Antigen Export Reduces Antigen Presentation and Limits T ...

Article

Antigen Export Reduces Antigen Presentation and
Limits T Cell Control of M. tuberculosis
Graphical Abstract

Highlights

d M. tuberculosis-infected host cells export bacterial antigens

by vesicular transport.

d Antigen export diverts antigens away from the MHC class II

pathway.

d Blocking antigen export improves CD4 T cell activation and

control of infection.

d Drugs that block antigen exportmay increase the efficacy of T

cell responses in TB.

Srivastava et al., 2016, Cell Host & Microbe 19, 44–54January 13, 2016 ª2016 Elsevier Inc.http://dx.doi.org/10.1016/j.chom.2015.12.003

Authors

Smita Srivastava, Patricia S. Grace,

Joel D. Ernst

[email protected] (S.S.),[email protected] (J.D.E.)

In Brief

Srivastava et al. report that M.

tuberculosis uses a vesicular transport

pathway to divert bacterial antigens from

the MHC class II pathway and out of

infected cells, minimizing CD4 T cell

activation and promoting bacterial

persistence. Drugs that block antigen

exportmay enhance the efficacy of T cells

in TB.

mailto:[email protected]


http://dx.doi.org/10.1016/j.chom.2015.12.003

http://crossmark.crossref.org/dialog/?doi=10.1016/j.chom.2015.12.003&domain=pdf

Cell Host & Microbe

Article

Antigen Export Reduces Antigen Presentationand Limits T Cell Control of M. tuberculosisSmita Srivastava,1,3,* Patricia S. Grace,1,3 and Joel D. Ernst1,2,3,*1Division of Infectious Diseases and Immunology, Department of Medicine2Department of Microbiology3Department of PathologyNew York University School of Medicine, New York, NY 10016, USA

*Correspondence: [email protected] (S.S.), [email protected] (J.D.E.)


SUMMARY

Persistence of Mycobacterium tuberculosis resultsfrom bacterial strategies that manipulate host adap-tive immune responses. Infected dendritic cells(DCs) transportM. tuberculosis to local lymph nodesbut activate CD4 T cells poorly, suggesting bacterialmanipulation of antigen presentation. However,M. tuberculosis antigens are also exported from in-fected DCs and taken up and presented by unin-fected DCs, possibly overcoming this blockade ofantigen presentation by infected cells. Here weshow that the first stage of this antigen transfer, anti-gen export, benefits M. tuberculosis by divertingbacterial proteins from the antigen presentationpathway. Kinesin-2 is required for antigen exportand depletion of this microtubule-based motor in-creases activation of antigen-specific CD4 T cellsby infected cells and improves control of intracellularinfection. Thus, although antigen transfer enablespresentation by bystander cells, it does not compen-sate for reduced antigen presentation by infectedcells and represents a bacterial strategy for CD4T cell evasion.

INTRODUCTION

Infection with Mycobacterium tuberculosis is a major global

health problem, due to its ease of transmission by the aerosol

route, the lack of an efficacious vaccine, and increasing emer-

gence of bacterial drug resistance (Philips and Ernst, 2012).

Even though the HIV pandemic has amplified the global problem

of tuberculosis (TB), most people with TB are immunocompe-

tent, indicating thatM. tuberculosis possesses effective mecha-

nisms for evasion of innate and adaptive immunity. Although

the onset of adaptive immune responses is delayed after

M. tuberculosis infection of humans (Poulsen, 1950) or mice

(Chackerian et al., 2002; Gallegos et al., 2008; Reiley et al.,

2008; Wolf et al., 2008), most infected individuals and experi-

mental animals develop antigen-specific CD4 and CD8 T cell

responses, and the resulting T cells have appropriate effector

functions as assessed by ex vivo restimulation (Ernst, 2012).

44 Cell Host & Microbe 19, 44–54, January 13, 2016 ª2016 Elsevier I

However, M. tuberculosis persists despite measurable T cell re-

sponses, suggesting that the bacteria manipulate the host to

prevent effector T cells from exerting their functions at the site

of infection (Urdahl et al., 2011).

SinceM. tuberculosis resides in macrophages and DCs in vivo

(Tailleux et al., 2003; Wolf et al., 2007), the suboptimal efficacy of

effector T cells may not be due to an intrinsic property of the an-

tigen-specific effector T cells themselves, but instead may be

secondary to bacterial manipulation of the infected antigen-pre-

senting cells. Indeed, multiple studies have reported that myco-

bacterial infection of antigen-presenting cells interferes with

MHC class II antigen presentation in vitro, although the mecha-

nisms that interfere with antigen presentation are poorly under-

stood (reviewed in Baena and Porcelli, 2009). Likewise, it is

unclear whether the phenomenon observed in vitro contributes

to suboptimal CD4 T cell efficacy in vivo (Ernst, 2012).

We recently reported that M. tuberculosis-infected host cells

export bacterial antigens for uptake and presentation by unin-

fected cells, and we and Pamer and colleagues found that anti-

gen transfer to uninfected DCs in the lymph node allows priming

of CD4 T cells in vivo (Samstein et al., 2013; Srivastava and Ernst,

2014), implying that this phenomenon helps to compensate for

impaired antigen presentation by M. tuberculosis-infected cells

(Hudrisier and Neyrolles, 2014). Because of the potential impor-

tance of antigen transfer for immunity to TB, we sought to further

understand its underlying mechanisms and significance. Here

we report the unexpected finding that the first stage of antigen

transfer, antigen export, benefits the pathogen by shunting bac-

terial antigens away from the MHC class II antigen presentation

pathway, thereby minimizing recognition of infected cells by an-

tigen-specific CD4 T cells. We also discovered that the second

stage of antigen transfer, antigen uptake and presentation by un-

infected cells, only partially compensates for the impaired anti-

gen presentation capacity of infected cells. Our results indicate

that antigen export is a potent and effective mechanism of im-

mune evasion that promotes persistenceM. tuberculosis despite

development of antigen-specific T cell responses.

RESULTS

Evidence for Antigen Export and Transfer to UninfectedLung Cells In VivoWe first confirmed and extended our finding thatM. tuberculosis

antigens can be acquired by uninfected cells in vivo in a process

of antigen export from infected cells followed by uptake and

nc.




http://crossmark.crossref.org/dialog/?doi=10.1016/j.chom.2015.12.003&domain=pdf

processing by uninfected bystander cells. In addition to the

earlier finding of antigen transfer in lymph nodes after aerosol

infection (Samstein et al., 2013; Srivastava and Ernst, 2014),

we found that antigen transfer occurs in the lungs. After infecting

mice with GFP-expressing M. tuberculosis, we isolated and

sorted live lung leukocytes on the basis of their phenotype and

presence (GFP+) or absence (GFP�) of intracellular bacteria

and used the sorted cells to stimulate M. tuberculosis Ag85B-

specific TCR transgenic (P25TCR-Tg) CD4 T cells. We found

that the distinct subsets of infected cells (lung DC, recruited

interstitial macrophages, andmonocytes) differed in their capac-

ity to activate CD4 T cells in vitro (Figure S1A). Notably, unin-

fected cells in each of the three subsets also activated CD4

T cells in this assay. Indeed, in all three subsets, the uninfected

cells were superior to the infected cells, indicating that they

had acquired bacterial antigen for processing and presentation

to CD4 T cells.

To assure that antigen acquisition by uninfected myeloid

cells in the lungs was not an artifact of cell isolation and sorting,

we stained frozen lung sections from mice that had been in-

fected with GFP-expressing M. tuberculosis. This demon-

strated that both infected (GFP+) and uninfected (GFP�)CD11c+ cells contain Ag85 and that the uninfected cells that

had acquired Ag85 were in proximity to infected cells, implying

that they had acquired antigen released from infected cells

(Figure S1B).

M. tuberculosis Ag85 Is Present in ExtraphagosomalVesicles in Infected CellsTo further understand the significance of antigen export, we first

sought to identify cellular mechanistic steps required for antigen

export from infected cells. Since our earlier studies indicated

that M. tuberculosis antigen export from infected cells did not

involve apoptosis or exosome shedding, and released unde-

graded bacterial proteins, we used immunofluorescence stain-

ing and confocal microscopy to localize M. tuberculosis antigen

85 (Ag85) in infected cells. Ag85, which consists of the closely

related proteins Ag85A, Ag85B, and Ag85C, is secreted by

M. tuberculosis and is a component of multiple TB vaccine can-

didates currently in development (Kaufmann et al., 2014). We

found Ag85 (principally Ag85B) staining in a punctate pattern

in DCs infected with wild-type, but not with Ag85B-null

M. tuberculosis (Figure 1A). Punctate staining of Ag85 was not

concentrated in the perinuclear region, and Ag85+ puncta

were present in the periphery of the infected cells. Since the

punctate pattern suggested that Ag85 is contained in vesicles,

we used immunogold staining and electron microscopy of cryo-

sectioned M. tuberculosis-infected bone-marrow-derived DCs

to further characterize the location of extraphagosomal

intracellular Ag85. This revealed Ag85 in bacteria-containing

phagosomes and in small membrane-bound vesicles devoid of

bacteria or electron-dense bacterial fragments (Figure 1B).

This finding resembles the reported location of Ag85 in

M. tuberculosis-infected human monocytes (Harth et al., 1996)

and murine macrophages (Beatty and Russell, 2000), deter-

mined using a different antibody to Ag85. While the fixation con-

ditions compatible with immunogold staining are not optimal for

preservation of membrane ultrastructure, the Ag85-containing

vesicles detected were bounded by a single bilayer (Figure 1B,

Cell H

inset). We refer to these vesicles as antigen export vesicles

(AEVs).

Phagosomes containing live mycobacteria, including

M. tuberculosis, have been reported to interact with recycling en-

dosomes containing transferrin and transferrin receptors (Clem-

ens and Horwitz, 1996; Sturgill-Koszycki et al., 1996). We there-

fore hypothesized that AEVs might be equivalent to recycling

endosomes that acquired bacterial proteins present from the

lumenal space of phagosomes, trafficked to the cell surface,

and released their cargo to the extracellular space. However,

immunofluorescence microscopy revealed lack of colocalization

of extraphagosomal Ag85 and transferrin receptor-bearing en-

dosomes (Figures 1C and 2D). Similar findings were obtained

when recycling endosomes were labeled using fluorescently

labeled transferrin (not shown). These findings indicate that

AEVs are not classical recycling endosomes that have inter-

sected with M. tuberculosis-containing phagosomes. They also

imply that an additional population of intracellular vesicles inter-

actswithmycobacterial phagosomes and is involved in transport

of bacterial products from phagosomes to the extracellular

space. We also found that most, but not all of the AEVs are

segregated from MHC class II-containing vesicles in infected

cells (Figure 1D), suggesting that a limited fraction of Ag85 is pre-

sent in theMHC-II pathway; AEVs also did not resemble early en-

dosomes, as they did not colocalize with EEA1 (not shown).

