| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Infectious Diseases and Microbiology and
Department of Immunology, Imperial College School of Medicine, St. Mary’s Campus, London, United Kingdom;
Glaxo-Wellcome Immunology Division, Stevenage, United Kingdom; and
Unité d’Oncologie Virale, Institut Pasteur, Paris, France
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The immature live mycobacterial phagosome is a dynamic compartment that remains accessible to the recycling endosomal pathway, allowing acquisition of transferrin receptor and exogenously added glycolipids (8, 9, 10). Mycobacterial glycolipids are also exported from the immature phagosome, trafficking within the infected cell and being transferred to adjacent cells by a process of exocytosis (11, 12). The exported glycolipids have been shown to colocalize with members of the CD1 family of Ag-presenting molecules (13) and may therefore provide targets for immune recognition by CD1-restricted T cells (14). Mycobacteria are particularly rich in lipid and glycolipid components, and it is attractive to speculate that CD1-restricted responses may make an important contribution to protective immunity (15).
Mycobacteria are also rich in lipoproteins; around 100 open reading frames identified in the M. tuberculosis genome possess a characteristic prokaryotic amino-terminal acylation motif (16). We were interested in examining the possibility that lipoproteins might also be exported from the immature mycobacterial phagosome and, if so, whether they could provide targets for early recognition by T lymphocytes. We focused our analysis on a well-characterized 19-kDa lipoprotein from M. tuberculosis (17, 18). The 19-kDa Ag is present in members of the M. tuberculosis complex (including BCG vaccine strains), with a related gene in some other slow-growing mycobacteria. Sequence analysis has uncovered no obvious homology with other known proteins, and its biological function remains to be determined. The protein is predominantly associated with the mycobacterial cell wall, and there is evidence of post-translational modification by acylation and glycosylation (18, 19, 20). The 19-kDa protein has been implicated in a range of immunological responses. In addition to stimulation of innate immunity via toll-like receptors (21), it is recognized by humoral (22, 23) and cell-mediated arms of the acquired immune system (24, 25). It was one of the first Ags to be defined as the target of specific CD8+ T cell responses during human infection (26).
Interest in recognition of mycobacterial infection by CD8+ T cells has been stimulated by a requirement for these cells to provide optimal protective immunity in murine models (27, 28) and by the potential for direct killing of intracellular bacteria as a consequence of their cytotoxic action (29). The class I pathway of Ag presentation is usually associated with proteins present in cytoplasm of cells, and the route by which Ags derived from exogenous mycobacteria become available for class I-restricted presentation to CD8+ T cells is an important subject of current research (30, 31, 32).
With the aim of gaining further understanding of the fate of lipophilic components of mycobacteria within infected cells and of exploring novel Ag processing pathways, we have analyzed trafficking of the 19-kDa Ag in infected macrophages and tested its ability to deliver a class I-restricted epitope for recognition by CD8+ T cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
M. bovis BCG (Montreal strain) and M. tuberculosis H37Rv were grown with shaking at 37°C in Middlebrook 7H9 broth containing 10% albumin-dextrose-catalase supplement (Difco, West Molesey, U.K.). Mycobacterium smegmatis (strain mc2 1-2c) (33) and M. vaccae NCTC11659 were similarly grown in 7H9 broth supplemented with 2% glucose. Mycobacterium vaccae was grown at 30°C. Recombinant mycobacteria were grown in the presence of 50 µg/ml hygromycin (Roche, Lewes, U.K.). Heat-killed bacteria were prepared by incubation for 20 min at 80°C. A luciferase reporter strain of BCG expressing luxA and luxB genes from Vibrio harveyi (pSMT1) was prepared as described previously (34). Before each experiment, the bacterial viability was assessed by propidium iodide exclusion. Bacterial populations containing <20% dead microorganisms were considered live. After heat-killing, all the bacteria were propidium iodide positive.
