The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Peng, Y.
Right arrow Articles by Elkon, K. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Peng, Y.
Right arrow Articles by Elkon, K. B.
The Journal of Immunology, 2000, 165: 1470-1478.
Copyright © 2000 by The American Association of Immunologists

Soluble CD8 Attenuates Cytotoxic T Cell Responses Against Replication-Defective Adenovirus Affording Transprotection of Transgenes In Vivo1

YuFeng Peng*, Erik Falck-Pedersen{dagger} and Keith B. Elkon2,*

Departments of * Medicine and {dagger} Microbiology and Immunology, Weill Medical College of Cornell University, New York, NY 10021


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
The T cell coreceptor, CD8, enhances T cell-APC interactions. Because soluble CD8{alpha} homodimers can antagonize CD8 T cell activation in vitro, we asked whether secretion of soluble CD8 would effect cytotoxic T cell responses in vivo. Production of soluble CD8 by a replication-defective adenovirus vector allowed persistent virus expression for up to 5 mo in C57BL/6 mice and protected a second foreign transgene from rapid deletion. Soluble CD8 selectively inhibited CD8 T cell proliferation and IFN-{gamma} production and could also attenuate peptide-specific CD8 T cell responses in vivo. These finding suggest that gene vector delivery of soluble CD8 may have therapeutic applications.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Innate and adaptive immunity are required for the elimination of virus-infected cells in vivo. Whereas CD4+ Th cells facilitate effector function through activation of APCs after ligation of CD40 (1, 2, 3), CD8+ T cells are predominantly responsible for cytotoxicity and elimination of replication-defective adenovirus (Ad)3 vectors that are used for experimental gene therapy applications (4, 5). CD8 is the coreceptor present on cytotoxic T cells. Engagement of CD8 by its natural binding partner, the {alpha}3 domain of MHC class I (MHC-I) present on all nucleated cells, lowers the threshold of T cell activation by enhancing the adhesion between CD8 T cells and their targets and by recruitment of lck kinase to the TCR/CD3 signaling complex, leading to phosphorylation of CD3 {zeta} chain (6, 7, 8, 9, 10, 11, 12, 13). We recently generated an Ad vector expressing the extracellular portion of murine CD8{alpha} that can be readily detected in the serum after Ad administration by the i.v. route (14). Despite the extremely low binding affinity of soluble CD8 (sCD8) for MHC-I, sCD has recently been shown to impair CD8 T cell activation and function in vitro (15). We report in this paper that an Ad vector expressing sCD8 attenuates the cytotoxic T cell response to Ad in vivo, thereby prolonging gene expression in C57BL/6 (B6) and C3H strains of mice.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Mice

B6 (H2b), BALB/c/J (H2d), and C3He/SnJ (H2k) mice were purchased from Jackson Laboratory, Bar Harbor, Maine. SCID mice were bred and maintained at the Hospital for Special Surgery (New York, NY) in a pathogen-free environment. OT-1 mice that express the transgenic TCR reactive with the OVA peptide (OVAp), OVA257–264, were kindly made available by Francis Carbone (Monash Medical School, Melbourne, Australia) (16).

Generation of replication-defective Ad expressing soluble murine proteins

The cDNAs encoding the ectodomain of murine CD8-{alpha} (Ly2.1) and Fas ectodomain were cloned from MRL+/+ mice. The chimeric Fas ectodomain was fused to the N terminus of the CD8 ectodomain by PCR, as described previously (14). The cDNAs were ligated between the HindIII and XhoI site of pAd.cmv-HS vector, and these viruses, as well as AdCAT, AdNull, produced as described elsewhere (14). Briefly, the pAd.cmv-HS vector containing the transgene was cotransfected with PJM17 into 293 cells. The cell lysate was used to infect 293 cells. Viral DNA was extracted by a modified Hirt assay, and recombination was verified by restriction enzyme digestion, PCR, and protein production (see below). The viruses were further plaque purified. Each clone was rescreened as above. Finally, large scale virus was purified by two-step CsCl concentration and stored in glycerol at -20°C or sucrose at -70°C. Quantitation of viral particles was measured at OD 260 nm.

Injection of mice with adenovirus vectors and evaluation of virus expression

Mice were injected with 2.5 x 10-10 particles of virus by the i.v. route unless indicated otherwise and were bled at 7- or 14-day intervals. Persistence of virus expression was determined by detection of the transgene product secreted into the serum (sCD8, FasCD8) or by protein expression in the liver (chloramphenicol acetyl transferase (CAT)) as described previously (17). In brief, CAT activity in liver homogenates was quantified by thin-layer chromatography of 14C-labeled substrate (17) and secreted transgene products by ELISA as described below. To verify active transcription of transgenes, total RNA was isolated from liver homogenates, and PCR amplification of the transgene was performed by using the following primers that were specific for Ad sequences immediately flanking the transgenes: 5'-CCCAGGTCCAACTGCAGCCC-3' and 5'-GGTACTTGTGAG CCAAGGCAG-3'. The amplification conditions were 95°C for 30 min, 55°C for 30 min, and 72°C for 50 min for 35 cycles.

ELISAs

The CD8 levels in culture supernatants or mouse serum were quantified by sandwich ELISAs utilizing two different mAbs to the mouse CD8-{alpha}-chain, TIB 105 hybridoma (American Type Culture Collection, Manassas, VA) and biotinylated mouse YST 169 (Caltag Laboratories, Burlingame, CA). ELISA plates were coated overnight with TIB 105 (5 µg/ml) at 4°C. The plates were blocked with PBS/3% BSA for 1 h at room temperature and then incubated with the 1/3 diluted serum sample for 4–5 h in room temperature. The plates were washed and sequentially incubated with biotinylated secondary Ab, avidin-alkaline phosphatase, and substrate. The OD was read at 405 nm. FasCD8 was quantified by a similar ELISA, except that plates were coated with the anti-Fas Ab Jo-2 (PharMingen, San Diego, CA) as described (14).

IFN-{gamma} was quantified using a similar sandwich ELISA (PharMingen) with R4-6A2 as capture Ab, biotinylated XMG1.2 as detecting Ab, and purified recombinant IFN-{gamma} as a standard (Life Technologies, Grand Island, NY). The levels of IgG anti-adenovirus Ab were determined by a solid-phase ELISA using purified Ad as Ag as described previously (17).

