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 Kersh, G. J.
Right arrow Articles by Allen, P. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kersh, G. J.
Right arrow Articles by Allen, P. M.
The Journal of Immunology, 2000, 164: 5675-5682.
Copyright © 2000 by The American Association of Immunologists

Ligand-Specific Selection of MHC Class II-Restricted Thymocytes in Fetal Thymic Organ Culture1

Gilbert J. Kersh2, Deborah L. Engle, Calvin B. Williams and Paul M. Allen3

Department of Pathology and Center for Immunology, Washington University School of Medicine, St. Louis, MO 63110


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Positive and negative selection of thymocytes is determined by the specificity of the TCR and signaling through its associated molecules. We have studied selection of thymocytes bearing a MHC class II-restricted TCR using fetal thymic organ culture. This system allows the addition of peptides to the already diverse panoply of endogenous peptide ligands and is useful for analyzing ligand-specific negative selection of CD4 single positive (CD4SP) thymocytes. The data reveal that the ability of a given ligand to mediate negative selection is related to its dissociation rate from the TCR. We find that negative selection is very sensitive, and only the weakest ligand that we can identify fails to induce negative selection. None of the numerous peptides tested were able to induce an increase in CD4SP thymocytes. In addition, the ligands that induce negative selection of CD4SP thymocytes also cause an increase in numbers of CD8SP thymocytes bearing high levels of the class II-restricted TCR. Although these cells have a cell surface phenotype consistent with positive selection, they most likely represent cells in the process of negative selection. Further analysis reveals that these cells are not induced by these ligands in intact adult animals and that their induction is probably only revealed in the organ culture system.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Selection of T cells in the thymus produces a T cell repertoire that is tolerant of self-peptide/MHC and contains receptors competent for recognition of almost all foreign Ag challenges (1). Ultimately, this selection is determined by the interaction of the TCR with the self-peptide/MHC complexes in the thymus (2). There are essentially three possible selection outcomes for a given thymocyte as it is exposed to the many thymic self-peptide/MHC complexes: 1) the thymocyte may die by neglect because its TCR may never find an interaction that is strong enough to result in any signal transduction, 2) the thymocyte may undergo negative selection if its TCR reacts strongly with self-peptide/MHC and transduces a signal that results in apoptosis, and 3) the thymocyte may experience positive selection if its TCR has an interaction with self-peptide/MHC that is sufficient to transduce a signal but is not strong enough to lead to apoptosis. Positive selection allows the thymocyte to continue its developmental program and become a mature T cell.

Although the cellular outcomes of thymocyte selection are well defined, the molecular mechanisms responsible for these events are much less well understood. To begin to understand these mechanisms, we would like to know the type of TCR interaction with peptide/MHC complexes that leads to either positive or negative selection and then compare these TCR-ligand interactions with recognition of the eventual foreign Ag. In addition, we would like to define the signals that lead to either positive or negative selection and then compare these signals with those induced by eventual foreign Ags.

To date, some progress has been made in answering the questions regarding the ligands for positive and negative selection. For the most part, these experiments have relied on murine TCR transgenic systems using a TCR specific for a peptide presented by MHC class I molecules. In these cases, the foreign peptide and MHC restriction are known, the TCR is fixed, and the variable component for selection in the thymus is the peptide. More progress has been made in the study of MHC class I-restricted TCRs because of methods available to limit the presentation of endogenous peptides. These include usage of the TAP and ß2-microglobulin knockout mice that have poor presentation of endogenous peptides on MHC class I but can be induced in fetal thymic organ culture (FTOC)4 systems to present exogenously added peptides (3, 4, 5, 6). These experiments have consistently shown that high concentrations of strong agonist ligands will induce negative selection. Another consistent finding of these studies is that peptide/MHC complexes that have a very weak interaction with the TCR can lead to positive selection. These complexes are likely to be antagonists for the mature T cell and have a relatively low affinity for the TCR (7). In addition, some studies have found that the agonist ligands that induce negative selection at high concentrations may be able to induce some positive selection at low concentrations (3, 4).

Studies of the role of peptides in thymic selection of MHC class II-restricted TCRs have been less decisive. We and others have expressed peptide ligands for a MHC class II-restricted TCR as transgenes (8, 9). These studies have either added an additional epitope to the endogenous peptides or have expressed a single peptide/MHC class II complex (10, 11, 12, 13, 14, 15, 16). The results are that negative selection is very sensitive and both agonist and antagonist ligands can induce negative selection. Drawbacks of these studies are the inability to titrate the added ligand and the limitations on numbers of peptides that can be analyzed. For MHC class II-restricted TCRs, studies using organ culture techniques have been reported, and these have added a small number of peptides to the endogenous peptide repertoire. These experiments have demonstrated peptide-specific negative selection (17), antagonism of positive selection (18), and selection of MHC class II-restricted thymocytes into the CD8 single positive (CD8SP) lineage (19). However, the nature of ligands that mediate positive selection of class II-restricted, CD4SP thymocytes is still unknown. In addition, a lower boundary for the level of stimulation needed to induce negative selection of class II-restricted TCRs has not been established. A limitation in most of these studies is the inability to eliminate the presentation of endogenous peptides on MHC class II and then add back a peptide of interest. It has been reported that in some cases the agonist ligand can induce positive selection (20, 21), but it is not clear whether negative selection is operational in these studies.

In the present study, we have employed FTOC to examine the selection of a particular MHC class II-restricted TCR. We have not taken any steps to limit the presentation of the endogenous peptides, although the selection of the particular TCR studied on the endogenous complexes is fairly inefficient. The advantages of this system over the previously employed transgenic system are that a greater number of peptides can be analyzed and a wide range of concentrations can be used. We find that negative selection is very efficient in this system, and a relative potency for negative selection can be assigned to the various ligands. Peptides that are very weak antagonists can induce negative selection, although some antagonists are below the lower boundary for negative selection. In addition, the potency of the different peptides in the negative selection assay is related to the off-rate of the TCR-ligand interaction. Furthermore, in some cases a population of CD8SP cells is induced by ligands that cause negative selection, but we provide evidence that the peptide-dependent induction of these cells only occurs in FTOC.


