Quantitative Proteomics Reveals Subset-Specific Viral ... - Cell Press

4 downloads 1454 Views 1MB Size Report
Nov 25, 2008 - proteomics, combined with label-free quantitation algorithms, to determine the ... from viruses activate IFN I secretion, which places the host in a ...... website. Source code for label-free algorithms incorporated into MaxQuant.
Immunity

Resource Quantitative Proteomics Reveals Subset-Specific Viral Recognition in Dendritic Cells Christian A. Luber,1,7 Ju¨rgen Cox,1,7 Henning Lauterbach,3 Ben Fancke,3 Matthias Selbach,4 Jurg Tschopp,5 Shizuo Akira,6 Marian Wiegand,2,8 Hubertus Hochrein,3 Meredith O’Keeffe,3,9,* and Matthias Mann1,* 1Proteomics

and Signal Transduction of Molecular Virology Max-Planck-Institute of Biochemistry, Am Klopferspitz 18, D-82152 Martinsried, Germany 3Research Immunology, Bavarian Nordic GmbH, Fraunhoferstr. 13, D-82152 Martinsried, Germany 4Intracellular Signalling and Mass Spectrometry, Max Delbru ¨ ck Center for Molecular Medicine, Robert-Ro¨ssle Str. 10, D-13125 Berlin, Germany 5Department of Biochemistry, University of Lausanne, CH-1066 Epalinges, Switzerland 6Department of Host Defense, Research Institute for Microbial Diseases, Osaka University and ERATO, Osaka 565-0871, Japan 7These authors contributed equally to this work 8Present Address: AmVac Research GmbH, Lochhamer Strasse 29, D-82152 Martinsried, Germany 9Present Address: Centre for Immunology, Burnet Institute, 85 Commercial Road, Melbourne Victoria, Australia 3004 *Correspondence: [email protected] (M.O.), [email protected] (M.M.) DOI 10.1016/j.immuni.2010.01.013 2Department

SUMMARY

Dendritic cell (DC) populations consist of multiple subsets that are essential orchestrators of the immune system. Technological limitations have so far prevented systems-wide accurate proteome comparison of rare cell populations in vivo. Here, we used high-resolution mass spectrometry-based proteomics, combined with label-free quantitation algorithms, to determine the proteome of mouse splenic conventional and plasmacytoid DC subsets to a depth of 5,780 and 6,664 proteins, respectively. We found mutually exclusive expression of pattern recognition pathways not previously known to be different among conventional DC subsets. Our experiments assigned key viral recognition functions to be exclusively expressed in CD4+ and double-negative DCs. The CD8a+ DCs largely lack the receptors required to sense certain viruses in the cytoplasm. By avoiding activation via cytoplasmic receptors, including retinoic acid-inducible gene I, CD8a+ DCs likely gain a window of opportunity to process and present viral antigens before activation-induced shutdown of antigen presentation pathways occurs.

INTRODUCTION Dendritic cells (DCs) are involved in immune responses ranging from resistance to infection to self-tolerance. Mouse spleen contains two types of DCs, plasmacytoid DCs (pDCs) and conventional DCs (cDCs). cDCs can be further segregated according to expression of surface markers into CD4+ cDCs, CD8a+ cDCs and CD4CD8a cDCs (double-negative [DN]) (Shortman and Naik, 2007; Vremec et al., 2000). Importantly, different DC subsets have specialized roles in immune

responses (Villadangos and Schnorrer, 2007; Villadangos and Young, 2008). pDCs uniquely secrete very large amounts of type I interferons (IFN I) directly upon activation and, therefore, play an important role in response to viral infections (Fuchsberger et al., 2005). Among the cDC subsets, CD8a+ cDCs are important in the presentation of antigen in the context of different viral infections (Allan et al., 2003; Belz et al., 2004, 2005). CD8a+ cDCs cross-present antigens, a process in which exogenous antigen is presented to T cells by MHC class I molecules to activate cytotoxic CD8+ T cells (Heath et al., 2004), so presentation of viral antigen by CD8a+ cDCs does not necessarily require direct infection. CD8a+ cDCs have been mainly associated with a T helper 1 (Th1) cell-inducing profile due to their ability to secrete extremely high amounts of the proinflammatory cytokine IL-12 (Maldonado-Lo´pez et al., 1999). CD4+ cDCs and DN cDCs are potent in stimulating CD4+ T cells via MHC II-antigen complexes. In contrast to CD8a+ cDCs, CD4+ and DN cDCs have been mainly associated with stimulating a Th2 cell response (Maldonado-Lo´pez et al., 1999). Innate immunity largely depends on the recognition of highly conserved structures of pathogens that are distinct from the host, so-called pathogen-associated molecular patterns (PAMPs) (Janeway and Medzhitov, 2002). PAMPs comprise two broad classes of biochemical compounds (Beutler et al., 2006): microbial products such as lipopolysaccharide and lipoteichoic acid and nucleic acids derived from pathogens, particularly viruses. Receptors involved in recognition of nucleic acids from viruses activate IFN I secretion, which places the host in a state of general alert (Stetson and Medzhitov, 2006). DCs express a variety of pattern recognition receptors (PRRs), such as toll-like receptors (TLRs), retinoic-acid-inducible gene (RIG)-like helicases (RLHs), or nucleotide-binding domain and leucine-rich repeat-containing molecules (NLRs) (Akira et al., 2006). The TLRs recognizing nucleic acids, namely TLR3 (sensing doublestranded RNA [dsRNA]), TLR7 and TLR8 (sensing singlestranded RNA [ssRNA]), and TLR9 (sensing CpG-DNA), are located in endosomal compartments within the cytosol (Alexopoulou et al., 2001; Heil et al., 2004; Hemmi et al., 2000). TLRs Immunity 32, 279–289, February 26, 2010 ª2010 Elsevier Inc. 279

Immunity

280 Immunity 32, 279–289, February 26, 2010 ª2010 Elsevier Inc.

CD4

A

CD11c

CD8α

Sirpα

B CD45RA

in general elicit cellular responses by recruiting several adaptor molecules, such as myeloid differentiation primary response gene 88 (Myd88) and toll-like receptor adaptor molecule 1 (Ticam1), leading to the expression of inflammatory genes (Kawai and Akira, 2008). DC subsets differ in their ability to sense viral infections. pDCs largely rely on the TLR system, namely TLR7 and TLR9, whereas cDCs mainly use the cytoplasmic virus sensors RIG-I and melanoma differentiation antigen 5 (MDA5) (Kato et al., 2005). Several studies using mice lacking RIG-I and MDA5 reported the importance of both cytoplasmic sensors for different virus infection models in vivo and in vitro (Kato et al., 2006, 2008). RIG-I and MDA5 recruit adaptor molecules like Mavs (also called Cardif, VISA, or IPS-1) (Meylan et al., 2005) that are distinct from those that mediate TLR signaling, although both pathways converge on IFN regulatory factors (IRFs) and NF-kB for the production of inflammatory cytokines and IFNs (Kawai and Akira, 2008). The molecular basis for the functional segregation of cDC subsets is still incompletely defined. We reasoned that differences in protein abundance might provide us with direct insights into functional differences among cDC subsets. Mass spectrometry (MS)-based proteomics can identify and quantify thousands of proteins in complex samples (Aebersold and Mann, 2003; Ong and Mann, 2005; Panchaud et al., 2008) Most quantitative MS methods rely on differential labeling of protein samples with stable isotopes. Our laboratory developed stable isotope labeling with amino acids in cell culture (SILAC) as a strategy that can accurately determine protein expression ratios (Ong et al., 2002). We showed recently that the SILAC approach can also be used to label entire mice (Kru¨ger et al., 2008). Unfortunately, comparison of DC subsets in vivo would require pooling of cells isolated from a large number of labeled animals and, therefore, be prohibitively expensive. An alternative to stable isotope labeling is ‘‘label-free’’ quantitation. In this case, peptide intensities measured during individual liquid chromatography (LC) runs are compared across runs (Bantscheff et al., 2007). Label-free quantitation is attractive because it can be applied to any proteomic sample without the need of introducing isotopes for quantitation. However, label-free methods are traditionally much less accurate than isotope-based methods, and proteome-wide quantitation has not yet been possible (Schiess et al., 2009; Xu et al., 2008). Major challenges in labelfree quantitation are difficulties in the matching of thousands of peptides across samples, variability in LC-MS resulting in retention time shifts, and errors introduced by slight differences in sample fractionation steps. Here, we studied the differences in abundance of proteins in DC subsets using an algorithm for label-free quantitation, which we have recently developed (J.C., C.A.L., N. Nagaraj, and M.M., unpublished data). We introduce this MS-based label-free quantitation approach to profile protein abundance differences of cDC subsets to a depth of more than 5,000 proteins, requiring only 1.5–2.0 3 106 purified cells. This technology now makes it possible to study closely related cell types in vivo. Expression profiles showed substantial overlap but also highly informative differences in protein composition among the three cDC subsets. Our analysis revealed mutually exclusive expression of pattern recognition pathways not previously known to be subset specific. We showed that members of the NLR, TLR,

