Temporal Regulation of Intracellular
Organelle Homeostasis in T Lymphocytes by
Autophagy
This information is current as
of June 18, 2017.
Subscription
Permissions
Email Alerts
J Immunol 2011; 186:5313-5322; Prepublished online 18
March 2011;
doi: 10.4049/jimmunol.1002404
http://www.jimmunol.org/content/186/9/5313
This article cites 38 articles, 18 of which you can access for free at:
http://www.jimmunol.org/content/186/9/5313.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2011 by The American Association of
Immunologists, Inc. All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
References
Wei Jia and You-Wen He
The Journal of Immunology
Temporal Regulation of Intracellular Organelle Homeostasis
in T Lymphocytes by Autophagy
Wei Jia and You-Wen He
T
he intracellular self-degradation homeostasis pathway
macroautophagy (hereafter referred to as autophagy) is evolutionarily conserved from yeast to mammals (1–5). Generally, amino acid starvation, growth factor withdrawal, and certain stress conditions activate the autophagy pathway. Autophagy is
initiated by a membrane structure termed an isolation membrane
or autophore. Through an elongation step, the autophore forms a
characteristic double-membrane structure called the autophagosome to surround long-lived cytosolic proteins, damaged or excess
organelles, or invading intracellular pathogens. Finally, the autophagosome matures by fusing with the lysosome to form the autolysosome, and the materials inside the autolysosome are degraded
by lysosomal proteases. The elongation step is generally directed
by autophagy-related (Atg) molecules. To date, 33 different autophagy-related genes have been identified in yeast (6), and most of
the corresponding homologous molecules have been confirmed in
mammalian cells (7). Two ubiqutin-like conjugation systems, the
Atg12 conjugation system and the microtubule-associated protein
L chain 3 (LC3; Atg8 in yeast) conjugation system, control the
elongation of the autophore to form the autophagosome. The molecular mechanisms of the autophagy-processing pathway have
Department of Immunology, Duke University Medical Center, Durham, NC 27710
Received for publication July 20, 2010. Accepted for publication February 14, 2011.
This work was supported by National Institutes of Health Grants AI074944 and
AI074754 (to Y.-W.H.).
Address correspondence and reprint requests to Dr. You-Wen He, Department of
Immunology, Box 3010, Duke University Medical Center, Durham, NC 27710.
E-mail address: [email protected]
Abbreviations used in this article: 7-AAD, 7-aminoactinomycin D; Atg, autophagyrelated; DHE, dihydroethidium; DN, double-negative; EM, electron microscopy; ER,
endoplasmic reticulum; ES, embryonic stem; Exon, exon fragment; FLICA, fluorochrome inhibitor of caspases; LA, long arm; LC3, microtubule-associated protein L
chain 3; LN, lymph node; MFI, mean fluorescence intensity; 4-OHT, 4-hydroxytamoxifen; ROS, reactive oxygen species; SA, short arm.
Copyright Ó 2011 by The American Association of Immunologists, Inc. 0022-1767/11/$16.00
www.jimmunol.org/cgi/doi/10.4049/jimmunol.1002404
been reviewed in detail (8). In both conjugation systems, Atg7
serves as a ubiquitin enzyme E1-like molecule. Atg10 serves as an
E2-like molecule in the Atg12 conjugation system. Atg7 and Atg10
catalyze Atg12 and Atg5 to form an Atg12–Atg5–Atg16L complex.
LC3 is cleaved by the protease Atg4 to generate the cytosolic LC3-I.
Atg7 and the other E2-like molecule, Atg3, then catalyze the conjugation of LC3-I to generate the membrane-bound lipid form, LC3phosphatidylethanolamine, which is also called LC3-II (9). The
Atg12 and LC3 conjugation systems interact and regulate each
other. The Atg5–Atg12–Atg16L complex functions as an E3-like
molecule for LC3 localization and lipidation (10, 11). Overexpression of both Atg5 and Atg12 inhibits the lipidation of LC3
(12).
Atg3 interacts with Atg12 (10), and in Atg3-deficient mice, the
formation of Atg12-Atg5 conjugates is dramatically decreased.
Furthermore, the elongation and closure of autophagosomes are
defective in the absence of Atg3 (13). Therefore, both conjugation
systems are essential for autophagy in mammalian cells. Interestingly, an Atg7/Atg5-independent alternative autophagy pathway
has been identified in mouse cells (14). In Atg7- or Atg5-deficient
cells, autophagosomes and autolysosomes formed and mediated
protein degradation (14).
Autophagy has critical physiological functions in the immune
system. Autophagy is the mechanism of clearance of intracellular
pathogen infections (15), presentation of exogenous Ags (16),
cross-presentation of endogenous Ags by MHC class II molecules
(17, 18), and recognition of ssRNA viruses and production of IFNa in plasmocytoid dendritic cells (19). Our group and others have
shown that autophagy is essential for organelle homeostasis
in lymphocytes (20, 21) and erythroid cells (22–24). Autophagy
developmentally regulates mitochondria (20) and endoplasmic
reticulum (ER) (21). In an Atg5 knockout fetal liver transfer
system and an Atg7f/fLck-Cre T cell-specific deletion system, we
have demonstrated that autophagy is essential for T lymphocyte
survival (25). To determine whether the phenotypes observed in
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
The highly conserved self-degradation pathway known as autophagy plays important roles in regulating T lymphocyte homeostasis.
Recently, we found that T lymphocytes lacking the autophagy-related gene Atg5 or Atg7 have defective survival and contain expanded mitochondria and endoplasmic reticulum (ER); however, whether these defects are caused by impaired autophagy or by
defects in their autophagy-independent signaling capacity of Atg5 or Atg7 in T lymphocytes remains unknown. Furthermore, the
function of the microtubule-associated protein L chain 3 (LC3) conjugation system in T lymphocytes remains unclear. To address
these questions, we generated conditional knockout mice with specific deletion of Atg3, a ubiquitin enzyme E2-like molecule involved
in the LC3 conjugation system, in T lymphocytes. Atg3-deficient T lymphocytes displayed a phenotype similar to those of Atg7- and
Atg5-deficient T cells. The survival of Atg3-deficient naive CD4+ and CD8+ T cells was defective. Furthermore, the mitochondria
and ER were expanded in Atg3-deficient T cells. Interestingly, mitochondrial and ER content did not change instantly upon
inducible deletion of Atg3 in mature T lymphocytes in vitro. Instead, it began to expand 10 d after inducible deletion of Atg3 in
mature T lymphocytes, and mitochondrial content continued to increase on day 18. Cell death began to increase 24 d after
inducible deletion of Atg3. These data show that the LC3 conjugation system is essential for autophagy in T lymphocytes. Our data
suggest that autophagy promotes T lymphocyte survival by regulating organelle homeostasis and that the decreased survival of
autophagy-deficient T cells is due to the temporal accumulation of these autophagy-related defects. The Journal of Immunology,
2011, 186: 5313–5322.
5314
Atg5- and Atg7-deficient T cells are due to impairments in
autophagy machinery or in the signaling capacity of the Atg5/
Atg7 molecules and to further investigate the functions of the
LC3 conjugation system in T lymphocytes, we generated Atg3
conditional knockout mice.
In this study, we show that the autophagy pathway is impaired in
Atg3-deficient T lymphocytes. Survival is defective and organelle
homeostasis is abnormal in Atg3-deficient T cells. Furthermore, we
use an inducible deletion system to demonstrate a temporal accumulation of mitochondria and ER beginning 10 d after the deletion of Atg3. Accordingly, we show that cell death does not begin
to increase until .3 wk after deletion of Atg3. Our results demonstrate that the LC3 conjugation system is indispensable for
autophagy in T lymphocytes. We also show that autophagy regulates organelle homeostasis and is essential for T cell survival.
Importantly, our findings provide clear evidence that the phenotype of autophagy-deficient T cells is due to the temporal accumulation of autophagy-related defects.