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

The 3.L2tg mice have been described previously (22). These mice were bred and maintained in a specific pathogen free mouse facility at Washington University. B6.AKR mice (H-2k) were originally obtained from The Jackson Laboratory (Bar Harbor, ME). The transgenic mice expressing the Hb(64–76) epitope or the altered peptide ligands derived from it have been described previously (9), with the exception of the mouse expressing the D73 epitope. This mouse was made in a manner similar to the other altered peptide ligand (APL)-expressing mice, with the D73 epitope engineered into the sequence of a membrane form of hen egg lysozyme. Expression of this molecule on all MHC class II-positive cells was driven by the I-E{alpha} promoter derived from the plasmid pDOI-5 (23).

Peptides

Peptides were synthesized using standard 9-fluorenylmethyloxycarbonyl chemistry on a Symphony/Multiplex multiple peptide synthesizer (Protein Technologies, Tuscon, AZ). The 9-fluorenylmethyloxycarbonyl-protected amino acids as well as Lys-coupled resin were purchased from Advanced Chemtech (Louisville, KY). Peptides were purified on a C18 reverse phase HPLC column. Amino acid content and accurate concentrations of all peptides were determined by analysis on the Beckman (Fullerton, CA) model 6300 amino acid analyzer and by comparison with a standard of known concentration. The purity and composition of each of the peptides was confirmed by mass spectrometry at the Washington University Mass Spectrometry Resource. Altered peptides of Hb(64–76) are referred to using the one-letter amino acid code for the substituted amino acid followed by its position. For example, A72 refers to an Hb(64–76) peptide that has Ala substituted for Asn at position 72. The sequences of the peptides used were as follows: Hb(64–76), GKKVITAFNEGLK; T72, GKKVITAFTEGLK; D73, GKKVITAFNDGLK; I72, GKKVITAFIEGLK; A72, GKKVITAFAEGLK; S72, GKKVITAFSEGLK; L72, GKKVITAFLEGLK; G72, GKKVITAFGEGLK; Q72, GKKVITAFQEGLK; and E72, GKKVITAFEEGLK.

FTOC

To obtain fetal thymic lobes, homozygous 3.L2tg male mice were mated with B6.AKR female mice. The morning that mating plugs were found was considered day 1 of gestation, and the fetal lobes were harvested at day 16. Lobes were placed on a 0.45-µm filter (HAWG01300; Millipore, Bedford, MA) that was resting on a gelfoam sponge (NDC 0009-0315-03; Pharmacia and Upjohn, Peapack, NJ). These items were placed in a well of a six-well culture dish containing 2 ml RPMI 1640 supplemented with 10% FCS (HyClone, Logan UT), 2 mM Glutamax (Life Technologies, Gaithersburg, MD), 50 µg/ml gentamicin (Life Technologies), 100 U/ml penicillin (Life Technologies), and 100 µg/ml streptomycin (Life Technologies). Media was exchanged daily, and in cases where peptide was present it was included throughout the culture. Lobes were incubated at 37°C, 5% CO2 and were harvested after 7 days of culture. Single-cell suspensions were then prepared for analysis.

Flow cytometry

Samples were stained at 4°C for 30 min with the Abs diluted in PBS with 0.5% BSA and 0.02% NaN3, and then they were washed. When necessary, cells were also stained in an identical manner with a second step reagent. The Abs used in this study were 53.6.7-FITC (rat anti-mouse CD8{alpha}; PharMingen, San Diego, CA), CT-CD8a-tricolor (rat anti-mouse CD8{alpha}; Caltag, South San Francisco, CA), H129.19-PE (rat anti-mouse CD4; PharMingen), streptavidin-FITC (Caltag), M1/69-biotin (rat anti-mouse CD24; PharMingen), H1.2F3-biotin (hamster anti-mouse CD69; PharMingen), 53-7.3-biotin (rat anti-mouse CD5; PharMingen), and clonotype Ab (CAb)-biotin (mouse anti-3.L2 clonotype) (22). Cells were analyzed on a FACSCalibur flow cytometer (Becton Dickinson, Mountain View, CA) using the CellQuest analysis software.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
The FTOC system

The Th1 clone 3.L2 is specific for a peptide derived from residues 64–76 of the d allele of murine Hb (Hb(64–76)) presented by I-Ek (24). This clone not only recognizes the Hb(64–76) peptide but also many APLs derived from Hb(64–76) (25). These APLs were made by single amino acid substitutions in the Hb(64–76) peptide. They are referred to by the single-letter code of the new amino acid followed by the position of the substitution. For example, S72 is a peptide that has a serine substitution at position 72 of Hb(64–76). The peptides used can be classified based on their effect on IL-2 production and include weak agonists (T72 and D73), antagonists (I72, A72, S72, L72, and G72), and null peptides (Q72 and E72). TCR transgenic mice have been generated that utilize the 3.L2 TCR (22), and this transgenic system is therefore well suited for the study of thymic selection using FTOC.

To obtain 3.L2 transgenic fetuses, homozygous 3.L2tg males were bred to B6.AKR females. Therefore, the resulting offspring were all H-2k and heterozygous for the 3.L2 transgene. A culture of embryonic day 16 (E16) lobes for 7 days in normal media produced thymocytes that were analyzed by staining for CD4, CD8, and the anti-3.L2 CAb. This analysis revealed a distribution of cells much like those found in adult 3.L2tg thymocytes (Fig. 1Go). We followed the selection of the 3.L2 TCR-bearing cells specifically by gating on only those cells with high levels of the 3.L2 TCR (CAbhigh). Therefore, the endogenous peptides resulted in the selection of CAbhigh CD4SP cells that made up about 6–7% of total thymocytes.



