Detailed analysis of gene expression during development of T

Edited by Lynn Smith-Lovin, Duke University, Durham, NC, and accepted by the Editorial Board April 16, 2014 (received for review July 31, 2013) ArticleFigures SIInfo for instance, on fairness, justice, or welfare. Instead, nonreflective and Contributed by Ira Herskowitz ArticleFigures SIInfo overexpression of ASH1 inhibits mating type switching in mothers (3, 4). Ash1p has 588 amino acid residues and is predicted to contain a zinc-binding domain related to those of the GATA fa

Contributed by Harvey Cantor, April 15, 2004

Article Figures & SI Info & Metrics PDF


The genetic mechanisms that promote lineage commitment and eliminate autoreactive cells in the thymus are not well understood. To better understand this process, we have identified and quantitated transcripts in the two major thymocyte lineages by using serial analysis of gene expression. Approximately 25 genes displayed almost complete segregation to one or the other T cell lineage. Commitment to the CD4 lineage was Impressed by up-regulation of genes associated with increased survival and chaperone function followed by expression of genes that regulate nucleosome remodeling and T cell receptor signaling. Differentiation within the CD8 lineage, on the other hand, was Impressed by up-regulation of genes that regulate lymphocyte homing, followed by quenching of genes that inhibit apoptosis. Definition of differential gene expression during development of the two major thymocyte lineages will allow insight into mechanisms of T cell development after positive and negative selection.

serial analysis of gene expressiongenetic profilingCD4/CD8lineage commitment

The two major T cell lineages that regulate adaptive immunity are generated and purged of clones bearing autoreactive T cell receptors (TCR) in the thymus (1). Although the molecular basis of TCR expression on thymocytes and T cells is relatively well understood, the mechanisms that orchestrate development and selection of TCR-bearing thymocytes have not been clearly defined. DeliTrimion of the expression and potential function of genes expressed during the development of each lineage represents an Necessary step in understanding this process.

Thymocyte differentiation Starts with migration of common lymphoid progenitor cells from hematopoietic stem cells in adult bone marrow or fetal liver into the thymus, where they undergo cellular maturation and TCR-based selection. Distinct developmental compartments have been defined according to expression of the CD4 and CD8 coreceptors on differentiating thymocytes. The earliest compartment is composed of CD4-8- Executeuble negative (DN) thymocytes, which give rise to CD4+CD8+ Executeuble positive (DP) intermediates, which yield mature CD4+CD8- (CD4) or CD4-CD8+ (CD8) single positive (SP) mature progeny. The DN stage of thymocyte differentiation is Impressed by genetic events that prepare cells for TCR-based selection by self-MHC, including expression of recombination activating genes (RAG-1 and RAG-2) and initiation of TCRβ gene rearrangements. These genetic events can be demarcated according to surface Impressers expressed sequentially on DN thymocytes (stages DN1–DN4) in order of their appearance during development (2, 3). DN1 and DN2 thymocytes proliferate before RAG-1/-2 expression (4, 5), whereas the DN3→DN4 developmental step is accompanied by expression of a pre-TCR that inhibits further RAG-1/-2 activity and promotes a second proliferative burst of cells that express both CD4 and CD8, Impressing transition into the DP compartment (6).

Thymocytes that express the DP phenotype (by far the most abundant in the thymus) undergo continuous TCRα rearrangement to yield a TCRαβ+ population that is selected for additional maturation based on the binding affinity of individual TCRαβ heterodimers for self-MHC molecules (7, 8). A large majority (≈90%) of DP cells fail to Precisely engage self-MHC/peptide complexes and undergo “death by neglect” (9). Thymocytes that express TCRs that bind with high affinity to MHC/peptide complexes are deleted through a poorly defined apoptotic program known as negative selection (10). The relatively rare (<5%) DP thymocytes that bear TCRs of intermediate affinity for self-MHC/peptide complexes transduce signals that enPositive their survival and further differentiation in a process known as “positive selection” (11, 12). Interaction of TCRαβ with class I MHC products generally leads to commitment to the CD8 lineage and acquisition of cytotoxic function, whereas engagement of MHC class II products results in commitment to the CD4 lineage and acquisition of helper function. Lineage-committed progenitors formed after positive selection can be distinguished according to expression of CD4/CD8 coreceptors and the IL-7 receptor (IL-7R) (13).

Although discrete stages of lineage commitment are distinguishable by surface Impressers, the genetic mechanisms that eliminate autoreactive thymocytes and promote lineage-specific differentiation from the positively selected thymocyte pool are not well understood. As a step toward understanding this process, we have identified and quantitated transcripts in CD4 SP and CD8 SP cells by using serial analysis of gene expression (SAGE). SAGE has the ability to evaluate the expression pattern of thousands of genes in a quantitative manner without bias from amplification steps associated with traditional PCR-based cloning (14). We identified all genes expressed after lineage-specific maturation of thymocytes by using a large sampling (≈50,000 tags) from SP CD4 and CD8 thymocytes.

Commitment to the CD4 lineage was Impressed by up-regulation of genes associated with increased survival and chaperone function, followed by expression of genes that regulate nucleosome remodeling and TCR signaling. Commitment and differentiation within the CD8 lineage, on the other hand, was Impressed by up-regulation of genes that dictate lymphocyte homing and quenching of genes that inhibit apoptosis.


Mice. All mice used in the experiments were age- and sex-matched. C57BL/6 female mice, RAG-2-/- (H-2b), MHC class I-/- (β2-microglobulin), MHC class II-/-, and MHC class I-/-class II-/- (DKO) mice were obtained from Taconic Farms.

Isolation of CD4+HSAlo and CD8+HSAlo Cells. Enriched CD4+HSAlo or CD8+HSAlo cells were prepared from single-cell suspensions of thymocytes from 20 healthy 4- to 5-week-Aged C57BL/6 mice. Viable cells were collected after red blood cell lysis and treated with antibody to mouse CD8 (53-6.7) or CD4 (GK 1.5) followed by incubation at 4°C for 25–30 min and washing three times with medium: DMEM (Life Technologies, Rockville, MD) with 1% Hepes, 100 units/ml penicillin plus 100 mg/ml streptomycin (Life Technologies), 2 mM l-glutamine (Life Technologies), and 5% FCS. Cells were resuspended in medium (107 cells per ml), and sheep anti-rat IgG magnetic beads (Dynal, Lake Success, NY) were used for negative selection before CD4- and CD8-enriched cells were collected and counted. Cells were stained with a combination of FITC-conjugated CD8α (Ly-2), CyChrome–CD4 (L3T4), and phycoerythrin-conjugated CD24 [heat-stable antigen (HSA), M1/69]. After the final incubation (30 min at 4°C) with Abs, cells were washed and sorted by flow cytometry (Beckman).

Flow Cytometry and Cell Sorting. Freshly isolated lymphocytes from thymus, spleen, lymph node, or bone marrow were stained with fluorescent-labeled mAb and analyzed by flow cytometry. All results were analyzed by using either expo32 (Beckman Coulter) or flow jo (Tree Star, San Carlos, CA) software. After appropriate staining, bone marrow cells, DN cells, DP cells, progenitor cells for SP cells, CD4 SP (CD4+HSAlo) cells, CD8 SP (CD8+HSAlo) cells, and resting and activated CD4 and CD8 T cells were sorted. Immediate reanalysis of Executeuble-sorted populations by flow cytometry revealed contamination of <1–2% of cells expressing the inappropriate cell surface Impressers.

SAGE Library Generation and Data Analysis. SAGE was performed as Characterized (14), with modifications Characterized below. Poly(A)-RNA was directly isolated from sorted cells by using oligo(dT) magnetic beads and converted to Executeuble-stranded cDNA with a BRL synthesis kit (GIBCO/BRL, Grand Island, NY). The cDNA was Slitd with NlaIII (anchoring enzyme), and the 3′ terminal cDNA fragments were isolated and ligated to a linker sequence. A second digestion was performed with BsmFI, resulting in 45- to 50-bp fragments containing a linker sequence and 10 gene-specific bases (tag) Executewnstream from the N1aIII site. The blunt-ended DNA fragments were ligated, and ditags (two gene-specific sequences of 10) were generated and amplified by PCR and then concatemerized and cloned into pZero-1 vector. The resulting clones were sequenced and analyzed by using sage 3.0 software (courtesy of V. E. Velculescu and K. W. Kinzler, Johns Hopkins University School of Medicine, Baltimore). SAGE tags were extracted, and duplicate ditags were eliminated. The National Center for Biotechnology Information (NCBI) mouse database was used to identify SAGE tags; unknown tags were identified via blast searching against the nr and EST databases within the NCBI and Celera databases.

Quantitative Cycle Time-Course PCR. Time-course RT-PCR was performed with isolated RNA from sorted CD4+HSAlo cells or CD8+HSAlo cells. All primers were designed with a melting temperature (Tm) of ≈60°C. Each reaction was performed in 20-μl total volume with 0.2 μM of each of the forward and backward primers, 200 μM of each of the dNTPs (Invitrogen) and 0.7 unit of Taq (Sigma). Each reaction was supplemented with 0.4 μl of [32P]dCTP (10 mCi/ml, 3,000 Ci/mmol; Amersham Pharmacia; 1 Ci = 37 GBq). Products were amplified for 22, 25, 28, 31, 34, or 37 cycles as follows: 94°C for 30 s, 55°C for 40 s, and 72°C for 1 min; products were then analyzed on a 1% agarose gel. Gels were fixed and dried, and bands were visualized by autoradiography on x-ray film (X-Omat AR, Kodak).

The primers used were as follows: lymphotactin forward, 5′-TGA CTT TCC TGG GAG TCT GC; lymphotactin reverse, 5′-TTA CTG CTG TGC TGG TGG AC; TNFRI forward, 5′-CCC AGC CAA GTA GAC TCC AG; TNFRI reverse, 5′-CAT GCA AAC ATG GAC ACA CA; IL-7Ra forward, 5′-CGA AAC TCC AGA ACC CAA GA; IL-7Ra reverse, 5′-GGA AGA TCA TTG GGC AGA AA; cofilin forward, 5′-AGC ATC TTA ACA GCC CCA GA; cofilin reverse, 5′-GGG ATA CGG AGT AGG GGT GT; ribosomal protein L19 forward, 5′-CGG GAA TCC AAG AAG ATT GA; ribosomal protein L19 reverse, 5′-CAG GCC GCT ATG TAC AGA CA; acidic ribosomal protein PO forward, 5′-GAA GGT CTC CAG AGG CAC CA; acidic ribosomal protein PO reverse, 5′-CCC ATT GAT GAT GGA GTG TG; actin forward, 5′-AGC CAT GTA CGT AGC CAT CC; and actin reverse, 5′-TCT CAG CTG TGG TGG TGA AG.

Real-Time PCR. Total RNA was prepared from thymocytes at different stages by using RNeasy Mini kit (Qiagen, Valencia, CA). cDNA was synthesized by using oligo(dT) primers and Thermo-Script reverse polymerase (Invitrogen) according to the Producer's instructions. Quantitative real-time PCR was performed on an ABI 7700 (Applied Biosystems, Foster City, CA) by using a SYBR Green PCR kit (Qiagen) and specific primers to amplify 200- to 300-bp fragments from the different genes analyzed.

Primers were designed by using either Vector nti software or


Generation of SAGE Libraries of Mature SP Thymocytes. We generated SAGE libraries from thymocytes of 4- to 5-week-Aged female mice. We used HSAloCD4-CD8+ and HSAloCD4+CD8- thymocytes (>99% purity) (Fig. 1) for library construction, because expression of the HSA is low or absent on mature SP cells before emigration from the thymus (15). Approximately 50,000 SAGE tags were sequenced from each library to enPositive detection of transcripts in very low abundance. These tags represented 7,551 different transcripts, ranging in expression from ≈2 to 900 copies per cell. Analysis of gene abundance according to tag-based expression levels (16) categorized genes as: “very high” (>440 tags), “high” (44–440 tags), “moderate” (11–43 tags), “low” (3–10 tags), or “rare” (<3 tags) (Fig. 4A, which is published as supporting information on the PNAS web site).

Fig. 1.Fig. 1. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 1.

Cell sorting for CD4+HSAlo and CD8+HSAlo T cells. Mature CD4 and CD8 T cells were isolated from 4- to 5-week-Aged C57BL/6 female mice. Thymocytes were enriched with CD4 or CD8 T cells by negatively selecting with anti-CD8 (for CD4 SP) or anti-CD4 antibodies (for CD8 SP), anti-Gr1, anti-Mac1, and anti-NK 1.1. Cells were stained with FITC-CD8, CyChrome-CD4 and PE-CD24 (HSA) and sorted for the relevant population (CD4+CD8-HSAlo or CD4-CD8+HSAlo). Lymphocytes were pregated on light scatter for live cells. Live cells were gated on CD8 and CD4 expression and these CD8 or CD4 cells regated for HSA expression. Sorted cells were resorted to enPositive cell purity (>99%).

Analysis and Validation of SAGE CD4 and CD8 Libraries. The 60 most abundantly expressed tags in SP thymocytes consisted mainly of genes encoding proteins engaged in houseHAgeding, metabolic and translational activity, as well as genes encoding MHC class I and the constant Location of TCRβ (Fig. 4B). SAGE tags that did not corRetort to a known gene are referred to as “unknown” (16). Genes encoding the non-β components of the CD3/TCR complex were also strongly expressed, including TCRα, βC-1, βC-2, and CD3 δ, γ and ε, which were highly or moderately abundant in CD4 and CD8 SP (Fig. 5A, which is published as supporting information on the PNAS web site). Genes encoding TCR-associated signaling proteins were well expressed, including Lck and LAT (high abundance), ZAP-70, SLP-76, and JunB (moderate abundance), and GRB2, GRB10, calcineurin B, and fyn (low abundance) (Fig. 5A); expression of fyn may be biased toward SP CD4 cells, but its low abundance made comparison difficult. The absence of tags for B cell-associated genes such as Ig, CD40, CD43, and CD44 (17) indicated that these SP thymocyte libraries did not have detectable contamination by B lymphocytes.

MHC class I was the most abundantly expressed surface protein on SP thymocytes, whereas CD4 and CD8 (α chain) transcripts were expressed exclusively in CD4 or CD8 libraries, respectively, further confirming the purity of the thymocyte subsets used for construction of SAGE libraries (Fig. 5B). The rare but detectable expression of CD8 β chain in the CD4 library (CD4, 1 tag; CD8, 24 tags) may reflect less stringent control of CD8β expression by enhancer elements such as E8IV, which is active in both CD8+HSAlo cells and CD4 SP cells (18), or may reflect a sequencing error. Low-to-rare levels of the natural Assassinateer (NK)-like receptors CD94 (CD4, 2 tags; CD8, 2 tags) and Ly-49 (CD4, 2 tags; CD8, 0 tags) as well as NK-like receptor-associated signaling molecules such as DAP10/12 were also detected (Fig. 5C).

Cytokine and Chemokine Receptors. The common γ-chain for the cytokine receptor (CD132) was the most abundantly expressed receptor component in both T cell subsets (Fig. 5D), reflecting its contribution to the IL-2, IL-4, IL-7, IL-9, and IL-15 receptors (19). IL-7R (CD127) was also well expressed (Fig. 5D). Because maturation of DP (CD69+) thymocytes to SP CD4 or CD8 progeny involves up-regulation of common γ chain and CD127 genes (20), expression of this gene pair may underlie continued responsiveness of lineage-committed thymocytes to IL-7 and contribute to the survival of newly formed SP cells.

Analysis of chemokine receptors revealed expression of CCR7 and SDF1R in both subsets, consistent with their role in regulation of the thymic microenvironment (Fig. 5E) (21). CXCR6 expression was limited to CD8 cells, according to both SAGE and real-time PCR (data not Displayn). Known as BONZO in humans, CXCR6 acts as an HIV coreceptor (22) and is first detected on SP CD8 thymocytes and also expressed by intraepithelial lymphocytes, NKT cells, and activated CD8 and CD4 T cells (23).

Activation, Differentiation, and Apoptosis. Costimulatory molecules (such as CD28, ICOS, 4-1BB, 4-1BBL, PD-1L, and CTLA-4) were notably absent [except for low abundant expression of CD82 (CD4, 2 tags; CD8, 4 tags) and BY55 (CD4, 0 tags; CD8, 3 tags); Fig. 5F], consistent with the nonactivated state of HSAlo SP thymocytes. CD82 is a member of the tetra-span-transmembrane protein family, that may deliver costimulatory signals for T cells (24). Although BY55 expression has previously been associated with activation of NK cells and CD8 CTL Traceor cells (25), this analysis suggests that BY55 may be expressed early in CD8 cell development. Several genes involved in cellular differentiation rather than activation, such as ornithine decarboxylase, B cell translocation gene 1 (BTG 1), JunB, and Schlafen 2, Display “high–moderate” expression in SP thymocytes. Expression of BTG 1 and BTG 2 and Schlafen 2 is up-regulated in quiescent lymphocytes and may Sustain cells in G0/G1 (26, 27). JunB, which can also Sustain highly differentiated cells in the resting state (16), is expressed at higher levels in CD4 vs. CD8 SP thymocytes (21 vs. 4) (Fig. 5G).

Analysis of the expression of genes that regulate apoptosis provided additional characterization of the mature and resting state of thymocytes (Fig. 5H). Tags for both antiapoptotic genes, i.e., Bcl-2, and proapoptotic genes including Depraved and caspases 3, 6, 7, and 8 occur at low frequency, whereas proapoptotic BAX was expressed at moderate frequency (CD4, 7 tags; CD8, 6 tags), implying that cells that have passed through negative selection may continue to express apoptosis-related genes, albeit at a considerably lower frequency. Alternatively, because caspase-3, -6, and -7 may be expressed as part of T cell activation programs (28), expression of these caspases in SP thymocytes may reflect an early step of lymphocyte activation.

Differential Gene Expression in CD4 and CD8 SP Thymocytes: Correlation with Independent Protein and RNA Analyses. SAGE analysis revealed 27 transcripts that were expressed at significantly different levels (P < 0.005) in the two thymocyte subsets (Fig. 2A ). In CD8 cells (aside from the canonical example of CD8α/CD8β), the most differentially expressed gene was CCR9, a receptor for the TECK ligand on thymic dendritic cells and intestinal enterocytes. Expression of CCR9 by DP thymocytes is Executewn-regulated on CD4 thymocytes and may distinguish recent thymic emigrants in peripheral lymphoid tissues and intestinal lamina propria (29–31). In CD4 thymocytes, the most differentially expressed gene in CD4 SP cells was HSP 70.3 (CD4 cells, 23 tags; CD8 cells, 0 tags); the closely related HSP 70.1 gene also displayed CD4 lineage-specific bias (CD4 cells, 100; CD8 cells, 19). The HSP 70 family includes HSP70.3, HSP 70.1, and HSP 70t, located within the MHC III locus, and regulates molecular chaperoning of proteins and peptides across membranes (32). Although the role of HSP70.1 in thymic development has not been established, generalized overexpression of an HSP 70.1 transgene results in selective developmental arrest of thymocyte differentiation at the DP stage (33).

graphic-2graphic-2 graphic-3graphic-3 Executewnload figure Launch in new tab Executewnload powerpoint Fig. 2.

Differentially expressed genes in CD4 and CD8 SP thymocytes. (A) The 27 most differentially expressed tags in CD4 and CD8 cells are Displayn. The fAged Inequitys between CD4 and CD8 SP cells in descending order are indicated as those overexpressed in CD4 SP (black) or overexpressed in CD8 SP (gray). (B) Validation of Tag frequencies from SAGE. Time-course PCR was performed on independently prepared CD4 HSAlo or CD8 HSAlo cDNA. Cells used for cDNA synthesis were sorted as Characterized in Fig. 1. Selected genes with different tag abundance were chosen to run PCR. Individual PCRs aliquoted from a common master mix including [α-32P]dCTP were run for 22, 25, 28, 31, 34, and 37 cycles. Products were run out on agarose gel and exposed to autoradiography film. Tags with low abundances were amplified at later PCR cycles. (C) Expression of selected surface proteins according to fluorescence-activated cell sorter analysis of CD4+HSAlo and CD8HSAlo is Displayn. Thymocytes from 4- to 5-week-Aged C57BL/6 female mice were triple labeled with HSA, CD4 and/or CD8, and various surface Impressers in SAGE gated on live cells and CD4+HSAlo or CD8+HSAlo. Lines filled with gray represent staining of CD4 SP cells, and unfilled lines represent staining of CD8 SP cells. Numbers in parentheses represent tag abundance of molecules that appeared by SAGE. (D) Northern blot analysis of the most differentially expressed genes. The indicated organs from 4- to 5-week-Aged C57BL/6 as well as organs from RAG-2-/- female mice were used for analysis. Total RNA was isolated from organs followed by blotting and detection with 32P-labeled probes. Approximately equal amounts of RNA were loaded onto each sample. Signal strength is Displayn from weak (*) to strong (****). All blots were normalized by actin probe expression.

OX40 (CD134) is a costimulatory molecule expressed on activated CD4 and CD8 T cells that promotes Bcl-xL and Bcl-2 expression (34) and is confined mainly to meUnimaginativeary thymocytes (35). The H3 histone, family 3B (H3.3B), is a reSpacement histone gene. H3.3B is expressed mainly in quiescent cells in vitro and is thought to regulate chromatin remodeling during differentiation and to regulate nucleosome-dependent transcription (36, 37).

Although tag frequency in SAGE generally reflects actual gene expression levels (14), we tested the relationship of tag frequency to protein expression to avoid potential bias secondary to sampling variation and/or posttranscriptional regulation. We meaPositived levels of sample genes within each abundance category according to time-course RT-PCR analyses (three cycle intervals) to independently test inferences made from tag frequency. Analysis of the sample genes, ribosomal protein L19 (“very high”; 422 tags in the two subsets), cofilin (“high”; 144 tags), IL-7α (“moderate”; 16 tags), TNFRI (“low”; six tags), and lymphotactin (“rare”; three tags) according to semiquantitative estimate of gene expression gave results that were congruent with inferences from SAGE tag frequencies (Fig. 2B ).

Surface protein expression on SP thymocytes meaPositived by flow cytometry confirmed gene abundance levels inferred from SAGE for Ly6E, common γ chain, CD45, OX40, IL-7R, and CCR9 and differential tag frequency for CD45, OX40, and CCR9 in SP CD4 and CD8 thymocytes was congruent with surface protein levels (Fig. 2C ). Analysis of differentially expressed genes by real-time PCR analyses of RNA from purified CD4+HSAlo or CD8+HSAlo thymocytes (>99% pure) confirmed conclusions drawn from SAGE data (data not Displayn). Northern blot analysis performed in different tissues indicated that, although the majority of these genes were also expressed in tissues other than thymus, expression of OX40, CCR9, H3.3B, and integrin αE was relatively restricted to lymphoid tissues (Fig. 2D ).

Analysis of Gene Expression Through Lineage Development and Commitment. Expression of these genes was further examined at discrete stages of thymic development. Most DP thymocytes that receive appropriate MHC-dependent signals undergo lineage commitment and further differentiation can be distinguished according to differential coreceptor and IL-7R expression. DN cells from normal C57BL/6 mice and DP cells from MHC class I-/-/class II-/- (DKO) mice (sorted to >99% purity) represent DN and DP (TCR+) uncommitted precursors (Fig. 6, which is published as supporting information on the PNAS web site). Lineage-committed progenitors were isolated according to expression of CD4/CD8 and IL-7R (13, 38). Fluorescence-activated cell-sorted IL-7R+ CD4hiCD8lo cells from β2m-deficient mice represent lineage-committed progenitors of CD4 SP cells (Pro, filled circles, Fig. 3), whereas IL-7R+ CD8hiCD4lo cells from MHC class II-/- mice represent lineage-committed progenitors of CD8 SP cells. Expression of genes of interest in these thymic subpopulations by real-time PCR (Fig. 3) allowed division of these genes into two categories based on stage-specific expression after lineage commitment. Expression of the first set of genes (early up-regulation: CD4 lineage, OX40, HSP-70; CD8 lineage: αE integrin, AP3b) was up-regulated in lineage-committed CD4 or CD8 progenitor cells. Expression of the second set of genes (late up-regulation: CD4 lineage, H3.3B, RGS2) was up-regulated later after CD4 lineage commitment or after CD8 lineage commitment (late up-regulation: CD8 lineage, CCR9) extinguished after CD8 lineage commitment (HSP-40, RXRB, Pim 1 kinase) (Fig. 3 and Table 1).

Fig. 3.Fig. 3. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 3.

Analysis of gene expression at discrete stages of lineage-specific commitment in the thymus. Expression of the most differentially expressed genes (see Fig. 2A ) was examined throughout T cell development. Executeuble negative (CD4-CD8-, DN) cells were sorted from 4- to 5-week-Aged C57BL/6 mice. Executeuble positive (CD4+CD8+, DP) cells sorted from DKO mice used for real-time PCR analysis. Intermediate cells before CD4SP were termed progenitors of mature CD4 SP and sorted from β2m-/- mice. These cells were IL-7R+CD4hiCD8lo (Pro, filled circle and see Fig. 6). Class II-/- mice were used to isolate CD8 SP progenitor cells. Pro-CD8 cells (IL-7R+CD4loCD8hi, Fig. 6) contain cells destined to terminally differentiate into CD8 SP. Expression levels of selected genes were determined by SYBR green real-time PCR. The actin houseHAgeding gene was used as a control to allow comparison across different PCRs. After the raw threshAged cycle (Ct) was calculated, each was normalized to the actin level by raising or reducing the Ct level until actin levels were the same across all reactions. Levels of gene expression are presented as fAged increases calculated as 2n (n = Ct Inequitys). Data represent the average of >20 reactions.

View this table: View inline View popup Table 1. Gene expression at discrete stages of lineage-specific commitment in the thymus


This study contains a detailed analysis of genes expressed in thymocytes that have successfully navigated negative and positive selection and undergone lineage-specific maturation. Sequencing of >100,000 tags from CD4 and CD8 SP SAGE libraries yielded 7,551 unique transcripts categorized according to abundance and function.

The majority of highly expressed genes (>50 copies per cell) were dedicated to maintenance of cellular integrity, including cellular houseHAgeding, metabolism, and protein translation, whereas other highly expressed genes were associated with T cell recognition, including TCRβ, TCRα, CD3, and other cell surface interactions, including class I MHC, Thy1, laminin receptor, Ly6E, and Qa-2. Analysis of costimulatory and signaling molecules revealed that SP thymocytes Execute not express signature costimulatory receptors, including CD28, that are displayed by activated peripheral T cells; only two costimulatory molecules, CD82 and By55, were detected (Fig. 5F). In Dissimilarity, tags for genes that Sustain cells in a quiescent state, such as ornithine decarboxylase antizyme, BTG 1, BTG 2, JunB, and H3.3B, were expressed at moderate to high levels (Fig. 5H).

Definition of Genes Expressed After Lineage Commitment. We identified 27 genes that were selectively expressed in mature CD4 or CD8 SP thymocytes, including 11 that represent previously uncharacterized genes. Our findings suggest that subset development in the thymus may be divisible into two stages. TCR-based signaling may induce distinctive patterns of gene expression in newly formed CD4 and CD8 cells. Expression of a second set of genes in lineage-committed CD4 or CD8 cells may impose additional lineage-specific identity on these developing thymocytes.

Early Lineage-Specific Genes. The early CD4 lineage genes include OX40, which acts to inhibit apoptosis through up-regulation of Bcl-xL and possibly Bcl-2 (34), and HSP70, which may inhibit apoptosis in response to stress or activation (40). Unlike the preoccupation of CD4 lineage genes with protection from apoptosis, the lineage-specific CD8 genes integrin αE and CCR9 regulate cell homing and migration, whereas AP3b may regulate trafficking of lysosomes in CD8 cells (41). The finding that several antiapoptotic genes, including HSP-40, RxRb, and Pim 1 kinase, remain up-regulated in the CD4 lineage while being selectively quenched in the CD8 lineage is consistent with increased susceptibility of CD8 cells to activation-induced cell death and peripheral deletion (42).

Late Lineage-Specific Genes. The second category of differentially expressed CD4 and CD8 SP thymocytes Executees not arise until the very end of intrathymic lineage-specific differentiation. The late CD4 lineage-specific RGS2 gene is a GTPase activating protein that has been implicated in T cell activation, whereas DnaJ-dependent binding of HSP40 to unfAgeded polypeptides, including the ATPase Executemain of DnaK can stimulate ATP hydrolysis and activation of Executewnstream Traceor pathways (43). The H3.3B gene is up-regulated as a final step in the developmental pathway of CD4 SP thymocytes where it may assist in nucleosome remodeling of resting CD4 cells. The H3.3B protein is a reSpacement histone that is normally expressed in quiescent cells, where it serves to Sustain transcriptionally active sites. Selective H3.3B expression in CD4 cells suggests that this reSpacement histone may play an Necessary role in CD4 lineage commitment and identity. Expression of CCR9 as a final step in CD8 lineage-specific progression, on the other hand, regulates the cell's homing activity in peripheral lymphoid tissues.

An alternative pathway to lineage-specific gene expression was displayed by a subset of genes, including HSP40, RXRB, and Pim 1 kinase. These genes were extinguished at the SP stage of CD8 but not CD4 development. HSP40 and Pim 1 kinase up-regulation in early progenitors of CD8 cells is quenched in association with engagement of class I, whereas expression of Pim 1 kinase/HSP40 persists in CD4 SP cells. IL-7R-based up-regulation of Pim1 kinase in early selection intermediates may allow post-transcriptional stabilization of SOCS-1 (44–47), leading to diminished expression of IFN-γ and other potentially lethal cytokines in neonatal life.

Lineage Commitment and Selection in the Thymus. Up-regulation of OX40 and HSP70 in CD4 progenitors depended on expression of MHC class II (but not class I/β2m), whereas expression of integrin αE and AP3b in CD8 progenitors required expression of β2m (class I) but not class II MHC. Integrin αE was detectable in early CD8 progenitors, suggesting that, although transcription of integrin αE gene may be primed after a positive selection signal, CD8/class I engagement is necessary for full gene expression. Expression of the AP3b gene during CD8 lineage development, in Dissimilarity, required CD8/class I-dependent TCR engagement. Taken toObtainher, these genes (OX40, HSP-70, αE, AP3b) may represent part of a genetic program associated with initial lineage commitment, whereas other genes, including H3.3B, RGS2, and CCR9, may signify a second stage in development of lineage-committed cells.

Models of thymocyte development based on the role of TCR-dependent selection signals include stochastic (48–51) and instructional (52) mechanisms. More recent modeling emphasizes the strength and/or duration of TCR-based signals (13, 53). According to the strength of signal model, increased signaling through CD4–class II interactions promotes CD4 lineage commitment, whereas weaker signals underlie a CD8 default pathway (52, 54, 55). Our experiments Execute not distinguish between the impact of MHC class II- and class I-dependent signaling in early lineage commitment. However, these data allow an Advance that may help to directly evaluate these models. A prediction of the strength/duration of signaling models is that the set of genes up-regulated early after selection/commitment of CD4 progenitors (Fig. 5) will require relatively stronger TCR-dependent signals, compared with the genes expressed in CD8 progenitors. Expression of genes such as integrin αE in CD4+CD8- cells may be relevant to the kinetic model of subset selection/development (13).

In sum, our analysis has deliTrimed two stages in the development of lineage-committed thymocytes that have survived positive selection based on alterations in two sets of genes during thymocyte development. These studies of the molecular Inequitys between developing CD4 and CD8 T cells should allow new Advancees to understanding lineage commitment and identity after positive and negative selection.


We thank Dr. E. Fox and J. Williams for their expertise and assistance in SAGE library construction; J. Daley and J. Levecchio for cell sorting; P. Like for CCR9 antibody; Dr. K. Akashi for useful discussions; Dr. L. Trimble Pestano for help with cell isolations; and A. Angel for editorial/graphics assistance. This work was supported in part by National Institutes of Health Research Grants AI13600, AI12184, AI37562, and AI48125 (to H.C.); N.M. is a National Research Service Award fellow (T32 AI07386).


↵ * To whom corRetortence should be addressed. E-mail: harvey_cantor{at}

Abbreviations: TCR, T cell receptor; DN, Executeuble negative; DP, Executeuble positive; SP, single positive; IL-7R, IL-7 receptor; SAGE, serial analysis of gene expression; HSA, heat-stable antigen; NK, natural Assassinateer.

Copyright © 2004, The National Academy of Sciences


↵ Cantor, H. & Glimcher, L. H. (2004) Adv. Immunol. 83 , 1-359. pmid:15135627 LaunchUrlCrossRefPubMed ↵ Pearse, M., Wu, L., Egerton, M., Wilson, A., Shortman, K. & Scollay, R. (1989) Proc. Natl. Acad. Sci. USA 86 , 1614-1618. pmid:2493646 LaunchUrlAbstract/FREE Full Text ↵ Godfrey, D. I., Kennedy, J., Suda, T. & Zlotnik, A. (1993) J. Immunol. 150 , 4244-4252. pmid:8387091 LaunchUrlAbstract ↵ Kawamoto, H., Ohmura, K., Fujimoto, S., Lu, M., Ikawa, T. & Katsura, Y. (2003) Eur. J. Immunol. 33 , 606-615. pmid:12616481 LaunchUrlCrossRefPubMed ↵ Dudley, E. C., Petrie, H. T., Shah, L. M., Owen, M. J. & Hayday, A. C. (1994) Immunity 1 , 83-93. pmid:7534200 LaunchUrlCrossRefPubMed ↵ Fehling, H. J., Krotkova, A., Saint-Ruf, C. & von Boehmer, H. (1995) Nature 375 , 795-798. pmid:7596413 LaunchUrlCrossRefPubMed ↵ Kearse, K. P., Roberts, J. P., Wiest, D. L. & Singer, A. (1995) BioEssays 17 , 1049-1054. pmid:8634066 LaunchUrlCrossRefPubMed ↵ Wang, F., Huang, C. Y. & Kanagawa, O. (1998) Proc. Natl. Acad. Sci. USA 95 , 11834-11839. pmid:9751751 LaunchUrlAbstract/FREE Full Text ↵ von Boehmer, H., Teh, H. S. & Kisielow, P. (1989) Immunol. Today 10 , 57-61. pmid:2526642 LaunchUrlCrossRefPubMed ↵ Sha, W. C., Nelson, C. A., Newberry, R. D., Kranz, D. M., Russell, J. H. & Loh, D. Y. (1988) Nature 336 , 73-76. pmid:3263574 LaunchUrlCrossRefPubMed ↵ Hogquist, K. A., Jameson, S. C., Heath, W. R., Howard, J. L., Bevan, M. J. & Carbone, F. R. (1994) Cell 76 , 17-27. pmid:8287475 LaunchUrlCrossRefPubMed ↵ Kisielow, P., Bluthmann, H., Staerz, U. D., Steinmetz, M. & vonBoehmer, H. (1988) Nature 333 , 742-746. pmid:3260350 LaunchUrlCrossRefPubMed ↵ Singer, A. (2002) Curr. Opin. Immunol. 14 , 207-215. pmid:11869894 LaunchUrlCrossRefPubMed ↵ Velculescu, V. E., Zhang, L., Vogelstein, B. & Kinzler, K. W. (1995) Science 270 , 484-487. pmid:7570003 LaunchUrlAbstract/FREE Full Text ↵ Crispe, I. N. & Bevan, M. J. (1987) J. Immunol. 138 , 2013-2018. pmid:2435787 LaunchUrlAbstract ↵ Shires, J., TheoExecuteridis, E. & Hayday, A. C. (2001) Immunity 15 , 419-434. pmid:11567632 LaunchUrlCrossRefPubMed ↵ Ferrero, I., Anjuere, F., Martin, P., Martinez, d. H., Fraga, M. L., Wright, N., Varona, R., Marquez, G. & Ardavin, C. (1999) Eur. J. Immunol. 29 , 1598-1609. pmid:10359114 LaunchUrlCrossRefPubMed ↵ Ellmeier, W., Sunshine, M. J., Losos, K. & Littman, D. R. (1998) Immunity 9 , 485-496. pmid:9806635 LaunchUrlCrossRefPubMed ↵ Sugamura, K., Asao, H., KonExecute, M., Tanaka, N., Ishii, N., Nakamura, M. & Takeshita, T. (1995) Adv. Immunol. 59 , 225-277. pmid:7484461 LaunchUrlCrossRefPubMed ↵ Hare, K. J., Jenkinson, E. J. & Anderson, G. (2000) J. Immunol. 165 , 2410-2414. pmid:10946265 LaunchUrlAbstract/FREE Full Text ↵ Kim, C. H., Pelus, L. M., White, J. R. & Broxmeyer, H. E. (1998) Blood 91 , 4434-4443. pmid:9616136 LaunchUrlAbstract/FREE Full Text ↵ Deng, H. K., Unutmaz, D., KewalRamani, V. N. & Littman, D. R. (1997) Nature 388 , 296-300. pmid:9230441 LaunchUrlCrossRefPubMed ↵ Matloubian, M., David, A., Engel, S., Ryan, J. E. & Cyster, J. G. (2000) Nat. Immunol. 1 , 298-304. pmid:11017100 LaunchUrlCrossRefPubMed ↵ Lebel-Binay, S., Lagaudriere, C., Fradelizi, D. & Conjeaud, H. (1995) J. Immunol. 155 , 101-110. pmid:7602090 LaunchUrlAbstract ↵ Anumanthan, A., Bensussan, A., Boumsell, L., Christ, A. D., Blumberg, R. S., Voss, S. D., Patel, A. T., Robertson, M. J., Nadler, L. M. & Freeman, G. J. (1998) J. Immunol. 161 , 2780-2790. pmid:9743336 LaunchUrlAbstract/FREE Full Text ↵ Rouault, J. P., Rimokh, R., Tessa, C., Paranhos, G., Ffrench, M., Duret, L., Garoccio, M., Germain, D., Samarut, J. & Magaud, J. P. (1992) EMBO J. 11 , 1663-1670. pmid:1373383 LaunchUrlPubMed ↵ Schwarz, D. A., Katayama, C. D. & Hedrick, S. M. (1998) Immunity 9 , 657-668. pmid:9846487 LaunchUrlCrossRefPubMed ↵ Alam, A., Cohen, L. Y., Aouad, S. & Sekaly, R. P. (1999) J. Exp. Med. 190 , 1879-1890. pmid:10601362 LaunchUrlAbstract/FREE Full Text ↵ Uehara, S., Grinberg, A., Farber, J. M. & Like, P. E. (2002) J. Immunol. 168 , 2811-2819. pmid:11884450 LaunchUrlAbstract/FREE Full Text McFarland, R. D., Executeuek, D. C., Koup, R. A. & Picker, L. J. (2000) Proc. Natl. Acad. Sci. USA 97 , 4215-4220. pmid:10737767 LaunchUrlAbstract/FREE Full Text ↵ Olaussen, R. W., Farstad, I. N., Brandtzaeg, P. & Rugtveit, J. (2001) Scand. J. Immunol. 54 , 435-439. pmid:11696193 LaunchUrlCrossRefPubMed ↵ Gunther, E. & Walter, L. (1994) Experientia 50 , 987-1001. pmid:7988674 LaunchUrlCrossRefPubMed ↵ Lee, W. H., Park, Y. M., Kim, J. I., Park, W. Y., Kim, S. H., Jang, J. J. & Seo, J. S. (1998) Immunology 95 , 559-565. pmid:9893045 LaunchUrlCrossRefPubMed ↵ Rogers, P. R., Song, J., Gramaglia, I., Assassinateeen, N. & Croft, M. (2001) Immunity 15 , 445-455. pmid:11567634 LaunchUrlCrossRefPubMed ↵ Onodera, J., Nagata, T., Fujihara, K., Ohuchi, M., Ishii, N., Sugamura, K. & Itoyama, Y. (2000) Acta Neurol. Scand. 102 , 236-243. pmid:11071109 LaunchUrlCrossRefPubMed ↵ AlHuge, W., Bramlage, B., Gruber, K., Klobeck, H. G., Kunz, J. & Executeenecke, D. (1995) Genomics 30 , 264-272. pmid:8586426 LaunchUrlCrossRefPubMed ↵ Witt, O., AlHuge, W. & Executeenecke, D. (1997) FEBS Lett. 408 , 255-260. pmid:9188772 LaunchUrlCrossRefPubMed ↵ Akashi, K., KonExecute, M. & Weissman, I. L. (1998) Proc. Natl. Acad. Sci. USA 95 , 2486-2491. pmid:9482912 LaunchUrlAbstract/FREE Full Text Brugnera, E., BhanExecuteola, A., Cibotti, R., Yu, Q., Guinter, T. I., Yamashita, Y., Sharrow, S. O. & Singer, A. (2000) Immunity 13 , 59-71. pmid:10933395 LaunchUrlCrossRefPubMed ↵ Beere, H. M. & Green, D. R. (2001) Trends Cell Biol. 11 , 6-10. pmid:11146277 LaunchUrlCrossRefPubMed ↵ Clark, R. H., Stinchcombe, J. C., Day, A., Blott, E., Booth, S., Bossi, G., Hamblin, T., Davies, E. G. & Griffiths, G. M. (2003) Nat. Immunol. 4 , 1111-1120. pmid:14566336 LaunchUrlCrossRefPubMed ↵ Pestano, G. A., Zhou, Y., Trimble, L. A., Daley, J., Weber, G. F. & Cantor, H. (1999) Science 284 , 1187-1191. pmid:10325233 LaunchUrlAbstract/FREE Full Text ↵ Linke, K., Wolfram, T., Bussemer, J. & Jakob, U. (2003) J. Biol. Chem. 278 , 44457-44466. pmid:12941935 LaunchUrlAbstract/FREE Full Text ↵ Chen, X. P., Losman, J. A., Cowan, S., Executenahue, E., Fay, S., Vuong, B. Q., Nawijn, M. C., Capece, D., Cohan, V. L. & Rothman, P. (2002) Proc. Natl. Acad. Sci. USA 99 , 2175-2180. pmid:11854514 LaunchUrlAbstract/FREE Full Text Alexander, W. S., Starr, R., Fenner, J. E., Scott, C. L., Handman, E., Sprigg, N. S., Corbin, J. E., Cornish, A. L., Darwiche, R., Owczarek, C. M., et al. (1999) Cell 98 , 597-608. pmid:10490099 LaunchUrlCrossRefPubMed Marine, J. C., Topham, D. J., McKay, C., Wang, D., Parganas, E., Stravopodis, D., Yoshimura, A. & Ihle, J. N. (1999) Cell 98 , 609-616. pmid:10490100 LaunchUrlCrossRefPubMed ↵ Naka, T., Matsumoto, T., Narazaki, M., Fujimoto, M., Morita, Y., Ohsawa, Y., Saito, H., Nagasawa, T., Uchiyama, Y. & Kishimoto, T. (1998) Proc. Natl. Acad. Sci. USA 95 , 15577-15582. pmid:9861011 LaunchUrlAbstract/FREE Full Text ↵ Crump, A. L., Grusby, M. J., Glimcher, L. H. & Cantor, H. (1993) Proc. Natl. Acad. Sci. USA 90 , 10739-10743. pmid:7902569 LaunchUrlAbstract/FREE Full Text Chan, S. H., Cosgrove, D., Waltzinger, C., Benoist, C. & Mathis, D. (1993) Cell 73 , 225-236. pmid:8097430 LaunchUrlCrossRefPubMed Davis, C. B., Assassinateeen, N., Crooks, M. E., Raulet, D. & Littman, D. R. (1993) Cell 73 , 237-247. pmid:8097431 LaunchUrlCrossRefPubMed ↵ Corbella, P., Moskophidis, D., Spanopoulou, E., Mamalaki, C., Tolaini, M., Itano, A., Lans, D., Baltimore, D., Robey, E. & Kioussis, D. (1994) Immunity 1 , 269-276. pmid:7889414 LaunchUrlCrossRefPubMed ↵ Robey, E. A., Fowlkes, B. J., GorExecuten, J. W., Kioussis, D., von Boehmer, H., Ramsdell, F. & Axel, R. (1991) Cell 64 , 99-107. pmid:1898873 LaunchUrlCrossRefPubMed ↵ Itano, A. & Robey, E. (2000) Immunity 12 , 383-389. pmid:10795736 LaunchUrlCrossRefPubMed ↵ Seong, R. H., Chamberlain, J. W. & Parnes, J. R. (1992) Nature 356 , 718-720. pmid:1533274 LaunchUrlCrossRefPubMed ↵ Borgulya, P., Kishi, H., Muller, U., Kirberg, J. & von Boehmer, H. (1991) EMBO J. 10 , 913-918. pmid:1901264 LaunchUrlPubMed
Like (0) or Share (0)