Atg3 IN T LYMPHOCYTE FUNCTION
Flow cytometry analysis
The phenotype of Atg3f/fLck-Cre mice was analyzed in sex- and agematched 5–8-wk-old mice. Thymus, lymph node (LN), and spleen were
isolated from Atg3f/fLck-Cre mice and Atg3f/f wild-type control mice, and
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Materials and Methods
Generation of Atg3 conditional knockout mice
The bacterial artificial chromosome clone RP23-444N1, which contains the
Atg3 genomic sequence, was obtained from BACPAC Resources Center at
Children’s Hospital Oakland Research Institute (Oakland, CA) and was
used as a template to clone the genomic fragments long arm (LA), short
arm (SA), and exon fragment (Exon) by PCR. The exon fragment is
flanked by two loxP sites, allowing for deletion in Cre-expressing cells at
the genomic level. The primer sequences are as follows: LA forward
primer with SacII restriction site, 59-GCCCGCGGCCTCTGAACTCCATACATGTGTACACACACCAAGT-39; LA reverse primer with NotI
restriction site, 59-GCGCGGCCGCCAGCTTTGTTCTCACCATAGAACCTACACCCCCGG-39; Exon forward primer with SmaI restriction site, 59GCCCCGGGTTTATTAAATTTCAAGAAGAAAAAAGATCATGCTACTTG-39; Exon reverse primer with SmaI restriction site, 59-GCCCCGGGGATGTTTACACATCTCCACACGGACAACACAGACA-39; SA forward
primer with SalI restriction site, 59-GCGTCGACAGTAAAAGAAAATTTTACTTTATAAGCAAAATTCTCTTTC-39; and SA reverse primer with
SalI restriction site, 59-GCGTCGACCGGTGCTATTTCAGTTCTTATCAACAATCCTTTAC-39. PCRs were performed using BIO-X-ACT Long DNA
Polymerase (Bioline, Taunton, MA). All cloned fragments were inserted
into the pGKneoF2L2DTA targeting vector (26). The resulting pGKneoF2L2DTA-Atg3 construct was sequenced. Three mutations were found in
the LA fragment in the intron region between exon 7 and exon 8. No mutations were found in the Exon or SA fragments. EF1 embryonic stem (ES)
cells were transfected with linearized pGK-Neo-FRT-Atg3. The targeted ES
cell clones were screened for homologous recombination at the Atg3 genomic locus by PCR. The primers used for screening are as follows: forward
ES cell screening primer Neo F2 (corresponding to the Neo region in
pGKneoF2L2DTA), 59-CTCTATGGCTTCTGAGGCGGAAAGA-39; and reverse ES cell screening primer R2 (corresponding to the sequence downstream of the SA fragment in the Atg3 genomic locus), 59-TGCTGCAAATAGAGTGAATCAGCTAAAGGG-39. The positive ES clones were further
confirmed by PCR amplification of the region containing the other LoxP site
between the LA fragment and the Exon fragment. The primers used for
screening are as follows: forward Atg3 screening primer F4500, 59-AACCATAGCCGTGGTGTCTGGTAA-39; and reverse Atg3 screening primer R450,
59-CGATGGCATCTTATGCTGAGCAATG-39. This pair of primers was later
used to screen the Atg3f/+or Atg3f/f mice. The targeted ES clone 3D11 was
injected into B6 black mice blastocysts, and the injected blastocysts were
transferred into the uterus of pseudopregnant female recipient mice. The
resulting chimeric mice were screened and mated. The germline transmitted
and genetically modified male Atg3 mice (Atg3f/+) were bred with female
FLPeR mice to delete the FRT-flanked neomycin cassette (27). Finally, the
Atg3f/f mice were bred with Lck-Cre (28) or ER-Cre mice (29) to generate
Atg3f/fLck-Cre or Atg3f/fER-Cre mice. The deletion efficiency in Atg3f/fLckCre mice was analyzed by PCR using forward primer FE8, 59-GAGAGTGGATTGTTGGAAACAGATGAGG-39; and reverse primer RE9, 59-CTCATCATAGCCAAACAACCATAGCCG-39, to amplify a 313-bp product from
the genomic Atg3 exon 8 and exon 9 regions. Sex- and age-matched 5–8-wkold mice were used in all experiments. All experiments were performed following protocols that were approved by the Duke University Institutional
Animal Care and Use Committee. Mice were housed in a specific pathogenfree animal facility at Duke University Medical Center.
FIGURE 1. Generation of Atg3 conditional knockout mice. A, Strategy
for generating Atg3 conditional knockout mice. The diagrams represent
the target construct and deleted allele. B, Targeted ES cell screening
results. The targeted ES cells were screened by PCR using primers Neo F2
and R2 and further confirmed by PCR using primers F4500 and R450. The
DNA gel shows results from the PCR using primers F4500 and R450 using
ES cell genomic DNA as templates (left panel). The right panel shows the
PCR screening results using genomic DNA from mouse tails as templates.
The PCR product of floxed Atg3 is 866 bp, and the PCR product of wildtype Atg3 is 748 bp. C, Deletion efficiency in Atg3f/fLck-Cre mice. CD4+
and CD8+ cells were sorted by flow cytometry to .98% purity, and genomic DNA was extracted. The fragment from Atg3 exon 8 and exon 9 was
amplified by PCR using primers FE8 and RE9. Atg7 exon 14 PCR was
used as a control. D, The expression of Atg3 in T cells was analyzed by
Western blot. Samples of thymocytes or enriched T cells from Atg3f/fLckCre and Atg3f/f control mice are shown in the left panel. The right panel
shows 4-OHT–treated splenocytes from Atg3f/fER-Cre and Atg3f/f control
mice. The numbers at the right side represent Mr of protein markers.
*Nonspecific bands. DT, diphtheria toxin; E, exon; F, flippase recognition
target (FRT); L, loxP; x, XbaI restriction site.
The Journal of Immunology
cell suspensions were prepared. RBCs were lysed with ACK buffer.
Thymocytes were suspended in anti-Fc receptor Ab supernatant (2.4G2)
and stained with anti-CD4–PE, anti-CD8–FITC, and 7-aminoactinomycin
D (7-AAD) (BD Pharmingen, San Jose, CA). For analysis of doublenegative (DN) thymocytes, thymocytes were stained with anti-CD4–allophycocyanin, anti-CD8–allophycocyanin, anti-CD3–allophycocyanin, antiCD44–FITC, anti-CD25–PE, and 7-AAD. Splenocytes and LN cells were
suspended in 2.4G2 supernatant and stained with anti-CD4–PE, anti-CD8–
FITC, and 7-AAD. Alternatively, splenocytes or LN cells were stained with
anti-CD4–allophycocyanin (or anti-CD8–allophycocyanin), anti-CD62L–
PE, anti-CD44–FITC, and 7-AAD on ice for 15 min. Cells were then
washed, and the samples were analyzed by flow cytometry using FACScan,
FACStar Plus, or FACSCanto flow cytometers (BD Biosciences). All Abs
were from Biolegend (San Diego, CA) or eBioscience (San Diego, CA).
5315
Autophagy analysis
Electron microscopy analysis
FACS-sorted CD4+ and CD8+ T cells from spleen and LN of Atg3f/fLckCre and Atg3f/f mice were fixed in 4% glutaraldehyde/0.1 M sodium
cacodylate buffer overnight, and electron microscopy (EM) samples were
prepared as described previously (20). Micrographs were taken with a
Philips LS 410 electron microscope. The cytosol membrane structures were
quantified and presented as the membrane score. The EM picture of each
cell was divided into 12 equal sections according to a clock face. The membrane score was counted as each section containing membrane structure(s)
for each cell.
Apoptosis analysis
The freshly isolated splenocytes were stained for surface markers using antiCD4–allophycocyanin, anti-CD8–PE/Cy7, and anti-CD44–allophycocyanin Abs in 2.4G2 supernatant on ice for 15 min. Cells were then washed,
resuspended in binding buffer, and stained with Annexin V-PE or Annexin
V-Pacific Blue (Biolegend) and 7-AAD (BD Biosciences). The samples
were analyzed using an FACSCanto flow cytometer (BD Biosciences).
Caspase-9 activity was detected by fluorochrome inhibitor of caspases
FIGURE 2. Impaired autophagy in Atg3-deficient T cells. A and B, LC3
puncta formation in Atg3-deficient T cells. Splenocytes from Atg3f/fLckCre mice and Atg3f/f control mice were stimulated with anti-CD3 Ab or
cultured in complete RPMI 1640 medium overnight. LC3 puncta were
intracellularly stained with rabbit anti-LC3 Ab and Cy3-labeled donkey
anti-rabbit Ab. Z-stack images were captured and spaced at 1 mm apart
using 1003 oil objective in a Zeiss Axio Observer D1-based imaging
station (Zeiss). A total of 30 cells in each group were analyzed. Representative images were shown in A, and the numbers of LC3 puncta were
quantified (cut off at pixel .4). C, The level of LC3 II in Atg3-deficient
T cells under starvation condition. Splenocytes from Atg3f/fLck-Cre and
Atg3f/f mice were cultured in HBSS in starvation (Starv) condition or
complete medium overnight. T lymphocytes were purified after culture,
and cell lysates were prepared. The LC3-II was detected with anti-LC3 Ab
by Western blot. D, The expression of p62 in Atg3-deficient T cells. The
enriched naive T cell lysates (CD44low) were analyzed by Western blot. E,
EM images of Atg3-deficient T cells. Sorted CD4+ or CD8+ cells from
Atg3f/fLck-Cre and Atg3f/f mice were fixed, and EM samples were prepared. The upper panels were magnified 315000. The lower panels are
detailed images of the upper panels and were magnified 325000. Scale
bars, 500 nm. The membrane structures in the EM pictures were quantified
as membrane scores and shown in F. The cell was divided into 12 equal
sections according to a clock face. One membrane score means that there
is membrane structure(s) in each section of one cell. The column figures
are mean 6 SD and from 50 cells of each group. *p = 1.46 3 10221, **p =
1.23 3 10222.
FIGURE 3. Thymocyte development in Atg3-deficient mice. A, Total of
five Atg3f/fLck-Cre, three Atg3f/+Lck-Cre, and four Atg3f/f control mice.
B, Thymocyte subpopulations in Atg3f/fLck-Cre mice. Thymocytes from
Atg3f/fLck-Cre, Atg3f/+Lck-Cre, and Atg3f/f mice were stained with antiCD8–FITC, anti-CD4–PE, and 7-AAD. The cells were gated on the 7AAD2 population. Numbers represent the frequency of each gated subpopulation. C, DN subpopulations in Atg3-deficient mice. Thymocytes
from Atg3f/fLck-Cre, Atg3f/+Lck-Cre, and Atg3f/f mice were stained with
anti-CD4–allophycocyanin, anti-CD8–allophycocyanin, anti-CD3–allophycocyanin, anti-CD44–FITC, anti-CD25–PE, and 7-AAD. The samples were
gated on CD4-, CD8-, CD3-, and 7-AAD–negative cells. Numbers represent
the frequency of each gated subpopulation. Both B and C flow cytometry
analyses were repeated three times. *p = 0.018, **p = 0.031.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Splenocytes from Atg3f/fLck-Cre and Atg3f/f mice were stimulated with
anti-CD3 mAb (2C11; 5 mg/ml) or cultured in RPMI 1640 medium
overnight. The CD3 activated and unstimulated control T cells were
stained with anti-CD4–FITC or anti-CD8–FITC. After washing, the cells
were treated with 0.1% saponin (Sigma-Aldrich, St. Louis, MO) in PBS
containing 0.5% BSA for 10 min on ice and then intracellularly stained
with rabbit anti-LC3 Ab (MBL International, Woburn, MA) for 30 min on
ice. After washing, anti-rabbit–Cy3 Ab (Jackson ImmunoResearch Laboratories, West Grove, PA) was added and incubated on ice for another 30
min. Z-stack images were acquired at 1 mm apart using 1003 oil objective
by a Zeiss Axio Observer D1-based imaging station (Zeiss) equipped with
a CoolSNAP HQ CCD camera (Roper Scientific, Tucson, AZ) and
recorded with MetaMorph 7.6 software (Universal Imaging, West Chester,
PA). Three-dimensional deconvolution of the images was performed using
AutoQuant X software (Media Cybernetics Bethesda, MD). LC3 puncta of
three-dimensional deconvolution images was quantified by MetaMorph 7.6
software (Universal Imageing). A 1.53 background fluorescence intensity
was used to set up the threshold for the images of each cells. Only the pixel
of the LC3 image, which is .4, was considered as an LC3 punctum. The
numbers of LC3 puncta per cell were counted, and a total of 30 cells from
each group were analyzed. Starvation was induced by culturing the splenocytes in HBSS (Invitrogen, Carlsbad, CA) overnight. Then, T lymphocytes
were purified using a mouse T cell negative enrichment kit (StemCell
Technologies, Vancouver, BC, Canada), and cell lysates were prepared
for LC3 Western blot analysis.
5316
(FLICA) apoptosis detection kit (ImmunoChemistry Technologies, Bloomington, MN) according to the manufacturer’s instructions. Briefly, the
splenocytes were incubated with caspase-9 FLICA (carboxyfluorescein
[FAM]-LEHD-fluoromethyl ketone) at 37˚C for 30 min. After washing, the
cells were further stained with surface markers as described above. The
samples were analyzed by flow cytometry after another washing.
In vitro inducible Atg3 deletion
The splenocytes or LN cells from Atg3f/fER-Cre or Atg3f/f control mice
were cultured with 200 nM 4-hydroxytamoxifen (4-OHT; Sigma-Aldrich)
for 3 d with IL-7 (1 ng/ml; PeproTech, Rocky Hill, NJ). For longer cultures, 1 ng/ml IL-7 was added every other day, and the medium was
changed when necessary.
Mitochondria and ER staining
Splenocytes from Atg3f/fLck-Cre or 4-OHT–treated Atg3f/fER-Cre splenocytes and Atg3f/f control splenocytes were incubated with 100 nM MitoTracker Green (Invitrogen, Molecular Probes) or 1 mM ER-Tracker BlueWhite DPX (Invitrogen, Molecular Probes) in RPMI 1640 medium at 37˚C
for 30 min. Cells were washed with RPMI 1640 medium and then stained
with anti-CD4–allophycocyanin, anti-CD8–PE/Cy7, anti-CD44–allophycocyanin/Cy7, and 7-AAD. The geometric mean intensity of fluorescence
Atg3 IN T LYMPHOCYTE FUNCTION
was used to represent the volume of mitochondria or ER. The mitochondria and ER of Atg3f/f ER-Cre splenocytes and Atg3f/f control splenocytes
were analyzed on different days after 4-OHT treatment. The samples were
analyzed using an FACSCanto flow cytometer (BD Biosciences).
Reactive oxygen species production analysis
Splenocytes from Atg3f/f Lck-Cre mice or Atg3f/f control mice were incubated with 2.5 mM dihydroethidium (DHE; Sigma-Aldrich) at 37˚C for
1 h. Cells were washed and stained with anti-CD44–allophycocyanin and
anti-CD4–FITC or anti-CD8–FITC. The mean fluorescence intensity of the
PE channel represented the DHE staining activity. The samples were analyzed using an FACStar flow cytometer (BD Biosciences).
Western blot
The T lymphocytes from Atg3f/f or Atg3f/fLck-Cre mice were enriched
using a mouse T cell enrichment kit (StemCell Technologies). Cell purity
was determined by flow cytometry to be .95%. Total cell lysates were
prepared in sample buffer (50 mM Tris-Cl [pH 6.8], 0.5% 2-ME, 2% SDS,
0.2% bromophenol blue, and 10% glycerol). The samples were separated
by SDS-PAGE and then transferred to Immobilon-FL polyvinylidene
difluoride membranes (Millipore, Billerica, MA). Membranes were blocked
with 5% nonfat milk and then incubated with primary Ab in PBS contain-
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 4. Decreased mature T cells in Atg3f/fLck-Cre mice. A, Total splenocytes in Atg3f/fLck-Cre mice. Total cell numbers from four Atg3f/fLck-Cre
mice, three Atg3f/+Lck-Cre, and five Atg3f/f mice (mean 6 SD). B, Frequency of CD4+ and CD8+ subpopulations in Atg3f/fLck-Cre mice. C, Numbers of
CD4+ and CD8+ T cells in Atg3f/fLck-Cre mice. Data are presented as mean 6 SD from three mice of each group. D, Memory-like cell populations in
Atg3f/fLck-Cre mice. Splenocytes or LN cells were stained with anti-CD4–allophycocyanin (or anti-CD8–allophycocyanin), anti-CD44–FITC, antiCD62L–PE, and 7-AAD. Samples were gated on 7-AAD2 cells. Numbers represent the frequency of each gated subpopulation. E, Numbers of CD44high
CD62Llow population in Atg3-deficient mice. A total of four Atg3f/f, four Atg3f/fLck-Cre, and five Atg3f/+Lck-Cre mice were analyzed. *p = 0.013, **p =
0.0067, ***p = 0.33 for CD4+ cells and p = 0.15 for CD8+ T cells.
The Journal of Immunology
ing 3% BSA (Sigma-Aldrich) and 0.5% Tween 20 at 4˚C overnight. The
next day, the membranes were washed and incubated with Alexa Fluor 680labeled anti-rabbit Ab (Invitrogen) or IRDye 800-labeled anti-goat Ab
(Rockland Immunochemicals, Gilbertsville, PA) at room temperature for 1 h
in PBS containing 3% BSA and 0.5% Tween 20. The blots were visualized
using the Odyssey Infrared Imaging System and analyzed using Odyssey
software (LI-COR Bioscience, Lincoln, NE). Atg3 rabbit polyclonal Ab was
from Sigma-Aldrich. Anti-actin Ab was from Santa Cruz Biotechnology
(Santa Cruz, CA). Anti-LC3 Ab (anti-LC3A/B, strong reactivity with LC3A/
B-II) was from Cell Signaling Technology (Danvers, MA).
Proliferation assay
The splenocytes from Atg3f/f or Atg3f/fLck-Cre mice were incubated with 5
mM CFSE (Invitrogen) in 5% FCS-PBS for 5 min at room temperature.
Cells were washed with 5% FCS-PBS three times and stimulated with
soluble anti-CD3 (2C11; 5 mg/ml) or anti-CD3 plus anti-CD28 (2 mg/ml;
Biolegend) for 72 h. The cells were then stained with anti-CD4–allophycocyanin, anti-CD8–allophycocyanin/Cy7, and 7-AAD in 2.4G2 supernatant on ice for 15 min. After washing, the samples were analyzed by flow
cytometry. The 7-AAD–negative cells were gated and analyzed. Unstimulated cells were used to gate the CFSE-diluted cells.
Two-tailed Student t tests were used to compare the means of different
samples.
Cre CD4+ or CD8+ T lymphocytes by PCR amplification of the
sequence from exon 8 and exon 9. The deletion efficiency was
.90% in CD4+ T cells and .80% in CD8+ T cells at the genomic
level (Fig. 1C). The deletion efficiency was further confirmed at
the protein level by Western blot. Atg3 protein was absent in
Atg3f/fLck-Cre thymocytes and purified Atg3f/fLck-Cre splenic
naive T cells (Fig. 1D, left panel). Furthermore, in the Atg3f/fERCre inducible deletion model, Atg3 was efficiently deleted in
Atg3f/fER-Cre splenocytes upon treatment with 4-OHT for
3 d (Fig. 1D, right panel). Together, these data demonstrate the
successful generation of T lymphocyte-specific (Atg3f/fLck-Cre)
and inducible (Atg3f/fER-Cre) models of deletion of Atg3.
Atg3 is essential for autophagy in T lymphocytes
Our previous data showed that TCR stimulation induces autophagy
in T lymphocytes (20). To determine whether the autophagy
pathway remains intact or is impaired in Atg3-deficient T cells,
splenocytes from Atg3f/fLck-Cre and control mice were stimulated
with anti-CD3, and LC3 puncta were quantified by intracellular
staining using anti-LC3 Ab. The number of LC3 puncta in CD3stimulated Atg3-deficient CD4+ T cells was much lower than that
in control CD4+ T cells (Fig. 2A, 2B). Similar results were observed in Atg3-deficient CD8+ T lymphocytes (data not shown).
Results
Generation of Atg3 conditional knockout mice
Our previous data from Atg5 knockout fetal liver transfer experiments and Atg7f/fLck-Cre mice suggested that autophagy is
essential for T lymphocyte survival (25) and that autophagy
mediates developmental regulation of organelle homeostasis (20,
21). Two ubiqutin-like pathways direct the process of autophagy,
and the results from studying Atg5- or Atg7-deficient T lymphocytes may only reflect the requirement of one of these pathways
in T lymphocyte survival. To investigate the function of the LC3
pathway in T lymphocytes, an Atg3 conditional knockout mouse
line was generated. The gene encoding Atg3, which is located on
chromosome 16, is 30,895 bp long, and the mRNA transcript is
2014 bp. Atg3 consists of 12 exons encoding 315 aas. Because
another gene, named Slc35a5, oriented in the opposite direction is
located 200 bp upstream of Atg3 on chromosome 16, the deletion
of exon 1 of Atg3 may affect the function of this upstream gene.
Atg3 is an E2-like molecule, and its E2-like conjugating enzyme
active site cysteine is encoded by exon 10 (30). We thus chose to
target the 1345 bp region beginning upstream of exon 8 and
ending in the intron region downstream of exon 10 (Fig. 1A). The
deletion of exon 8 to exon 10 results in an out-of-frame mutation
in exon 11 and the introduction of a stop codon in exon 11. In
total, this targeting strategy results in the deletion of ∼100 aa.
As shown in Fig. 1A, two loxP sites were inserted upstream of
exon 8 and downstream of exon 10. As a result, we expected
a genomic deletion of 1345 bp of Atg3 in Cre-expressing cells.
EF1 ES cells in which positive homologous recombination occurred between the targeting construct and the Atg3 genomic locus were selected by PCR screening for the loxP site downstream
of exon 10. Six positive ES clones were further confirmed by PCR
amplification of products containing the loxP site upstream of
exon 8 (Fig. 1B, left panel). The germline transmitted mice were
screened and bred with FLPeR mice to delete the neomycin resistance cassette. Mice homologous for the floxed Atg3 allele
(Atg3f/f, Fig. 1B, right panel) were then bred with Lck-Cre mice to
generate Atg3f/fLck-Cre mice, in which Atg3 is specifically deleted in T lymphocytes, or bred with ER-Cre mice to generate
Atg3f/fER-Cre mice, in which Atg3 can be inducibly deleted. The
efficiency of deletion of Atg3 was analyzed in sorted Atg3f/fLck-
FIGURE 5. Survival defect in Atg3-deficient T cells. A, Freshly isolated
splenocytes from Atg3f/fLck-Cre and Atg3f/f control mice were stained
with anti-CD4–allophycocyanin, anti-CD8–PE/Cy7, anti-CD44–FITC,
Annexin V-PE, and 7-AAD. Annexin V and 7-AAD staining was analyzed
in CD4+ and CD8+ subpopulations. Numbers represent the percentage of
each gated population. Data are representative of three experiments. B, Increased caspase-9 activity in Atg3-deficient T cells. Splenocytes from
Atg3f/fLck-Cre and Atg3f/f mice were incubated with caspase-9 FLICA
and analyzed by flow cytometry. The cells were gated on CD44low cell
population. Data are representative of three experiments. C, The survival
defect in Atg3-deficient T cells is partially rescued by zVAD. Splenocytes
from Atg3f/fLck-Cre and Atg3f/f mice were cultured in RPMI 1640 medium containing 10% FBS with or without IL-7 (1 ng/ml) or IL-7 in
combination with zVAD (20 mM) for 48 h. The cells were stained with
anti-CD4–allophycocyanin, anti-CD8–FITC, anti-CD44–allophycocyanin/
Cy7, and 7-AAD. The 7-AAD+ cells are dead cells (*p , 0.05, zVAD
treatment versus RPMI culture). CD44Hi, CD44high; CD44Lo, CD44low.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Statistical analysis
5317
5318
Consistent with the LC3 punctum formation results, the level of
LC3 II was also reduced in Atg3-deficient T cells under starvation
conditions (Fig. 2C).
It was reported that the LC3-binding protein p62/SQSTM1 is
constantly degraded by the autophagy pathway and accumulates in
autophagy-deficient cells (31, 32). p62 was accumulated in Atg3deficient T lymphocytes, whereas Atg3 wild-type naive T cells
expressed background levels of p62 (Fig. 2D). The ultrastructure
of Atg3-deficient T cells was further analyzed by EM. EM pictures show that some membrane structures that are not characteristic of autophagosomes accumulated in Atg3-deficient T cells
(Fig. 2E, 2F). The above results demonstrate that Atg3 is required
for autophagy in T lymphocytes.
Atg3 IN T LYMPHOCYTE FUNCTION
Atg3 is essential for T lymphocyte survival
Atg3f/fLck-Cre mice appear similar to Atg3f/f littermates. No
obvious abnormalities were found in Atg3f/f mice, suggesting that
the mutations introduced in the targeting construct in the intron
region between exon 7 and exon 8 do not affect the function of
Atg3. The thymic cellularity of Atg3f/fLck-Cre mice was ∼50% of
that of Atg3f/f and Atg3f/+Lck-Cre mice (Fig. 3A); however, the
frequencies of DN, double-positive, and single-positive thymocytes in Atg3f/f Lck-Cre mice were comparable to those in Atg3f/f
control mice (Fig. 3B, 3C). These results suggest that the loss of
Atg3 did not obviously alter thymocyte development. Together
with our previous findings in Atg7f/fLck-Cre mice (20) and Atg5
knockout fetal liver transfer experiments (25), these results are
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 6. Abnormal organelle homeostasis in Atg3-deficient T cells. A–C, Mitochondria in Atg3-deficient T cells. Splenocytes from Atg3f/fLck-Cre
and Atg3f/f control mice were incubated with 100 nM Mito-Tracker Green at 37˚C for 30 min. After washing, the splenocytes were stained with anti-CD4–
allophycocyanin, anti-CD8–PE/Cy7, anti-CD44–allophycocyanin/Cy7, and 7-AAD. Each subpopulation was gated on 7-AAD2 cells. A, FACS profiles of
Mito-Tracker Green staining of Atg3-deficient T cells. Shaded profiles, Atg3f/f T cells; open profiles, Atg3f/fLck-Cre T cells. B, The mean fluorescence
intensity (MFI) of Mito-Tracker Green from 3 pairs of Atg3f/fLck-Cre and Atg3f/f mice. C, The ratio of Mito Tracker MFIs in Atg3-deficient cell populations to those in corresponding wild-type cell populations. CD19+ cells were used as a negative control. The bar figures are mean 6 SD from six pairs of
Atg3f/f and Atg3f/fLck-Cre mice. D–F, Atg3-deficient T cells produce elevated levels of ROS. Splenocytes from Atg3f/fLck-Cre and Atg3f/f control mice
were incubated with DHE for 1 h at 37˚C. After washing, the cells were stained with Abs described above. The fluorescence intensity of the PE channel
represents ROS production. All samples were gated on 7-AAD2 cells. D, FACS profiles of DHE staining. Shaded profiles, Atg3f/f T cells; open profiles,
Atg3f/fLck-Cre T cells. E, MFIs of DHE staining. Shown is mean 6 SD from three pairs of Atg3f/fLck-Cre and Atg3f/f mice. F, The ratio of DHE MFIs in
Atg3-deficient cells to those in corresponding wild-type cells. Data are mean 6 SD from five pairs of mice. CD19+ cell population was used as a negative
control. G–I, Expanded ER in Atg3-deficient mice. Splenocytes from Atg3f/fLck-Cre and Atg3f/f control mice were incubated with 1 mM ER-Tracker BlueWhite DPX at 37˚C for 30 min. The cells were stained and analyzed as described above. G, FACS profiles of ER-Tracker Blue-White staining. Shaded
profiles, Atg3f/f T cells; open profiles, Atg3f/fLck-Cre T cells. H, The MFIs of ER-Tracker Blue-White staining. Data are mean 6 SD from three pairs of
Atg3f/fLck-Cre and Atg3f/f mice. I, The ratio of ER-Tracker Blue-white MFIs in Atg3-deficeint cells to those in corresponding wild-type cell populations.
Data are mean 6 SD from five pair of Atg3f/fLck-Cre and Atg3f/f mice. CD19+ cell population is used as a negative control. *p , 0.01, **p , 0.05, ***p .
0.05, Atg3f/fER-Cre to Atg3f/f.
The Journal of Immunology
Atg3-deficient T cells, we next analyzed reactive oxygen species
(ROS) production in Atg3f/fLck-Cre mice. Naive Atg3-deficient
CD4+ or CD8+ T cells produced more ROS than corresponding
wild-type control cells (Fig. 6D, 6F), suggesting that the abnormal
expansion of mitochondria in the absence of Atg3 may drive
overproduction of ROS.
Atg3-deficient T cells cannot proliferate efficiently
To further assess the function of Atg3-deficient T cells, splenocytes
from Atg3f/fLck-Cre and Atg3f/f control mice were stimulated
with anti-CD3 Ab. All analyzed cell populations were gated on 7AAD–negative cells to exclude dead cells. CFSE dilution was
only analyzed in gated live cells. Both CD4+-AAD2 and CD8+7AAD2 cells from wild-type Atg3f/f mice proliferated efficiently
(Fig. 7A). The frequency of proliferating cells was .90% when
wild-type cells were stimulated with soluble anti-CD3 Ab alone or
in combination with anti-CD28 Ab. In contrast, Atg3-deficient
CD4+7-AAD2 and CD8+7-AAD2 T cells did not proliferate efficiently; the frequency of proliferating cells among these populations was ,60% (Fig. 7A). Although Atg3-deficient T cells did
Atg3 regulate organelle homeostasis in T lymphocytes
Using a T cell-specific Atg7 deletion model, we previously showed
that autophagy mediates the developmental regulation of mitochondrial content (20) and ER homeostasis (21). We thus analyzed
mitochondrial and ER homeostasis in the Atg3 T cell-specific
deletion system. Splenocytes from Atg3f/fLck-Cre mice were
stained with the cell permeable mitochondria- or ER-specific
staining reagents Mito-Tracker Green or ER-Tracker Blue-White,
respectively. As shown in Fig. 6A, 6B, 6G, and 6H, both mitochondria and ER were expanded in Atg3-deficient mature CD4+
or CD8+ naive T lymphocytes. The mitochondria and ER in Atg3deficient naive T cells were increased to .150% of wild-type
control cells as shown in Fig. 6C and 6I. These data suggest that
inhibition of autophagy in these cells affected mitochondrial and
ER homeostasis. Because the mitochondria were expanded in
FIGURE 7. Defective proliferation in Atg3-deficient T lymphocytes. A,
Splenocytes from Atg3f/f and Atg3f/fLck-Cre mice were loaded with CFSE
and stimulated with soluble anti-CD3 (5 mg/ml) or soluble anti-CD3 plus
anti-CD28 (2 mg/ml) for 72 h or were left unstimulated as resting cell
CFSE control. After staining for surface markers and 7-AAD, the samples
were analyzed by flow cytometry. The cells without any stimulation were
used to define the CFSE dilution gate. Numbers inside the figures represent
the frequency of CFSE-diluted cells. B, Activation of Atg3-deficient
T cells. Splenocytes from Atg3f/f and Atg3f/fLck-Cre mice were stimulated
with soluble anti-CD3 Ab (5 mg/ml) plus anti-CD28 Ab overnight. The
cells were stained with anti-CD4–allophycocyanin, anti-CD8–allophycocyanin/Cy7, anti-CD25–FITC, and anti-CD69–PE and analyzed by flow
cytometry. Numbers represent the frequency of cells within each gate.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
consistent with the observations that thymocyte development is
largely normal in the absence of autophagy.
Splenic cellularity was decreased in Atg3f/fLck-Cre mice. The
total splenic cellularity of Atg3f/fLck-Cre mice was ∼60% of that
of control Atg3f/f mice (Fig. 4A). Furthermore, the frequency and
absolute number of both CD4+ and CD8+ splenic T cells were
decreased in Atg3f/fLck-Cre mice. The frequency and number of
CD4+ T cells in Atg3f/fLck-Cre mice were 50 and 30%, respectively, of those in the control mice. The frequency and number
of CD8+ T cells in Atg3f/fLck-Cre mice were 20 and 10%, respectively, of those in the control mice (Fig. 4B, 4C). The splenic
cellularity and the number of CD4+ or CD8+ T cells in Atg3f/+
Lck-Cre mice were not significantly different from those in Atg3f/f
mice (Fig. 4A, 4C).
We next examined the CD44 and CD62L expression patterns
in Atg3f/fLck-Cre splenocytes and LN cells. The frequency of
memory-like T cells (CD62Llow and CD44high) among splenocytes
and LN cells was increased within both the CD4+ and CD8+
compartment of Atg3f/fLck-Cre mice but not significantly changed
in Atg3f/+Lck-Cre mice (Fig. 4D, 4E). These results were consistent with the lymphopenic environment in Atg3f/fLck-Cre mice.
The above results suggest that survival is defective in Atg3deficient T cells. To test this hypothesis, freshly isolated splenocytes were stained with Annexin V and 7-AAD. As shown in Fig.
5A, the frequency of Annexin V+ cells among naive (CD44low)
CD4+ or CD8+ T cells was two times higher in Atg3f/fLck-Cre
mice than that in Atg3f/f mice. To further examine the apoptotic
defect in Atg3-deficient T cells, the active form of caspase-9 was
analyzed in freshly isolated T cells from Atg3f/fLck-Cre and
Atg3f/f mice. Increased active caspase-9 was found in Atg3f/fLckCre T cells (Fig. 5B). These data suggest that the impaired survival
of Atg3-deficient T cells is related to caspase activation.
The impaired survival in Atg3-deficient T cells was further
characterized in vitro. To decrease the spontaneous T lymphocyte
death, IL-7 was added in the culture. IL-7 greatly enhanced the
survival of Atg3f/f wild-type T cells in vitro compared with RPMI
1640 medium only culture (Fig. 5C). More dead cells were found
in Atg3-deficient T cells in in vitro culture. The frequency of
apoptosis among Atg3-deficient T cells was greater than that
among wild-type cells even in the presence of IL-7 (Fig. 5C). To
further characterize the cell death of Atg3-deficient T cells, the
pan-caspase inhibitor zVAD was used in cell survival assay. As
shown in Fig. 5C, zVAD partially rescued the survival defect in
Atg3-deficient T cells in the presence of IL-7, suggesting that the
death of Atg3-deficient T cells may be mediated by the apoptosis
pathway (Fig. 5C). Collectively, these results demonstrate that
Atg3 and autophagy are essential for the survival of T lymphocytes.
5319
5320
not proliferate as well as wild-type T cells, Atg3-deficient CD4+
and CD8+ T cells expressed CD69 and CD25 after anti-CD3 Ab
stimulation (Fig. 7B). These data indicate that autophagy may
regulate T lymphocyte proliferation but that it does not affect their
activation.
Cumulative defects in autophagy-deficient T cells
FIGURE 8. Cumulative defects in Atg3-deficient
T cells. A, Survival of T lymphocytes in short-term
culture after inducible deletion of Atg3. Splenocytes
from Atg3f/fER-Cre and Atg3f/f control mice were
treated with 200 nM 4-OHT or equal volume of ethanol
(ETOH) in vitro for 3 d with or without IL-7 (1 ng/ml).
The cells were then stained with anti-CD4–allophycocyanin, anti-CD8–PE/Cy7, and anti-CD44–APC/Cy7.
After washing, the cells were resuspended in binding
buffer and stained with Annexin V-PE and 7-AAD and
analyzed by flow cytometry. B, Increased T cell death
in long-term culture after inducible deletion of Atg3.
4-OHT–treated Atg3f/fER-Cre and control splenocytes
were continually cultured in complete RPMI 1640
medium. IL-7 (1 ng/ml) was added every other day. On
day 24, the cells were stained for surface markers and
7-AAD as described in A. Samples were gated on CD4+
CD44low naive T cells, and the frequency of dead cells
was calculated based on 7-AAD staining. The results
from three pairs of Atg3f/fER-Cre and Atg3f/f wild-type
control mice are presented as mean 6 SD. *p , 0.05,
Atg3f/fER-Cre to Atg3f/f.
deletion of Atg3 in the presence of IL-7 showed no difference in
Annexin V and 7-AAD staining between CD4+ and CD8+ T cells
from Atg3f/f and Atg3f/fER-Cre mice (data not shown). In contrast,
when the cells were cultured for .24 d, ∼13% of Atg3-deficient
CD4+ T cells and 25% of Atg3-deficient CD8+ T cells were 7AAD+, whereas ,3% of control CD4+ and 5% of CD8+ T cells
were 7-AAD+ (Fig. 8B). These results suggest that although acute
loss of Atg3 does not affect T cell survival, accumulation of defects during long-term cultures may result in the death of autophagy-deficient T cells.
Experiments using the Lck-Cre system have demonstrated that
autophagy is essential in maintaining the homeostasis of mitochondria and ER in T cells. As the Lck-Cre induced deletion of
autophagy genes occurs at early stage of thymocyte development,
the defects in autophagy-deficient T cells may be due to accumulative effects. To investigate the accumulation of autophagy
defect, the organelle homeostasis was thus analyzed in the ER-Cre
inducible deletion system. No change in mitochondrial volume was
detected on day 3 after inducible deletion of Atg3; however, mitochondrial volume began to increase on day 10 and further expanded by days 18 and 28 after inducible deletion of Atg3 in naive
CD4+ or CD8+ T cells. Similarly, no change in ER content was
detected on day 3 after inducible deletion of Atg3; however, ER
volume was expanded in Atg3f/fER-Cre CD4+ and CD8+ T cells
on day 10 after inducible deletion of Atg3 (Fig. 9). These results
demonstrate that excess mitochondria and ER accumulate over
time in Atg3-deficient T cells and suggest that these cumulative
defects in organelle homeostasis may cause cell death in Atg3and autophagy-deficient naive T cells.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Our previous data and the above results indicate that autophagy
is essential for T cell survival and that autophagy mediates the
regulation of organelle homeostasis. These data were all obtained
using T cell-specific deletion models or Atg5 knockout fetal liver
transfer experiments. To further investigate the functions of
autophagy in mature T cells, Atg3f/f mice were bred with ER-Cre
mice to generate Atg3f/fER-Cre mice, in which Atg3 can be
inducibly deleted. As shown in Fig. 1D (right panel), Atg3 was
deleted in ER-Cre+ cells upon in vitro treatment with 200 nM 4OH tamoxifen for 3 d. After 4-OHT treatment, apoptosis was
analyzed in Atg3-deficient naive CD4+CD44low or CD8+ CD44low
T cells. On day 3 after 4-OHT treatment, the majority of naive
T cells from both Atg3f/f and Atg3f/fER-Cre mice were apoptotic.
More than 50% of the naive CD4+ or CD8+ T cells were Annexin
V+, and ∼50% of the naive CD4+ cells and 60% of the naive CD8+
T cells were 7-AAD+ (Fig. 8A, upper left panel). No significant
differences were found between 4-OHT–treated Atg3f/f cells and
Atg3f/fER-Cre cells (Fig. 8A). This spontaneous death of naive
T cells during the 3-d culture period was rescued by the addition
of 1 ng/ml IL-7 (Fig. 8A, upper right panels). The results of carrier
control ethanol treatment were shown in Fig. 8A, lower panels.
Analysis of apoptosis on days 7, 10, and 18 after 4-OHT–mediated
Atg3 IN T LYMPHOCYTE FUNCTION
The Journal of Immunology
Discussion
We have previously shown that T lymphocytes lacking Atg5 or
Atg7 have impaired survival and disrupted organelle homeostasis
(20, 21, 25), suggesting that autophagy regulates T lymphocyte
homeostasis (33). However, these data could reflect autophagyindependent functions of Atg5 or Atg7. Furthermore, Atg5 and
Atg7 are both involved in the same conjugation pathway even
though Atg7 also participates in the second conjugation pathway.
Therefore, the defects in Atg5- or Atg7-deficient T cells might be
due to the defect from one conjugation pathway. Thus, it is important to test the function of the second conjugation pathway of
autophagy in T cell survival. To this end, we have generated Atg3
conditional knockout mice.
Our results show that Atg3-deficient T cells have an in vivo
phenotype similar to that of Atg5- and Atg7-deficient T cells.
Mature Atg3-deficient T cells display increased cell death and
expansion of mitochondria and ER. Atg3 conditional KO mice have
more memory-like phenotype T lymphocytes expressing CD62Llow
CD44high. This is consistent with a well-described phenomenon
termed lymphopenia-driven proliferation that occurs in many
T cell-deficient mouse strains (34). Similar to other mammalian
cells, the Atg3 is indispensable for autophagy in T lymphocytes.
In this study, we used an ER-Cre inducible deletion system to
address the acute and cumulative effects of deletion of Atg3 in
naive T cells. Interestingly, we found no survival defect in naive
T cells until 24 d after deletion of Atg3, suggesting that the acute
deletion of Atg3 does not affect these cells; rather, accumulation
of abnormally expanded organelles causes the increased cell death
observed in Atg3-deficient naive T cells.
Previous studies have shown that the Atg12 and LC3 conjugation
systems regulate each other. The Atg12 pathway regulates the LC3
pathway, and the LC3 pathway is required for the maturation of the
autophagosome (35). Atg12–Atg5 conjugates function as a ubiquitin-protein ligase (E3)-like enzyme and enhance the generation
of LC3- phosphatidylethanolamine (10, 12). Atg3 interacts with
Atg12 and facilitates the formation of the Atg12–Atg5 complex
(30). Our data suggest that Atg3 deficiency impairs autophagy in
T lymphocytes, resulting in dysregulation of mitochondrial and
ER homeostasis. A recent publication showed that Atg12 conjugates with Atg3, and Atg12–Atg3 complex regulates mitochondria homeostasis through a mitophagy-independent pathway
in mouse embryonic fibroblasts (36). Our results reported in this
study, together with our previous findings in Atg5- or Atg7deficient T cells, indicate that autophagy pathways are essential
for maintaining lymphocyte physiological function.
The Atg3f/fER-Cre inducible deletion system provides a novel
method to further investigate the functions of autophagy in
T lymphocytes. After inducible deletion of Atg3, no acute defects
were observed in T cell function. T cells in which Atg3 had been
inducibly deleted survived similar to wild-type T cells, and mitochondria and ER were not expanded at early time points;
however, if the Atg3-deficient T lymphocytes were cultured for
longer periods of time (.2 to 3 wk), the organelles were gradually
expanded, and the frequency of cell death began to increase. These
effects may be related to the basic functions of autophagy. Generally, the physiological function of autophagy is to provide cells
with energy during starvation, growth factor withdrawal, or other
stress conditions and to remove long-lived proteins or damaged,
aged, or excess organelles. The temporal accumulation of mitochondria and ER after deletion of Atg3 suggests that the survival
and proliferation defects observed in Atg5-, Atg7-, and Atg3deficient T cells are indirect effects of impaired autophagy. After expanded organelles accumulate to a certain level, the cells
may not be able to compensate for these abnormalities, and the
survival defects become apparent.
The relationships among autophagy, cell death, and cell survival
are complex (37). Our previous data and the results presented in
this study demonstrate that autophagy is essential for the survival
of T lymphocytes (20, 25). IL-7 upregulates the expression of Bcl2 to promote the survival of T lymphocytes (38); however, IL-7
does not rescue the survival defect in Atg3-deficient T cells. The
enhanced activation of caspase-9 and the fact that pan-caspase
inhibitor zVAD partially rescues the survival defect in Atg3deficient T cells suggest that the cell death in autophagydeficient T cells is due to the abnormal activation of the apoptosis pathway; however, other factors related to impaired autophagy
may also affect the survival of T lymphocytes. Interestingly,
abolishing the Atg12–Atg3 conjugates inhibits cell death by increasing the expression of Bcl-2 member Bcl-xL in mouse embryonic fibroblasts (36). However, this is independent of autophagy because the disruption of the conjugates of the Atg12 and
Atg3 complex does not affect the starvation-induced autophagy
(36).
In summary, our results indicate that Atg3 and LC3 conjugation
are essential for autophagy in T lymphocytes. The homeostasis of
organelles in Atg3-deficient T cells is abnormal, and the survival
of Atg3-deficient T cells is defective. By using an inducible deletion model, we have provided clear evidence that the survival defect
and organelle expansion phenotypes in Atg3-deficient T cells are
temporally accumulated effects of impaired autophagy.
Acknowledgments
We thank Dr. Michael Cook, Lynn Martinek, and Nancy Martin for helping
with the flow cytometry and sorting the cells, Philip Christopher for help in
transmission EM, and Claire L. Gordy for editing and critically reading the
manuscript.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
FIGURE 9. Organelles expand in a temporal manner after inducible
deletion of Atg3 in T cells. Splenocytes from Atg3f/fER-Cre and Atg3f/f
control mice were treated with 200 nM 4-OHT or carrier control ethanol
(ETOH) for 3 d in the presence of IL-7 (1 ng/ml). The cells were continually cultured in IL-7-containing medium. The mitochondria (A) and ER
(B) were stained as described in Fig. 6 at different time points. All cells
were gated on CD44low population. Data are mean 6 SD from three pairs
of Atg3f/fER-Cre and Atg3f/f mice and representative of two experiments.
*p , 0.05, **p , 0.01 Atg3f/f ER-Cre to Atg3f/f.
5321
5322
Disclosures
The authors have no financial conflicts of interest.
References
20. Pua, H. H., J. Guo, M. Komatsu, and Y. W. He. 2009. Autophagy is essential for
mitochondrial clearance in mature T lymphocytes. J. Immunol. 182: 4046–4055.
21. Jia, W., H. H. Pua, Q. J. Li, and Y. W. He. 2011. Autophagy regulates endoplasmic reticulum homeostasis and calcium mobilization in T lymphocytes. J.
Immunol. 186: 1564–1574.
22. Kundu, M., T. Lindsten, C. Y. Yang, J. Wu, F. Zhao, J. Zhang, M. A. Selak,
P. A. Ney, and C. B. Thompson. 2008. Ulk1 plays a critical role in the autophagic
clearance of mitochondria and ribosomes during reticulocyte maturation. Blood
112: 1493–1502.
23. Zhang, J., M. S. Randall, M. R. Loyd, F. C. Dorsey, M. Kundu, J. L. Cleveland,
and P. A. Ney. 2009. Mitochondrial clearance is regulated by Atg7-dependent
and -independent mechanisms during reticulocyte maturation. Blood 114: 157–
164.
24. Mortensen, M., D. J. Ferguson, M. Edelmann, B. Kessler, K. J. Morten,
M. Komatsu, and A. K. Simon. 2010. Loss of autophagy in erythroid cells leads
to defective removal of mitochondria and severe anemia in vivo. Proc. Natl.