View larger version (64K):
[in this window]
[in a new window]
  		FIGURE 1. Negative selection of 3.L2 thymocytes in FTOC is induced by Hb(64–76). H-2k/k thymic lobes heterozygous for the 3.L2 TCR transgene were taken from an embryonic day 16 fetus. After culture for 7 days in the presence of the Hb(64–76) peptide, the lobes were harvested and stained for CD4, CD8, and CAb (anti-3.L2 clonotype). CD4 vs CD8 dot plots are displayed for those cells that express high levels of CAb. The concentration of peptide included throughout the culture is indicated above each dot plot, and for each concentration two lobes taken from different fetuses were pooled. The numbers in the upper left quadrants indicate the percentage of total thymocytes that are CD4+ CD8- CAbhigh. The number of recovered cells varied between 150,000 and 250,000 cells per lobe, with no significant decrease in cases of negative selection.

 
Ligand-specific negative selection

The addition of the strong agonist peptide Hb(64–76) to the organ culture resulted in the efficient deletion of nearly all CAbhigh CD4SP thymocytes and most CAbhigh double-positive thymocytes (Fig. 1Go). The percentage of CAbhigh CD4SP cells went from 6.8% in the media control to 0.08% with 1 µM Hb(64–76), representing greater than 98% deletion. There was considerable variability in the sizes of both the media-treated and peptide-treated lobes, but overall there was not a large decrease in thymic cellularity when negative selection was ongoing. Therefore, the decreases observed in percentage of Cabhigh CD4SP cells also represented a decrease in absolute number. The negative selection in FTOC (4) was concentration-dependent, and some negative selection was seen as low as 0.001 µM Hb(64–76), but at 0.0001 µM Hb(64–76), the selection looked very similar to selection without added peptide (Fig. 1Go). In fact, peptide concentrations as low as 1 pM were tested and did not result in selection that was significantly different from the media control (data not shown). Interestingly, at 0.1 µM Hb(64–76) there was an induction of cells that had high levels of the 3.L2 TCR but were CD8SP or CD4, CD8 double negative. These cells will be discussed further below. Therefore, the sensitivity of this FTOC system to variations in concentration allowed us to use this system to address the potency of various APLs in negative selection.

APLs were first tested by culturing pairs of lobes derived from two different fetuses in the presence of various concentrations of peptides. The two lobes were pooled and analyzed for CD4, CD8, and CAb expression. The percentage of Cabhigh CD4SP cells remaining after culture with a titration of Hb(64–76) or four APLs is depicted in Fig. 2GoA. For comparison, stimulation of mature 3.L2 T cells by the peptides shown causing negative selection in Fig. 2GoA is displayed in Fig. 2GoB. This stimulation was measured by assaying for the apoptosis of the APC used in the assay (the B cell lymphoma CH27). This apoptosis was TCR ligand-dependent and mediated by interaction with stimulated T cell clones. The concentration of peptide at which apoptosis was observed can be used as a measure of the potency of the ligand (25).



View larger version (29K):
[in this window]
[in a new window]
  		FIGURE 2. The ability of altered peptides to induce negative selection in FTOC. A, Embryonic day 16 thymic lobes were cultured as in Fig. 1Go. The percentage of cells that were CD4+ CD8- CAbhigh after 7 days in culture is plotted vs the concentrations of the various peptides. Controls that used only media and no peptide usually had between 6.5 and 7.5% CD4+ CD8- CAbhigh cells. B, For comparison, the ability of these same peptides to induce killing of the CH27 B cell lymphoma by the mature 3.L2 Th1 clone is displayed. CH27 cells were loaded with [3H]thymidine and were added to the 3.L2 Th1 clone in the presence of the indicated peptides. A reduction in counts is due to the apoptosis of the CH27 cells. These data are the same as those reported in Ref. 25 .

 
Three main points can be made from the data in Fig. 2Go. First, there is a hierarchy of peptides, in which some are more potent at inducing negative selection than others. The hierarchy of the peptides acting on thymocytes as measured by negative selection is the same as that observed for the peptides acting on mature T cells as measured by the induction of apoptosis of the APC (25). Other assays using mature 3.L2 T cells also have the same hierarchy, but the apoptosis assay is shown because it is the only assay in which all of these peptides can be compared directly. Second, the mature T cells are more responsive to low doses of the highly potent peptide Hb(64–76), whereas the thymocytes are more responsive to low doses of the weaker peptides (I72 and A72) (Table IGo). This results in a compression of the range of peptide doses that are effective in the negative selection assay compared with the mature T cell assay. This is very likely the result of differences in the biochemical pathways induced by these peptides in thymocytes vs mature T cells. Third, the potency of the peptides at inducing negative selection is related to the off-rate of the interaction of the TCR with peptide/MHC ligand. We have previously measured the off-rates of the 3.L2 TCR interacting with many of these peptide/MHC ligands and found that Hb(64–76)/I-Ek has the slowest off-rate and that the off-rate increases as the potency of these ligands in T cell activation decreases (26). The results presented here demonstrate that this correlation also extends to negative selection of thymocytes. The relationship between thymocyte deletion, induction of apoptosis by the 3.L2 clone, and TCR-ligand half-life is summarized in Table IGo.


View this table:
[in this window]
[in a new window]
  		Table I. Comparison of thymocyte deletion, induction of apoptosis by a T cell clone, and TCR-ligand half-life1

 
Other ligands exist for the 3.L2 TCR that are unable to cause apoptosis of an APC but can be identified as ligands for this TCR by their ability to act as TCR antagonists for IL-2 production and by their ability to induce phosphorylation of the TCR-associated {zeta}-chain (26, 27). To study the effect of these weaker peptides on selection in FTOC, the experiment was set up in a slightly different manner. For these peptides, pairs of lobes were removed from an individual E16 fetus and were separated into a peptide-treated culture for one lobe and a media control culture for the other lobe. This technique eliminated potential mouse-to-mouse variability and allowed examination of more subtle effects of these peptides on selection. In this case, selection was measured by dividing the percentage of cells that were CAbhigh CD4SP in the peptide-treated lobe by the percentage in the media control lobe. By this technique, the A72 peptide also demonstrated negative selection at high concentrations (Fig. 3Go).



View larger version (24K):
[in this window]
[in a new window]
  		FIGURE 3. Weak antagonist peptides can either induce weak negative selection or have no apparent effect on the 3.L2-positive cells in FTOC. For each data point, the two thymic lobes of an embryonic day 16 fetus were removed. One lobe was put into culture in the presence of peptide, and the other lobe was cultured with media only. The percentage of CD4+ CD8- CAbhigh cells was determined for peptide-treated and control lobes, and the ratio of peptide-treated to control is plotted vs the concentrations of the peptides used.

