A Polypyrimidine Tract-Binding Protein-Dependent Pathway of mRNA Stability Initiates with CpG Activation of Primary B Cells¹

Thymus-dependent B cell activation requires both cognate and noncognate CD4⁺ T cell interactions to induce a comprehensive humoral immune response against a wide range of pathogens. The ligand for CD40 (CD154 or CD40L) is critical for orchestrating this process by binding to CD40 on Ag-activated B cells (1, 2, 3, 4). Interaction between CD154 expressed on CD4⁺ T cells and CD40 on B cells results in proliferation, differentiation, isotype switching, affinity maturation, and the development of the memory cell compartment (3, 5).

In contrast, T cell-independent mechanisms are not dependent on direct T cell interactions and are generally mediated by pathogenic surface Ags that express repetitive epitopes capable of cross-linking the BCR (6). Additionally, B cells express TLR that recognize conserved pathogen-associated molecular patterns such as LPS, peptidoglycan, dsRNA, flagellin, and unmethylated CpG repetitive DNA (6, 7, 8). The TLR9 protein has been shown to bind specifically to unmethylated-CpG DNA and is expressed on endosomal membranes of both primary and transformed B cells (9). TLR9 engagement induces proliferation, IgG class switch recombination, and the expression of cytokines IL-6 and IL-12 (10, 11, 12), whereas dual signaling through CD40 and TLR9 provides synergistic activation of tonsillar B cells into Ab-producing cells (13).

Activation-induced genes are often regulated via transcriptional processes, although posttranscriptional control represents a second major pathway of regulation (reviewed in Refs. 14, 15, 16, 17). Differential RNA decay acts to quickly integrate distinct cellular signals to change the expression of a particular gene or set of genes necessary for an appropriate immune response. A major pathway of RNA turnover involves the interaction of several RNA binding proteins with AU-rich elements (AREs)⁴ located in the 3′ untranslated region (UTR) of cytokine and growth factor mRNAs (reviewed in Refs. 17, 18). These RNA binding proteins include AUF1 and tristetraprolin that act to destabilize the associated transcript and HuR that has been shown to stabilize the targeted message. The mRNA decay initiates with the degradation of the poly(A) tail, and the remaining transcript is recognized and degraded either by a 5′ or 3′ exonuclease pathway (reviewed in Ref. 17). In contrast, non-ARE pathways often involve a stabilizing protein that binds to a specific sequence and the absence of binding leads to increased turnover of the message. For example, one important class of stability elements is distinguished by a high (CU) content and the ability to extend the half-life of long-lived messages (reviewed in Ref. 19). The most widely studied element in this class is the erythroid-specific mRNA stability determinant, composed of a C-rich sequence in a highly pyrimidine-rich region of the α-globin 3′ UTR. This element is functionally linked to the stability of the α-globin transcript and binds a set of proteins, termed the α-complex, which contains α-CP-1 and α-CP-2 as well as additional cofactors (20, 21).

Over the past several years, our lab has demonstrated the importance of a non-ARE pathway of posttranscriptional control in regulating CD154 expression in differentially activated primary CD4⁺ T cells. At early times of anti-CD3 activation, the CD154 message is highly unstable with a half-life of less than 40 min. However, at later times of activation, the half-life is increased to more than 2 h (22). A sequence was identified in the CD154 3′ UTR that contained three adjacent CU-rich binding sites that bound either a dominant RNA binding complex termed complex I, or a minor complex, termed complex II (23). Subsequent work revealed that the major protein in complex I was polypyrimidine tract-binding protein (PTB) or hnRNP I (24). Nucleolin was a second component that bound both to PTB and the CD154 transcript (25). Importantly we demonstrated that complex I played a critical role in stabilizing the CD154 transcript at late times of activation (24).

In this report we extend our previous work on the role of complex I-mediated mRNA stability in activated T cells by identifying a PTB-containing complex, B-cpx I (B cell-complex I), which is regulated in response to B cell activation. Specifically we show that B-cpx I is expressed in response to engagement of TLR9 and only minimally to CD40 signals. Microarray analysis identified a subset of transcripts that interacted with PTB in activated B cells and further work identified a transcript, Rab8A, which was stabilized in a PTB-dependent process in response to CpG activation. Together our findings reveal a PTB-regulated pathway of mRNA decay common to both activated T cells and B cells. Furthermore, in B cells, this pathway appears to be highly dependent on a specific activation pathway because PTB-regulated decay is most prominently observed with TLR9 but not CD40 signals.

