Alternative Splicing Controlled by Heterogeneous Nuclear Ribonucleoprotein L Regulates Development, Proliferation, and Migration of Thymic Pre-T Cells This information is current as of October 2, 2021. Marie-Claude Gaudreau, Florian Heyd, Rachel Bastien, Brian Wilhelm and Tarik Möröy J Immunol published online 20 April 2012 http://www.jimmunol.org/content/early/2012/04/20/jimmun

ol.1103142 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2012/04/20/jimmunol.110314 Material 2.DC1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on October 2, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2012 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published April 20, 2012, doi:10.4049/jimmunol.1103142 The Journal of Immunology

Alternative Splicing Controlled by Heterogeneous Nuclear Ribonucleoprotein L Regulates Development, Proliferation, and Migration of Thymic Pre-T Cells

Marie-Claude Gaudreau,*,† Florian Heyd,‡ Rachel Bastien,* Brian Wilhelm,x and Tarik Mo¨ro¨y*,†

The regulation of posttranscriptional modifications of pre-mRNA by alternative splicing is important for cellular function, devel- opment, and immunity. The receptor tyrosine phosphatase CD45, which is expressed on all hematopoietic cells, is known for its role in the development and activation of T cells. CD45 is known to be alternatively spliced, a process that is partially regulated by heterogeneous nuclear ribonucleoprotein (hnRNP) L. To investigate the role of hnRNP L further, we have generated conditional hnRNP L knockout mice and found that LckCre-mediated deletion of hnRNP L results in a decreased thymic cellularity caused by a partial block at the transition stage between double-negative 4 and double-positive cells. In addition, hnRNP L2/2 thymocytes Downloaded from express aberrant levels of the CD45RA splice isoforms and show high levels of phosphorylated Lck at the activator tyrosine Y394, but lack phosphorylation of the inhibitory tyrosine Y505. This indicated an increased basal Lck activity and correlated with higher proliferation rates of double-negative 4 cells in hnRNP L2/2 mice. Deletion of hnRNP L also blocked the migration and egress of single-positive thymocytes to peripheral lymphoid organs in response to sphingosine-1-phosphate and the chemokines CCL21 and CXCL12 very likely as a result of aberrant splicing of genes encoding GTPase regulators and proteins affecting cytoskeletal organization. Our results indicate that hnRNP L regulates T cell differentiation and migration by regulating pre-TCR http://www.jimmunol.org/ and chemokine receptor signaling. The Journal of Immunology, 2012, 188: 000–000.