Recombinant constructs
Plasmids used for construction of recombinant mycobacteria were based on p16R1, a shuttle vector carrying a hygromycin resistance determinant and origins of replication suitable for maintenance in mycobacteria and in Escherichia coli (35). Constructs expressing the M. tuberculosis 19-kDa Ag, and a mutant form of the protein defective in O-linked glycosylation have been described previously (19, 20). Additional 19-kDa variants were generated using a similar two-step mutagenesis strategy. In the first step, two PCR products were generated using primers Mut-1 (5'-TCGACGGTATCGATAAGCTT-3') and Mut-2 (5'-GCGGTGGCGGCCGCTCTAGA-3') derived from the p16R1 multiple cloning site and flanking the 19-kDa encoding gene together with a pair of overlapping primers, P1 and P2, specific for each mutation. A single product was then generated by a second PCR amplification using Mut-1 and Mut-2. PCR reactions were conducted in a DNA thermal cycler (Perkin-Elmer, Bucks, U.K.) using Taq DNA polymerase (Perkin-Elmer) in reactions consisting of 25 cycles of denaturation at 96°C (30 s), hybridization at 55°C (30 s), and elongation at 72°C (1–2 min). To generate a nonacylated variant, the cysteine-encoding triplet TGT at the amino terminus of the native 19-kDa protein was changed into the alanine-encoding GCG, using the primers P1-NA (5'-CTTTCCGGAGCGTCAAGCAAC-3') and P2-NA (reverse complementary sequence of P1-NA). This mutation generates a hybrid signal sequence cleavage site (GLSGAS), which would be anticipated to provide a suboptimal match for either type 1 or type 2 signal peptidases (36). However, it was expressed at a level comparable to the wild-type protein in M. vaccae, in contrast to a second mutant with an optimized type 1 site (GLAGAS), which was poorly expressed (data not shown). To generate a nonsecreted variant, the primers were P1-NS (5'-AAGGAGCTGCAGATGTCAAGCAACAAGTCG-3') and P2-NS (5'-CATCTGCAGCTCCTTCCTGTGCTCCTTTGCCTGTCG-3'). This generates a 19-kDa variant in which the signal sequence is replaced by an ATG start codon immediately upstream of the serine residue at position 2 in the wild-type protein. A third variant was generated by mutation of the AAC triplet encoding asparagine 128, to a lysine residue (AAA) using the primers P1-NNG (5'-GGTGAAAAAGTCGTTCGAAATCGAGG-3') and P2-NNG (reverse complementary sequence of P1-NNG).
For epitope tagging, PCR-directed mutagenesis was used to introduce BamHI restriction sites into the 19-kDa Ag at positions corresponding to aa 42–43 and 106–107 of the mature protein. These nucleotide changes were designed to avoid additional changes to the 19-kDa amino acid sequence. An oligonucleotide encoding a Db-restricted CD8+ T cell epitope from influenza virus nucleoprotein (NP; residues 366-ASNENMDAM-374) (37) was then introduced at each of the two BamHI sites, generating 19-kDa Ag:NP1 and 19-kDa Ag:NP2, respectively. The same sequence was also cloned into a BamHI site within a permissive loop of the M. tuberculosis superoxide dismutase (SOD) molecule as described previously (38) to generate SOD:NP. To construct a nonacylated 19-kDa Ag:NP fusion, a PvuI conservative DNA fragment was cloned from the plasmid encoding the nonacylated mutant and exchanged with the corresponding fragment from the 19-kDa Ag:NP2 fusion protein, generating 19-kDa Ag (NA):NP2.
Recombinant constructs were prepared in E. coli DH5
, and
the identity of each construct was confirmed by restriction profile and
nucleotide sequence analysis using an ABI 310 Genetic Analyzer and ABI
PRISM Dye Terminator Cycle Sequencing Ready Reaction kits (Perkin-Elmer
Applied Biosystems Division). Each construct was then introduced into
M. vaccae or M. smegmatis by electroporation as
described previously (35).
Gel electrophoresis and Western blotting
To analyze expression of recombinant proteins, sonicated mycobacterial preparations were boiled in the presence of SDS and 2-ME, and samples (2–5 µg total protein) were analyzed by PAGE in the presence of SDS, followed by transfer onto nitrocellulose membrane (Hybond C; Amersham, Aylesbury, U.K.). For visualization of Ag expression, blots were probed with polyclonal sera at a dilution of 1/3000. Antisera to the M. tuberculosis 19-kDa Ag and SOD were obtained by immunizing rabbits with recombinant proteins purified from E. coli (39) and M. vaccae (40), respectively; the Abs recognized single bands during Western blot analysis of M. tuberculosis. Blots were then incubated with HRP-conjugated goat anti-rabbit Abs (Dako, Cambridge, U.K.) and stained for peroxidase activity using chemiluminescence detection as described by the supplier (Pierce, Chester, U.K.).
Cell cultures
Bone marrow-derived macrophages (BMMs) were obtained from
femoral bones of 6- to 8-wk-old female wild-type or
TAP1-/- C57BL/6 mice (The Jackson Laboratory,
Bar Harbor, ME) and cultivated at 37°C for 4 days in the presence of
5% CO2 in serum-free macrophage-serum-free
medium (Life Technologies, Paisley, U.K.) supplemented with 10 ng/ml
recombinant mouse IL-3 (mIL-3; PharMingen, San Diego, CA)
(41). Fresh medium containing mIL-3 was added on day 4,
and cells were cultivated for an additional 4–6 days. Analysis by flow
cytometry demonstrated that the majority of cells obtained using this
procedure were positive for the macrophage markers F4/80 and
Mac-1/CD11b. The murine macrophage-like cell line J774A.1 (ATCC TIB-67;
American Type Culture Collection, Manassas, VA) was maintained in the
medium described above, with 1% FCS (Labtech, Uckfield, U.K.). The
activation status of macrophages was assessed before each infection by
measuring both the expression level of CD54 (ICAM-1) by flow cytometry
(see below) and the induction of iNOS using the Griess
reagent (Sigma, Poole, U.K.). Cells expressing 10 times less CD54 and
producing no detectable NO, compared with fully IFN-
and
LPS-activated cells, were considered resting. Dendritic cells (DCs)
were used for Ag presentation in some experiments. DCs were cultured
from bone marrow of C57BL/6 mice in the presence of recombinant mouse
GM-CSF as described previously (42) and characterized by
flow cytometry.
For flow cytometry, BMMs, DCs, and F5 T cells were stained with FITC- or PE-conjugated Abs for CD11b, I-Ab, Kb, CD80, CD86, CD54, CD25, or CD69 (PharMingen) and analyzed on a Becton Dickinson FACScan flow cytometer (San Diego, CA).
Immunofluorescence (IF) analysis of intracellular trafficking
The intracellular pathways followed by mycobacteria and mycobacterial Ags in infected macrophages were observed by IF using a Zeiss LSM-510 confocal microscope (Carl Zeiss, Herts, U.K.). Series of optical sections at <0.8-µm intervals were recorded and mounted using the Adobe Photoshop software (Adobe Systems, Amsterdam, The Netherlands). Macrophages were cultured on 13-mm glass coverslips and infected with mycobacteria at a multiplicity of infection (MOI) of 10–20. For infection times longer than 1 h, the incubation medium was removed after 1-h infection at 37°C and replaced with fresh medium containing 0.2 mg/ml amikacin (Sigma) for 2 h to prevent growth of extracellular mycobacteria. After various periods of time (see Results), infected cells were washed with PBS, fixed in PBS-4% formaldehyde for 20 min, incubated in PBS-10 mM NH4Cl for 10 min, perforated in PBS-0.1% Triton X-100 for 10 min, and treated with PBS-1% BSA-5% goat serum for 30 min. Coverslips were sequentially incubated in 20 µl of PBS-1% BSA-5% goat serum containing primary Abs for 1 h, and then with the corresponding FITC-conjugated secondary Ab (Calbiochem, La Jolla, CA) at 1/100 for an additional 1 h. Each incubation was preceded by three extensive washes in PBS. To detect intracellular bacteria, DNA was stained with 20 µl of PBS containing 2 mg/ml RNase A and 5 µg/ml propidium iodide for 1 h. In some experiments mycobacteria were stained with FITC or rhodamine isothiocyanate before infection by incubation in 1 µg/ml dye in 10 mM Na2CO3 buffer, pH 9.1, for 15 min at 37°C. The coverslips were finally mounted on glass slides using Mowiol 4-88 (Calbiochem) as mounting buffer. No IF was detectable in samples stained with secondary Abs alone or in uninfected cells stained with both primary and secondary Abs. A rabbit polyclonal serum directed against the 116-kDa subunit of the mouse vesicular H+-ATPase (v-ATPase) was provided by Synaptic Systems (Gottingen, Germany), and used at a 1:100 dilution in IF experiments. The rabbit anti-19-kDa polyclonal serum was used at a 1:500 dilution in IF experiments. The acidotropic dye Lysotracker DND-99 was obtained from Molecular Probes (Junction City, OR) and used at 50 nM for 5 min after infection.
Immunoelectron microscopy
To locate the 19-kDa Ag in mycobacteria-infected cells, immuno-gold labeling of ultrathin cryosections was conducted as described previously (43), using the rabbit anti-19-kDa polyclonal serum at a dilution of 1:200 and 20 nm gold-conjugated anti-rabbit Ab (Amersham, Little Chalfont, U.K.) without Tween 20. The preparations were examined with a JEOL EX 1200 electron microscope operating at an accelerating voltage of 80 kV (JEOL, Peabody, MA). Under these conditions, no gold particles were detected in uninfected cells stained with both primary and gold-conjugated secondary Abs.
Fractionation by density gradient electrophoresis (DGE)
Macrophages were infected with BCG at an MOI of one to five
bacteria per cell, and intracellular compartments were resolved by DGE
as described previously (44), except that the infecting
bacteria were transformed with the plasmid pSMT1 and expressed the
luciferase enzyme from V. harveyi (34). The
19-kDa Ag was detected in DGE fractions by dot-blot analysis using mAb
F29-47 (20) or rabbit anti-19-kDa serum (identical
results were obtained with either Ab). Bacteria were detected by
measuring luminescence in DGE fractions; substrate (100 µl of 1%
N-decylaldehyde in ethanol; Sigma) was injected into 1-ml
samples, and luminescence measured over a 20-s period using a Berthold
AutoLumat LB953 apparatus (Wokingham, U.K.). As detailed previously
(44), the DGE profile of early and late endosomal
compartments was monitored using HRP type IV (Sigma) as a marker, and
lysosomes were detected by measurement of 
-hexosaminidase
activity.
CD8+ T cell responses
F5 TCR transgenic mice on a recombinase-activating gene 1-/- background were used as a source of NP-specific CD8+ T cells as described previously (45). The F5 mice were originally generated by cloning cDNA encoding the TCR from a murine CTL clone recognizing aa 365–379 from the 1968 influenza A NP (46). Spleens were removed aseptically, teased into cell suspension in PBS, treated briefly with lysis buffer for 1 min, and centrifuged. Cell pellets were resuspended in complete RPMI medium (RPMI 1640 containing 10% FCS, 50 µM 2-ME, and 2 mM glutamine; Life Technologies). Cells (106/well) were incubated with influenza NP peptide (NP366–374) in 24-well tissue culture plates for 6–7 days at 37°C. Recombinant mIL-2 (10 ng/ml; PharMingen) was added on day 1, and cultures were split 1 in 4 with additional mIL-2 on days 3–4. These cells were then used as effectors on days 6–7. Naive CD8+ T cells were purified by negative selection by incubating spleen cells with a cocktail of biotinylated Abs (anti-CD11b, anti-NK1.1, anti-I-Ab; PharMingen) for 10 min at room temperature followed by washing and incubation with 100 µl of streptavidin Dynal beads (Dynal, Skogen, Norway) for 15 min, and then removal of Ab-coated cells with the magnet. The remaining cells were washed in complete medium and counted. The purity was checked by flow cytometry and was about 90%.
CTL effectors that recognized the Db-restricted NP epitope were also derived from virus-primed mice as described previously (47) and used in cytotoxicity assays. Briefly, a polyclonal T cell line was generated by priming C57BL/6 mice with a recombinant vaccinia virus encoding a rapidly degraded form of NP and subsequent weekly restimulation of splenocytes with influenza virus-infected feeder splenocytes in medium containing rIL-2. CTL were used in cytotoxicity assays on day 5 after restimulation. In these assays, BMMs were infected with recombinant mycobacteria at an MOI of 5–10 for 1 h at 37°C. Cells were then recovered, resuspended in 0.3 ml of culture medium containing 50 µCi of 51Cr (Amersham), and incubated for 1 h at 37°C. Cells were then washed once in medium without chromium and incubated for an additional 3 h. Cells were washed once more and plated out in 96-well plates at 104 cells/well. Controls (in quadruplicate) contained 0.1 ml of medium alone or 5% Triton X-100. Experimental wells (in duplicate) contained effector CTL in 0.1 ml of medium to give E:T cell ratios of 10:1, 5:1, 2.5:1, and 1.25:1. Where peptide was included, the peptide ASNENMDAM was present at 3 µM. The assay was left for 5 h at 37°C, and 50 µl of supernatant was then taken for counting. The percentage of specific lysis was calculated as follows: % lysis = 100 x [(release by CTL - medium release)/(Triton X-100 release - medium release)]. Spontaneous release in the absence of CTL was <30% for all targets.
To measure T cell activation by release of IFN-, nonadherent bone
marrow-derived DCs (2 x 106/group) were
incubated in antibiotic-free medium and pulsed with either the
ASNENMDAM peptide (1 µM) or recombinant mycobacteria at an MOI of
5–10 for 2 h at 37°C. The cells were washed three times to
remove external Ags and added to 24-well plates (5 x
104/well) together with naive or effector F5 T
cells at a ratio 10:1, T cell:DC. The cells were incubated at 37°C,
and supernatants were harvested after 24 h for effector cells and
after 48 h for naive cells. Ab pairs against murine IFN- were
obtained from PharMingen and used to measure IFN- release in
supernatants, following the recommendations of the manufacturer.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To test whether the 19-kDa lipoprotein resembles other lipophilic
components in being exported from the mycobacterial phagosome during
live infection (12, 13), we used microscopy and
biochemical fractionation approaches to study the interaction of
macrophages with live and heat-killed M. bovis BCG.
Mycobacteria were added to monolayer cultures of the J774A.1
macrophage-like cell line, and at various time points samples were
analyzed by confocal microscopy, using propidium iodide to stain the
bacteria and specific Ab to localize the 19-kDa Ag (Fig. 1
A). When phagocytosis was
prevented by carrying out incubations at 0–4°C, the 19-kDa Ag was
found closely associated with mycobacteria adhering to the surface of
host cells. When the temperature was increased to 37°C, bacteria were
observed within the cells, and there was a dramatic change in
localization of the 19-kDa Ag. Within the first hour after
phagocytosis, staining for the 19-kDa Ag was observed throughout the
infected cell. The 19-kDa signal diminished at later time points; by
3 h after initial infection, only faint staining was detected with
the anti-19-kDa Ab. In contrast to the observations with live
mycobacteria, when cells were allowed to phagocytose BCG that had been
killed by heating before infection, the 19-kDa Ag trafficked together
with the dead bacilli. An identical pattern of differential trafficking
of the 19-kDa Ag was observed when infection experiments were repeated
using murine BMMs in place of J774A.1 cells and also when BCG was
substituted by the virulent strain, M. tuberculosis H37Rv.
Differential trafficking was not observed using Abs to a second
mycobacterial Ag, SOD. SOD is found in the filtered medium and in cell
lysates of mycobacterial cultures (33, 48). During
macrophage infection, SOD staining colocalized with live BCG bacilli
throughout a 3-h infection.
|
B). The figure shows an image
obtained after 1 h of infection. A mycobacterial phagosome is seen
in the upper part of the figure; arrows indicate 19-kDa Ag
detectable outside of the phagosome. The definition obtained in these
images did not allow us to determine whether the extraphagosomal Ag was
present within a membrane-bound compartment.
Finally, we used a biochemical fractionation approach to analyze the
relationship between the 19-kDa Ag and live BCG in infected macrophages
(Fig. 1C). DGE allows separation of subcellular compartments
on the basis of charge and density and has previously been used to
monitor the location of mycobacteria (44, 49) and of
mycobacterial glycolipids (12) during macrophage
infection. The migration positions of early and late endosomes were
determined by measuring peroxidase activity after a 5-min pulse (early
endosomes) or a 5-min pulse followed by a 30-min chase (late endosomes)
with HRP. Lysosomes were identified by measuring -hexosaminidase
activity. In confirmation of a previous report (44),
immunoblot analysis for the 19-kDa Ag in extracts from cells that had
been infected with BCG for 1 h identified a predominant peak
(fractions 25–34) migrating in a region overlapping that of the major
cell protein peak. The migration position of live mycobacteria was
identified by luminescence assay, taking advantage of a
luciferase-reporter strain of BCG (34). Two peaks were
observed; one (fractions 21–27) migrating in the region
characteristic of early endosomes, and the second (fractions
11–15) on the anodic side of the lysosomal peak. Staining for
acid-fast mycobacteria was similarly localized to these two regions.
The positions of the two peaks are similar to those reported by Ferrari
et al. (49) for immature and mature phagosomes,
respectively. The quantitative distribution of luciferase-positive
mycobacteria between the two peaks varied in a series of experiments
using activated and quiescent BMMs and J774A.1 cells (M.-P. Gares, P.
Tormay, and D. Young, unpublished observations), but the differential
migration of the 19-kDa Ag was a consistent observation.
In summary, analysis by confocal IF microscopy, immunoelectron microscopy, and subcellular fractionation demonstrated that the bulk of the 19-kDa Ag is dissociated from live BCG within 1 h of macrophage infection. A rapid decline in 19-kDa Ab signal precluded monitoring of subsequent intracellular trafficking of the Ag.
Trafficking of rapidly growing mycobacteria expressing recombinant 19-kDa Ag
Trafficking of the 19-kDa Ag was further explored by cloning the
M. tuberculosis gene in rapidly growing mycobacterial host
strains, M. smegmatis and M. vaccae. These
strains lack an endogenous homologue of the 19-kDa Ag, but are able to
replicate the post-translational modification associated with the
native M. tuberculosis protein (19, 20). When
the recombinant rapidly growing strains were used to infect
macrophages, the pattern of 19-kDa staining was identical with that
observed with BCG (see Fig. 3 below), with the Ag trafficking
separately from live, but not heat-killed, bacilli.
|
). For several hours after infection,
live M. vaccae delayed the acquisition of the v-ATPase on
the membrane of their phagosome. Three hours after infection only
16 ± 6% of the phagosomes stained positively for v-ATPase. This
observation was correlated with an inability of the M.
vaccae phagosome to accumulate the acidotropic dye Lysotracker.
When bacteria were killed before uptake, the opposite pattern was seen,
with >90% of the phagosomes becoming positive for acidification
markers within 15 min. The alteration in kinetics of phagosome
maturation during the early stages of infection with live and
heat-killed M. vaccae mirrors that observed with BCG, but a
significant difference was observed between rapid- and slow-growing
mycobacteria at later time points. After 24 h phagosomes generated
during infection with live M. vaccae were predominantly
v-ATPase positive and acidified (58 ± 6%), in contrast to the
situation with live BCG, for which only 6 ± 1% of phagosomes
stained positively for v-ATPase. The similarities in phagosome
maturation and distribution of the 19-kDa Ag during the first few hours
of infection with rapid- and slow-growing mycobacteria encouraged us to
make use of the recombinant system to explore the structural
determinants involved in 19-kDa trafficking.
|
The 19-kDa Ag has an amino-terminal cysteine motif
characteristic of bacterial lipoproteins (18). To
investigate the contribution of acylation to subcellular trafficking,
the cysteine residue was modified by site-directed mutagenesis, and the
nonacylated variant was introduced into M. vaccae. This was
compared with a second 19-kDa variant in which two threonine clusters
implicated in O-linked glycosylation of the protein were
replaced by valine residues. This mutant is characterized by loss of
lectin binding and by susceptibility to a proteolytic cleavage that
results in detachment of the lipid tail from the main part of the
protein (20). In contrast to the wild-type 19-kDa Ag, both
nonacylated mutant forms remained with the mycobacteria during
macrophage infection with M. vaccae recombinants (Fig. 3). Two additional mutants were analyzed.
Mutation of a potential N-linked glycosylation site
(asparagine 128) resulted in a small reduction in apparent molecular
mass, but no change intracellular trafficking. A 19-kDa mutant lacking
the signal peptide required for protein secretion was expressed inside
the mycobacterial cell. It was readily detectable in lysed
mycobacterial extracts, but was not detected in culture
supernatants or during IF analysis of infected cells. These
results show that the lipid tail of the 19-kDa Ag is required for
differential trafficking in macrophages.
Access to an MHC class I presentation pathway
We next wished to assess whether differential trafficking of the 19-kDa Ag was associated with differential availability for immune recognition. In particular, the apparent extraphagosomal location led us to focus on possible access to the MHC class I pathway. To facilitate comparison of different Ags expressed in different bacterial compartments using a single well-defined T cell readout, we adopted an approach based on epitope tagging (50). An oligonucleotide cassette encoding a 9-aa peptide corresponding to residues 366-ASNENMDAM-374 of influenza virus NP was inserted within the coding sequences of intact and nonacylated forms of the 19-kDa Ag and within a permissive loop of the M. tuberculosis SOD molecule (38). Each epitope-tagged recombinant protein was expressed in M. vaccae at a level comparable to the corresponding wild-type protein. Subcellular trafficking of the 19-kDa Ag was unaffected by the epitope tag.
Class I presentation of the NP epitope delivered by the different
recombinant mycobacteria was assayed using NP-specific T cells derived
from virus-primed mice and from F5 TCR transgenic
mice (45). BMMs were infected with recombinant M.
vaccae for 1 h and then radiolabeled with
51Cr. Effector CTL were added, and chromium
release was measured after 5 h. Cytolytic activity was observed
only in the case of macrophages infected with M. vaccae
expressing the acylated 19-kDa fusion protein with the NP epitope
inserted at either of two sites in the protein (Fig. 4). The extent of cell lysis after
infection at a ratio of 10 bacteria/cell was comparable to that
observed using macrophages incubated with purified NP peptide;
infection at a ratio of 1:1 resulted in a corresponding reduction in
the specific lysis. CTL recognition of the 19-kDa Ag:NP recombinant was
completely inhibited when M. vaccae preparations were killed
by heating before infection or when the nonacylated variant of the
19-kDa Ag was used to carry the NP epitope. The same pattern was
observed using an alternative bone marrow-derived DC system for Ag
presentation, with IFN- expression or induction of cell surface
markers CD25 and CD69 as readout for T cell activation (Fig. 5). Again, 19-kDa Ag:NP, but not SOD:NP
triggered a T cell response, which was ablated in heat-killed samples.
In this assay system activation of naive splenocytes from
F5 TCR transgenic mice as well as activation of
primed effector cells was observed.
|
|
To explore the pathway involved in class I presentation of the
19-kDa Ag:NP fusions, experiments were repeated using BMMs from mice
lacking the TAP1 transporter. The level of CTL activity using BMMs from
TAP1-/- mice was comparable to that observed in
wild-type C57BL/6 macrophages, demonstrating a TAP-independent
mechanism (Fig. 6). Exocytosis of
processed peptides (peptide regurgitation) has been shown to result in
class I loading in some systems (51, 52, 53). To test for this
mechanism, J774A.1 macrophages, which express the H-2
Dd determinant and are not recognized by the
effector CTL, were infected with M. vaccae expressing 19-kDa
Ag:NP and then mixed with 51Cr-labeled
L-Db fibroblasts, which are routinely used as
target cells in the NP model (47). No CTL activity was
detected on subsequent addition of the
Db-restricted T cells (Fig. 6), demonstrating
that peptide regurgitation does not make a major contribution to
19-kDa-mediated presentation of the NP epitope.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Building on the observation that mycobacterial glycolipids are exported from the immature phagosome during live infection (11, 12, 13), we investigated whether an analogous pathway might influence trafficking of a lipoprotein Ag. Evidence from confocal microscopy, electron microscopy, and subcellular fractionation suggests that the 19-kDa lipoprotein is indeed exported from the mycobacterial phagosome during the first hour after phagocytosis of live mycobacteria. Experiments with mutated forms of the protein demonstrated that 19-kDa trafficking is dependent on the presence of the lipid tail on the protein, and it is possible that this promotes interaction and cotrafficking together with other lipophilic components. Further work will be required to confirm whether this is the case. The Ab-based approach we have used to track the 19-kDa protein is limited by the fact that the signal disappears a few hours after macrophage infection, perhaps as a result of proteolytic degradation. Within the limitations of the assay system, we are able to conclude that the 19-kDa Ag trafficks separately from the live mycobacteria during the early stages of infection. However, we are unable to determine whether, as has been demonstrated for the glycolipids (12), the 19-kDa Ag is subsequently transferred to late endosomal compartments and exported to adjacent cells. Additional experiments using alternative tracking techniques to follow the fate of mycobacterial lipoproteins and lipoprotein fragments during infection will be required to characterize the trafficking pathway. This is of particular importance in light of the observed association of bacterial lipoproteins, including the 19-kDa Ag (21), with signaling through toll-like receptors (56). Lipoprotein trafficking may influence innate as well as acquired immune responses during mycobacterial infection.
Differential trafficking of the 19-kDa Ag was prevented by heat-killing of mycobacteria before infection. This could be due to physical trapping of the protein as a result of denaturation of cell wall components. Alternatively, the rapid maturation of phagosomes containing the killed bacteria may reduce the opportunities for lipoprotein export. It is of interest that, at least during the first few hours after infection, the delay in phagosome maturation following live infection is similar in the case of both rapid- and slow-growing mycobacteria. The delayed acidification observed for the M. vaccae phagosome over a 24-h timeframe parallels previous results with M. smegmatis, for which it was found that phagosome acidification proceeded over a period of several days in J774 cells (57). A delay in phagosome maturation has also been observed following uptake of hydrophobic particles (58), and it is possible that the slow kinetics of mycobacterial phagosome development reflect in part an inherent property of their lipid-rich cell wall. The more prolonged resistance to acidification at later stages of infection is characteristic of the slow-growing pathogenic mycobacteria and presumably involves mechanisms that enhance or synergize with this general hydrophobicity effect. The effect of heat-killing on phagosome maturation might, in turn, at least partially reflect a change in physical properties of the mycobacterial surface. Analysis of phagosome maturation and export of lipophilic components following uptake of mycobacteria that have been killed by different methods will assist in clarification of the exact relationship between viability and trafficking.
The second question addressed in the study concerned the availability of the 19-kDa protein for Ag processing. When delivered as a fusion with the intact acylated 19-kDa protein, a peptide epitope from influenza virus NP was rapidly and efficiently delivered for recognition by MHC class I-restricted CD8+ T cells during live M. vaccae infection. As had been observed in the trafficking experiments, Ag presentation was blocked by removal of the lipid tail or by heat-killing before infection.
The class I pathway is primarily involved in sampling of Ags present in the cytoplasm of APCs (59). Cytoplasmic proteins, generated during viral infection for example, are digested by proteasomes, and the resulting peptides are conveyed by the TAP transporter to the endoplasmic reticulum where they can associate with nascent class I molecules. However, the class I pathway can also play a role in the immune response to exogenous proteins during infection with intracellular bacteria (60), and experiments in knockout mouse strains have shown that CD8+ T cells are essential for optimal resistance to M. tuberculosis (27). Therefore, the mechanisms by which mycobacterial Ags gain access to the class I presentation pathway are of considerable interest. Class I recognition of exogenous Ags can occur by two general mechanisms (61). Firstly, the Ag can gain access to the cytoplasm, for example as a result of disruption of the phagosomal membrane. Once in the cytoplasm, the Ag can then be processed by the conventional TAP-dependent pathway. Alternatively, Ags trafficking within the endosomal network can associate with class I molecules either within the cell or on the cell surface following exocytosis of processed peptides. The second mechanism, a phenomenon initially described as cross-priming, is independent of the TAP transporter. Previous studies have provided evidence of a TAP-dependent pathway during mycobacterial infection. TAP-knockout mice have enhanced susceptibility to tuberculosis (28, 62), and phagosomes in macrophages infected for 18–36 h with live BCG show enhanced permeability to macromolecules (30, 32). In contrast, the 19-kDa pathway identified in the present study belongs to the TAP-independent category and is novel in its dependence on acylation. The site at which loading of class I molecules with 19-kDa-derived peptides occurs remains to be determined. The lack of CTL activity during coculture of infected histoincompatible macrophages with histocompatible target cells suggests that exocytosis of processed peptide is unlikely to play a role, at least in the short timeframe used in the assay. Access to an MIIC-like compartment (as demonstrated for mycobacterial glycolipids (12, 13)) would provide a potential site for 19-kDa processing and loading of both class I and class II molecules (63). However, as discussed above, alternative experimental approaches will be required to evaluate this as a possible route of lipoprotein trafficking.
Enhanced susceptibility in knockout mice has been used to demonstrate a role for TAP-dependent processing in the immune response to M. tuberculosis (28, 62), and it remains to be investigated whether the TAP-independent lipoprotein pathway makes some additional contribution. The rapid kinetics of the TAP-independent presentation would have obvious attractions in terms of mounting an early protective response. We did not observe the early CTL response following introduction of the 19-kDa Ag:NP construct into BCG. Quantitative effects related to competition between expression of recombinant forms of the 19-kDa Ag and expression of the endogenous BCG homologue could contribute to this observation, and additional experiments in 19-kDa deletion strains or using a 19-kDa-specific T cell readout are required to determine whether this pathway is functional in slow-growing mycobacteria. It is intriguing to speculate that the mycobacterial pathogens, which have been shown to interfere with MHC class II trafficking within infected cells (64), might somehow disrupt this early display of class I-associated Ag. The paradoxical observation that despite inducing an Ag-specific Th1 response, addition of the 19-kDa Ag has a detrimental effect on protective efficacy of rapid-growing mycobacterial vaccines (40), provides a further indication of a complex biology underlying the immune response to mycobacterial lipoproteins.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. O. Neyrolles, Service de Microbiologie du Prof. Lagrange, Hopital Saint Louis, 1 avenue Claude Vellefaux, 75475 Paris Cedex 10, France.
3 Abbreviations used in this paper: bacillus Calmette-Guérin, BCG; NP, nucleoprotein; BMM, bone marrow-derived macrophage; mIL, mouse IL; DC, dendritic cell; IF, immunofluorescence; MOI, multiplicity of infection; DGE, density gradient electrophoresis; SOD, superoxide dismutase; v-ATPase, vesicular H+-ATPase.
Received for publication June 8, 2000. Accepted for publication October 4, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2-microglobulin in the processing of bacterial or particulate antigens via an alternate class I MHC processing pathway. J. Immunol. 156:4182.[Abstract]
interferon-activated bone marrow macrophages: role of reactive nitrogen intermediates. Infect. Immun. 59:3213.This article has been cited by other articles:
|
|
 |
|
  M. Bastian, T. Braun, H. Bruns, M. Rollinghoff, and S. Stenger Mycobacterial Lipopeptides Elicit CD4+ CTLs in Mycobacterium tuberculosis-Infected Humans J. Immunol., March 1, 2008; 180(5): 3436 - 3446. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Barrionuevo, J. Cassataro, M. V. Delpino, A. Zwerdling, K. A. Pasquevich, C. G. Samartino, J. C. Wallach, C. A. Fossati, and G. H. Giambartolomei Brucella abortus Inhibits Major Histocompatibility Complex Class II Expression and Antigen Processing through Interleukin-6 Secretion via Toll-Like Receptor 2 Infect. Immun., January 1, 2008; 76(1): 250 - 262. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Gandotra, S. Jang, P. J. Murray, P. Salgame, and S. Ehrt Nucleotide-Binding Oligomerization Domain Protein 2-Deficient Mice Control Infection with Mycobacterium tuberculosis Infect. Immun., November 1, 2007; 75(11): 5127 - 5134. [Abstract] [Full Text] |
), late endosomes (
), and total
proteins (•). The right panel shows the migration
position of luciferase-expressing mycobacteria (•) and 19-kDa Ag
(