Export of M. tuberculosis Ag85B from Infected CellsInvolves the Motor Protein Kinesin-2Although the mechanisms of exocytosis and protein secretion

from macrophages and DCs are incompletely characterized,

microtubule-dependent vesicle transport promotes secretory

vesicle transport and exocytosis in diverse cell types, from neu-

rons to cytotoxic T lymphocytes (Kuhn and Poenie, 2002; Stinch-

combe et al., 2006; Vale et al., 1985). Immunofluorescence

confocal microscopy revealed a high frequency of localization

of AEVs with microtubules in M. tuberculosis-infected cells (Fig-

ures 2A and 2D). AEVs were closely aligned with microtubules,

including at the periphery of infected cells (Figure 2A). AEVs

(red) frequently colocalized with tubulin (green), as indicated by

yellow fluorescence; in addition, AEVs were frequently in close

proximity to microtubules. To determine the functional signifi-

cance of colocalization of AEVs and microtubules, we inhibited

tubulin polymerization using nocodazole and observed a dose-

dependent decrease in the extracellular release of Ag85 (Fig-

ure 2B). The maximum tolerated concentration of nocodazole

decreased Ag85B export by approximately 50%. These results

are consistent with microtubule-dependent vesicular transport

as an essential step in export of an M. tuberculosis antigen

from infected cells.

Microtubule-dependent organelle transport toward the cell

periphery depends on one or more members of the kinesin

family of motor proteins (reviewed in Hirokawa et al., 2009).

We examined kinesin-2, since it is implicated in transport of ve-

sicular organelles; its subunits, KIF3A, KIF3B, and KIF-associ-

ated protein 3 (KAP3 or KIFAP3) are expressed in macrophages

and myeloid DCs (Gautier et al., 2012; Miller et al., 2012),

including CD11bhi lung macrophages and CD103�CD11b+

lung DCs (http://www.immgen.org); and its KIF3A subunit has

been implicated in virus release from HIV-1-infected human

ost & Microbe 19, 44–54, January 13, 2016 ª2016 Elsevier Inc. 45

http://www.immgen.org

Figure 1. M. tuberculosis Ag85 Is Located in

Extraphagosomal Vesicles Distinct from

Classical Recycling Endosomes

(A) Confocal microscopy of DCs infected with

GFP-expressing M. tuberculosis (pseudo-colored

blue) and stained for surface CD45.2 (pseudo-

colored green), fixed, permeabilized, and stained

with an antibody to Ag85B (red). Top row: DCs

infected with wild-type M. tuberculosis H37Rv;

bottom row: DCs infected with H37Rv:D85B.

Although the antibody to Ag85B cross reacts with

Ag85A and Ag85C on immunoblots, the predomi-

nant signal detected by immunofluorescence is

from Ag85B. Scale bars: 10 mm.

(B) Immunoelectron microscopy of Ag85 in cry-

osections of DCs infected with wild-type

M. tuberculosis H37Rv. Ag85 staining associated

with bacteria (arrow heads) and membrane-bound

small vesicles (arrows with long tail; enlarged inset

image). Scale bar: 500 nm.

(C) DCs infected with M. tuberculosis H37Rv ex-

pressing cyan fluorescent protein (CFP) (blue) and

analyzed for transferrin receptor-containing re-

cycling endosomes. Transferrin receptor (green)

and Ag85B (red). Scale bars: 10 mm.

(D) DCs infected with M. tuberculosis H37Rv ex-

pressing CFP (blue), stained for MHC-II (green)

and Ag85B (red). Scale bars: 10 mm. Data in (A),

(C), and (D) represent two to four independent

staining experiments. (B) is representative of

20 images.

macrophages (Gaudin et al., 2012). Immunofluorescence

confocal microscopy revealed that AEVs colocalized with the

kinesin-2 component KIF3A in infected cells (Figure 2C). We

used automated image analysis to quantitate the fraction of

AEVs that colocalized with microtubules and/or with KIF3A in

100–300 infected cells and found that in the majority of infected

cells, >60% of AEVs colocalized with tubulin or KIF3A (Fig-

ure 2D), suggesting that microtubule associated kinesin-2 con-

tributes to trafficking of AEV. In contrast, less than 20% of

Ag85+ AEVs colocalized with transferrin receptor-containing

vesicles in any infected cell (Figure 2D).

Since members of the dynamin protein family mediate vesicle

budding and membrane scission in diverse intracellular pro-

cesses (Ferguson and De Camilli, 2012; Praefcke andMcMahon,

2004), we hypothesized that dynamin-2 (the form expressed in

myeloid cells) might contribute to generation and/or maturation

of AEV and found that in the majority of cells 40%–60% of

46 Cell Host & Microbe 19, 44–54, January 13, 2016 ª2016 Elsevier Inc.

Ag85+ AEVs colocalized with dynamin-2

(Figure 2D), consistent with a partial or

transient association of AEVs and dyna-

min 2.

Kinesin-2 Depletion Blocks AntigenExport from Infected CellsSince disruption of microtubule polymer-

ization decreased export of Ag85 from

infected cells, and AEVs colocalized

with microtubules and with KIF3A, we

examined the functional importance of

kinesin-2 in antigen export. Kinesin-2 typically functions as a

heterotrimer of KIF3A, KIF3B, and KIF3 adaptor protein

(KAP3 or KIFAP3) (Cole et al., 1993; Hirokawa et al., 2009),

so we used siRNAs targeting KIF3A and KAP3 to deplete kine-

sin-2. siRNA treatment of infected cells resulted in marked

depletion of both KIF3A and KAP3 as indicated by immuno-

blots of cell lysates (Figures 3A and S2A). Depletion of these

components of kinesin-2 decreased export of Ag85B from in-

fected cells by approximately 70%, as assayed by transferring

conditioned medium to naive uninfected DC, which were then

used to activate Th1-polarized Ag85B-specific TCR transgenic

(P25TCR-Tg) CD4 T cells (Figures 3B and S2B). Kinesin-2

depletion decreased the concentration of Ag85B protein in

conditioned medium by a similar amount as detected by immu-

noblotting, without a difference in the number of bacteria in ki-

nesin-2-depleted and control cells (Figures 3C, S2C, and S2D).

Kinesin-2 depletion also decreased export of at least three

Figure 2. Ag85B Export from Infected Cells

Depends on Microtubules

(A) AEVs colocalize with tubulin. DCs infected with

CFP-expressing M. tuberculosis H37Rv (blue)

were stained for tubulin (green) and Ag85B (red).

Scale bars: 10 mm.

(B) Inhibition of tubulin polymerization decreases

Ag85B export. Upper panel: immunoblot of Ag85

in conditioned medium (CM) collected from DCs

that had been infected and cultured in the pres-

ence of the designated concentrations of noco-

dazole for 20 hr. The same DCs that were used to

generate CM were lysed and probed for actin.

Lower panel: quantitation of Ag85B band intensity

by Image J of conditioned medium of

M. tuberculosis-infected cells cultured in the

absence or presence of 10 mM nocodazole,

determined using the immunoblot in the upper

panel.

(C) Colocalization of M. tuberculosis Ag85 and

kinesin-2 component KIF3A. DCs infected with

CFP-expressing M. tuberculosis (blue) were

stained for KIF3A (pseudo-colored green) and

Ag85 (red). Scale bars: 10 mm.

(D) Quantitative analysis of the fraction of AEVs

that colocalize with tubulin, KIF3A, dynamin 2, or

transferrin receptor (TfR) in a given cell. Confocal z

stack images were acquired and analyzed using

theManders’ overlap coefficient (Dunn et al., 2011)

analysis module of the Volocity image analysis

software. Data in (A)–(D) represent two to four in-

dependent experiments where 100–300 cells were

analyzed to obtain the Manders’ colocalization

coefficient. Data shown are mean ± SD. A similar

trend with respect to colocalization was confirmed

by analysis with the JACoP (Just another colocal-

ization plugin) of the ImageJ software (Costes

et al., 2004).

other M. tuberculosis antigenic proteins, ESAT-6, CFP-10, and

EsxH (Figures 3C, S2C, and S2D), indicating that the export

pathway of these proteins from M. tuberculosis-infected cells

resembles that of Ag85B. A similar decrease of exported bac-

terial antigens was observed when infected macrophages were

used instead of DC (Figures 3D and 3E). The finding that export

of M. tuberculosis antigens from infected cells depends on ki-

nesin-2 is consistent with the results that implicated an intact

microtubule cytoskeleton in antigen export. Moreover, the

finding that kinesin-2 is required for export of at least four

M. tuberculosis proteins, which are secreted by the bacteria us-

ing three distinct secretion systems (Ag85: SecA1; ESAT-6 and

CFP-10: the type VII secretion system, Esx-1; and EsxH: Esx-3)

indicates that AEVs export multiple bacterial proteins that are

targets of T cell recognition.

To further characterize AEVs and the process of antigen

export, we examined the functional role of dynamin 2, which

mediates budding and scission of diverse populations of intra-

Cell Host & Microbe 19, 44–5

cellular vesicles (Ferguson and De Ca-

milli, 2012), and since immunofluores-

cence microscopy revealed evidence

for association of AEVs and dynamin 2.

We first depleted dynamin 2 from DCs

using siRNA prior to infection, but this reduced uptake of the

bacteria, consistent with the previously reported role of dyna-

min 2 in phagocytosis (Gold et al., 1999). As an alternative, we

took advantage of the rapid kinetics of dynasore, a cell-perme-

able chemical inhibitor of dynamin GTPase activity (Macia

et al., 2006), and added it to DCs after infection with

M. tuberculosis. Dynamin inhibition reduced export of Ag85B

by �60%, as reflected by immunoblotting of CM and by acti-

vation of P25TCR-Tg Th1 cells after transfer of CM to fresh

DCs (Figure 3F, left panel). The decreased responses of

P25TCR-Tg cells to CM from dynasore-treated infected DCs

were due to decreased availability of antigen in CM and not

due to dynasore carried over into the T cell response assay,

since P25TCR-Tg cells were able to respond when recombi-

nant Ag85B was added to CM from dynasore-treated DCs

(Figure 3F; right panel). Together, the findings that dynamin 2

associates with AEVs and that dynamin inhibition blocks anti-

gen export imply a role for dynamin 2 in one or more vesicle

4, January 13, 2016 ª2016 Elsevier Inc. 47

Figure 3. Kinesin 2 Is Required for Antigen

Export from M. tuberculosis-Infected DCs

and Macrophages

(A) Efficacy of siRNA knockdown of kinesin-2

components KIF3A and KAP3 in DCs.