hymocytes must pass through a tightly regulated devel- by upregulating both CD4 and CD8 markers to become double- opmental process to mature into effector T cells. First, positive (DP) cells (4–9). These cells then undergo a positive and T early lymphocyte precursors migrate from bone marrow to negative selection to eliminate autoreactive T cells and to produce the peripheral blood and enter the thymus, where they undergo the final single-positive (SP) CD4- and CD8-expressing T effector differentiation (1, 2). The most immature thymocytes, termed cell populations. During these two selection processes, thymo- double negative (DN), are characterized by the absence of both cytes receive different signals from the pre-TCR, the TCR, the by guest on October 2, 2021 CD4 and CD8 coreceptor surface markers and are subdivided coreceptors, and probably other receptors that promote cell sur- based on the expression of CD44 and CD25 into four fractions vival and proliferation (4, 10–16). called DN1 to DN4 (3). In DN3 cells, b-selection takes place, One of the critical factors that regulate the strength of TCR or which assures the selection of cells with a productive rearrange- pre-TCR signaling is the transmembrane tyrosine phosphatase ment of the TCRb gene and the correct assembly of a surface pre- CD45, which is not only expressed on T cells, but is found on TCR complex. From this, DN4 cells emerge and develop further a wide variety of hematopoietic cells, except platelets and eryth- rocytes (17, 18). This protein exerts its regulatory function by modulating the activity of the src kinases Lck and Fyn (18–22). *Institut de Recherches Cliniques de Montre´al, Montre´al, Que´bec H2W 1R7, Can- Multiple isoforms of CD45 can be generated by alternative † ada; De´partement de Microbiologie et Immunologie, Universite´ de Montre´al, Mon- splicing of the variable exons 4–6, also called A, B, and C, which tre´al, Que´bec H3T 1J4, Canada; ‡Institut fu¨r Molekularbiologie und Tumorforschung, x Philipps Universita¨t Marburg, D-35032 Marburg, Germany; and Institut de code for different extracellular domains of the protein (23). The Recherches en Immunologie et Cancer, Universite´ de Montre´al, Montre´al, Que´bec expression of a specific isoform of CD45 is cell-type specific and H3C 3J7, Canada changes during thymocyte development. Immature DPs predom- Received for publication November 2, 2011. Accepted for publication March 22, inantly express CD45RO, which lacks the domains encoded by 2012. exons 4–6, whereas CD4SP or CD8SP cells express the high m.w. This work was supported by a Canadian Institutes of Health Research (CIHR) Ph.D. fellowship (to M.-C.G.), the Institut de Recherches Cliniques de Montre´al, the Ca- isoform CD45RB, which contains the domain encoded by exon 5 nadian Foundation for Innovation, CIHR Operating Grant MOP-86516, and a Canada (24–26). The smaller isoforms of CD45 form predominantly Research Chair (Tier 1; to T.M.). homodimers, whereas the high m.w. isoforms lose this potential, The sequences presented in this article have been submitted to Gene Expression resulting in a less efficient signal transduction (27). A number of Omnibus (http://www.ncbi.nlm.nih.gov/geo/) under accession number GSE33306. studies using CD45-deficient mice have demonstrated a crucial Address correspondence and reprint requests to Dr. Tarik Mo¨ro¨y, Institut de role for this protein, because its absence results in a severely Recherches Cliniques de Montre´al, 110 des Pins West Avenue, Montre´al, Que´bec H2W 1R7, Canada. E-mail address: [email protected] impaired TCR signal transduction and in a differentiation block The online version of this article contains supplemental material. during positive and negative selection that occurs during the dif- Abbreviations used in this article: CT, cycle threshold; DN, double negative; DP, ferentiation of DP thymocytes to mature SP cells (28–30). double positive; GO, gene ontology; hnRNP, heterogeneous nuclear ribonucleopro- Several lines of evidence provided by in vitro studies on al- tein; SP, single positive; S1P, sphingosine-1-phosphate; S1P1, S1P receptor 1; wt, ternative splicing of CD45 reveal the implication of heterogeneous wild-type. nuclear ribonucleoprotein (hnRNP) L in mediating this process. Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00 The hnRNP proteins belong to a family of RNA-binding factors

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1103142 2 hnRNP L IN T CELL DEVELOPMENT AND MIGRATION that regulate alternative splicing by binding exonic splicing silencer and immediately incubated at 37˚C for the indicated time. Cells were elements, resulting in exon exclusion from the mature mRNA (31). processed for surface and intracellular staining according to standard In the case of hnRNP L, its binding to the exonic splicing silencer procedure. elements 1 sequence in the alternatively spliced CD45 exons Proliferation assays results in their exclusion from the mature, spliced RNA (23, 31). For in vivo proliferation assays, mice were injected i.p. with 200 ml 10 mg/ We show in this study that deletion of hnRNP L in the mouse ml solution 5-bromo-2-deoxyuridine (Sigma-Aldrich) 16 h before sacri- results in a very early block of embryonic development, empha- fice. Thymocytes were stained with specific fluorescent Abs, fixed, and sizing its crucial role in morphogenesis. To study a role of hnRNP L treated with Perm/Wash (BD Biosciences). DNase I (Sigma-Aldrich) in T cell development and function, we have restricted hnRNP L treatment at 30 mg/ml was applied for 1 h at 37˚C. Anti-BrdU FITC conjugated (BD Biosciences) was added for 30 min at room temperature, deletion to the T cell compartment by using a T cell-specific Cre and events were acquired on a LSR (BD Biosciences). recombinase transgene (LckCre). This strategy allowed us to re- veal new and important roles of alternative splicing mediated by RT-PCR hnRNP L in T cell maturation and migration. Total RNA was extracted from thymocytes using RNeasy mini kit (Qiagen), according to the manufacturer’s instructions. For RT-PCR analysis, RNA Materials and Methods was used to prepare cDNA using SuperScript II reverse transcriptase (Invitrogen). PCR for CD45 and GAPDH were done according to previous Mice report (32). To analyze CD45 alternative splicing, an established radio- C57BL/6, LckCre transgenic, Gt(ROSA)26Sortm4(ACTB-tdTomato,-EGFP)Luo, active RT-PCR protocol was used (33). Briefly, 0.5 mg purified RNA of sorted thymic subpopulations was used for reverse transcription with a and CD45.1 mice were purchased from Jackson ImmunoResearch Labo- 32 ratories or were maintained at the Institut de Recherches Cliniques de gene-specific primer, and PCR was performed with a P-labeled forward Downloaded from Montreal. For the hnRNP L targeting construct, a phage screen was per- primer. Products were separated on a denaturing PAGE and quantified by formed that yielded a clone containing ∼18 kb of the genomic hnRNP L Phosphoimager analysis (Typhoon 9400, ImageQuant; both GE Health- locus (performed and provided by R. Waldschu¨tz, University of Duisburg- care). Essen). An 11-kb EcoRV-NdeI fragment was cloned into pcDNA3 for further manipulation. A single loxP site was inserted into the EcoRI site, Quantitative PCR and a neocassette flanked by two loxP sites was introduced into the Sfi1 DN3 and DN4 or DN, DP, CD4SP, and CD8SP thymocyte subsets were site (thereby destroying this site). This construct was found to efficiently sorted using a MoFlo (Cytomation), and RNA was extracted using RNeasy recombine in bacteria stably expressing Cre-recombinase. For homologous mini kit (Qiagen), according to the manufacturer’s instructions. cDNA was http://www.jimmunol.org/ recombination in embryonic stem cells, the targeting construct was freed prepared from RNA using SuperScript II reverse transcriptase (Invitrogen) from vector sequences by EcoRV-NdeI digestion. After transfection into and detected using TaqMan probes (Applied Biosystems) for Ptcra and embryonic stem cells, homologous recombination was confirmed by ver- Gapdh as an internal control. A Mx-3005 system (Stratagene) was used, ifying presence of the neocassette (StuI digestion) as well as the third loxP and relative expression was calculated via the 2-DD cycle threshold (CT) site (XmnI digestion) by Southern blot. A positive clone was transiently method. Exon usage of genes identified in the RNA-Seq analysis was transfected with a plasmid encoding Cre-recombinase, and partially validated using quantitative PCR. Validation was performed using two recombined clones with the neocassette removed, leaving the endogenous sets of primers spanning a variable exon junction and a constant exon locus with only two loxP sites inserted, were identified by PCR and con- junction from the same transcript. Primer sequences will be made avail- firmed by Southern blot (Bcl1/Sfi1 digestion). One clone was used to able on request. Quantitative PCR was performed using a MasterMix for generate hnRNP L-floxed mice using standard procedures. hnRNP L- SYBR Green (Quanta Biosciences) and a Mx-3005 system. Melting curve by guest on October 2, 2021 floxed mice were backcrossed to C57BL/6 for more than eight gener- was used to confirm the amplicon specificity. The exon usage was cal- ations. All animals were housed under specific pathogen-free conditions at culated via the DDCT (variable exon-constant exon) and the 2-DDCT the Institut de Recherches Cliniques de Montreal animal facilities, and all method. experiments conformed to ethical principles and guidelines approved by the institutional animal care committee. Migration assay Abs Total thymocytes were harvested and resuspended in migration media (RPMI 1640 containing 0.5% BSA) at 2.5 3 106 cells/ml. Migration assays Anti-mouse CD4 (RM4-5), CD8a (53-6.7), CD25 (PC61), CD44 (1M7), were performed in Transwell plates (Corning; 5 mM pore). Chemokines TCRg/d (GL3), TCRa/b (H57-597), CD3 (145-2C11), B220 (RA3-6B2), CXCL12 (PeproTech), CCL21 (PeproTech), and sphingosine-1-phosphate CD45RA (14.8), and CD45RB (16A) Abs and streptavidin fluorescent (S1P; Avanti Polar Lipids) were diluted in migration media at different conjugated form were purchased from BD Biosciences. The lineage concentrations and added to the lower chamber. After resting for 2 h at cocktail was made from B220, Ter119, CD3, Mac1, Gr1, NK1.1, CD49b, 37˚C, 100 ml cell suspension was added to the upper chamber and cells were TCRg/d, and CD8, all biotinylated, and were from either BD Biosciences allowed to migrate for 3 h at 37˚C. Cells were collected, stained for CD4 or eBioscience. Phospho-specific Abs were LckY505 from BD Bio- and CD8, and resuspended in 200 ml PBS supplemented with 2% FCS, and sciences, LckY394 from Santa Cruz Biotechnology, ERK and Akt from a fixed number of countbright beads (Beckman Coulter) was added. The Cell Signaling Technology. Goat anti-hamster Ig, HRP- and FITC- percentage of migrating cells was calculated by dividing the number of conjugated anti-rabbit and anti-goat were purchased from Jackson cells that migrated into lower chamber by the input cell number. ImmunoResearch Laboratories, and anti-actin was from Santa Cruz Bio- technology. Actin polymerization Immunoblotting Thymocytes were collected and resuspended in RPMI 1640 supplemented with 0.5% BSA and rest at 37˚C for 2 h. Cells were then activated with Cell lysates from equal number of cells were prepared by lysis in buffer (20 different concentrations of S1P, CXCL12, or CCL21 for 30 s at 37˚C. Cells mM Tris [pH 7.5], 150 mM NaCl, and 5 mM EDTA), supplemented with were fixed with Cytofix/Cytoperm (BD Biosciences) for 10 min on ice and complete mini protease inhibitors (Roche Applied Science), 1 mM Na3VO4, Perm/Wash (BD Biosciences). Cells were then incubated for 30 min at and 1% Nonidet P-40. Western blot analysis was done following SDS room temperature with 1 U phalloidin-Alexa488 (Invitrogen), followed by polyacrylamide electrophoresis and transferred onto nitrocellulose mem- surface staining for CD4 and CD8. Events were acquired on a LSR (BD branes (GE Healthcare). General and specific tyrosine-phosphorylated Biosciences). Relative F-actin polymerization was calculated as the ratio proteins were detected with Abs, followed by goat anti-rabbit or donkey of mean fluorescence of phalloidin-Alexa488 of activated cells over non- anti-goat–coupled HRP. All immunoblots were visualized with ECL stimulated cells. chemiluminescent (Thermo Scientific) detection system, and images were taken on film. Recent thymic emigrants Cell stimulation Four- to 5-wk-old mice were anesthetized with isofluorane. The shaved skin at the thoracic inlet was cut open by a 1-cm incision, one-third of the Thymocytes were preincubated with 10 mg/ml anti-CD3 (2C11) for 10 min sternum was longitudinally bisected, and the thymus was exposed. Each on ice. Bound Abs were cross-linked with goat anti-hamster Ig at 20 mg/ml thymic lobe was injected with 10 ml FITC (Sigma-Aldrich) dissolved in The Journal of Immunology 3

PBS at 10 mg/ml, and the wound was closed by suture. Mice were sac- mus (Fig. 1A, 1B). However, loss of hnRNP L protein led to a rificed 4 d later. decrease in overall thymic cellularity by 75% compared with wt RNA-Seq mice (Fig. 1C). Without affecting the absolute DN cell number, deletion of hnRNP L resulted in the loss of DP thymocytes and the RNA was extracted from total thymocytes using Tri Reagent (Molecular more mature CD4 and CD8 SP cells in about the same proportions Research Center), followed by a purification using RNeasy mini kit and + RNase-free DNase on column (Qiagen) for 15 min at room temperature, (Fig. 1C). In addition, the loss of thymic cellularity in LckCre both following manufacturer’s instructions. RNA integrity and quality have hnRNP Lfl/fl mice was specific to TCRa/b cells and did not affect been confirmed using a bioanalyzer (Agilent). rRNA from each biological TCRg/d cells because their absolute number remained unchanged sample was depleted from total RNA using a RiboMinus kit (Invitrogen), (Fig. 1D, 1E). Consistent with this, the DN4 population of and the treated RNA was then fragmented using RNase III. Ligation of the + fl/fl adaptor mix A and reverse transcription were performed following the LckCre hnRNP L mice had a reduced frequency of cells manufacturer’s protocol. Libraries were size selected for fragments be- expressing surface TCRb, but all hnRNP L-deficient DN4s tween 150 and 300 bp, amplified for 12 cycles of PCR, and purified using expressed the cytoplasmic form of TCRb. In addition, the ex- PureLink PCR micro kit (Invitrogen). Barcoded library concentrations pression of Ptcra was not disturbed by the ablation of hnRNP L, were determined by quantitative PCR using a standard curve of template at suggesting that a pre-TCR can be made in hnRNP L-deficient known concentrations (DH10B), and ∼0.25 ng each library was used for each full-emulsion PCR (4 full-emulsion PCR/sample). Approximately mice (Fig. 1F, 1G). 200 million beads from each two samples were deposited on single full To test whether the reduced thymic cellularity of LckCre+hnRNP slides (2 slides for 4 samples in total) and sequenced using the Opti Lfl/fl mice could be due to an effect of hnRNP L deletion in thymic Fragment Library Sequencing kit-Master Mix 50 on a SOLiD machine epithelial cells, we generated mixed bone marrow chimeras. (version 3+). Mapping of DNA sequence reads was performed using Congenic CD45.1+ recipient mice that were reconstituted with

Bioscope (version 1.3) against both the mouse genome (mm9) and all Downloaded from + fl/fl + annotated splice junctions to calculate gene expression and splice site bone marrow cells from LckCre hnRNP L mice (CD45.2 ) usage counts. RNA-Seq data have been deposited in the public database showed the same defect in thymocyte development as was ob- Gene Expression Omnibus (National Center for Biotechnology Informa- served in the original hnRNP L-deficient mice (data not shown), tion; accession number GSE33306;http://www.ncbi.nlm.nih.gov/geo/query/ suggesting that the phenotype observed upon hnRNP L deletion acc.cgi?token=pnwhfuowmqwiqto&acc=GSE33306). is cell autonomous. Statistical analysis To discriminate between cells before and after positive and http://www.jimmunol.org/ Two-tailed Student t test was used to calculate p values, where indicated. A negative selection, we stained thymocytes from both wt and hnRNP p value #0.05 was considered statistically significant, as follows: *p # L-deficient mice for CD4, CD8, CD69, and TCRb expression, 0.05, **p # 0.01, ***p # 0.001. Two-way ANOVAwas used to calculated which enables the definition of four maturation stages (35). This p value for the migration assay and the actin polymerization. analysis showed that the relative percentages of the mature sub- sets as defined by CD69 and TCRb expression were not different Results with regard to the relative frequency of cells between wt and Constitutive deletion of hnRNP L impedes early embryonic hnRNP L-deficient mice. In this analysis, CD4 cells appear before development CD8 cells, and this sequential appearance is also unaffected by hnRNP L deletion (Supplemental Fig. 2A, 2B). In addition, to To be able to analyze the consequences of a constitutive or in- by guest on October 2, 2021 ducible ablation of hnRNP L for the immune system and hema- confirm whether the absence of hnRNP L affects positive se- topoietic differentiation, we used a previously published strategy lection, we introduced the OTI or OTII TCR transgene into + fl/fl (34) to generate mice carrying loxP (fl) sites flanking coding LckCre hnRNP L mice. We observed that CD8SP cells or + fl/fl exons 2–6 of the hnRNP L gene (Supplemental Fig. 1A, 1B). CD4SP cells were formed in OTI tg LckCre hnRNP L and in + fl/fl Animals carrying two hnRNP L-floxed alleles (hnRNP Lfl/fl) OTII tg LckCre hnRNP L mice, respectively, at a similar rate were bred with germline deleter strains (CMV Cre) to generate than in controls (Supplemental Fig. 2C, 2D). This suggests that constitutively hnRNP L-deficient mice. However, live pups or there are no major defects in either positive or negative selection embryos beyond stage 3.5 days postcoitus could not be gener- or in CD4/CD8 lineage choice due to the loss of hnRNP L. ated with this strategy (data not shown), suggesting that the hnRNP L regulates the splicing of CD45 in vivo ablation of hnRNP L is lethal at a very early stage of murine development. Because it has been suggested that hnRNP L regulates the al- To overcome this problem and because RT-PCR analyses re- ternative splicing of CD45 in vitro (23, 31), we tested whether the vealed expression of hnRNP L in a wide range of hematopoietic expression of this transmembrane phosphatase was affected by cells, including lymphoid progenitors and thymocyte subsets hnRNP L ablation in vivo. It has been described that CD45RO and (Supplemental Fig. 1C), we chose to restrict our analysis to the CD45RB isoforms are expressed in thymocytes, whereas the T cell compartment. To determine the function of hnRNP L in CD45RABC isoform is absent in these cells (26). We detected an 2/2 T lymphocytes, we deleted the hnRNP L-floxed alleles using mice increased expression of CD45RB on hnRNP L DP and CD8SP expressing Cre recombinase driven by the Lck proximal promoter. cells compared with the respective wt controls and found the In these mice, Cre expression starts at the pro-T cell DN2/DN3 expression of CD45RA upregulated 4-fold over the wt in all 2/2 stages and is most efficient at and after the DN4 stage. As ex- hnRNP L thymocyte subsets except DN3 (Fig. 2A, 2B). An pected, the floxed alleles in LckCre, hnRNP Lfl/fl mice efficiently assessment of the mRNA expression level of all possible iso- deleted in pro-T cells, and hnRNP L protein expression is almost forms of CD45 by RT-PCR indicated that the mRNAs for the eliminated in the thymus of these animals (Supplemental Fig. 1D, CD45RABC isoform that is absent from wt thymus are now 1E). present in hnRNP L-deficient DN4 and CD8SP cells and to a lesser extent also in DP and CD4SP cells compared with their Loss of hnRNP L expression affects thymic cellularity in respective controls (Fig. 2C). In addition, the same analysis also a cell-autonomous manner suggested that the mRNA encoding CD45RB and CD45RBC Flow cytometric analysis revealed a normal frequency of the are enriched in hnRNP L2/2 CD4 and DP cells over wt controls different thymocyte subsets in LckCre+hnRNP Lfl/fl mice with (Fig. 2C). A more quantitative RT-PCR method using the incorpo- a slight increase of DN cells compared with wild-type (wt) thy- ration of radioactive nucleotides confirmed an enrichment of the 4 hnRNP L IN T CELL DEVELOPMENT AND MIGRATION