CD8α

Label-free Quantitative Proteomics of DC Subsets

CD11c

CD45RA

Figure 1. FACS Separation of cDC Subsets and pDCs (A) Enriched CD11chigh DCs were separated by flow cytometry into the three cDC subsets, according to CD4 and CD8a expression. (B) Enriched CD11cint DCs were separated by flow cytometry into pDCs, according to CD45RA and Sirpa expression.

and RLH signaling pathways were differentially expressed in cDC subsets. Among these, RIG-I and MDA5 were significantly higher or uniquely expressed by CD4+ and DN cDCs (p % 0.01, standard t test). Only CD4+ and DN cDC subsets, by virtue of their selective expression of RLH, were activated in response to direct infection with certain ssRNA viruses and were able to produce key antiviral cytokines, such as IFN-a. Although infected, CD8a+ cDCs were not themselves activated and retained their ability to cross-present in the presence of virus. Thus, labelfree, proteome-wide quantitation assigns and clarifies key viral recognition functions to cDC subsets. RESULTS Quantitative Proteomic Comparison of DC Subsets To identify differentially expressed proteins among spleen DC subsets, we used a MS-based proteomics approach. The cDCs were sorted into subsets by flow cytometry, according to the expression of CD8a and CD4 surface molecules (Figure 1A). cDC preparations from pooled spleens of 32 mice consistently yielded more than 2.5 3 106 cDCs per subset with purity higher than 95% (data not shown). Spleen pDCs were also sorted in separate experiments based on the expression of CD11c, CD45RA, and Sirpa (Figure 1B). Sorted pDCs and CD4+, CD8a+, and DN cDCs, resulting in 15 to 20 mg total protein each, were separated by one dimensional (1D) SDS-PAGE and, after tryptic in-gel digestion, analyzed by online LC-MS. We repeated this experiment twice, with different pools of spleens resulting in biological triplicates. MS data from all 126 gel slices of the three independent large scale cDC subset experiments were combined and analyzed by MaxQuant (Cox and Mann, 2008). We incorporated new algorithms for label-free

Immunity Label-free Quantitative Proteomics of DC Subsets

Figure 2. Quantitative Differences between cDC Subsets

8

A

N fkb 2

-log10[p]

6

Itg a m

4

2

0

P yg l

P p t1

T lr1 2 F u ca 1 T lr3

B

Irf8

L cn 2 Cd8a

Cd36

C ysC

D d r1 S irp a

L a ss6 T g m 2 N ca ld

R ig -I

Cd205 M p e g 1 P d ia 5

- ∞ -8

O a s1

-4

0

4

-log10[p]

6



Cd1d

(99,228 nonredundant sequences), corresponding to 5359, 5642, and 4830 proteins in CD4+, CD8a+, and DN cDCs E m r1 with 99% certainty, respectively. Cd4 Summed peptide intensity is a good proxy for absolute protein abundance (de Godoy et al., 2008), and by this measure, we covered the cDC proteomes across more than four orders of magni8 ∞ tude. The cDC proteome of 5,780 proteins shows no bias against low-level regulatory proteins such as kinases (209 identified or 3.6% of our proteome versus 4.2% in the genome). Relative label-free quantitation was highly reproducible between biological replicates and correlation between normalized protein intensities was between 0.84 and 0.96 (Figures S1A– S1C). It thus provides an accurate means of comparing differences in protein expression between cDC subsets in vivo 8 ∞ for thousands of proteins (Table S1). Expression of most proteins was similar, and only a relatively small number showed statistically significant and highly DN reproducible differences between any two subsets (standard t test p value % 0.01) (Figures 2A–2C and Figures S1D– S1F). Overall, differences in protein expression patterns showed that CD4+ and DN cDCs were more closely related to each other than to CD8a+ cDCs (judged by the median of the outliers, Figures 2A–2C) as already suggested by microarray analysis (Edwards et al., 2003a). We then compared the label-free quantitation data with known marker proteins of cDC subsets. All subsets expressed CD11c, the pan-marker of DCs. Markers for other lymphocyte lineages, such as CD19, CD3 subunits, or Nk1.1, were not detected, M ica l3

2

0

P sa

Irf8

4

T lr3 F u ca 1

- ∞ -8

N fkb 2 L cp 2 T rim 3 4 P kcb

F ln b A rsb C ysC F ln a

G im a p 4

-4

0

4

log2[CD4+ / DN] ratio

8

6

Cd22 T a f6 S e rp in b 6 b

-log10[p]

8

log2[CD4+ / CD8α+] ratio

8

C

Il1 rn C d4

(A–C) Volcano plot of protein expression differences between cDC subsets as a function of statistical significance (standard t test p value % 0.01) as indicated for CD4+ (red), CD8a+ (green), and DN (blue) cDCs. Proteins with no statistically significant difference in expression between subsets (p > 0.01) are in gray. Proteins with no detectable signal in one of the subsets were assigned a ratio of infinity. Values are median fold changes of three separate DC preparation and MS experiments. (D) Filled histograms show expression of CD97 by cDC subsets as indicated. Unfilled histograms show unstained background controls. Data is representative of two experiments.

Cd8a

4

2

0

Cd1d1

Irf8 S e rp in b 9 P p t1 F u ca 2

Fgr Cd11b Tgm 2 R ig -I S irp a

Cd36 Cd205

- ∞ -8

-4

0

4

log2[DN / CD8α+] ratio

D CD4+

CD8α+

CD97

quantitation into MaxQuant, which enabled us to quantify peptides across individual MS runs (J.C., C.A.L., N. Nagaraj, and M.M., unpublished data). Analysis of the entire cDC data set using uniform statistical criteria identified 1,283,676 peptides

Immunity 32, 279–289, February 26, 2010 ª2010 Elsevier Inc. 281

Immunity Label-free Quantitative Proteomics of DC Subsets

Table 1. Expression of Known Marker Proteins Determined by Mass Spectrometry CD8a+ cDC

CD4+ cDC

CD11c

57,823,000

78,828,000

55,484,000

CD8a

3,876,600

0

0

CD205

6,461,300

381,840

443,580

CD36

3,421,300

92,491

445,720

Necl2

950,650

0

0

CD4

0

1,096,000

0

Sirpa

139,680

6,585,700

4,044,100

33D1

0

950,990

694,500

CD11b

1,211,600

17,304,000

16,806,000

DN cDC

Numbers are the median of summed peptide intensities in ion counts per second over three biological replicates.