Acad. Sci. USA 107: 832–837.
25. Pua, H. H., I. Dzhagalov, M. Chuck, N. Mizushima, and Y. W. He. 2007. A
critical role for the autophagy gene Atg5 in T cell survival and proliferation. J.
Exp. Med. 204: 25–31.
26. Zhang, N., and Y. W. He. 2005. An essential role for c-FLIP in the efficient
development of mature T lymphocytes. J. Exp. Med. 202: 395–404.
27. Farley, F. W., P. Soriano, L. S. Steffen, and S. M. Dymecki. 2000. Widespread
recombinase expression using FLPeR (flipper) mice. Genesis 28: 106–110.
28. Hennet, T., F. K. Hagen, L. A. Tabak, and J. D. Marth. 1995. T-cell-specific
deletion of a polypeptide N-acetylgalactosaminyl-transferase gene by sitedirected recombination. Proc. Natl. Acad. Sci. USA 92: 12070–12074.
29. Vooijs, M., J. Jonkers, and A. Berns. 2001. A highly efficient ligand-regulated
Cre recombinase mouse line shows that LoxP recombination is position dependent. EMBO Rep. 2: 292–297.
30. Tanida, I., E. Tanida-Miyake, M. Komatsu, T. Ueno, and E. Kominami. 2002.
Human Apg3p/Aut1p homologue is an authentic E2 enzyme for multiple substrates, GATE-16, GABARAP, and MAP-LC3, and facilitates the conjugation of
hApg12p to hApg5p. J. Biol. Chem. 277: 13739–13744.
31. Bjørkøy, G., T. Lamark, A. Brech, H. Outzen, M. Perander, A. Overvatn,
H. Stenmark, and T. Johansen. 2005. p62/SQSTM1 forms protein aggregates
degraded by autophagy and has a protective effect on huntingtin-induced cell
death. J. Cell Biol. 171: 603–614.
32. Komatsu, M., and Y. Ichimura. 2010. Physiological significance of selective
degradation of p62 by autophagy. FEBS Lett. 584: 1374–1378.
33. Pua, H. H., and Y. W. He. 2009. Autophagy and lymphocyte homeostasis. Curr.
Top. Microbiol. Immunol. 335: 85–105.
34. Takada, K., and S. C. Jameson. 2009. Naive T cell homeostasis: from awareness
of space to a sense of place. Nat. Rev. Immunol. 9: 823–832.
35. Mizushima, N., A. Yamamoto, M. Hatano, Y. Kobayashi, Y. Kabeya, K. Suzuki,
T. Tokuhisa, Y. Ohsumi, and T. Yoshimori. 2001. Dissection of autophagosome
formation using Apg5-deficient mouse embryonic stem cells. J. Cell Biol. 152:
657–668.
36. Radoshevich, L., L. Murrow, N. Chen, E. Fernandez, S. Roy, C. Fung, and
J. Debnath. 2010. ATG12 conjugation to ATG3 regulates mitochondrial homeostasis and cell death. Cell 142: 590–600.
37. Baehrecke, E. H. 2005. Autophagy: dual roles in life and death? Nat. Rev. Mol.
Cell Biol. 6: 505–510.
38. Chetoui, N., M. Boisvert, S. Gendron, and F. Aoudjit. 2010. Interleukin-7 promotes the survival of human CD4+ effector/memory T cells by up-regulating
Bcl-2 proteins and activating the JAK/STAT signalling pathway. Immunology
130: 418–426.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
1. Klionsky, D. J., and S. D. Emr. 2000. Autophagy as a regulated pathway of
cellular degradation. Science 290: 1717–1721.
2. Ohsumi, Y. 2001. Molecular dissection of autophagy: two ubiquitin-like systems.
Nat. Rev. Mol. Cell Biol. 2: 211–216.
3. Mizushima, N., Y. Ohsumi, and T. Yoshimori. 2002. Autophagosome formation
in mammalian cells. Cell Struct. Funct. 27: 421–429.
4. Levine, B., and D. J. Klionsky. 2004. Development by self-digestion: molecular
mechanisms and biological functions of autophagy. Dev. Cell 6: 463–477.
5. Nakatogawa, H., K. Suzuki, Y. Kamada, and Y. Ohsumi. 2009. Dynamics and
diversity in autophagy mechanisms: lessons from yeast. Nat. Rev. Mol. Cell Biol.
10: 458–467.
6. Kanki, T., and D. J. Klionsky. 2010. The molecular mechanism of mitochondria
autophagy in yeast. Mol. Microbiol. 75: 795–800.
7. Glick, D., S. Barth, and K. F. Macleod. 2010. Autophagy: cellular and molecular
mechanisms. J. Pathol. 221: 3–12.
8. Geng, J., and D. J. Klionsky. 2008. The Atg8 and Atg12 ubiquitin-like conjugation systems in macroautophagy. ‘Protein modifications: beyond the usual
suspects’ review series. EMBO Rep. 9: 859–864.
9. Ichimura, Y., T. Kirisako, T. Takao, Y. Satomi, Y. Shimonishi, N. Ishihara,
N. Mizushima, I. Tanida, E. Kominami, M. Ohsumi, et al. 2000. A ubiquitin-like
system mediates protein lipidation. Nature 408: 488–492.
10. Hanada, T., N. N. Noda, Y. Satomi, Y. Ichimura, Y. Fujioka, T. Takao, F. Inagaki,
and Y. Ohsumi. 2007. The Atg12-Atg5 conjugate has a novel E3-like activity for
protein lipidation in autophagy. J. Biol. Chem. 282: 37298–37302.
11. Fujita, N., T. Itoh, H. Omori, M. Fukuda, T. Noda, and T. Yoshimori. 2008. The
Atg16L complex specifies the site of LC3 lipidation for membrane biogenesis in
autophagy. Mol. Biol. Cell 19: 2092–2100.
12. Tanida, I., T. Nishitani, T. Nemoto, T. Ueno, and E. Kominami. 2002. Mammalian Apg12p, but not the Apg12p.Apg5p conjugate, facilitates LC3 processing. Biochem. Biophys. Res. Commun. 296: 1164–1170.
13. Sou, Y. S., S. Waguri, J. Iwata, T. Ueno, T. Fujimura, T. Hara, N. Sawada,
A. Yamada, N. Mizushima, Y. Uchiyama, et al. 2008. The Atg8 conjugation
system is indispensable for proper development of autophagic isolation membranes in mice. Mol. Biol. Cell 19: 4762–4775.
14. Nishida, Y., S. Arakawa, K. Fujitani, H. Yamaguchi, T. Mizuta, T. Kanaseki,
M. Komatsu, K. Otsu, Y. Tsujimoto, and S. Shimizu. 2009. Discovery of Atg5/
Atg7-independent alternative macroautophagy. Nature 461: 654–658.
15. Deretic, V. 2006. Autophagy as an immune defense mechanism. Curr. Opin.
Immunol. 18: 375–382.
16. Lee, H. K., L. M. Mattei, B. E. Steinberg, P. Alberts, Y. H. Lee, A. Chervonsky,
N. Mizushima, S. Grinstein, and A. Iwasaki. 2010. In vivo requirement for Atg5
in antigen presentation by dendritic cells. Immunity 32: 227–239.
17. Paludan, C., D. Schmid, M. Landthaler, M. Vockerodt, D. Kube, T. Tuschl, and
C. Münz. 2005. Endogenous MHC class II processing of a viral nuclear antigen
after autophagy. Science 307: 593–596.
18. Münz, C. 2010. Antigen processing via autophagy—not only for MHC class II
presentation anymore? Curr. Opin. Immunol. 22: 89–93.
19. Lee, H. K., J. M. Lund, B. Ramanathan, N. Mizushima, and A. Iwasaki. 2007.
Autophagy-dependent viral recognition by plasmacytoid dendritic cells. Science
315: 1398–1401.
Atg3 IN T LYMPHOCYTE FUNCTION