Skip to main content
PMC home page
Journal of Leukocyte Biology logoLink to Journal of Leukocyte Biology
. 2008 Jun 17;84(3):842–851. doi: 10.1189/jlb.0208087

Identification of a new transmembrane adaptor protein that constitutively binds Grb2 in B cells

Yan Liu 1, Weiguo Zhang 1,1
PMCID: PMC2516900  PMID: 18559951

Abstract

Transmembrane adaptor proteins couple antigen receptor engagement to downstream signaling cascades in lymphocytes. One example of these proteins is the linker for activation of T cells (LAT), which plays an indispensable role in T cell activation and development. Here, we report identification of a new transmembrane adaptor molecule, namely growth factor receptor-bound protein 2 (Grb2)-binding adaptor protein, transmembrane (GAPT), which is expressed in B cells and myeloid cells. Similar to LAT, GAPT has an extracellular domain, a transmembrane domain, and a cytoplasmic tail with multiple Grb2-binding motifs. In contrast to other transmembrane adaptor proteins, GAPT is not phosphorylated upon BCR ligation but associates with Grb2 constitutively through its proline-rich region. Targeted disruption of the gapt gene in mice affects neither B cell development nor a nitrophenylacetyl-specific antibody response. However, in the absence of GAPT, B cell proliferation after BCR cross-linking is enhanced. In aged GAPT−/− mice, the number of marginal zone (MZ) B cells is increased, and other B cell subsets are normal. The serum concentrations of IgM, IgG2b, and IgG3 are also elevated in these mice. These data indicate that GAPT might play an important role in control of B cell activation and proper maintenance of MZ B cells.

Keywords: marginal zone B cells, BCR signaling, B cell proliferation

INTRODUCTION

Transmembrane adaptor proteins play important roles in lymphocyte development and activation by coupling receptor engagement to activation of downstream pathways [1,2,3,4]. One of these proteins is the linker for activation of T cells (LAT). Upon TCR ligation, LAT is phosphorylated and assembles a complex of signaling proteins containing growth factor receptor-bound protein 2 (Grb2), son of sevenless (Sos), Grb2-related adaptor downstream of Shc (Gads), Src homology 2 (SH2) domain-containing leukocyte-specific phosphoprotein of 76 kDa, phospholipase C (PLC)-γ1, and other proteins. TCR-mediated Ca2+ mobilization, Ras-MAPK activation, and IL-2 production are defective in LAT-deficient Jurkat cells [5, 6]. Thymocyte development is arrested at the double-negative 3 (DN3) stage in LAT−/− mice [7]. These data indicate that LAT is essential in T cell activation and development [6, 7].

As a result of the pivotal role of LAT in TCR signaling, much work has been done to identify LAT-like adaptor proteins in other cell types. Several LAT-like molecules, such as linker for activation of X cells (LAX), linker for activation of B cells [LAB; also known as non-T cell activation linker (NTAL)], and Lck-interacting membrane protein (LIME), were subsequently discovered [8,9,10,11,12]. Among these adaptor proteins, LAX is expressed in T and B cells. Upon stimulation via the TCR or BCR, LAX is tyrosine-phosphorylated and associates with Grb2, Gads, and p85. In contrast to the positive role of LAT, LAX functions to negatively regulate lymphocyte signaling. TCR-mediated p38 MAPK activation and Akt are enhanced in LAX−/− cells [8, 13]. How LAX regulates TCR-mediated signaling is still not clear. LIME is an adaptor protein that recruits a different set of signaling proteins from LAT or LAX. Instead of binding cytosolic adaptors, it interacts with the Src family tyrosine kinase Lck and its negative regulator, Csk. Ectopic expression of LIME in Jurkat T cells enhances Lck phosphorylation and TCR-mediated signaling [11, 12]. LIME is also expressed in B cells. Reduced expression of LIME by small interfering RNA affects BCR-mediated MAPK activation, calcium flux, and other signaling events [14]. Among the aforementioned three adaptor proteins, LAB is most similar to LAT. Like LAT, it has a short, extracellular domain, a transmembrane domain, and a cytoplasmic tail, which contains multiple Grb2-binding sites and is palmitoylated. It is expressed highly in B cells, NK cells, and mast cells (MCs). Upon engagement of the BCR or FcRs, LAB is tyrosine-phosphorylated and interacts with Grb2. Data from the studies with LAB-deficient DT40 chicken B cells indicate that the Grb2–LAB complex can positively regulate BCR-mediated calcium signaling. In addition, LAB is capable of rescuing TCR-mediated signaling in LAT-deficient Jurkat cells and thymocyte development in LAT−/− mice [9, 10, 15]. However, deletion of LAB in mice has no substantial effect on B cell development or BCR-mediated signaling [16, 17]. Studies of LAB in other cell types suggest that LAB is a negative regulator of cellular activation. FcεRI-mediated Erk activation, calcium mobilization, degranulation, and cytokine production are enhanced in LAB−/− MCs [16, 18]. Recent data also show that LAB is up-regulated in T cells upon engagement of the TCR and negatively regulates TCR-mediated MAPK activation and calcium mobilization [19]. Together, studies about these transmembrane adaptor proteins indicate that in addition to LAT, other transmembrane adaptors are important in regulation of lymphocyte activation.

In this study, we reported cloning of a gene that encodes a new transmembrane adaptor protein, Grb2-binding adaptor protein, transmembrane (GAPT). We examined expression of GAPT in different tissues and cell types, characterized this protein biochemically, and examined its association with Grb2. We also generated GAPT-deficient mice to investigate GAPT function in lymphocyte development, activation, and immune responses.

MATERIALS AND METHODS

Cloning of GAPT

Human GAPT (hGAPT) was identified by searching the National Center for Biotechnology Information (NCBI) database for novel proteins containing multiple YXN motifs, as described previously [8]. Mouse GAPT (mGAPT) was identified by BLAST searching a hGAPT sequence. The GAPT GenBank accession numbers are AK090960 (for human) and AK036534 (for mouse).

hGAPT was amplified from BJAB cells by RT-PCR and was cloned into a retroviral vector for subsequent experiments. RT-PCR amplification of GAPT was used to detect GAPT expression in different human tissues and cell lines. The G3PDH transcript was also amplified to normalize the amounts of cDNAs used in each reaction. cDNAs from different human tissues were purchased from Clontech (Palo Alto, CA, USA). For different human cell lines, RNAs were prepared and reverse-transcribed.

Antibodies, Western blotting, and FACS analysis

A GST fusion protein containing the cytoplasmic domain of mGAPT (residues 43–157) was used to immunize rabbits to generate polyclonal GAPT antisera. Other antibodies used in this study include: Myc (9E10), Lyn, and Grb2 (Santa Cruz Biotechnology, Santa Cruz, CA, USA); PLC-γ2 and 4G10 (Upstate Biotechnology, Lake Placid, NY, USA); heat shock protein (HSP)70 (BD Transduction Laboratories, Franklin Lakes, NJ, USA); and phospho-Erk (pErk), Erk, pJNK, JNK, pp38, and p38 (Cell Signaling Technology, Beverly, MA, USA). For Western blotting, samples were separated by SDS-PAGE and transferred onto nitrocellulose membranes. After incubation with primary antibodies, membranes were probed with goat anti-mouse or anti-rabbit Ig conjugated with Alexa Fluor 680 (Molecular Probes, Eugene, OR, USA) or IRDye 800 (Rockland Immunochemicals, Gilbertsville, PA, USA). Membranes were then visualized and quantified with an infrared fluorescence imaging system (Odyssey system, LI-COR Biosciences, Lincoln, NE, USA).

For FACS analysis, single-cell suspensions from spleen, thymus, and bone marrow (BM) were prepared after removal of erythrocytes with ACK buffer. The following FITC-, PE-, or allophycocyanin-conjugated antibodies were used in this study: CD4, CD8, TCRβ, Thy1.2, B220, IgM, CD43, CD23, CD21/35, IgD, DX5, membrane-activated complex 1 (Mac-1), Gr-1, CD11c, and GL-7 (eBioscience, San Diego, CA, USA). The FACS data were collected on a FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA, USA) and analyzed by the FlowJo software.

Cell stimulation, lysis, and fractionation

Human BJAB cells were cultured in RPMI 1640 with 10% FBS. Cells were resuspended in RPMI 1640 at 2 × 107 cells/ml before stimulation with anti-human IgM F(ab′)2 (Jackson ImmunoResearch, West Grove, PA, USA) and were lysed in 1% Brij lysis buffer. Primary mouse B cells were stimulated with anti-mouse IgM F(ab′)2 (ICN/Cappel Laboratories, Oxford, PA, USA). Isolation of the membrane and cytosolic fractions by Dounce homogenization and preparation of lipid raft fractions by sucrose gradient were performed as described previously [20].

Retroviral transduction

Retroviral vectors expressing hGAPT or hGAPT with three conserved prolines mutated into alanines (3PA) with a C-terminal Myc epitope tag were used to transfect human embryo kidney 293T cells for virus packaging. Viral supernatants were collected and used to transduce human BJAB cells by spin inoculation in the presence of polybrene (8 μg/ml). Forty-eight hours after transduction, cells were selected by blasticidin (10 μg/ml) to establish stable cell lines expressing GAPT.

Cell purification

For purification of CD4+, CD8+, and B220+ splenocytes, single-cell suspensions were incubated with biotin-conjugated CD4, CD8, and B220 antibodies (BD Biosciences, San Diego, CA, USA). Streptavidin-conjugated microbeads were then used to positively enrich these cell subsets by AutoMACS (Miltenyi Biotec, Auburn, CA, USA). Purified cells were subjected to FACS analysis to monitor purity before being used in subsequent experiments. For isolation of dendritic cells (DCs), spleens were digested with Liberase CI (Roche, Nutley, NJ, USA) for 30 min before preparation of single-cell suspensions and staining. BM-derived MCs (BMMCs) were generated by culturing total BM cells in the presence of IL-3 for 3 weeks [16]. BM-derived DCs (BMDCs) were generated by culturing total BM in GM-CSF containing medium for 7–9 days [21].

Generation of GAPT−/− mice

The targeting construct for disruption of the gapt gene is shown (see Fig. 3A). After successful, homologous recombination, embryonic stem cells carrying the floxed gapt exon were identified by PCR and confirmed by Southern blot analysis before being injected into blastocysts of C57BL/6 mice. The germ-line transmission of the floxed gapt gene was confirmed again by Southern blotting (see Fig. 3B). To generate GAPT−/− mice, GAPTf/+ mice were crossed with β-actin-Cre mice to delete the gapt gene. Genotyping of these mice was performed by PCR using the following primers: 5′-GTG ATC CAC CAA GGG TAA AG-3′ and 5′-TTA GCC CCT CAG CAC AGG A-3′. GAPT−/− mice were born at a normal Mendelian frequency and sex ratio. All experiments were performed in accordance with protocols approved by Duke University Medical Center Animal Care and Use Committees (Durham, NC, USA) and National Institutes of Health guidelines (Bethesda, MD, USA).

Fig. 1.

Fig. 1.

Open in a new tab

Comparison of GAPT with LAT family adaptor proteins. (A) Comparison of conserved tyrosine residues between GAPT and other LAT family members. Similar to LAT, GAPT has a short, extracellular domain, a transmembrane (TM) domain, and multiple cytoplasmic tyrosine residues. Among these tyrosine residues, four of them are within a Grb2-binding motif (YxN; gray). (B) Sequence alignment of hGAPT and mGAPT, which share 58% homology of amino acid sequences. The conserved transmembrane domain is labeled, and potential palmitoylation sites are boxed. Potential Grb2-binding motifs are underlined. Arrows show the conserved prolines. The molecular weights of hGAPT and mGAPT are estimated to be 17.8 and 17.6 kDa, respectively.

Fig. 2.

Fig. 2.

Open in a new tab

GAPT is expressed in human immune tissues and associated with Grb2 in B cells. (A) GAPT expression in human cell lines and tissues was determined by RT-PCR. cDNAs from various human tissues or cell lines were used to amplify the GAPT transcript. The G3PDH transcript was also amplified to indicate that similar amounts of cDNAs were used. The figure shown is a representative of three experiments with similar results. (B) Membrane localization of GAPT. Cytosolic and membrane fractions of transduced BJAB cells were isolated by Dounce homogenization in hypotonic buffer followed by ultracentrifugation. Membrane fractions were solubilized with 1% Brij 97 lysis buffer before Western blot analysis. WCL, Whole cell lysate; WCL*, WCL without nuclear and cytoskeletal debris. (C) GAPT is not localized in lipid rafts. Triton-soluble and lipid raft fractions were separated by sucrose gradient centrifugation from cells expressing Myc-tagged GAPT. The presence of Lyn and GAPT was detected by Western blotting with anti-Lyn and anti-Myc antibodies, respectively. (D) Association of GAPT with Grb2 in B cells. Transduced BJAB cells were stimulated for 0, 0.75, 2, 5, and 10 min before GAPT was immunoprecipitated (IP) by anti-Myc antibody. Immunoprecipitates were then blotted with anti-p-tyrosine (pTyr) and anti-Grb2. LC, Ig light chain. (E) Importance of conserved proline residues in GAPT-Grb2 association. EV, Empty vector.

Fig. 3.

Fig. 3.

Open in a new tab

Targeted disruption of GAPT in mice. (A) GAPT gene-targeting strategy. The targeting construct was designed to replace the GAPT exon with the floxed exon. P1 and P2 indicate the primers used in PCR genotyping. (B) Southern blot analysis. The genomic DNAs from mouse tails were digested with EcoRI and BamHI (E and B, respectively, in A) and analyzed by Southern blotting using a probe shown in A. The expected sizes of detected DNA fragments are 12 kb for the WT allele and 9 kb for the floxed allele. (C) PCR genotyping of GAPT−/− mice. (D) Absence of GAPT protein in GAPT−/− mice. Protein lysates of BM cells and splenocytes from WT and GAPT−/− mice were analyzed by anti-GAPT Western blot. SPL, Spleen. (E) Distribution of GAPT in the mouse immune system. Different cell populations from the mouse immune system were purified and analyzed further by Western blotting with GAPT antisera. Immunoblotting with anti-HSP70 was used as a loading control. Mφ, Macrophage.

Cell proliferation assay

CD4+ and CD8+ T cells (>96% purity) were cultured at 4 × 105 cells/ml with medium only, anti-CD3 (5 μg/ml), and PMA (40 ng/ml) plus ionomycin (500 ng/ml) for 48 h in 96-well plates. B220+ B cells (>99% purity) were incubated with medium only, anti-IgM (5, 10, 20, 40, and 80 μg/ml), anti-CD40 (1 μg/ml), and LPS (1 μg/ml) for 48 h. Cells were then pulsed with 1 mCi/ml [3H] thymidine for 6 h before being assayed for thymidine incorporation using a liquid scintillation luminescence counter (PerkinElmer, Wellesley, MA, USA).

Immunization, germinal center formation, and ELISA

To evaluate T-independent antibody responses, 6- to 7-week-old wild-type (WT) and GAPT−/− mice were immunized with 50 μg nitrophenylacetyl (NP)38-Ficoll. These mice were bled on Days 7, 14, and 21, and sera were analyzed by ELISA. For T-dependent antibody responses, mice were immunized with 50 μg NP21-chicken γ-globulin (CGG) and boosted with the same amount of NP21-CGG on Day 21. Sera were collected at Days 0, 7, 14, and 21 after immunization and Days 5 and 10 after boost. NP-specific IgM or IgG was measured by ELISA. NP5-BSA or NP25-BSA was used to coat plates to measure high-affinity or low-affinity NP-specific antibodies, respectively. For germinal center formation, mice were immunized with 50 μg NP21-CGG, and 12 days later, splenocytes were isolated and analyzed for GL-7 and B220 double-positive (DP) cells using FACS. Spleen sections were also examined by immunostaining with GL-7 and B220 antibodies.

RESULTS

Molecular cloning of GAPT

A self-developed MotifFinder program was used to search for novel transmembrane adaptor proteins in the human genome database [9]. The candidates were selected based on the following criteria: multiple Grb2-binding motifs (YxN, where x is any amino acid), a putative transmembrane domain, and a potential palmitoylation site. One of the proteins we identified is a hypothetical protein encoded by a gene located at 5q11.2 (GenBank Accession No. AK090960). This protein contains seven tyrosine residues at its cytosolic region, four of them within a Grb2-binding motif. We named this protein GAPT after consulting with the Human Gene Organization (HUGO) Gene Nomenclature Committee (London, UK).

Blast search of the NCBI database, using the human gapt sequence, showed that the GAPT transcript is present in germinal center B cells, DCs, CD34+ hematopoietic stem cells, and myeloid cells. We also identified mouse gapt (GenBank Accession No. AK036534) by blast-searching the NCBI database with the hgapt sequence. Conceptual translation of gapt sequences revealed that hgapt and mgapt contain 157-aa residues and share 58% homology with molecular masses of 17.8 kDa and 17.6 kDa, respectively (Fig. 1, A and B).

Comparison of GAPT with LAT family proteins

Transmembrane adaptor proteins LAT, LAX, and LAB contain multiple conserved tyrosine residues in the cytoplasmic domain, five of which are within Grb2-binding motifs. Similar to these proteins, GAPT has five conserved tyrosine residues in the cytoplasmic tail, and four of them (Y93, Y112, Y126, Y133) are within Grb2-binding motifs (Fig. 1A). The Y93ENV and Y112ENT motifs in GAPT are similar to Y136ENV in LAB and Y226ENL in LAT, two critical motifs involved in Grb2 binding [22,23,24]. Although GAPT, LAT, and LAB do not have a significant homology in amino acid sequences, they share a similar domain structure. GAPT also has a short extracellular domain followed by a hydrophobic region (putative transmembrane domain; Fig. 1, A and B). Similar to LAT and LAB, there are two cysteine residues (C25GIGC29 in hGAPT and mGAPT) in the juxtamembrane region (Fig. 1B). In LAT, these cysteines are palmitoylated and required for its localization to lipid rafts and subsequent tyrosine phosphorylation [20].

Expression of GAPT in human tissues and cell lines

RT-PCR was used to examine expression of GAPT in different human tissues and cell lines. As shown in Fig. 2A, GAPT was expressed highly in human spleen and PBL. A small amount of GAPT was also detected in the thymus. No obvious expression of GAPT was detected in other tissues. When different human cell lines were examined, GAPT expression was seen in human B cell lines (BJAB, Daudi, Raji, and Jiyoye), monocytic line (THP1), and NK-like cells (YT), but not in the human T cell line Jurkat or Hela cells (Fig. 2A). These results indicate that GAPT is mainly expressed in hematopoietic tissues and B cell lines.

Subcellular localization of GAPT

GAPT has a putative transmembrane domain and a palmitoylation site similar to those in LAT [20]. Next, we asked whether GAPT is also localized in the membrane and lipid rafts. The full length of hGAPT cDNA was cloned into a retroviral vector with a C-terminal Myc tag. Recombinant retroviruses were made and used to transduce the human BJAB cell line to perform biochemical analysis. To determine the subcellular localization of GAPT, transduced BJAB cells were fractionated, and the cytosolic and membrane fractions were analyzed by Western blotting with an anti-Myc antibody. Myc-tagged GAPT migrated at the molecular weight of ∼25 kDa on SDS-PAGE. As shown in Fig. 2B, GAPT was detected exclusively in the membrane fraction, indicating that GAPT is associated with the membrane. As controls, Lyn was detected exclusively in the membrane fraction, and PLC-γ2 was only in the cytosolic fraction. As there are two cysteines, which could be palmitoylated in the juxtamembrane region of GAPT, we next determined whether GAPT is localized in lipid rafts by sucrose gradient centrifugation. As shown in Fig. 2C, GAPT was only detected in Triton X-100-soluble fractions, and Lyn was detected mostly in lipid raft fractions. These data indicated that although GAPT has two membrane-proximal cysteines, GAPT is excluded from lipid rafts and only associated with the plasma membrane.

Association of GAPT with Grb2

hGAPT and mGAPT have four tyrosine residues within a Grb2-binding motif in their cytoplasmic tails. Phosphorylation of the cytoplasmic tyrosine residues in LAT is important for its interaction with other proteins [22]. We next examined whether GAPT is phosphorylated and interacts with Grb2 through tyrosine residues upon BCR stimulation. BJAB cells expressing Myc-tagged GAPT were stimulated with anti-human IgM F(ab′)2 before lysis. GAPT was immunoprecipitated with an anti-Myc antibody and analyzed by Western blotting with Myc, pTyr, and Grb2 antibodies. Surprisingly, there was no apparent phosphorylation of GAPT upon stimulation via the BCR (Fig. 2D). Although GAPT was not tyrosine-phosphorylated, it was constitutively associated with Grb2, and there was no increase in Grb2 association after stimulation (Fig. 2D). This association was not a result of nonspecific binding, as Grb2 was not detected in anti-Myc immunoprecipitates from untransduced BJAB cells. These data indicated that GAPT binds to Grb2 through mechanisms other than the pTyr–SH2 domain interaction.

As there are multiple prolines conserved between hGAPT and mGAPT (Fig. 1B), we speculated that GAPT may interact with the SH3 domains of Grb2 through this potential proline-rich region. To examine this possibility, we generated a GAPT mutant with all three conserved prolines (P140, P141, and P145) mutated to alanines (GAPT 3PA). The mutant and WT GAPT were introduced into the BJAB cell line by retroviral transduction. Similar amounts of GAPT were expressed in WT and mutant cell lines, as detected by anti-Myc immunoblotting (Fig. 2E). The association between GAPT and Grb2 was abolished completely by praline-to-alanine mutations (Fig. 2E), indicating that the GAPT–Grb2 interaction is mediated via the proline residues and most likely, the SH3 domains of Grb2.

Targeted disruption of the gapt gene in mouse

The mgapt gene only has one exon. A construct containing the floxed exon, and FRT-flanked phosphoglycerine kinase (PGK)-Neo was used to disrupt the gapt gene. Successful targeting resulted in replacement of the exon with the floxed exon and FRT-flanked Neo (Fig. 3, A and B). Targeted mice were crossed with β-actin-Cre mice to generate complete GAPT knockout mice (Fig. 3A). Deletion of the gapt exon was detected by PCR (Fig. 3C) using two primers as indicated in Figure 3A. The absence of GAPT protein was further confirmed by Western blotting with rabbit anti-GAPT sera (Fig. 3D). GAPT−/− mice were normal in appearance, size, and fertility and were born at the expected Mendelian frequency and sex ratio.

GAPT distribution in primary murine immune cells

To compare the expression profile of GAPT in mouse and human tissues and to further determine its distribution in purified immune cells, we next examined the expression of GAPT in different cell types in the mouse immune system using GAPT antisera. Cells from GAPT−/− mice were used as negative controls to ensure positive identification of GAPT protein in Western blotting. Total BM cells and thymocytes were isolated directly from WT and GAPT−/− mice. CD4+, CD8+ T cells, and B220+ B cells were purified using AutoMACS. BMMCs and DCs were generated by culturing BM cells in the presence of IL-3 and GM-CSF, respectively. Peritoneal macrophages were also examined for GAPT expression. As shown in Fig. 3E, GAPT was expressed primarily in murine B220+ cells and total BM cells. It was expressed less abundantly in MCs and DCs, and it was absent in mouse T cells and macrophages.

Development of B cells and other immune cells in GAPT−/− mice

As GAPT is expressed primarily in B220+ cells, we first examined B cell development in GAPT−/− mice. B cell development occurs mainly in BM, and B cell maturation continues in peripheral lymphoid organs. The developmental process starts with pro-B cells, which are the first irreversible B cell precursors [25, 26]. Pro-B cells differentiate into pre-B cells and later, immature B cells in BM. Immature B cells have to go through negative selection before they progress to the periphery, where they mature further into different B cell subsets: transitional 1 (T1), transitional 2 (T2), marginal zone (MZ) B cells, follicular B cells, and CD5+ B1 B cells [27]. We first examined B cell precursors in BM, as shown in Fig. 4, A and 4B. The percentages and absolute numbers of pro-B (IgM−CD43+B220+), pre-B (IgM−CD43−B220+), immature B (B220loIgM+), and mature, recirculating B cells (B220hiIgM+) were similar in WT and GAPT−/− mice. The percentages and numbers of peripheral B (B220+) and T (Thy1.2+) cells were also similar between the WT and GAPT−/− mouse (Fig. 4D, and data not shown). Moreover, the percentages and absolute numbers of mature (IgMloIgDhi), immature (IgMhiIgDlo), T2 (IgMhiIgDhi or CD23+CD21+IgM+), T1 (CD23−CD21−IgM+), and MZ B cells (CD23−CD21+) in spleen were all normal (Fig. 4C). B1-B (CD5+B220+) cells from the peritoneal cavity also developed normally in GAPT−/− mice (data not shown). Together, these data indicated that GAPT is not required during B cell development, despite its abundant expression in these cells. We also examined T cell development in GAPT−/− mice. The percentages of DN (CD4−CD8−), DP (CD4+CD8+), and single-positive thymocytes (CD4+ or CD8+) were comparable between WT and GAPT−/− mice. The TCR expression was normal in GAPT−/− thymocytes. In the spleens, normal numbers of CD4+ and CD8+ T cells were seen in GAPT−/− mice (data not shown). These results were expected, as GAPT was not expressed in thymocytes or peripheral T cells.

Fig. 4.

Fig. 4.

Open in a new tab

Lymphocyte development in GAPT−/− mice. (A and B) B cell development in the BM of WT and GAPT−/− mice. B cell precursor subsets are characterized as the following: pro-B (IgM−CD43+B220+), pre-B (IgM−CD43−B220+), immature B (B220loIgM+), and mature, recirculating B cells (B220hiIgM+). The absolute numbers of these subsets per mouse are shown in B. (C and D) B cell differentiation in spleen of WT and GAPT−/− mice. FACS profiles of B cell subsets are in C, mature (IgMloIgDhi), immature (IgMhiIgDlo), T2 (IgMhiIgDhi or CD23+CD21+IgM+), T1 (CD23−CD21−IgM+), and MZ B cells (CD23−CD21+B220+). The absolute numbers of T cells and B cell subsets per mouse spleen are shown in D. (E) The numbers of NK and NKT cells, DCs, granulocytes, and macrophages per spleen in WT and GAPT−/− mice. PDCs, Plasmacytoid DCs.

GAPT was also expressed in NK-like cells (YT, Fig. 2A) and myeloid cells (Figs. 2A and 3E); so, we then examined the development of these cell types in GAPT−/− mice. Similar numbers of NK (DX5+TCRβ−) and NKT (DX5+TCRβ+) cells, Mac-1+Gr-1+ granulocytes, Mac1+Gr-1− macrophages, myeloid DCs (CD11chighCD8α−), lymphoid DCs (CD11chighCD8α+), and plasmacytoid DCs (CD11clowCD8αint) were found in GAPT−/− and WT mice (Fig. 4E, and data not shown). We also examined in vitro differentiation of DCs and MCs from the BM of WT and GAPT−/− mice in the presence of GM-CSF and IL-3, respectively. Similar numbers of BMMCs and BMDCs were obtained from WT and GAPT−/− BMs after culture (data not shown). Together, these data indicated that GAPT is not required for the development of NK and myeloid cells.

Humoral responses in GAPT−/− mice

To determine the role of GAPT in antibody responses, we first determined the concentration of serum Ig isotypes in naive mice by ELISA. Although serum IgG1 amounts appeared to be increased slightly in GAPT−/− mice (P=0.054; t-test), there were no significant differences in the concentrations of serum Ig isotypes between WT and GAPT−/− mice (Fig. 5A). We next examined whether antigen-induced germinal center formation is normal in GAPT−/− mice. Six- to 7-week-old mice were immunized with NP-CGG, and 12 days later, mice were examined for germinal center formation. FACS analysis showed normal numbers of GL-7+B220+ germinal center B cells in WT and GAPT−/− mice after immunization (Fig. 5B). Similar numbers of germinal centers were also observed in immunized WT and GAPT−/− mice by immunohistochemistry staining with GL-7 antibody (data not shown). We further determined T-dependent and -independent antibody responses in GAPT−/− mice. NP-CGG (for T-dependent) and NP-Ficoll (T-independent) were used for immunization, and mouse sera were collected at different time-points for subsequent ELISA assay. For T-dependent antibody response, high-affinity (Fig. 5C, top panel) and low-affinity (Fig. 5C, middle panel) antibody titers were similar in WT and GAPT−/− mice. T-independent response was also normal in GAPT−/− mice (Fig. 5C, bottom panel). These data indicated that GAPT deficiency did not affect humoral immune responses.

Fig. 5.

Fig. 5.

Open in a new tab

Humoral responses in GAPT−/− mice. (A) Concentrations of serum Ig isotypes in naive GAPT−/− mice. (B) Generation of germinal center B cells. Twelve days after immunization with NP-CGG, splenocytes were isolated and analyzed by FACS. B220+GL-7+ cells represent germinal center B cells. (C) Antibody responses in WT and GAPT−/− mice, which were immunized at Day 0 and boosted at Day 21 with NP-CGG or NP-Ficoll. T-dependent antibody production (top panel, high-affinity; middle panel, low-affinity) was determined by measuring NP-specific IgG with ELISA after NP-CGG immunization. T-independent antibody production (bottom panel) was determined by measuring NP-specific IgM after NP-Ficoll immunization. B5 and B10 indicate Day 5 or 10 after boost. (D) BCR-induced protein tyrosine phosphorylation and MAPK activation in GAPT−/− B cells. Purified B220+ splenocytes were stimulated by anti-mouse IgM F(ab′)2 for 0, 0.75, 2, 5, and 15 min before analysis by Western blot with anti-pTyr, -pErk, -pJNK, and -pp38. HSP70, Erk2, JNK, and p38 antibodies were used for reblotting stripped membranes to verify equal loading.

BCR-mediated signaling in GAPT−/− B cells

To examine the potential function of GAPT in BCR-mediated signaling, we determined total tyrosine phosphorylation and MAPK activation upon anti-IgM stimulation. B220+ splenocytes were purified from WT and GAPT−/− mice by positive selection with AutoMACS and were subsequently stimulated by anti-mouse IgM F(ab′)2 (20 μg/ml) for 0, 0.75, 2, 5, and 15 min before whole cell lysates were analyzed by Western blotting. As shown in Figure 5D, total protein tyrosine phosphorylation of GAPT−/− B cells was similar to that of WT. BCR-induced Erk, JNK, and p38 activation was also normal (Fig. 5D). These data indicated that GAPT is not required for BCR-induced Ras/MAPK activation.

Enhanced B cell proliferation in the absence of GAPT

To determine whether GAPT could be involved in lymphocyte proliferation or survival, we measured T and B cell proliferation in response to different stimuli by the [3H] thymidine incorporation assay. T cell proliferation induced by anti-CD3 and anti-CD28 cross-linking and PMA plus ionomycin were normal in the absence of GAPT (Fig. 6A). GAPT−/− B cell proliferated normally upon anti-CD40 or LPS stimulation; however, 48 h after stimulation through the BCR, [3H]-thymidine incorporation into proliferating B cells was increased significantly in the absence of GAPT. We further examined proliferation of GAPT−/− B cells in response to anti-IgM at different concentrations and found that B cell proliferation was enhanced at all concentrations, but the most significant differences were found at lower anti-IgM concentrations (Fig. 6A). These results suggested that GAPT might be important in regulation of B cell proliferation after persistent low-grade BCR ligation.

Fig. 6.

Fig. 6.

Open in a new tab

Enhanced B cell proliferation and expansion of MZ B cells. (A) B and T cells purified from spleens of WT and GAPT−/− mice were incubated with medium (–), anti-IgM F(ab′)2, anti-CD40, and LPS or anti-CD3 and PMA plus ionomycin (P+I), respectively, for 48 h. Proliferating cells were then examined by measuring [3H] thymidine incorporation. Experiments were performed in triplicate, and figures represent three independent experiments with similar results. Left panel, B cells were stimulated by anti-IgM at 10 μg/ml; right panel, B cells were stimulated by anti-IgM at different concentrations shown in the figure. (B) The absolute numbers of total T cells, B cells, and B cell subsets per mouse spleen. (C) Percentages of MZ B cells in WT and GAPT−/− mice. MZ B cells were characterized as CD23−CD21+B220+ cells. Figure also shows the increased percentage of MZ and T2 B cells as a result of increased MZ B cell percentage. (D) Concentrations of circulating serum Ig isotypes in aged GAPT−/− mice. *, P < 0.05; **, P < 0.01.

Increased MZ B cells and antibody concentrations in aged GAPT−/− mice

BCRs can be exposed to low-grade stimulation by self-antigens or infectious agents for a prolonged time during aging. Considering the enhanced B cell proliferation upon anti-IgM stimulation in the absence of GAPT, we set out to examine whether different B cell populations are normally maintained in aged GAPT−/− mice. Splenocytes from 8-month-old GAPT−/− mice and age, sex-matched WT mice were analyzed by FACS after staining with anti-CD23, CD21, and IgM. As shown in Figure 6B, aged WT and GAPT−/− mice had similar numbers of total peripheral T and B cells. The numbers of NK cells, DCs, and macrophages were also similar (data not shown). However, we found a significant increase (more than two-fold) of the MZ B cell population in GAPT−/− mice. Other B cell subsets, such as T1, T2, and follicular B cells, were normal in these mice (Fig. 6, B and C). As MZ B cells only account for a small percentage of total peripheral B cells, the increase in MZ B cells did not lead to a higher number of total B cells in GAPT−/− mice. Although this difference was only observed in old but not young GAPT−/− mice, these results suggested that GAPT might play a role in maintenance of MZ B cells.

In light of the increased number of MZ B cells in aged GAPT−/− mice, we also examined antibody concentrations of different isotypes in these aged mice. It is possible that abnormal expansion of MZ B cells might be accompanied by enhanced Ig production. Indeed, we found that the concentrations of IgM, IgG2b, and IgG3 in GAPT−/− mice were increased significantly compared with those in WT mice (Fig. 6D). There were no differences in the levels of IgG1 and IgA between aged WT and GAPT−/− mice. Together, these data suggest that GAPT might have a role in regulation of B cell activation and antibody production.

DISCUSSION

GAPT is a novel transmembrane adaptor protein. It shares several structural characteristics with LAT family adaptor proteins. GAPT possesses a transmembrane domain, a pair of membrane proximal cysteines, and multiple Grb2-binding motifs. However, GAPT likely functions differently from LAT family proteins as a result of its biochemical characteristics. LAT and LAB are palmitoylated and localized in lipid rafts [9, 20]. GAPT, similar to LAX [8], is localized in the membrane but excluded from lipid rafts. LAT, LAB, and LAX are phosphorylated upon stimulation through the TCR or BCR and bind to Grb2 and other SH2 domain-containing adaptor molecules [8, 9, 20]. In contrast, GAPT was not tyrosine-phosphorylated upon BCR ligation. Instead, GAPT constitutively associated with Grb2 in B cells via a proline-rich region. Mutation of GAPT proline residues abolished the association between GAPT and Grb2, indicating that instead of binding to the Grb2 SH2 domain, GAPT likely interacts with the SH3 domains of Grb2 through this proline-rich region. We have also examined the function of GAPT in MCs after engagement of the FcεRI. GAPT deficiency did not significantly affect FcεRI-mediated signaling and degranulation in MCs (data not shown). In addition, we did not detect phosphorylation of GAPT after stimulation via the FcεRI. Although we did not observe tyrosine phosphorylation of GAPT upon engagement of the BCR, it is possible that GAPT might be phosphorylated after activation or engagement of other receptors in B cells and interacts with other SH2 domain-containing proteins. This possibility remains to be explored in the near future.

Our results indicated that GAPT binds Grb2 constitutively; however, the deletion of the GAPT protein did not affect Ras/MAPK activation upon BCR stimulation, suggesting that GAPT plays a minimal role in activation of this pathway. This is consistent with our results, showing that GAPT is not required during B cell development, as BCR-induced Ras/MAPK activation is important for this process [28]. Although BCR-mediated MAPK activation was normal, GAPT−/− B cells were hyper-responsive to stimulation via the BCR, suggesting that GAPT might have a negative role in regulating B cell proliferation. GAPT might recruit other negative regulators, such as phosphatases, to down-regulate B cell activation. It is also possible that GAPT can regulate the survival of B cells rather than the activation of B cells.

Although there were no major defects in the development and function of the immune system in young GAPT-deficient mice, there were more MZ B cells in aged GAPT−/− mice. It is not clear why only MZ B cells were affected; however, the special, functional characteristics of this B cell subset might account for the observed phenotype. MZ B cells, as a result of their unique localization in the spleen, are more readily activated and proliferate more rapidly in response to blood-borne pathogens than follicular B cells and immature T1 and T2 B cells [29, 30]. Although follicular B cells are the major, mature B cell type that initiates T-dependent responses and mainly secretes antibodies of the IgG1 isotype, MZ B cells primarily respond to T-independent pathogens and preferentially produce IgM and IgG3 isotypes [31,32,33]. Therefore, the elevated concentrations of IgM, IgG2b, and IgG3, but not IgG1 and other Ig isotypes in old GAPT−/− mice, further demonstrated that the expanded MZ B cells might be responsible for the increased basal Ig concentrations in aged GAPT−/− mice. Besides being the first line of B cell defense, MZ B cells also differ from follicular B cells in their lifespan. These cells are relatively long-lived compared with follicular B cells [30, 34]. The extended lifespan might expose them to accumulative, low-grade stimulation through the BCR, which causes increased proliferation and secretion of antibodies in the absence of GAPT.

In summary, we identified GAPT, a novel Grb2-binding transmembrane adaptor protein that is mainly expressed in B cells. Functional studies of GAPT by targeted disruption of the gapt gene in mice revealed that it negatively regulates B cell proliferation after stimulation through the BCR and might play an important role in normal maintenance of MZ B cells. How GAPT regulates B cell activation remains to be determined. Although GAPT was not required for the development of DCs, NK cells, and granulocytes, it is still possible that GAPT might play a role in the activation of these cells through other receptors. This possibility remains to be explored in the future.

Acknowledgments

This work was supported by National Institutes of Heath grants AI048674 and AI056156. Y. L. performed research, analyzed data, and wrote the paper. W. Z. designed the project, analyzed data, and edited the paper. W. Z. is a scholar of the Leukemia and Lymphoma Society. The authors declare no competing financial interests. The authors thank the Duke University Cancer Center Flow Cytometry, DNA Sequencing, Human Vaccine Institute, and Transgenic Mouse facilities for their excellent services and Eva Chung, Kathleen O'Hara, and Josh Mahlios for careful reading of the manuscript and their valuable advice.

References

  1. Horejsi V, Zhang W, Schraven B. Transmembrane adaptor proteins: organizers of immunoreceptor signaling. Nat Rev Immunol. 2004;4:603–616. doi: 10.1038/nri1414. [DOI] [PubMed] [Google Scholar]
  2. Samelson L E. Signal transduction mediated by the T cell antigen receptor: the role of adapter proteins. Annu Rev Immunol. 2002;20:371–394. doi: 10.1146/annurev.immunol.20.092601.111357. [DOI] [PubMed] [Google Scholar]
  3. Jordan M S, Singer A L, Koretzky G A. Adaptors as central mediators of signal transduction in immune cells. Nat Immunol. 2003;4:110–116. doi: 10.1038/ni0203-110. [DOI] [PubMed] [Google Scholar]
  4. Yablonski D, Weiss A. Mechanisms of signaling by the hematopoietic-specific adaptor proteins, SLP-76 and LAT and their B cell counterpart, BLNK/SLP-65. Adv Immunol. 2001;79:93–128. doi: 10.1016/s0065-2776(01)79003-7. [DOI] [PubMed] [Google Scholar]
  5. Finco T S, Kadlecek T, Zhang W, Samelson L E, Weiss A. LAT is required for TCR-mediated activation of PLCγ1 and the Ras pathway. Immunity. 1998;9:617–626. doi: 10.1016/s1074-7613(00)80659-7. [DOI] [PubMed] [Google Scholar]
  6. Zhang W, Irvin B J, Trible R P, Abraham R T, Samelson L E. Functional analysis of LAT in TCR-mediated signaling pathways using a LAT-deficient Jurkat cell line. Int Immunol. 1999;11:943–950. doi: 10.1093/intimm/11.6.943. [DOI] [PubMed] [Google Scholar]
  7. Zhang W, Sommers C L, Burshtyn D N, Stebbins C C, DeJarnette J B, Trible R P, Grinberg A, Tsay H C, Jacobs H M, Kessler C M, Long E O, Love P E, Samelson L E. Essential role of LAT in T cell development. Immunity. 1999;10:323–332. doi: 10.1016/s1074-7613(00)80032-1. [DOI] [PubMed] [Google Scholar]
  8. Zhu M, Janssen E, Leung K, Zhang W. Molecular cloning of a novel gene encoding a membrane-associated adaptor protein (LAX) in lymphocyte signaling. J Biol Chem. 2002;277:46151–46158. doi: 10.1074/jbc.M208946200. [DOI] [PubMed] [Google Scholar]
  9. Janssen E, Zhu M, Zhang W, Koonpaew S, Zhang W. LAB: a new membrane-associated adaptor molecule in B cell activation. Nat Immunol. 2003;4:117–123. doi: 10.1038/ni882. [DOI] [PubMed] [Google Scholar]
  10. Brdicka T, Imrich M, Angelisova P, Brdickova N, Horvath O, Spicka J, Hilgert I, Luskova P, Draber P, Novak P, Engels N, Wienands J, Simeoni L, Osterreicher J, Aguado E, Malissen M, Schraven B, Horejsi V. Non-T cell activation linker (NTAL): a transmembrane adaptor protein involved in immunoreceptor signaling. J Exp Med. 2002;196:1617–1626. doi: 10.1084/jem.20021405. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Brdickova N, Brdicka T, Angelisova P, Horvath O, Spicka J, Hilgert I, Paces J, Simeoni L, Kliche S, Merten C, Schraven B, Horejsi V. LIME: a new membrane Raft-associated adaptor protein involved in CD4 and CD8 coreceptor signaling. J Exp Med. 2003;198:1453–1462. doi: 10.1084/jem.20031484. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hur E M, Son M, Lee O H, Choi Y B, Park C, Lee H, Yun Y. LIME, a novel transmembrane adaptor protein, associates with p56lck and mediates T cell activation. J Exp Med. 2003;198:1463–1473. doi: 10.1084/jem.20030232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Zhu M, Granillo O, Wen R, Yang K, Dai X, Wang D, Zhang W. Negative regulation of lymphocyte activation by the adaptor protein LAX. J Immunol. 2005;174:5612–5619. doi: 10.4049/jimmunol.174.9.5612. [DOI] [PubMed] [Google Scholar]
  14. Ahn E, Lee H, Yun Y. LIME acts as a transmembrane adapter mediating BCR-dependent B-cell activation. Blood. 2006;107:1521–1527. doi: 10.1182/blood-2005-05-1859. [DOI] [PubMed] [Google Scholar]
  15. Janssen E, Zhu M, Craven B, Zhang W. Linker for activation of B cells: a functional equivalent of a mutant linker for activation of T cells deficient in phospholipase C-γ1 binding. J Immunol. 2004;172:6810–6819. doi: 10.4049/jimmunol.172.11.6810. [DOI] [PubMed] [Google Scholar]
  16. Zhu M, Liu Y, Koonpaew S, Granillo O, Zhang W. Positive and negative regulation of FcεRI-mediated signaling by the adaptor protein LAB/NTAL. J Exp Med. 2004;200:991–1000. doi: 10.1084/jem.20041223. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Wang Y, Horvath O, Hamm-Baarke A, Richelme M, Gregoire C, Guinamard R, Horejsi V, Angelisova P, Spicka J, Schraven B, Malissen B, Malissen M. Single and combined deletions of the NTAL/LAB and LAT adaptors minimally affect B-cell development and function. Mol Cell Biol. 2005;25:4455–4465. doi: 10.1128/MCB.25.11.4455-4465.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Volna P, Lebduska P, Draberova L, Simova S, Heneberg P, Boubelik M, Bugajev V, Malissen B, Wilson B S, Horejsi V, Malissen M, Draber P. Negative regulation of mast cell signaling and function by the adaptor LAB/NTAL. J Exp Med. 2004;200:1001–1013. doi: 10.1084/jem.20041213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Zhu M, Koonpaew S, Liu Y, Shen S, Denning T, Dzhagalov I, Rhee I, Zhang W. Negative regulation of T cell activation and autoimmunity by the transmembrane adaptor protein LAB. Immunity. 2006;25:757–768. doi: 10.1016/j.immuni.2006.08.025. [DOI] [PubMed] [Google Scholar]
  20. Zhang W, Trible R P, Samelson L E. LAT palmitoylation: its essential role in membrane microdomain targeting and tyrosine phosphorylation during T cell activation. Immunity. 1998;9:239–246. doi: 10.1016/s1074-7613(00)80606-8. [DOI] [PubMed] [Google Scholar]
  21. Inaba K, Inaba M, Romani N, Aya H, Deguchi M, Ikehara S, Muramatsu S, Steinman R M. Generation of large numbers of dendritic cells from mouse bone marrow cultures supplemented with granulocyte/macrophage colony-stimulating factor. J Exp Med. 1992;176:1693–1702. doi: 10.1084/jem.176.6.1693. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Zhang W, Trible R P, Zhu M, Liu S K, McGlade C J, Samelson L E. Association of Grb2, Gads, and phospholipase C-γ 1 with phosphorylated LAT tyrosine residues. Effect of LAT tyrosine mutations on T cell antigen receptor-mediated signaling. J Biol Chem. 2000;275:23355–23361. doi: 10.1074/jbc.M000404200. [DOI] [PubMed] [Google Scholar]
  23. Liu S K, Fang N, Koretzky G A, McGlade C J. The hematopoietic-specific adaptor protein gads functions in T-cell signaling via interactions with the SLP-76 and LAT adaptors. Curr Biol. 1999;9:67–75. doi: 10.1016/s0960-9822(99)80017-7. [DOI] [PubMed] [Google Scholar]
  24. Koonpaew S, Janssen E, Zhu M, Zhang W. The importance of three membrane-distal tyrosines in the adaptor protein NTAL/LAB. J Biol Chem. 2004;279:11229–11235. doi: 10.1074/jbc.M311394200. [DOI] [PubMed] [Google Scholar]
  25. Nutt S L, Heavey B, Rolink A G, Busslinger M. Commitment to the B-lymphoid lineage depends on the transcription factor Pax5. Nature. 1999;401:556–562. doi: 10.1038/44076. [DOI] [PubMed] [Google Scholar]
  26. Allman D, Li J, Hardy R R. Commitment to the B lymphoid lineage occurs before DH-JH recombination. J Exp Med. 1999;189:735–740. doi: 10.1084/jem.189.4.735. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Hardy R R, Shinton S A. Characterization of B lymphopoiesis in mouse bone marrow and spleen. Methods Mol Biol. 2004;271:1–24. doi: 10.1385/1-59259-796-3:001. [DOI] [PubMed] [Google Scholar]
  28. Genot E, Cantrell D A. Ras regulation and function in lymphocytes. Curr Opin Immunol. 2000;12:289–294. doi: 10.1016/s0952-7915(00)00089-3. [DOI] [PubMed] [Google Scholar]
  29. Oliver A M, Martin F, Kearney J F. IgMhighCD21high lymphocytes enriched in the splenic marginal zone generate effector cells more rapidly than the bulk of follicular B cells. J Immunol. 1999;162:7198–7207. [PubMed] [Google Scholar]
  30. Pillai S, Cariappa A, Moran S T. Marginal zone B cells. Annu Rev Immunol. 2005;23:161–196. doi: 10.1146/annurev.immunol.23.021704.115728. [DOI] [PubMed] [Google Scholar]
  31. Zinkernagel R M, Ehl S, Aichele P, Oehen S, Kundig T, Hengartner H. Antigen localization regulates immune responses in a dose- and time-dependent fashion: a geographical view of immune reactivity. Immunol Rev. 1997;156:199–209. doi: 10.1111/j.1600-065x.1997.tb00969.x. [DOI] [PubMed] [Google Scholar]
  32. Cyster J G. B cells on the front line. Nat Immunol. 2000;1:9–10. doi: 10.1038/76859. [DOI] [PubMed] [Google Scholar]
  33. Martin F, Oliver A M, Kearney J F. Marginal zone and B1 B cells unite in the early response against T-independent blood-borne particulate antigens. Immunity. 2001;14:617–629. doi: 10.1016/s1074-7613(01)00129-7. [DOI] [PubMed] [Google Scholar]
  34. Hao Z, Rajewsky K. Homeostasis of peripheral B cells in the absence of B cell influx from the bone marrow. J Exp Med. 2001;194:1151–1164. doi: 10.1084/jem.194.8.1151. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Journal of Leukocyte Biology are provided here courtesy of The Society for Leukocyte Biology

RESOURCES