene sets (Fig. 4E). Antigen processing and presentation (e.g., Lgmn and Tabpb) and DC maturation (e.g., Lgal9 and Ptpn1) gene sets were also enriched in day 9 iDCs (fig. S9D). Top biological processes GO and pathway analysis for the differentially expressed genes between day 9 iDCs and cDC1s highlighted the enrichment for antigen pro-cessing and presentation in day 9 iDCs (Fig. 4F). Stat6 associated with immature DCs is up-regulated in cDC1s, whereas Stat1, described to increase with maturation, is up-regulated in day 9 iDCs (Fig. 4G) (24). Furthermore, the high levels of Ciita expression were driven mainly by promoter I in day 9 iDCs, a feature of cDCs (Fig. 4H). Collectively, these data suggest that iDCs are intrinsically more mature than nonstimulated cDC1 cells, recapitulating a transcrip-tion profile of IFN-stimulated DCs.

Functional reprogramming into APCsWe confirmed that iDCs express Tlr3 (Fig. 3E), Tlr4, and the medi-ators of TLR signaling Ticam2, Traf6, and Ikk (fig. S10A). Accord-ingly, upon TLR3 [polyinosinic-polycytidylic acid (PIC)] or TLR4 (LPS) (Fig. 5A) challenge, we observed an increase in IL-6 secretion (2.1- and 3.1-fold, respectively) by purified populations of tdT+ iDCs at day 9. Tumor necrosis factor– (TNF-) secretion also increased 1.3- and 1.4-fold after PIC and LPS stimuli. In contrast, iDCs did not secrete IL-10 with or without stimulation. M2rtTA-transduced MEFs only secreted IL-6 upon stimulation with bacterial LPS. Similar re-sults were observed in nonpurified PIB-transduced cells (fig. S10B). We observed that 31.2% of iDCs produce IL12p40 in response to LPS/profilin/CD40L stimulation (Fig. 5B). These results suggest that iDCs underwent maturation toward a proinflammatory profile and support the cDC1 affiliation of PIB-induced cells.

We also confirmed that iDCs express key mediators of receptor- mediated endocytosis (Fcgr2b, Tfr2, and Mrc1) and macropinocytosis of dead cells (Axl, Lrp1, and Scarf1) (fig. S10C). Thus, we evaluated the capacity of iDCs to capture exogenous antigens. First, we con-firmed the ability of iDCs to engulf fluorescein isothiocyanate– labeled latex beads (fig. S10D) and fluorescent-labeled ovalbumin (OVA) (fig. S10E), suggesting that iDCs have established the com-petence for phagocytosis. Next, we evaluated whether iDCs were able to internalize dead cell material in vitro (Fig. 5, C and D, and fig. S10F), a key feature of cross-presenting DCs (25). After over-night incubation with labeled dead cells, 65.7% of purified tdT+ cells incorporated dead cell material (Fig. 5C). Efficient uptake was confirmed by time-lapse microscopy (Fig. 5D and movie S2), high-lighting projected cellular protrusions to incorporate and engulf dead cells.

We then addressed the functional capacity of iDCs to promote antigen-specific proliferation and activation of CD4+ T cells (Fig. 5E) using OT-II CD4+ T cells expressing a T cell receptor specific for the OVA 323-339 peptide presented in the context of MHC-II and splenic cDCs as control. When given OVA protein in the presence of LPS, iDCs display comparable ability to induce proliferation and CD44 expression in OT-II T cells when compared with cDCs (52.2% versus 63.8%) (Fig. 5, E and F, and fig. S10G). As expected, the same was observed with the preprocessed peptide. In the absence of pro-inflammatory stimulation, 42.8 ± 8.84% of CD4+ T cells diluted car-boxyfluorescein diacetate succinimidyl ester (CFSE) when cocultured with iDCs. This ability of iDCs to induce antigen-specific T cell responses even in the absence of stimuli further supports the IFN- matured state of iDCs (Fig. 4).

We then assessed cross-presentation to CD8+ T cells, a key func-tional feature of cDC1s. We confirmed that iDCs express key regu-lators of the cross-presentation process, including Cybb, Atg7, Tap1, Tap2 (fig. S10H), and Cd74 (Fig. 3E) (26). We evaluated antigen ex-port from endosomes to the cytosol, a feature of cross-presentation via the cytosolic pathway, using a cytofluorimetry-based assay (Fig. 5G) (27). After 90-min incubation with -lactamase, about 80% of CCF4-loaded iDCs contained cleaved CCF4 (fig. S10I). tdT+ iDCs displayed similar levels of cleaved CCF4 to Flt3l/granulocyte- macrophage colony-stimulating factor (GM-CSF) BM-DCs (Fig. 5H), confirming that iDCs have the ability to uptake -lactamase and export it into the cytoplasm, leading to the generation of cleaved CCF4. We then evaluated cross-presentation of OVA in the context of MHC-I molecules by coculturing iDCs with B3Z T cell hybridoma cells. iDCs induced antigen-specific T cell activation in a concentration- dependent manner (Fig. 5I, left) that was further increased with PIC but not with LPS stimulation (Fig. 5I, middle). Accordingly, it has been described that maturation of cDC1s with PIC enhances cross- presentation (28). After OVA pulse incubation, purified tdT+ cells induced T cell activation at comparable levels to BM-DCs (Fig. 5I, right). MEFs were included as control and did not induce T cell activation.

Last, we evaluated cross-presentation ability using MHC-I–restricted OVA-specific T cells (OT-I cells) after a short pulse with OVA protein (Fig. 5J and fig. S10J). In the presence of PIC, 47.7% of CD8+ T cells diluted CellTrace Violet (CTV) and up-regulated CD44 expression after 3 days of coculture with purified iDCs (sorted tdT+ at day 7). Splenic cDC1s induced 84.5% of proliferative and ac-tivated T cells. When quantifying proliferation, coculture with pu-rified tdT+ cells induced 79.2 ± 3.7% of proliferative CD8+ T cells

by guest on Novem


munology.sciencem

ag.org/D

ownloaded from




7 of 15

when compared with 93.8 ± 1.9% and 99.1 ± 0.2% for cDC1s and BM-DCs, respectively (Fig. 5K and fig. S10K). Residual levels of stimulation were detected without antigen or by OVA-pulsed MEFs and tdT− cells. Together, these results show that reprogrammed

DCs with PIB are efficient in capturing, processing, and per-forming presentation and cross-presentation of exogenous antigens in the context of MHC-I and MHC-II, key features of functional cDC1 cells.

Fig. 4. Reconstruction of single-cell reprogramming trajectory highlights different maturation states of iDCs. (A) Pseudotime ordering of single cells (line) in a two-dimensional independent component space during reprogramming. MEF; iDC at day 3, day 7, and day 9; and cDC1 are shown. (B) Individual cells are colored by cell state. The number of cells in each cell state is depicted inside parentheses. (C) Top five GO biological processes (BP), mouse phenotypes, and pathway enrichment analysis of genes differentially expressed between state 2 and state 3. (D) Gene expression levels of Ciita, H2-Eb1, Lgmn, and Tnfrsf1a in single cells from the three cell states. (E) GSEA between day 9 iDCs and cDC1 was performed against the immunologic signatures collection. Gene sets were ordered by normalized enrichment score (NES). False discovery rate (FDR) q values are shown (<0.25). Black lines represent DC gene sets. Bottom right panel shows enrichment plots for IFN-stimulated DC gene sets. (F) Top five GO BP, cellular components (CC), and pathway enrichment analysis of genes differentially expressed between day 9 iDC and cDC1. (G and H) Violin plots showing expres-sion distribution of (G) the maturation regulators Stat1 and Stat6 and (H) Ciita with corresponding normalized promoter usage in day 9 iDC and cDC1 (right).

Com

pone

nt 2

Component 1

Exp

ress

ion

(log 10

)

Branch kinetic curvesState 2 State 3

State1 32

BA

1

–20

–10

0

10

20

–40 –20 0 20

3etatS1etatS State 2

MEF d3 d7

d7 d9 cDC1

d7 d9 cDC1

(27)(16)(26)

(1)(16)(11)

(3)(13)(50)

E

G

0

5

Stat1Stat6

0

2

4

6

Exp

ress

ion

(log 2)

C

Pat

hway

s

Estrogen signaling pathway

Regulation of actin cytoskeleton

Breast cancer

MAPK signaling

Gastric cancer

G13 signaling

Alzheimer's disease

heart hypertrophy

IL-3 signaling pathway

Phe

noty

pes Lethality at weaning

Premature death

Abnormal mineral homeostasis

Mammalian phenotype

Abnormal startle reflex

Abnormal antigen presenting cell

Abnormal immune system

Abnormal blood cell

Abnormal response to infection

Abnormal bone marrow

GO

BP

Phosphatidylinositol dephosph.

Regulation of translesion synthe.

Actin filament capping

Actin filament reorganization

Peptidyl-serine autophosph.

signaling

Response to

Cellular response to

Neutrophil degranulation

Type I interferon signaling

3etatS2etatS

MEFd3d7d9 cDC1

1

–20

–10

0

10

20

–40 –20 0 20Component 1

Com

pone

nt 2

D

Ciita

110

100

Lgmn

0 25 50 75 100

110

100

Pseudotime

Tnfrsf1a

110

100

Reprogramming

H2-Eb1

110

100

d9 iDC cDC1

d9 iDC

Enrichment profile Hits

vs

0 2

FDR q value

NES1 0

Enr

iche

d ge

ne s

ets

(imm

unol

ogic

sig

natu

res)

cDC1

Enr

ichm

ent s

core

(ES

) 0.20.1

0

NES: 1.53 FDR q: 0.01

0.2

0.1

0

NES: 1.25 FDR q: 0.09

0

2

4

6

8Ciita

Exp

ress

ion

(log 2)

H P1 P3 P4

Pro

mot

er u

sage

ME

F d3 d7

cDC

1d9

ME

F d3 d7

cDC

1d9

ME

F d3 d7

cDC

1d9

FAntigen processing and presentationIntracellular transportProtein localizationProtein transportEstablishment of protein localization

VacuoleLysosomeLytic vacuoleEndoplasmic reticulumEndomembrane system

LysosomeAntigen processing and presentationFocal adhesionB cell receptor signaling pathwayLeukocyte transendothelial migration

Actin filament-based processCell morphogenesisRegulation of small GTPaseResponse to DNA damageCell projection organization

Non–membrane-bounded organelleIntracellular organelleCytoskeletonCell junctionPlasma membrane part

RibosomePathways in cancerWnt signaling pathwayGliomaThyroid cancer

syawhtaPCCOGPBOG

d9 iD

CcD

C1

0

50

100cD

C1d9

by guest on Novem


munology.sciencem

ag.org/D

ownloaded from




8 of 15

Increased reprogramming efficiency with polycistronic vectorsWe hypothesized that reprogramming efficiency may be limited by cotransduc-tion of MEFs. Thus, we generated two alternative polycistronic vectors encod-ing the three reprogramming factors separated by self-cleaving peptides in different orders: Spi1 followed by Irf8 and then Batf3 [PIBpoly (PIB polycistronic)] or Irf8 followed by Spi1 and Batf3 [IPBpoly (IPB polycistronic)] (Fig. 6A). We ob-served an increase in the reprogramming efficiency for both vectors, reaching 10.08 ± 0.75% of tdT+ cells with PIBpoly (Fig. 6B). The total MHC-II+ population and the proportion of tdT+ cells con-tained within were also increased 7.2- and 3.8-fold with PIBpoly (Fig. 6C). We con-firmed that this difference in reporter activation was correlated with higher levels of expression for the first gene in the polycistronic cassette (Fig. 6D), indi-cating that high levels of PU.1 are impor-tant for DC reprogramming efficiency and Clec9a reporter activation.

Then, to confirm that the polycis-tronic vector induced DC transcriptional reprogramming, we performed popula-tion mRNA sequencing (mRNA-seq) ex-periments of sorted tdT+ MHC-II+ CD45+ cells isolated at reprogramming days 5, 7, 8, and 9 (Fig. 6E and fig. S11). MEFs, splenic cDC1s, cDC2, and pDCs were in-cluded as controls. First, clustering analy-sis between differentially expressed genes between the three DC subsets was per-formed to define cDC1, cDC2, and pDC signature genes (fig. S11A). When in-tersecting the genes up-regulated during reprogramming with the subset signa-ture gene list, day 9 iDCs shared 58% of the cDC1-specific genes and only 38 and 28% of the cDC2 and pDC genes, re-spectively (fig. S11B). Correlation matrix analysis also highlighted that iDCs’ tran-scriptional profile maps closer to cDC1 than to the other DCs subsets and MEFs (fig. S11C). These results are consistent with activation of the cDC1-specific genes Xcr1, Clec9a, Rbpj, Ifi205, and Naaa and the lack of expression of the cDC2 and pDC markers Sirpa, Esam, Cd4, and Siglech (fig. S11D). Together, these data confirm the cDC1-like affilia-tion of iDCs. In addition, by profiling the same surface phenotype at different time points, we sought to characterize

CFSE

iDC

52.2%

S

PL +

AV

O

CD4+ T-cells

0.1%

0.2%

CD

44

CFSE

+ L

PS

- LP

S

SPL

+ editpeP

OVA 323-339 OVA

LPS + +––

––

+ + - + + - -+ +- - - - -

iDC

- + + +

OT

-II C

D4+

CF

SE

low

T c

ells

(%

)

cDC

20

40

60

80

100

iDC

cDC

iDC

cDC

F

CA

Cel

lVue

far

red

tdTomato

tdT

+td

T–

0%10.5%

0%89.5%

65.7%1.0%

27.0%6.2%

2.50

5 7.5

10 12.5

63.8%

60.9%cDC

7 days

CD4+

T celliDC +

OT-II/Rag2KO

E57.6%

D

+/– LPS

IL-6

(ng/

ml)

+LPS

+PIC

+LPS

+PIC

MEF tdT+

ME

F

LPS/profilin/CD40LUnstimulated

tdTo

mat

o

IL12p40

31.2%2.8%

0.0%0.0%

tdT

+

TN

F-

(pg/

ml)

B

0

50

100

OT-

I CD

8+ CTV

low

T c

ells

(%

)

G

MEF

iDC +

LPS

+PIC

1.0

1.5

2.0

2.5

T c

ell a

ctiv

atio

n(A

bs)

CD

44N

o an

tigen

+ O

VA

pul

seC

D4

C4

0.0%

0.5%

0.0%

0.0%

CTV

0.4%

0.0%

47.7%

22.1%

0.0%

2.0%

84.5%

5.6%

J

3 days

CD8+

T celltdT+

+

OT-I

+PIC+OVApulse

MEF tdT+ cDC1 K

Cleaved CCF4CCF4

Export to cytosol

CCF4

Cle

aved

CC

F4

0′ 30′ 0′ 30

′0

20

40

60

80

Cle

aved

CC

F4 (%

)

BM-DC

tdT+ BM-DC0 min

1.4%30 min

57.7%0 min

6.8%30 min

62.2%

00.2

5 0.5 1

1.5

2.0

2.5 iDCs

I

1.0

1.5

2.0 No antigenOVA pulse

H

MEF

tdT+

BM-D

C

tdT+

IL-1

0(p

g/m

l)

-

0

00.51.01.52.0

0100200300400

020406080

100

- -

– – – – ––

– – –

+

Fig. 5. Induced DCs are functional APCs. (A) Cytokine secretion of sorted tdT+ cells after stimulation with LPS or PIC and (B) IL12p40 expression after stimulation with LPS/profilin/CD40L. (C) Purified tdT+ cells were incubated with CellVue far red–labeled dead cells or (D) DAPI-labeled dead cells and analyzed by time-lapse microscopy for 12.5 hours. Scale bar, 50 m. (E and F) iDCs at day 8 and splenic CD11c+ MHC-II+ (cDC) were cocultured with CFSE-labeled OT-II Rag2KO CD4+ T cells. CD44 activation and CFSE dilution were quantified (n = 2 to 4, mean ± SD). (G and H) -Lactamase’s export to cytosol of iDCs at day 7 (gated in tdT+) and BM-DCs measured as CCF4 cleavage. (I) iDCs at day 16 were cocultured with B3Z T cell hybridoma for 18 hours with increasing concentrations of OVA protein (left) and LPS or PIC stimulation (middle). Purified tdT+ cells at day 8 and BM-DCs were pulsed with OVA for 10 hours before coculture with B3Z in the presence of PIC (right). (J and K) MEFs, purified tdT+ and tdT− cells at day 8, BM-DCs, and cDC1 were pulsed with OVA for 10 hours before coculture with CTV-labeled OT-I CD8+ T cells. CD44 activation and CTV dilution were quantified (n = 3, mean ± SD).

by guest on Novem


munology.sciencem

ag.org/D

ownloaded from




9 of 15

transcriptional transitions at later stages of reprogramming. PCA revealed that a transcriptional shift occurred from days 5 to 7 (fig. S11E), whereas day 7, day 8, and day 9 iDCs clustered closer together, suggesting more subtle transitions. Top biological process and cellular component GO categories include defense and inflammatory responses and plasma membrane and cell surface, respectively (fig. S11F). The major transcriptional shift occurs from MEFs to day 5 tdT+ MHC-II+ CD45+ cells. GO analysis for this transition revealed enrichment for immune response and antigen presentation.

Then, we have used the PIBpoly vector to transduce wild-type MEFs that do not carry any reporter (Fig. 6F). At day 9, about 8% of PIB-transduced cells express MHC-II at the cell surface. We also

detected CD45 expression in the MHC-II+ population, indicating that iDC reprogramming occurs and can be traced in the absence of a reporter. We have observed a similar surface phenotype in wild-type adult tail-tip fibroblasts (TTFs), showing that adult cells are also permissive for iDC reprogramming (Fig. 6G).

Induction of human DCsLast, since the expression of CLEC9A, SPI1, IRF8, and BATF3 was also enriched in human DC type 1 (hDC1) cells (fig. S12A) (29), we asked whether the transcription factor network governing DC fate was con-served between mouse and human. Human embryonic fibroblasts (HEFs) were transduced with PIBpoly (Fig. 7A). After 9 days, human cells acquired DC morphology with cytoplasmic ruffles and protusions (Fig. 7B). At day 3, CD45+ cells were already detected, suggesting that CD45 is an early marker of human iDC (hiDC) reprogramming (Fig. 7C and fig. S12B). Histocompatibility antigen HLA-DR expression was only robustly detected later at days 7 and 9 (Fig. 7D), and it was not increased in cultures with Fl3tl and GM-CSF (fig. S12C). At day 9, hiDCs displayed expression of the hDC1-specific markers CD141 and CLEC9A (Fig. 7E) and functional ability to incorporate beads (fig. S12D), labeled OVA (Fig. 7F), and dead cells (Fig. 7G). More-over, we extended our proof-of-concept experiments to adult human dermal fibroblasts (HDFs) that were also amenable for DC1-like reprogramming with PIB showing acquisition of DC morphology (Fig. 7H) and CD45, HLA-DR, CD141, and CLEC9A (Fig. 7, I and J). Together, our data strongly support the feasibility of using a direct reprogramming approach to generate patient-specific DCs from easily accessible human fibroblasts. Overall, iDC reprogramming might contribute to the future development of novel immunotherapies.

DISCUSSIONHere, we have shown that the combination of PIB efficiently repro-grams mouse and human fibroblasts into functional DCs, termed iDCs. Conditional gene deletion on hematopoietic progenitors has shown that the E26 transformation-specific (ETS) family transcrip-tion factor PU.1 is essential for generation of cDCs and pDCs (8). High levels of PU.1 in monocytes antagonize the macrophage factor MAFB and favors DC development (30). Irf8−/− animals develop a myeloproliferative syndrome characterized by the lack of pDC and cDC1 subsets (31), and studies using conditional Irf8- deficient mice highlighted its role as a terminal selector of the cDC1 lineage (32). BATF3 is a member of the activator protien-1 (AP-1) family of transcription factors expressed in both cDC subsets, but it is only required for the development of cDC1 cells (33), thus considered BATF3-dependent DCs (7). PIB have been shown to cooperate during DC specification. High levels of PU.1 are necessary to drive DC specification in an Irf8-dependent manner (8, 34), consistent with the polycistronic data suggesting that high levels of PU.1 in-crease the efficiency of DC reprogramming. The initiation of IRF8 expression in CDPs is affected by a point mutation that affects the interaction between PU.1 and IRF8 (9) and impairs cDC1 but not pDC development (35). Because of its low affinity to IFN- response elements, IRF8 can be recruited to target sites through interaction with ETS transcription factors. AP-1 transcription factors, such as BATF3, were also shown to interact with IRF4 and IRF8 at AP-1–IRF consensus elements (AICEs) (36). In this context, it was recently shown that BATF3 binds to the Irf8 super- enhancer region, establish-ing an autoregulatory loop to maintain IRF8 levels after commitment

BA

poly

Spi1 Irf8 Batf3

P2A T2A

Spi1Irf8 Batf3

P2A T2A

12

10

8

6

4

2

0

P+I+B PIBpo

ly

IPBpo

ly

C****

****

M2r

tTA

E

0.1%

M2rtTA

0

20

40

60

80

100

0

20

40

60

80

100

Mod

al

0.0% 8.1%C57

BL/

6 M

EF

Mod

alMHC-II+ tdT+MHC-II tdT

0.0% 0.0%

0.0%100%

3.7% 6.2%

2.0%88.2%

tdTo

mat

o

Cle

c9a-

tdT

ME

F

Mod

al

C57

BL/

6 T

TF

DA

PI

MHC-II0 10

310

410

5

CD45

26.0%

8.9%

F

G

DA

PI

MHC-II

MHC-II

MHC-II+MHC-II–

M2rtTA

M2rtTA

CD45

CD45

FUW-tetO-PIBpoly

FUW-tetO-IPBpoly

MHC-II+MHC-II–

0

20

40

60

80

10044.6%

0 103

104

105

0 103

104

105

0

20

40

60

80

100 40.1%

0

5

10

15

20TdT+TdT

MH

C-I

I+ c

ells

(%

)

P+I+BPIBpo

ly

IPBpo

ly

M2r

tTA

Dtd

Tom

ato+

cel

ls (

%)

PIBpolyIPBpolyM2

CANX

IRF8

PU.1

–

––PIBpoly

PIBpoly

PIBpoly

Fig. 6. Polycistronic vectors encoding PIB increase reprogramming efficiency in mouse embryonic and adult fibroblasts. (A) Schematic representation of the two polycistronic (poly) lentiviral plasmids encoding Spi1, Irf8, and Batf3 in differ-ent orders separated by P2A and T2A self-cleaving peptides. (B) Flow cytometry quantification of tdT+ cells after transduction with PIB individual vectors (P + I + B) and PIBpoly and IPBpoly vectors at day 8 after the addition of Dox (n = 7, mean ± SD). (C) Quantification of MHC-II+ cells in tdT+ and tdT− populations (n = 6, mean ± SD). (D) Western blot analysis of PU.1 and IRF8 protein levels in MEFs transduced with PIBpoly and IPBpoly plasmids at day 2. M2rtTA (M2)–transduced cells and calnexin (CANX) levels were included as controls. (E to G) MHC-II and CD45 expression in (E) PIBpoly- transduced Clec9a-tdT MEFs, (F) C57BL/6 MEFs, and (G) adult tail-tip fibro-blasts (TTFs) at day 9. ****P < 0.0001, one-way ANOVA with Bonferroni’s test.

by guest on Novem


munology.sciencem

ag.org/D

ownloaded from




10 of 15

to the cDC1 lineage (9). However, because PU.1 and IRF8 over-expression is not sufficient to induce reprogramming, our results suggest that BATF3 may play additional roles during cDC1 specifi-cation (9). In-depth mechanistic studies on the direct reprogram-ming events governing conversion from fibroblasts into DCs may help in elucidating specification and segregation of DC subsets. For instance, it would be interesting to investigate whether IRF8-PU.1 and IRF8-BATF3 complexes cooperatively bind to DNA at ETS-IRF and AICEs to kick-start the reprogramming process.

Overexpression of PIB induces rapid reporter activation, DC morphology, surface phenotype, and genome-wide transcriptional reprogramming favoring a cDC1-like fate. Single-cell and popula-tion transcriptomes showed high resemblance to bona fide splenic cDC1 cells, validating the fidelity of the reprogramming process. As previously described in reprogramming to cardiomyocytes (37), population mRNA-seq results did not capture the asynchronous na-ture of the reprogramming process by revealing major transcriptional changes at early time points and clustering together day 5, day 7, day 8, and day 9 iDCs. In contrast, full-transcript single-cell mRNA-seq highlighted stepwise transcriptome transitions as verified by path-way and transcription factor network analysis. After 9 days, a robust transcription factor network is established, weakly correlated with

PIB, suggesting reprogramming into a stable cell fate, in agreement with the activation of endogenous Spi1, Irf8, and Batf3 expression. During reprogramming to induced neurons, an intermediate tran-scriptional state resembling progenitor cells is transiently induced (38). However, during DC reprogramming, induced transcription factor networks of mature cDC1s and the absence of colony forma-tion potential support the concept of a direct reprogramming pro-cess that matures over time.

Collectively, our results show that a small combination of tran-scription factors is sufficient to induce DC-like cells that function as APCs. We showed that iDCs secrete inflammatory cytokines upon TLR3/4 stimulation and acquire the competence for phagocytosis of small particles, proteins, and dead cells. We have provided proof of principle that immune responses can be induced by direct cell re-programming, showing that iDCs present antigens to CD4+ and cross- present antigens to CD8+ T cells. These results represent a unique opportunity to merge the field of cell reprogramming and immuno-therapy. cDC1 cells are essential for inducing T-cytotoxic responses and tumor clearance, making iDCs attractive as novel cancer immuno-therapies (39). Methods to generate BM-DCs have been extensively explored; however, BM cultures give rise to cDC1-like cells contained in a mix population of cDC2- and pDC-like cells (20, 40, 41). In our

100 um100 um 50 um1010101010100001001000000000000000 umumumumumumumumuu

A M2rtTA d9

HE

F b

right

fiel

d

PIB d9

1010000 umumumumummmu

Cd3 d5 d7

HLA

-DR

CD45

d90.0%

7.5%

0.1%

9.2%

0.2%

7.1%

0.6%

10.8%

D

G72.4%0.8%

HLA-DR–

Far

red

HLA-DR

M2rtTA d9 PIBpoly d9H I HDF + PIB d9HDF + M2rtTA d9

HLA

-DR

CD45

0.0%

0.0%

0.8%

0.2%

J

ECD45

0.0%

0.0%

HEF + PIBpoly

HEF + M2rtTA d9

0-103

103

104

105

0

CD141

20

40

60

80

100

CLEC9A

PIB-gated HLA-DR+FMO control

Mod

al

0-103

103

104

105

020

40

60

80

100

Mod

alCD141 CLEC9A

F

PIB-gated HLA-DR+FMO control

5050000 uuuuuummmm

HD

F B

right

field

CD

45+

HLA

-DR

+ (%

) CD45+ HLA-DR+

d0 d3 d5 d7 d90.0

0.2

0.4

0.6

0.8

M2rtTAPIBpoly

Days

B M2rtTA d9 PIB d9

HE

F S

EM

30.3%

26.3%

0 103

104

105

17.0%

0 103

104

105

23.7%

Auto

0.2% 1.2%O

VA

-A64

74ºC 37ºC

Humanfibroblast hiDC

P I B

HLA-DR+

Fig. 7. PIB induce reprogramming of human embryonic and adult fibroblasts to DC-like cells. (A) Strategy to generate human iDCs (hiDCs) upon transduction of fibroblasts with the PIBpoly vector. (B) Bright-field microscopy (left) and scanning electron microscopy (SEM; right) of PIBpoly-transduced human embryonic fibroblasts (HEFs) at day 9 after the addition of Dox. Scale bars, 50 µm (left) and 10 µm (right). (C and D) Kinetics of CD45 and HLA-DR expression of PIBpoly- and M2rtTA-transduced HEFs analyzed by flow cytometry (n = 2 to 5, mean ± SD). (E) CD141 and CLEC9A expression in HLA-DR+ cells at day 9. Fluorescence minus one (FMO) was included as control. (F) Incorporation of Alexa Fluor 647–labeled OVA (OVA-A647) by PIBpoly-transduced HEFs after 20-min incubation at 37° or 4°C. (G) PIBpoly-transduced HEFs were incubated overnight with CellVue far red–labeled dead cells at day 8. HLA-DR− and HLA-DR+ populations are shown. (H) Bright-field microscopy of PIBpoly-transduced human dermal fibro-blasts (HDFs) at day 9. (I) CD45 and HLA-DR expression of PIB- and M2rtTA-transduced HDFs at day 9. (J) CD141 and CLEC9A expression in HLA-DR+ cells. Scale bars, 100 µm.

by guest on Novem


munology.sciencem

ag.org/D

ownloaded from




11 of 15

system, we have generated cDC1-like cells using a transcription factor– based reprogramming methodology. We have observed minimal to residual expression of surface markers reminiscent of cDC2 and pDC lineage commitment. As such, we believe that our system consti-tutes an instructive system that cell-intrinsically promotes the gener-ation of cDC1-like cells. In addition, BM-based methods to generate DCs for therapy have limited translation utility because of scarce accessibility to starting material. Currently, human “DCs” are gener-ated from peripheral blood mononuclear cells (PBMCs), often asso-ciated with compromised migration and cross-presentation ability. Therefore, adult fibroblasts could offer a viable alternative to generate human bona fide DC. The two approaches could potentially be syner-gistic because polycistronic delivery of PIB factors could be used in other cell types including BM progenitors and PBMCs to direct the generation of cDC1 cells.

This study sets the foundation for a future thorough characteri-zation of the in vivo stability and function of iDCs. It is conceivable that DCs in vivo receive tissue-specific cues that are difficult to re-produce during reprogramming. iDCs’ stability, migration, antigen presentation, and cross-presentation should be addressed in wild-type, DC-depleted, and tumor-bearing animal models to fully understand the potential of DC reprogramming. Cell transfer exper-iments or inducing DC reprogramming in vivo (42) may help clarify the requirements for extrinsic and intrinsic regulation during DC reprogramming. This may be particularly important for antitumor vaccination where the utility of human cDC1 cells has been limited by their rarity and low yield.

Overall, we provide evidence that the combined action of PIB programs antigen-presenting cDC1-like cell fate in fibroblasts. These results provide insight into the molecular mechanisms regulating DC specification and constitute the foundation for the develop-ment of powerful cancer immunotherapies based on direct DC reprogramming.

MATERIALS AND METHODSStudy designThis study aimed at defining the minimal transcription factor com-bination necessary to reprogram fibroblasts in functional antigen- presenting DCs. All experiments were repeated at least once. The number of biological replicates for each experiment is indicated in the figure legends.

MiceClec9aCre/Cre animals (11) were crossed with Rosa26-stopflox-tdT re-porter mice to generate double homozygous Clec9aCre/Cre RosatdT/tdT (Clec9a-tdT) mice in a C57BL/6 background. C57BL/6 mice were acquired from Charles River Laboratories, OT-II transgenic/Rag2 constitutive knockout (KO) (OT-II/Rag2KO) mice were provided by L. Graça, and OT-I mice were acquired from Scanbur. All ani-mals were housed under controlled temperature (23° ± 2°C) and subjected to a fixed 12-hour light/12-hour dark cycle, with free access to food and water. Animal care and experimental procedures were performed in accordance with Portuguese and Swedish guidelines and regulations after approval from local committees.

Isolation and culture of MEFsMEFs were isolated from E13.5 embryos of Clec9a-tdT or C57BL/6 mice as previously described (3, 11). A single-cell suspension was

obtained and plated in 0.1% gelatin-coated 10-cm tissue cul-ture dishes in growth media. Cells were grown for 2 to 3 days until confluence, dissociated with TrypLE Express, and frozen in fetal bovine serum (FBS) and 10% dimethyl sulfoxide (Sigma). MEFs used for screening and in the following experiments were sorted for tdT− CD45− with a purity of >99% and expanded up to four passages.

BM and spleen isolationsTotal BM cells were harvested from long bones (tibias and femurs) by crushing with pestle and mortar. Freshly isolated spleens were homogenized using the frosted ends of two sterile slides. Cells were harvested in phosphate-buffered saline (PBS) supplemented with 2% FBS and filtered through a 70-m cell strainer (BD Biosciences). Red blood cells were lysed with BD Pharm Lyse (BD Biosciences) for 8 min at room temperature. Lysis was stopped by the addition of ≥5 volumes of PBS with 2% FBS.

Cell cultureHuman embryonic kidney (HEK) 293T cells, MEFs, HEFs, and HDFs (ScienCell) were maintained in growth medium [Dulbecco’s modi-fied Eagle’s medium supplemented with 10% (v/v) FBS, 2 mM l- glutamine, and antibiotics (penicillin and streptomycin, 10 g/ml)]. OP9 and OP9-DL1 cell lines were cultured in minimum essential medium alpha containing 20% FBS, 1 mM l-glutamine, and anti-biotics (43, 44). B3Z hybridoma cells with a T cell receptor specific to the Kb/OVA257–264 peptide complex (45) were grown in RPMI 1640, supplemented with 10% FBS, 2 mM GlutaMax, 10 mM Hepes, 1 mM sodium pyruvate, 1× nonessential amino acids, antibiotics, and 50 M -mercaptoethanol. All cells were maintained at 37°C and 5% (v/v) CO2. All tissue culture reagents were from Thermo Fisher Scientific unless stated otherwise.

Generation of BM-DCsTotal BM cells were plated in petri dishes (5 × 106 cells per plate) in RPMI complete media supplemented with GM-CSF (20 ng/ml). After 3 days of culture, 5 ml of fresh RPMI media with GM-CSF was added. BM-DCs were used after 8 days of culture. cDC1-like BM-DCs were generated by supplementing RPMI media with Flt3l (200 ng/ml) and GM-CSF (5 ng/ml) as previ-ously described (41).

Molecular cloning and lentiviral productionCoding regions of each candidate transcription factor (table S1) were individually cloned into the pFUW-TetO vector where expres-sion is under the control of the tetracycline operator and a minimal cytomegalovirus promoter (FUW-TetO-TF) (3, 46). For the poly-cistronic vectors, coding sequences for Spi1, Irf8, and Batf3 were cloned together in the pFUW-TetO plasmid interspaced with 2A self-cleaving peptides (47). The first two coding sequences lacked the stop codon. A lentiviral vector containing the reverse tetracy-cline transactivator M2rtTA under the control of a constitutively active human ubiquitin C promoter (FUW-M2rtTA) was used for cotransductions (3, 48). HEK293T cells were transfected with a mixture of transfer plasmid and packaging constructs expressing the viral packaging functions and the envelope VSV-G protein. Viral supernatants were harvested after 36, 48, and 72 hours; fil-tered (0.45 m); and concentrated 40-fold with Amicon ultracen-trifugal filters (Millipore).

by guest on Novem


munology.sciencem

ag.org/D

ownloaded from




12 of 15

Viral transduction and reprogrammingClec9a-tdT, C57BL/6 MEFs, and TTFs, HEFs, and HDFs were seeded at a density of 40,000 cells per well on 0.1% gelatin-coated six-well plates. Cells were incubated overnight with a ratio of 1:1 FUW-TetO-TFs/FUW-M2rtTA lentiviral particles in media supplemented with polybrene (8 g/ml). When testing combinations of transcription factors, equal multiplicities of infection of each individual viral par-ticle were applied. Cells were transduced twice in consecutive days, and media were replaced in between. After the second transduction, growth media were supplemented with Dox (1 g/ml), and this was considered day 0. Media were changed every 2 to 3 days for the duration of the cultures. When stated, variations of culture condi-tions were applied, namely, RPMI complete media, LPS (100 ng/ml; Sigma), 2-mercaptoethanol (2-ME; 1 × 104 M), l-glutamine (2 mol/ml), GM-CSF (10 ng/ml), IL-4 (20 ng/ml), and Flt3l (100 ng/ml; STEMCELL Technologies).

Fluorescent microscopy and immunofluorescenceClec9a-driven tdT in MEFs was visualized directly on six-well plates under an inverted microscope (Zeiss AxioVert 200 M), and images were processed with AxioVision and Adobe Photoshop software. DAPI (4ʹ,6-diamidino-2-phenylindole; 1 g/ml; Sigma) and phalloidin–Alexa Fluor 488 (50 g/ml; Sigma) were used to stain nuclei and F-actin, respectively. For time-lapse microscopy, fluorescent pictures were acquired immediately after adding Dox every hour for 6 days using an IN Cell Analyzer 2200 (GE Healthcare) coupled with a KiNEDx robot (PAA) and a Cytomat2 incubator (Thermo Fisher Scientific). Time-lapse imaging of dead cell phagocytosis was per-formed every 1 hour for 35 hours.

Flow cytometry analysisAntibodies are listed in table S1. For screening of candidate factors, transduced Clec9a-tdT MEFs were dissociated with TrypLE Ex-press and analyzed in Accuri C6 (BD Biosciences). For the analysis of surface marker expression, dissociated mouse and human cells were incubated with adequate antibodies diluted in PBS with 5% FBS at 4°C for 30 min in the presence of rat or mouse serum (1/100, GeneTex) to block unspecific binding. Cells were washed and re-suspended in PBS with 5% FBS and analyzed in Accuri C6 or in FACSAria III (BD Biosciences). Flow cytometry data were analyzed using FlowJo software (FLOWJO LLC, version 7.6). Unless stated otherwise, all flow cytometry analyses were performed in single live cell gates.

Fluorescence-activated cell sortingTo purify C57BL/6 and Clec9a-tdT MEFs, we incubated cells at 4°C for 30 min with anti-CD45 antibody and purified them in FACSAria III. For isolation of DCs, splenocytes were purified according to gating strategies defined in fig. S13. For functional experiments, DCs were first enriched using Pan-DC Enrichment beads (Miltenyi). For isolation of iDCs, cells were dissociated using TrypLE Express and resuspended in PBS with 5% FBS, and tdT− or tdT+ cells were purified. When mentioned, cells were incubated with rat anti- mouse I-A/I-E and CD45 antibody in the presence of rat serum for 30 min at 4°C.

Single-cell mRNA-seq library preparationMEFs and PIB-transduced MEFs were FACS sorted and collected at day 3 (tdT+), day 7 (tdT+), and day 9 (tdT+ MHC-II+). cDC1s were iso-

lated from C56BL/6 mice (fig. S13). Single- cell capture was performed on a C1 Auto Prep System (Fluidigm) according to the manufacturer’s instructions. Briefly, 500 to 1000 cells/l in C1 Cell Suspension Reagent (3:2 ratio) were loaded onto a 5- to 10-m or a 10- to 17-m C1 Single Cell mRNA-seq IFC chip. A microscope was used to exclude samples with no cell, more than one cell, or samples with cellular debris. Sin-gle cells were lysed, and reverse transcription and complementary DNA (cDNA) preamplification were performed in the chip using the SMART-Seq v4 Ultra Low RNA Kit for Illumina Sequencing (Clontech). During reverse transcription, adaptors were incorporated within the primers, allowing amplification of full-length transcripts by polymerase chain reaction (PCR). cDNA was harvested immedi-ately after finishing the program and stored at −20°C.

cDNA was quantified using the Quant-iT PicoGreen double- stranded DNA Assay Kit (Life Technologies) and read on a fluores-cence MicroPlate Reader (Synergy H1, BioTek). Samples were then diluted to a final concentration of 0.15 to 0.30 ng/l using C1 Harvest Reagent. Sequencing libraries were prepared using the Nextera XT DNA Library Prep Kit (Illumina) modified for single-cell mRNA-seq. Tagmented cDNA was amplified using the dual index primers (Nextera XT DNA Index kit 96 indices, Illumina), and 96 single-cell libraries were pooled together. Clean up reactions were done with AMPure XP beads (Beckman Coulter Genomics). The concentration of the pooled libraries was determined using Agilent Bioanalyzer using the High Sensitivity DNA Analysis Kit (Agilent). Libraries were sequenced, yielding, on average, ~4.8 million 75-nucleotide single-end reads per sample on a HiSeq 2000 platform.

Single-cell mRNA-seq analysisSingle-end reads were mapped to the mm10 mouse genome (Ensembl annotation, release 89) using Salmon v0.8.1 with k = 21. The resulting transcripts per million were imported into R using tximport library and converted into mRNA counts using the Census algorithm im-plemented in Monocle library (49). Scater library was used to in-clude cells and genes that pass quality control thresholds, according to the following criteria: (i) total number of mRNA census counts detected per sample >50,000, (ii) number of genes detected in each single cell >1000, and (iii) percentage of counts in mitochondrial genes <9%. From the 192 cells initially profiled, 163 individual cells passed quality control filters and were used for analysis. Custom R scripts were used to perform t-SNE (Monocle and Scater pack-age), PCA (Monocle and Scater package), clustering (SC3 package), and analysis of variance (ANOVA) and to construct heat maps, box plots, scatterplots, violin plots, dendrograms, bar graphs, and histo-grams. Generally, ggplot2, gplots, graphics, and pheatmap packages were used to generate data graphs.

Differential expression analysis was performed using Monocle, selecting genes with Benjamini and Hochberg (BH)–corrected P values <0.05. The resulting genes were next filtered by variance (genes with variance ≥1 across all conditions were selected). Last, the re-sulting 6525 genes were grouped into four distinct clusters based on complete- linkage hierarchical clustering (pheatmap package).

To estimate the endogenous expression of genes, the mapping pro-cedure was repeated using STAR v2.5.3a with default settings. The number of reads in the 5′ untranslated region (5′UTR) and 3′UTR was calculated using multicov from bedtools v2.27.0. For the total expression, multicov software was also used considering the window (start of the gene, end of the gene). Log counts were displayed as box plots with whiskers extending to ±1.5 × interquartile range.

by guest on Novem


munology.sciencem

ag.org/D

ownloaded from




13 of 15

cDC1 and cDC2 gene signaturescDC1 and cDC2 gene signatures were defined using splenic DC (10) and BM-DC data (20). Genes were excluded from the analysis ac-cording to the following criteria: (i) genes in which expression in MEFs was significantly higher compared with day 3, day 7, or day 9 iDCs (logFC > 0, BH-corrected P value <0.05); (ii) genes from the cDC2 list that we found highly expressed in cDC1 compared with day 3, day 7, or day 9 iDCs (logFC > 0, BH-corrected P value <0.05). Next, for each selected gene, we calculated the median of gene expression across cells from each biological sample group. Last, we plotted the median of gene expression across the curated cDC1 and cDC2 gene signatures.

Pseudotime reconstructionThe Monocle package was used to order cells on a pseudotime course during MEF to iDC reprogramming (50). Monocle analysis was per-formed based on cDC1 and cDC2 gene signatures (12). Because the resulting single-cell trajectories include branches, BEAM was imple-mented to find differentially expressed genes between the branches (49). Alternatively, pseudotime reconstruction was performed using TSCAN software (51).

Ciita promoter analysisThe three different promoters described in literature (52) were defined based on UCSC mm10 genes. The same length for each promoter was defined to account for effect associated with different window sizes. The number of reads for each promoter was calcu-lated using multicov from bedtools v2.27.0. Next, we calculated the average per cell condition and estimated the proportion of pro-moter preferences.

Inflammatory cytokine assaysLevels of IL-6, IL-10, and TNF- were assessed in supernatants of purified tdT+ cells at day 9 or in PIB-transduced MEFs 10 days after Dox. LPS (100 ng/ml) or PIC (25 g/ml) (Invivogen) was added for overnight stimulation. Fifty microliters of culture supernatants was collected and analyzed by the CBA Mouse Inflammation Kit (BD Biosciences) or by the LEGENDplex Mouse Anti-Virus Response Panel (BioLegend), according to the manufacturer’s instructions. Acquisition was performed with Accuri C6 or FACSCanto, and data were analyzed using FCAP (BD Biosciences) or LEGENDplex (Bio-Legend) softwares. For IL12p40, PIB-transduced Clec9a- tdT MEFs at day 9 were incubated overnight in the presence of LPS (100 ng/ml), profilin (100 ng/ml) (Sigma), and CD40L (1 g/ml; BioLegend). On the following day, Golgiplug (1 l/ml; BD Biosciences) was added and incubated at 37°C for 4 hours. Cells were then harvested and intracellular staining of IL-12p40 was performed (Cytofix/Cytoperm; BD Biosciences).

Incorporation of dead cellsHEK293T cells were exposed to ultraviolet irradiation (50 J/m2) to induce cell death and labeled with the CellVue Claret Far Red Fluo-rescent Cell Linker Kit (Sigma). Purified tdT+ and tdT− populations at day 10 and PIBpoly-transduced HEFs at day 9 were incubated with far red–labeled dead cells overnight, washed with PBS with 5% FBS, and analyzed in FACSAria III. DAPI staining was used to exclude floating or membrane-adherent dead cells. Dead cell incorporation was quantified in live tdT− and tdT+ cells or HLA-DR+ and HLA-DR− cells using far red staining. For time-lapse imaging of dead cell

phagocytosis, dead HEK293T cells were labeled with DAPI and added to FACS-sorted PIB- transduced tdT+ cell cultures immedi-ately before starting image acquisition.

CD4+ T cell isolation and antigen-presenting assaysCD4+ T cells from spleen of OT-II/Rag2KO mice were enriched using the Dynabeads Untouched Mouse CD4 Cells Kit (BD Biosciences). Enriched CD4+ T cells (purity, ≥85%) were labeled with 5 M CFSE at room temperature for 10 min, washed, and counted. Twenty thou-sand PIB-transduced MEFs or 20,000 splenic CD11c+ MHC-II+ cells were incubated with 20,000 CFSE-labeled OT-II CD4+ T cells in 96-well round-bottom tissue culture plates with OVA protein (10 g/ml) or the OVA323-339 peptide (10 g/ml) in the presence or absence of LPS (100 ng/ml). After 7 days of coculture, T cells were collected and stained for CD44. T cell proliferation (dilution of CFSE staining) and activation (CD44 expression) were analyzed in Accuri C6.

Export to cytosolThe efficiency of antigen export to the cytosol by Clec9a-tdT+ cells was analyzed by a cytofluorimetry-based assay as previously de-scribed (27). Briefly, purified tdT+ cells at day 7, unpurified popula-tion at day 16, or Flt3l/GM-CSF BM-DCs were resuspended in loading buffer and loaded with 1 M CCF4-AM for 30 min at room temperature. Cells were then washed and incubated with -lactamase (2 mg/ml) at 37°C for 30, 60, and 90 min. To stop the reaction, we transferred cells to ice-cold PBS. Immediately before flow cytometry analysis in FACSAria III, cells were stained with Fixable Viability Dye eFluor 780 (eBioscience).

CD8+ T cell isolation and antigen cross-presentation assaysCD8+ T cells from spleen of OT-I mice were enriched using a naïve mouse CD8+ T cell Isolation kit (Miltenyi). Enriched CD8+ T cells were labeled with 5 M CTV (Thermo Fisher) at room temperature for 20 min, washed, and counted. FACS-sorted tdT+ and tdT− cells at day 8, MEFs, freshly isolated splenic cDC1 cells (CD11c+ MHC-II+ CD8+), and Flt3l/GM-CSF BM-DCs were incubated at 37°C with OVA protein (100 g/ml) in the presence of PIC (25 g/ml) for 10 hours. After extensive washing, 20,000 APCs were incubated with 100,000 CTV-labeled OT-I CD8+ T cells in 96-well round- bottom tissue culture plates with PIC (25 g/ml). After 3 days of coculture, T cells were collected, stained, and analyzed in BD LSRFortessa. T cell proliferation (dilution of CTV staining) and activation (CD44 ex-pression) were determined by gating live single TCR+ CD8+ T cells. For hybridoma cross-presentation assays, PIB-transduced Clec9a-tdT MEFs at day 16 after the addition of Dox were dissociated with TrypLE Express, resuspended in growth media, and incubated for 4 hours with different concentrations of OVA protein. After washing three times with PBS with 5% FBS, 100,000 PIB-transduced MEFs were co-cultured with 100,000 B3Z cells in 96-well round-bottom tissue cul-ture plates in the presence or absence of LPS (100 ng/ml) or PIC (25 g/ml). Alternatively, FACS-sorted tdT+ cells at day 8, nontrans-duced MEFs, and Flt3l/GM-CSF BM-DCs were incubated at 37°C with OVA protein (100 g/ml) in the presence of PIC (25 g/ml) for 10 hours. Fifty thousand APCs were cocultured with 100,000 B3Z cells in 96-well round-bottom tissue culture plates in the presence of PIC. After 18 hours, cells were lysed in 0.125% NP-40 (Tergitol, Sigma), 9 mM MgCl2, and chlorophenol red--D-galactopyranoside - galactosidase substrate (Roche). -Galactosidase activity was mea-sured on a MicroPlate Reader at an absorbance of 590 nm.

by guest on Novem


munology.sciencem

ag.org/D

ownloaded from




14 of 15

Western blotC57BL/6 MEFs transduced with polycistronic vectors were harvested 48 hours after the addition of Dox. Cells were resuspended in radio-immunoprecipitation assay buffer (Thermo Scientific) for 20 min, and protein extracts were diluted 1:2 in Laemmli buffer (Bio-Rad) with 5% 2-ME (Sigma) and boiled at 98°C for 8 min. Samples were run in NuPAGE 4 to 12% bis-tris (Invitrogen) SDS-PAGE gels using XCell Sure Lock (Invitrogen) and MOPS SDS running buffer (Invitrogen). Transfer was done using iBlot (Thermo Scientific) dry system for 7 min. Membranes were incubated overnight with unconjugated primary antibodies against PU.1, IRF8, or calnexin and with donkey anti-rabbit horseradish peroxidase–conjugated secondary antibody diluted at 1:4000. Membranes were incubated with ECL prime (Amersham) for 5 min, and data were acquired in ChemiDoc (Bio-Rad).

Statistical analysisComparisons among groups were performed by one-way ANOVA followed by Bonferroni’s multiple comparison test with GraphPad Prism 5 software. P values are shown when relevant (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns, nonsignificant).

SUPPLEMENTARY MATERIALSSupplementary material for this article is available at http://immunology.sciencemag.org/cgi/content/full/3/30/eaau4292/DC1MethodsFig. S1. Candidate transcription factors to instruct DC cell fate.Fig. S2. Clec9a-based reporter to identify transcription factors to induce DC fate.Fig. S3. Screening for transcription factors to activate Clec9a-tdT.Fig. S4. DC morphology and dendrites are established in mouse fibroblasts.Fig. S5. Spi1, Irf8, and Batf3 are enriched in cDC1 cells.Fig. S6. Global single-cell gene expression analysis during iDC reprogramming.Fig. S7. Stepwise transitions during iDC reprogramming.Fig. S8. iDC reprogramming does not induce alterations associated with tumorigenesis.Fig. S9. Analysis of iDC maturation states.Fig. S10. Induced DCs are functional APCs.Fig. S11. Polycistronic vector induces cDC1-like transcriptional program.Fig. S12. PIB induce human iDC reprogramming.Fig. S13. FACS gating strategies.Table S1. Analysis of candidate transcription factors.Table S2. Single-cell mRNA-seq data analysis.Table S3. Pseudotime single-cell mRNA-seq data analysis.Table S4. Population mRNA-seq data analysis.Table S5. Raw data.Movie S1. Time lapse showing Clec9a-tdT+ cell emergence.Movie S2. Time lapse displaying dead cell incorporation.References (53–55)

REFERENCES AND NOTES 1. K. Takahashi, S. Yamanaka, Induction of pluripotent stem cells from mouse embryonic

and adult fibroblast cultures by defined factors. Cell 126, 663–676 (2006). 2. C. F. Pereira, I. R. Lemischka, K. Moore, Reprogramming cell fates: Insights from

combinatorial approaches. Ann. N. Y. Acad. Sci. 1266, 7–17 (2012). 3. C. F. Pereira, B. Chang, J. Qiu, X. Niu, D. Papatsenko, C. E. Hendry, N. R. Clark,

A. Nomura-Kitabayashi, J. C. Kovacic, A. Ma’ayan, C. Schaniel, I. R. Lemischka, K. Moore, Induction of a hemogenic program in mouse fibroblasts. Cell Stem Cell 13, 205–218 (2013).

4. R. Feng, S. C. Desbordes, H. Xie, E. S. Tillo, F. Pixley, E. R. Stanley, T. Graf, PU.1 and C/EBP/ convert fibroblasts into macrophage-like cells. Proc. Natl. Acad. Sci. U.S.A. 105, 6057–6062 (2008).

5. H. Xie, M. Ye, R. Feng, T. Graf, Stepwise reprogramming of B cells into macrophages. Cell 117, 663–676 (2004).

6. M. Merad, P. Sathe, J. Helft, J. Miller, A. Mortha, The dendritic cell lineage: Ontogeny and function of dendritic cells and their subsets in the steady state and the inflamed setting. Annu. Rev. Immunol. 31, 563–604 (2013).

7. B. T. Edelson, W. KC, R. Juang, M. Kohyama, L. A. Benoit, P. A. Klekotka, C. Moon, J. C. Albring, W. Ise, D. G. Michael, D. Bhattacharya, T. S. Stappenbeck, M. J. Holtzman,

S.-S. J. Sung, T. L. Murphy, K. Hildner, K. M. Murphy, Peripheral CD103+ dendritic cells form a unified subset developmentally related to CD8+ conventional dendritic cells. J. Exp. Med. 207, 823–836 (2010).

8. S. Carotta, A. Dakic, A. D'Amico, S. H. M. Pang, K. T. Greig, S. L. Nutt, L. Wu, The transcription factor PU.1 controls dendritic cell development and Flt3 cytokine receptor expression in a dose-dependent manner. Immunity 32, 628–641 (2010).

9. G. E. Grajales-Reyes, A. Iwata, J. Albring, X. Wu, R. Tussiwand, W. Kc, N. M. Kretzer, C. G. Briseno, V. Durai, P. Bagadia, M. Haldar, J. Schonheit, F. Rosenbauer, T. L. Murphy, K. M. Murphy, Batf3 maintains autoactivation of Irf8 for commitment of a CD8+ conventional DC clonogenic progenitor. Nat. Immunol. 16, 708–717 (2015).

10. J. Helft, J. Bottcher, P. Chakravarty, S. Zelenay, J. Huotari, B. U. Schraml, D. Goubau, C. Reis e Sousa, GM-CSF mouse bone marrow cultures comprise a heterogeneous population of CD11c(+)MHCII(+) macrophages and dendritic cells. Immunity 42, 1197–1211 (2015).

11. B. U. Schraml, J. van Blijswijk, S. Zelenay, P. G. Whitney, A. Filby, S. E. Acton, N. C. Rogers, N. Moncaut, J. J. Carvajal, C. Reis e Sousa, Genetic tracing via DNGR-1 expression history defines dendritic cells as a hematopoietic lineage. Cell 154, 843–858 (2013).

12. A. Schlitzer, V. Sivakamasundari, J. Chen, H. R. B. Sumatoh, J. Schreuder, J. Lum, B. Malleret, S. Zhang, A. Larbi, F. Zolezzi, L. Renia, M. Poidinger, S. Naik, E. W. Newell, P. Robson, F. Ginhoux, Identification of cDC1- and cDC2-committed DC progenitors reveals early lineage priming at the common DC progenitor stage in the bone marrow. Nat. Immunol. 16, 718–728 (2015).

13. D. Sancho, D. Mourão-Sá, O. P. Joffre, O. Schulz, N. C. Rogers, D. J. Pennington, J. R. Carlyle, C. Reis e Sousa, Tumor therapy in mice via antigen targeting to a novel, DC-restricted C-type lectin. J. Clin. Invest. 118, 2098–2110 (2008).

14. S. Senju, S. Hirata, H. Matsuyoshi, M. Masuda, Y. Uemura, K. Araki, K. Yamamura, Y. Nishimura, Generation and genetic modification of dendritic cells derived from mouse embryonic stem cells. Blood 101, 3501–3508 (2002).

15. N. Kitamura, H. Yokoyama, T. Yashiro, N. Nakano, M. Nishiyama, S. Kanada, T. Fukai, M. Hara, S. Ikeda, H. Ogawa, K. Okumura, C. Nishiyama, Role of PU.1 in MHC class II expression through transcriptional regulation of class II transactivator pI in dendritic cells. J. Allergy Clin. Immunol. 129, 814–824.e6 (2012).

16. W. Reith, S. LeibundGut-Landmann, J. M. Waldburger, Regulation of MHC class II gene expression by the class II transactivator. Nat. Rev. Immunol. 5, 793–806 (2005).

17. G. Basha, K. Omilusik, A. Chavez-Steenbock, A. T. Reinicke, N. Lack, K. B. Choi, W. A. Jefferies, A CD74-dependent MHC class I endolysosomal cross-presentation pathway. Nat. Immunol. 13, 237–245 (2012).

18. M. Bretou, P. J. Sáez, D. Sanséau, M. Maurin, D. Lankar, M. Chabaud, C. Spampanato, O. Malbec, L. Barbier, S. Muallem, P. Maiuri, A. Ballabio, J. Helft, M. Piel, P. Vargas, A.-M. Lennon-Duménil, Lysosome signaling controls the migration of dendritic cells. Sci. Immunol. 2, eaak9573 (2017).

19. K. S. Kobayashi, P. J. van den Elsen, NLRC5: A key regulator of MHC class I-dependent immune responses. Nat. Rev. Immunol. 12, 813–820 (2012).

20. M. E. Kirkling, U. Cytlak, C. M. Lau, K. L. Lewis, A. Resteu, A. Khodadadi-Jamayran, C. W. Siebel, H. Salmon, M. Merad, A. Tsirigos, M. Collin, V. Bigley, B. Reizis, Notch signaling facilitates in vitro generation of cross-presenting classical dendritic cells. Cell Rep. 23, 3658–3672.e6 (2018).

21. A. T. Satpathy, W. KC, J. C. Albring, B. T. Edelson, N. M. Kretzer, D. Bhattacharya, T. L. Murphy, K. M. Murphy, Zbtb46 expression distinguishes classical dendritic cells and their committed progenitors from other immune lineages. J. Exp. Med. 209, 1135–1152 (2012).

22. L. Frasca, M. Nasso, F. Spensieri, G. Fedele, R. Palazzo, F. Malavasi, C. M. Ausiello, IFN- arms human dendritic cells to perform multiple effector functions. J. Immunol. 180, 1471–1481 (2008).

23. M. Montoya, G. Schiavoni, F. Mattei, I. Gresser, F. Belardelli, P. Borrow, D. F. Tough, Type I interferons produced by dendritic cells promote their phenotypic and functional activation. Blood 99, 3263–3271 (2002).

24. S. H. Jackson, C.-R. Yu, R. M. Mahdi, S. Ebong, C. E. Egwuagu, Dendritic cell maturation requires STAT1 and is under feedback regulation by suppressors of cytokine signaling. J. Immunol. 172, 2307–2315 (2004).

25. O. Schulz, C. Reis e Sousa, Cross-presentation of cell-associated antigens by CD8+ dendritic cells is attributable to their ability to internalize dead cells. Immunology 107, 183–189 (2002).

26. O. P. Joffre, E. Segura, A. Savina, S. Amigorena, Cross-presentation by dendritic cells. Nat. Rev. Immunol. 12, 557–569 (2012).

27. O. I. Vivar, J. G. Magalhaes, S. Amigorena, Measurement of export to the cytosol in dendritic cells using a cytofluorimetry-based assay, in Dendritic Cell Protocols, E. Segura, N. Onai, Eds. (Springer, 2016), pp. 183–188.

28. L. Delamarre, H. Holcombe, I. Mellman, Presentation of exogenous antigens on major histocompatibility complex (MHC) class I and MHC class II molecules is differentially regulated during dendritic cell maturation. J. Exp. Med. 198, 111–122 (2003).

29. A.-C. Villani, R. Satija, G. Reynolds, S. Sarkizova, K. Shekhar, J. Fletcher, M. Griesbeck, A. Butler, S. Zheng, S. Lazo, L. Jardine, D. Dixon, E. Stephenson, E. Nilsson, I. Grundberg,

by guest on Novem


munology.sciencem

ag.org/D

ownloaded from

http://immunology.sciencemag.org/cgi/content/full/3/30/eaau4292/DC1

http://immunology.sciencemag.org/cgi/content/full/3/30/eaau4292/DC1




15 of 15

D. McDonald, A. Filby, W. Li, P. L. De Jager, O. Rozenblatt-Rosen, A. A. Lane, M. Haniffa, A. Regev, N. Hacohen, Single-cell RNA-seq reveals new types of human blood dendriticcells, monocytes, and progenitors. Science 356, eaah4573 (2017).

30. Y. Bakri, S. Sarrazin, U. P. Mayer, S. Tillmanns, C. Nerlov, A. Boned, M. H. Sieweke, Balance of MafB and PU.1 specifies alternative macrophage or dendritic cell fate. Blood 105, 2707–2716 (2005).

31. G. Schiavoni, F. Mattei, P. Sestili, P. Borghi, M. Venditti, H. C. Morse III, F. Belardelli, L. Gabriele, ICSBP is essential for the development of mouse type I interferon-producing cells and for the generation and activation of CD8 (+) dendritic cells. J. Exp. Med. 196, 1415–1425 (2002).

32. D. Sichien, C. L. Scott, L. Martens, M. Vanderkerken, S. Van Gassen, M. Plantinga, T. Joeris, S. De Prijck, L. Vanhoutte, M. Vanheerswynghels, G. Van Isterdael, W. Toussaint, F. B. Madeira, K. Vergote, W. W. Agace, B. E. Clausen, H. Hammad, M. Dalod, Y. Saeys, B. N. Lambrecht, M. Guilliams, IRF8 transcription factor controls survival and function of terminally differentiated conventional and plasmacytoid dendritic cells, respectively. Immunity 45, 626–640 (2016).

33. K. Hildner, B. T. Edelson, W. E. Purtha, M. Diamond, H. Matsushita, M. Kohyama, B. Calderon, B. U. Schraml, E. R. Unanue, M. S. Diamond, R. D. Schreiber, T. L. Murphy, K. M. Murphy, Batf3 deficiency reveals a critical role for CD8+ dendritic cells in cytotoxic T cell immunity. Science 322, 1097–1100 (2008).

34. J. Schönheit, C. Kuhl, M. L. Gebhardt, F. F. Klett, P. Riemke, M. Scheller, G. Huang, R. Naumann, A. Leutz, C. Stocking, J. Priller, M. A. Andrade-Navarro, F. Rosenbauer, PU.1 level-directed chromatin structure remodeling at the Irf8 gene drives dendritic cell commitment. Cell Rep. 3, 1617–1628 (2013).

35. P. Tailor, T. Tamura, H. C. Morse, K. Ozato, The BXH2 mutation in IRF8 differentially impairs dendritic cell subset development in the mouse. Blood 111, 1942–1945 (2008).

36. T. L. Murphy, R. Tussiwand, K. M. Murphy, Specificity through cooperation: BATF–IRFinteractions control immune-regulatory networks. Nat. Rev. Immunol. 13, 499–509(2013).

37. Z. Liu, L. Wang, J. D. Welch, H. Ma, Y. Zhou, H. R. Vaseghi, S. Yu, J. B. Wall, S. Alimohamadi, M. Zheng, C. Yin, W. Shen, J. F. Prins, J. Liu, L. Qian, Single-cell transcriptomics reconstructs fate conversion from fibroblast to cardiomyocyte. Nature 551, 100–104 (2017).

38. B. Treutlein, Q. Y. Lee, J. G. Camp, M. Mall, W. Koh, S. A. M. Shariati, S. Sim, N. F. Neff, J. M. Skotheim, M. Wernig, S. R. Quake, Dissecting direct reprogramming from fibroblast to neuron using single-cell RNA-seq. Nature 534, 391–395 (2016).

39. H. Salmon, J. Idoyaga, A. Rahman, M. Leboeuf, R. Remark, S. Jordan, M. Casanova-Acebes, M. Khudoynazarova, J. Agudo, N. Tung, S. Chakarov, C. Rivera, B. Hogstad, M. Bosenberg, D. Hashimoto, S. Gnjatic, N. Bhardwaj, A. K. Palucka, B. D. Brown, J. Brody, F. Ginhoux, M. Merad, Expansion and activation of CD103+ dendritic cell progenitors at the tumor site enhances tumor responses to therapeutic PD-L1 and BRAF inhibition. Immunity 44, 924–938 (2016).

40. S. H. Naik, A. I. Proietto, N. S. Wilson, A. Dakic, P. Schnorrer, M. Fuchsberger, M. H. Lahoud, M. O'Keeffe, Q. X. Shao, W. F. Chen, J. A. Villadangos, K. Shortman, L. Wu, Cutting edge: Generation of splenic CD8+ and CD8− dendritic cell equivalents in fms-like tyrosine kinase 3 ligand bone marrow cultures. J. Immunol. 174, 6592–6597 (2005).

41. C. T. Mayer, P. Ghorbani, A. Nandan, M. Dudek, C. Arnold-Schrauf, C. Hesse, L. Berod, P. Stüve, F. Puttur, M. Merad, T. Sparwasser, Selective and efficient generation of functional Batf3-dependent CD103+ dendritic cells from mouse bone marrow. Blood 124, 3081–3091 (2014).

42. L. Qian, Y. Huang, C. I. Spencer, A. Foley, V. Vedantham, L. Liu, S. J. Conway, J. D. Fu, D. Srivastava, In vivo reprogramming of murine cardiac fibroblasts into induced cardiomyocytes. Nature 485, 593–598 (2012).

43. T. Nakano, H. Kodama, T. Honjo, Generation of lymphohematopoietic cells from embryonic stem cells in culture. Science 265, 1098–1101 (1994).

44. R. N. La Motte-Mohs, E. Herer, J. C. Zúñiga-Pflücker, Induction of T-cell development from human cord blood hematopoietic stem cells by Delta-like 1 in vitro. Blood 105, 1431–1439 (2005).

45. J. Karttunen, S. Sanderson, N. Shastri, Detection of rare antigen-presenting cells by the lacZ T-cell activation assay suggests an expression cloning strategy for T-cell antigens. Proc. Natl. Acad. Sci. 89, 6020–6024 (1992).

46. T. Brambrink, R. Foreman, G. G. Welstead, C. J. Lengner, M. Wernig, H. Suh, R. Jaenisch, Sequential expression of pluripotency markers during direct reprogramming of mouse somatic cells. Cell Stem Cell 2, 151–159 (2008).

47. Z. Liu, O. Chen, J. B. J. Wall, M. Zheng, Y. Zhou, L. Wang, H. Ruth Vaseghi, L. Qian, J. Liu, Systematic comparison of 2A peptides for cloning multi-genes in a polycistronic vector. Sci. Rep. 7, 2193 (2017).

48. D. Hockemeyer, F. Soldner, E. G. Cook, Q. Gao, M. Mitalipova, R. Jaenisch, A drug-inducible system for direct reprogramming of human somatic cells to pluripotency. Cell Stem Cell 3, 346–353 (2008).

49. X. Qiu, A. Hill, J. Packer, D. Lin, Y.-A. Ma, C. Trapnell, Single-cell mRNA quantification and differential analysis with Census. Nat. Methods 14, 309–315 (2017).

50. C. Trapnell, D. Cacchiarelli, J. Grimsby, P. Pokharel, S. Li, M. Morse, N. J. Lennon, K. J. Livak, T. S. Mikkelsen, J. L. Rinn, The dynamics and regulators of cell fate decisions are revealed by pseudotemporal ordering of single cells. Nat. Biotechnol. 32, 381–386 (2014).

51. Z. Ji, H. Ji, TSCAN: Pseudo-time reconstruction and evaluation in single-cell RNA-seq analysis. Nucleic Acids Res. 44, e117–e117 (2016).

52. A. Muhlethaler-Mottet, L. A. Otten, V. Steimle, B. Mach, Expression of MHC class IImolecules in different cellular and functional compartments is controlled by differential usage of multiple promoters of the transactivator CIITA. EMBO J. 16, 2851–2860 (1997).

53. A. Dobin, C. A. Davis, F. Schlesinger, J. Drenkow, C. Zaleski, S. Jha, P. Batut, M. Chaisson, T. R. Gingeras, STAR: Ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).

54. M. I. Love, W. Huber, S. Anders, Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).

55. J. T. Leek, svaseq: Removing batch effects and other unwanted noise from sequencing data. Nucleic Acids Res. 42, e161–e161 (2014).

Acknowledgments: We thank the members of the Pereira laboratory for useful discussions and S. Pedreiro for technical assistance. We thank L. Graça (iMM, Lisboa, Portugal) for OT-II/Rag2KO mice, S. Amigorena (Institute Curie, Paris, France) for B3Z hybridoma cells, and H. Ahlenius and E. Quist (LSCC, Lund, Sweden) for HEF cells. We also thank LBIC and S. Wasserstrom for SEM assistance. We thank F. Granucci (University of Milano Bicocca, Milan, Italy) and S. Hugues (University of Geneva, Geneva, Switzerland) for helpful discussions. Funding: F.F.R. and C.F.P. were supported by Foundation for Science and Technology (FCT) fellowships SFRH/BD/130845/2017 and SFRH/BPD/121445/2016, respectively. This project was co-funded by FCT, PAC CANCEL_STEM/2016, CENTRO-01-0145-FEDER-030013, Cancerfonden CAN 2017/745, Crafoord foundation 20180864, and Swedish Research Council 2018-02442. The Knut and Alice Wallenberg foundation is acknowledged for generous support. Author contributions: F.F.R., C.F.P., A.G.F., L.G.P., K.S., L.S., and C.A. conducted reprogramming experiments. F.F.R. and C.F.P. analyzed the data. F.F.R., C.F.P., I.K., A.M.G., and D.P. analyzed mRNA-seq datasets. O.S. and C.R.e.S. generated the Clec9a-tdT reporter mouse model and provided input. F.F.R., C.F.P., and C.-F.P. designed the experiments and wrote the manuscript. Competing interests: F.F.R., C.F.P., and C.-F.P. have filed a Patent Cooperation Treaty (PCT)(on 5 April 2018) protecting the intellectual property described here. Data and materials availability: The data reported in this paper are tabulated in the Supplementary Materials and deposited in the Gene Expression Omnibus database under accession number GSE103618.

Submitted 9 June 2018Accepted 3 October 2018Published 7 December 201810.1126/sciimmunol.aau4292

Citation: F. F. Rosa, C. F. Pires, I. Kurochkin, A. G. Ferreira, A. M. Gomes, L. G. Palma, K. Shaiv, L. Solanas, C. Azenha, D. Papatsenko, O. Schulz, C. Reis e Sousa, C.-F. Pereira, Direct reprogramming of fibroblasts into antigen-presenting dendritic cells. Sci. Immunol. 3, eaau4292 (2018).

by guest on Novem


munology.sciencem

ag.org/D

ownloaded from


Direct reprogramming of fibroblasts into antigen-presenting dendritic cells

Solanas, Cláudia Azenha, Dmitri Papatsenko, Oliver Schulz, Caetano Reis e Sousa and Carlos-Filipe PereiraFábio F. Rosa, Cristiana F. Pires, Ilia Kurochkin, Alexandra G. Ferreira, Andreia M. Gomes, Luís G. Palma, Kritika Shaiv, Laura

DOI: 10.1126/sciimmunol.aau4292, eaau4292.3Sci. Immunol.

generating other DC subsets by reprogramming.system could be useful in the clinic for generation of patient-specific DCs, and their studies open up the possibility of

T cells. Their+and had the ability to engulf and present antigens; murine iDCs could also cross-present antigens to CD8fibroblast cells to ''induced'' dendritic cells (iDCs). Human and murine iDCs were comparable to conventional type 1 DCs

found that ectopic expression of transcription factors PU.1, IRF8, and BATF3 reprogrammed mouse and humanal.etfunctions and in the development of immunotherapies. Taking a leaf from the regenerative medicine handbook, Rosa

In vitro systems that culture immune cells have contributed greatly in shaping our understanding of immune cell''Induced'' dendritic cells

ARTICLE TOOLS http://immunology.sciencemag.org/content/3/30/eaau4292

MATERIALSSUPPLEMENTARY http://immunology.sciencemag.org/content/suppl/2018/12/03/3.30.eaau4292.DC1

REFERENCES

http://immunology.sciencemag.org/content/3/30/eaau4292#BIBLThis article cites 54 articles, 19 of which you can access for free

Terms of ServiceUse of this article is subject to the

is a registered trademark of AAAS.Science ImmunologyNew York Avenue NW, Washington, DC 20005. The title (ISSN 2470-9468) is published by the American Association for the Advancement of Science, 1200Science Immunology

Science. No claim to original U.S. Government WorksCopyright © 2018 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of

by guest on Novem


munology.sciencem

ag.org/D

ownloaded from

http://immunology.sciencemag.org/content/3/30/eaau4292

http://immunology.sciencemag.org/content/suppl/2018/12/03/3.30.eaau4292.DC1

http://immunology.sciencemag.org/content/3/30/eaau4292#BIBL

http://www.sciencemag.org/about/terms-service


dendritic cells copyright © 2018 direct reprogramming of ...€¦ · in both dc development and...

Documents