indicating that sorted cDC populations were indeed pure. The surface markers CD4 and CD8a were expressed by the respective population and correctly assigned by our bioinformatic algorithms. Importantly, this was also true for other known surface markers, such as DEC-205, CD36, and Necl2 for CD8a+ cDCs and CD11b, 33D1, and SIRPa for CD4+ and DN cDCs (Table 1). Clec9a was recently described as a C-type lectin-like molecule with restricted expression on CD8a+ cDCs in mice and a subset of human blood DCs (Caminschi et al., 2008; Huysamen et al., 2008; Sancho et al., 2009). Thus, similarly to Necl2 (Galibert et al., 2005), it may be useful to align human and mouse cDC subsets. We detected two peptides of Clec9a exclusively in CD8a+ cDCs (however, despite unambiguous identification, we could not quantify this detectable Clec9a expression as differentially significant because of its low signal to noise). Correct assignment of these key surface markers by hypothesis-free quantitative label-free proteomics demonstrates the robustness of our approach. CD97 was 4-fold more abundant in CD8a+ compared to CD4+ cDCs. Flow cytometry with antibodies against CD97 independently confirmed our proteomic finding (Figure 2D). Previous microarray data did not classify CD97 as differentially expressed among cDCs (Dudziak et al., 2007; Edwards et al., 2003a). Our data illustrates that proteome and transcriptome measurements do not necessarily agree, either due to technical factors or due to additional regulation at the protein level (Bonaldi et al., 2008). Mice that lack IFN regulatory factor 8 (IRF-8) essentially lack CD8a+ cDCs and pDCs (Tamura et al., 2005), and we found that this transcription factor was specifically expressed in CD8a+ compared to CD4+ cDCs (Table S1). Mice that lack IRF-4 (Tamura et al., 2005) and the NF-kB subunit RelB (Wu et al., 1998) have reduced numbers of CD4+ and DN cDCs. In concordance with this genetic finding, these transcription factors were much more abundant in CD4+ cDCs (Table S1). The NF-kB family members NfkB1 and NfkB2 have a similar expression pattern to RelB, whereas RelA and c-Rel have no subtype specificity, suggesting a functional specialization of the NF-kB pathway in cDC subtypes. This was previously only suggested by differential effects on DC subsets in mutant mice lacking members of the NF-kB family (O’Keeffe et al., 2005; Wu et al., 1998). 282 Immunity 32, 279–289, February 26, 2010 ª2010 Elsevier Inc.

Protein Expression Reflects Different Functionality of cDC Subsets PRRs endow DCs and other cells with the ability to sense PAMPs (Akira et al., 2006). We reasoned that differences in protein abundance of individual PRRs might provide us with direct insights into functional differences between cDC subsets. It was previously known that certain TLRs are differentially expressed by cDCs (TLR7 only in CD4+ and DN cDCs; TLR3 mainly in CD8a+ cDCs), and this was confirmed by our data (Table S1) (Edwards et al., 2003b). Analysis of the expression of other TLRs identified in our analysis revealed that the expression of two poorly characterized members of the TLR family, TLR12 and TLR13, were mainly restricted to CD8a+ cDCs. Conversely, members of the NLRs, NOD1, and IPAF were much more highly expressed in CD4+ and DN cDCs (Figure 3). CD8a+ cDCs are known to be equipped with antiviral functions via their ability to present viral antigens, including their unique ability to cross-present, their ability to produce high amounts of IL-12, which activates antiviral NK cells and T cells (Allan et al., 2003; Belz et al., 2004, 2005; Maldonado-Lo´pez et al., 1999), and their high expression of TLR3 and TLR9 (Table S1) (Edwards et al., 2003b). In contrast, the CD4+ and DN cDC subsets have been associated with the ability to induce Th2 cell responses and, besides their expression of TLR7 and -9, have not been shown to play a major role in antiviral defense. Our data showed a higher expression of the viral RNA recognition molecules RIG-I and MDA5 (Kato et al., 2006, 2008) in CD4+ and DN DCs (Figure 3). These RLHs signal through interaction with Mavs, which relays the signal to downstream activation of transcription factors and IFN I response and proinflammatory responses (Meylan et al., 2005). Consistently, NLRX1, a potent regulator of Mavs, is also more highly expressed in CD4+ cDCs (Figure 3; Table S1). RIG-I was described to be essential for induction of IFN I upon RNA virus infection of all cDCs (Kato et al., 2005). Our results now suggest that not all cDC subsets are equally equipped to sense RNA viruses. This is further supported by the higher expression in CD4+ and DN cDCs of OAS1 and OAS3, which are 20 ,50 -oligoadenylate synthetases (OAS) involved in antiviral defense (Silverman, 2007). The restriction of cytoplasmic viral recognition to CD4+ and DN cDCs does not appear to be limited to RNA viruses or RNA intermediates because we also identified DAI (also known as DLM1 or ZBP1), a cytosolic PRR activated by dsDNA (Takaoka et al., 2007), to be more abundant in CD4+ and DN cDCs than in CD8a+ cDCs (Figure 3; Table S1).

The Proteome of pDCs Having established an in-depth proteome of cDCs, we also wished to provide the proteome of the other main population of DCs, the plasmacytoid DCs. In a triplicate series of experiments, we separately determined the proteome of pDCs to a depth of 6,664 proteins with 99% certainty (Table S2). Our measurements between biological triplicate experiments were highly reproducible (Figure S2). As expected, pDCs express IRF-4, IRF-7, and IRF-8 (Table S2) (Tailor et al., 2006). PRRs detected in the proteome of pDCs include TLR7 and TLR9 and, like CD4+ and DN cDC subsets, this DC population also expressed RIG-I and MDA5 (Table S2). In common with CD8a+ cDCs, pDCs have TLR12 (Table S2). Thus, the pDCs show expression

Immunity Label-free Quantitative Proteomics of DC Subsets

Figure 3. Differential Expression of PRRs between CD4+ and CD8a+ Splenic cDCs Members of PRR pathways are color-coded according to statistical significance (standard t test p value), with red denoting higher expression in CD4+ cDCs with p % 0.01. Green denotes higher expression in CD8a+ cDCs with p % 0.01. Gray filled boxes designate proteins with no statistically significant change between subsets (p > 0.01). Open boxes with an asterisk indicate proteins that have been detected but were not quantified. Remaining open boxes represent undetected proteins. Numbers in parentheses are fold changes determined by label-free proteomics. Pathways are adapted from Hara et al. (2008), Ishikawa et al. (2009), Kawai and Akira (2007), Komuro et al. (2008), Moore et al. (2008), Robinson et al. (2009), Saitoh and Miyake (2009), Silverman (2007); Tabeta et al. (2004), Takaoka et al. (2007), and Yamamoto-Furusho et al. (2006).

of both endosomal and cytoplasmic PRRs, overlapping with the CD8a+, CD4+, and DN cDCs. Differential Sensing of Virus Infection among DC Subsets We directly tested the hypothesis arising from our unbiased quantitative proteomics experiment that cDC subsets differ in their ability to sense viral infection. We challenged purified cDC subsets with Sendai virus, a known activator of RIG-I (Kato et al., 2006). The activation-specific marker CD86 was increased in CD4+ and DN cDCs after viral infection, whereas CD8a+ cDCs did not show any upregulation of CD86 (Figure 4A). We measured IFN-a production after viral infection and found that— corresponding to cell surface maturation—only CD4+ cDCs and DN cDCs produced IFN-a. The production of IL-6 and IFN-b in response to Sendai virus was also restricted to CD4+ and DN cDC subsets (data not shown). Thus only CD4+ and DN cDCs, but not CD8a+ cDCs, respond to direct Sendai virus infection. Culture of cDC subsets with Sendai-GFP virus revealed that all three subsets highly expressed GFP, demonstrating they were similarly infected (Figures S3A–S3C). It was possible that differences in virus detection depend on differential TLR signaling rather than on differential expression of RLHs. We ruled out this possibility using mice lacking Myd88 (Adachi et al., 1998), an adaptor protein for most TLR

signaling pathways. After viral infection of cDCs from these mice, both CD4+ and DN cDCs, but not CD8a+ cDCs, upregulated CD86 and produced IFN-a, showing that responsiveness to Sendai virus is functional in the absence of TLR signaling (Figure 4B). To further verify that the RIG-I antiviral pathway is indeed the unique feature enabling CD4+ and DN cDCs, but not CD8a+ cDCs, to recognize direct viral infection, we tested responses of DCs derived from Mavs/ mice (Michallet et al., 2008). Without Mavs, none of the cDC subsets is expected to recognize Sendai virus infection via RIG-I. Virus infection indeed did not result in any detectable surface activation or IFN-a production from any of the cDC subsets of Mavs/ mice (Figure 4C). In contrast, the production of IFN-a and other cytokines by pDCs in response to different viruses was not affected in Mavs/ mice (Figure S4). Cultures of bone marrow (bm) with fms-like tyrosine kinase 3 ligand (FL) to generate FLDCs containing phenotypic and functional equivalents of CD8a+ cDCs (eCD8+) and equivalents of both CD4+ and DN cDCs (eCD8) (Naik et al., 2005) were also prepared from WT, Myd88/, and Mavs/ mice. eCD8 DCs, but not eCD8+ DCs of WT and Myd88/ mice, were still able to produce IFN-a in response to Sendai virus, but this response was completely abrogated in Mavs/ mice (Figure 4D). Thus, in the bm model, Sendai virus recognition by eCD8 DCs is also specifically dependent on the RLH pathway, but not on the Immunity 32, 279–289, February 26, 2010 ª2010 Elsevier Inc. 283

Immunity Label-free Quantitative Proteomics of DC Subsets

WT

A

CD8α

1200

+

CD8α CD4+ DN

IFN-α (U/ml)

1000 CD4+

DN

Figure 4. Response of WT, Myd88/, and Mavs/ DC Subsets to Sendai Virus Infection In Vitro

+

(A–C) Histograms showing expression of CD86 of sorted splenic cDC subsets of (A) WT, (B) Myd88-, and (C) Mavs-deficient mice after incubation with Sendai virus (filled histograms) or without stimulation (open histograms). Sorted splenic cDC subsets of (A) WT, (B) Myd88- and (C) Mavsdeficient mice, as indicated, were stimulated with Sendai virus, and supernatants were analyzed for IFN-a. (D) Sorted spleen cDC equivalents of bonemarrow-derived FLDCs from WT, Myd88-, and Mavs-deficient mice were infected with Sendai virus, and supernatants were analyzed for IFN-a by ELISA. Representative results of at least three experiments are shown. Data represent mean ± SD. *not detected.

800 600 400 200

*

0

*

Myd88-/-

400

CD8α+

IFN-α (U/ml)

DN

Mavs

200 100

*

0

CD86

C

*

400

IFN-α (U/ml)

DN

600 400 200 *

*

Control

*

SeV

*

*

Myd88-/-

75 50 25 *

*

Control

*

CD8a+ cDCs Are Functionally Active in the Presence of Virus The lack of CD8a+ cDC activation in response to infection with either Sendai Mavs or Flu virus could, in principle, be due to eCD8 rapid killing or inactivation by these eCD8 viruses. To exclude this possibility, we infected CD8a+ cDCs with Sendai virus in the presence of phosphorothioated * * * * CpG-motif containing oligonucleotides Control SeV (CpG1668) and polyionosinic-polycytidylic acid (poly I:C). This TLR9 and TLR3 ligand combination is a stimulus known to induce high amounts of cytokines from CD8a+ cDCs (Hochrein et al., 2001). The production of IL-12p70, IL-6, TNF-a, MIP-1a, and MIP-1b by the CD8a+ cDCs were not affected by the additional presence of Sendai virus (Figure S5A). Of note, and in line with these observations, viral infection also did not enhance the production of any of these cytokines. The direct and cross-presentation capabilities of CD8a+ cDCs are important in the course of viral infection (Allan et al., 2003). To test whether the nonresponsiveness of CD8a+ cDCs to Sendai virus was concomitant with an impaired ability to cross-present antigen, we tested if CD8a+ cDCs could cross-present antigen in the presence of Sendai virus. As shown in Figure S5B, the CD8a+ cDCs cross-presentation of ovalbumin (OVA) peptides derived from OVA-coated spleen cells, to OT-I T cells, is still functional in the presence of Sendai virus. The number of proliferating OT-I T cells is reduced by about 20% but was still about ten times greater than the proliferation induced by the other cDC subsets. -/-

150

eCD8 eCD8-

100

* SeV

+

125

0

*

Control

150

+

IFN-α (U/ml)

IFN-α (U/ml)

*

0

eCD8 eCD8-

Collectively, these results demonstrate that CD4+ and DN cDCs, in contrast to CD8a+ cDCs, are uniquely equipped with a functional RLH pathway to sense viral infection with ssRNA viruses.

200

WT

0

SeV

100

CD86

800

*

CD8α+ CD4+ DN

300 CD4+

1000

*

Control

-/-

CD8α+

D

SeV

CD8α+ CD4+ DN

300 CD4+

*

IFN-α (U/ml)

B

*

Control

CD86

*

SeV

125 100 75 50 25 0

TLR pathway. It should be noted that although the IFN-a response of CD4+, DN cDCs, and eCD8 DCs in response to Sendai virus was completely dependent upon Mavs, the amounts of IFN-a produced by these cells from Myd88/ mice were less than 50% of WT amounts (Figure 4). This suggests, as previously shown in fibroblasts and macrophages (Rasmussen et al., 2009), that TLR and RLH may synergistically cooperate in CD8 DC subsets to induce maximal cytokine responses to viral infection. Next, we wanted to demonstrate that the subset specific viral recognition was not restricted to a particular virus and employed another ssRNA virus, NS1-deleted mutant influenza A virus (Flu). As with Sendai virus, CD8a+ cDCs were not responsive to Flu virus, as demonstrated by a lack of IFN-a, IL-6, MIP-1a, and MIP-1b production (Figure 5). Likewise, in concordance with Sendai virus infection, CD4+ and DN cDCs produced these cytokines in response to Flu virus, and this response was absolutely dependent upon the RLH pathway, as demonstrated by its abrogation in Mavs/ mice (Figure 5). 284 Immunity 32, 279–289, February 26, 2010 ª2010 Elsevier Inc.

+ -

Immunity Label-free Quantitative Proteomics of DC Subsets

WT

CD8α+ CD4+ DN

40

30 IFN-α (U/ml)

IFN-α (U/ml)

30 20 10

*

*

*

*

Control WT

CD8α+ CD4+ DN

250

150 100 50

50

Control

0

Flu WT

CD8α+ CD4+ DN

250

Control

MIP-1α (pg/ml)

100 50

150 100 50

Control

0

Flu

WT

CD8α+ CD4+ DN

400

200

Control

Flu

Control

*

CD8α+ CD4+ DN

To our knowledge, our data set is the most comprehensive quantitative proteome comparison of different in vivo cell populations obtained so far. We achieved reliable and reproducible quantitation for more than 5,000 proteins from a few micrograms of rare cell types directly isolated from mice. The approach does not involve metabolic labeling and Flu can thus be used to quantify in vivo Mavs CD8α protein levels in all organisms, including CD4 humans. The approach was validated by DN its ability to identify known subsetspecific markers. Subsetspecific proteins were detected and validated, including CD97, which was not classified as differentially expressed by microarray studies (Dudziak et al., 2007; Edwards et al., 2003a). Our results demonstrate that Flu label-free quantitation can be used to directly compare amounts of thousands Mavs CD8α of proteins in vivo. CD4 Previous functional and microarray DN data demonstrated selective expression of PRRs among cDC subsets. Our proteomic analyses additionally revealed that the orphan receptors TLR12 and TLR13 were found to be more abundant in CD8a+ cDCs. Thus, among cDCs, the ligands of these TLR would only stimulate Flu CD8a+ cDCs. The pDC proteome also reveals their expression of TLR12. Identifying TLR12 and -13 ligands may provide further insights into the functional capacity of these DC subsets. The cytosolic, inflammasome-associated NLRs are involved in the activation of caspase-1 and the secretion of proinflammatory cytokines, such as IL-1 and IL-18 (Kanneganti et al., 2007). We found higher expression of the NLR-associated molecules NOD1, IPAF, CARD9, and CARD11 in CD4+ and DN cDCs, suggesting further functional separation among cDC subsets. Finally, we show that RLHs, cytoplasmic RNA detectors, and some key RLH signaling molecules are expressed more abundantly or even exclusively in CD4+ and DN cDCs. Only these cDC subsets produce IFN I and other cytokines and display an activated surface phenotype upon infection with the ssRNA Sendai or flu viruses. This viral recognition was specific for the RLH pathway in these cDCs because it was functional in Myd88 mutant mice but abolished in mice lacking the essential RLH +

+

-/-

+

600

MIP-1β (pg/ml)

600

* Flu

+

200

150

*

-/-

250

200

0

*

Mavs-/-

IL-6 (pg/ml)

IL-6 (pg/ml)

100

0

* Control

200

150

0

*

250

200

MIP-1α (pg/ml)

20

0

Flu

Figure 5. Cytokine Secretion of WT and Mavs/ DC Subsets to Flu Virus Infection In Vitro Sorted splenic cDC subsets of WT and Mavs-deficient mice, as indicated, were stimulated with Flu virus, and supernatants were analyzed for IFN-a, IL-6, MIP-1a, and MIP-1b. Representative results of at least three experiments are shown. Data represent mean ± SD. *not detected.

10

0

MIP-1β (pg/ml)

CD8α+ CD4+ DN

Mavs-/-

40

400

200

0

Control

Our data examining the response of CD4+ and DN cDC subsets to ssRNA viruses, such as Sendai and Flu virus, that require the RLH pathway for recognition indicated that, indeed, these cells are activated via the RLH pathway. CD8a+ cDCs, on the other hand, do not express these molecules and, indeed, when infected with either Sendai or Flu virus, were not activated. However, in presence of these viruses, CD8a+ cDCs maintain the ability to cross-present antigen and respond to TLR3 and TLR9 stimuli.

DISCUSSION Here, we showed that comprehensive MS-based proteomics, combined with label-free quantitation algorithms, can determine the differences in protein abundance for a substantial part of the proteome of cDC subsets in vivo.

Immunity 32, 279–289, February 26, 2010 ª2010 Elsevier Inc. 285

Immunity Label-free Quantitative Proteomics of DC Subsets

signaling molecule Mavs. pDCs also expressed RLHs and downstream signaling molecules, but the response of pDCs to either Sendai or Flu virus was not affected in Mavs/ mice. Thus, in line with previous studies (Kato et al., 2005), pDCs utilize mainly TLR-dependent mechanisms for viral recognition. It has recently been shown that pDCs utilize the RLH pathway in response to viruses replicating in the cytoplasm in the absence of IFN I receptor signaling (Kumagai et al., 2009). The expression of RLHs and the downstream signaling molecules by pDCs appear particularly important in response to viruses that may escape TLR-dependent recognition. Unlike in macrophages (Imaizumi et al., 2006), the high levels of RIG-I expressed by CD4+ and DN cDC subsets of mouse spleen were constitutive and did not require prior cell activation, suggesting that the mouse CD4+ and DN cDCs are already ‘‘primed’’ for rapid cytoplasmic virus recognition. This is analogous to the high expression of TLR7 and TLR9 by pDCs, allowing rapid recognition of endosomal pathogens. Thus, the CD4+ and DN cDCs are equipped as ‘‘professional’’ cytoplasmic virus sensors. Given the fact that these subsets also preferentially express NLR, our data raise the possibility that these cells are specialized for the rapid response to cytoplasmic pathogens. The RLHs are expressed nearly ubiquitously in hematopoietic and nonimmune cells (Yoneyama and Fujita, 2009), so the absence in CD8a+ cDCs may be important for the biology of these cells. CD8a+ cDCs are well known for their cross-presentation capacity. This capacity is abrogated within a short time frame upon stimulation (Wilson et al., 2006). Lack of responsiveness to RNA and lack of, or at least delay in, CD8a+ cDC activation might be necessary for effective cross-priming in RNA virus infections. Moreover, the RLH pathway has been implicated in the recognition of DNA (Choi et al., 2009), possibly, at least in the case of EBV, via RNA polymerase III (Ablasser et al., 2009). We hypothesize that cross-presentation by CD8a+ cDCs and direct presentation by virally infected CD8a+ cDCs and consequent CTL priming is given a ‘‘head start’’ during viral infection by avoiding the activation of CD8a+ cDCs by infecting virus. The presence of infecting virus does not dramatically affect the ability of CD8a+ cDCs to cross-present. Therefore, a non-activated CD8a+ cDC, even if itself is infected, would retain the ability to present antigen during early infection. Activation of CD8a+ cDCs would increase in an environment of viral-induced cell death because the expression of receptors such as Clec9a (Sancho et al., 2009), equipping CD8a+ cDCs for the uptake of dying cells, allows the subsequent transport of the viral-loaded dead cell cargo to the endosomes containing TLR3 and TLR9. As an interesting consequence of the lack of RLHs, CD8a+ cDCs, or their immediate precursors (Bedoui et al., 2009) could potentially be manipulated by certain siRNAs without activating these cells. Specifically, short hairpin RNAs (shRNAs) activate cells via RIG-I and not via TLR3 (Kenworthy et al., 2009), making shRNA an attractive tool for selective ablation of genes of interest in CD8a+ cDCs. In conclusion, our study demonstrates the feasibility of quantifying rare in vivo cell subpopulations at the protein level. The proteomic expression data provided immediate insights into the functional segregation of DC subsets. Direct infection of DC subsets ex vivo using Sendai or Flu virus confirmed our findings. The data produced in this study will provide extensive 286 Immunity 32, 279–289, February 26, 2010 ª2010 Elsevier Inc.

information for future detailed comparison of the proteomes of mouse cDC subsets with human cDC subsets. Such a comparison would be a powerful means for dissecting similarities and differences between the DCs across species. EXPERIMENTAL PROCEDURES Mice All mice were bred and maintained either in the animal facility at Bavarian Nordic GmbH or the Max Planck Institute of Biochemistry (MPIB), according to institutional guidelines. Experiments were generally done between 8–12 weeks of age. Mice lacking Myd88 were generated by S. Akira (Adachi et al., 1998) and backcrossed on a C57BL/6 background by H. Wagner. Mavs-deficient mice on a C57BL/6 background were provided by J. Tschopp (Meylan et al., 2005). b2M-deficient mice and transgenic OT-I mice were obtained from M. Suter (University of Zurich). C57BL/6 mice were provided from the animal facility of MPIB. Cells, Flow Cytometric Analysis, and Sorting DC subsets were isolated from mouse spleens from Myd88/, Mavs/, and C57BL/6 mice, as described (Vremec et al., 2007). cDC populations were segregated from pre-enriched DC preparations based on the expression of CD11c, CD45RA, CD4, and CD8a; pDCs were purified based on CD11c, CD45RA, and Sirpa (all BD Biosciences) expression. Cell sorting was performed on a FACS Aria instrument (BD Biosciences). Activation of DC subsets was determined by CD86 expression using a FACS Calibur instrument (BD Biosciences). Analysis of FACS data was performed with WEASEL software (WEHI). FLDCs were prepared as described (Naik et al., 2005). Recombinant murine FL was expressed in Chinese hamster ovary cells and purified in house as described previously (O’Keeffe et al., 2002). Fluorescent cell sorting into both eCD8+ and eCD8-DC subsets was done based on expression of CD11c, CD11b, CD24, CD103, and CD45R (all BD Biosciences). Sample Preparation for MS FACS-purified DC subsets were washed in PBS and immediately frozen on dry ice. Lysates were boiled in 23 sample buffer (NuPAGE, Invitrogen) and separated by 1D-SDS PAGE (4%–12% Bis-Tris Mini-Gel, Invitrogen). After Colloidal Blue Staining (Invitrogen), gel pieces were excised from the gel and subjected to reduction, alkylation and in-gel digestion with sequence grade modified trypsin (Promega) as described (Shevchenko et al., 1996). After digestion, peptides were extracted by 30% acetonitrile in 3% TFA in water, reduced in a speedvac, and desalted using StageTips before analysis by MS (Rappsilber et al., 2003). MS Analysis of cDC Subsets MS experiments of cDCs were performed on a nanoflow HPLC system (Agilent Technologies 1100 or 1200) connected to a hybrid LTQ-Orbitrap (Thermo Fisher Scientific), equipped with a nanoelectrospray ion source (Proxeon Biosystems). Peptide mixtures were separated by reverse phase chromatography using in-house-made C18 microcolumns (75 mm ID packed with ReproSil-Pur C18-AQ 3 mm resin, Dr. Maisch GmbH) with a 2H gradient from 5% to 60% acetonitrile in 0.5% acetic acid at a flow rate of 200 nl/min and directly electrosprayed into the mass spectrometer. The LTQ-Orbitrap was operated in the data dependent mode to simultaneously measure full scan MS spectra in the Orbitrap and the five most intense ions in the LTQ part by collisionally induced dissociation, respectively. MS Analysis of pDCs Peptide mixtures of pDCs were separated by reverse phase chromatography using in house-made C18 microcolumns (75mm ID packed with ReproSil-Pur C18-AQ 3 mm resin, Dr. Maisch, GmbH) in a 2.5H gradient from 5% to 60% acetonitrile in 0.5% acetic acid at a flow rate of 200 nl/min using a Proxeon easy-nLC (Proxeon Biosystems), which was directly connected to a LTQOrbitrap VELOS mass spectrometer (Thermo Fisher Scientific) via a nanoelectrospray ion source (Proxeon Biosystems). The LTQ-Orbitrap VELOS was operated in the data-dependent mode to simultaneously measure full scan

Immunity Label-free Quantitative Proteomics of DC Subsets

MS spectra in the Orbitrap and the ten most intense ions in the LTQ part by collisionally induced dissociation, respectively. Background ions were reduced by using an ABIRD device (ESI Source Solutions). Data Processing and Analysis The data analysis was performed with MaxQuant software, supported by Mascot as a database search engine for peptide identification (mouse database IPI 3.46) as described (Cox and Mann, 2008). Label-free quantitation algorithms were added to MaxQuant by extracting isotope patterns for each peptide in each run (J.C., C.A.L., N. Nagaraj, and M.M., unpublished data). These isotope patterns were matched to each other across runs using peptide identifications, very high mass accuracy, and nonlinearly remapped retention time. Total peptide signals within each run were normalized in order to make experiments comparable that were performed months apart. For label-free quantitation, we compared the maximum number of peptides between any two samples, resulting in a matrix of protein ratios, calculated as the median of all ratios for common peptides. We used least-squares regression to solve the overdetermined system of equations to obtain the best estimate for the protein ratios. Details of the algorithm will be described elsewhere (J.C., C.A.L., N. Nagaraj, and M.M., unpublished data). The proportion of kinases in the proteome and genome were determined using the gene ontology (GO) term 0004672. Virus Infection Purified Sendai virus (Cantell strain, Charles River Laboratories) and delNS1 Influenza A virus (A/PR/8/34, H1N1, AVIR Green Hills Biotechnology) were used for in vitro infection of sorted DC subsets from C57BL/6, Myd88/ and Mavs/ mice. 1 3 106 DCs per ml were infected with virus in RPMI 1640 medium supplemented with 10% FCS and antibiotics in the presence of IL-3 and GM-CSF. Supernatants were harvested after 18 hr and IFN-a, IFN-b, IL-6, TNF-a, MIP-1a, and MIP-1b were measured by ELISA as described (Hochrein et al., 2004) or by flow cytometric bead assay (FlowCytomix Simplex, Bender Medsystems) according to manufacturer’s protocol. In Vitro Cross-presentation Assay In vitro cross-presentation assay was performed as described (Schnorrer et al., 2006) with slight modifications. In short, we incubated ex vivo sorted spleen DC (1.5 3 104 cells/well) with 5 3 104 irradiated OVA-coated spleen cells (OCS) from b2M-deficient mice and 5 3 104 CFSE-labeled OT-1 Ly5.1 cells in the presence of a known number of Calibrite beads (Becton Dickinson) in complete medium containing 10 ng/ml GM-CSF. 500 nM CpG1668 and Sendai virus were added as indicated. T cell proliferation was analyzed after 58 hr by FACS (gated on live CD8+ CD45.1+ CFSElo cells). Data and Software Access Notes Raw data and supplemental tables were uploaded to the Tranche database (http://www.proteomecommons.org) and can be downloaded freely from the website. Source code for label-free algorithms incorporated into MaxQuant are available at the MaxQuant.org website upon publication. SUPPLEMENTAL INFORMATION The Supplemental Information include five figures and three tables and can be found with this article online at doi:10.1016/j.immuni.2010.01.013. ACKNOWLEDGMENTS We thank K. Shortman (WEHI, Melbourne, Australia) and M. Suter (University of Zurich) for providing critical reagents; J. Hamann (University of Amsterdam) for CD97 antibody; A. Egorov (AVIR Green Hills Biotechnology) for delNS1 influenza virus; B. Bathke, M. Dodel, and J. Paetzold for excellent technical assistance; the animal services facilities at the Max Planck Institute of Biochemistry and Bavarian Nordic for animal husbandry; C. Kumar for gene ontology (GO) support; and M. Schmidt-Supprian, M. Sixt, and M. Vermeulen for critical reading of the manuscript. The Max-Planck Society and the DC Thera 6th framework project of the European Union provided funding. C.A.L. performed most experiments. C.A.L, M.O., H.H, M.S., and M.M. conceived and designed the study and analyzed data. H.L., B.F., M.O., and H.H. per-

formed some experiments. J.C. provided early access to label-free software. J.T. and S.A. provided mice. M.W provided Sendai-GFP virus. C.A.L., M.O., H.H., and M.M. wrote the paper. B.F., H.L., H.H., and M.O. were employees of Bavarian Nordic GmbH. M.O. and M.M. share senior authorship. Received: March 26, 2009 Revised: December 18, 2009 Accepted: January 26, 2010 Published online: February 18, 2010 REFERENCES Ablasser, A., Bauernfeind, F., Hartmann, G., Latz, E., Fitzgerald, K.A., and Hornung, V. (2009). RIG-I-dependent sensing of poly(dA:dT) through the induction of an RNA polymerase III-transcribed RNA intermediate. Nat. Immunol. 10, 1065–1072. Adachi, O., Kawai, T., Takeda, K., Matsumoto, M., Tsutsui, H., Sakagami, M., Nakanishi, K., and Akira, S. (1998). Targeted disruption of the MyD88 gene results in loss of IL-1- and IL-18-mediated function. Immunity 9, 143–150. Aebersold, R., and Mann, M. (2003). Mass spectrometry-based proteomics. Nature 422, 198–207. Akira, S., Uematsu, S., and Takeuchi, O. (2006). Pathogen recognition and innate immunity. Cell 124, 783–801. Alexopoulou, L., Holt, A.C., Medzhitov, R., and Flavell, R.A. (2001). Recognition of double-stranded RNA and activation of NF-kappaB by Toll-like receptor 3. Nature 413, 732–738. Allan, R.S., Smith, C.M., Belz, G.T., van Lint, A.L., Wakim, L.M., Heath, W.R., and Carbone, F.R. (2003). Epidermal viral immunity induced by CD8alpha+ dendritic cells but not by Langerhans cells. Science 301, 1925–1928. Bantscheff, M., Schirle, M., Sweetman, G., Rick, J., and Kuster, B. (2007). Quantitative mass spectrometry in proteomics: a critical review. Anal. Bioanal. Chem. 389, 1017–1031. Bedoui, S., Prato, S., Mintern, J., Gebhardt, T., Zhan, Y., Lew, A.M., Heath, W.R., Villadangos, J.A., and Segura, E. (2009). Characterization of an immediate splenic precursor of CD8+ dendritic cells capable of inducing antiviral T cell responses. J. Immunol. 182, 4200–4207. Belz, G.T., Smith, C.M., Eichner, D., Shortman, K., Karupiah, G., Carbone, F.R., and Heath, W.R. (2004). Cutting edge: conventional CD8 alpha+ dendritic cells are generally involved in priming CTL immunity to viruses. J. Immunol. 172, 1996–2000. Belz, G.T., Shortman, K., Bevan, M.J., and Heath, W.R. (2005). CD8alpha+ dendritic cells selectively present MHC class I-restricted noncytolytic viral and intracellular bacterial antigens in vivo. J. Immunol. 175, 196–200. Beutler, B., Jiang, Z., Georgel, P., Crozat, K., Croker, B., Rutschmann, S., Du, X., and Hoebe, K. (2006). Genetic analysis of host resistance: Toll-like receptor signaling and immunity at large. Annu. Rev. Immunol. 24, 353–389. Bonaldi, T., Straub, T., Cox, J., Kumar, C., Becker, P.B., and Mann, M. (2008). Combined use of RNAi and quantitative proteomics to study gene function in Drosophila. Mol. Cell 31, 762–772. Caminschi, I., Proietto, A.I., Ahmet, F., Kitsoulis, S., Shin Teh, J., Lo, J.C., Rizzitelli, A., Wu, L., Vremec, D., van Dommelen, S.L., et al. (2008). The dendritic cell subtype-restricted C-type lectin Clec9A is a target for vaccine enhancement. Blood 112, 3264–3273. Choi, M.K., Wang, Z., Ban, T., Yanai, H., Lu, Y., Koshiba, R., Nakaima, Y., Hangai, S., Savitsky, D., Nakasato, M., et al. (2009). A selective contribution of the RIG-I-like receptor pathway to type I interferon responses activated by cytosolic DNA. Proc. Natl. Acad. Sci. USA 106, 17927–17932. Cox, J., and Mann, M. (2008). MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat. Biotechnol. 26, 1367–1372. de Godoy, L.M., Olsen, J.V., Cox, J., Nielsen, M.L., Hubner, N.C., Fro¨hlich, F., Walther, T.C., and Mann, M. (2008). Comprehensive mass-spectrometrybased proteome quantification of haploid versus diploid yeast. Nature 455, 1251–1254.

Immunity 32, 279–289, February 26, 2010 ª2010 Elsevier Inc. 287

Immunity Label-free Quantitative Proteomics of DC Subsets

Dudziak, D., Kamphorst, A.O., Heidkamp, G.F., Buchholz, V.R., Trumpfheller, C., Yamazaki, S., Cheong, C., Liu, K., Lee, H.W., Park, C.G., et al. (2007). Differential antigen processing by dendritic cell subsets in vivo. Science 315, 107–111. Edwards, A.D., Chaussabel, D., Tomlinson, S., Schulz, O., Sher, A., and Reis e Sousa, C. (2003a). Relationships among murine CD11c(high) dendritic cell subsets as revealed by baseline gene expression patterns. J. Immunol. 171, 47–60. Edwards, A.D., Diebold, S.S., Slack, E.M., Tomizawa, H., Hemmi, H., Kaisho, T., Akira, S., and Reis e Sousa, C. (2003b). Toll-like receptor expression in murine DC subsets: lack of TLR7 expression by CD8 alpha+ DC correlates with unresponsiveness to imidazoquinolines. Eur. J. Immunol. 33, 827–833. Fuchsberger, M., Hochrein, H., and O’Keeffe, M. (2005). Activation of plasmacytoid dendritic cells. Immunol. Cell Biol. 83, 571–577. Galibert, L., Diemer, G.S., Liu, Z., Johnson, R.S., Smith, J.L., Walzer, T., Comeau, M.R., Rauch, C.T., Wolfson, M.F., Sorensen, R.A., et al. (2005). Nectin-like protein 2 defines a subset of T-cell zone dendritic cells and is a ligand for class-I-restricted T-cell-associated molecule. J. Biol. Chem. 280, 21955–21964. Hara, H., Ishihara, C., Takeuchi, A., Xue, L., Morris, S.W., Penninger, J.M., Yoshida, H., and Saito, T. (2008). Cell type-specific regulation of ITAM-mediated NF-kappaB activation by the adaptors, CARMA1 and CARD9. J. Immunol. 181, 918–930. Heath, W.R., Belz, G.T., Behrens, G.M., Smith, C.M., Forehan, S.P., Parish, I.A., Davey, G.M., Wilson, N.S., Carbone, F.R., and Villadangos, J.A. (2004). Cross-presentation, dendritic cell subsets, and the generation of immunity to cellular antigens. Immunol. Rev. 199, 9–26. Heil, F., Hemmi, H., Hochrein, H., Ampenberger, F., Kirschning, C., Akira, S., Lipford, G., Wagner, H., and Bauer, S. (2004). Species-specific recognition of single-stranded RNA via toll-like receptor 7 and 8. Science 303, 1526–1529. Hemmi, H., Takeuchi, O., Kawai, T., Kaisho, T., Sato, S., Sanjo, H., Matsumoto, M., Hoshino, K., Wagner, H., Takeda, K., and Akira, S. (2000). A Toll-like receptor recognizes bacterial DNA. Nature 408, 740–745. Hochrein, H., Shortman, K., Vremec, D., Scott, B., Hertzog, P., and O’Keeffe, M. (2001). Differential production of IL-12, IFN-alpha, and IFN-gamma by mouse dendritic cell subsets. J. Immunol. 166, 5448–5455. Hochrein, H., Schlatter, B., O’Keeffe, M., Wagner, C., Schmitz, F., Schiemann, M., Bauer, S., Suter, M., and Wagner, H. (2004). Herpes simplex virus type-1 induces IFN-alpha production via Toll-like receptor 9-dependent and -independent pathways. Proc. Natl. Acad. Sci. USA 101, 11416–11421. Huysamen, C., Willment, J.A., Dennehy, K.M., and Brown, G.D. (2008). CLEC9A is a novel activation C-type lectin-like receptor expressed on BDCA3+ dendritic cells and a subset of monocytes. J. Biol. Chem. 283, 16693–16701. Imaizumi, T., Sashinami, H., Mori, F., Matsumiya, T., Yoshida, H., Nakane, A., Wakabayashi, K., Oyama, C., and Satoh, K. (2006). Listeria monocytogenes induces the expression of retinoic acid-inducible gene-I. Microbiol. Immunol. 50, 811–815. Ishikawa, H., Ma, Z., and Barber, G.N. (2009). STING regulates intracellular DNA-mediated, type I interferon-dependent innate immunity. Nature 461, 788–792. Janeway, C.A., Jr., and Medzhitov, R. (2002). Innate immune recognition. Annu. Rev. Immunol. 20, 197–216. Kanneganti, T.D., Lamkanfi, M., and Nu´n˜ez, G. (2007). Intracellular NOD-like receptors in host defense and disease. Immunity 27, 549–559. Kato, H., Sato, S., Yoneyama, M., Yamamoto, M., Uematsu, S., Matsui, K., Tsujimura, T., Takeda, K., Fujita, T., Takeuchi, O., and Akira, S. (2005). Cell type-specific involvement of RIG-I in antiviral response. Immunity 23, 19–28. Kato, H., Takeuchi, O., Sato, S., Yoneyama, M., Yamamoto, M., Matsui, K., Uematsu, S., Jung, A., Kawai, T., Ishii, K.J., et al. (2006). Differential roles of MDA5 and RIG-I helicases in the recognition of RNA viruses. Nature 441, 101–105. Kato, H., Takeuchi, O., Mikamo-Satoh, E., Hirai, R., Kawai, T., Matsushita, K., Hiiragi, A., Dermody, T.S., Fujita, T., and Akira, S. (2008). Length-dependent

288 Immunity 32, 279–289, February 26, 2010 ª2010 Elsevier Inc.

recognition of double-stranded ribonucleic acids by retinoic acid-inducible gene-I and melanoma differentiation-associated gene 5. J. Exp. Med. 205, 1601–1610. Kawai, T., and Akira, S. (2007). SnapShot: Pattern-recognition receptors. Cell 129, 1024. Kawai, T., and Akira, S. (2008). Toll-like receptor and RIG-I-like receptor signaling. Ann. N Y Acad. Sci. 1143, 1–20. Kenworthy, R., Lambert, D., Yang, F., Wang, N., Chen, Z., Zhu, H., Zhu, F., Liu, C., Li, K., and Tang, H. (2009). Short-hairpin RNAs delivered by lentiviral vector transduction trigger RIG-I-mediated IFN activation. Nucleic Acids Res. 37, 6587–6599. Komuro, A., Bamming, D., and Horvath, C.M. (2008). Negative regulation of cytoplasmic RNA-mediated antiviral signaling. Cytokine 43, 350–358. Kru¨ger, M., Moser, M., Ussar, S., Thievessen, I., Luber, C.A., Forner, F., Schmidt, S., Zanivan, S., Fa¨ssler, R., and Mann, M. (2008). SILAC mouse for quantitative proteomics uncovers kindlin-3 as an essential factor for red blood cell function. Cell 134, 353–364. Kumagai, Y., Kumar, H., Koyama, S., Kawai, T., Takeuchi, O., and Akira, S. (2009). Cutting Edge: TLR-Dependent viral recognition along with type I IFN positive feedback signaling masks the requirement of viral replication for IFN-alpha production in plasmacytoid dendritic cells. J. Immunol. 182, 3960–3964. Maldonado-Lo´pez, R., De Smedt, T., Michel, P., Godfroid, J., Pajak, B., Heirman, C., Thielemans, K., Leo, O., Urbain, J., and Moser, M. (1999). CD8alpha+ and CD8alpha- subclasses of dendritic cells direct the development of distinct T helper cells in vivo. J. Exp. Med. 189, 587–592. Meylan, E., Curran, J., Hofmann, K., Moradpour, D., Binder, M., Bartenschlager, R., and Tschopp, J. (2005). Cardif is an adaptor protein in the RIG-I antiviral pathway and is targeted by hepatitis C virus. Nature 437, 1167–1172. Michallet, M.C., Meylan, E., Ermolaeva, M.A., Vazquez, J., Rebsamen, M., Curran, J., Poeck, H., Bscheider, M., Hartmann, G., Ko¨nig, M., et al. (2008). TRADD protein is an essential component of the RIG-like helicase antiviral pathway. Immunity 28, 651–661. Moore, C.B., Bergstralh, D.T., Duncan, J.A., Lei, Y., Morrison, T.E., Zimmermann, A.G., Accavitti-Loper, M.A., Madden, V.J., Sun, L., Ye, Z., et al. (2008). NLRX1 is a regulator of mitochondrial antiviral immunity. Nature 451, 573–577. Naik, S.H., Proietto, A.I., Wilson, N.S., Dakic, A., Schnorrer, P., Fuchsberger, M., Lahoud, M.H., O’Keeffe, M., Shao, Q.X., Chen, W.F., et al. (2005). 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. O’Keeffe, M., Hochrein, H., Vremec, D., Pooley, J., Evans, R., Woulfe, S., and Shortman, K. (2002). Effects of administration of progenipoietin 1, Flt-3 ligand, granulocyte colony-stimulating factor, and pegylated granulocyte-macrophage colony-stimulating factor on dendritic cell subsets in mice. Blood 99, 2122–2130. O’Keeffe, M., Grumont, R.J., Hochrein, H., Fuchsberger, M., Gugasyan, R., Vremec, D., Shortman, K., and Gerondakis, S. (2005). Distinct roles for the NF-kappaB1 and c-Rel transcription factors in the differentiation and survival of plasmacytoid and conventional dendritic cells activated by TLR-9 signals. Blood 106, 3457–3464. Ong, S.E., and Mann, M. (2005). Mass spectrometry-based proteomics turns quantitative. Nat. Chem. Biol. 1, 252–262. Ong, S.E., Blagoev, B., Kratchmarova, I., Kristensen, D.B., Steen, H., Pandey, A., and Mann, M. (2002). Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol. Cell. Proteomics 1, 376–386. Panchaud, A., Affolter, M., Moreillon, P., and Kussmann, M. (2008). Experimental and computational approaches to quantitative proteomics: status quo and outlook. J. Proteomics 71, 19–33. Rappsilber, J., Ishihama, Y., and Mann, M. (2003). Stop and go extraction tips for matrix-assisted laser desorption/ionization, nanoelectrospray, and LC/MS sample pretreatment in proteomics. Anal. Chem. 75, 663–670.

Immunity Label-free Quantitative Proteomics of DC Subsets

Rasmussen, S.B., Jensen, S.B., Nielsen, C., Quartin, E., Kato, H., Chen, Z.J., Silverman, R.H., Akira, S., and Paludan, S.R. (2009). Herpes simplex virus infection is sensed by both Toll-like receptors and retinoic acid-inducible gene- like receptors, which synergize to induce type I interferon production. J. Gen. Virol. 90, 74–78. Robinson, M.J., Osorio, F., Rosas, M., Freitas, R.P., Schweighoffer, E., Gross, O., Verbeek, J.S., Ruland, J., Tybulewicz, V., Brown, G.D., et al. (2009). Dectin-2 is a Syk-coupled pattern recognition receptor crucial for Th17 responses to fungal infection. J. Exp. Med. 206, 2037–2051. Saitoh, S., and Miyake, K. (2009). Regulatory molecules required for nucleotide-sensing Toll-like receptors. Immunol. Rev. 227, 32–43. Sancho, D., Joffre, O.P., Keller, A.M., Rogers, N.C., Martı´nez, D., HernanzFalco´n, P., Rosewell, I., and Reis e Sousa, C. (2009). Identification of a dendritic cell receptor that couples sensing of necrosis to immunity. Nature 458, 899– 903. Schiess, R., Mueller, L.N., Schmidt, A., Mueller, M., Wollscheid, B., and Aebersold, R. (2009). Analysis of cell surface proteome changes via labelfree, quantitative mass spectrometry. Mol. Cell Proteomics. 8, 624–638. Published online November 25, 2008. 10.1074/mcp.M800172-MCP200. Schnorrer, P., Behrens, G.M., Wilson, N.S., Pooley, J.L., Smith, C.M., El-Sukkari, D., Davey, G., Kupresanin, F., Li, M., Maraskovsky, E., et al. (2006). The dominant role of CD8+ dendritic cells in cross-presentation is not dictated by antigen capture. Proc. Natl. Acad. Sci. USA 103, 10729–10734. Shevchenko, A., Wilm, M., Vorm, O., and Mann, M. (1996). Mass spectrometric sequencing of proteins silver-stained polyacrylamide gels. Anal. Chem. 68, 850–858. Shortman, K., and Naik, S.H. (2007). Steady-state and inflammatory dendriticcell development. Nat. Rev. Immunol. 7, 19–30. Silverman, R.H. (2007). Viral encounters with 20 ,50 -oligoadenylate synthetase and RNase L during the interferon antiviral response. J. Virol. 81, 12720– 12729.

Takaoka, A., Wang, Z., Choi, M.K., Yanai, H., Negishi, H., Ban, T., Lu, Y., Miyagishi, M., Kodama, T., Honda, K., et al. (2007). DAI (DLM-1/ZBP1) is a cytosolic DNA sensor and an activator of innate immune response. Nature 448, 501–505. Tamura, T., Tailor, P., Yamaoka, K., Kong, H.J., Tsujimura, H., O’Shea, J.J., Singh, H., and Ozato, K. (2005). IFN regulatory factor-4 and -8 govern dendritic cell subset development and their functional diversity. J. Immunol. 174, 2573– 2581. Villadangos, J.A., and Schnorrer, P. (2007). Intrinsic and cooperative antigenpresenting functions of dendritic-cell subsets in vivo. Nat. Rev. Immunol. 7, 543–555. Villadangos, J.A., and Young, L. (2008). Antigen-presentation properties of plasmacytoid dendritic cells. Immunity 29, 352–361. Vremec, D., Pooley, J., Hochrein, H., Wu, L., and Shortman, K. (2000). CD4 and CD8 expression by dendritic cell subtypes in mouse thymus and spleen. J. Immunol. 164, 2978–2986. Vremec, D., O’Keeffe, M., Hochrein, H., Fuchsberger, M., Caminschi, I., Lahoud, M., and Shortman, K. (2007). Production of interferons by dendritic cells, plasmacytoid cells, natural killer cells, and interferon-producing killer dendritic cells. Blood 109, 1165–1173. Wilson, N.S., Behrens, G.M., Lundie, R.J., Smith, C.M., Waithman, J., Young, L., Forehan, S.P., Mount, A., Steptoe, R.J., Shortman, K.D., et al. (2006). Systemic activation of dendritic cells by Toll-like receptor ligands or malaria infection impairs cross-presentation and antiviral immunity. Nat. Immunol. 7, 165–172. Wu, L., D’Amico, A., Winkel, K.D., Suter, M., Lo, D., and Shortman, K. (1998). RelB is essential for the development of myeloid-related CD8alpha- dendritic cells but not of lymphoid-related CD8alpha+ dendritic cells. Immunity 9, 839– 847.

Stetson, D.B., and Medzhitov, R. (2006). Type I interferons in host defense. Immunity 25, 373–381.

Xu, D., Suenaga, N., Edelmann, M.J., Fridman, R., Muschel, R.J., and Kessler, B.M. (2008). Novel MMP-9 substrates in cancer cells revealed by a label-free quantitative proteomics approach. Mol. Cell. Proteomics 7, 2215–2228.

Tabeta, K., Georgel, P., Janssen, E., Du, X., Hoebe, K., Crozat, K., Mudd, S., Shamel, L., Sovath, S., Goode, J., et al. (2004). Toll-like receptors 9 and 3 as essential components of innate immune defense against mouse cytomegalovirus infection. Proc. Natl. Acad. Sci. USA 101, 3516–3521.

Yamamoto-Furusho, J.K., Barnich, N., Xavier, R., Hisamatsu, T., and Podolsky, D.K. (2006). Centaurin beta1 down-regulates nucleotide-binding oligomerization domains 1- and 2-dependent NF-kappaB activation. J. Biol. Chem. 281, 36060–36070.

Tailor, P., Tamura, T., and Ozato, K. (2006). IRF family proteins and type I interferon induction in dendritic cells. Cell Res. 16, 134–140.

Yoneyama, M., and Fujita, T. (2009). RNA recognition and signal transduction by RIG-I-like receptors. Immunol. Rev. 227, 54–65.

Immunity 32, 279–289, February 26, 2010 ª2010 Elsevier Inc. 289