Abs and flow cytometry analysis

Flow cytometry analysis was performed using a FACscan with Cell Quest software (Becton Dickinson, Mountain View, CA) as described previously (18). mAbs against the following Ags were used for staining: CD4 (PharMingen), CD8 (PharMingen), CD44 (Caltag Laboratories), and V{alpha}2 (PharMingen). To detect intracellular cytokines, spleen cells from infected mice were restimulated (see below) for 4 days. On day 4, brefeldin A (10 µg/ml) was added for 5 h and then the cells were fixed and permeabilized according the manufacturer’s instructions (cytofix/cytoper plus kit; PharMingen). Cytokine expression was evaluated by two-color flow cytometry using anti-CD4 or -CD8 and anti-IFN-{gamma} (Caltag Laboratories).

Isolation of intrahepatic lymphocytes (IHLs) and limiting dilution assay

Intrahepatic lymphocytes were isolated according to Watanabe et al. (19). Briefly, livers from day 10-infected mice were perfused with 10 ml HBSS and cut into small pieces. The liver fragments were forced through a metal mesh, and the slurry was digested with liver digestion medium (Life Technologies) at 37°C for 40 min before washing. The cell pellet was resuspended in medium (RPMI 1640 supplemented with 10% FCS), and the IHLs were isolated by Ficoll density centrifugation.

To estimate the frequency of specific percursor CTLs in the spleen, single-cell suspensions were prepared from day 10-infected mice, and 2-fold dilutions of responder cells were incubated in 24 wells of 96-well U-bottom plates. Irradiated day 7-infected syngeneic spleen cells were added at 2 x 105/well. Mixed lymphocyte cultures were restimulated with AdCAT (multiplicity of infection (MOI) = 1) for 5 days in complete medium supplemented with IL-2 (5 U/ml). Individual cultures were tested by the cytotoxicity assay described below. A well was considered negative if the specific lysis was <10%. The percentage of cultures that were negative was logarithmically plotted against the number of responder cells per well for each dilution. The frequency of precursor CTLs was determined by linear regression corresponding to 37% negative wells (20).

Cytotoxicity assays

Spleen cells were harvested from mice infected with Ad transgenes 10 days postinfection and cultured at 5 x 106 cells/ml in a 24-well plate in 1 ml RPMI 1640 supplemented with 10% FBS (HyClone Laboratories, Logan, UT), penicillin/streptomycin (5000 U/50 µg/ml), and 5 x 10-5 M 2-ME plus 1 ml viral supernatant at 37°C. The cells were restimulated with AdCAT or AdCD8 (MOI = 10). On day 4, the appropriate MHC-matched target cells (C57SV (H2b)for B6, L929 (H2k) for C3H, and SVBALB (H2d) for BALB/c (21, 22)) were infected with AdCAT (MOI = 100) and labeled with [3H]thymidine (5 µCi/ml) (23) because 51Cr labeling gave unacceptably high backgrounds in a 7-h cytotoxicity assay. The next day, the labeled infected cells were seeded at 2000 cell/well in a 96-well plate and used as targets for CTL at different E:T ratios. Lysis was allowed to proceed for 7 h, after which the plates were extensively washed and the remaining cells detached with trypsin/EDTA (Life Technologies). The cells were harvested and counted in a Microbeta Trilux Scintillation Counter (Wallac, Gaithersburg, MD). The percentage of specific lysis was calculated from the following formula: 100 x [(S - E)/S], where S and E are the spontaneous and experimental cpm, respectively.

CTL activity of OT-1 mice injected with OVAp was evaluated by a Cr release assay. Briefly, splenocytes were incubated with 51Cr-labeled EL4 cells at varying E:T ratios, and Cr release was quantified on a gamma counter. Percentage lysis was calculated according to the following formula: [(cpm sample - cpm spontaneous)/(cpm maximum - cpm spontaneous)] x 100. SDs were derived from triplicates within experiments.

T cell proliferation and 5-bromo-2'-deoxyuridine (BrdU) labeling

Spleen cells from mice infected with Ad transgenes were prepared as described above, except that 2.5 x 105 cells per well were cultured in 96-well plates. Splenocytes were restimulated with AdCAT or AdCD8 in the presence of 100 µl viral supernatant plus 100 µl fresh medium. At day 3, [3H]thymidine (1 uCi/well) was added, and the cells were harvested 16 h later for scintillation counting. BrdU labeling was performed as described (24). Briefly, splenocytes were incubated with 10 µM BrdU (Boehringer Mannheim, Indianapolis, IN) for 2 days and then stained for surface CD4 or CD8. The cells were then fixed with 1.2 ml ethanol on ice for 30 min and then with 1% paraformaldehyde at room temperature for 30 min. After DNA digestion, the cells were stained with FITC-conjugated anti-BrdU mAb (Boehringer Mannheim), and the percentage of stained cells was quantified by flow cytometry.

Statistical analysis

Test samples were analyzed for normal distribution and then compared by either Student’s t test (normal distribution) or the Mann-Whitney rank sum test (nonparametric data). Where appropriate, the paired t test was applied.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Expression of sCD8 protects against immune elimination in vivo

We previously reported the construction of soluble murine CD8 and sCD8 fusion proteins comprising the ectodomain of murine CD8-{alpha} alone (sCD8) or sCD8 fused to the ectodomain of either murine TNFR1 or Fas (APO-1/CD95) (14). To determine whether expression of sCD8 would influence the immune response to Ad, three different strains of mice (B6, C3H, and BALB/c) were injected i.v. with Ad expressing sCD8, and the level of serum expression was quantified weekly for up to 5 mo postinfection by ELISA. As shown in Fig. 1GoA, a striking persistence of CD8 expression was observed in B6 (up to 20 wk) and C3H mice (2–3 mo), but not in BALB/c mice. Prolonged expression could not be explained by differences in clearance of sCD8 from the circulation because the peak serum levels at 1 wk were similar in the three different strains, and continued mRNA expression in the liver was demonstrable in B6 and C3H but not in BALB/c mice at day 30 postinfection (Fig. 1GoB). These findings suggested that sCD8 interfered with the immune response to the Ad transgene.



View larger version (41K):
[in this window]
[in a new window]
  		FIGURE 1. A, Expression of sCD8 is prolonged in B6 and C3H mice. B6, C3H, or BALB/c mice were injected with 2.5 x 1010 particles of either AdsCD8 or AdFasCD8 by the i.v. route. Serum levels of sCD8 and sFasCD8 were quantified every 2 wk by ELISA as described in Materials and Methods. The level of expression of each protein in the serum was normalized to the peak level of the corresponding protein in SCID mouse serum as calculated by the formula relative expression = OD of test sample/OD in SCID serum, and it is expressed as the mean ± SD. In C3H mice, the relative level of sCD8 is significantly higher than FasCD8. *, p < 0.0001. B, Adenovirus transcripts were detected in the livers of B6 and C3H but not of BALB/c mice infected with AdsCD8. Mice were injected with 2.5 x 1010 particles AdsCD8 or AdFasCD8 (as in A) or AdCAT. Mice were sacrificed at the times indicated, and RNA was harvested from the liver. RT-PCR was performed as described in Materials and Methods using primers common to Ad vector sequences flanking the transgene (Ad5' and Ad3') and using ß-actin as a control. P, Plasmid containing sCD8 as a positive control; N, negative control from uninfected mice.

 
sCD8 provides transprotection for a second transgene

To determine the effect of co-injection of AdsCD8 with other transgenes, we first co-injected AdsCD8 with AdFasCD8. As shown in Fig. 2GoA, AdFasCD8 co-injected with AdCAT was rapidly cleared, as reported previously (14), whereas co-injection of AdFasCD8 with AdsCD8 resulted in sustained expression of FasCD8 until week 8–10 in B6 mice (Fig. 2GoA). We previously demonstrated that blockade of TNF-{alpha} by an Ad-encoded sTNFR fusion protein could attenuate the immune response to a second transgene (transprotection). AdTNFR prolonged the expression of the highly immunogenic bacterial protein CAT expressed by AdCAT (14). To determine whether sCD8 could also exert an inhibitory effect on immune clearance of a foreign protein, AdCD8 was co-injected with AdCAT and CAT levels quantified in the liver at d28. As shown in Fig. 2GoB, a substantially higher level of CAT expression was observed in B6 and C3H mice, but not in BALB/c mice, when the two transgenes were co-injected compared with mice injected with AdCAT alone. Because the serum level of sCD8 in the co-injection experiment was similar to AdCD8 alone (data not shown), these findings confirm that sCD8 has a potent effect on the immune clearance of Ad-encoded transgenes. Co-injection of an Ad null vector together with AdCAT had no effect on CAT expression at day 28. Although the sCD8 was derived from a mouse with the CD8 Ly2.1 allele, we found no evidence for Ab responses to the sCD8 in the Ly2.2 strains, BALB/c, or B6 (not shown).



View larger version (16K):
[in this window]
[in a new window]
  		FIGURE 2. sCD8 attenuates the immune response to a second transgene. A, B6 mice (n = 5) were injected with 2 x 1010 AdFasCD8 together with 2 x 1010 particles AdsCD8 or AdCAT as in Fig. 1Go. sFasCD8 in serum was captured using anti-Fas (Jo-2) and was quantified by ELISA as described in Materials and Methods. Error bar = SD. B, The three strains of mice indicated were injected i.v. with 2.5 x 1010 particles of AdCAT alone or a combination of AdCAT and AdsCD8 or AdCAT and AdNull (2.5 x 1010 particles of each virus). The livers were harvested at days 7 or 28 and were analyzed for CAT activity (mean ± SD) as indicated. Five mice of each strain were analyzed at day 7, and 10 mice from each strain were analyzed at day 28. As a positive control for CD8 inhibition, five AdCAT-infected C3H mice were injected weekly with 150 µg of anti-CD8 mAb (anti-CD8).

 
sCD8 inhibits IHL cytotoxicity

To directly assess the effect of sCD8 on anti-Ad CTL response in vivo, we isolated intrahepatic lymphocytes from day 10-infected B6 mice. As shown in Fig. 3Go, ex vivo cytotoxicity of mice infected with AdsCD8 + AdCAT was substantially lower than that of mice infected with Adnull + AdCAT. These findings indicate that sCD8 inhibited the anti-Ad CTL response in vivo. An inhibitory effect of sCD8 on CTL was confirmed by evaluation of the frequency of Ad-specific CTL precursors (CTLp). CTLp from AdsCD8 + AdCAT-infected mice was 2- to 3-fold lower than that from Adnull + AdCAT-infected mice (Expt. 1, 1/70,000 vs 1/190,000; Expt. 2, 1/30,000 vs 1/110,000).



View larger version (16K):
[in this window]
[in a new window]
  		FIGURE 3. sCD8 inhibits intrahepatic lymphocyte cytotoxicity. Intrahepatic lymphocytes were isolated from B6 mice injected with AdCAT + AdNull or with AdCAT + AdsCD8 (2 x 1010 particles of each virus) at day 10 postinfection. IHLs were pooled from three mice in each group and were analyzed for cytotoxicity against AdCAT-infected or uninfected 51Cr-labeled C57SV targets. The results are presented as the percentage of specific lysis (mean ± SD of triplicates). A similar result was obtained in a second experiment.

 
sCD8 attenuates the anti-Ad CTL response in vitro

Further in vitro studies were undertaken to define whether sCD8 anergized the CTL or interfered with lysis. Pilot studies to evaluate the CTL activity of B6 mice infected with AdsCD8 failed to reveal a significant defect in cytotoxicity after restimulation in vitro (not shown). Because very little sCD8 would be expected to be present in these cytotoxicity assays, we attempted to mimic the in vivo situation more accurately by including supernatants containing sCD8 in the restimulation phase, the cytotoxicity phase, or both phases of the experiment. As shown in Fig. 4Go, sCD8 significantly reduced cytotoxic effector function not only when present at the time of target cell lysis, but also when present during T cell restimulation. These findings suggest that sCD8 impairs cytotoxic T cell function both by reducing T cell activation as well as by directly interfering with lysis.



View larger version (11K):
[in this window]
[in a new window]
  		FIGURE 4. sCD8 attenuates in vitro cytotoxicity. B6 (n = 5), C3H (n = 4), and BALB/c (n = 5) mice were injected with 2.5 x 1010 particles of AdCAT i.v. Ten days later, the mice were sacrificed and spleen cells were restimulated with AdCAT (MOI = 1) in the presence of conditioned medium from AdCAT- or AdsCD8-infected HeLa cells. Five days later, cells were analyzed for their capacity to lyse the appropriate MHC-matched AdCAT-infected target C57SV (H2b) for B6, L929 (H2k) for C3H, and SVBALB (H2d) for BALB/c effectors. R, sCD8 present only in the restimulation step; C, sCD8 present only in the cytotoxicity step; C + R, sCD8 present at both steps of the assay; error bar = SD of the triplicate.

 
sCD8 inhibits T cell restimulation in vitro

Because sCD8 attenuated CTL responses in vitro when included in the restimulation step of the assay, we asked whether T cells obtained from B6 mice that had been infected with AdsCD8 were equally responsive to in vitro stimulation by Ad virus compared with T cells from mice infected with AdCAT. As shown in Fig. 5GoA, the T cell proliferative responses were similar in the two experimental infections, indicating that these T cells had not been rendered anergic and could be inhibited to the same extent by medium containing sCD8. To determine whether the sCD8 could inhibit T cell proliferation in all three strains of mice, mice were infected with AdCAT as before and then restimulated with AdCAT in the presence of medium from AdCAT- or AdsCD8-infected HeLa cells. sCD8 significantly reduced T cell proliferation in all three strains of mice (Fig. 5GoB).



View larger version (28K):
[in this window]
[in a new window]
  		FIGURE 5. sCD8 inhibits T cell activation. A and C, B6 mice were injected with 2.5 x 1010 particles of AdCAT or AdCD8 (n = 4 mice/group). Ten days later, the mice were sacrificed, and spleen cells were stimulated with AdCAT or AdsCD8 (MOI = 10) in fresh medium. T cell proliferation was quantified by [3H]thymidine incorporation (A) and IFN-{gamma} production by ELISA (C) 48 h after restimulation. The results between AdCAT- and AdCD8-infected mice were compared by the paired t test. W/O, Medium without virus; error bar = SD. B and D, Three strains of mice (n = 5 in each group) were infected with AdCAT and sacrificed at day 10 as in A. Spleen cells were restimulated with AdCAT (MOI = 10) in the presence of conditioned medium from AdCAT (CAT)- or AdsCD8 (sCD8)-infected HeLa cells. T cell proliferation (B) and IFN-{gamma} production (D) were quantified at day 2 after restimulation. The stimulation index was calculated from [3H]thymidine incorporation of virus-stimulated sample/[3H]thymidine incorporation samples without viral restimulation. Statistical analysis was performed by the paired t test. Error bar = SD.

 
One of the most important cytokines that promotes a cytotoxic T cell response is {gamma}-IFN (25, 26). Therefore, we assessed {gamma}-IFN production under the same conditions described for T cell proliferation in Fig. 5Go, A and B. As shown in Fig. 5Go, C and D, when compared with AdCAT-infected splenocytes, sCD8 inhibited {gamma}-IFN production after restimulation in all three strains of mice tested. Both proliferation and IFN-{gamma} production were lower in BALB/c mice compared with B6 mice (p < 0.001 and p < 0.002, respectively).

sCD8 selectively impairs sensitization of CD8 T cells

Virus-specific cytotoxic T cells are derived from the CD8 population but require help from CD4 T cells for optimal effector function (1, 2). To confirm that sCD8 selectively blocked CD8 T cell activation, we compared the effect of sCD8 on proliferation and IFN-{gamma} production by CD4 and CD8 subsets. As shown in Fig. 6Go, sCD8 blocked CD8 but not CD4 proliferation. Similarly, sCD8 blocked IFN-{gamma} production by CD8 but not by CD4 T cells (Fig. 7Go). To further test whether sCD8 has an impact on CD4 T cell function, we quantified anti-Ad Abs 28 days postinfection. No differences were observed in anti-Ad IgG between AdCAT or sCD8 in any of the three strains tested (not shown).



View larger version (53K):
[in this window]
[in a new window]
  		FIGURE 6. sCD8 blocks CD8 but not CD4 cell proliferation. B6 mice (n = 4) were injected with 2.5 x 1010 particles AdCAT, and spleen cells were collected at day 10 and restimulated in the presence of conditioned medium from either AdCAT- (CAT; A and D) or AdsCD8- (sCD8; B and E) infected HeLa cells for 2 days. BrdU (10 µM) was added at days 1 and 2 of restimulation. Cells were analyzed for BrdU incorporation. Cells were double stained with anti-CD8 (A, B, and C) or anti-CD4 Abs (C–F). Naive mice (n = 3) were used as controls. G, Summary of the results obtained from four individual mice. Error bar = SD.

 


View larger version (47K):
[in this window]
[in a new window]
  		FIGURE 7. sCD8 blocks {gamma}-IFN production by CD8 cells, but not by CD4 cells. A, Splenocytes were isolated from B6 mice (n = 5) previously injected with 2.5 x 1010 particles AdCAT as in Fig. 6Go. The cells were restimulated with AdCAT (MOI = 10) for 3 days in vitro in the presence of conditioned medium from AdCAT (CAT)- or AdsCD8 (sCD8)-infected HeLa cells. Brefeldin-A (10 µM) was added together with fresh AdCAT (MOI = 10) and incubated for an additional 5 h. Pooled cells from five mice were analyzed for intracellular {gamma}-IFN production by flow cytometry. The result is representative of two experiments.

 
sCD8 attenuates Ag-specific CD8 T cell activation in vivo

Injection of OVAp into OT-1 transgenic mice induces activation, proliferation, and subsequent apoptosis of the transgenic CD8+V{alpha}2+ T cells (27). To determine whether sCD8 could impede Ag-specific CD8 T cell activation in another in vivo experimental model, but one where Ag-specific T cells can be readily quantified, OT-1 transgenic mice were injected with AdCAT or AdsCD8 at day 1 as above. On day 7, the mice were injected i.v. with OVAp (4 nM/mouse), and the thymus, lymph nodes, and spleen were harvested 2 days later. As shown in Fig. 8GoA, three-color flow cytometry analysis revealed that mice that previously received AdCD8 expressed 2- to 3-fold less CD44 on lymph node-derived CD8+V{alpha}2+ T cells compared with mice that received AdCAT. When the total numbers of CD8+V{alpha}2+ T cells were compared between the two groups of mice, there was substantial elimination of thymocytes in both AdCAT- and AdsCD8-treated mice as expected (27), but significant decrease (p = 0.0046) in splenic CD8+V{alpha}2+ T cells in AdsCD8 compared with AdCAT-injected mice (Fig. 8GoB). The reduced number of CD8+V{alpha}2+ T cells in AdsCD8 mice was associated with reduced cytotoxicity (Fig. 8GoC). Cytotoxicity was Ag-specific and CD8-dependent because there was no specific lysis of EL4 cells without peptide (not shown), and cytotoxicity was blocked by anti-CD8 Ab. These results demonstrate that sCD8 impairs CTL activity in vivo, at least in part through attenuation of T cell activation.



View larger version (16K):
[in this window]
[in a new window]
  		FIGURE 8. SCD8 attenuates T cell activation in vivo. OT-1 T cell transgenic mice were injected i.v. with AdCAT (thin line in A) or AdsCD8 (thick line in A) at day 1 and challenged with 4 nM OVA peptide by the i.v. route at day 7. Two days later, the mice were sacrificed, and the degree of T cell activation (as determined by CD44 expression on CD8+V{alpha}2+ lymph node-derived T cells (A)), proliferation (as determined by the absolute number of CD8+V{alpha}2+ T cells in thymocytes and the spleen (B)), and cytotoxic effector function (evaluated by a Cr release assay with EL4 targets (C)) were determined. In A, one of three representative histograms is shown. In C, anti-CD8 mAb was used as a positive control. The results in B and C are the means ± SD for the number of mice indicated. Similar results were obtained in a second independent experiment.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Cytotoxic T cells engage their targets through specific recognition between the TCR and peptide Ag located in the groove of the ß-pleated sheet of the MHC-I molecule (28). In addition, the {alpha}-chain of the CD8 coreceptor binds to the {alpha}3 domain of MHC-I (28, 29, 30, 31, 32) with very low affinity (~200 µM in solution (33)). The precise way in which CD8 facilitates T cell activation is unclear, but it includes stabilization of the T cell-target contact and activation of lck (10, 12, 34, 35). One model suggests that the TCR first engages the peptide MHC complex; this leads to lck activation and subsequent recruitment of CD8 into the multimolecular TCR/CD3/MHC-I complex (33, 36). Only when all molecules are simultaneously engaged can full activation of the T cell occur (33).

Anti-CD8 Abs or soluble CD8{alpha}{alpha} homodimers have previously been shown to inhibit some, but not all, cytotoxic T cell functions in vitro (15). Because cytotoxic CD8+ T cells play a critical role in the clearance of virus from mice infected with replication-defective Ad vectors (4, 5), we asked whether sCD8 could attenuate the cytotoxic responses to the Ad vector that expressed murine sCD8. A soluble murine FasCD8 fusion protein (14) was used as a control for these experiments because this protein does not attenuate the immune response to Ad in normal animals (14) and, in contrast to sCD8, the N terminus of CD8 is not free to interact with MHC-I. We observed that not only did sCD8 persist for much longer periods of time compared with FasCD8, but it could also extend the expression of FasCD8 as well as a highly immunogenic bacterial protein, CAT, in B6 and C3H mice. Detection of AdsCD8 mRNA expression in hepatocytes at 28 days postinfection confirmed that persistence of sCD8 in the serum of infected mice was caused by reduced elimination of virus at the site of production.

The serum concentration of sCD8 used in these experiments was of the order of 5 µg/ml (14). Although this concentration is unlikely to saturate MHC-I in vivo, the requirement for simultaneous engagement of TCR/MHC-I and CD8 for full agonist T cell responses, as described above, may explain why low concentrations of sCD8 can attenuate the cytotoxic T cell response in vitro (15) or in vivo. Delon et al. (37) have shown that TCR/CD8 heterodimerization is required for full Ca2+ activation similar to that obtained by anti-CD3 Abs and that this is prevented if lck cannot be recruited to the intracellular domain of CD8 or if TCR/CD8 apposition is prevented. It has also been suggested that interference with a small number of the multimolecular complexes might alter the quality of the T cell response from an agonist to partial agonist (38, 39). Similar mechanisms may explain how a soluble CD8 peptide corresponding to amino acids 54–59 of CD8 extends the survival of allogeneic skin grafts in vivo (40).

Direct analysis of lymphocytes extracted from infected livers confirmed that the CTL activity from mice injected with AdsCD8 had impaired killing of virus-infected target cells. In vitro studies indicated that sCD8 impeded not only the effector phase of T cell cytotoxicity, but also reactivation of T cells as determined by lower proliferation and IFN-{gamma} production. These effects were specific for the CD8 T cell subpopulation in that CD4 T cells were unaffected and anti-virus Ab titers were equivalent in AdCAT- and AdsCD8-infected mice. The striking persistence of sCD8, particularly in the B6 strain of mice, is therefore likely to be explained by attenuation of T cell activation as well as cytotoxic effector function. Because the low frequency and polyclonal nature of the T cell response to viruses precluded analysis of the effect of sCD8 on virus-specific CD8 T cell activation in vivo, we studied the effect of sCD8 on T cell activation in OT-1 mice. OT-1 mice are transgenic for a TCR that is expressed on ~90% of CD8-positive T cells and can readily be quantified with an anti-V{alpha}2 mAb (16). Whereas immunization of OT-1 mice with the OVA peptide Ag induces activation and proliferation of the CD8 V{alpha}2+ T cells (27), mice previously injected with AdsCD8 demonstrated less activation and expansion of these cells compared with AdCAT-infected controls. These findings confirm that sCD8 can inhibit T cell activation and proliferation in vivo in a second experimental model.

Together, the observations in this report suggest that gene therapy utilizing Ad encoding sCD8 may be a useful strategy to impair the immune response to Ad vectors expressing a therapeutic gene product. Whether sCD8 can attenuate strong cytotoxic responses to replication-competent viruses, where reliance on coreceptor function may be less crucial, is uncertain because the sCD8 has not inhibited cytotoxicity of all T cell clones tested in vitro (our unpublished observations). However, this selectivity could be beneficial in gene therapy applications whereby sCD8 would substantially reduce CD8 T cell responses to relatively low levels of Ag produced by replication-defective adenovirus but would not lead to suppression of cytotoxic responses to infectious microorganisms.

It is well-established that immune responses to infectious agents in mice vary from strain to strain. Some (41, 42) but not all (21, 43) studies demonstrate that Ad-encoded transgenes persist longer in B6, compared with other strains of mice. The reasons for variation are not clear, but the precise Ad vector, nature of the transgene, and the host immune response genes need to be taken into account. Differences in expression of molecules directly responsible for cytotoxic effector function (such as perforin and Fas ligand) or innate immune response elements (such as TNF-{alpha}) can markedly influence the rate of elimination of the transgene (4, 14, 41, 44). Whereas immune responses to individual proteins may be correlated with MHC differences, responses to infectious agents are, in part, dictated by differences in cytokine production such as the Th1/Th2 paradigm (reviewed in Ref. 45). In view of the Th2 skew-ing of the immune response in BALB/c mice to Leishmania major, the rapid clearance of Ad vectors in this series of experiments and total failure of sCD8 to attenuate the cytotoxic response in vivo were surprising. However, in vitro studies revealed that BALB/c mice paradoxically demonstrated reduced T cell cytotoxicity, proliferative, and IFN-{gamma} responses compared with the other two strains, suggesting that CD8+ T cells may play a lesser role in clearance of Ad in this strain. Our previous demonstration that AdsCD8 is rapidly cleared from the lung even in B6 mice (14) may also indicate that either {gamma}{delta} T cells or NK cells (46, 47) have important effector functions toward Ad vectors.

We recently reported that blockade of the cytokine TNF-{alpha} by expression of a soluble TNFR-CD8 interfered with immune elimination of a foreign transgene encoded by Ad. The extended expression of Ad transgenes by soluble TNFR-CD8 was explained, in part, by a reduction in recruitment of mononuclear cells to the site of infection (14). The current study demonstrates that Ad persistence in the liver is also effected by sCD8, presumably by interference with CD8-MHC interactions, in several strains of mice. It will be useful to determine whether the combined use of strategies targeted toward the innate (TNF-{alpha}) and adaptive (sCD8) immune responses will prove to be additive and practical for use in humans. Manipulation of Ag-specific CD8 responses using sCD8 could also be envisaged as a new therapeutic approach in autoimmune diseases where CTLs play an important role in tissue injury, such as in insulin-dependent diabetes mellitus (48, 49).


   Acknowledgments
 
We thank Francis Carbone for providing OT-1 mice, H. Daniel Lacorazza for advice, and Janko Nikolic-Zugic for helpful discussions.


   Footnotes
 
1 This work was supported by Grant HL-9308-L (to K.B.E. and E.F.-P.) from the National Institutes of Health. Back

2 Address correspondence and reprint requests to Dr. Keith B. Elkon, Research Division, Hospital for Special Surgery, Cornell University Medical Center, 535 East 70th Street, New York, New York 10021. Back

3 Abbreviations used in this paper: Ad, adenovirus; s, soluble; MHC-I, MHC class I; OVAp, OVA peptide; CAT, chloramphenicol acetyl transferase; IHL, intrahepatic lymphocyte; MOI, multiplicity of infection; BrdU, 5-bromo-2'-deoxyuridine; B6, C57BL/6. Back

Received for publication February 1, 2000. Accepted for publication May 16, 2000.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Bennett, S. R., F. R. Carbone, F. Karamalis, R. A. Flavell, J. F. Miller, W. R. Heath. 1998. Help for cytotoxic-T-cell responses is mediated by CD40 signalling. Nature 4:478.
  2. Schoenberger, S. P., R. E. Toes, E. I. van der Voort, R. Offringa, C. J. Melief. 1998. T-cell help for cytotoxic T lymphocytes is mediated by CD40-CD40L interactions. Nature 393:480.[Medline]
  3. Ridge, J. P., F. Di Rosa, P. Matzinger. 1998. A conditioned dendritic cell can be a temporal bridge between a CD4+ T-helper and a T-killer cell. Nature 393:474.[Medline]
  4. Yang, Y., H. C. Ertl, J. M. Wilson. 1994. MHC class I-restricted cytotoxic T lymphocytes to viral antigens destroy hepatocytes in mice infected with E1-deleted recombinant adenoviruses. Immunity 1:433.[Medline]
  5. Yang, Y., Z. Xiang, H. C. Ertl, J. M. Wilson. 1995. Upregulation of class I major histocompatibility complex antigens by interferon {gamma} is necessary for T-cell-mediated elimination of recombinant adenovirus-infected hepatocytes in vivo. Proc. Natl. Acad. Sci. USA 92:7257.[Abstract/Free Full Text]
  6. Turner, J. M., M. H. Brodsky, B. A. Irving, S. D. Levin, R. M. Perlmutter, D. R. Littman. 1990. Interaction of the unique N-terminal region of tyrosine kinase p56lck with cytoplasmic domains of CD4 and CD8 is mediated by cysteine motifs. Cell 60:755.[Medline]
  7. Anel, A., A. M. O’Rourke, A. M. Kleinfeld, M. F. Mescher. 1996. T cell receptor and CD8-dependent tyrosine phosphorylation events in cytotoxic T lymphocytes: activation of p56lck by CD8 binding to class I protein. Eur. J. Immunol. 6:2310.
  8. Kersh, E. N., A. S. Shaw, P. M. Allen. 1998. Fidelity of T cell activation through multistep T cell receptor {zeta} phosphorylation. Science 281:572.[Abstract/Free Full Text]
  9. Veillette, A., M. A. Bookman, E. M. Horak, J. B. Bolen. 1988. The CD4 and CD8 T cell surface antigens are associated with the internal membrane tyrosine-protein kinase p56lck. Cell 55:301.[Medline]
  10. Zamoyska, R., P. Derham, S. D. Gorman, P. von Hoegen, J. B. Bolen, A. Veillette, J. R. Parnes. 1989. Inability of CD8 {alpha}' polypeptides to associate with p56lck correlates with impaired function in vitro and lack of expression in vivo. Nature 342:278.[Medline]
  11. Parnes, J. R.. 1989. Molecular biology and function of CD4 and CD8. Adv. Immunol. 44:265.[Medline]
  12. Fung-Leung, W. P., M. C. Louie, A. Limmer, P. S. Ohashi, K. Ngo, L. Chen, K. Kawai, E. Lacy, D. Y. Loh, T. W. Mak. 1993. The lack of CD8 {alpha} cytoplasmic domain resulted in a dramatic decrease in efficiency in thymic maturation but only a moderate reduction in cytotoxic function of CD8+ T lymphocytes. Eur. J. Immunol. 23:2834.[Medline]
  13. Araham, N., M. C. Miceli, J. R. Parnes, A. Veillette. 1991. Enhancement of T-cell responsiveness by the lymphocyte-specific tyrosine protein kinase p56lck. Nature 350:62.[Medline]
  14. Peng, Y., J. Trevejo, J. Zhou, M. W. Marino, R. G. Crystal, E. Falck-Pederson, K. B. Elkon. 1999. Inhibition of tumour necrosis factor {alpha} by an adenovirus-encoded soluble fusion protein extends transgene expression in the liver and lung. J. Virol. 73:5098.[Abstract/Free Full Text]
  15. Sewell, A. K., U. C. Gerth, D. A. Price, M. A. Purbhoo, J. M. Boulter, G. F. Gao, J. I. Bell, R. E. Phillips, B. K. Jakobsen. 1999. Antagonism of cytotoxic T-lymphocyte activation by soluble CD8. Nat. Med. 5:399.[Medline]
  16. Hogquist, K. A., S. C. Jameson, W. R. Heath, J. L. Howard, M. J. Bevan, F. R. Carbone. 1994. T cell receptor antagonist peptides induce positive selection. Cell 76:17.[Medline]
  17. Elkon, K. B., C. C. Liu, J. G. Gall, J. Trevejo, M. W. Marino, K. A. Abrahamsen, L. J. Old, R. G. Crystal, X. Song, J. L. Zhou, E. Falck-Pedersen. 1997. TNF-{alpha} plays a central role in immune mediated clearance of adenoviral vectors. Proc. Natl. Acad. Sci. USA 94:9814.[Abstract/Free Full Text]
  18. Drappa, J., N. Brot, K. B. Elkon. 1993. The Fas protein is expressed at high levels on double positive thymocytes and activated mature T cells in normal but not MRL/lpr mice. Proc. Natl. Acad. Sci. USA 90:10340.[Abstract/Free Full Text]
  19. Watanabe, H., K. Ohtsuka, M. Kimura, Y. Ikarashi, K. Ohmori, A. Kusumi, T. Ohteki, S. Seki, T. Abo. 1992. Details of an isolation method for hepatic lymphocytes in mice. J. Immunol. Methods 146:145.[Medline]
  20. Bot, A., S. Bot, C. A. Bona. 1998. Protective role of {gamma} interferon during the recall response to influenza virus. J. Virol. 72:6637.[Abstract/Free Full Text]
  21. Suzuki, M., R. Singh, M. A. Moore, W. R. Song, R. G. Crystal. 1998. Similarity of strain- and route-dependent murine responses to an adenovirus vector using the homologous thrombopoietin cDNA as the reporter genes. Hum. Gene Ther. 20:1223.
  22. Krajcsi, P., T. Dimitrov, T. W. Hermiston, A. E. Tollefson, T. S. Ranheim, S. B. Vande Pol, A. H. Stephenson, W. S. M. Wold. 1996. The adenovirus E3-14.7K protein and the E3-10.4K/14.5K complex of proteins, which independently inhibit tumor necrosis factor (TNF)-induced apoptosis, also independently inhibit TNF-{alpha} induced release of arachidonic acid. J. Virol. 70:4904.[Abstract/Free Full Text]
  23. Matzinger, P.. 1991. The JAM test: a simple assay for DNA fragmentation and cell death. J. Immunol. Methods 145:185.[Medline]
  24. Tough, D. F., J. Sprent. 1994. Turnover of naive- and memory-phenotype T cells. J. Exp. Med. 179:1127.[Abstract/Free Full Text]
  25. Butz, E. A., M. J. Bevan. 1998. Massive expansion of antigen-specific CD8+ T cells during an acute virus infection. Immunity 8:167.[Medline]
  26. Murali-Krishna, K., J. D. Altman, M. Suresh, D. J. Sourdive, A. J. Zajac, J. D. Miller, J. Slansky, R. Ahmed. 1998. Counting antigen-specific CD8 T cells: a reevaluation of bystander activation during viral infection. Immunity 8:177.[Medline]
  27. Koniaras, C., S. R. Bennett, F. R. Carbone, W. R. Heath, A. Lew. 1997. Peptide-induced deletion of CD8 T cells in vivo occurs via apoptosis in situ. Int. Immunol. 9:1601.[Abstract/Free Full Text]
  28. Kern, P. S., M. K. Teng, A. Smolyar, J. H. Liu, J. Liu, R. E. Hussey, R. Spoerl, H. C. Chang, E. L. Reinherz, J. H. Wang. 1998. Structural basis of CD8 coreceptor function revealed by crystallographic analysis of a murine CD8{alpha}{alpha} ectodomain fragment in complex with H-2Kb. Immunity 9:519.[Medline]
  29. Salter, R. D., R. J. Benjamin, P. K. Wesley, S. E. Buxton, T. P. Garrett, C. Clayberger, A. M. Krensky, A. M. Norment, D. R. Littman, P. Parham. 1990. A binding site for the T-cell co-receptor CD8 on the {alpha}3 domain of HLA-A2. Nature 345:41.[Medline]
  30. Potter, T. A., T. V. Rajan, Jr R. F. Dick, J. A. Bluestone. 1989. Substitution at residue 227 of H-2 class I molecules abrogates recognition by CD8-dependent, but not CD8-independent, cytotoxic T lymphocytes. Nature 337:73.[Medline]
  31. Connolly, J. M., T. H. Hansen, A. L. Ingold, T. A. Potter. 1990. Recognition by CD8 on cytotoxic T lymphocytes is ablated by several substitutions in the class I{alpha}3 domain: CD8 and the T-cell receptor recognize the same class I molecule. Proc. Natl. Acad. Sci. USA 87:2137.[Abstract/Free Full Text]
  32. Aldrich, C. J., R. E. Hammer, S. Jones-Youngblood, U. Koszinowski, L. Hood, I. Stroynowski, J. Forman. 1991. Negative and positive selection of antigen-specific cytotoxic T lymphocytes affected by the {alpha}3 domain of MHC I molecules. Nature 352:718.[Medline]
  33. Wyer, J. R., B. E. Willcox, G. F. Gao, U. C. Gerth, S. J. Davis, J. I. Bell, P. A. van der Merwe, B. K. Jakobsen. 1999. T cell receptor and coreceptor CD8{alpha}{alpha} bind peptide-MHC independently and with distinct kinetics. Immunity 10:219.[Medline]
  34. Norment, A. M., R. D. Salter, P. Parham, V. H. Engelhard, D. R. Littman. 1988. Cell-cell adhesion mediated by CD8 and MHC class I molecules. Nature 336:79.[Medline]
  35. Garcia, K. C., C. A. Scott, A. Brunmark, F. R. Carbone, P. A. Peterson, I. A. Wilson, L. Teyton. 1996. CD8 enhances formation of stable T-cell receptor/MHC class I molecule complexes. Nature 384:577.[Medline]
  36. Thome, M., V. Germain, J. P. DiSanto, O. Acuto. 1996. The p56lck SH2 domain mediates recruitment of CD8/p56lck to the activated T cell receptor/CD3/{zeta} complex. Eur. J. Immunol. 26:2093.[Medline]
  37. Delon, J., C. Gregoire, B. Malissen, S. Darche, F. Lemaitre, P. Kourilsky, J. P. Abastado, A. Trautmann. 1998. CD8 expression allows T cell signaling by monomeric peptide-MHC complexes. Immunity 9:467.[Medline]
  38. Purbhoo, M. A., A. K. Sewell, P. Klenerman, P. J. Goulder, K. L. Hilyard, J. I. Bell, B. K. Jakobsen, R. E. Phillips. 1998. Copresentation of natural HIV-1 agonist and antagonist ligands fails to induce the T cell receptor signaling cascade. Proc. Natl. Acad. Sci. USA 95:4527.[Abstract/Free Full Text]
  39. Bertoletti, A., A. Sette, F. V. Chisari, A. Penna, M. Levrero, M. De Carli, F. Fiaccadori, C. Ferrari. 1994. Natural variants of cytotoxic epitopes are T-cell receptor antagonists for antiviral cytotoxic T cells. Nature 369:407.[Medline]
  40. Choksi, S., B. A. Jameson, R. Korngold. 1998. A structure-based approach to designing synthetic CD8{alpha} peptides that can inhibit cytotoxic T-lymphocyte responses. Nat. Med. 4:309.[Medline]
  41. Barr, D., J. Tubb, D. Ferguson, A. Scaria, A. Lieber, C. Wilson, J. Perkins, M. A. Kay. 1995. Strain related variations in adenovirally mediated transgene expression from mouse hepatocytes in vivo: comparisons between immunocompetent and immunodeficient inbred strains. Gene Ther. 2:151.[Medline]
  42. Morral, N., W. O’Neal, H. Zhou, C. Langston, A. Beaudet. 1997. Immune responses to reporter proteins and high viral dose limit duration of expression with adenoviral vectors: comparison of E2a wild type and E2a deleted vectors. Hum. Gene Ther. 8:1275.[Medline]
  43. Muzzin, P., R. C. Eisensmith, K. C. Copeland, S. L. Woo. 1996. Correction of obesity and diabetes in genetically obese mice by leptin gene therapy. Proc. Natl. Acad. Sci. USA 93:14804.[Abstract/Free Full Text]
  44. Yang, Y., J. M. Wilson. 1995. Clearance of adenovirus-infected hepatocytes by MHC class I-restricted CD4+ CTLs in vivo. J. Immunol. 155:2564.[Abstract]
  45. Mosmann, T. R., R. L. Coffman. 1989. Th1 and Th2 cells: different patterns of lymphokine secretion lead to different functional properties. Annu. Rev. Immunol. 7:145.[Medline]
  46. Zuany-Amorim, C., C. Ruffie, S. Haile, B. B. Vargaftig, P. Pereira, M. Pretolani. 1998. Requirement for {gamma}{delta} T cells in allergic airway inflammation. Science 280:1265.[Abstract/Free Full Text]
  47. Korsgren, M., C. G. Persson, F. Sundler, T. Bjerke, T. Hansson, B. J. Chambers, S. Hong, L. Van Kaer, H. G. Ljunggren, O. Korsgren. 1999. Natural killer cells determine development of allergen-induced eosinophilic airway inflammation in mice. J. Exp. Med. 189:553.[Abstract/Free Full Text]
  48. Tisch, R., H. McDevitt. 1996. Insulin-dependent diabetes mellitus. Cell 85:291.[Medline]
  49. Delovitch, T. L., B. Singh. 1997. The non-obese diabetic mouse as a model of autoimmune diabetes: immune dysregulation gets the NOD. Immunity 7:727.[Medline]



This article has been cited by other articles:


Home page
 J. Virol.Home page
Y. Peng, E. Falck-Pedersen, and K. B. Elkon
Variation in Adenovirus Transgene Expression between BALB/c and C57BL/6 Mice Is Associated with Differences in Interleukin-12 and Gamma Interferon Production and NK Cell Activation
J. Virol., May 15, 2001; 75(10): 4540 - 4550.
[Abstract] [Full Text]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Peng, Y.
Right arrow Articles by Elkon, K. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Peng, Y.
Right arrow Articles by Elkon, K. B.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS