volume 11 number 1 january 2010 nature immunology
1Harvard Medical School and Dana Farber Cancer Institute, Boston,
Massachusetts, USA. 2Max Planck Institute for Infection Biology, Campus
Charite, Berlin, Germany. Correspondence should be addressed to H.v.B.
(email@example.com) or F.M. (firstname.lastname@example.org).
Published online 17 December 2009; doi:10.1038/ni.1794
Checkpoints in lymphocyte development
and autoimmune disease
Harald von Boehmer1 & Fritz Melchers2
Antigen receptor–controlled checkpoints in B lymphocyte development are crucial for the prevention of autoimmune diseases such
as systemic lupus erythematosus. Checkpoints at the stage of pre–B cell receptor (pre-BCR) and BCR expression can eliminate
certain autoreactive BCRs either by deletion of or anergy induction in cells expressing autoreactive BCRs or by receptor editing. For
T cells, the picture is more complex because there are regulatory T (Treg) cells that mediate dominant tolerance, which differs from
the recessive tolerance mediated by deletion and anergy. Negative selection of thymocytes may be as essential as Treg cell generation
in preventing autoimmune diseases such as type 1 diabetes, but supporting evidence is scarce. Here we discuss several scenarios in
which failures at developmental checkpoints result in autoimmunity.
T lymphocytes develop in the thymus, whereas B lymphocytes develop in
the bursa of Fabricius in birds and in fetal liver or adult bone marrow in
mammals. Undifferentiated lymphocyte precursors enter these primary
lymphoid organs and begin to differentiate and rearrange antigen recep-
tor loci. T and B cell antigen receptors (TCR and BCR) are generated by
joining of V, D and J segments at the Tcr and Ig loci. Rearrangements
of these segments are ordered according to time and cell differentia-
tion stage. At the Tcrb, Tcrd and Igh loci, D-to-J rearrangements occur
before V-to-DJ rearrangements. V-to-J rearrangements at the Tcra loci
occur later, as do rearrangements at the Igk and Igl loci. Furthermore,
N-region sequences of differing lengths are inserted at most joining
sites. Although some V(D)J segments rearrange more frequently than
others1,2, the emerging TCR and BCR repertoires contain unpredictably
diverse V segments and certainly include receptors capable of binding
to autoantigens. As lymphocytes are continuously generated through-
out life from receptor-negative progenitors, there is a lifelong need to
establish tolerance to self. In this review we discuss the ways in which
checkpoints during T and B cell development facilitate the induction
of self tolerance.
B cells exist in at least three distinct lineages: B1a, B1b and B2 cells.
Most B1a cells develop during the prenatal period in the fetal liver from
precursors that lack lg rearrangements and are distinct from those that
give rise to B1b and B2 cells3,4. B1a cells persist after birth, even though
they are derived from precursors that disappear during the neonatal
period. B1a cells contribute a large part of the ‘natural’ immunoglobu-
lin M (IgM) antibodies in serum, respond to T cell–independent anti-
gens, express unmutated V regions, contain very few, if any, N-region
sequences in their complementarity-determining region 3 (CDR3)
sequences and need the spleen to develop5. B1a cells can also be found
in the gut-associated lymphoid tissues, where some of them secrete
IgA, which can influence host responses to commensal bacterial flora6.
B1b cells, by contrast, contain N-regions in their CDR3 sequences and
develop largely from precursors in the bone marrow. Some can also
express mutated IgG. B1b cells also respond to T cell–independent anti-
gens and seem essential for the production of antibodies against some
pathogenic bacteria and parasites. B2 cells are generated in the bone
marrow throughout life, can respond to T cell–dependent antigens in
germinal centers and are required for adaptive immunity.
T cell lineages are more numerous, but all lineages, with the possible
exception of fetal γδ T cells7, are derived from the same intrathymic
precursors. Tcrb rearrangement results in the generation of the pre–T
cell receptor (pre-TCR)8, which together with Notch commits progeni-
tor cells to the αβ lineage9. Productive Tcrg and Tcrd rearrangement
results in expression of a γδ TCR, which usually commits cells to the
γδ lineage10. Tcra rearrangement in αβ-lineage thymocytes results in
expression of the αβ TCR, whose specificity is crucial in determining
whether cells die from neglect or are positively or negatively selected11.
TCR specificity also directs differentiation of positively selected thymo-
cytes into CD4 helper, CD8 cytotoxic or CD4 Treg cell lineages12.
Further lineage subdivisions generate relatively small subsets of cells
with innate properties, such as natural killer T (NKT) cells, again most
likely in a manner instructed by TCR signals11. Thus, in developing T
cells the TCR determines lineage fate. In contrast, the decision to enter
B1a versus B1b or B2 lineage fate can apparently precede immuno-
globulin expression, although a contribution of the BCR has not been
Existence of autoreactive BCRs
Binding to ‘self’ antigens such as DNA or insulin has been used as a
readout to designate BCRs as autoreactive. Similarly, binding to viral or
bacterial antigens has been considered indicative of non-self-reactive
BCRs13. It is estimated that more than half of all newly generated BCRs
© 2010 Nature America, Inc. All rights reserved.
nature immunology volume 11 number 1 january 2010
are capable of binding autoantigen14.The reactivity or responsiveness
of B cells to antigens and autoantigens remains imprecisely defined,
however, as the correlation between the binding affinity of an antigen to
a BCR and the functional response of the B cell expressing this BCR can-
not be given as a quantitative term. Furthermore, such antigen-sensing
mechanisms are not necessarily the same in developmentally distinct
subsets of lymphocytes.
In addition, not all autoantigen-binding BCRs are necessarily det-
rimental to the organism. For example, a certain proportion of BCRs
may bind autoantigen with too low an affinity to trigger an autoim-
mune response, but may bind strongly enough to invading pathogens to
exert a protective host-defense effect. In addition, newly generated BCR
repertoires include a large number of BCRs that seem to defy the immu-
nological dogma of ‘one BCR, one antigen binding specificity’; each of
these BCRs binds several antigens and autoantigens (for example, DNA,
different intracytoplasmic structures, lipopolysaccharides) and hence
are called ‘polyreactive’15. Many of these autoreactive and polyreactive
antibodies have long CDR3 regions in immunoglobulin heavy chain V
domains that contain positively charged and/or aromatic amino acid
residues16,17. Although not all BCRs that bind autoantigen in one assay
are potentially dangerous, mechanisms are in place to purge or control
those autoreactive B cells that are harmful.
Purging and controlling autoreactive BCRs
More than 98% of mature B2 cells are in the G0-phase of the cell cycle and
thus are probably not activated or functionally autoreactive. However, the
B1 population seems to be in a G1-like state but not in S phase. Weakly
autoreactive and polyreactive B cells are detectable in this population18,
and it was argued that such B cells are positively selected, with the poten-
tial to generate a first line of defense against infectious organisms. Such
cells may become more easily active in autoimmune disease.
Hence, the mature B1 and B2 cells emerge only after passing through
checkpoints, several of which enforce self-tolerance. During the first
checkpoint, newly generated µ heavy chains encoded by productively
rearranged Igh alleles are probed for their ability to form a pre-BCR by
binding to a surrogate light chain consisting of λ5 and VpreB proteins on
the surface of precursor B cells called pre-B2 cells19 (Fig. 1). Signals ema-
nating from the resulting pre-BCRs induce proliferation. Cross-linking
of pre-BCRs through positively charged arginine residues in the non-
immunoglobulin portion of λ5, possibly mediated by repetitive nega-
tive charges on molecules such as nucleic acids and other molecules on
stromal cells20,21, initiates pre-BCR signaling (Fig. 1). The pre-BCR also
subsequently induces the downregulation of surrogate light chain and
recombinase activating gene (RAG) expression, thereby limiting pro-
liferation of µ-chain-expressing pre-B2 cells and preventing VDJ rear-
rangements on the second Igh allele. At least 50% to 70% of all originally
generated heavy chains are unfit to pair with surrogate light chain and
hence do not induce proliferation of pre-B2 cells through pre-BCRs. In
situations where pre-BCRs cannot be formed, such nonproliferating pre-
B2 cells nevertheless proceed in differentiation22–24. However, because
such pre-BCR-defective pre-B cells do not proliferate, their contribution
to the developing B cell compartment should be at least 20- to 40-fold
lower than that of their pre-BCR-expressing counterparts25.
In human B cell development, many polyreactive cells13 and some
but not all autoreactive cells are lost during the transition from pre-B2
to immature B cells. In people with systemic lupus erythematosus (SLE)
and rheumatoid arthritis, however, these autoreactive and polyreactive
BCRs are not lost at this checkpoint26, providing strong evidence of the
role of this checkpoint in preventing autoimmune disease. In apparent
contrast to these results, another study reported that a subset of poly-
reactive pre-BCRs expressed in pre-B2-like cells induces a proliferative
burst in tissue culture, suggesting positive rather than negative selection
by polyreactive pre-BCRs27. However, other experiments indicate that
pre-BCR cross-linking in fetal liver organ culture does not boost pre-B
Pre-BCR formation leads to a ‘shift’ in the originally generated heavy
chain repertoire3, and this shift is delayed in surrogate light chain–defi-
cient bone marrow. This shift facilitates a loss of polyreactive µ-chains
containing positively charged and aromatic amino acid residues and thus
may represent negative selection of µ-chain repertoires by the pre-BCR29.
It is not clear how much of this negative selection of pre-BCRs is medi-
ated by binding of autoantigens to CDRs of µ-chains and how much is
due to lack of pairing with surrogate light chain. A quantitative analysis
of the relative numbers of originally generated heavy chains that undergo
positive, negative or no selection through the pre-BCR might help to
resolve some of the apparent discrepancies mentioned above.
Autoreactive BCRs can apparently also be purged by receptor edit-
ing, a process through which antigen binding induces continued rear-
rangement of immunoglobulin gene segments; this process results in a
change in the specificity of a previously autoreactive BCR. V regions of
both heavy and light immunoglobulin chains can be edited by exchange
with other V regions. VH replacement30–33 occurs predominantly in pre-
BCR-expressing pre-B cells and hence influences tolerance at that check-
point34. During the process of light chain replacement, evolutionarily
selected ‘editor’ immunoglobulin light chains that have low isoelectric
points (owing, for example, to the presence of aspartate residues) seem
to ‘neutralize’ the DNA-binding properties of certain heavy chains; this
Figure 1 Pre-BCR cross-linking. In the pre-BCR, the VH domain of the
µ-chain associates with the Vpre-B subunit of the surrogate light chain
through classical VH–VL contacts (horizontal arrows). The arginine-rich
non-immunoglobulin portion of the λ5 subunit of SL chain (red line)
interacts with oppositely charged residues on stromal cells or with charged
molecules such as DNA. These interactions cross-link pre-BCRs and induce
pre-BCR signaling. Some CDR3 regions of µ-chains may contain several
arginine residues and may take over the cross-linking function of the non-
immunoglobulin portion of λ5. An accumulation of tyrosine residues in
CDR3 regions may function in a similar manner.
© 2010 Nature America, Inc. All rights reserved.
volume 11 number 1 january 2010 nature immunology
may facilitate avoidance of diseases, such as systemic lupus erythematosus
(SLE), mediated by antibody–DNA complexes35. Continued rearrange-
ment can also lead to the expression of two or even more different light
chains (and possibly also two different heavy chains) in a single immature
B cell36–38. If V-gene replacement results in the expression of one auto-
reactive and one non-autoreactive BCR, immature B cells can become
unreactive to autoantigen because of dilution of the autoreactive with the
non-autoreactive BCR. These dual-expressor B cells can enter the mature
B cell pool while remaining potentially autoreactive. Light chain replace-
ment can also lead to polyreactivity26. Interestingly, some human B cells
coexpress Vpre-B and conventional light chains together with µ-chains
containing CDR3 regions enriched in positively charged and/or aromatic
amino acids. Two-thirds of these cells are autoreactive and hence seem
to have escaped central tolerance. Although human Vpre-B, which has an
isoelectric point of 5.67, may act as an ‘editor’ by neutralizing positively
charged CDR3 regions, mouse Vpre-B proteins have an isoelectric point
of 9.37 and thus may not be able to perform this function39.
While VH repertoires expressed in immature B cells in the bone mar-
row and in immature and mature B cells in spleen are not significantly
different, almost 90% of the newly formed immature B cells never
leave the bone marrow40. Therefore they should be subject to addi-
tional deletion mechanisms. However, hardly any cell loss is detected
during the transition from immature (including transitional T1 and
T2 cells) to mature B cells in the spleen. In humans there is a further
reduction of the percentage of autoreactive cells (40% to 20%) across
this transition, whereas the percentage of polyreactive cells remains at
the low level (6%) already achieved during the transition from pre-B
to immature B cells13. Again, individuals with SLE and rheumatoid
arthritis fail to establish recessive tolerance even at this second check-
point, indicating that failure of several mechanisms for establishing
tolerance may contribute to these diseases41.
It is remarkable that almost nothing is known with regard to the
autoantigen-presenting modes that establish central B cell tolerance.
However, genetic deficiencies in the complement components C1q, C4,
serum amyloid protein, complement receptor-2 or secreted natural serum
IgM lead to systemic autoimmune disease characterized by production
of autoantibodies to DNA and other nuclear antigens42. Two models
have been proposed to explain the emergence of autoantigen-reactive
B cell repertoires in these knockout mice. In one model43, macrophages
expressing the appropriate complement receptors (C1qR, CR1) efficiently
remove apoptotic cells that are bound by natural IgM, C1q and C4b,
thereby preventing accumulation of these cells and subsequent activation
of mature B cells. The other model42 suggests that autoantigens from
apoptotic cells are presented to immature B cells by immune complexes
containing C1q, C4b and IgM on as-yet-unidentified cells expressing the
appropriate receptors (C1qR, CR1, possibly FcRµ) (Fig. 2). In the latter
model, this antigen presentation can result in negative selection and/or
anergy. A better understanding of the modes of central tolerance induced
by apoptosis, editing, anergy and ignorance requires identification of the
relevant antigen-presenting cell populations.