The anti-PTB hybridoma CRL-2501 was obtained from American Type Culture Collection (ATCC). Anti-PTB mAb was purified from conditioned medium using protein A beads. The Jurkat/D1.1 cell line was obtained from Dr. S. Lederman (Columbia University, New York, NY) and cultured as previously described (25). The D11-LCL^tet line was generated using a mini-EBV vector expressing the LMP1 gene under the control of a tetracycline responsive promoter (26) and from PBMCs obtained from a healthy volunteer. Generation and maintenance of this line has recently been described (27). The 10103 CpG ODN was obtained from Coley Pharmaceutical.

Heparinized blood from healthy volunteers was obtained as fractionated buffy coats from the New York Blood Center (New York, NY). PBMCs were purified by centrifugation over Ficoll as previously described (24). Primary B cells were isolated using Dynabeads CD19 Pan B magnetic beads according to the manufacturer’s instructions (Invitrogen). For a subset of experiments, CD19 selected cells were incubated for 30 min on a 10-cm petri dish coated with 5 ml of 25 μg/ml anti-IgG (Southern Biotechnology Associates) to remove IgG⁺ B cells. The unattached cells (CD19⁺/IgG⁻) were removed by pipetting, and isolated primary B cells were cultured in RPMI 1640 supplemented with 10% FCS, 2 mM glutamine, and 1 mM sodium pyruvate at a concentration of 2.5 × 10⁶ cells/ml. B cells were activated for 48 h with the following stimulators alone or in combination: 3 μM 10103 CpG ODN, membranes purified from the 293 cell line expressing CD40L (293/CD40L) as previously described (27) and 200 U/ml IL-4 (PeproTech).

Primary CD19⁺/IgG⁻ or D11-LMP^tet B cells were washed in PBS and resuspended in extraction buffer (0.2% Nonidet P-40, 40 mM KCl, 10 mM HEPES (pH 7.9), 3 mM MgCl₂, 5% glycerol, 1 mM DTT, and 1X protease inhibitor cocktail (Sigma-Aldrich)) as previously described (28). Cells were incubated on ice for 5 min and centrifuged at 12,000 × g for 10 min, and the supernatant collected.

For RNA immunoprecipitation, protein A beads (Santa Cruz Biotechnology) were coated with anti-PTB mAb (ATCC) or IgG2_b (Southern Biotechnology Associates). Cells were washed two times in PBS, and cytoplasmic extracts were prepared as described. Cellular equivalents of DNase I-treated cytoplasmic extract (∼100–200 μg from resting or CpG-activated primary B cells) were loaded onto the Ab-coated beads resuspended in NT2 buffer (50 mM Tris (pH 7.4), 150 mM NaCl, 1 mM MgCl₂, 0.05% Nonidet P-40, 20 mM EDTA, 1 mM DTT, 200 U/ml rRNasin (Promega)). The bead or extract mixture was incubated on a rotator at 4°C for 4 h, washed six times with ice-cold NT2 buffer, resuspended in NT2 buffer plus 0.1% SDS and 30 μg proteinase K followed by incubation at 55°C for 2–3 h. The supernatant was collected and RNA isolated by phenol-chloroform extraction and ethanol precipitation.

RNA quantity and quality were assessed by NanoDrop spectrophotometer and electrophoresis using the Agilent Bioanalyzer 2100 (Quantum Analytics) before cDNA synthesis. Immunoprecipitated RNA (1 μg) was used to synthesize cDNA using Superscript II and oligo(dT)₁₅ following the manufacturer’s protocol (Invitrogen). Biotin-labeled cRNA was generated using the ENZO BioArray RNA transcript labeling kit and fragmented at 94°C for 35 min. All samples (15 μg) were subjected to gene expression analysis via the Affymetrix Human U133A 2.0 high-density oligonucleotide array, which queried 18,400 human transcripts. Processing, detection, and quantification was performed according to the manufacturer’s directions. Array analysis was conducted with two independently isolated and immunoprecipitated RNA samples at the Transcriptional Profiling Shared Resource of the Cancer Institute of New Jersey (New Brunswick, NJ). The third microarray was performed in a similar fashion using the Affymetrix Human U133 Plus 2.0 high-density oligonucleotide array, which queried 47,400 human transcripts including the 18,400 on the U133A array according to the manufacturer’s protocol at the Bionomics Research and Technology Center, Environmental and Occupational Health Sciences Institute (Rutgers University, Piscataway, NJ).

Each set of arrays was normalized to each other and a comparison analysis was performed using Affymetrix MAS 5.0 to determine which transcripts were significantly increased (2-fold or more). Wilcoxon signed-rank test was used to generate change p values to indicate the likelihood of change in transcript expression level.

RNAs were immunoprecipitated as described with IgG2b or anti-PTB mAb from cytoplasmic extracts of CpG-activated CD19⁺ B cells. Reverse transcription reactions were conducted using Transcriptor First Strand cDNA Synthesis Kit (Roche). A one-tenth volume equivalent of cytoplasmic extracts was also reverse transcribed and used as a positive control for each reaction. PCR (25 μl) was performed with 0.3 μM 3′ and 5′ primers and FastStart TaqDNA polymerase (Roche). The cycling parameters were as follows: 94°C for 5 min; 40 cycles of 94°C for 10 s; 10 s at 60°C; and 72°C for 20 s. Sets of intron-spanning primers were designed using the Universal Probe Library Assay Design Center (http://www.roche-applied-science.com/sis/rtpcr/upl/adc.jsp) with the exception of primers used for β–actin transcript detection (see primers used in quantitative PCR).

The complex I XbaI-HaeIII probe was synthesized as previously described (24). Primers used for PCR amplification of gene-specific RNA probes and mapping of the Rab8A binding site are as follows: cyclin D₂ (forward) 5′-CGTAATACGACTCACTATAGGGAAGGAATCCTGGATTTTG-3′, (reverse) 5′-TAGGATTTGGCCAAAAGAGGGGAAGGATCA-3′; CD20 (forward) 5′-CGTAATACGACTCACTATAGGGGTGATTTCTTCTGTTTTC-3′, (reverse) 5′-AAGCTATGCTGTTTTTACAAAAAACAGATG-3′; SET (forward) 5′-CGTAATACGACTCACTATAGGGATAGAACACTGATGGATT-3′, (reverse) 5′-CAGAGCGAAGGGAGCAGGTTTTTCTTTTTT-3′; RAB8 (forward) 5′-CGTAATACGACTCACTATAGGGCCTTTTCTTTTTTTCTTT-3′, (reverse) 5′-AAAAGAAACGATGATGCCAATGGTTTGGAT-3′; RAB8 M5′-9 (forward) 5′-CGTAATACGACTCACTATAGGGTTTTTCTTTCTTTTTTTT-3′; RAB8 M5′-21 (forward) 5′-CGTAATACGACTCACTATAGGGTTTTTTTTTTTCCTCCTT-3′; RAB8 M3′-8 (reverse) 5′-CGATGATGCCAATGGTTTGGATTGACAGCA-3′; RAB8 M3′-29 (reverse) 5′-TTGACAGCAGCTTAAGGAGGAAAA-3′; and RAB8 M3′-34 (reverse) 5′-AGCAGCTTAAGGAGGAAAAAAAAA-3′. RNA probes were synthesized and labeled with [α-³²P]rUTP (PerkinElmer Life Sciences) using 100 ng of purified PCR product and T7 polymerase (Promega). The RNA probe was purified by elution through a G25 column (Amersham Biosciences).

Reactions were prepared in RNA binding buffer with 4 ng of Escherichia coli tRNA, 5 μg of cytoplasmic extract and 40,000 cpm of RNA probe for primary B cells or 5000 cpm for D11-LCL^tet cells. The reactions were incubated for 20 min at room temperature to allow for complex formation followed by the addition of 1 μl of RNase mix containing 40 U RNase T₁, 10 ng of RNaseA, 0.01 U RNase V₁ (Ambion) for primary B cells, and a 1/8 dilution of this RNase mix for D11-LCL^tet cells. Samples were incubated at 37°C for 30 min, 2 μl of 50 μg/ml heparin was added and reactions were incubated on ice for 10 min. Protein-RNA complexes were separated on a 7% native acrylamide gel in 0.25X TBE. Supershift experiments were performed with 1 μg of anti-PTB or control IgG2_b Abs added to the reactions for 1 h at room temperature before addition of the probe.

The 52 bp minimum binding site for B-cpx I identified in the 3′ UTR of Rab8A was amplified using PfuUltra Fusion HS DNA polymerase (Stratagene) and D11-LCL^tet cDNA using the following primers: Rabd52 3′ forward 5′-GGGTGTCACCAGTCCAAACCATTGGCATCA-3′ and Rabd52 3′ reverse 5′-TGATGCCAATGGTTTGGACTGGTGACACCC-3′. The amplified region was cloned into the XbaI site located in the 3′ UTR of the Renilla luciferase gene of the pRLSV40 plasmid (Promega). A control vector was constructed that contained the 52-bp sequence ligated into the XbaI site in the reverse orientation. The pGL23a plasmid containing the firefly luciferase under the control of the CD23a promoter from Dr. S. Lederman (Columbia University, New York, NY) was used as a transfection control in all luciferase assays performed.

Approximately, 5 × 10⁶ D11-LCL^tet cells were grown in 1 μg/ml tetracycline and incubated with 2.5 μg of the binding site construct and 25 μg of pGL23a control plasmid. The cells were transfected with 1 pulse of 250 mV and 960 μF capacitance using an electroporator (Bio-Rad), resuspended in medium plus tetracycline and 3 μM CpG ODN and incubated for 48 h. After incubation, cells were harvested and extracts prepared for analysis of luciferase activity using a Dual Luciferase Assay kit (Promega) and a Glomax Luminometer (Promega). Results were normalized to firefly luciferase activity to account for differences in transfection efficiency.

To analyze the decay rate of Rab8A in primary B cells, 2 × 10⁷ CD19⁺/IgG⁻ B cells were incubated with 3 μM CpG ODN for 48 h followed by 50 μg/ml DRB (5,6-Dichlorobenzimidizole 1-β-D-ribofuranoside). A total of 5 × 10⁶ cells were removed at each time point over a 4 h time course. RNA was isolated using the High Pure RNA Isolation kit (Roche).

The effect of the Rab8A B-cpx I binding region on mRNA stability was determined by transiently transfecting 3 × 10⁷ D11-LCL^tet cells with the different Rab8A luciferase constructs using the Amaxa Biosystems Transfection System followed by incubation with 3 μM CpG for 24 h. Treatment with DRB over a time course was conducted as described. RNA was prepared using TRIzol (Invitrogen) followed by treatment with Turbo DNase (Ambion).

cDNA was synthesized from DRB-treated RNA samples using the Transcriptor First Strand cDNA Synthesis kit (Roche). Quantitative PCR was performed on the cDNA using the following primers: 5′-GCATCCTCACCCTGAAGTA-3′ together with 5′-TGTGGTGCCAGATTTTGTCC-3′ to detect the β–actin transcript and 5′-CCAAGACACAAGGCATTCCA-3′ and 5′-GTCCCAGTCGCAGTCCCTAT-3′ to detect the Rab8A transcript. Amplification reactions were conducted using both the Mx4000 Multiplex PCR System and FastStart SYBR Green Master mix (Rox) according to the manufacturer’s directions (Stratagene and Roche, respectively). After the amplification a melt curve was performed to ensure amplicon homogeneity. Decay of a specific RNA transcript was determined by the average threshold cycle (2^−ΔΔCt) method comparing each time point to the zero time point to quantify the percentage of RNA remaining (29, 30) using the MxPro-Mx3000P software.

To generate the pLVTHM-U6-shPTB and pLVTHM-U6-shCTRL vectors the following primers were annealed and cloned into the HindIII and EcoRI sites of the pSilencer2.1-U6 hygro (Ambion). Primers for U6-shPTB 5′-GATCCAACTTCCATCATTCCAGAGAACTTGCTTCTTCTCTGGAATGATGGAAGTTTTTTTTGGAAA-3′ and 5′-AGCTTTTCCAAAAAAAACTTCCATCATTCCAGAGAAGAAGCAAGTTCTCTGGAATGATGGAAGTTG-3′ and for U6-shControl 5′-GATCCAATCAGACGTGGACCAGAAGAGAGATCTTCTGGTCCACGTCTGATTTTTTTTGGAAA-3′ and 5′-AGCTTTTCCAAAAAAAATCAGACGTGGACCAGAAGATCTCTCT TCTGGTCCACGTCTGATTG-3′. The sequence containing the U6 promoter and shRNA was then PCR amplified from these vectors using the primer set 5′-CCATCGATGGAGCTTTTCCAAAAAAAACTT-3′ and 5′-CAGAAAGGTGACCCCTTAAGCTTCTAGAAG-3′. These PCR products were then cut with the restriction enzymes ClaI and BglII and cloned into the corresponding sites of the pLVTHM vector (Addgene plasmid 12247).

Viral particles were packaged by transfection of 293T cells with either pLVTHM-U6-shPTB or pLVTHM-U6-shCTRL together with the virus packaging plasmids pSPAX2 (Addgene plasmid 12260) and pCI-VSVG (Addgene plasmid 1733), both obtained from D. Trono (Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland), using FuGENE HD Transfection Reagent (Roche). Culture supernatants were collected every 12 h. A total of 5 × 10⁷ D11-LCL^tet cells were infected with 5 ml of viral supernatant in a final volume of 20 ml. At 48 h postinfection, an aliquot of 1 × 10⁷ cells was incubated in medium containing 10 mg/ml doxycycline for 24 h. These cells were then analyzed for GFP expression by FACS analysis to determine the percentage of infected cells. Cytoplasmic extracts were also analyzed by Western blot to determine the level of PTB between cells expressing a control sequence containing a scrambled shRNA (shCTRL) and cells containing a shRNA against PTB (shPTB). The remaining cells were incubated with 50 μg/ml DRB over a 3-h time course and RNA was isolated and analyzed by quantitative PCR as described.

For comparison of two samples, a two-tailed Student’s t test was used. Significance was set at p < 0.05. Data shown are mean ± SEM unless otherwise indicated.

Primary CD19⁺ B cells were isolated from human peripheral blood and activated with CD154-expressing 293 cells or CpG. RNA EMSA was conducted with cytoplasmic extracts from CpG, CD154, CpG plus CD154, or CD154 plus IL-4-stimulated cells and an in vitro synthesized and labeled RNA probe containing the CD154 stability element (termed XbaI-HaeIII) (Fig. 1,A). This probe binds complex I in extracts from both late activated CD4⁺ T cells and Jurkat/D1.1 T cells (24). As shown in Fig. 1,B, strong complex binding was observed with extracts from B cells stimulated with CpG (Fig. 1,B, lanes 5, 6, and 8), although there was a low level of complex formed with extracts from cells stimulated both with CD154 and CD154 plus IL-4 (Fig. 1,B, lanes 4 and 7). Little to no complex binding activity was seen with extracts from resting B cells (Fig. 1,B, lane 3). Because we had previously shown that PTB is a major protein in complex I from activated T cells, RNA EMSA was conducted using anti-PTB Abs to determine whether PTB is present in the B cell complex. Accordingly, PTB was also found to be a major component of the complex in activated B cells (Fig. 1 B, lane 9) and this result was confirmed by UV cross-linking (data not shown). Based on the similarity of this complex to complex I, we designated it B cell-complex I or B-cpx I.

To identify transcripts bound by B-cpx I, cytoplasmic extracts from CpG-treated and resting CD19⁺ B cells were immunoprecipitated with anti-PTB Abs. Bound transcripts were isolated, reversed transcribed, and used as probes on the U133A 2.0 human oligonucleotide array. Results revealed that 48 transcripts gave positive signals on two independent arrays with an ∼2-fold or greater increase over resting B cells on each array. To eliminate any contribution from the small percentage (<10%) of IgG⁺ B cells, another set of RNAs was immunoprecipitated from cytoplasmic extracts of CpG-treated and resting CD19⁺/IgG⁻ B cells. These RNAs were analyzed on the U133 Plus 2.0 high-density oligonucleotide array and 31 transcripts were shown to be increased by 2-fold or greater on all three arrays (Table I).

Confirmation of the compiled microarray results was conducted by immunoprecipitating cytoplasmic extracts from CpG-activated CD19⁺ B cells with either anti-PTB or control (IgG2_b) Abs followed by assessing the subset of identified mRNAs for specific binding. To show primer specificity, the same reactions were performed using cytoplasmic extracts with 10-fold less starting material (Table II, lane 3). PCR was continued to saturation. As shown in Table II, all transcripts tested (25 of 31 total) gave positive signals in the PTB-selected fraction compared with the Ig control fraction. To show that the positively selected transcripts represented a true subpopulation, PCR amplification of the PTB selected and control samples was conducted with primers to GAPDH and β-actin. Notably, there was no amplification of these transcripts in the selected or control populations confirming the specificity of the immunoprecipitation. These findings demonstrated that a subset of transcripts in B cells activated with CpG bound specifically to PTB and therefore were potential targets for B-cpx I binding.

Because of inherent limitations associated with obtaining appropriate cell numbers to carry out biochemical assays, we sought to identify a human B cell line that showed similar characteristics relative to B-cpx I binding of specific target RNAs. The D11-LCL^tet line is an IgM⁺ B cell line that undergoes cell division in response to LMP1 expression in the presence of tetracycline (26, 27). Using RNA binding assays we found that the D11-LCL^tet line constitutively expressed B-cpx I in the presence and absence of tetracycline however, there was an increase in complex binding when cells were activated with CpG (Fig. 2,A). To determine whether B-cpx I bound specifically to motifs within the 3′ UTR of array-identified transcripts, the 3′ UTR was scanned for potential PTB binding motifs (UCUU) that were grouped together in the context of a pyrimidine-rich tract similar to the complex I binding region in the CD154 transcript (23). Initial experiments were conducted with four transcripts, Rab8A, cyclin D2, CD20, and SET, which respectively contained five, nine, nine, and four potential 3′ UTR PTB binding sites (data not shown). Extracts from D11-LCL^tet cells were stimulated with CpG for 48 h and PTB-specific complexes were identified by the addition of anti-PTB Abs to a subset of in vitro binding reactions. Regions of each 3′ UTR that span putative PTB binding sites were used as in vitro-labeled probes. As shown in Fig. 2,B, only RNA probes generated from Rab8A and cyclin D2 bound a potential B-cpx I-like complex (Fig. 2 B, lanes 9–11 and lanes 15–17, respectively). The lack of binding to the subregions of the CD20 and the SET 3′ UTR suggested that either PTB binding occurred at sites outside the targeted regions or that PTB-specific binding to these transcripts was unstable during the RNA EMSA reaction.

To extend the analysis of B-cpx I to include functional properties similar to complex I, we monitored the patterns of Rab8A and cyclin D2 RNA stability in resting vs 48-h CpG-activated CD19⁺ B cells. Following activation, cells were incubated with DRB to arrest transcription for different times over a 4-h time course and RNA was isolated from each time point. Similar to what was observed with CD154 in late-activated CD4⁺ T cells, CpG activation of B cells was found to induce an increase in Rab8A RNA stability from a t_1/2 ∼1.3 h to ∼2.6 h. In contrast, the cyclin D2 transcript showed no activation-induced stabilization over the same time course (Fig. 3). These findings suggested that only a subset of transcripts that bind PTB within their 3′ UTRs are regulated in response to activation. One difference in the binding of PTB to the Rab8A and cyclin D2 transcripts was that the avidity of binding was much higher with the Rab8A transcript and this property may influence whether B-cpx I can stabilize a transcript (Fig. 2 B). The fact that CpG increases Rab8A transcript stability extends the level of similarity between the regulation of the Rab8A and CD154 transcripts and strongly suggests that B-cpx I is involved in modulating the turnover of the Rab8A mRNA in an activation-dependent manner.

Analysis of the 258 nt sequence within the Rab8A 3′ UTR that binds PTB revealed a CU-rich sequence of 81 nt that was highly pyrimidine rich and indicative of PTB binding sites (Fig. 4,A). To identify minimum B-cpx I binding sites within this region, RNA EMSA was conducted with a set of in vitro transcribed RNA probes. As shown in Fig. 4,B, the probes initiating with nt 1999, M (Fig. 4,B, lane 13), M3′-8 (Fig. 4,B, lane 16), M3′-29 (Fig. 4,B, lane 17), and M3′-34 (Fig. 4,B, lane 18) formed complex, whereas probes M5′-9 and M5′-21, which began 9 nt and 21 nt downstream of 1999 and terminated with nt 2079 did not. This suggested that the 5′ boundary of the binding site was between nt 1999 and 2008. The 3′ boundary required for optimal binding was located downstream of nt 2050 (Fig. 4 B, lanes 16–18) based on the decrease in complex formation with the M3′-34. Therefore, the minimal B-cpx I binding region in Rab8A mRNA was defined as a 52 nt region between nt 1999 and nt 2051.

To characterize the functional capacity of the B-cpx I binding site we established a luciferase-based expression system in which the 52-bp minimum binding site was introduced into the 3′ noncoding region of the Renilla luciferase operon in the pRLSV40 vector (termed, pRLRab52). A vector containing the 52-bp insert in the reverse orientation was used as control (termed, pRLComp). The constructs were transiently transfected into D11-LCL^tet cells, treated with CpG, and analyzed for luciferase activity 48 h later. Insertion of the Rab8A 52-bp minimum binding site increased luciferase activity ∼4-fold over the control construct suggesting that the introduced element was having an effect either at the level of message stability or translation (Fig. 5,A). Fig. 5 B shows the results of three independent experiments that analyzed decay at 0, 1, and 4 h post-addition of DRB and 48 h after transfection of the pRLRab52 and pRLComp plasmids. Renilla RNA containing the Rab52 insert was more stable than RNA containing the control insert, lending strong support for a role of the Rab52 insert in mRNA turnover.

The effect of PTB on Rab8A transcript stabilization in vivo was assessed by infecting D11-LCL^tet cells with the retroviral vector (pLVTHM-U6) constitutively expressing either shRNA against PTB (shPTB) or a control sequence containing a scrambled shRNA (shCTRL). The pLVTHM-U6 vector also expresses enhanced GFP under the control of a doxycycline-inducible promoter to allow the assessment of infection efficiency by FACS. Because the level of B-cpx I was increased when D11-LCL^tet cells were cultured in the presence of CpG (Fig. 2,A) we wanted to assess the effect of Rab8A RNA stability in these cells grown either with tetracycline or 3 μM CpG. As shown in Fig. 6,A, a significant majority of the cells were GFP-positive 4 days following infection with the packaged pLVTHM-U6 virus, and infection corresponded to a distinct down-regulation of the three isoforms of cytoplasmic PTB (PTB-1, PTB-2, and PTB-4) even after 2 days of CpG treatment (Fig. 6,B). To establish whether this level of PTB down-regulation affected Rab8A transcript stability, cells were infected with virus and grown continuously either with or without tetracycline for 4 days followed by activation of one culture with 3 μM CpG for 48 h. Both cell populations were treated with DRB over a 3-h time course, and RNA was isolated at intervening time points. Overall CpG activation of the D11-LCL^tet cells led to an increase in the half-life from ∼95 min to more than 3 h (Fig. 6, C and D, compare shCTRL curves). This increase was very similar to the increase in Rab8A stability that was observed in primary B cells (Fig. 3) and confirmed that the pathways leading to B-cpx I stability are present in the D11-LCL^tet cells. Importantly, introduction of the shPTB into either population decreased Rab8A RNA stability, although the change was much greater in cells activated with CpG where the half-life decreased from 3 h to ∼90 min (Fig. 6 D). This change closely mirrors the difference in Rab8A transcript stability in CpG vs tetracycline-treated D11-LCL^tet cells and supports a model whereby transcript stability is a direct function of PTB binding. Thus, these findings strongly support a critical role for PTB in actively stabilizing the Rab8A transcript in CpG-activated B cells. Importantly, B cells and T cells clearly maintain an activation-induced RNA stability pathway that is dependent on PTB.

Posttranscriptional regulation via RNA turnover is a critical factor that maintains appropriate levels of specific gene products during an ongoing immune response. Although ARE-mediated decay has been extensively characterized and shown to be important for the expression of different inflammatory cytokines and growth factors, other pathways of regulated decay are much less well described. RNA-focused programs of gene expression have become increasingly important areas of study, particularly with respect to how these programs interface with more well-characterized transcriptional and translational processes. Our past work defined both cis- and trans-acting factors required for an activation-dependent RNA decay program that regulates the expression of CD154 in activated CD4⁺ T cells. The goal of this current study was to establish whether B cells also use a program of PTB- and activation-dependent mRNA turnover and if they do, to identify specific transcripts that are regulated by this mechanism.

The identification and characterization of the Rab8A mRNA as a B-cpx I-regulated transcript demonstrates that an activation-regulated process involving PTB is functioning in B cells. Rab proteins are members of the Ras GTPase superfamily and are known to function in both constitutive and regulated exocytosis, endocytosis, and transcytosis (reviewed in Ref. 31). Although the exact role for Rab proteins in lymphocyte function is unknown, these processes are closely linked to lymphocyte activation, and increased expression of Rab8A would likely be required during activation-induced proliferation, motility, and Ag processing. Our current work adds Rab8A to a growing list of transcripts, including CD154 (24, 32), insulin (33), vascular endothelial growth factor (34), and inducible NO synthase (35), which are regulated by PTB. Based on the fact that one of the four transcripts tested (of 31 identified) showed PTB-dependent activation-induced turnover leads us to expect that a number of additional transcripts in B cells would also be regulated in an analogous manner. Additionally, because many more transcripts were identified on the larger array (Array no. 3) as showing enhanced PTB binding, we predict that the initial 31 transcripts identified is a conservative estimate. Although some of the transcripts identified on the arrays could merely reflect an overall increase in transcript binding that is unrelated to either B-cpx I or activation-induced decay, our expectation is that ∼25% would demonstrate an activation-linked correspondence between B-cpx I and mRNA decay. Also, as we observed with cyclin D2, some transcripts may bind B-cpx I, but do not show activation-induced stability. We hypothesize that the failure of cyclin D2 to be stabilized by B-cpx I may be related to the relatively low level of binding in CpG-activated cells (Fig. 2). This type of relationship has been demonstrated with tristetraprolin, an ARE-binding protein, and transcripts containing variations of the ARE motif. Specifically, a direct link was observed between binding affinity and mRNA decay using a number of distinct transcripts that varied in their affinity for tristetraprolin (36). In future work, we will explore the underlying differences in the Rab8A and cyclin D2 transcripts that lead to distinct modes of regulation upon activation.

PTB exists as three isoforms generated by differential splicing and the expression of each isoform can vary according to the tissue or cell type (37, 38). PTB is monomeric in the reducing intracellular environment (39) and the binding sites defined for the RNA binding domains are UCU (RBD1 and RBD4), UCNU (RBD2), and UCUCU (RBD3) (40). Sites similar if not identical to these are present within the defined pyrimidine-rich minimum binding sites for both complex I (CD154) and B-cpx I (Rab8A). In fact, the Rab8A site contains 3 UCU binding motifs in close proximity to each other. It is possible that monomeric PTB binds to these motifs, particularly the first motif given the fact that removal of the first UCU motif, as in the M5′-9 probe (Fig. 4), completely eliminates B-cpx I binding. Interestingly, the last 13 nt of the minimal binding region lack a pyrimidine-rich motif, suggesting that binding of a second RNA binding protein may be required for efficient complex formation. This possibility is supported by other work demonstrating the association of PTB with such RNA binding proteins as nucleolin (25) and cold shock domain (CSD or Y-box) proteins (34). Alternatively, the 3′ sequences may establish a secondary structure important for PTB binding.

The RNA interference results clearly demonstrate that PTB confers significant stability on the Rab8A transcript in CpG-activated D11-LCL^tet cells and supports a model whereby induction by TLR9 signaling of B-cpx I activity directly regulates message stability. Because we observe PTB in the cytoplasm of both activated and nonactivated B cells (data not shown), TLR9 signaling may induce a modification of PTB that allows it to form B-cpx I upon activation. This type of modification may be reminiscent of phosphorylation changes associated with the nucleocytoplasmic transport of PTB. Specifically, the 3,5-cAMP-dependent protein kinase A was shown to directly phosphorylate PTB on a conserved Ser¹⁶ and this modification led to the accumulation of PTB in the cytoplasm (41).

The segregation of B-cpx I activity with CpG and not CD40 activation was unexpected given that we had observed complex I activity in differentially activated CD4⁺ T cells (22). This absence of B-cpx-I activity was not due to a lack of CD40 responsiveness because signaling through either TLR9 or CD40 resulted in B cell proliferation and the expression of multiple activation markers (data not shown). The one marker that was positively associated with B-cpx I activity was up-regulation of CD20. Enhanced expression of CD20 in human B cells has been shown to be controlled by ERK-dependent mechanisms (42), suggesting that B-cpx I may also be regulated by this pathway. This possibility is consistent with our findings relative to CD40 activation because CD40-mediated ERK or MAPK activity appears to require prior stimulation through the BCR (43). In our experiments, B cells were activated through CD40 without BCR stimulation, suggesting an absence of both ERK activation and B-cpx-I activity.

In conclusion, these findings expand our understanding of regulated mRNA decay in activated lymphocytes by demonstrating a common pathway of RNA turnover in both T cells and B cells. This pathway is dependent on the expression of a cytoplasmic PTB-containing complex that stabilizes mRNAs upon cellular activation. The unexpected finding that PTB-mediated stability was dependent on specific activation programs allows us to focus future work on identifying signaling pathways that directly engage the B-cpx I activity and deciphering mechanisms that relate mRNA stability to complex formation.

We acknowledge Lindsay Ordower for expert technical assistance and thank the other members of the Covey laboratory for helpful discussions and critical comments on the manuscript. We are very grateful to Dr. Terri Goss Kinzy (Robert Wood Johnson Medical School, Piscataway, NJ) for valuable comments and criticism.

The authors have no financial conflict of interest.