(B) Kinesin-2 depletion decreases export of Ag85B

antigenic activity from infected DCs. CM from in-

fected DCs treated with either a non-targeting

siRNA (Control) or siRNAs targeting KIF3A and

KAP3 (Kinesin-2-depleted) was transferred to

fresh uninfected DC, which were used to stimulate

Ag85B-specific P25TCR-Tg CD4 T cells. CD4

T cell activation was quantitated as IFNg secretion.

Data shown are mean ± SEM.

(C) Kinesin-2 depletion blocks export of Ag85B,

CFP-10, ESAT-6, and EsxH.

(D) Kinesin-2 (KIF3A andKAP3) depletion by siRNA

treatment of bone-marrow-derived macrophages

(MØ) decreases secretion of Ag85B protein. Upper

panels show efficacy of siRNA depletion of KIF3A

and KAP3; the lower panel shows Ag85B in CM

generated from macrophages from the same

experiment.

(E) Kinesin-2 depletion blocks export of Ag85B

antigenic activity from macrophages. Methods

were identical to those used in (B), using bone-

marrow-derived macrophages instead of DCs.


(F) Dynasore treatment of infected DCs decreases

Ag85B export to CM as assayed by immunoblot or

T cell activation assay (upper and lower left

panels); addition of recombinant Ag85B to CM

from dynasore-treated DCs restored T cell acti-

vation (right panel). Data are shown as mean ±

SEM and represent four to ten independent ex-

periments *p < 0.05; **p < 0.01; ***p < 0.001

(Student’s t test).

membrane budding/scission events in the antigen export

pathway.

Blocking Antigen Export Decreases CD4 T Cell Primingby Bystander Cells and Increases T Cell Priming byInfected Cells In VivoTo determine whether kinesin-2-dependent antigen export im-

pacts activation of Ag85B-specific CD4 T cells in vivo, we used

intratracheal transfer of M. tuberculosis-infected bone-marrow-

derived DC, followed by assay of P25TCR-Tg CD4 T cell prolifer-

ation in the mediastinal lymph node of recipient mice (Srivastava

and Ernst, 2014). Transfer ofM. tuberculosis-infected kinesin-2-

depleted MHC II�/� DCs (that are impaired in antigen export and

also cannot directly activate CD4 T cells) to naive MHC II+/+ re-

cipients resulted in less proliferation of antigen-specific CD4

T cells than observed when infected kinesin-2 replete

MHC II�/� DCs were transferred (Figures 4A–4D). These results

support our in vitro observations that kinesin-2 depletion in in-

fected DCs decreases export of bacterial antigens and de-

creases availability of antigen that can be acquired and pre-

sented by bystander cells.

To determine the impact of blocking antigen export on the abil-

ity of infected cells to directly present antigen and activate anti-

gen-specific CD4 T cells, we transferred M. tuberculosis-in-

fected kinesin-2-depleted or kinesin-2-replete MHC class II+/+


DCs into MHC class II�/� recipients and assayed proliferation

of adoptively transferred naive P25TCR-Tg CD4 T cells. This

yielded the unexpected result that kinesin-2-depleted DCs

(which are impaired in antigen export) stimulatedmore T cell pro-

liferation than did DCs that were treated with a control siRNA

(Figures 4E–4H); this was not due to differences in the number

of bacteria in the transferred DCs nor to enhanced migration of

kinesin-2-depleted DCs to the local lymph node (Figures S3A–

S3C). Since the only cells that can activate CD4 T cells under

these conditions are the infected, transferred DCs, this result in-

dicates that disrupting antigen export increases the antigen pre-

senting efficacy of M. tuberculosis-infected cells. These results

indicate that antigen export is a mechanism that decreases

MHC class II antigen presentation by M. tuberculosis-infected

cells.

Antigen Export Limits Overall CD4 T Cell ActivationIn VivoOur previously published finding that M. tuberculosis-infected

cells export antigen for uptake and presentation by uninfected

bystander cells implied that antigen transfer compensates for

impaired antigen presentation by infected cells (Srivastava

and Ernst, 2014). In the context of those results, our finding

that blocking antigen export could increase antigen presenta-

tion by infected cells raised the question whether antigen

nc.

Figure 4. Blockade of Antigen Export Decreases CD4 T Cell Priming by Bystander Cells and Increases Priming by Infected Cells In Vivo

(A–D) Kinesin-2 depletion in infected donor MHC-II�/� DCs blocks antigen transfer to resident MHC-II+/+ cells reflected as decreased CD4 T cell priming in vivo.


(A) Frequency of Ag85B-specific P25TCR-Tg CD4+ T cells in lymph nodes of MHC II+/+ recipients following transfer of infected MHC-II�/� control (left panel) or

kinesin-2-depleted (right panel) DC.

(B) Quantitation of P25TCR-Tg cells in lymph nodes of the groups of mice in (A).

(C) Proportion of P25TCR-Tg cells that underwent R 1 cycle of proliferation (CFSEdim) in lymph nodes of groups of mice shown in (A).

(D) Quantitation of P25TCR-Tg cells that underwent R 1 cycle of proliferation (CFSEdim) in lymph nodes of groups of mice in (C).

(E–H) Kinesin-2 depletion in infected donor MHC-II+/+ DCs increases their ability to directly prime antigen-specific CD4 T cells in vivo in MHC-II�/� mice.

(E) Frequency of P25TCR-Tg cells in lymph nodes of MHC II�/� recipients following transfer of infected MHC II+/+ control (left panel) or kinesin-2-depleted (right

panel) DCs.

(F) Quantitation of P25TCR-Tg cells in lymph nodes of groups of mice in (E).

(G) Proportion of P25TCR-Tg cells that underwent R 1 cycle of proliferation (CFSEdim) in lymph nodes of mice shown in (E).

(H) Quantitation of P25TCR-Tg cells that underwent R 1 cycle of proliferation (CFSEdim) in lymph nodes of groups of mice in (G). Data are mean± SEM of three

pools of mice (n = 6) per experimental group and represent two independent experiments. *p < 0.05; **p < 0.01; ***p < 0.001 (Student’s t test).

transfer to bystander cells fully compensates for impaired anti-

gen presentation by infected cells. To address this question,

we transferred wild-type M. tuberculosis-infected kinesin-2-

depleted (impaired for antigen transfer) or control (competent

for antigen transfer) DCs into wild-type mice and quantitated

proliferation of adoptively transferred naive P25TCR-Tg CD4

T cells. This revealed quantitatively greater proliferation of anti-

gen-specific CD4 T cells in recipients of kinesin-2-depleted

DCs compared with that in recipients of control DCs (Figures

5A–5D); this was not due to a difference in the number of

antigen-producing bacteria in the two groups of cells or a dif-

ference in their ability to migrate to the local lymph node (Fig-

ures S4A–S4D). These results indicate that blocking antigen

export enhances CD4 T cell activation to a greater extent

than can be accomplished by antigen export followed by anti-

gen uptake and presentation by bystander cells. These results

also indicate that antigen export is a quantitatively significant

mechanism that reduces MHC class II antigen presentation

by M. tuberculosis-infected cells.

Cell H

Blocking Antigen Export Enhances Activation ofAntigen-Specific Th1 Effector T Cells and T Cell-Dependent Antimycobacterial ActivityThe unexpected finding that blocking antigen transfer in vivo

enhances priming and proliferation of naive CD4 T cells sug-

gested that blocking antigen transfer might also enhance acti-

vation of CD4 effector T cells. When varying numbers of kine-

sin-2-depleted or control DCs infected with M. tuberculosis

were cultured with a constant number of Ag85B-specific Th1-

polarized P25TCR-Tg CD4 T cells, kinesin-2-depleted DCs

were superior to kinesin-2-replete DCs at all DC:T cell ratios

examined, with 3-fold greater T cell activation by kinesin-2-

depleted cells at the lowest DC:T cell ratio examined (Fig-

ure 6A). Enhanced activation of antigen-specific CD4 effector

cells was due to enhanced presentation of antigen produced

by intracellular bacteria and not due to increased surface

expression of MHC II or costimulatory molecules, differential

apoptosis, or cytokine secretion; depletion of kinesin-2 in

DCs did not affect activation of Ag85B-specific CD4 T cells


Figure 5. Blockade of Antigen Export from

Infected DCs Enhances Overall Antigen-

Specific CD4 T Cell Activation In Vivo

Comparison of Ag85B-specific CD4 T cell prolif-

eration in MHC II+/+ recipients following transfer of

infected kinesin-2-depleted or control MHC II+/+

DC.

(A) Frequency of Ag85B-specific P25TCR-Tg

CD4+ T cells in lymph nodes.

(B) Quantitation of total P25TCR-Tg CD4+ T cells in

lymph nodes of groups of mice shown in (A).

(C) Proportion of P25TCR-Tg cells that underwent

R 1 cycle of proliferation (CFSEdim) in lymph nodes

of groups of mice shown in (A).

(D) Quantitation of P25TCR-Tg cells that under-

went R 1 cycle of proliferation (CFSEdim) in lymph

nodes of groups of mice in (C). Data are mean±

SEM of four pools of mice (n = 8) per experimental

group and represent two independent experi-

ments. *p < 0.05; **p < 0.01; ***p < 0.001 (Student’s

t test).

when antigen was provided as recombinant Ag85B or as pep-

tide 25 (Figures S5A–S5E). At a constant ratio of DC:effector

T cells, enhanced activation of CD4 effector cells by kinesin-

2-depleted DCs depended on the dose of siRNA used; as little

as 25 nM siRNA caused a significant increase in T cell secretion

of IFNg (Figure 6B). Kinesin-2 depletion did not affect uptake of

M. tuberculosis, as reflected by the number of viable bacteria

present immediately following infection, prior to addition of

P25TCR-Tg Th1 cells (Figure S5F).

The evidence that blocking antigen export from

M. tuberculosis-infected cells caused increased CD4 effector

T cell activation suggested that blocking antigen export might

also enhance T-cell-dependent control of intracellular

M. tuberculosis infection. When Ag85B-specific P25TCR-Tg

Th1 cells were cocultured with M. tuberculosis-infected bone

marrow DCs that were depleted of kinesin-2 by treatment with

KIF3A and KAP3 siRNAs, there was an siRNA dose-dependent

decrease in the number of viable bacteria recovered after

20 hr, compared with that in export-competent control cells (Fig-

ure 6C). The decreased recovery of viable bacteria was not a

consequence of decreased DCs viability or bacterial uptake (Fig-

ure S5F), indicating that inhibition of antigen export not only

increased activation of antigen-specific CD4 T cells but also

improved T-cell-dependent control of intracellular infection.

The apparent threshold for improved bacterial control was higher

than that required for increased CD4 T cell activation: treatment

of DCs with 25 nM siRNA was sufficient to increase activation of

antigen-specific CD4 T cells, while 50 nM was required to

enhance control of intracellular M. tuberculosis. In these experi-

ments, CD4 T cell activation and bacterial control required both

MHC class II and specific antigen, as MHC-II�/� DCs infected

50 Cell Host & Microbe 19, 44–54, January 13, 2016 ª2016 Elsevier Inc.

with wild-type M. tuberculosis (H37Rv)

and MHC-II+/+ DCs infected with Ag85B-

null M. tuberculosis (H37Rv:D85B) did

not activate P25TCR-Tg Th1 cells and

did not enhance control of bacterial

growth (Figures 6D, 6E, and S5G). These

results indicate that antigen export fromM. tuberculosis-infected

cells contributes to survival of the bacteria in the presence of an-

tigen-specific Th1 cells.

DISCUSSION

The studies reported here reveal that DCs and macrophages in-

fected with Mycobacterium tuberculosis divert mycobacterial

antigens away from the MHC class II antigen presentation

pathway and to the extracellular space, and by doing so reduce

their activation of CD4 T cells, leading to improved survival of the

intracellular bacteria. Althoughwe and others have previously re-

ported that antigen transfer from infected to uninfected DCs after

aerosol infection provides a mechanism to compensate for

impaired antigen presentation by infected cells (Samstein

et al., 2013; Srivastava and Ernst, 2014), our present results pro-

vide twomajor insights. They reveal that the first stage of antigen

transfer, antigen export from infected cells, clearly benefits the

pathogen, by reducing antigen presentation and CD4 T cell acti-

vation by infected cells. They also reveal that the second stage,

antigen uptake and presentation by uninfected cells, provides

incomplete compensation for the decreased antigen presenta-

tion by infected cells, indicating that the net effect of antigen

export is to diminish activation of antigen-specific CD4 T cells

at the site of infection.

Antigen export is a quantitatively important route for trafficking

of mycobacterial antigens produced by intraphagosomal bacte-

ria, as indicated by the 3-fold increase in activation of Ag85B-

specific CD4 T cells when antigen export was blocked by

depletion of kinesin-2 (Figure 6A). This implies that more antigen

enters the export pathway than enters the MHC class II pathway

Figure 6. Blockade of Antigen Export by Ki-

nesin 2 Depletion Increases Activation of

Ag85B-Specific CD4 T Cells and Enhances

Control of Intracellular M. tuberculosis

(A) Kinesin-2 depletion in DCs increases their

ability to activate P25TCR-Tg Th1 CD4 effector

T cells (IFNg secretion) at a range of DC: T cell

ratios.

(B) Increased activation of P25TCR-Tg Th1 CD4

effector T cells by kinesin-2-depleted cells is

siRNA dose dependent.

(C) Increased activation of P25TCR-Tg Th1 CD4

effector T cells by kinesin-2-depleted DCs results

in enhanced control of intracellular bacteria.

(D) Increased activation of P25TCR-Tg Th1 CD4

effector T cells by kinesin-2-depleted DCs de-

pends on MHC class II and bacterial antigen.

(E) Enhancedcontrol of intracellularM. tuberculosis

in kinesin 2-depleted DCs by P25TCR-Tg Th1 CD4

effector T cells requires MHC II and bacterial

antigen. Data are shown as mean ± SEM and

represent two to four independent experiments.

*p < 0.05; **p < 0.01; ***p < 0.001 (Student’s

t test).

in M. tuberculosis-infected cells. Diverting antigen to the export

pathway has functional significance for survival of the bacteria,

as blockade of antigen export also improves CD4 T-cell-depen-

dent control of intracellular infection. This finding is especially

relevant in light of our previous results using mixed (MHC II�/�

and MCH II+/+) bone marrow chimeric mice that revealed that

CD4 T cells must directly recognize infected lung DCs and mac-

rophages to control intracellular infection in vivo (Srivastava and

Ernst, 2013). By exporting bacterial antigens,M. tuberculosis-in-

fected cells reduce the availability of antigen for presentation on

MHC class II, diminishing their ability to be recognized by and to

activate CD4 effector cells, thereby promoting bacterial persis-

tence. A limitation of the present work is that we have not yet

quantitated the impact of antigen export on bacterial persistence

in vivo; additional experiments are needed to reveal this impor-

tant information.

In this and our previous report (Srivastava and Ernst, 2014), we

found that all of the M. tuberculosis antigens that we examined

are exported from infected cells and share a kinesin-2-depen-

dent export pathway. While this establishes that antigen export

is not unique to Ag85B, the antigen we studied in most detail,

it raises the question whether all mycobacterial antigens are di-

verted from the MHC class II pathway and exported to a similar

extent. One common property of the antigens we have examined

so far is that they are secreted by the bacteria. These proteins

may accumulate to especially high concentrations in the intra-

Cell Host & Microbe 19, 44–5

phagosomal space and therefore might

be particularly prone to export by intracel-

lular vesicles that bud from the phago-

some membrane and enter the export

pathway. Further studies of the intracel-

lular trafficking of antigens that are not

secreted by the bacteria are warranted,

as our findings imply that antigens that

are not exported from infected cells might

induce more productive CD4 T cell activation and thus might be

especially useful TB vaccine antigens.

Our finding that antigen export provides a mechanism for

evasion of CD4 T cell activation by infected cells may also help

explain the observation that few, if any, human T cell epitopes

ofM. tuberculosis exhibit evidence of escape mutations (Comas

et al., 2010; Copin et al., 2014; Coscolla et al., 2015). The effi-

ciency of antigen export and evidence that it limits activation of

CD4 T cells by infected cells implies that it may reduce selection

pressure from T cell recognition so that epitope escape muta-

tions are neither selected nor necessary for persistence and pro-

liferation of M. tuberculosis.

The results reported here provide initial insight into key mech-

anistic elements of antigen export. We found that Ag85B is pre-

sent in extraphagosomal membrane vesicles, that one or more

steps in antigen export depends on the GTPase activity of dy-

namin-2, and that AEVs are associated with microtubules and

the motor protein, kinesin-2. We also determined that disruption

of microtubules or depletion of kinesin-2 decreases antigen

export. Notably, we found that decreasing antigen export by ki-

nesin-2 depletion improved the ability of M. tuberculosis-in-

fected cells to activate Ag85B-specific CD4 T cells and the abil-

ity of CD4 T cells to control infection. These findings provide a

strong rationale for efforts to identify additional mechanistic

steps of antigen export and their essential mediators, including

mechanisms for AEVs budding from phagosomes, AEVs

4, January 13, 2016 ª2016 Elsevier Inc. 51

targeting through a nondegradative pathway, and AEVs mem-

brane fusion with the plasma membrane for extracellular deliv-

ery of bacterial antigens. Such efforts will also clarify the identity

of AEVs and their relationship to previously characterized vesic-

ular compartments involved in intracellular protein transport,

and will provide insight into whether AEVs are an existing

vesicle population that is co-opted by M. tuberculosis, or

whether AEVs are generated in response to infection. Identifica-

tion of specific steps and mediators in the antigen export

pathway may guide pharmacological intervention to block anti-

gen export and enhance the efficacy of CD4 T cell responses in

control of TB. As the interactions of mycobacterial phagosomes

with intracellular vesicle networks have been thought to be

limited, the further characterization of AEVs will also shed light

on a fundamental interaction of intracellular M. tuberculosis

with host cells.

Another important matter is whether antigen export or a

similar process contributes to the pathogenesis of other infec-

tions. One related phenomenon is the export of typhoid toxin

from epithelial cells infected with the agent of typhoid fever,

Salmonella enterica serovar Typhi (S. Typhi) (Spano and Galan,

2008). Like M. tuberculosis, S. Typhi resides in vesicles and ex-

ports a heterotrimeric toxin from infected cells to intoxicate

neighboring uninfected cells. Typhoid toxin appears as intracel-

lular puncta in infected cells; upon release to the extracellular

media, typhoid toxin is neutralized by an antibody, indicating

that it is no longer contained in a membrane-bound compart-

ment (Spano and Galan, 2008). Export of typhoid toxin requires

assembly of the trimeric complex of CdtB, PltA, and PltB, indi-

cating that structural features of PltA and/or PltB are required

to target the complex to export vesicles. Our finding that mul-

tiple structurally unrelated M. tuberculosis antigens are ex-

ported suggests that the export pathway in M. tuberculosis-in-

fected cells does not require specific structural features for a

protein to enter and to be transported. Together, the findings

with typhoid toxin and M. tuberculosis antigen export provide

evidence for a mechanism beyond cytokine secretion, bacterial

membrane vesicle shedding (Rath et al., 2013), exosome shed-

ding (Bhatnagar et al., 2007), and apoptotic vesicles (Divangahi

et al., 2010; Winau et al., 2006) for pathogen-infected cells to

communicate with and affect the function of other cells, by

release of soluble pathogen proteins through an intracellular

nondegradative pathway. Further characterization of the anti-

gen export pathway of M. tuberculosis-infected cells will pro-

vide opportunities and targets to enhance the efficacy of natu-

ral and vaccine-induced T cells in TB and may also provide

means of intervening in the pathogenesis of other infectious

diseases.

EXPERIMENTAL PROCEDURES

Mice

Mice were bred and housed in specific pathogen-free facilities. Intratracheal

administration of infected bone-marrow-derived DCs was done as we

recently described (Srivastava and Ernst, 2014) followed by harvest and

processing of MDLN and lungs into single cell suspensions. The staining

and quantitation of antigen-specific CD4 T cell proliferation and bacteria

in tissues was performed as previously described (Wolf et al., 2008; Wolf

et al., 2007). All procedures were approved by the NYU School of Medicine

IACUC.


Bone-Marrow-Derived DCs and Macrophages

CD11c+ DCs were sorted following culture of fresh or cryopreserved bone

marrow in bone-marrow-derived DC (BMDC) medium (RPMI 1640 with 10%

heat-inactivated FBS, 2 mM L-glutamine, 1 mM sodium pyruvate, 13 b-ME,

10 mM HEPES) containing 10 ng/ml recombinant murine GM-CSF for

7 days. For siRNA treatments, CD11c+ DCs were sorted on day 5. Macro-

phages were generated by culture of bone marrow cells in bone-marrow-

derived macrophage (BMDM) medium (DMEM with 10% heat-inactivated

FBS, 2 mM L-glutamine, 1 mM sodium pyruvate, 13 b-ME, 10 mM HEPES,

and 20 ng/ml recombinant murine M-CSF from Peprotech) for 6 days prior

to use in the designated assays.

Immunofluorescence and Confocal microscopy

BMDCs on coverslips were stained with biotinylated anti-CD45.2 (Biolegend)

followed by streptavidin Alexa-647 (Invitrogen) and fixed overnight in PFA.

Fixed cells were permeabilized in cold PBS/0.5% Triton X-100 (5 min on

ice), washed, and stained for Ag85, and transferrin receptor, tubulin, KIF3A,

dynamin 2, or MHC-II (see Supplemental Experimental Procedures). Finally,

coverslips were washed and mounted in Vectashield mounting medium

(Vector Labs). An automated TCS SP5 II confocal microscope (Leica) with a

633/1.4 oil immersion lens was used to acquire 3D image stacks (0.3–

0.5 mm) using appropriate laser combinations. Representative image stacks

of average intensity from relevant channels were merged and converted to

RGB format using Image J. All related groups were subjected to similar color

settings for each color channel.

Quantitation of Colocalization

Colocalization analysis was performed on raw images in Volocity image anal-

ysis software using the built-in Manders’ overlap coefficient colocalization

feature. Multiple confocal images were subjected to background subtraction

followed by colocalization analysis on a per-cell basis. The data was presented

as Manders’ colocalization coefficient of Ag85 with respect to other proteins;

that is, the fraction of Ag85 that colocalized with the other protein of interest in

each analyzed cell (Dunn et al., 2011). The data represent two to three inde-

pendent experiments and 100–300 cells per experiment.

Cryo-Immunoelectron Microscopy

BMDC infected with wild-type M. tuberculosis H37Rv or Ag85B-deficient

M. tuberculosis (Wolf et al., 2008) were washed and fixed (PBS/2% PFA/

0.2% glutaraldehyde) as pellets that were then processed for cryo-immunoe-

lectron microscopy as described in the Supplemental Experimental

Procedures.

siRNA

ON-Target plus SMART pool Kif3a siRNA (L-042111-01), Kifap3 (KAP3) siRNA

(L-047278-01), and ON-Target plus non-targeting control siRNA (D-001810-

01) were obtained from GE Dharmacon (details in Table S1).

Generation of Conditioned Media and In Vitro Antigen Transfer

Assay

BMDMs or CD11c+ DCs were treated with control or targeting siRNAs begin-

ning on day 5 of culture. After 48 hr of exposure to siRNAs, the cells were

infected overnight with M. tuberculosis (see Supplemental Experimental Pro-

cedures for details). After overnight infection, cells were washed, treated

with amikacin, washed again, and cultured further in fresh BMDC or BMDM

medium that was harvested after 20–24 hr as conditioned medium (CM). The

CM was sterile filtered and added to uninfected BMDC in the presence of

P25TCR-Tg CD4+ Th1 effector cells for 60 hr, as described previously (Srivas-

tava and Ernst, 2014). Culture supernatants were collected at 60 hr and used

for IFNg quantitation by ELISA (BD Biosciences).

In Vitro T Cell Activation by Infected DCs and Bacterial Quantitation

CD11c+ bone-marrow-derived DCs were cultured, treated with siRNAs, and

infected withM. tuberculosis as in the preceding section. After amikacin treat-

ment andwashing, the DCswere cultured further in fresh BMDCmedium in the

absence or presence of Th1 polarized P25-TCRTg CD4 T cells for 20–24 hr.

After 20–24 hr, plates were spun to pellet cells, and the supernatants were

aspirated carefully, sterile filtered, and assayed for IFNg by ELISA (BD

nc.

Biosciences). The cells in the wells were lysed in PBS-0.5% Tween 80 at a cell

density of 2 3 105/ml and plated on 7H11 agar for bacterial quantitation (see

Supplemental Experimental Procedures for details).

In Vivo T Cell Activation following Intratracheal Transfer of Infected,

siRNA-Treated DC

CD11c+ bone-marrow-derived DCs were cultured and treated with siRNAs as

for in vitro experiments, with the variation that siRNAs were used at a final con-

centration of 80 nM and siRNA treatment was for 3 days prior to overnight

infection withM. tuberculosis (MOI, 1). After washing and amikacin treatment,

the siRNA-treated infected DCs were administered intratracheally (3–5 3 105

DC/mouse) 1 day after adoptive transfer of CFSE-labeled naive P25TCR-Tg

CD4 cells (1–3 3 106/mouse). To assay the effect of blocking antigen export

by transferred DCs in the absence of direct T cell priming, MHC-II�/� DCs

were treated with relevant siRNAs, infected, and administered intratracheally

toMHC-II+/+ mice. To assay the effect of antigen export on direct T cell priming

by transferred DCs in the absence of antigen presentation by resident cells,

MHC-II+/+ DCs were treated with relevant siRNAs, infected, and administered

intratracheally to MHC-II�/� mice. To assay the net effect of antigen export

when antigen-specific CD4 T cells could be activated by direct antigen presen-

tation by infected cells and by indirect activation after exported antigen is

taken up and processed by resident uninfected cells, MHC-II+/+ DCs were

treated with relevant siRNAs, infected, and administered intratracheally to

MHC-II+/+ mice. In all experiments, MDLN were harvested and processed

60 hr after BMDC transfer for bacterial quantitation and flow cytometry while

lungs were processed for bacterial quantitation only.

SUPPLEMENTAL INFORMATION

Supplemental Information includes five figures, one table, and Supplemental

Experimental Procedures and can be found with this article online at http://

dx.doi.org/10.1016/j.chom.2015.12.003.

AUTHOR CONTRIBUTIONS

S.S., P.S.G., and J.D.E. planned the experiments; S.S. and P.S.G. performed

the experiments; S.S., P.S.G., and J.D.E. interpreted the results and wrote the

manuscript.

ACKNOWLEDGMENTS

We thank the NYU imaging core facility (Feng-Xia [Alice] Liang for cryo-immu-

noelectron microscopy and Yan Deng for assistance with confocal micro-

scopy) (both supported by NCRR S10 RR024708). We greatly appreciate early

guidance by Sergio Grinstein, Ph.D. and are grateful to members of the Ernst

Lab for helpful discussions and suggestions on early drafts of the manuscript.

We also thank Jennifer Philips, M.D., Ph.D. and members of her laboratory for

sharing their siRNA expertise with primary cells.

Supported by grants from the NIH (AI 051242 and AI 084041). The authors

have no conflicts of interest relevant to this work.

Received: August 8, 2015

Revised: October 4, 2015

Accepted: December 15, 2015

Published: January 13, 2016

REFERENCES

Baena, A., and Porcelli, S.A. (2009). Evasion and subversion of antigen presen-

tation by Mycobacterium tuberculosis. Tissue Antigens 74, 189–204.

Beatty, W.L., and Russell, D.G. (2000). Identification of mycobacterial surface

proteins released into subcellular compartments of infected macrophages.

Infect. Immun. 68, 6997–7002.

Bhatnagar, S., Shinagawa, K., Castellino, F.J., and Schorey, J.S. (2007).

Exosomes released from macrophages infected with intracellular pathogens

stimulate a proinflammatory response in vitro and in vivo. Blood 110, 3234–

3244.

Cell H

Chackerian, A.A., Alt, J.M., Perera, T.V., Dascher, C.C., and Behar, S.M.

(2002). Dissemination of Mycobacterium tuberculosis is influenced by host

factors and precedes the initiation of T-cell immunity. Infect. Immun. 70,

4501–4509.

Clemens, D.L., and Horwitz, M.A. (1996). The Mycobacterium tuberculosis

phagosome interacts with early endosomes and is accessible to exogenously

administered transferrin. J. Exp. Med. 184, 1349–1355.

Cole, D.G., Chinn, S.W.,Wedaman, K.P., Hall, K., Vuong, T., and Scholey, J.M.

(1993). Novel heterotrimeric kinesin-related protein purified from sea urchin

eggs. Nature 366, 268–270.

Comas, I., Chakravartti, J., Small, P.M., Galagan, J., Niemann, S., Kremer, K.,

Ernst, J.D., and Gagneux, S. (2010). Human T cell epitopes of Mycobacterium

tuberculosis are evolutionarily hyperconserved. Nat. Genet. 42, 498–503.

Copin, R., Coscolla, M., Seiffert, S.N., Bothamley, G., Sutherland, J., Mbayo,

G., Gagneux, S., and Ernst, J.D. (2014). Sequence diversity in the pe_pgrs

genes of Mycobacterium tuberculosis is independent of human T cell recogni-

tion. MBio 5, http://dx.doi.org/10.1128/mBio.00960-13.

Coscolla, M., Copin, R., Sutherland, J., Gehre, F., de Jong, B., Owolabi, O.,

Mbayo, G., Giardina, F., Ernst, J.D., and Gagneux, S. (2015). M. tuberculosis

T Cell Epitope Analysis Reveals Paucity of Antigenic Variation and Identifies

Rare Variable TB Antigens. Cell Host Microbe 18, 538–548.

Costes, S.V., Daelemans, D., Cho, E.H., Dobbin, Z., Pavlakis, G., and Lockett,

S. (2004). Automatic and quantitative measurement of protein-protein colocal-

ization in live cells. Biophys. J. 86, 3993–4003.

Divangahi, M., Desjardins, D., Nunes-Alves, C., Remold, H.G., andBehar, S.M.

(2010). Eicosanoid pathways regulate adaptive immunity to Mycobacterium

tuberculosis. Nat. Immunol. 11, 751–758.

Dunn, K.W., Kamocka, M.M., and McDonald, J.H. (2011). A practical guide to

evaluating colocalization in biological microscopy. Am. J. Physiol. Cell Physiol.

300, C723–C742.

Ernst, J.D. (2012). The immunological life cycle of tuberculosis. Nat. Rev.

Immunol. 12, 581–591.

Ferguson, S.M., and De Camilli, P. (2012). Dynamin, a membrane-remodelling

GTPase. Nat. Rev. Mol. Cell Biol. 13, 75–88.

Gallegos, A.M., Pamer, E.G., and Glickman, M.S. (2008). Delayed protection

by ESAT-6-specific effector CD4+ T cells after airborne M. tuberculosis infec-

tion. J. Exp. Med. 205, 2359–2368.

Gaudin, R., de Alencar, B.C., Jouve,M., Berre, S., Le Bouder, E., Schindler, M.,

Varthaman, A., Gobert, F.X., and Benaroch, P. (2012). Critical role for the kine-

sin KIF3A in the HIV life cycle in primary humanmacrophages. J. Cell Biol. 199,

467–479.

Gautier, E.L., Shay, T., Miller, J., Greter, M., Jakubzick, C., Ivanov, S., Helft, J.,

Chow, A., Elpek, K.G., Gordonov, S., et al.; Immunological Genome

Consortium (2012). Gene-expression profiles and transcriptional regulatory

pathways that underlie the identity and diversity of mouse tissue macro-

phages. Nat. Immunol. 13, 1118–1128.

Gold, E.S., Underhill, D.M., Morrissette, N.S., Guo, J., McNiven, M.A., and

Aderem, A. (1999). Dynamin 2 is required for phagocytosis in macrophages.

J. Exp. Med. 190, 1849–1856.

Harth, G., Lee, B.Y., Wang, J., Clemens, D.L., and Horwitz, M.A. (1996). Novel

insights into the genetics, biochemistry, and immunocytochemistry of the 30-

kilodalton major extracellular protein of Mycobacterium tuberculosis. Infect.

Immun. 64, 3038–3047.

Hirokawa, N., Noda, Y., Tanaka, Y., and Niwa, S. (2009). Kinesin superfamily

motor proteins and intracellular transport. Nat. Rev. Mol. Cell Biol. 10,

682–696.

Hudrisier, D., and Neyrolles, O. (2014). Antigen smuggling in tuberculosis. Cell

Host Microbe 15, 657–659.

Kaufmann, S.H., Lange, C., Rao, M., Balaji, K.N., Lotze, M., Schito, M., Zumla,

A.I., and Maeurer, M. (2014). Progress in tuberculosis vaccine development

and host-directed therapies–a state of the art review. Lancet Respir. Med. 2,

301–320.

Kuhn, J.R., and Poenie, M. (2002). Dynamic polarization of the microtubule

cytoskeleton during CTL-mediated killing. Immunity 16, 111–121.




http://refhub.elsevier.com/S1931-3128(15)00495-3/sref1






















http://dx.doi.org/10.1128/mBio.00960-13
















































Macia, E., Ehrlich, M., Massol, R., Boucrot, E., Brunner, C., and Kirchhausen,

T. (2006). Dynasore, a cell-permeable inhibitor of dynamin. Dev. Cell 10,

839–850.

Miller, J.C., Brown, B.D., Shay, T., Gautier, E.L., Jojic, V., Cohain, A., Pandey,

G., Leboeuf, M., Elpek, K.G., Helft, J., et al.; Immunological Genome

Consortium (2012). Deciphering the transcriptional network of the dendritic

cell lineage. Nat. Immunol. 13, 888–899.

Philips, J.A., and Ernst, J.D. (2012). Tuberculosis pathogenesis and immunity.

Annu. Rev. Pathol. 7, 353–384.

Poulsen, A. (1950). Some clinical features of tuberculosis. 1. Incubation period.

Acta Tuberc. Pneumol. Scand. 24, 311–346.

Praefcke, G.J., andMcMahon, H.T. (2004). The dynamin superfamily: universal

membrane tubulation and fission molecules? Nat. Rev. Mol. Cell Biol. 5,

133–147.

Rath, P., Huang, C., Wang, T., Wang, T., Li, H., Prados-Rosales, R., Elemento,

O., Casadevall, A., and Nathan, C.F. (2013). Genetic regulation of vesiculogen-

esis and immunomodulation in Mycobacterium tuberculosis. Proc. Natl. Acad.

Sci. USA 110, E4790–E4797.

Reiley, W.W., Calayag, M.D., Wittmer, S.T., Huntington, J.L., Pearl, J.E.,

Fountain, J.J., Martino, C.A., Roberts, A.D., Cooper, A.M., Winslow, G.M.,

and Woodland, D.L. (2008). ESAT-6-specific CD4 T cell responses to aerosol

Mycobacterium tuberculosis infection are initiated in the mediastinal lymph

nodes. Proc. Natl. Acad. Sci. USA 105, 10961–10966.

Samstein, M., Schreiber, H.A., Leiner, I.M., Susac, B., Glickman, M.S., and

Pamer, E.G. (2013). Essential yet limited role for CCR2+ inflammatory mono-

cytes during Mycobacterium tuberculosis-specific T cell priming. eLife 2,

e01086.

Spano, S., and Galan, J.E. (2008). A novel pathway for exotoxin delivery by an

intracellular pathogen. Curr. Opin. Microbiol. 11, 15–20.

Srivastava, S., and Ernst, J.D. (2013). Cutting edge: Direct recognition of in-

fected cells by CD4 T cells is required for control of intracellular

Mycobacterium tuberculosis in vivo. J. Immunol. 191, 1016–1020.


Srivastava, S., and Ernst, J.D. (2014). Cell-to-cell transfer of M. tuberculosis

antigens optimizes CD4 T cell priming. Cell Host Microbe 15, 741–752.

Stinchcombe, J.C., Majorovits, E., Bossi, G., Fuller, S., and Griffiths, G.M.

(2006). Centrosome polarization delivers secretory granules to the immunolog-

ical synapse. Nature 443, 462–465.

Sturgill-Koszycki, S., Schaible, U.E., and Russell, D.G. (1996).

Mycobacterium-containing phagosomes are accessible to early endosomes

and reflect a transitional state in normal phagosome biogenesis. EMBO J.

15, 6960–6968.

Tailleux, L., Schwartz, O., Herrmann, J.L., Pivert, E., Jackson, M., Amara, A.,

Legres, L., Dreher, D., Nicod, L.P., Gluckman, J.C., et al. (2003). DC-SIGN is

the major Mycobacterium tuberculosis receptor on human dendritic cells.

J. Exp. Med. 197, 121–127.

Urdahl, K.B., Shafiani, S., and Ernst, J.D. (2011). Initiation and regulation of T-

cell responses in tuberculosis. Mucosal Immunol. 4, 288–293.

Vale, R.D., Schnapp, B.J., Mitchison, T., Steuer, E., Reese, T.S., and Sheetz,

M.P. (1985). Different axoplasmic proteins generate movement in opposite di-

rections along microtubules in vitro. Cell 43, 623–632.

Winau, F., Weber, S., Sad, S., de Diego, J., Hoops, S.L., Breiden, B., Sandhoff,

K., Brinkmann, V., Kaufmann, S.H., and Schaible, U.E. (2006). Apoptotic ves-

icles crossprime CD8 T cells and protect against tuberculosis. Immunity 24,

105–117.

Wolf, A.J., Linas, B., Trevejo-Nunez, G.J., Kincaid, E., Tamura, T., Takatsu, K.,

and Ernst, J.D. (2007). Mycobacterium tuberculosis infects dendritic cells with

high frequency and impairs their function in vivo. J. Immunol. 179, 2509–2519.

Wolf, A.J., Desvignes, L., Linas, B., Banaiee, N., Tamura, T., Takatsu, K., and

Ernst, J.D. (2008). Initiation of the adaptive immune response to

Mycobacterium tuberculosis depends on antigen production in the local lymph

node, not the lungs. J. Exp. Med. 205, 105–115.

nc.































































Cell Host & Microbe, Volume 19

Supplemental Information

Antigen Export Reduces Antigen Presentation

and Limits T Cell Control of M. tuberculosis Smita Srivastava, Patricia S. Grace, and Joel D. Ernst

Supplemental Figure 1

Figure S1 related to Figure 1. Ag85B is acquired by uninfected lung cells in vivo. (A) Left panel: CD11c/CD11b-defined subsets (mDC: myeloid dendritic cells; RM: recruited macrophages; Mono: monocytes) were sorted by FACS from GFP-positive (i.e, infected) and GFP-negative (i.e., uninfected) lung cells obtained from mice infected with GFP-expressing M. tuberculosis H37Rv 27 days earlier and subsequently used to stimulate P25TCR-Tg Th1 CD4 cells without addition of exogenous antigen. Four days later supernatants were collected and used to quantitate IFNγ by ELISA. The GFP-negative (i.e., uninfected) cell subsets contained fewer than 1 bacterial colony-forming unit (CFU) per 1000 sorted cells. The bar marked "none" indicates IFNγ secretion by P25TCR-Tg Th1 cells incubated in parallel in the absence of any antigen-presenting cells. In the same experiment, separate wells of the sorted uninfected cells received added peptide 25 at the same time as P25TCR-Tg Th1 cells, which stimulated release of increased amounts of IFNγ (mDC: 18,935 ± 740; RM: 8002 ± 2453; Monos: 6334 ± 2813; all as pg/ml, mean ± SD). Right panel: As a control, lung myeloid cell subsets were sorted from uninfected mice that received LPS intratracheally 3 days before harvest and sorting. After cells from each sorted subset were incubated with P25TCR-Tg Th1 cells, IFNγ secretion was quantitated by ELISA. In the same experiment, separate wells of the sorted uninfected cells received added peptide 25 at the same time as P25TCR-Tg Th1 cells, which stimulated release of increased amounts of IFNγ (mDC: 11,954 ± 1402; RM: 3.547 ± 72; Monos: 1,126 ± 90; all as pg/ml, mean ± SD). (B) Lungs from mice infected with GFP-expressing M. tuberculosis H37Rv (green; left panel) or H37Rv:ΔAg85B (green; right panel) were frozen in

A GFP-positive (infected)

GFP-negative (uninfected)

0

50

100

150

200

250

300

350

APC: mDC RM Mono none

**

**

***

IFNγ

(p

g/m

l)

B

11c

H37Rv

Ag85

11c

Ag85

H37Rv

:Δ85B

H37Rv H37Rv:Δ85B

mDC RM Mono0

10

20

30

IFN!

(pg/

ml)

Cells isolated from uninfected mice

OCT at day 24 post infection, sectioned, fixed and stained with anti-CD11c (blue) and anti-Ag85B (red). White arrows in the left panel point to bacteria-free CD11c+ DCs that contain Ag85. Scale bars 20 µm. Data represent 2-3 independent experiments *p < 0.05; **p < 0.01; ***p < 0.001 (Student’s t-test). Residual staining of Ag85 seen in the section from the mouse infected with H37Rv:ΔAg85B is attributable to recognition of Ag85A and Ag85C, which are recognized by the antibody and which are exported from infected cells (Figures 1A and 3C, and Srivastava and Ernst, 2014). Supplemental Figure 2

Figure S2 related to Figure 3. (A-D) Kinesin-2 depletion decreases export multiple secreted M. tuberculosis antigens from infected DC without affecting bacterial uptake. (A) Western blot showing the extent of knockdown of KIF3A in DC that were treated with siRNA targeting the kinesin-2 components KIF3A alone (lane 2), its adaptor protein KAP3 alone (lane 4) or both KIF3A and KAP3 (KIF3A lane 3) prior to

100 KD 75 KD KIF3A

Actin

Contro

l

KIF3A

KAP3siRNA: KIF

3A+

KAP3****

CM from Control DC +

+

+

CM from KAP3 depleted DC +

CM from KIF3A+KAP3 depleted DC

CM from KIF3A depleted DC

7

6

5

4

3

2

1

0IF

Na��(ng/

ml)

********

****

****

1 2 3 4

A B

C

Ag85B

CFP-10

ESAT6Esx

H

0.00

0.25

0.50

0.75

1.00

CM from Kinesin-2 depleted DC

CM

from

Contro

l DC

Ba

nd

inte

nsi

ties

0

2

4

6

Contro

l DC

KIn

esin-

2

deple

ted

DC

Bac

teria

(X

10 5

per

106

cells

)

D

infection with M. tuberculosis. DC were treated with a non-targeting siRNA as control (lane 1). (B) IFNγ secretion by P25TCR-Tg CD4+ effector T cells when co-cultured with uninfected DC that received conditioned medium (CM) generated from infected, siRNA treated, DC as shown in panel A. Control wells did not receive any CM. Culture supernatants were collected at 60 h and assayed for IFNγ by ELISA. Data are expressed as mean ± SEM *p < 0.05; **p < 0.01; ***p < 0.001 (Student’s t-test). (C) Band intensities for Ag85B, CFP-10, ESAT-6 and EsxH proteins from the western blot in Figure 3D were calculated using Image J to show the relative decrease in their export to conditioned medium generated from DC treated with siRNA targeting kinesin-2 and infected with M. tuberculosis (Fig 3D, lane 2). The band intensity for each of the proteins present in CM from control siRNA treated DC (Fig 3D, lane1) was normalized to one for comparison. (D) Live bacteria (CFU) in control DC and kInesin-2 depleted DC following overnight infection. Data in panels B and D are shown as mean ± SEM. Data represent 2-6 independent experiments *p < 0.05; **p < 0.01; ***p < 0.001 (Student’s t-test).


Figure S3 related to Figure 4. (A-C) Quantitation of bacteria in siRNA treated, M. tuberculosis-infected donor DC and in mediastinal draining lymph nodes (MDLN) and lungs of mice that received the siRNA-treated M. tuberculosis-infected DC. (A) Western blot showing KIF3A in control or kinesin-2 depleted wild type and MHC-II-/- BMDC following infection with M. tuberculosis prior to their intra-tracheal transfer to naïve MHC-II-/- mice and wild type mice respectively (B) Quantitation of bacteria in control or kinesin-2-depleted wild type and MHC-II-/- BMDC that were infected with H37Rv and administered intratracheally to naïve MHC-II-/- and wild-type mice respectively. Data are expressed as mean ± SEM of siRNA treated infected donor BMDC plated in triplicate. (C) Bacteria in MDLN (left panel) and lungs (right panel) of MHC-II-/- and wild-type mice 60 h after intratracheal transfer of siRNA-treated H37Rv-infected wild type and MHC-II-/- BMDC (from panels A and B) respectively. Data represent 2 independent experiments and are expressed as mean ± SEM. NS not significant (Student’s t-test).

CF

U (

X10

4 )/m

ouse

CF

U (

X10

6 )/m

ouse

WT mice

MHC-II-/- mice

B

MDLN Lung

Bac

teria

(x1

06)

/0.5

mill

ion

DC

A

DC

Contro

l

Kinesin

-2

dep

leted

WT

Contro

l

Kinesin

-2

dep

leted

MHC-II-/-

DC

KIF3A

Actin 0

1

2

3

4NS NS

NS

0.0

0.4

0.8

1.2

1.6 NSNS

NS

WT DC

Contro

l

Kinesin

-2

dep

leted

MHC-II-/-

DC

Contro

l

Kinesin

-2

dep

leted

0

0.2

0.4

0.6

0.8

1.0

WT mice

MHC-II-/- mice

C

WT DC

Contro

l

Kinesin

-2

dep

leted

MHC-II-/-

DC

Contro

l

Kinesin

-2

dep

leted

WT DC

Contro

l

Kinesin

-2

dep

leted

MHC-II-/-

DC

Contro

l

Kinesin

-2

dep

leted


Figure S4 related to Figure 5. (A-D). Kinesin 2 depletion does not alter phagocytosis of M. tuberculosis or in vivo migration of infected DC. (A) Western blot showing KIF3A levels in wild type infected BMDC that were treated with either non-targeting siRNA (control siRNA) or siRNAs targeting the kinesin-2 components KIF3A and KAP3 prior to intratracheal transfer to naïve wild type mice (B) Quantitation of bacteria in control and kinesin-2 depleted wild type infected BMDC given intratracheally to naïve wild-type mice. Data are expressed as mean ± SEM of siRNA treated infected donor BMDC plated in triplicate. (C) Quantitation of bacteria in MDLN (left panel) and lungs (right panel) of wild-type mice 60 h after intratracheal transfer of siRNA-treated, H37Rv-infected wild type DC (from panel A-B). (D) Frequency of cell trace-labeled control kinesin-2 depleted M. tuberculosis-infected donor DCs in the MDLNs of recipient mice following intratracheal transfer. Cells from MDLNs of mice that received non-labeled infected DC were used as controls. Data represent 2 independent experiments and are expressed as mean ± SEM. NS not significant (Student’s t-test).

CF

U (

x10

5 )/m

ouse

B

MDLN Lung

A

0

0.5

1.0

1.5

2.0

B

acte

ria (

X10

5 ) /

0.5

mill

ion

DC

CF

U (

x10

3 )/m

ouse

0

2

4

6

8

0

2

4

6

8

Control Kinesin-2 depleted



C

Contro

l

Kinesin

-2

dep

leted

KIF3A

Actin

1 2

DCNS NS

NS

CD

11

c

CFSE


Wild type mice Wild type DC

CFSE labeled infected DC

Non labeled infected DC

0.18 + 0.02_ 0.15 + 0.027_0.063

D


Figure S5 related to Figure 6: Kinesin-2 depleted DC are functionally similar to control DC, and MHC class II antigen presentation by DC is essential for the anti-mycobactericidal activity of P25TCR-Tg CD4 effector T cells. (A) FACS plots show MHC class II (left panel), CD86 (middle panel), and CD80 surface expression (right panel) in control or kinesin-2 depleted CD11c+ DC following overnight infection with M. tuberculosis (MOI 1). (B) IFNγ production by P25TCR-Tg CD4 effector T cells following uptake, processing, and presentation of recombinant Ag85B protein by control and kinesin-2 depleted DC. (C) IFNγ production by P25TCR-Tg CD4 effector T cells following stimulation with control and kinesin-2 depleted CD11c+ DC that had been infected with Ag85B deficient H37Rv (H37RvΔ85B) and pulsed with Ag85B peptide 25. (D) FACS plots

MHC-II

% o

f M

ax

CD 80 CD 86

0

10

20

30

40

Ag85B protein (10 +g/ml)

IFNa�

(ng/

ml)

0

0.6

1.2

1.8

2.4

0

0.5

1.0

1.5

2.0

IL-1

�(ng/

ml)

TN

F-_�(n

g/m

l)

Kine

sin-2

deple

ted

DC

Contro

l DC

H37Rv infected

DCFS

C

Activated Caspase 3

Control Kinesin-2 depleted Isotype control

M. tuberculosis infected DC

Kine

sin-2

deple

ted

DC

Contro

l DC

H37Rv infected

DC

C ontrolKinsein-2 depleted

Fluorescence intensity

A B C

D E

F

Bac

teria

/ 10

00 c

ells

0

500

1000

1500

2000

25 50 100 siRNA(nM): H37Rv685B infected

0

1000

2000

3000

G

Bac

teria

/ 10

00 c

ells

Control

Kinesin-2 depleted

14% 15%1.1%


+ +

Peptide 25 (2 +g/ml)

+ + _ __

H37Rv685B infected

0

10

20

30

IFNa�

(ng/

ml)

NSNS

NS

NS

H37Rvinfected

NSNS

NS NSNS

Wild type DC

MHC-II-/- DC


show expression of activated caspase 3 in control or kinesin-2 depleted CD11c+ DC following overnight infection with M. tuberculosis (MOI 1), indicating that kinesin-2 depletion does not accelerate or retard apoptosis of DC. (E) IL1-β (left panel) and TNF (right panel) release by control or kinesin-2 depleted CD11c+ DC following 24 hours of infection with M. tuberculosis (MOI 1). (F) Quantitation of live bacteria per 1000 infected DC (MOI 1) treated with varying amounts of either a non-targeting siRNA or kinesin-2 siRNA and cultured in the absence of P25TCR-Tg CD4 effector T cells. (G) Quantitation of live bacteria per 1000 wild-type and MHC-II-/- DC that were infected (MOI 1) with either H37Rv or Ag85B-deficient H37Rv (H37RvΔ85B) and cultured for 20 h in the absence of P25TCR-Tg CD4 effector T cells. Data are expressed as mean ± SEM *p < 0.05; **p < 0.01; ***p < 0.001, NS not significant (Student’s t-test) Table S1 related to Figures 3, 4, 5, and 6: Names and target sequences of siRNAs used in these studies.

ON-TARGET plus Target Sequence KIF3A SMART pool siRNA

# 09

CGACUAAUAUGAACGAGCA

# 10

GCAAAGCCUGAGACCGUAA

# 11

CGAUCAAUAAGUCGGAGAA

# 12

GGUGGAAAUGCAAGCGAAA

KIFAP3 (KAP3) SMART pool siRNA

# 12

GAGACUCAUUGCCGGGAAA

# 11

CCAUGUUGGGAGAACGAAA

# 10

GGAUGGAGAUUGACGAAGU

# 09

UUAAUGACAUGGACGAAUA

Non targeting siRNA (Control siRNA)

# 01

UGGUUUACAUGUCGACUAA

Supplemental Experimental Procedures

Immunofluorescence and confocal microscopy

BMDC on coverslips were stained with biotinylated anti-CD45.2 (Biolegend) followed by streptavidin Alexa-647

(Invitrogen) and fixed overnight in PFA. Fixed cells were permeabilized in cold PBS/0.5% Triton X-100 (5 min

on ice), washed, and incubated with anti-Ag85B (Bold et al., 2012) followed by Cy3-conjugated donkey anti-

rabbit IgG (Jackson ImmunoResearch). Transferrin receptor staining was by the same protocol with the

difference that after fixation, coverslips were briefly blocked for non-specific endogenous mouse IgG (M.O.M

kit; BMK-2202; Vector Laboratories) followed by addition of transferrin receptor antibody (136800; Invitrogen)

and anti-Ag85B antibody. Coverslips were washed and incubated with AlexaFluor-488 conjugated goat anti

mouse IgG (Invitrogen) and Cy3-conjugated donkey anti-rabbit IgG (Jackson ImmunoResearch). For tubulin

staining, coverslips were permeabilized and incubated with rabbit anti-Ag85B and mouse anti-α-tubulin

(Sigma-Aldrich), followed by Cy3-conjugated donkey anti-rabbit IgG (Jackson ImmunoResearch) and

AlexaFluor-488 conjugated goat anti-mouse IgG (Invitrogen). For KIF3A staining, fixed and permeabilized

coverslips were incubated with rabbit anti-Ag85B and goat anti-KIF3A (Santa Cruz Biotech) followed by

incubation with Cy3-conjugated donkey anti-rabbit IgG, after which the cells were washed extensively and

further incubated with AlexaFluor-647 conjugated rabbit anti-goat IgG (Invitrogen). For dynamin 2 staining,

coverslips were fixed in ice cold methanol at -20c for 10 min, washed and incubated with rabbit anti-Ag85B and

goat anti-dynamin 2 (Santa Cruz Biotech) followed by Cy3-conjugated donkey anti-rabbit IgG after which the

cells were washed extensively and further incubated with AlexaFluor-647 conjugated rabbit anti-goat IgG

(Invitrogen). After completion of staining, coverslips were washed and mounted in Vectashield mounting

medium (Vector Labs). An automated TCS SP5 II confocal microscope (Leica) with a 63x/1.4 oil immersion

lens was used to acquire 3-D image stacks (0.3-0.5 mm) using appropriate laser combinations. Representative

image stacks of average intensity from relevant channels were merged and converted to RGB format using

Image J. Color levels were enhanced using Photoshop and in some cases were done equally for all three

channels while in other cases were done differently between channels for optimal visibility in the merged

image. All related groups were subjected to similar color settings for each color channel.

Cryoimmunoelectron microscopy

BMDC infected with wild-type M. tuberculosis H37Rv or Ag85B-null M. tuberculosis (Wolf et al., 2008) were

washed and fixed (PBS/2% PFA/0.2% glutaraldehyde) as pellets that were then processed for

cryoimmunoelectron microscopy embedded in 10% gelatin, infused with sucrose, and 90 nm cryosections

obtained using a cryo-ultratome (Leica EMUC6). Grids were blocked with 1% coldwater fish skin gelatin

(Sigma), incubated with anti-Ag85B and either 15 nm Protein A-gold (University Medical Centre Utrecht, The

Netherlands) or 18 nm colloidal gold-affinity purified goat anti-rabbit IgG (Jackson ImmunoResearch) for 1 h

followed by 1% glutaraldehyde. Grids were contrasted and embedded in 3% uranyl acetate and 2% methyl

cellulose mix and examined on a Philips CM-12 electron microscope and photographed with a Gatan (4K

x2.7k) digital camera (Gatan Inc. CA).

siRNA treatment for in vitro direct T cell activation by infected cells

For assays involving in vitro T cell activation, siRNA stocks (20 µM) were diluted in 100 µl of serum-free DMEM

mixed with 6 µl of Hiperfect (Qiagen) reagent. The complexes were then added drop-wise onto BMDC (8 x 104

cell/ 0.4ml/ well of a 24-well plate) making the total volume of culture 0.5 ml per well. After forty-eight hours,

cells were infected overnight with M. tuberculosis (making the total volume 1 ml/well), washed, treated with

amikacin, washed and cultured further in fresh BMDC medium in the absence or presence of Th1 polarized

P25-TCRTg CD4 T cells for 20-24 hours. After 20-24 hrs, the plate was spun to pellet cells, supernatants were

aspirated carefully, sterile filtered, and saved for IFNγ quantitation by ELISA (BD Biosciences). The cells in the

wells were lysed in PBS-0.5% Tween 80 at a cell density of 2 x 105/ml, and lysates were plated on 7H11 agar

for bacterial quantitation.

siRNA treatment for generation of conditioned medium (CM) and the use of CM in the in vitro antigen

transfer assay

For conditioned medium generation, siRNA stocks (20 µM) were diluted in 400 µl of serum-free DMEM mixed

with 30 µl of Hiperfect (Qiagen) reagent, vortexed and left to stand for 10 min at room temperature. The

complexes were then added drop-wise onto BMDC (8 x 105 cell/ 1.6ml/ well of a 6-well plate) making the total

volume of culture 2 ml and final siRNA concentration as specified in the figures. Forty-eight hours later,

bacteria were added for a total cell culture volume of 2.5-3 ml per well. After overnight infection, cells were

washed, treated with amikacin, washed again and cultured further in fresh BMDC medium that was harvested

after 20-24 hours as conditioned medium (CM). The CM was sterile filtered and added to uninfected BMDC in

the presence of P25TCR-Tg CD4+ Th1 effector cells for 60 h, as described previously (Srivastava and Ernst,

2014). Samples were collected at 60 h and used for IFNγ quantitation by ELISA (BD Biosciences). For assays

involving macrophages, BMDM medium was used at all times.

Protein sample preparation and immunoblotting

Conditioned medium was depleted of albumin using 50K Centricon filters and the <50K filtrates were

concentrated using a 3K filter. The resulting samples were used for immunoblotting. The infected cells used to

generate the conditioned medium (CM) were washed with cold PBS and lysed in lysis buffer (150 mM NaCl, 50

mM Tris-Cl pH 7.5, 10% glycerol and 1% Triton-X; protease inhibitor tablet was added at the time of use)

collected and centrifuged at 10,000 x g for 15 mins. The resulting supernatant was collected, sterile filtered,

concentrated using a 3K filter and stored at -70°C for later use.

Samples (CM and cell lysates) were separated on either 10% or 18% polyacrylamide-SDS gels. Following

transfer to nitrocellulose (BioRad), membranes were probed with following antibodies: rabbit anti-Ag85B

(prepared by our lab (Bold et al., 2012)), mouse anti-KIF3A (1:200; Santa Cruz Biotech) or goat anti-KIF3A

(Santa Cruz Biotech), rabbit anti-KAP3 (1:500; Abcam), rabbit anti-EsxH (1:3000; inhouse), mouse anti-ESAT6

(1:1000; Thermoscientific), mouse anti-CFP10 (1:2000; BEI) and mouse anti-β-actin (1:1000; Santa Cruz

Biotech). Bound antibodies were detected using either HRP-conjugated goat anti-rabbit IgG (Invitrogen), HRP-

conjugated anti-goat IgG or HRP-conjugated anti mouse IgG (Santa Cruz Biotech).

Lung immunofluorescence

Lungs from mice infected with GFP-H37Rv or GFP-H37Rv:ΔAg85B were perfused with PBS and embedded in

OCT. 6 mm sections were cut on a Cryostat (HM505N; Leica), acetone-fixed (10 min, -20°C), and stained with

AlexaFluor-647-conjugated CD11c (N418) in PBS/1% BSA. Sections were fixed (3% PFA, RT), stained with

anti-Ag85B (1:2000) and Cy3-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch; 1:1000) in PBS/1%

BSA/0.01% Triton-X100. Sections were washed with PBS and mounted in Vectashield (Vector Laboratories).

Image stacks (0.8-1 mm) were acquired for lung sections using an automated TCS SP5 II confocal microscope

(Leica) with a 63x/1.4 oil immersion lens.

Cell sorting from infected mice and stimulation of P25TCR-Tg CD4 T lymphocytes

Lung cells from mice infected with GFP-expressing H37Rv were stained with CD11c, CD11b and Gr-1

antibodies as described (Wolf et al., 2007). Following staining, samples were pooled (7-10 mice/pool) and live

cells were sorted using an iCyt SynergyTM sorter in BSL-3 containment. Neutrophils were gated out, followed

by separation of GFP+ cells from GFP- cells. Each set (GFP+ and GFP-) was further sorted based on CD11c

and C11b staining and cultured (~10,000/well) with P25TCR-Tg Th1 effector cells (2 x 105/well) in V-bottom 96

well plates for 4 days. IFNγ in supernatants was quantitated by ELISA. An aliquot of sorted subsets (GFP+ and

GFP-) was cultured to quantitate live bacteria; GFP- subsets were found to contain fewer than 1 live

bacterium/1,000 sorted cells.

Cell sorting from LPS-treated mice and stimulation of P25TCR-Tg CD4 T lymphocytes

Lungs were harvested from C57BL/6 mice treated with 1 µg of LPS in 100 µL PBS intratracheally 3 days prior

to harvest. Samples were pooled (2-4 mice mice/pool) prior to staining with CD11c, CD11b, and Gr-1. Antigen-

presenting cells were sorted on an iCyt SynergyTM sorter using the same parameters as used for sorting cells

from lungs of M. tuberculosis-infected mice, with the exception that the FITC channel was used for

autofluorescence, rather than to detect GFP-expressing bacteria. Neutrophils were gated and alveolar

macrophages (CD11chigh, autoflouresence/FITC+) were gated out. Myeloid DCs, recruited macrophages, and

monocytes were sorted. The sorted populations were cultured (104/well) with Th1-P25TCR-Tg cells (105/well)

for 3 days. ELISA was performed to determine levels of secreted IFNγ in supernatants.

Flow cytometry staining

BMDCs were treated with control siRNA or siRNA targeting kinesin-2 for 2 days, then they were infected

overnight with M. tuberculosis (MOI 1:1). Following infection, cells were washed, treated with amikacin,

washed and stained for MHC class II, CD80, CD86 and intracellular activated Caspase-3. Samples were fixed

overnight in 1% paraformaldehyde solution before acquisition on a flow cytometer (BD LSR II).

Antigen Export Reduces Antigen Presentation and Limits T ...

Documents

Transcript of Antigen Export Reduces Antigen Presentation and Limits T ...