How can B cell tolerance fail?
Prolongation of the short half-life of immature B cells, brought about
either by a decrease in expression and/or activity of proapoptotic genes or
by an increase in prosurvival gene expression (for example, B lymphocyte-
activating factor, BAFF) favors B cell–driven autoimmunity that manifests
as SLE-like syndromes44,45. Elevated expression of the genes encoding Toll-
like receptors46,47 can promote similar autoimmune manifestations.
Signaling defects can also impair central tolerance induction in imma-
ture B cells. For example, the z alleles of Sle1 and its sublocus Sle1b derived
from the NZM2410 mouse strain, as well as of the orthologous human
locus, impair B cell anergy, receptor editing and deletion48,49 Members
of the SLAM family of costimulatory molecules are candidate Sle1b gene
products, among which the Ly108.1 isoform is highly expressed in imma-
ture B cells50. The Ly108.2 isoform, but not the SLE-associated Ly108.1
isoform, sensitizes immature B cells to deletion and RAG reexpression.
When the z allele of Sle1 is combined with the Fas lpr mutation, lym-
phoproliferative autoimmunity induced by the phosphatidylinositol-
3-kinase–AKT–mTOR pathway is induced. As the signaling pathways
in immature B cells become known in greater detail, more genes that
interfere with the proper establishment of B cell central tolerance are
likely to be identified.
Selection of γδ and αβ lineage cells
Successful Tcrb rearrangement results in the formation of a pre-TCR,
which signals in a cell-autonomous fashion without any apparent require-
ment for a ligand. Polar residues in the pre-TCR α-chain extracellular
domain likely contribute to cell autonomous signaling51, as does the cyto-
plasmic tail52, but these features may not represent the only mechanisms
facilitating ligand-independent signaling, as they are not evolutionarily
conserved. The only conserved portion of the pre-TCR α-chain is the
transmembrane region. Notably, the pre-TCR terminates further Tcrb
rearrangement but, unlike the pre-BCR, does not select cells expressing
particular Tcrb chains; the distribution of Tcrb V regions in unselected
and pre-TCR–selected thymocytes is the same2. All pre-TCR–expressing
cells commit to the αβ lineage with the help of Notch signaling and sub-
sequently undergo Tcra rearrangement.
Successful Tcra rearrangements can lead to expression of an αβ TCR
on the cell surface. In immature B cells, expression of a surface immuno-
globulin eventually halts further Ig rearrangement. In contrast, spontane-
Figure 2 Model of involvement of complement and natural serum IgM in B
cell central tolerance. Genetic defects resulting in lupus-like autoimmune
disease suggest the (hypothetical) picture in which, at the transition
of immature (Bimm) to mature (Bmat) B cells, BCRs could interact with
autoantigen (pink ellipses), which, in turn, would be bound by natural serum
IgM and complement components (C1q, C4). High-avidity BCR–autoantigen
(auto-Ag) interactions would induce light-chain editing, whereas low-
avidity interactions would induce B cell anergy. As-yet-unknown antigen-
presenting stromal cells expressing receptors for these immune complexes
would influence immature B cells through cell–cell contacts and cytokines,
including B cell activating factor (BAFF), produced by dendritic cells.
BAFF-R, BAFF receptor; CR1 and C1qR, complement receptors; FcRµ, IgM
constant region receptor; TLR, Toll-like receptor.
© 2010 Nature America, Inc. All rights reserved.
nature immunology volume 11 number 1 january 2010
ous continued Tcra rearrangement is halted only when the expressed αβ
TCR binds to an intrathymic ligand53. As a consequence, all mature αβ
T cells contain two rearranged Tcra alleles, and 30% of such cells express
two productive Tcra rearrangements54. This does not mean, however,
that that 30% of αβ T cells express two TCRs on the cell surface55, as
there are heavy preferences among TCRα chains for pairing with certain
TCRβ chains56, so that in the mouse only about 10% of T cells express
two different TCRα chains on the cell surface57. When a particular αβ
TCR binds with relatively low affinity to an intrathymic ligand, Tcra
rearrangement is halted and the cell is selected for survival and further
maturation. Stronger and/or longer positively selecting signals generated
as a result of co-ligation of the TCR and the CD4 co-receptor results in
instruction of CD4 helper lineage fate while weaker and/or shorter signals
initiated by the TCR and the CD8 co-receptor result in CD8 cytotoxic
lineage fate determination58,59. Very strong signaling, however, can result
in apoptosis60—referred to as negative selection—of αβ-TCR+ thymo-
cytes or in diversion of developing αβ-TCR+ cells into unconventional
lineages involved in innate immunity.
Unconventional T cell populations
Most cells with productive Tcrd and Tcrg rearrangements commit to
the γδ lineage without any allelic exclusion of TCRγ or TCRδ chains61.
It is in fact not the isotype of the TCR but the signaling strength62,63
that commits cells to either the αβ or γδ lineage. Some 20% of γδ
TCR–expressing cells that undergo weak signaling can still enter the
αβ lineage by silencing TCRγ expression and deleting the Tcrd locus
by Tcra rearrangement10. Little is known about tolerance mechanisms
in γδ T cells except that TCR ligands seem to have a role in committing
these cells to the γδ lineage62,63 rather than in causing apoptosis. In
fact, TCR ligation results in the generation of a subset of γδ T cells that
assume the characteristics of innate lymphocytes in that they express
the PLZF transcription factor, which is also is expressed in NKT cells64.
Conceivably such cells may exert so-far-unknown regulatory functions
and could thereby influence autoimmune disease.
Apart from the ‘conventional’ αβ T cell development pathways described
above, some immature, αβ-TCR-expressing CD4+CD8+ (double positive)
cells receive signals that result in their assuming an NKT cell phenotype65.
In this case, the presumed TCR ligand is expressed on lymphoid cells
rather than stromal cells65. The latter instruct lineage fate of conventional
CD4+ and CD8+ T cell lineages. NKT cells have been postulated to be able
to prevent autoimmunity such as that in type 1 diabetes66, but this issue is
somewhat controversial and will not be further discussed here67. In addi-
tion, whereas conventional cytotoxic T cells express CD8-αβ heterodim-
ers, some unconventional T cells express the CD8-αα homodimer68: the
latter population may have a regulatory role in immunity of the gut tissue,
but such a role remains hypothetical69.
Thus, notably, intrathymic confrontation with relatively strong TCR
ligands does not always lead to negative selection through anergy or apop-
tosis. Strong TCR signals can lead to the generation of subsets of T cells,
usually with a restricted TCR repertoire, that often show innate properties
and regulate immunity in peripheral lymphoid tissue.
Selection and generation of Treg cells
High-affinity TCR ligands can also induce the generation of Treg cells that
express the transcription factor Foxp3 and have an essential role in pre-
venting autoimmune disease70,71. Several independent studies indicate that
the lineage fate of such cells is instructed by TCR binding to agonist ligands
expressed on stromal cells within the thymus71,72. Cross-presentation of
such ligands by hematopoietic cells is not required72. The Treg cell lineage
differentiation process can begin with CD4+CD8– single-positive cells
in the thymic medulla, but a contribution of Foxp3-expressing double-
positive cells in the thymic cortex is not entirely excluded73.
If Treg cell development were instructed by TCR signals, then one might
predict that a TCR from a Treg cell should instruct Treg cell development
when expressed as a transgene in mice. Surprisingly, this is not the case.
Recently it was determined that this observation is due to the existence
of saturable ‘niches’ for Treg cell development; reducing the number of
precursors with the transgenic Treg cell–derived TCR results in increased
frequencies of Treg cells expressing that particular TCR74. These findings
are analogous to earlier experiments with positively selectable transgenic
TCRs where decreasing the numbers of cells expressing such TCRs results
in an increased efficacy of positive selection75. Experiments in various
TCR-transgenic mice leave little doubt that the Treg cell lineage fate is
instructed by TCR signals induced by intrathymic76 and/or extrathy-
mic77,78 ligands. Considering that intrathymic ligands are able to instruct
Treg cell development, one might think that such ligands include peripheral
antigens whose intrathymic expression is regulated by the Aire (autoim-
mune regulator) transcriptional regulator79 or a functional equivalent. In
any case, Foxp3+-cell differentiation is likely induced by intrathymically
expressed autoantigens80. Of note, the TCR repertoires of Treg cells and
conventional CD4+ αβ T cells seem to be overlapping, which could be
explained by the observation that saturable thymic niches for Treg cell
selection do not permit all cells with a suitable TCR to be selected into the
Treg cell lineage74. The mechanisms by which Treg cells regulate autoim-
mune disease is the subject of a different review in this supplement81.
Recessive versus dominant T cell tolerance
Negative selection occurring in primary lymphoid organs, also called
recessive or central tolerance, was thoroughly addressed in the late 1980s
by analyzing the fate of developing thymocytes in response to stimu-
lation by conventional peptide–MHC complexes in TCR-transgenic
mice82, or in response to superantigens in wild-type mice83. These
studies concluded that there was a period during the ontogeny of lym-
phocytes when binding of receptors to antigen resulted in cell death or
anergy84 rather than proliferation and development into effector cells.
The conclusions were consistent with early proposals on immunologi-
cal tolerance by Burnet85 and Lederberg86, even though information
on primary lymphoid organs and lymphocyte turnover was scarce at
the time those proposals were made.
The existence of recessive T cell tolerance mechanisms involving apop-
totic cell death of immature lymphocytes60 suggested to some83, but not
others82, that dominant tolerance did not exist. Although central toler-
ance was the first to be mechanistically defined, at the same time strong
evidence for the existence of dominant tolerance was presented87,88. This
evidence culminated in the characterization of Treg cells89 and an eluci-
dation of the importance of the Foxp3 gene in both the generation and
the function of at least one subset of Treg cells90–92. Mutations in Foxp3
result in severe autoimmune disease in man and mouse93,94, providing
immediate genetic evidence for the essential role of dominant tolerance
in avoiding autoimmunity. This clear evidence put recessive tolerance in
a defensive spot, as similarly persuasive evidence for an essential role of
recessive tolerance was not available. Nevertheless, we will argue below
that failed recessive tolerance is likely to be pivotal in some of the most
prevalent autoimmune diseases.
Recessive T cell tolerance by deletion
Transgenic mice expressing a TCR specific for a ligand (HY) ubiq-
uitously expressed in male mice are characterized by the absence of
double-positive thymocytes in the thymic cortex95. This finding initially
could have been misinterpreted because the precocious expression of
the transgenic TCR leads to lineage diversion of some double-nega-
tive cells into the double-negative γδ lineage96–98. At the time, this was
© 2010 Nature America, Inc. All rights reserved.
volume 11 number 1 january 2010 nature immunology
taken as evidence by some, but not others, that in these mice there was
a developmental arrest at the double-negative stage rather than deletion
of double-positive thymocytes99. However, independent experiments
with double-positive cells from female mice—which do not express the
ligand for the HY TCR—confirmed that such cells could be deleted by
antigenic stimulation in suspension culture60 as well as organ culture100.
Furthermore, analysis of HY male lacking the proapoptotic protein
Bim101 revealed a marked increase in double-positive thymocytes. In
addition, recent experiments using HY transgenic mice in which the
TCR-α chain is expressed with physiologically relevant kinetics con-
firm antigen-induced cell death of double-positive thymocytes102, even
though the authors of that study initially disputed death of cortical thy-
mocytes as an important mechanism of immunological tolerance103.
These experiments seem to confirm very early experiments in which
CD3-specific antibodies in organ culture were shown to delete double-
positive thymocytes104. The results of these experiments, unlike their
in vivo equivalents105, cannot be explained by stress-induced cell death
of cortical thymocytes102 and hence represent further support for the
concept that negative selection can affect cortical thymocytes even before
their positive selection106.
In light of these studies, one could in fact argue that proposals that
belittle negative selection of cortical thymocytes as transgenic arti-
fact107,108 and instead invoke a role for TCR editing107 are somewhat
unrealistic. Double-positive cortical thymocytes as well as immature
medullary single-positive thymocytes can in fact be deleted106,109.
Deletion of medullary thymocytes is important for antigens that are
present in the medulla only. Presentation of tissue-specific antigens from
the periphery in the thymus is made possible either by immigration
of peripheral antigen-presenting dendritic cells110 or by promiscuous
antigen expression by medullary epithelial cells facilitated by Aire111.
Considering the phenotype of Aire-deficient mice, negative selection
in the medulla is presumably important in preventing autoimmune
However, as explained above, there exists straightforward genetic evi-
dence for a role of dominant tolerance in preventing autoimmunity93,94
Because Treg cells are generated intrathymically by antigenic stimula-
tion70,71 and because the same antigens can cause deletion71 of cells
expressing the same TCRs as Treg cells (a surprising notion that may
depend on encounter of antigen in different thymic niches), it is not
easy to determine whether certain experimental manipulations affect
recessive versus dominant tolerance. Certainly there is no formal genetic
analysis that examines whether genetic factors that contribute to type 1
diabetes, such as Ptpn22, Iddm2 or CD25 (ref.
112) do so by affecting recessive or dominant
tolerance. Strong evidence for an essential role
of recessive tolerance seems to be confined to a
monogenic autoimmune disease caused by the
disruption of the mouse Aire gene, for which
it was concluded that intrathymic deletion but
not generation of Treg cells was impaired113.
Even these particular studies could be made
more convincing if they were carried out in
genetically deficient mice that would defini-
tively exclude a role of Treg cells. More studies
addressing the contribution of recessive and
dominant tolerance in different systems are
required to establish recessive T cell tolerance
as a key contributor to the prevention of auto-
How can recessive T cell tolerance fail?
Several reports point to failed central tolerance in diabetes-prone mouse
strains. Some analyses have reported failed deletion of antigen-specific
thymocytes due to diminished expression of proapoptotic genes and
enhanced expression of prosurvival factors114–116. Furthermore, the
mode of antigen presentation seems to be important. Human suscepti-
bility factors include those that affect the amount of insulin expressed
intrathymically117, and, accordingly, in the non-obese diabetic (NOD)
mouse, thymic overexpression of insulin reduces the incidence of dia-
betes by a recessive tolerance mechanism118.
How else can tolerance to certain autoantigens fail? Mutations in
signal-transducing molecules, such as ZAP-70, result in autoimmune
disease119. This could be explained by the assumption that the mutations
disturb the balance of positive and negative selection and allow certain
T cells to escape negative selection; these T cells can then be activated
by their self ligands in peripheral lymphoid tissue. These mutants con-
ceivably could also disturb the balance of positive selection and Treg cell
generation. In analogy to expression of more than one BCR in a B cell,
expression of a second TCR in addition to an autoreactive TCR can
‘dilute’ the signaling capacity of the autoreactive TCR such that these
cells can escape negative selection120.
Of considerable interest is the notion that several class II MHC epitopes
that are recognized by disease-causing T cell clones are presented in an
MHC binding register that is unusual because it minimizes binding to
class II MHC molecules121,122. This holds true for epitopes recognized
by T cells causing experimental allergic encephalomyelitis (EAE) or mul-
tiple sclerosis122 as well as epitopes recognized in type 1 diabetes123. The
proposed scenario is that the limited availability of these epitopes in the
thymus results in incomplete recessive tolerance in the form of negative
selection, such that escaping T cell clones can be activated in antigen-
draining lymph nodes where these antigens are more abundant121 (Fig.
3). The fact that poor MHC binding was established with several inde-
pendent epitopes recognized by autoimmune disease–causing T cell
clones make this scenario persuasive even though one could argue that
the poor intrathymic presentation may also result in poor generation of
Treg cells. Thus formal genetic studies comparing recessive and dominant
tolerance in well defined experimental systems are warranted in order to
obtain conclusive evidence for whether recessive tolerance generally and
essentially contributes to the prevention of autoimmune disease.
Central tolerance is enforced by antigen receptor–controlled checkpoints
in lymphocyte development. Checkpoints acting during the stages of
ThymusBlood Pancreatic LNPancreas
Naive T cell
Figure 3 weak epitopes fail to negatively select thymocytes. Preproinsulin (PPI)-2 expression in thymic
medullary epithelial cells (MeC) is regulated by Aire. Peptide 9–23 makes a weak epitope when bound
to class II MHC I-Ag7 on dendritic cells (DC) in non-obese diabetic (NOD) mice, such that the limited
supply of this peptide results in partial failure of negative selection. In pancreatic lymph nodes (LN),
peptide 9–23, this time derived from PPI-1 expressed in pancreatic β cells, is more abundant and
hence can activate T cells specific for peptide 9–23 that escaped negative selection in the thymus.
T cell activation can result in β-cell destruction and type 1 diabetes (T1D). In accordance with this
model, mice deficient in PPI-2 show accelerated and more frequent T1D, whereas mice deficient in
PPI-1 show diminished T1D.
© 2010 Nature America, Inc. All rights reserved.
nature immunology volume 11 number 1 january 2010
pre-BCR and BCR expression purge many autoreactive BCRs by block-
ing development or by triggering receptor editing. There is convinc-
ing evidence that such checkpoints fail in autoimmune diseases such
as SLE or rheumatoid arthritis. Developing T lymphocytes are policed
by checkpoints that result in the generation of Treg cells in addition to
those that delete or anergize immature T cells expressing autoreactive
TCRs. More evidence is needed to confirm that negative selection of
thymocytes is as crucial as Treg cell generation in the induction of T
cell self tolerance. It remains challenging to interfere with the processes
that generate autoimmunity, and the identification of disease-causing
epitopes seems a good starting point in this challenge.
Supported by the Deutsche Forschungsgesellschaft (Koselleck grant P.S.ME
2764/1-1 to F.M.) and the US National Institutes of Health (ROI AI 045846, R37 AI
053102, ROI AI 051378 and POI CA 10990 to H.v.B.).
COMPETING FINANCIAL INTERESTS
The authors declare competing financial interests: details accompany the full-text
HTML version of the paper at http://www.nature.com/natureimmunology/.
Published online at http://www.nature.com/natureimmunology/.
reprints and permissions information is available online at http://npg.nature.com/
1. Yamagami, T., ten Boekel, e., Andersson, J., Rolink, A. & Melchers, F. Frequencies of
multiple IgL chain gene rearrangements in single normal or kappaL chain-deficient B
lineage cells. Immunity 11, 317–327 (1999).
von Boehmer, H. et al. Thymic selection revisited: how essential is it? Immunol. rev.
191, 62–78 (2003).
ten Boekel, e., Melchers, F. & Rolink, A.G. Changes in the V(H) gene repertoire of
developing precursor B lymphocytes in mouse bone marrow mediated by the pre-B
cell receptor. Immunity 7, 357–368 (1997).
Tung, J.w. & Herzenberg, L.A. Unraveling B-1 progenitors. curr. opin. Immunol. 19,
wardemann, H., Boehm, T., Dear, N. & Carsetti, R. B-1a B cells that link the innate
and adaptive immune responses are lacking in the absence of the spleen. J. exp. Med.
195, 771–780 (2002).
Loder, F. et al. B cell development in the spleen takes place in discrete steps and is
determined by the quality of B cell receptor-derived signals. J. exp. Med. 190, 75–89
Havran, w.L. & Allison, J.P. Developmentally ordered appearance of thymocytes
expressing different T-cell antigen receptors. nature 335, 443–445 (1988).
von Boehmer, H. & Fehling, H.J. Structure and function of the pre-T cell receptor.
annu. rev. Immunol. 15, 433–452 (1997).
Ciofani, M. et al. Obligatory role for cooperative signaling by pre-TCR and Notch during
thymocyte differentiation. J. Immunol. 172, 5230–5239 (2004).
10. Kreslavsky, T., Garbe, A.I., Krueger, A. & von Boehmer, H. T cell receptor-instructed
αβ versus γδ lineage commitment revealed by single-cell analysis. J. exp. Med. 205,
11. von Boehmer, H. Selection of the T-cell repertoire: receptor-controlled checkpoints in
T-cell development. adv. Immunol. 84, 201–238 (2004).
12. von Boehmer, H. Positive and negative selection in Basel. nat. Immunol. 9, 571–573
13. wardemann, H. & Nussenzweig, M.C. B-cell self-tolerance in humans. adv. Immunol.
95, 83–110 (2007).
14. Nemazee, D. Does immunological tolerance explain the waste in the B-lymphocyte
immune system? experiment and theory. ann. nY acad. Sci. 764, 397–401 (1995).
15. Haspel, M.V. et al. Multiple organ-reactive monoclonal autoantibodies. nature 304,
16. Radic, M.Z. et al. Residues that mediate DNA binding of autoimmune antibodies. J.
Immunol. 150, 4966–4977 (1993).
17. Barbas, S.M. et al. Human autoantibody recognition of DNA. Proc. natl. acad. Sci.
USa 92, 2529–2533 (1995).
18. Rowley, B., Tang, L., Shinton, S., Hayakawa, K. & Hardy, R.R. Autoreactive B-1 B
cells: constraints on natural autoantibody B cell antigen receptors. J. autoimmun. 29,
19. Melchers, F. The pre-B-cell receptor: selector of fitting immunoglobulin heavy chains
for the B-cell repertoire. nat. rev. Immunol. 5, 578–584 (2005).
20. Ohnishi, K. & Melchers, F. The nonimmunoglobulin portion of λ5 mediates cell-auton-
omous pre-B cell receptor signaling. nat. Immunol. 4, 849–856 (2003).
21. Bradl, H.w.J., Milius, D., Vettermann, C. & Jack, H.M. Interaction of murine precur-
sors B cell receptor with stroma cells is controlled by the unique tail of lambda 5 and
stroma cell-associated heparan sulfate. J. Immunol. 171, 2338–2348 (2003).
22. Grawunder, U., Haasner, D., Melchers, F. & Rolink, A. Rearrangement and expres-
sion of kappa light chain genes can occur without mu heavy chain expression during
differentiation of pre-B cells. Int. Immunol. 5, 1609–1618 (1993).
23. Rolink, A., Melchers, F. & Andersson, J. The SCID but not the RAG-2 gene product
is required for S mu-S epsilon heavy chain class switching. Immunity 5, 319–330
24. Grawunder, U., Rolink, A. & Melchers, F. Induction of sterile transcription from the
kappa L chain gene locus in V(D)J recombinase-deficient progenitor B cells. Int.
Immunol. 7, 1915–1925 (1995).
25. Rolink, A. et al. B cell development in mice with a defective lambda 5 gene. eur. J.
Immunol. 23, 1284–1288 (1993).
26. witsch, e.J., Cao, H., Fukuyama, H. & weigert, M. Light chain editing generates
polyreactive antibodies in chronic graft-versus-host reaction. J. exp. Med. 203,
27. Köhler, F. et al. Autoreactive B cell receptors mimic autonomous pre-B cell receptor
signaling and induce proliferation of early B cells. Immunity 29, 912–921 (2008).
28. Ceredig, R., ten Boekel, e., Rolink, A., Melchers, F. & Andersson, J. Fetal liver organ
cultures allow the proliferative expansion of pre-B receptor-expressing pre-B-II cells
and the differentiation of immature and mature B cells in vitro. Int. Immunol. 10,
29. Keenan, R.A. et al. Censoring of autoreactive B cell development by the pre-B cell
receptor. Science 321, 696–699 (2008).
30. Lutz, J., Muller, w. & Jack, H.M. VH replacement rescues progenitor B cells with two
nonproductive VDJ alleles. J. Immunol. 177, 7007–7014 (2006).
31. Koralov, S.B., Novobrantseva, T.I., Konigsmann, J., ehlich, A. & Rajewsky, K.
Antibody repertoires generated by VH replacement and direct VH to JH joining.
Immunity 25, 43–53 (2006).
32. Zhang, Z. et al. Contribution of Vh gene replacement to the primary B cell repertoire.
Immunity 19, 21–31 (2003).
33. Chen, C., Nagy, Z., Prak, e.L. & weigert, M. Immunoglobulin heavy chain gene
replacement: a mechanism of receptor editing. Immunity 3, 747–755 (1995).
34. Nakajima, P.B., Kiefer, K., Price, A., Bosma, G.C. & Bosma, M.J. Two distinct popula-
tions of H chain-edited B cells show differential surrogate L chain dependence. J.
Immunol. 182, 3583–3596 (2009).
35. Li, H., Jiang, Y., Prak, e.L., Radic, M. & weigert, M. editors and editing of anti-DNA
receptors. Immunity 15, 947–957 (2001).
36. Gerdes, T. & wabl, M. Autoreactivity and allelic inclusion in a B cell nuclear transfer
mouse. nat. Immunol. 5, 1282–1287 (2004).
37. Khan, S.N. et al. editing and escape from editing in anti-DNA B cells. Proc. natl.
acad. Sci. USa 105, 3861–3866 (2008).
38. Doyle, C.M., Han, J., weigert, M.G. & Prak, e.T. Consequences of receptor editing
at the lambda locus: multireactivity and light chain secretion. Proc. natl. acad. Sci.
USa 103, 11264–11269 (2006).
39. Meffre, e. et al. Surrogate light chain expressing human peripheral B cells produce
self-reactive antibodies. J. exp. Med. 199, 145–150 (2004).
40. Rolink, A.G. et al. Mutations affecting either generation or survival of cells influence
the pool size of mature B cells. Immunity 10, 619–628 (1999).
41. Yurasov, S. et al. Defective B cell tolerance checkpoints in systemic lupus erythema-
tosus. J. exp. Med. 201, 703–711 (2005).
42. Melchers, F. & Rolink, A.R. B cell tolerance–how to make it and how to break it. curr.
Top. Microbiol. Immunol. 305, 1–23 (2006).
43. Carroll, M.C. A protective role for innate immunity in systemic lupus erythematosus.
nat. rev. Immunol. 4, 825–831 (2004).
44. Mackay, F. et al. Mice transgenic for BAFF develop lymphocytic disorders along with
autoimmune manifestations. J. exp. Med. 190, 1697–1710 (1999).
45. Rolink, A.G., Tschopp, J., Schneider, P. & Melchers, F. BAFF is a survival and matura-
tion factor for mouse B cells. eur. J. Immunol. 32, 2004–2010 (2002).
46. Deane, J.A. et al. Control of toll-like receptor 7 expression is essential to restrict
autoimmunity and dendritic cell proliferation. Immunity 27, 801–810 (2007).
47. Shlomchik, M.J. Sites and stages of autoreactive B cell activation and regulation.
Immunity 28, 18–28 (2008).
48. Liu, Y. et al. Lupus susceptibility genes may breach tolerance to DNA by impairing
receptor editing of nuclear antigen-reactive B cells. J. Immunol. 179, 1340–1352
49. Kumar, K.R. et al. Regulation of B cell tolerance by the lupus susceptibility gene
Ly108. Science 312, 1665–1669 (2006).
50. Xie, C. et al. PI3K/AKT/mTOR hypersignaling in autoimmune lymphoproliferative
disease engendered by the epistatic interplay of Sle1b and FASlpr. Int. Immunol.
19, 509–522 (2007).
51. Yamasaki, S. et al. Mechanistic basis of pre–T cell receptor–mediated autonomous
signaling critical for thymocyte development. nat. Immunol. 7, 67–75 (2006).
52. Aifantis, I. et al. A critical role for the cytoplasmic tail of pTα in T lymphocyte devel-
opment. nat. Immunol. 3, 483–488 (2002).
53. Borgulya, P., Kishi, H., Uematsu, Y. & von Boehmer, H. exclusion and inclusion of
α and β T cell receptor alleles. cell 69, 529–537 (1992).
54. Casanova, J.L., Romero, P., widmann, C., Kourilsky, P. & Maryanski, J.L. T cell recep-
tor genes in a series of class I major histocompatibility complex-restricted cytotoxic T
lymphocyte clones specific for a Plasmodium berghei nonapeptide: implications for T
cell allelic exclusion and antigen-specific repertoire. J. exp. Med. 174, 1371–1383
55. Padovan, e. et al. expression of two T cell receptor α chains: dual receptor T cells.
Science 262, 422–424 (1993).
56. Saito, T., Sussman, J.L., Ashwell, J.D. & Germain, R.N. Marked differences in the
efficiency of expression of distinct α β T cell receptor heterodimers. J. Immunol. 143,
57. Heath, w.R. et al. expression of two T cell receptor α chains on the surface of normal
murine T cells. eur. J. Immunol. 25, 1617–1623 (1995).
© 2010 Nature America, Inc. All rights reserved.
20 Download full-text
volume 11 number 1 january 2010 nature immunology
58. Yasutomo, K., Doyle, C., Miele, L., Fuchs, C. & Germain, R.N. The duration of antigen
receptor signalling determines CD4+ versus CD8+ T-cell lineage fate. nature 404,
59. Brugnera, e. et al. Coreceptor reversal in the thymus: signaled CD4+8+ thymocytes
initially terminate CD8 transcription even when differentiating into CD8+ T cells.
Immunity 13, 59–71 (2000).
60. Swat, w., Ignatowicz, L., von Boehmer, H. & Kisielow, P. Clonal deletion of immature
CD4+8+ thymocytes in suspension culture by extrathymic antigen-presenting cells.
nature 351, 150–153 (1991).
61. Sleckman, B.P., Khor, B., Monroe, R. & Alt, F.w. Assembly of productive T cell recep-
tor δ variable region genes exhibits allelic inclusion. J. exp. Med. 188, 1465–1471
62. Haks, M.C. et al. Attenuation of γδTCR signaling efficiently diverts thymocytes to the
αβ lineage. Immunity 22, 595–606 (2005).
63. Hayes, S.M., Li, L. & Love, P.e. TCR signal strength influences αβ/γδ lineage fate.
Immunity 22, 583–593 (2005).
64. Kreslavsky, T. et al. TCR-inducible PLZF transcription factor required for innate phe-
notype of a subset of γδ T cells with restricted TCR diversity. Proc. natl. acad. Sci.
USa 106, 12453–12458 (2009).
65. Bendelac, A., Savage, P.B. & Teyton, L. The biology of NKT cells. annu. rev. Immunol.
25, 297–336 (2007).
66. Chatenoud, L. NKT cells control autoimmunity. J. clin. Invest. 110, 747–748
67. Lee, P.P. et al. Testing the NKT cell hypothesis of human IDDM pathogenesis. J. clin.
Invest. 110, 793–800 (2002).
68. Yamagata, T., Mathis, D. & Benoist, C. Self-reactivity in thymic double-positive cells
commits cells to a CD8 αα lineage with characteristics of innate immune cells. nat.
Immunol. 5, 597–605 (2004).
69. Guy-Grand, D. & Vassalli, P. Immunology. Tracing an orphan’s genealogy. Science 305,
70. Jordan, M.S. et al. Thymic selection of CD4+CD25+ regulatory T cells induced by an
agonist self-peptide. nat. Immunol. 2, 301–306 (2001).
71. Apostolou, I., Sarukhan, A., Klein, L. & von Boehmer, H. Origin of regulatory T cells
with known specificity for antigen. nat. Immunol. 3, 756–763 (2002).
72. Aschenbrenner, K. et al. Selection of Foxp3+ regulatory T cells specific for self antigen
expressed and presented by Aire+ medullary thymic epithelial cells. nat. Immunol. 8,
73. Liston, A. et al. Differentiation of regulatory Foxp3+ T cells in the thymic cortex. Proc.
natl. acad. Sci. USa 105, 11903–11908 (2008).
74. Bautista, J.L. et al. Intraclonal competition limits the fate determination of regulatory
T cells in the thymus. nat. Immunol. 10, 610–617 (2009).
75. Huesmann, M., Scott, B., Kisielow, P. & von Boehmer, H. Kinetics and efficacy of
positive selection in the thymus of normal and T cell receptor transgenic mice. cell
66, 533–540 (1991).
76. Tai, X., Cowan, M., Feigenbaum, L. & Singer, A. CD28 costimulation of developing
thymocytes induces Foxp3 expression and regulatory T cell differentiation indepen-
dently of interleukin 2. nat. Immunol. 6, 152–162 (2005).
77. Apostolou, I. & von Boehmer, H. In vivo instruction of suppressor commitment in naive
T cells. J. exp. Med. 199, 1401–1408 (2004).
78. Kretschmer, K. et al. Inducing and expanding regulatory T cell populations by foreign
antigen. nat. Immunol. 6, 1219–1227 (2005).
79. Mathis, D. & Benoist, C. A decade of AIRe. nat. rev. Immunol. 7, 645–650
80. Hsieh, C.S. A The role of TCR specificity in naturally arising CD25+ CD4+ regulatory
T cell biology. curr. Top. Microbiol. Immunol. 293, 25–42 (2005).
81. wing, K. & Sakaguchi, S. Regulatory T cells exert checks and balances on self-
tolerance and autoimmunity. nat. Immunol. 11, 7–13 (2009).
82. Kisielow, P., Bluthmann, H., Staerz, U.D., Steinmetz, M. & von Boehmer, H. Tolerance
in T-cell-receptor transgenic mice involves deletion of nonmature CD4+8+ thymocytes.
nature 333, 742–746 (1988).
83. Kappler, J.w., Roehm, N. & Marrack, P. T cell tolerance by clonal elimination in the
thymus. cell 49, 273–280 (1987).
84. Ramsdell, F. & Fowlkes, B.J. Clonal deletion versus clonal anergy: the role of the
thymus in inducing self tolerance. Science 248, 1342–1348 (1990).
85. Burnet, F.M. The clonal Selection Theory (Cambridge Univ. Press, London, 1959).
86. Lederberg, J. Genes and antibodies: do antigens bear instructions for antibody speci-
ficity or do they select cell lines that arise by mutation? Science 129, 1649–1653
87. Ohki, H., Martin, C., Corbel, C., Coltey, M. & Le Douarin, N.M. Tolerance induced by
thymic epithelial grafts in birds. Science 237, 1032–1035 (1987).
88. Salaün, J. et al. Thymic epithelium tolerizes for histocompatibility antigens. Science
247, 1471–1474 (1990).
89. Sakaguchi, S., Sakaguchi, N., Asano, M., Itoh, M. & Toda, M. Immunologic self-
tolerance maintained by activated T cells expressing IL-2 receptor α-chains (CD25).
Breakdown of a single mechanism of self-tolerance causes various autoimmune
diseases. J. Immunol. 155, 1151–1164 (1995).
90. Fontenot, J.D., Gavin, M.A. & Rudensky, A.Y. Foxp3 programs the development and
function of CD4+CD25+ regulatory T cells. nat. Immunol. 4, 330–336 (2003).
91. Khattri, R., Cox, T., Yasayko, S.A. & Ramsdell, F. An essential role for Scurfin in
CD4+CD25+ T regulatory cells. nat. Immunol. 4, 337–342 (2003).
92. Hori, S., Nomura, T. & Sakaguchi, S. Control of regulatory T cell development by
the transcription factor Foxp3. Science 299, 1057–1061 (2003).
93. Brunkow, M.e. et al. Disruption of a new forkhead/winged-helix protein, scurfin,
results in the fatal lymphoproliferative disorder of the scurfy mouse. nat. Genet.
27, 68–73 (2001).
94. Chatila, T.A. et al. JM2, encoding a fork head-related protein, is mutated in X-linked
autoimmunity-allergic disregulation syndrome. J. clin. Invest. 106, R75–R81
95. Kisielow, P., Teh, H.S., Bluthmann, H. & von Boehmer, H. Positive selection of
antigen-specific T cells in thymus by restricting MHC molecules. nature 335,
96. von Boehmer, H., Kirberg, J. & Rocha, B. An unusual lineage of α/β T cells that
contains autoreactive cells. J. exp. Med. 174, 1001–1008 (1991).
97. Bruno, L., Fehling, H.J. & von Boehmer, H. The α β T cell receptor can replace
the γδ receptor in the development of γδ lineage cells. Immunity 5, 343–352
98. egawa, T., Kreslavsky, T., Littman, D.R. & von Boehmer, H. Lineage diversion of T
cell receptor transgenic thymocytes revealed by lineage fate mapping. PloS one
3, e1512 (2008).
99. Takahama, Y., Shores, e.w. & Singer, A. Negative selection of precursor thymocytes
before their differentiation into CD4+CD8+ cells. Science 258, 653–656 (1992).
100. Buch, T., Rieux-Laucat, F., Forster, I. & Rajewsky, K. Failure of HY-specific thy-
mocytes to escape negative selection by receptor editing. Immunity 16, 707–718
101. Bouillet, P. et al. BH3-only Bcl-2 family member Bim is required for apoptosis of
autoreactive thymocytes. nature 415, 922–926 (2002).
102. McCaughtry, T.M., Baldwin, T.A., wilken, M.S. & Hogquist, K.A. Clonal deletion
of thymocytes can occur in the cortex with no involvement of the medulla. J. exp.
Med. 205, 2575–2584 (2008).
103. Baldwin, T.A., Sandau, M.M., Jameson, S.C. & Hogquist, K.A. The timing of TCR
α expression critically influences T cell development and selection. J. exp. Med.
202, 111–121 (2005).
104. Smith, C.A., williams, G.T., Kingston, R., Jenkinson, e.J. & Owen, J.J. Antibodies
to CD3/T-cell receptor complex induce death by apoptosis in immature T cells in
thymic cultures. nature 337, 181–184 (1989).
105. wack, A. et al. Direct visualization of thymocyte apoptosis in neglect, acute and
steady-state negative selection. Int. Immunol. 8, 1537–1548 (1996).
106. von Boehmer, H. Developmental biology of T cells in T cell-receptor transgenic
mice. annu. rev. Immunol. 8, 531–556 (1990).
107. Nemazee, D. Receptor editing in lymphocyte development and central tolerance.
nat. rev. Immunol. 6, 728–740 (2006).
108. Sprent, J. & Kishimoto, H. The thymus and negative selection. Immunol. rev. 185,
109. Swat, w., Dessing, M., von Boehmer, H. & Kisielow, P. CD69 expression during
selection and maturation of CD4+8+ thymocytes. eur. J. Immunol. 23, 739–746
110. Bonasio, R. et al. Clonal deletion of thymocytes by circulating dendritic cells hom-
ing to the thymus. nat. Immunol. 7, 1092–1100 (2006).
111. Anderson, M.S. et al. Projection of an immunological self shadow within the thymus
by the aire protein. Science 298, 1395–1401 (2002).
112. wicker, L.S. et al. Type 1 diabetes genes and pathways shared by humans and NOD
mice. J. autoimmun. 25 (suppl.), 29–33 (2005).
113. Anderson, M.S. et al. The cellular mechanism of Aire control of T cell tolerance.
Immunity 23, 227–239 (2005).
114. Kishimoto, H. & Sprent, J. A defect in central tolerance in NOD mice. nat. Immunol.
2, 1025–1031 (2001).
115. Zucchelli, S. et al. Defective central tolerance induction in NOD mice: genomics
and genetics. Immunity 22, 385–396 (2005).
116. Liston, A. et al. Impairment of organ-specific T cell negative selection by diabetes
susceptibility genes: genomic analysis by mRNA profiling. Genome Biol. 8, R12
117. Pugliese, A. et al. The insulin gene is transcribed in the human thymus and tran-
scription levels correlated with allelic variation at the InS VNTR-IddM2 suscepti-
bility locus for type 1 diabetes. nat. Genet. 15, 293–297 (1997).
118. Jaeckel, e., Lipes, M.A. & von Boehmer, H. Recessive tolerance to preproinsulin
2 reduces but does not abolish type 1 diabetes. nat. Immunol. 5, 1028–1035
119. Sakaguchi, N. et al. Altered thymic T-cell selection due to a mutation of the ZAP-70
gene causes autoimmune arthritis in mice. nature 426, 454–460 (2003).
120. Sarukhan, A., Garcia, C., Lanoue, A. & von Boehmer, H. Allelic inclusion of T
cell receptor α genes poses an autoimmune hazard due to low-level expression of
autospecific receptors. Immunity 8, 563–570 (1998).
121. Hahn, M., Nicholson, M.J., Pyrdol, J. & wucherpfennig, K.w. Unconventional
topology of self peptide-major histocompatibility complex binding by a human
autoimmune T cell receptor. nat. Immunol. 6, 490–496 (2005).
122. Garcia, K.C., Teyton, L. & wilson, I.A. Structural basis of T cell recognition. annu.
rev. Immunol. 17, 369–397 (1999).
123. Suri, A., Levisetti, M.G. & Unanue, e.R. Do the peptide-binding properties of
diabetogenic class II molecules explain autoreactivity? curr. opin. Immunol. 20,
© 2010 Nature America, Inc. All rights reserved.