 
Interestingly, the S72 and L72 peptides also induced a small degree of negative selection (Fig. 3Go). These ligands are very weak antagonists and score negative in the apoptosis assay (25). This result demonstrates the extreme sensitivity of negative selection. The G72 peptide is the weakest ligand that we have identified in assays on mature 3.L2 T cells, and it is able to antagonize IL-2 production only at very high concentrations. In addition, the G72 peptide will induce a small amount of TCR-associated {zeta} phosphorylation when used at high concentration (27). As shown in Fig. 3Go, this is the only ligand that clearly gives a signal to mature T cells but is not able to induce any negative selection. The Q72 and E72 peptides have no apparent ability to stimulate mature T cells, and in this assay, they also seem to have no effect on thymocytes. The E72 peptide was compared numerous times to lobes that had no peptide added and was not found to alter development in a significant manner. Therefore, ratios of 0.8–1.25 shown in Fig. 3Go fall within the range of variability that we have observed using media alone.

No ligand-specific positive selection of Cabhigh CD4SP cells was observed

In none of these assays did we see a peptide-dependent increase in the percentage of CAbhigh CD4SP cells. In addition to the data presented ( Figs. 1–3GoGoGo), much lower concentrations of the high-potency peptides were used (Hb(64–76) and T72) as well as much lower concentrations of the low-potency peptides (S72 and G72) (data not shown). In none of these cases was an increase in positive selection observed that was above the variation seen with the E72 control peptide or with media alone. This could be due to the fact that positive selection in this system is already at a maximum and increasing the number of available ligands cannot lead to an increase over the selection observed on endogenous ligands. An alternative possibility is that the added peptides are not loading the MHC class II on the proper APCs required for positive selection into the CD4 lineage.

Induction of CD8SP cells with high levels of the 3.L2 TCR

The result of T cell development is that almost all cells that have a MHC class II-restricted TCR coexpress the CD4 molecule, and those that are MHC class I-restricted coexpress the CD8 molecule. However, exceptions have been found, and these have been most notably observed as MHC class II-restricted TCRs expressed on CD8SP cells. In some MHC class II-restricted TCR transgenic mice, a small percentage of CD8SP cells have been observed, even when endogenous receptors were eliminated by recombinase-activating gene-2 deficiency (28). This small percentage has been found to increase dramatically under special conditions: CD4 deficiency (29), Bcl-2 overexpression (30), expression of an activated form of Notch-1 (31), and administration of a TCR antagonist in newborn thymic organ culture (19). These conditions have been hypothesized to increase the numbers of CD8SP cells with a class II-restricted TCR by influencing the mechanisms of lineage commitment; however, it has also been proposed that these cells may appear as a result of interference with the mechanisms of apoptosis (32).

In this study, we have also observed a peptide-dependent increase in the percentage of CAbhigh CD8SP thymocytes. As noted above, a dose of 0.1 µM Hb(64–76) resulted in increased numbers of CAbhigh CD8SP cells. Surprisingly, all of the ligands that were able to induce negative selection were also able to induce an increase in these cells (Fig. 4Go). This was an increase not only in percentage but also in absolute number and could be as much as 6-fold more than the CAbhigh CD8SP cells obtained with media alone. For all of these ligands, the induction was concentration-dependent and occurred at a concentration that also resulted in a significant deletion of CAbhigh CD4SP cells. The actual concentration of a particular peptide that resulted in good induction of the CAbhigh CD8SP cells was related to its potency in the negative selection assay. Lower potency ligands induced the CAbhigh CD8SP cells only at higher concentrations.



View larger version (69K):
[in this window]
[in a new window]
  		FIGURE 4. All of the peptides that induce negative selection can also induce a population of CD4- CD8+ CAbhigh cells. Embryonic day 16 thymic lobes were cultured as in Fig. 1Go. CD4 vs CD8 dot plots are displayed for those cells that express high levels of CAb. The numbers in the lower right quadrants indicate the percentage of total thymocytes that are CD4- CD8+ CAbhigh. The appearance of the CD4- CD8+ CAbhigh cells was concentration dependent, and for Hb(64–76) and T72, these cells were largely absent at higher concentrations of peptide. The increase in percentage also represents an increase in absolute number of cells because the sizes of the peptide-treated lobes were not significantly decreased compared with those treated with media alone.

 
Because it was unusual to observe the induction of CD8SP cells by a peptide/MHC class II interaction with the TCR, we examined these cells further for phenotypic markers of mature cells (Fig. 5Go). For comparison, CD8SP and CD4SP cells induced on the endogenous peptide repertoire were also examined. CAbhigh CD8SP thymocytes were also present in the absence of added peptide, but a clear increase was observed with added 5 µM T72 peptide (Fig. 5Go). In both cases, these CD8SP cells mostly expressed low levels of CD24 (heat-stable Ag), an indicator that these are mature cells and have probably been through the process of positive selection (33). The markers CD69 and CD5 are also up-regulated during both positive and negative selection (34, 35). High levels of these markers were also found on some of the CD8SP cells, and in the case of the CD8SP cells induced by T72 peptide, the percentage of positive cells was similar to the CD4SP cells selected on endogenous peptides. Overall, these results show that the CAbhigh CD8SP cells induced by the negatively selecting peptides have the cell surface phenotype of mature, positively selected cells. Similar cells were also present when selection occurred on the endogenous peptides, but a large increase was observed when peptide was added.



View larger version (23K):
[in this window]
[in a new window]
  		FIGURE 5. The CD4- CD8+ cells induced by peptides have the cell surface phenotype of positively selected cells. Ten embryonic day 16 thymic lobes were cultured in the presence of 5 µM T72 peptide, and ten lobes were cultured in media only. After 7 days in culture, the lobes from each condition were pooled and analyzed by flow cytometry. Because the lobes were pooled, 600,000 cells were recovered from each condition. The panels show staining for CAb, CD24 (heat-stable Ag), CD69, and CD5. The panels display expression of these different cell surface markers on either CD4+ CD8- or CD4- CD8+ cells generated by the endogenous peptide repertoire or on CD4- CD8+ cells generated when the T72 peptide was present. The numbers in the upper right corner of each histogram represent the percentage of total thymocytes that are coreceptor single positive and fall within the indicated marker.

 
To test whether these CAbhigh CD8SP cells were only present in FTOC, thymocytes from adult animals were also examined. In Fig. 6Go, CAb staining on CD8SP thymocytes from adult 3.L2tg mice is shown. Although perhaps fewer CAbhigh CD8SP cells were present in the adult thymus, they were still present in significant numbers. Furthermore, we have shown previously that some CAbhigh CD8SP thymocytes were present in 3.L2tg mice on the recombinase-activating gene-1-/- background (22). These data show that the existence of CAbhigh CD8SP cells is not restricted to FTOC. To put these results in the context of nontransgenic T cell development, Fig. 6Go also displays a comparison of CD24 staining on CD8SP thymocytes between a 3.L2tg mouse and a normal B6.AKR mouse. The result shows that the 3.L2tg mouse does have significantly more CD24low CD8SP cells, suggesting that selection into the CD8 lineage is somehow partly increased by the presence of the 3.L2 transgene. The conclusion is that small numbers of Cabhigh CD8SP cells with a mature phenotype are selected on the endogenous peptides and that their numbers are greatly increased in FTOC by the presence of specific peptides that can mediate negative selection.



View larger version (25K):
[in this window]
[in a new window]
  		FIGURE 6. CAbhigh CD8SP thymocytes are present in the adult thymus of 3.L2tg mice. Thymocytes from an adult 3.L2tg mouse on the B6.AKR background and from a nontransgenic B6.AKR mouse were analyzed by flow cytometry. Histograms for CD24 (3.L2tg and B6.AKR) and CAb (3.L2tg) staining on CD8SP thymocytes are shown. The numbers represent the percentage of total thymocytes that are CD8SP and fall within the indicated marker.

 
The increase in the number of CAbhigh CD8SP cells by the addition of a concentration of peptide that induces negative selection of CAbhigh CD4SP cells was difficult to understand. It is entirely possible that this induction is peculiar to FTOC and is not part of the normal T cell biology. To test this possibility, we examined adult 3.L2tg thymocytes from mice that also express different Hb(64–76) epitopes in vivo. These mice have been described previously and express on all MHC class II-positive cells a membrane form of hen egg lysozyme into which the Hb(64–76) peptide or its derivatives have been inserted (9). As we reported previously, the expression of Hb(64–76), T72, or I72 in vivo resulted in almost complete deletion of CAbhigh CD4SP thymocytes. The E72 peptide had no effect, and the A72 peptide resulted in a slight decrease in CD4SP thymocytes with the highest levels of CAb staining. A mouse expressing the D73 epitope was also constructed and also had efficient deletion of CAbhigh CD4SP thymocytes (data not shown). When these mice were examined for CAb expression on CD8SP thymocytes, none of these ligands led to an increase in CAbhigh CD8SP cells (Fig. 7Go). In fact, these cells were efficiently deleted, much like the CD4SP thymocytes. Because all of these ligands were expressed at a similar level in vivo, there is an expectation that if any of these ligands could induce CAbhigh CD8SP cells, one of them would have been at a proper concentration to do so. The conclusion is that although CD8SP thymocytes with a MHC class II-restricted TCR exist in vivo, their peptide-dependent induction is peculiar to the FTOC system.



View larger version (18K):
[in this window]
[in a new window]
  		FIGURE 7. In vivo peptide/MHC ligands that cause negative selection of 3.L2tg thymocytes do not induce CAbhigh CD8SP cells. Mice expressing the indicated Hb peptide epitopes on the cell surface of all MHC class II-positive cells were bred to the 3.L2tg mice. The expression of CAb on the CD8SP cells from adult thymocytes is displayed on the histograms. The numbers represent the percentage of total thymocytes that are CAbhigh and CD8SP. The indicated markers delineate the CAbhigh cells and were placed based on a 3.L2tg control mouse done on the same day. The markers are not all in the same position because not all of these analyses were done at the same time. As shown previously for all of these ligands except D73, these epitopes result in the negative selection of almost all CAbhigh CD4SP cells. The expression of D73 similarly eliminates all of the CAbhigh CD4SP cells (data not shown).

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
In this study, we have examined selection of a MHC class II-restricted TCR in FTOC. We have done this not only using the endogenous peptide repertoire but also with a wide potency range of added peptides. The results are that almost all of the peptides that have any effect on mature cells will induce negative selection of the CAbhigh CD4SP cells. The hierarchy of these ligands for negative selection is the same as the hierarchy for mature T cell activation and correlates with the off-rate of the TCR-ligand interaction. However, one peptide was found that is an antagonist for the mature 3.L2 T cell and induces phosphorylation of the TCR-associated {zeta} chain, but induces no negative selection in FTOC. No evidence of positive selection was observed. Suprisingly, the negatively selecting peptides induced CAbhigh CD8SP cells in FTOC, but this effect was not observed in vivo.

Kinetic proofreading in negative selection

Models for T cell activation have been proposed that incorporate the idea of kinetic proofreading to explain the fidelity of T cell activation and the ability of the TCR to distinguish between subtly different ligands (36, 37). In these models, the response of the T cell depends on the lifetime of the interaction between the TCR and ligand, with longer interaction times leading to greater T cell activation. The postulate is that signaling events, such as the phosphorylation of TCR-associated proteins and binding of Src homology-2 domain-containing proteins require bound TCR and time to occur. We and others have provided evidence for such a model by showing that low-potency ligands have faster off-rates from the TCR compared with those of strong agonist ligands (26, 38, 39).

It has previously been unclear whether this model should apply to negative selection of thymocytes, particularly those bearing class II-restricted TCRs. The TCR on thymocytes is associated with the same CD3 and {zeta} signaling components as the TCR on mature T cells, but there is evidence that the signal transduction is different in thymocytes. For example, TCR-associated {zeta} molecules have much more constitutive {zeta} phosphorylation in the thymus than in mature T cells (40, 41). The data provided in this study show that negative selection can also fit into the kinetic proofreading model. The potency for negative selection of the various ligands studied correlates with their off-rates for interaction with the 3.L2 TCR. However, the data also show that the process of negative selection has some important biochemical differences from activation of mature T cells.

These differences are revealed by comparing multiple ligands at various concentrations. Based on theoretical considerations, thymocytes are expected to be more sensitive to Ag stimulation than mature T cells are. This is because it would make sense to have a lower threshold for deletion of autoreactive clones, thereby providing a safety margin against autostimulation of mature T cells (42). Several studies have demonstrated that thymocytes are indeed more sensitive to stimulation than mature T cells are (42, 43, 44), but in some assays thymocytes are found to be less responsive (45, 46). What we have observed is that the relative sensitivity of thymocytes to stimulation depends on the strength of the ligand considered. For negative selection, there is only a 3000-fold difference between concentrations of Hb(64–76) and A72 peptides required for 50% negative selection, whereas for induction of B cell apoptosis by the 3.L2 T cell clone this difference is greater than 100,000-fold. This results from a similar sensitivity to the Hb(64–76) ligand but a much greater sensitivity of thymocytes to the A72 ligand. Thus, we can conclude that intracellular events induced by ligands in the thymus are likely to be different from events induced by ligands in the 3.L2 clone. The result is a greater sensitivity of thymocytes to very weak ligands presented on thymic APCs. This enables thymocytes to be deleted not only by agonist ligands but also by ligands that are far too weak to cause any subsequent activation in peripheral T cells.

Positive selection of class II-restricted thymocytes

Not all of the peptides that act as ligands for the 3.L2 TCR were able to induce negative selection in this system. Numerous attempts using the G72 peptide revealed that this ligand does not induce any negative selection. Because in assays of TCR antagonism and {zeta} phosphorylation G72 is the weakest ligand we have found for the 3.L2 TCR, it might be expected that this ligand is the best candidate for induction of positive selection. In this FTOC system, we never observed any increase in the number of Cabhigh CD4SP cells with any of the ligands tested. This included G72 as well as Q72. The Q72 peptide does not stimulate mature T cells by any assay that we have, but when expressed in vivo, we have found it to induce a slight increase in the number of Cabhigh CD4SP cells (9). It could be that ligands that induce positive selection of thymocytes bearing MHC class II-restricted TCRs do not have effects on mature T cells that we can measure, but in FTOC, neither G72 nor Q72 induced any positive selection, and the reason for this may be because the selection on the endogenous repertoire is at a maximum. This is not likely because the selection of these 3.L2-bearing thymocytes is fairly inefficient and we have observed a slight increase when Q72 was expressed in vivo. A more likely explanation is that the delivery of peptides to the thymic lobe in FTOC does not result in presentation of the peptide on the proper APCs and/or at a high enough density required for positive selection of MHC class II-restricted thymocytes. Thus, analysis of peptides in positive selection of thymocytes bearing class II-restricted TCRs may require more tedious in vivo expression systems.

CD8SP cells with a class II-restricted TCR

Surprisingly, when added to the FTOC system, many of the ligands we have studied result in an increase in the number of CD8SP thymocytes that express high levels of a class II-restricted TCR. Cells that have such a mismatch between TCR specificity and coreceptor can sometimes be observed in vivo, but usually only under unique circumstances: ectopic expression of Bcl-2 and an intracellular form of Notch are two examples (30, 31). Interestingly, the Bcl-2 and Notch gene products are known to interfere with apoptosis pathways (30, 32), and the ligands that induced CD8SP cells in the FTOC system also induce negative selection (both in vivo and in FTOC). Our finding that none of these ligands induces CD8SP cells in vivo suggests that they are revealed only by the FTOC system. These cells then are probably in the process of negative selection and are unable to efficiently complete that process in the FTOC system. However, negative selection is functional in FTOC because at high concentrations of Hb(64–76), almost all CAbhigh cells are deleted. A correlate exists in vivo because when in vivo apoptosis pathways are compromised (due to Bcl-2 or intracellular Notch expression (30, 31)), CD8SP thymocytes with a class II-restricted TCR appear. This is only a correlation, and the relationship between the cells in FTOC and those in vivo is not known. However, it does seem clear that the presence of CD8SP CAbhigh cells in our FTOC experiments suggests that the appearance of CD8SP cells in FTOC does not necessarily indicate positive selection into that lineage.


   Acknowledgments
 
We thank Jerri Smith for the manuscript preparation, Steve Horvath for peptide synthesis, and Darren Kreamalmeyer for animal husbandry. We also thank Kris Hogquist for her advice on FTOC.


   Footnotes
 
1 This work was supported by grants from the National Institutes of Health. Back

2 Current address: Department of Pathology and Laboratory Medicine, 7301 Woodruff Memorial Building, Emory University, 1639 Pierce Drive, Atlanta, GA 30322. Back

3 Address correspondence and reprint requests to Dr. Paul M. Allen, Department of Pathology, Campus Box 8118, Washington University School of Medicine, 660 South Euclid Avenue, St. Louis, MO 63110. Back

4 Abbreviations used in this paper: FTOC, fetal thymic organ culture; SP, single positive; Hb, hemoglobin; APL, altered peptide ligand; CAb, clonotype Ab. Back

Received for publication January 7, 2000. Accepted for publication March 20, 2000.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Sebzda, E., S. Mariathasan, T. Ohteki, R. Jones, M. F. Bachmann, P. S. Ohashi. 1999. Selection of the T cell repertoire. Annu. Rev. Immunol. 17:829.[Medline]
  2. Robey, E., B. J. Fowlkes. 1994. Selective events in T cell development. Annu. Rev. Immunol. 12:675.[Medline]
  3. Sebzda, E., V. A. Wallace, J. Mayer, R. S. M. Yeung, T. W. Mak, P. S. Ohashi. 1994. Positive and negative thymocyte selection induced by different concentrations of a single peptide. Science 263:1615.[Abstract/Free Full Text]
  4. Ashton-Rickardt, P. G., A. Bandeira, J. R. Delaney, L. Van Kaer, H.-P. Pircher, R. M. Zinkernagel, S. Tonegawa. 1994. Evidence for a differential avidity model of T cell selection in the thymus. Cell 76:651.[Medline]
  5. 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]
  6. Hogquist, K. A., M. A. Gavin, M. J. Bevan. 1993. Positive selection of CD8+ T cells induced by major histocompatibility complex binding peptides in fetal thymic organ culture. J. Exp. Med. 177:1469.[Abstract/Free Full Text]
  7. Alam, S. M., P. J. Travers, J. L. Wung, W. Nasholds, S. Redpath, S. C. Jameson, N. R. J. Gascoigne. 1996. T-cell-receptor affinity and thymocyte positive selection. Nature 381:616.[Medline]
  8. Oehen, S., L. Feng, Y. Xia, C. D. Surh, S. M. Hedrick. 1996. Antigen compartmentation and T helper cell tolerance induction. J. Exp. Med. 183:2617.[Abstract/Free Full Text]
  9. Williams, C. B., D. L. Engle, G. J. Kersh, M. J. White, P. M. Allen. 1999. A kinetic threshold between negative and positive selection based on the longevity of the T cell receptor-ligand complex. J. Exp. Med. 189:1531.[Abstract/Free Full Text]
  10. Ignatowicz, L., J. Kappler, P. Marrack. 1996. The repertoire of T cells shaped by a single MHC/peptide ligand. Cell 84:521.[Medline]
  11. Liu, C.-P., D. Parker, J. Kappler, P. Marrack. 1997. Selection of antigen-specific T cells by a single IEk peptide combination. J. Exp. Med. 186:1441.[Abstract/Free Full Text]
  12. Fukui, Y., T. Ishimoto, M. Utsuyama, T. Gyotoku, T. Koga, K. Nakao, K. Hirokawa, M. Katsuki, T. Sasazuki. 1997. Positive and negative CD4+ thymocyte selection by a single MHC class II/peptide ligand affected by its expression level in the thymus. Immunity 6:401.[Medline]
  13. Barton, G. M., A. Y. Rudensky. 1999. Requirement for diverse, low-abundance peptides in positive selection of T cells. Science 283:67.[Abstract/Free Full Text]
  14. Tourne, S., T. Miyazaki, A. Oxenius, L. Klein, T. Fehr, B. Kyewski, C. Benoist, D. Mathis. 1997. Selection of a broad repertoire of CD4+ T cells in H-2 Ma0/0 mice. Immunity 7:187.[Medline]
  15. Martin, W. D., G. G. Hicks, S. K. Mendiratta, H. I. Leva, H. E. Ruley, L. Van Kaer. 1996. H2-M mutant mice are defective in the peptide loading of class II molecules, antigen presentation, and T cell repertoire selection. Cell 84:543.[Medline]
  16. Surh, C. D., D.-S. Lee, W.-P. Fung-Leung, L. Karlsson, J. Sprent. 1997. Thymic selection by a single MHC/peptide ligand produces a semidiverse repertoire of CD4+ T cells. Immunity 7:209.[Medline]
  17. Spain, L. M., L. J. Berg. 1992. Developmental regulation of thymocyte susceptibility to deletion by "self"-peptide. J. Exp. Med. 176:213.[Abstract/Free Full Text]
  18. Spain, L. M., J. L. Jorgensen, M. M. Davis, L. J. Berg. 1994. A peptide antigen antagonist prevents the differentiation of T cell receptor transgenic thymocytes. J. Immunol. 152:1709.[Abstract]
  19. Volkmann, A., T. Barthlott, S. Weiss, R. Frank, B. Stockinger. 1998. Antagonist peptide selects thymocytes expressing a class II major histocompatibility complex-restricted T cell receptor into the CD8 lineage. J. Exp. Med. 188:1083.[Abstract/Free Full Text]
  20. Wang, R., A. Nelson, K. Kimachi, H. M. Grey, A. G. Farr. 1998. The role of peptides in thymic positive selection of class II major histocompatibility complex-restricted T cells. Proc. Natl. Acad. Sci. USA 95:3804.[Abstract/Free Full Text]
  21. Nakano, N., R. Rooke, C. Benoist, D. Mathis. 1997. Positive selection of T cells induced by viral delivery of neopeptides to the thymus. Science 275:678.[Abstract/Free Full Text]
  22. Kersh, G. J., D. L. Donermeyer, K. E. Frederick, J. M. White, B. L. Hsu, P. M. Allen. 1998. TCR transgenic mice in which usage of transgenic {alpha}- and ß-chains is highly dependent on the level of selecting ligand. J. Immunol. 161:585.[Abstract/Free Full Text]
  23. Kouskoff, V., H.-J. Fehling, M. Lemeur, C. Benoist, D. Mathis. 1993. A vector driving the expression of foreign cDNAs in the MHC class II-positive cells of transgenic mice. J. Immunol. Methods 166:287.[Medline]
  24. Evavold, B. D., S. G. Williams, B. L. Hsu, S. Buus, P. M. Allen. 1992. Complete dissection of the Hb(64–76) determinant using Th1, Th2 clones, and T cell hybridomas. J. Immunol. 148:347.[Abstract]
  25. Kersh, G. J., P. M. Allen. 1996. Structural basis for T cell recognition of altered peptide ligands: a single T cell receptor can productively recognize a large continuum of related ligands. J. Exp. Med. 184:1259.[Abstract/Free Full Text]
  26. Kersh, G. J., E. N. Kersh, D. H. Fremont, P. M. Allen. 1998. High- and low-potency ligands with similar affinities for the TCR: the importance of kinetics in TCR signaling. Immunity 9:817.[Medline]
  27. 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]
  28. Kirberg, J., A. Baron, S. Jakob, A. Rolink, K. Karjalainen, H. von Boehmer. 1994. Thymic selection of CD8+ single positive cells with a class II major histocompatibility complex-restricted receptor. J. Exp. Med. 180:25.[Abstract/Free Full Text]
  29. Matechak, E. O., N. Killeen, S. M. Hedrick, B. J. Fowlkes. 1996. MHC class II-specific T cells can develop in the CD8 lineage when CD4 is absent. Immunity 4:337.[Medline]
  30. Linette, G. P., M. J. Grusby, S. M. Hedrick, T. H. Hansen, L. H. Glimcher, S. J. Korsmeyer. 1994. Bcl-2 is upregulated at the CD4+ CD8+ stage during positive selection and promotes thymocyte differentiation at several control points. Immunity 1:197.[Medline]
  31. Robey, E., D. Chang, A. Itano, D. Cado, H. Alexander, D. Lans, G. Weinmaster, P. Salmon. 1996. An activation form of notch influences the choice between CD4 and CD8 T cell lineages. Cell 87:483.[Medline]
  32. Deftos, M. L., Y. W. He, E. W. Ojala, M. J. Bevan. 1998. Correlating notch signaling with thymocyte maturation. Immunity 9:777.[Medline]
  33. van Meerwijk, J. P., R. N. Germain. 1993. Development of mature CD8+ thymocytes: selection rather than instruction?. Science 261:911.[Abstract/Free Full Text]
  34. Bendelac, A., P. Matzinger, R. A. Seder, W. E. Paul, R. H. Schwartz. 1992. Activation events during thymic selection. J. Exp. Med. 175:731.[Abstract/Free Full Text]
  35. Kersh, G. J., S. M. Hedrick. 1995. Role of TCR specificity in CD4 versus CD8 lineage commitment. J. Immunol. 154:1027.
  36. McKeithan, T. W.. 1995. Kinetic proofreading in T-cell receptor signal transduction. Proc. Natl. Acad. Sci. USA 92:5042.[Abstract/Free Full Text]
  37. Rabinowitz, J. D., C. Beeson, D. S. Lyons, M. M. Davis, H. M. McConnell. 1996. Kinetic discrimination in T-cell activation. Proc. Natl. Acad. Sci. USA 93:1401.[Abstract/Free Full Text]
  38. Lyons, D. S., S. A. Lieberman, J. Hampl, J. J. Boniface, Y.-H. Chien, L. J. Berg, M. M. Davis. 1996. A TCR binds to antagonist ligands with lower affinities and faster dissociation rates than to agonists. Immunity 5:53.[Medline]
  39. Davis, M. M., J. J. Boniface, Z. Reich, D. Lyons, J. Hampl, B. Arden, Y.-H. Chien. 1998. Ligand recognition by {alpha}ß T cell receptors. Annu. Rev. Immunol. 16:523.[Medline]
  40. Nakayama, T., A. Singer, E. D. Hsi, L. E. Samelson. 1989. Intrathymic signalling in immature CD4+CD8+ thymocytes results in tyrosine phosphorylation of the TCR {zeta} chain. Nature 341:651.[Medline]
  41. van Oers, N. S. C., N. Killeen, A. Weiss. 1994. Zap-70 is constitutively associated with tyrosine-phosphorylated TCR {zeta} in murine thymocytes and lymph node T cells. Immunity 1:675.[Medline]
  42. Peterson, D. A., R. J. DiPaolo, O. Kanagawa, E. R. Unanue. 1999. Negative selection of immature thymocytes by a few peptide-MHC complexes: differential sensitivity of immature and mature T cells. J. Immunol. 162:3117.[Abstract/Free Full Text]
  43. Davey, G. M., S. L. Schober, B. T. Endrizzi, A. K. Dutcher, S. C. Jameson, K. A. Hogquist. 1998. Preselection thymocytes are more sensitive to T cell receptor stimulation than mature T cells. J. Exp. Med. 188:1867.[Abstract/Free Full Text]
  44. Vasquez, N. J., L. P. Kane, S. M. Hedrick. 1994. Intracellular signals that mediate thymic negative selection. Immunity 1:45.[Medline]
  45. Finkel, T. H., M. McDuffie, J. W. Kappler, P. Marrack, J. C. Cambier. 1987. Both immature and mature T cells mobilize Ca2+ in response to antigen receptor crosslinking. Nature 330:179.[Medline]
  46. Havran, W. L., M. Poenie, J. Kimura, R. Tsien, A. Weiss, J. P. Allison. 1987. Expression and function of the CD3-antigen receptor on murine CD4+8+ thymocytes. Nature 330:170.[Medline]



This article has been cited by other articles:


Home page
 J. Immunol.Home page
L. M. Relland, M. K. Mishra, D. Haribhai, B. Edwards, J. Ziegelbauer, and C. B. Williams
Affinity-Based Selection of Regulatory T Cells Occurs Independent of Agonist-Mediated Induction of Foxp3 Expression
J. Immunol., February 1, 2009; 182(3): 1341 - 1350.
[Abstract] [Full Text] [PDF]

Home page
 JEMHome page
H. Kao and P. M. Allen
An antagonist peptide mediates positive selection and CD4 lineage commitment of MHC class II-restricted T cells in the absence of CD4
J. Exp. Med., January 3, 2005; 201(1): 149 - 158.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
T. A. Baldwin, K. A. Hogquist, and S. C. Jameson
The Fourth Way? Harnessing Aggressive Tendencies in the Thymus
J. Immunol., December 1, 2004; 173(11): 6515 - 6520.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
E. Maverakis, J. Beech, D. B. Stevens, A. Ametani, L. Brossay, P. van den Elzen, R. Mendoza, Q. Thai, L. H. Macias, D. Ethell, et al.
Autoreactive T cells can be protected from tolerance induction through competition by flanking determinants for access to class II MHC
PNAS, April 29, 2003; 100(9): 5342 - 5347.
[Abstract] [Full Text] [PDF]

Home page
 JEMHome page
P. Kraj, R. Pacholczyk, H. Ignatowicz, P. Kisielow, P. Jensen, and L. Ignatowicz
Positive Selection of CD4+ T Cells Is Induced In Vivo by Agonist and Inhibited by Antagonist Peptides
J. Exp. Med., August 13, 2001; 194(4): 407 - 416.
[Abstract] [Full Text] [PDF]

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 Kersh, G. J.
Right arrow Articles by Allen, P. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kersh, G. J.
Right arrow Articles by Allen, P. M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS