Divergent TLR7 and TLR9 signaling and type I interferon production ...

© 2008 Nature Publishing Group http://www.nature.com/naturemedicine

ARTICLES

Divergent TLR7 and TLR9 signaling and type I interferon production distinguish pathogenic and nonpathogenic AIDS virus infections Judith N Mandl1,2,6, Ashley P Barry2,6, Thomas H Vanderford2, Natalia Kozyr2, Rahul Chavan2, Sara Klucking2, Franck J Barrat3, Robert L Coffman3, Silvija I Staprans2,4,5 & Mark B Feinberg2,4,5 Pathogenic HIV infections of humans and simian immunodeficiency virus (SIV) infections of rhesus macaques are characterized by generalized immune activation and progressive CD4+ T cell depletion. In contrast, natural reservoir hosts for SIV, such as sooty mangabeys, do not progress to AIDS and show a lack of aberrant immune activation and preserved CD4+ T cell populations, despite high levels of SIV replication. Here we show that sooty mangabeys have substantially reduced levels of innate immune system activation in vivo during acute and chronic SIV infection and that sooty mangabey plasmacytoid dendritic cells (pDCs) produce markedly less interferon-a in response to SIV and other Toll-like receptor 7 and 9 ligands ex vivo. We propose that chronic stimulation of pDCs by SIV and HIV in non-natural hosts may drive the unrelenting immune system activation and dysfunction underlying AIDS progression. Such a vicious cycle of continuous virus replication and immunopathology is absent in natural sooty mangabey hosts.

A hallmark of HIV infection is chronic activation of the immune system in association with dysfunction of cellular and humoral immune responses and failure to effectively control virus replication. After HIV infection, turnover rates of CD4+ and CD8+ T cells are elevated, high levels of activation-induced cell death are seen in both T cell subsets independent of their infection by HIV1–3, B cells are polyclonally activated with consequent hypergammaglobulinemia4, natural killer (NK) cell activation and turnover is increased5 and dendritic cell (DC) numbers are diminished in the peripheral blood6. This chronic inflammatory environment compromises CD4+ T cell regenerative capacity as a result of suppression of bone marrow, reduction of thymic function and impairment of the structural and functional integrity of peripheral lymphoid tissues7. Indeed, it is now recognized that chronic, generalized immune activation is a major driving force for CD4+ T cell depletion and AIDS progression7,8. We have studied one of the natural African primate reservoir hosts for SIV, the sooty mangabey. Sooty mangabeys are infected with SIVsm strains documented to be the origin of HIV-2 in humans, as well as the source of SIVs used in experimental AIDS pathogenesis and vaccine studies in rhesus macaques9. Rather surprisingly, key features proposed to have central roles in driving AIDS progression in HIV-1–infected humans and SIV-infected rhesus macaques, including chronic high levels of viremia, preferential tropism of SIV and HIV for CD4+ T cells, short half-lives of virus-infected cells and severe depletion of mucosal CD4+ T cells, have also been found to characterize nonpathogenic SIV

infections of sooty mangabeys10–13. However, SIV-infected sooty mangabeys show far lower levels of chronic immune activation than SIV-infected rhesus macaques or HIV-infected humans13. To explore how sooty mangabeys avoid aberrant immune activation, we developed a comparative experimental infection model in which sooty mangabeys and non-natural rhesus macaque hosts are inoculated with SIVsm obtained directly from a naturally infected sooty mangabey14. Only SIVsm-infected rhesus macaques develop progressive CD4+ T cell depletion and AIDS, indicating that it is the host response to infection, rather than properties inherent to the virus itself, that causes immunodeficiency in disease-susceptible primate hosts14. Notably, divergent host responses to SIV are manifested early after infection, with significantly attenuated adaptive cellular immune responses seen in sooty mangabeys compared with rhesus macaques14. Here we identify host-specific differences in innate immune responses that influence the induction of adaptive antiviral immune responses and that may represent primary determinants of whether or not immune activation and immunodeficiency disease follow AIDS virus infection. RESULTS Limited NK cell expansion in SIV-infected sooty mangabeys We sought to identify the earliest evidence of a divergent host response to SIV infection in sooty mangabeys compared to rhesus macaques by studying innate immune cell populations in vivo during acute SIVsm infection. Uncloned SIVsm replicated well in the three animals

1Graduate Program in Population Biology, Ecology and Evolution, Emory University, 1510 Clifton Road, Atlanta, Georgia 30322, USA. 2Emory Vaccine Center and Yerkes National Primate Research Center, 954 Gatewood Road, Atlanta, Georgia 30329, USA. 3Dynavax Technologies, 2929 Seventh Street, Berkeley, California 94710, USA. 4Department of Microbiology and Immunology and Department of Medicine, Emory University School of Medicine, 954 Gatewood Road, Atlanta, Georgia 30329, USA. 5Current address: Merck Vaccines and Infectious Diseases, Merck & Co., Inc., WP97-A337, 770 Sumneytown Pike, PO Box 4, West Point, Pennsylvania 19486, USA. 6These authors contributed equally to this work. Correspondence should be addressed to M.B.F. ([email protected]).

Received 2 April; accepted 21 August; published online 14 September 2008; doi:10.1038/nm.1871

NATURE MEDICINE

VOLUME 14

[

NUMBER 10

[

OCTOBER 2008

1077

infected from both species, with comparable virus replication (Fig. 1a). As we observed in an earlier study14, there were notable differences in the extent of CD4+ and CD8+ T cell proliferation, as measured by expression of the Ki67 proliferation marker after SIV inoculation, with rhesus macaques showing considerably greater and sustained proliferation, while sooty mangabeys showed limited or transient T cell proliferation (Supplementary Fig. 1 online). To explore whether the differences in adaptive immune response were preceded by differences in the innate immune response, we followed the proliferative response of NK cells (Fig. 1b,c). Similar to the limited changes in T cell proliferation that we observed in sooty mangabeys, the percentage of Ki67+ NK cells in sooty mangabeys did not increase substantially after infection (Fig. 1b,c). This was in

RM1 RM2 RM3

SM1 SM2 SM3

c

107

105

3

10

contrast to the large NK cell expansion observed in rhesus macaques (Fig. 1b,c). Thus, the divergent sooty mangabey response to SIVsm is manifested very early after infection, at the level of the innate immune response. Lack of sooty mangabey pDC activation in acute SIV infection As DCs are key initiators of immune responses, we next explored the in vivo response of sooty mangabey and rhesus macaque DCs to SIV infection. Two subsets of blood DCs have been identified in humans and rhesus macaques15,16: CD11c+ myeloid DCs (mDCs) and CD123+ pDCs. Of particular importance in eliciting host responses to viral infections, pDCs produce large amounts of type I interferons (IFNs), resulting in the maturation of mDCs into efficient antigen-presenting

25

e

RM SM

20 15 10 5

101

40 30 20 10

0 20 30 40 Time after infection (d)

50

d

b SSC

SM 2.09

20 30 40 Time after infection (d)

2.3

50

60

Plasmacytoid DC

FSC

1.4 µm

35.8

Myeloid DC

NK cells

Day –14

g

Day 10

Percentage lymph node pDCs

CD123

1.4

0.5

9.4

6.3

0.7

8.7

PB

1.2

1.7 CD11c

10 15 20 Time after infection (d)

25

30

Day –14 Day 10

30

Day 14

LN

11

30

10

28

12

1.2

1.6

25

1.4 µm

RM Day 10

8.6

20

20

CD123 SM Day –14

5

15

0

CD11c

f

0

10

30

Day 10 CD16

5

40

Lineage

CD16

0 50

HLA-DR

Ki67

10

RM 1.39

Day 0 FSC

0 0

60

+ Percentage CCR7 mDCs

10

SSC

0

SM1 SM2 SM3

RM1 RM2 RM3

50

Percentage CCR7 + pDCs

109

Percentage Ki67 + NK cells

SIV RNA copies per milliliter

a

CD8


ARTICLES

28

2.3

20

10

0

ND

ND

RM1

RM2

ND RM3

SM1

ND SM2

ND SM3

Figure 1 Muted NK cell activation and DC maturation and homing during acute SIVsm infection of sooty mangabeys compared to rhesus macaques. (a) Plasma SIV viral load in three rhesus macaques (RMs) and three sooty mangabeys (SMs) infected intravenously with the same uncloned SIVsm. (b) NK cells in RMs and SMs were identified by first gating on lymphocytes and then gating on CD16+CD8+ cells. Percentage Ki67+ NK cells on day 0 and day 14 after SIV infection are shown in a representative RM and SM. SSC, side scatter; FSC, forward scatter. (c) Percentage proliferating (Ki67+) NK cells in RMs and SMs. Means ± s.e.m. are shown. (d) DCs in RMs and SMs were identified by first excluding granulocytes (based on FSC and SSC), then gating on HLADR+ lineage (CD14, CD3, CD20)-negative cells. pDCs express CD123 and mDCs express CD11c. Representative electron micrographs of sorted blood pDCs and mDCs from SMs are shown. (e) CCR7 expression on pDCs (top) and mDCs (bottom) in peripheral blood of SMs and RMs after inoculation with SIV. (f) DC populations in lymph node (LN) and peripheral blood (PB) samples taken on day 0 and day 10 after SIV infection in representative RMs and SMs, shown gated on HLA-DR+ lineage-negative cells. (g) Percentage lymph node pDCs at day –14, day 10 and day 14 after SIV infection in SMs (blue) and RMs (red). ND, not done. Numbers on FACS plots indicate percentage of cells in the gates shown.

1078

VOLUME 14

[

NUMBER 10

[

OCTOBER 2008

NATURE MEDICINE

ARTICLES

c

b

4,000

PBMC pDC-depleted

2,000

–1 IFN-α (pg ml )

IFN-α (pg ml–1)

IFN-α (pg ml–1)

3,000

2,000

*

1,000

** ** 0

es ic N le s o in tim hi bi to D r O V05 D N 6 cn trl

**

ov

ic N le s o in tim hi bi D tor V0 IR 56 S9 IR 54 S IR 66 O S8 1 D 6 N 9 cn trl

ov

iSIV stimulation

M ic r

iHIV stimulation

** ** 20 5 Chloroquine (µM)

es

es i N cle o in stim hi bi t IR or S9 IR 54 S IR 661 O S8 D 69 N cn trl

ov

**

0

**

M ic r

** * *

0

M ic r


1,000

*

1,000

100

75

* 50

25

***

0 iS IV

2,000 3,000

4,000

Percentage of pre–pDC depletion IFN-α production

5,000

**

***

iH IV In (li flu ve e ai nza ch i) C pG A

a

iSIV stim

Figure 2 SIV and HIV stimulate IFN-a production via TLR7 and TLR9 on pDCs. (a,b) IFN-a production by HIV– human PBMCs (a) or SIV– RM PBMCs (b) after stimulation with microvesicle controls, iSIV or iHIV. The inhibitory effects of specific oligodeoxynucleotide inhibitors of TLR7 (5.6 mM IRS661), TLR9 (5.6 mM IRS869), both TLR7 and TLR9 (5.6 mM IRS954 or 2 mM DV056), a nonspecific oligodeoxynucleotide control (ODN cntrl) or the endosome acidification inhibitor chloroquine (20 mM and 5 mM) are shown. Histograms represent means ± s.e.m. of seven humans (iSIV) or three humans (iHIV) in a and of three RMs in b. (c) IFN-a production by HIV– human PBMCs stimulated with iSIV, iHIV, live influenza (aichi) or CpG A2336 (CpG A) before and after having been depleted of pDCs. *P o 0.05, **P o 0.01 and ***P o 0.001.

cells15,17 and facilitating the induction of antiviral CD8+ T cell, CD4+ T cell, NK cell and B cell responses18–20. In both sooty mangabeys and rhesus macaques, DCs could be divided into pDCs and mDCs with antibodies for CD123 and CD11c, respectively (Fig. 1d). As neither pDC nor mDC populations have previously been described in sooty mangabeys, we expanded DCs in vivo in SIV-uninfected sooty mangabeys by treating them with fms-like tyrosine kinase ligand (Flt3L), sorted the DCs for both subsets and then examined their morphology by electron microscopy (Fig. 1d). Sooty mangabey pDCs and mDCs showed the characteristic features of these cells described in humans and rhesus macaques15,16,20. The migration of activated and maturing DCs to lymph nodes is pivotal to the generation of immune responses15. Exposure of human DCs to HIV-1 results in upregulation of maturation markers on pDCs, as well as expression of CCR7 that enables pDCs to migrate to lymph nodes in response to its ligands CCL19 and CCL2115,17. Therefore, we investigated the in vivo maturation and trafficking of DCs in rhesus macaques versus sooty mangabeys after infection with SIVsm. Rhesus macaque pDCs upregulated CCR7 expression after SIVsm infection, and, in some rhesus macaques, CCR7 expression was also increased in mDCs (Fig. 1e). However, there was little change in pDC or mDC CCR7 expression in sooty mangabeys (Fig. 1e). Consistent with the known role of CCR7 in immune cell homing to lymphoid tissue, we observed an association between CCR7 expression on blood pDCs and the accumulation of pDCs in lymph nodes (Fig. 1e,g). In rhesus macaques, an increase in the frequency of lymph node pDCs was seen on days 10 and 14 after infection, whereas such pDC migration was lacking in sooty mangabeys (Fig. 1f,g). SIV stimulates IFN-a production by pDCs via TLR7 and TLR9 Plasmacytoid DCs are known to express Toll-like receptor 7 (TLR7) and TLR9 and to secrete IFN-a upon TLR7 and TLR9 signaling21. TLR7 recognizes single-stranded RNA, and TLR9 recognizes unmethylated CpG-oligodeoxynucleotide–containing DNA21. TLR7 and TLR9 have also been implicated in the recognition of HIV by pDCs and in their consequent IFN-a production after HIV stimulation22,23. However, TLR recognition of SIV has not been investigated in humans, rhesus macaques or sooty mangabeys. We observed that both

NATURE MEDICINE

VOLUME 14

[

NUMBER 10

[

OCTOBER 2008

aldrithiol-2–inactivated HIV (iHIV) and SIV (iSIV) elicit the production of large amounts of IFN-a from human peripheral blood mononuclear cells (PBMCs; Fig. 2a). We next made use of defined nonstimulatory DNA sequences, termed immunoregulatory sequences (IRS), able to inhibit cytokine production by human PBMCs after stimulation with TLR7 or TLR9 ligands24, to assess the impact of inhibition of TLR7, TLR9 or both on the production of type I interferon upon PBMCs with iHIV or iSIV stimulation of PBMCs. Adding antagonists of TLR7 (IRS661), TLR9 (IRS869) or both TLR7 and TLR9 (IRS954) to iHIV-stimulated human PBMCs significantly reduced IFN-a production (Fig. 2a). This is consistent with previous results22, and, similar to those results, we were also unable to rule out a role for TLR9 in the recognition of HIV-1 (Fig. 2a). Likewise, iSIV-induced IFN-a production by human PBMCs was significantly inhibited by the addition of IRS661, IRS869 or IRS954 (Fig. 2a). Furthermore, chloroquine, an endosome acidification inhibitor that blocks both TLR7 and TLR9 signaling, inhibited SIV-stimulated IFN-a secretion to below detection (Fig. 2a). Of note, the greater inhibition of SIV-stimulated type I interferon production with the dual TLR7 and TLR9 inhibitor compared to each single TLR antagonist alone lends support to a possible role for both TLR7 and TLR9 in the recognition of SIV. The development of a potent dual TLR7 and TLR9 antagonist, DV056, a 25-base single-stranded phosphorothioate oligodeoxynucleotide that retains the key inhibitory motifs previously described24 but has been modified for optimum activity in both rhesus macaques and humans (data not shown), allowed us to ask whether the same TLR7 and TLR9 pathway responsible for recognition of SIV and HIV in humans is responsible for recognition of SIV in rhesus macaques. As with IRS954, stimulation of human PBMCs with iSIV in the presence of DV056 markedly inhibited IFN-a production (Fig. 2a). The addition of DV056 to rhesus macaque PBMCs stimulated with iSIV resulted in 490% inhibition of IFN-a production (Fig. 2b). Thus, SIV, like HIV, requires TLR7, TLR9 or both to induce a type I interferon response in both rhesus macaques and humans. To investigate whether pDCs were producing the IFN-a induced by the recognition of HIV and SIV via TLR7 and TLR9 signaling pathways, we depleted pDCs from human PBMCs and measured IFN-a

1079

ARTICLES P < 0.001

a

NS

7,000

b

P < 0.001

c

6

6,000 5,000 4,000 3,000

1,400 1,200 1,000 800 600

2,000

2 1.0

0.5

400

1,000

200

0

0.0

0

d

RM iSIV

SM

HU

P < 0.001 NS

10,000

Microvesicles

RM SM Microvesicles

Influenza (live aichi)

P < 0.001

P < 0.001

NS

1,000

7,500

500

NS

3,500



2,500

250

e

RM SM Influenza PR8 (heat-inact.) SM1

HU

RM R-848

P < 0.001

3,500

3,000

3,000

2,500

2,500

2,000 1,500 1,000

2,000 1,500 1,000 500

0

0 HU

iSIV

P < 0.001 NS

P < 0.001

500 0


P < 0.001

P < 0.001

750

5,000

Microvesicles

iSIV


HU

IFN-α (pg ml –1)

SM SM mean

4

IFN-α (pg per pDC)

IFN-α (pg ml –1)

IFN-α (pg ml –1)

2,000

0 HU

SM

RM HSV (UV-inact.)

HU

SM

RM CpG C

SM

P < 0.0001

SM2

RM1

Human

RM2

70

f

P = 0.02

Microvesicles

0.066

iSIV

9.73

0.51

0

43.5

6.09

0.23

39.7

Microvesicles

iSIV

0.07

31.1

Percentage IFN-α + cells

60 50

P = 0.004

40 30 20 10 0

9.6

26.1

8.17

21.3

iHIV

30.7

IFN-α

CpG C


0.13 1.52

39.3

30.4

20

49.6

g

38.7

44

IFNα

RM SM Microvesicles 6 Fold change in IFN-α1 expression relative to 0 h (log10)

HSV (UV-inact.)

CD123


RM RM mean

8

4,000

5

RM

SM

RM

iSIV

SM

CpG C NS

*

RM SM Influenza (live aichi) SM RM

**

4 3 2 1 0 –1

iSIV

CpG C


Figure 3 IFN-a production by SM pDCs upon TLR7 or TLR9 ligand and iSIV stimulation is lower than that by RM pDCs. (a) IFN-a production by human, RM or SM PBMCs stimulated with microvesicle controls or iSIV. (b,c) PBMCs from RMs or SMs were stimulated with live influenza (aichi) or iSIV. IFN-a production is shown in pg ml–1 (b) and as IFN-a produced per pDC (corrected for the number of pDCs in each animal, as measured by FACS) (c). Error bars depict group means ± s.e.m. (d) IFN-a production by human, RM or SM PBMCs stimulated with heat-inactivated influenza (PR8), R-848, ultraviolet light (UV)-inactivated HSV or CpG C2395. (e,f) IFN-a production by pDCs (gated as in Fig. 1b) after stimulation with microvesicle controls, iSIV, iHIV, UVinactivated HSV, CpG C2395, or live influenza (aichi). Two representative SMs, two RMs and one human are shown in e, and a summary graph is shown in f. (g) IFN-a1 expression assessed by real-time quantitative RT-PCR after stimulation of SM and RM PBMCs with iSIV, CpG C2395 or live influenza (aichi) for 8 h, calculated relative to a housekeeping gene for each individual animal and shown as log10 fold change in expression from the value before stimulation (0 h). Histograms represent means ± s.e.m. of five sooty mangabeys and three rhesus macaques. *P o 0.05 and **P o 0.01. NS, not significant. In scatter graphs, points represent individual animals; group means are denoted by a line throughout. All sooty mangabeys and rhesus macaques shown are SIV negative; all humans shown are HIV negative.

1080

VOLUME 14

[

NUMBER 10

[

OCTOBER 2008

NATURE MEDICINE

ARTICLES surveyed produced significantly lower amounts of IFN-a (Fig. 3a). The diminished IFN-a production in the sooty mangabeys was not explained by differences in pDC numbers, as correction for the number of pDCs showed that IFN-a production per sooty mangabey pDC was also substantially lower than that in rhesus macaques (Fig. 3b,c). We did not detect differences in the frequency of blood pDCs or in the expression of CD4 by pDCs, which is known to be required for SIV- and HIV-induced type I interferon production22, that could have explained the attenuated type I interferon response of sooty mangabey PBMCs to SIV (Supplementary Fig. 2 online). Although a significantly lower fraction of pDCs in sooty mangabeys expressed CCR5 (Supplementary Fig. 2b), one of the co-receptors for HIV and SIV, CCR5 has been shown not to be required for HIV endocytosis and TLR7 and TLR9 signaling in pDCs22,26. The low IFN-a response by sooty mangabey pDCs was not due to failure of the IFN-a antibody used to detect sooty mangabey IFN-a, as live influenza A/aichi stimulation revealed the production of 41,000 pg ml–1 IFN-a in both rhesus macaques and sooty mangabeys (Fig. 3b,c).

Patterns of IFN-a production by sooty mangabey pDCs We next investigated whether the in vivo differences in pDC activation and homing between rhesus macaques and sooty mangabeys were due to differences in pDC function. To do so, we stimulated PBMCs from HIV- or SIV-uninfected humans, rhesus macaques and sooty mangabeys with iSIV and measured IFN-a production. Notably, whereas human and rhesus macaque PBMCs produced large amounts of IFN-a after iSIV stimulation, PBMCs from the 23 sooty mangabeys

a

RM2

RM1 0

0.51

SM1 0.1

0.97

b

SM2 0

0.7

0.66

Microvesicles

1.21 10.5

10.4

3.29

16.3 33.3

12.9

1.52

0.82

1.58

0.29

14.8

78.9

80.6

63.5

54.1

72.8

0 iSIV

18.8

SM

0.19

0.23 9.67

9.23

IFN-α


0

20.5 52.8

3.32

2.45

0.65 1.07

2.56

CD123

0.17 36.4

RM

Human

0

Microvesicles

CpG C iSIV

16.7

8.46

14.2

16.2 TNF-α

TNF-α

c

d

NS

e

NS

3,000

4

NS

NS

SM RM

80

40

2,000

Fold change in IL-12

60

1,000

20

0

0 RM SM Microvesicles

RM

SM

RM

iSIV

SM R-848

expression relative to 0 h (log10)

NS

IL-12 (pg ml–1)

Percentage TNF-α+ cells


production upon iSIV, iHIV, live influenza A/aichi and CpG stimulation in pDC-depleted versus unfractionated PBMCs (Fig. 2c). Whereas the depletion of pDCs reduced the production of IFN-a after stimulation with live influenza A/aichi, a virus that activates multiple innate signaling pathways including retinoic acid–inducible gene I (RIG-I) and TLR3 (ref. 25) in addition to TLR7, by only B50%, it resulted in the near complete abrogation of IFN-a responses to SIV, HIV and CpG (Fig. 2c). Hence, the majority of IFN-a produced in response to SIV and HIV upon TLR7 and TLR9 activation is generated by pDCs.

3

2

1

0 iSIV

CpG C

influenza (live aichi)

Figure 4 The levels of inflammatory cytokine production upon iSIV and TLR7 or TLR9 ligand stimulation of PBMCs in RMs and SMs are similar. (a,b) IFN-a and TNF-a production (a) or TNF-a production only (b) by pDCs after stimulation with microvesicle controls, iSIV, CpG C2395 or live influenza (aichi). Representative animals and humans are shown. In all FACS plots, cells were gated on pDCs (HLA-DR+, lineage-negative, CD123+ cells as in Fig. 1b), and numbers indicate percentage of cells in the gates shown. (c) Percentage pDCs expressing TNF-a in RMs and SMs, as measured by intracellular cytokine staining after stimulation with microvesicle controls or iSIV. (d) R-848–induced IL-12 production above background (IL-12 production with media alone is subtracted) by RM or SM PBMCs. In all scatter plots, points represent individual animals; group means are denoted by a line. (e) IL-12 expression, as assessed by real-time quantitative RT-PCR after stimulation of SM and RM PBMCs with iSIV, CpG C2395 or live influenza (aichi) for 8 h, calculated relative to a housekeeping gene for each individual animal and shown as log10 fold change in expression from before stimulation (0 h). Histograms represent means ± s.e.m. of three SMs and three RMs. *P o 0.05 and **P o 0.01. NS, not significant. All SMs and RMs shown are SIV negative; all humans shown are HIV negative.

NATURE MEDICINE

VOLUME 14

[

NUMBER 10

[

OCTOBER 2008

1081

ARTICLES 10

20

30

40

50

60

70

80

90

100

110

120

130

140

....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|

Human MALAPERAAPRVLFGEWLLGEISSGCYEGLQWLDEARTCFRVPWKHFARKDLSEADARIFKAWAVARGRWPPSSRGGGPPP-EAETAERAGWKTNFRCALRSTRRFVMLRDNSGDPADPHKVYALSRELCWREGPGTDQT RM .............................................................................D...P...A........................................P..G.......... SM .............................................................................D...P...A........................................P..G.......... 150 160 170 180 190 200 210 220 230 240 250 260 270 280 ....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|

Human EAEAPAAVPPPQGGPPGPFLAHTHAGLQAPGPLPAPAGDKGDLLLQAVQQSCLADHLLTASWGADPVPTKAPGEGQEGLPLTGACAGGPGLPAGELYGWAVETTPSPGPQPAALTTGEAAAPESPHQAEPYLSPSPSACT RM ...G....R....R........RD........F...................................AQ..........................CT....A........T..M....T...P........A....... SM ........R....R........RDG.........................G...........A.....AQ..........................CT....A........V..MA...T...P...V....A.......


290 300 310 320 330 340 350 360 370 380 390 400 410 420 ....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|

Human AVQEPSPGALDVTIMYKGRTVLQKVVGHPSCTFLYGPPDPAVRATDPQQVAFPSPAELPDQKQLRYTEELLRHVAPGLHLELRGPQLWARRMGKCKVYWEVGGPPGSASPSTPACLLPRNCDTPIFDFRVFFQELVEFRA RM V..............................M..............................................Q............................................................. SM ...............................M..............................................Q.....................................................R....... 430 440 450 460 470 480 490 500 ....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....|....

Human RQRRGSPRYTIYLGFGQDLSAGRPKEKSLVLVKLEPWLCRVHLEGTQREGVSSLDSSSLSLCLSSANSLYDDIECFLMELEQPA RM .......C.............R...........................................T.........L.......V SM .......C.............R...........................................T.........L.......V

DNA-binding domain

Transactivation domain

Autoinhibitory domain

Phosphorylation domain

IRF phosphorylation motif

Figure 5 Multiple SM-specific amino-acid substitutions are present in the transactivation domain of IRF-7. Amino-acid translation of SM and RM IRF-7 mRNA sequence. Open reading frames are aligned with the human IRF-7 transcript variant A (accession number NM_001572). IRF-7 functional domains are annotated in color as previously defined38, and SM-specific substitutions within these domains are highlighted in white. Gaps are indicated with dashes, and identity with the human sequence is indicated with a period.

To determine whether the reduced IFN-a production capacity of sooty mangabey PBMCs is also seen after exposure to prototypic TLR7 and TLR9 ligands, we stimulated PBMCs from SIV- and HIV-uninfected humans, rhesus macaques and sooty mangabeys with the TLR7 ligands R-848 and heat-inactivated influenza A/PR8 and with TLR9 ligands CpG and ultraviolet light–inactivated herpes simplex virus (HSV). Whereas human and rhesus macaque PBMCs produced similar amounts of IFN-a upon stimulation, sooty mangabey PBMCs made only very limited amounts of IFN-a, regardless of the TLR7 or TLR9 ligand used, suggesting that sooty mangabey PBMCs have a reduced capacity to produce IFN-a that is associated with altered TLR7 and TLR9 signaling (Fig. 3d). Marked attenuation of IFN-a production after exposure of sooty mangabey PBMCs to TLR7 and TLR9 agonists is highly reproducible and has been observed in all (430) sooty mangabeys surveyed to date. Intracellular staining confirmed that there were substantially fewer IFN-a–producing sooty mangabey pDCs upon SIV, TLR7 or TLR9 ligand stimulation compared to rhesus macaque and human pDCs (Fig. 3e,f). Stimulation with live influenza A/aichi showed that the antibody used was able to detect sooty mangabey IFN-a, although there was a significantly lower proportion of IFN-a+ pDCs in sooty mangabeys, perhaps owing to the contribution of TLR7 and TLR9 signaling to the production of type I interferon in response to this virus (Fig. 2c and Fig. 3f). At the peak of IFN-a mRNA induction in rhesus macaques, 8 h after stimulation, there was also a significant difference between rhesus macaques and sooty mangabeys in SIV- and CpG-induced IFN-a mRNA production, but not in live influenza A/aichi-induced IFN-a mRNA production (Fig. 3g). The difference between rhesus macaque and sooty mangabey IFN-a transcription was most striking after stimulation with CpG, which signals solely through TLR9, similar to what we observed at the protein level (Fig. 3g). Whereas there was some upregulation in IFN-a RNA transcription in sooty mangabeys upon stimulation with iSIV (perhaps accounting for the small amount of type I interferon detected at the protein level in sooty mangabeys), it peaked by 4 h after stimulation and then rapidly declined (Supplementary Fig. 3 online). Intact NF-jB–dependent signaling in sooty mangabey pDCs After TLR7 or TLR9 ligand engagement, the downstream signaling pathway bifurcates, leading on the one hand to robust IFN-a

1082

production via a pathway that requires interferon-regulatory factor-7 (IRF-7) and on the other to the production of proinflammatory cytokines such as tumor necrosis factor-a (TNF-a) and interleukin-12 (IL-12) via a nuclear factor-kB (NF-kB)-dependent pathway27–29. To further characterize the difference in TLR7 and TLR9 signaling between sooty mangabey pDCs and rhesus macaque and human pDCs and to assess whether the TLRs themselves or downstream signaling pathways are responsible for the reduced IFN-a production in sooty mangabeys, we examined whether sooty mangabey PBMCs were able to produce proinflammatory cytokines in response to TLR7 and TLR9 signaling. After SIV or CpG stimulation, we observed a similar fraction of TNF-a+ pDCs in both rhesus macaques and sooty mangabeys, despite the fact that sooty mangabey pDCs did not produce IFN-a (Fig. 4a–c). Furthermore, after stimulation of sooty mangabey and rhesus macaque PBMCs with R-848, similar amounts of IL-12 could be detected by ELISA and also by quantitative PCR for IL-12 mRNA in both species (Fig. 4d,e). These data indicate that the pathways involved in SIV, single-stranded RNA and CpG recognition are intact in sooty mangabeys in terms of ligand binding, proximal TLR signaling, NF-kB activation and proinflammatory cytokine production. Consistent with the apparent impairment of the IRF-7 signaling pathway leading to IFN-a production in sooty mangabeys, we observed very little induction of IFN-b after CpG stimulation in sooty mangabeys (Supplementary Fig. 3c). This difference in IFN-b expression, although less pronounced, is also seen upon stimulation with iSIV (Supplementary Fig. 3c). Notably, IRF-7 mRNA, itself induced via an NF-kB–dependent pathway, is expressed at similar levels in sooty mangabeys and rhesus macaques upon TLR7 or TLR9 stimulation by iSIV, live influenza A/aichi or CpG (Supplementary Fig. 3d), suggesting that IRF-7 induction, although perhaps not its function, is intact. Genetic polymorphisms in TLR7 and TLR9 signaling pathways Seeking to identify specific genetic polymorphisms that may be responsible for the altered TLR7 and TLR9 signaling phenotype in sooty mangabeys, we sequenced genes encoding proteins involved in TLR7 and TLR9 signaling from sooty mangabeys and rhesus macaques and compared them with human sequences (Supplementary Table 1 online). Most of these genes were highly conserved

VOLUME 14

[

NUMBER 10

[

OCTOBER 2008

NATURE MEDICINE

***

ARTICLES

SM

1.5

RM

**

*

* ** *

0.5

*

* **

*

**

*

1.0

**

Fold change in expression relative to uninfected controls (log10)


Human

0.0

–0.5

IFN-α1

IFN-α2

IL-6

MX-1

IP-10

TNF-α

IL-12

Figure 6 Elevations in IFN-a and type I interferon responses occur in chronically SIV-infected RMs and HIV-infected humans, but not in SIVinfected SMs. IFN-a1, IFN-a2, IL-6, MX-1, IP-10, TNF-a and IL-12 expression in SIV+ SMs, SIV+ RMs and HIV+ humans, as assessed by realtime quantitative RT-PCR, calculated relative to a housekeeping gene for each individual and shown as log10 fold change in expression compared to uninfected SMs, RMs and humans, respectively. Histograms represent means ± s.e.m. of six SMs, eight RMs and twelve humans. P values indicate comparisons of SM gene expression to that of humans and RMs. *P o 0.05, **P o 0.01 and ***P o 0.001.

between all three species. Given that there was no difference in the production of proinflammatory cytokines after TLR7 or TLR9 stimulation between sooty mangabeys and rhesus macaques, it is not surprising that there were few amino acid changes in TLR7, TLR9 and MyD88 (an adaptor protein required for TLR7 and TLR9 signaling) between sooty mangabeys and rhesus macaques (Supplementary Fig. 4 online). In fact, TLR7 and TLR9 were the most conserved of the ten TLRs sequenced (Supplementary Table 1). However, IRF-7 had a number of sooty mangabey–specific amino acid substitutions within its transactivation domain (Fig. 5). To determine whether transcription factor binding motifs involved in the regulation of interferon mRNA production were altered in sooty mangabeys, we sequenced the promoters and open reading frames of IFN-a1 and IFN-a2, as well as of IFN-b1, which are the interferon subtypes most highly induced by viral infection or TLR stimulation (Supplementary Fig. 4). Except for one polymorphism found in the IFN-a1 promoter in some sooty mangabeys, interferon promoters and coding regions were identical between rhesus macaques and sooty mangabeys. Hence, on the basis of the relative degree and character of IRF-7 sequence polymorphisms and of its functional role in the IFN-a response, IRF-7 is the most probable candidate responsible for the observed reduction in sooty mangabey IFN-a production after TLR7 and TLR9 stimulation. Type I interferon signatures and AIDS susceptibility To assess the in vivo consequences of the distinctive patterns of TLR7 and TLR9 signaling and IFN-a production in sooty mangabey, rhesus macaque and human pDCs, we examined the IFN-a and type I interferon response gene expression profiles in infected and uninfected sooty mangabeys, rhesus macaques and humans by quantitative

NATURE MEDICINE

VOLUME 14

[

NUMBER 10

[

OCTOBER 2008

real-time PCR. We quantified the expression of mRNAs encoding IFN-a1 and IFN-a2, the interferon response proteins MX-1, IL-6 and IP-10, and TNF-a and IL-12 in uninfected sooty mangabeys, SIV+ sooty mangabeys, uninfected rhesus macaques, SIV+ rhesus macaques, uninfected humans and HIV+ humans (CD4 counts and SIV or HIV viral loads are shown in Supplementary Table 2 online). Whereas these mRNAs were upregulated by 1.3–18-fold in SIV+ rhesus macaques and HIV+ humans compared to uninfected controls, they were expressed at similar (or lower) levels in infected compared to uninfected sooty mangabeys (Fig. 6). Thus, the reduced capacity of sooty mangabey pDCs to produce IFN-a upon ex vivo stimulation with SIV translates to a profound absence in chronically SIV-infected sooty mangabeys of the characteristic type I interferon gene expression profile observed in pathogenic SIV or HIV infections of rhesus macaques or humans. DISCUSSION Pathogenic HIV or SIV infection has numerous effects on the host immune system, including the induction of increased turnover, apoptosis and dysfunction of NK, B and T cells1,2,4,5—all of which are probably the consequence of the prevailing environment of chronic immune activation. Determining how sooty mangabeys and other natural hosts avoid chronic immune activation could provide insight into the pathogenic mechanisms responsible for AIDS progression in non-natural hosts13,14. Here we describe key differences in the maturation and trafficking of pDCs in sooty mangabeys after infection with SIV that are associated with their reduced IFN-a production after TLR7 or TLR9 stimulation. The markedly attenuated production of IFN-a by sooty mangabey pDCs after SIV engagement of TLR7 and TLR9 may not only result in the generation of very limited innate and adaptive antiviral cellular immune responses during acute SIV infection, but also enable sooty mangabeys to avoid generalized immune activation during chronic infection. We propose a model of AIDS pathogenesis in non-natural hosts wherein the failure of adaptive immune responses to durably control virus replication permits ongoing innate immune system stimulation by the substantial amounts of both infectious and noninfectious virions produced each day in infected hosts30. This vicious cycle of continuous virus replication, innate immune stimulation and multifactorial immunopathology may represent the primary force driving AIDS progression in pathogenic HIV and SIV infections. Our study identified the lack of pDC maturation and lymphoid tissue homing as an early point of divergence in the natural sooty mangabey host response to SIV infection. Furthermore, we showed that NK cells, another key innate effector population, do not proliferate in sooty mangabeys during acute SIV infection, in sharp contrast to NK cell expansions seen in acutely HIV-infected humans5 or SIVinfected rhesus macaques (this study and ref. 31). The diminished NK cell proliferation in acutely SIV-infected sooty mangabeys probably stems from the absence of activated DCs and DC-elaborated cytokines to signal NK cell proliferation32. Given the importance of reciprocal interactions between NK cells and DCs in achieving full functioning of both cell types, as well as the importance of these interactions in generating cellular immune responses32, it is likely that the reduced NK cell proliferation in infected sooty mangabeys not only reinforces the continued absence of DC activation despite considerable viremia, but also affects the magnitude and nature of the antiviral cellular immune responses that are generated33,34. Plasmacytoid DCs have a central role in activating both innate and adaptive responses, as evidenced by pDC activation leading to the

1083


ARTICLES direct or indirect activation of many cell types, including monocytes, mDCs, B cells, NK cells and T cells17,35,36. IFN-a, produced primarily from pDCs, in addition to having direct antiviral activity, provides an important signal for T helper precursor differentiation in favor of a T helper type 1 immune response18. The pDC-induced maturation of mDCs then enables efficient antigen presentation and stimulation of naive T cells, and IFN-a itself can directly amplify CD8+ T cell expansion after viral infections19. Although sooty mangabey pDCs produce less IFN-a upon in vitro stimulation with SIV or TLR7 and TLR9 ligands, they do produce proinflammatory cytokines known to result from a distinct downstream signaling pathway involving NF-kB, which is also activated after TLR engagement29. These results, together with the conserved TLR7 and TLR9 sequences in sooty mangabeys, indicate that the muted type I interferon response does not result from an inability of the initial upstream receptors to detect SIV or other TLR7 and TLR9 ligands, but rather from divergent propagation of activation signals along signaling pathways downstream of receptor binding. A potential molecular mediator that is crucial for type I interferon production and that might be involved in this divergent response is IRF-7. Indeed, after stimulation of sooty mangabey pDCs with CpG DNA (known to signal solely through TLR9 (ref. 37)), near-absent IFN-a and IFN-b production in concert with preserved production of other NF-kB–regulated proinflammatory cytokines greatly resembles the phenotype observed in Irf7 –/– mice27. Furthermore, reminiscent of the sooty mangabey response to SIV, Irf7 –/– mice show marked impairment in the generation of CD8+ T cell responses to antigens whose immunogenicity depends on IRF-7–dependent TLR signaling27. We have found that, unlike other proteins involved in TLR7 and TLR9 signaling that are highly conserved between sooty mangabeys, rhesus macaques and humans, the IRF-7 gene sequence contains numerous sooty mangabey–specific amino acid substitutions. The majority of these changes in IRF-7 are located in the transactivation domain, which recruits other transcriptional coactivators to the IFN-a and IFN-b promoters38,39. Although these amino acid changes are relatively conservative in nature, the cumulative effect of the substitutions might impede the phosphorylation or nuclear translocation of IRF-7 or compromise its stability40 or ability to engage transcriptional co-activators—thus limiting type I interferon production after TLR7 and TLR9 stimulation in sooty mangabeys. Additional analyses will be needed to evaluate the role of IRF-7 in the specific pattern of type I interferon production by sooty mangabey pDCs. The preserved ability of sooty mangabey pDCs to produce TNF-a and IL-12 after TLR7 and TLR9 stimulation, along with apparent redundancies in innate responses inferred from humans with rare inherited polymorphisms leading to impaired TLR signaling, who show discrete rather than generalized susceptibility to specific infections41, may explain why sooty mangabeys are able to mount effective immune responses to most potential pathogens and remain healthy in the wild, despite a reduced ability to express IFN-a in response to engagement of TLR7 and TLR9 by specific agonists. Furthermore, we can infer that because the signaling pathway leading to IL-12 and TNF-a production in response to TLR7 and TLR9 stimulation in sooty mangabeys is intact, neither of these cytokines by itself is responsible for the chronic activation of the immune system seen during SIV or HIV infection. However, the reduced type I interferon response in sooty mangabeys might also promote a more complicated phenotype where potentially damaging innate and adaptive immune responses to SIV are also actively downmodulated. Although TLR recognition of viral components and DC activation is crucial for antiviral immunity, TLR stimulation has also been

1084

reported to directly contribute to deleterious inflammatory responses. In mouse models, high levels of IFN-a result in bone marrow suppression, reduced thymic cellularity and thymocyte proliferation, thereby inhibiting B and T lymphopoiesis42,43. In addition, ongoing specific TLR9 stimulation by chronic administration of CpG has been shown to damage the microarchitecture and function of lymphoid organs via IFN-a production44. Of note, mice deficient in TLR3 have paradoxically attenuated disease, diminished pathologic inflammatory responses and improved survival after infection with West Nile virus, phlebo virus or influenza A virus, despite having equivalent or greater levels of virus replication than those seen in wild-type controls45–47. HIV-induced type I interferon production by human pDCs can enhance the activation-induced death of primary CD4+ T cells via tumor necrosis factor–related apoptosis-inducing ligand (TRAIL)- or Fas-mediated pathways, suggesting a direct mechanism for the increased CD4+ T cell apoptosis observed in HIV-infected individuals48,49. The association between a type I interferon gene expression signature indicative of persistent and ongoing IFN-a production in vivo during pathogenic AIDS virus infections has been reported50–53, consistent with our demonstration of increased IFN-a and type I interferon response mRNA expression in HIV-infected humans and SIV-infected rhesus macaques. Of note, we show here that chronically SIV-infected sooty mangabeys do not show upregulation of type I interferon response mRNAs. Taken together, these data suggest that ongoing pDC activation and IFN-a production may lead to both the accelerated destruction and the impaired regeneration of CD4+ T cells, as well as dysregulation of cellular and humoral immune responses in pathogenic SIV and HIV infections. Other studies have posited that the early and severe depletion of mucosal CD4+ T cells in pathogenic AIDS virus infections underlies immune activation and disease progression in non-natural hosts54. According to this model, damage to gut-associated lymphoid tissue by HIV infection and depletion of resident CD4+ T cells leads to the compromise of gut mucosal integrity and the influx of microbial components that themselves activate the immune system54. The positive correlation between plasma lipopolysaccharide (LPS) abundance and the extent of immune activation in HIV-infected people, as well as the lower amounts of LPS in plasma of SIV-infected sooty mangabeys has been cited as evidence supporting this hypothesis54. However, it is difficult to reconcile this hypothesis with the observation that the same magnitude of mucosal CD4+ T cell depletion occurs in nonpathogenic infections of sooty mangabeys and African green monkeys as is observed in rhesus macaques and humans10,55. In the absence of aberrant immune activation in these natural hosts, the depletion of CD4+ T cells in gut-associated lymphoid tissues is not sufficient to compromise the gut epithelium and points to the primary role of immune activation in this process. Although LPS translocation might potentially augment immune activation in pathogenic AIDS virus infections, it is not likely to be the primary cause of it. An alternative hypothesis consistent with the observed correlation between LPS translocation, immune activation and disease susceptibility is that active virus replication in gut mucosal tissues, along with the recruitment and chronic in situ activation of pDCs, precipitates inflammatory damage to mucosal epithelial surfaces, enabling LPS and other microbial products in the gut lumen to traverse the normally impermeable mucosal barrier and gain access to the bloodstream. CD4+ T cell–tropic lentiviruses may be uniquely associated with chronic generalized immune activation during infections of nonnatural hosts not only because they establish persistent infections that cannot be cleared by host adaptive immune responses, but also

VOLUME 14

[

NUMBER 10

[

OCTOBER 2008

NATURE MEDICINE


ARTICLES because they preferentially target cells that express cell surface CD4. Although conventional models of AIDS pathogenesis have highlighted direct infection of CD4+ T cells as the primary event driving disease progression, efficient binding to and activation of pDCs by HIV as a result of their expression of CD4 may be an additional key determinant. Through facilitated binding to CD4, efficient endocytosis and TLR7 and TLR9 signaling, HIV and SIV engender ongoing pDC activation and type I interferon production. In comparison to HIV, hepatitis C virus also causes chronic high-level viremia in humans but does not target or stimulate pDCs56. As a result, among virus infections of humans, HIV may be uniquely able to result in the chronic activation of multiple distinct branches of the immune system. On the basis of the model of HIV pathogenesis that we propose, there are a number of immunomodulatory interventions that might ameliorate pathologic immune activation in non-natural hosts for SIV and HIV. These include interventions that interrupt virus binding to CD4 on pDCs (using antibodies to CD4 or CD4-immunoglobulin immunoadhesins) or that diminish the elicitation of type I interferon production by pDCs. Potential approaches to limit virus-activated IFN-a production include the use of TLR-targeted antagonists to mitigate signaling after HIV-1 engagement of TLR7 and TLR9 (ref. 24) or pharmacological interference with endososomal acidification and signaling by TLR7 and TLR9 (for example, using chloroquine)21. Each of these approaches would broadly, and potentially undesirably, affect all TLR7- and TLR9-mediated pDC responses, including inhibition of production of IFN-a, TNF-a and IL-12 by pDCs. Pharmacological interference with IRF-7 might, if feasible, confer greater specificity. Alternatively, inhibition of the deleterious consequences of chronic IFN-a production by pDCs after HIV-1 infection might be accomplished by administration of antibodies capable of neutralizing IFN-a (ref. 57). Such unique immunomodulatory strategies, used in conjunction with antiretroviral drugs, may be of particular value in individuals who experience impaired recovery of CD4+ T cells despite effective antiretroviral suppression of virus replication. To our knowledge, this work represents the first description of a naturally occurring, species-wide polymorphism in TLR signaling that has major effects on innate immune function and a substantial impact on host responses and disease susceptibility to a prevalent pathogen. If host immune responses are fundamentally unable to durably control HIV or SIV replication, and chronic viremia drives pathological innate immune activation, then selection for decreased innate immune responses may represent an effective adaptive evolutionary response to avoid susceptibility to immunodeficiency disease. Understanding the commonalities between sooty mangabeys and other natural hosts for nonpathogenic SIV infections may enable us to anticipate the future evolutionary trajectory of human populations in response to the profound selective pressures imposed by the AIDS pandemic. Likewise, identification of genetic polymorphisms associated with variable thresholds for activation of innate immune responses may illuminate host factors contributing to the heretofore unexplained diversity in disease courses observed in HIV infected humans58. METHODS Human subjects. We obtained blood samples from HIV+ or HIV– humans after they had provided written informed consent. Inclusion criteria for HIVinfected humans were viral loads of 41 103 RNA copies ml–1, CD4 counts of 4120 cells ml–1, no active opportunistic infections and antiretroviral-naive or off antiretroviral treatment for 41.5 years (Supplementary Table 2). Protocols were approved by the Emory University Institutional Review Board.

NATURE MEDICINE

VOLUME 14

[

NUMBER 10

[

OCTOBER 2008

Animals. We sampled 30 rhesus macaques (Macaca mulatta, Indian genetic background) and 26 sooty mangabeys (Cercocebus atys) and housed them at the Yerkes National Primate Research Center. All animals were SIV negative, with the exception of the six animals experimentally infected with SIVsm as described below, as well as six naturally SIV-infected sooty mangabeys and eight SIVmac239-infected rhesus macaques cross-sectionally sampled for gene expression analyses. We performed SIVmac239 infections at least one year before sampling, as previously described59. In comparative sooty mangabey and rhesus macaque studies, animals were age- and sex-matched wherever possible. Animal housing, care and research was in accordance with the Guide for the Care and Use of Laboratory Animals, and all studies were approved by the Emory University Institutional Animal Use and Care Committee. Uncloned SIVsm stock. We intravenously inoculated 1 ml of plasma from a naturally SIV-infected sooty mangabey into an uninfected sooty mangabey. We stored plasma from blood drawn at the peak of viremia (day 11) at –80 1C. SIV infections. We intravenously inoculated three rhesus macaques and three sooty mangabeys with 0.5 ml SIVsm stock (5 106 SIV RNA copies). We measured immunological and virological parameters on day –14, day 0, day 2, day 4, day 7, day 10, day 14, day 21 and day 28. We took lymph node biopsies on day –14 and on day 10 or day 14 of SIV infection as previously described11. SIVsm viral loads. We quantified SIVsm viral loads as previously described12. Immunophenotyping. We stained cells as previously described33 with monoclonal antibodies to the following proteins (antibody names in parentheses): CD3 (SP34-2), CD4 (L-200), CD8 (SK1), CD11c (S-HCL-3), CD20 (L27), CD123 (73G), CCR5 (3A9) and HLA-DR (L243), all from BD; CD14 (My4) and CD16 (2H4) from Coulter; and CCR7 from R&D Systems. We acquired at least 100,000 gated events for lymphocytes or 700,000 total events for DCs on a FACSCalibur or LSRII flow cytometer (BD). We performed analyses with FlowJo (Treestar). Electron microscopy. To mobilize DCs in sooty mangabeys in vivo, we administered 100 mg kg–1 Flt3L-IgG2 (R.C. and M.B.F., unpublished data) subcutaneously for 5 d. We isolated PBMCs on day 8 after treatment and enriched for DCs by depleting CD20+ and CD3+ cells on a MACS LD column (Miltenyi). We then stained the cells with antibodies to CD14, HLA-DR, CD123 and CD11c (BD) and sorted them with a FACSARIA (BD). We fixed sorted pDCs and mDCs with 2.5% glutaraldehyde in 0.1 M cocadylate buffer, post-fixed them with buffered 1% osmium tetroxide for 1 h, dehydrated them in a graded ethanol series to 100%, embedded them in low-viscosity epoxy resin and hardened them at 60 1C. We thin-sectioned the resin blocks at 60–70 nm, stained them with uranyl acetate and lead citrate and examined them on a Hitachi H7500 transmission electron microscope. Ex vivo peripheral blood mononuclear cell stimulation. We stimulated PBMCs in complete RPMI (Invitrogen) at 400,000 cells per well in duplicate wells with 1 mM R-848 (3M Pharmaceuticals), 6 mg ml–1 CpG C2395 or CpG A2336 (Coley Pharmaceuticals), multiplicity of infection (MOI) 0.25 ultraviolet light–inactivated HSV (described in ref. 24), 1 105 hemagglutination (HA) units ml–1 live influenza virus A/aichi H3N2 (Charles River Laboratories), MOI 2.0 heat-inactivated (56 1C for 30 min) influenza virus A/PR/8 H1N1 (American Type Culture Collection) or 500 ng ml–1 (total protein) aldrithiol2–inactivated SIVmac239, aldrithiol-2–inactivated HIV-1ADA or SUPT1 cell–derived microvesicles as a control (the latter three were a gift, see Acknowledgments) for 17 h at 37 1C. Briefly, the aldrithiol-2 treatment of SIV and HIV particles (described in ref. 60) covalently modifies thiol groups in internal viral proteins, rendering the particles noninfectious while leaving envelope glycoproteins intact. For some stimulations, we added inhibitors of TLR7, TLR9 or both: 5.6 mM IRS954, IRS661, IRS869 or oligodeoxynucleotide control (Dynavax Technologies), 2 mM DV056 (Dynavax Technologies), or 20 mM or 5 mM chloroquine (Sigma). After stimulation, we harvested supernatants for cytokine detection with the Human Interferon Alpha Multi-Subtype ELISA Kit (PBL Biomedical Laboratories) or IL-12 p70 ELISA Kit MK (BioSource). All kits were cross-reactive for rhesus macaques, sooty mangabeys and humans.

1085

ARTICLES


Ex vivo pDC depletions. We magnetically labeled PBMCs with a BDCA-4 microbead kit (Miltenyi) and depleted them of B90% of pDCs by passing them three times through an autoMACS separator (Miltenyi). Intracellular cytokine detection. We stimulated 4 106 PBMCs with 1 105 HA units ml–1 live influenza virus A/aichi H3N2, 1 mg ml–1 iSIV, iHIV or microvesicles, MOI 0.25 ultraviolet light–inactivated HSV or 24 mg ml–1 CpG C2395 for 8 h at 37 1C. We added 5 mg ml–1 brefeldin-A (Sigma) and continued the incubation for 4 h. We fixed cells in 2% paraformaldehyde for 5 min, washed them and permeabilized them overnight in PBS containing saponin (2 g l–1), 0.08% BSA, 10 mM HEPES, 1 mM CaCl2, 0.2 mM MgSO4 and 5% nonfat milk (PBS-S-M) at 4 1C. We stained cells in fresh PBS-S-M containing monoclonal antibodies to CD3, CD20, CD14, HLA-DR, CD123, TNF-a (MAB11, BD) and IFN-a (225.C, Chromaprobe). We acquired at least 1 106 total events on a LSRII flow cytometer. Gene expression analysis. For mRNA expression analyses after TLR stimulation, we first stimulated PBMCs as for intracellular cytokine detection. Before adding stimulant (0 h) and 2 h, 4 h, 6 h, 8 h and 16 h after stimulation, we lysed cells in RLT buffer (Qiagen). For mRNA expression analyses of uninfected or chronically SIV- or HIV-infected humans, rhesus macaques or sooty mangabeys, we isolated PBMCs, washed them and lysed them. We extracted RNA was extracted with RNeasy kits (Qiagen) and reverse-transcribed the RNA with a High-Capacity cDNA Archive Kit (Applied Biosystems). We carried out the relative quantitative real-time PCRs for nine selected genes (Supplementary Table 3 online) following TaqMan Applied Biosystems protocols. We tested samples in duplicate in parallel with the housekeeping gene GUSB. We validated assays for rhesus macaques and sooty mangabeys where appropriate. We used StatMiner software (Integromics) to perform quality controls for all runs and relative quantification DDCt analyses to calculate the fold differences between samples. Gene sequencing. See Supplementary Methods online for a more detailed description. Briefly, we reverse-transcribed cellular RNA extracted from HIVor SIV-negative human, rhesus macaque or sooty mangabey PBMCs with either Powerscript (Clontech) or Superscript II (Invitrogen) and an oligo-dT primer. We amplified the gene regions coding for IFN-a1, IFN-a2 and IFN-b1 with the Takara LA kit (Chemicon International) from genomic DNA extracted from sooty mangabey and rhesus macaque PBMCs with the QIAamp DNA MiniKit (Qiagen). We designed PCR primers with available human or nonhuman primate sequences. After gel purification and TOPO-cloning (Invitrogen), we sequenced multiple independent clones from each PCR reaction (Lark Technologies) to identify potential PCR-induced mutations. Statistical analyses. We analyzed data by t-tests, Mann-Whitney tests (for nonparametric data) or, in cases where more than one group was compared, by Kruskal-Wallis analysis of variance (GraphPad Prism). A P value below 0.05 was considered significant for all analyses. Accession codes. Sooty mangabey and rhesus macaque gene sequences encoding the proteins studied are deposited in GenBank with the following accession codes (SM; RM): IKK-a (EU204926; EU204927), IRAK1 (EU204925; EU204924), IRAK4 (EU204923; EU204922), IRF2 (EU204920; EU204921), IRF3 (EU204918; EU204919), IRF7 (EU204916; EU204917), MyD88 (EU204915; EU204914), TRAF6 (EU204928; EU204929), TLR1 (EU204931; EU204930), TLR2 (EU204932; EU204933), TLR3 (EU204935; EU204934), TLR4 (EU204937; EU204936), TLR5 (EU204938; EU204939), TLR6 (EU204940; EU204941), TLR7 (EU204942; EU204943), TLR8 (EU204945; EU204944), TLR9 (EU204946; EU204947). For details, see Supplementary Table 1 and Supplementary Figure 4. Note: Supplementary information is available on the Nature Medicine website. ACKNOWLEDGMENTS The authors would like to thank B. Weaver, J. Skvarich, B. O’Hara and M. Mulligan for their help coordinating blood draws from HIV-infected humans, S. Ehnert and E. Strobert for their care of the study animals, B. Lawson and D. Lee for performing the SIVsm and SIVmac239 viral load assays, J. Ingersoll for performing the HIV viral load assays, A. McCrary for assistance with cloning

1086

and sequencing, K. Dalbey for help with running ELISAs, H. Yi for help with electron microscopy, J. Bess and J. Lifson at the US National Cancer Institute for providing the aldrithiol-2–inactivated SIV and HIV strains and microvesicle controls, 3M Pharmaceuticals for providing the R-848 and anonymous human volunteers for providing blood samples for these studies. We would also like to thank G. Silvestri for his input. We gratefully acknowledge the support of the US National Institutes of Health grants R01 HL075766 and R01 AI049155, the Yerkes National Primate Research Center Grant RR000165 and the Emory Center for AIDS Research Grant P30-AI-50409. The authors apologize for not citing all relevant publications due to space limitations. AUTHOR CONTRIBUTIONS J.N.M., A.P.B. and M.B.F. designed the experiments, and J.N.M. and A.P.B. conducted most of them. T.H.V. sequenced genes involved in the TLR signaling pathway under the supervision of S.I.S., N.K. performed gene expression analyses, R.C. and S.K. developed assays and reagents that paved the way for this work, F.J.B. and R.L.C. provided the TLR antagonists and contributed to planning inhibition experiments, and M.B.F. supervised the overall project. J.N.M., A.P.B. and M.B.F. analyzed the data and J.N.M. and M.B.F. wrote the manuscript. COMPETING INTERESTS STATEMENT The authors declare competing financial interests: details accompany the full-text HTML version of the paper at http://www.nature.com/naturemedicine/. Published online at http://www.nature.com/naturemedicine/ Reprints and permissions information is available online at http://npg.nature.com/ reprintsandpermissions/ 1. Hazenberg, M.D. et al. T-cell division in human immunodeficiency virus (HIV)-1 infection is mainly due to immune activation: a longitudinal analysis in patients before and during highly active antiretroviral therapy (HAART). Blood 95, 249–255 (2000). 2. Hellerstein, M.K. et al. Subpopulations of long-lived and short-lived T cells in advanced HIV-1 infection. J. Clin. Invest. 112, 956–966 (2003). 3. Finkel, T.H. et al. Apoptosis occurs predominantly in bystander cells and not in productively infected cells of HIV- and SIV-infected lymph nodes. Nat. Med. 1, 129– 134 (1995). 4. Moir, S. et al. Decreased survival of B cells of HIV-viremic patients mediated by altered expression of receptors of the TNF superfamily. J. Exp. Med. 200, 587–599 (2004). 5. Kottilil, S. et al. Enhanced susceptibility to CD95-mediated natural killer cell death and turnover induced by HIV viremia. J. Acquir. Immune Defic. Syndr. 46, 151–159 (2007). 6. Barron, M.A., Blyveis, N., Palmer, B.E., MaWhinney, S. & Wilson, C.C. Influence of plasma viremia on defects in number and immunophenotype of blood dendritic cell subsets in human immunodeficiency virus 1–infected individuals. J. Infect. Dis. 187, 26–37 (2003). 7. Grossman, Z., Meier-Schellersheim, M., Paul, W.E. & Picker, L.J. Pathogenesis of HIV infection: what the virus spares is as important as what it destroys. Nat. Med. 12, 289–295 (2006). 8. Giorgi, J.V. et al. Shorter survival in advanced human immunodeficiency virus type 1 infection is more closely associated with T lymphocyte activation than with plasma virus burden or virus chemokine coreceptor usage. J. Infect. Dis. 179, 859–870 (1999). 9. Hahn, B.H., Shaw, G.M., De Cock, K.M. & Sharp, P.M. AIDS as a zoonosis: scientific and public health implications. Science 287, 607–614 (2000). 10. Gordon, S.N. et al. Short-lived infected cells support virus replication in naturally SIV-infected sooty mangabeys: implications for AIDS pathogenesis. J. Virol. 82, 3725–3735 (2008). 11. Gordon, S.N. et al. Severe depletion of mucosal CD4+ T cells in AIDS-free simian immunodeficiency virus–infected sooty mangabeys. J. Immunol. 179, 3026–3034 (2007). 12. Sumpter, B. et al. Correlates of preserved CD4+ T cell homeostasis during natural, nonpathogenic simian immunodeficiency virus infection of sooty mangabeys: implications for AIDS pathogenesis. J. Immunol. 178, 1680–1691 (2007). 13. Silvestri, G. et al. Nonpathogenic SIV infection of sooty mangabeys is characterized by limited bystander immunopathology despite chronic high-level viremia. Immunity 18, 441–452 (2003). 14. Silvestri, G. et al. Divergent host responses during primary simian immunodeficiency virus SIVsm infection of natural sooty mangabey and nonnatural rhesus macaque hosts. J. Virol. 79, 4043–4054 (2005). 15. Banchereau, J. et al. Immunobiology of dendritic cells. Annu. Rev. Immunol. 18, 767–811 (2000). 16. Coates, P.T. et al. Dendritic cell subsets in blood and lymphoid tissue of rhesus monkeys and their mobilization with Flt3 ligand. Blood 102, 2513–2521 (2003). 17. Fonteneau, J.F. et al. Human immunodeficiency virus type 1 activates plasmacytoid dendritic cells and concomitantly induces the bystander maturation of myeloid dendritic cells. J. Virol. 78, 5223–5232 (2004). 18. Biron, C.A. Interferons a and b as immune regulators—a new look. Immunity 14, 661–664 (2001).

VOLUME 14

[

NUMBER 10

[

OCTOBER 2008

NATURE MEDICINE


ARTICLES 19. Kolumam, G.A., Thomas, S., Thompson, L.J., Sprent, J. & Murali-Krishna, K. Type I interferons act directly on CD8 T cells to allow clonal expansion and memory formation in response to viral infection. J. Exp. Med. 202, 637–650 (2005). 20. Liu, Y.J. IPC: professional type 1 interferon–producing cells and plasmacytoid dendritic cell precursors. Annu. Rev. Immunol. 23, 275–306 (2005). 21. Iwasaki, A. & Medzhitov, R. Toll-like receptor control of the adaptive immune responses. Nat. Immunol. 5, 987–995 (2004). 22. Beignon, A.S. et al. Endocytosis of HIV-1 activates plasmacytoid dendritic cells via Toll-like receptor–viral RNA interactions. J. Clin. Invest. 115, 3265–3275 (2005). 23. Meier, A. et al. MyD88-dependent immune activation mediated by human immunodeficiency virus type 1–encoded Toll-like receptor ligands. J. Virol. 81, 8180–8191 (2007). 24. Barrat, F.J. et al. Nucleic acids of mammalian origin can act as endogenous ligands for Toll-like receptors and may promote systemic lupus erythematosus. J. Exp. Med. 202, 1131–1139 (2005). 25. Pichlmair, A. & Reis e Sousa, C. Innate recognition of viruses. Immunity 27, 370–383 (2007). 26. Herbeuval, J.P. et al. Regulation of TNF-related apoptosis-inducing ligand on primary CD4+ T cells by HIV-1: role of type I IFN–producing plasmacytoid dendritic cells. Proc. Natl. Acad. Sci. USA 102, 13974–13979 (2005). 27. Honda, K. et al. IRF-7 is the master regulator of type I interferon–dependent immune responses. Nature 434, 772–777 (2005). 28. Tamura, T., Yanai, H., Savitsky, D. & Taniguchi, T. The IRF family transcription factors in immunity and oncogenesis. Annu. Rev. Immunol. 26, 535–584 (2008). 29. Akira, S., Uematsu, S. & Takeuchi, O. Pathogen recognition and innate immunity. Cell 124, 783–801 (2006). 30. Markowitz, M. et al. A novel antiviral intervention results in more accurate assessment of human immunodeficiency virus type 1 replication dynamics and T cell decay in vivo. J. Virol. 77, 5037–5038 (2003). 31. Giavedoni, L.D., Velasquillo, M.C., Parodi, L.M., Hubbard, G.B. & Hodara, V.L. Cytokine expression, natural killer cell activation and phenotypic changes in lymphoid cells from rhesus macaques during acute infection with pathogenic simian immunodeficiency virus. J. Virol. 74, 1648–1657 (2000). 32. Ferlazzo, G. Natural killer and dendritic cell liaison: recent insights and open questions. Immunol. Lett. 101, 12–17 (2005). 33. Barry, A.P. et al. Depletion of CD8+ cells in sooty mangabey monkeys naturally infected with simian immunodeficiency virus reveals limited role for immune control of virus replication in a natural host species. J. Immunol. 178, 8002–8012 (2007). 34. Dunham, R. et al. The AIDS resistance of naturally SIV-infected sooty mangabeys is independent of cellular immunity to the virus. Blood 108, 209–217 (2006). 35. Kamath, A.T., Sheasby, C.E. & Tough, D.F. Dendritic cells and NK cells stimulate bystander t cell activation in response to TLR agonists through secretion of IFN-a, IFN-b and IFN-g. J. Immunol. 174, 767–776 (2005). 36. Jego, G. et al. Plasmacytoid dendritic cells induce plasma cell differentiation through type I interferon and interleukin 6. Immunity 19, 225–234 (2003). 37. Reis e Sousa, C. Toll-like receptors and dendritic cells: for whom the bug tolls. Semin. Immunol. 16, 27–34 (2004). 38. Marie, I., Smith, E., Prakash, A. & Levy, D.E. Phosphorylation-induced dimerization of interferon regulatory factor 7 unmasks DNA binding and a bipartite transactivation domain. Mol. Cell. Biol. 20, 8803–8814 (2000). 39. Honda, K., Yanai, H., Takaoka, A. & Taniguchi, T. Regulation of the type I IFN induction: a current view. Int. Immunol. 17, 1367–1378 (2005).

NATURE MEDICINE

VOLUME 14

[

NUMBER 10

[

OCTOBER 2008

40. Sato, M. et al. Distinct and essential roles of transcription factors IRF-3 and IRF-7 in response to viruses for IFN-a/b gene induction. Immunity 13, 539–548 (2000). 41. Zhang, S.Y. et al. Human Toll-like receptor–dependent induction of interferons in protective immunity to viruses. Immunol. Rev. 220, 225–236 (2007). 42. Binder, D., Fehr, J., Hengartner, H. & Zinkernagel, R.M. Virus-induced transient bone marrow aplasia: major role of interferon-ab during acute infection with the noncytopathic lymphocytic choriomeningitis virus. J. Exp. Med. 185, 517–530 (1997). 43. Lin, Q., Dong, C. & Cooper, M.D. Impairment of T and B cell development by treatment with a type I interferon. J. Exp. Med. 187, 79–87 (1998). 44. Heikenwalder, M. et al. Lymphoid follicle destruction and immunosuppression after repeated CpG oligodeoxynucleotide administration. Nat. Med. 10, 187–192 (2004). 45. Wang, T. et al. Toll-like receptor 3 mediates West Nile virus entry into the brain causing lethal encephalitis. Nat. Med. 10, 1366–1373 (2004). 46. Gowen, B.B. et al. TLR3 deletion limits mortality and disease severity due to phlebovirus infection. J. Immunol. 177, 6301–6307 (2006). 47. Le Goffic, R. et al. Detrimental contribution of the Toll-like receptor (TLR)3 to influenza A virus–induced acute pneumonia. PLoS Pathog. 2, e53 (2006). 48. Herbeuval, J.P. & Shearer, G.M. HIV-1 immunopathogenesis: how good interferon turns bad. Clin. Immunol. 123, 121–128 (2007). 49. Kaser, A., Nagata, S. & Tilg, H. Interferon a augments activation-induced T cell death by upregulation of Fas (CD95/APO-1) and Fas ligand expression. Cytokine 11, 736–743 (1999). 50. Hyrcza, M.D. et al. Distinct transcriptional profiles in ex vivo CD4+ and CD8+ T cells are established early in human immunodeficiency virus type 1 infection and are characterized by a chronic interferon response as well as extensive transcriptional changes in CD8+ T cells. J. Virol. 81, 3477–3486 (2007). 51. Sedaghat, A.R. et al. Chronic CD4+ T cell activation and depletion in human immunodeficiency virus type I infection: type I interferon-mediated disruption of T cell dynamics. J. Virol. 82, 1870–1883 (2008). 52. Tilton, J.C. et al. Human immunodeficiency virus viremia induces plasmacytoid dendritic cell activation in vivo and diminished a interferon production in vitro. J. Virol. 82, 3997–4006 (2008). 53. Bosinger, S.E. et al. Gene expression profiling of host response in models of acute HIV infection. J. Immunol. 173, 6858–6863 (2004). 54. Brenchley, J.M. et al. Microbial translocation is a cause of systemic immune activation in chronic HIV infection. Nat. Med. 12, 1365–1371 (2006). 55. Pandrea, I.V. et al. Acute loss of intestinal CD4+ T cells is not predictive of simian immunodeficiency virus virulence. J. Immunol. 179, 3035–3046 (2007). 56. Decalf, J. et al. Plasmacytoid dendritic cells initiate a complex chemokine and cytokine network and are a viable drug target in chronic HCV patients. J. Exp. Med. 204, 2423–2437 (2007). 57. Stewart, T.A. Neutralizing interferon a as a therapeutic approach to autoimmune diseases. Cytokine Growth Factor Rev. 14, 139–154 (2003). 58. Bochud, P.-Y.M.B., Telenti, A. & Calandra, T. Innate immunogenetics: a tool for exploring new frontiers of host defence. Lancet Infect. Dis. 7, 531–542 (2007). 59. Picker, L.J. et al. Insufficient production and tissue delivery of CD4+ memory T cells in rapidly progressive simian immunodeficiency virus infection. J. Exp. Med. 200, 1299–1314 (2004). 60. Rossio, J.L. et al. Inactivation of human immunodeficiency virus type 1 infectivity with preservation of conformational and functional integrity of virion surface proteins. J. Virol. 72, 7992–8001 (1998).

1087