FIGURE 1. Deletion of hnRNP L and its effect on thymocyte pop- ulations. Flow cytometric analysis of wt and hnRNP L-deficient thymocyte populations assessed by Lin2/CD44/ CD25 (A) and CD4/CD8 staining (B). (C) Histograms representing the absolute cell number of total thymus and different thymocyte subsets from LckCre+hnRNP Lfl/fl and wt mice at 6 wk of age. A minimum of six mice per group was analyzed. Flow cytome- Downloaded from try (D) and absolute cell number (E) of TCRa/b and TCRg/d thymocyte populations. (F) Intracellular and sur- face expression of TCRb by flow cytometry performed on DN3- and DN4-gated cells. The plot is repre- sentative for three independent anal- http://www.jimmunol.org/ yses. (G) Expression of Ptcra mea- sured by quantitative RT-PCR on DN3 and DN4 cells sorted by flow cytometry. Expression was measured and normalized to the expression of the Gapdh gene and is presented as the fold increase relative to wt cells (set as 1-fold). Data represent three

independent experiments, each done in by guest on October 2, 2021 triplicate. All error bars are means 6 SEM (*p , 0.01, **p , 0.001, ***p , 0.0001).

CD45RB mRNA over the CD45ROD7 mRNA isoforms in DN4 LckCre+hnRNP Lfl/fl thymocytes, in particular DN4 cells, and DP cells (Fig. 2D). These data indicate that absence of expressed aberrant level of the larger CD45 isoforms, we tested hnRNP L leads to an enrichment of mRNA isoforms that encode whether this affected the phosphorylation status of LckY505. We the larger isoforms of CD45, which is consistent with the role of found a lower level of Lck phosphorylation at the inhibitory Y505 hnRNP L as one of the proteins that regulate the exclusion of in hnRNP L-deficient DN4 cells both at steady state and after anti- CD45 exons. CD3 activation (Fig. 3B, 3C). Interestingly, the phosphorylation of the activating tyrosine Y394 of Lck as well as the downstream hnRNP L modulates the TCR signal in thymocytes tyrosine kinases ERK and Akt was increased in these DN4 cells It is known that CD45 is a tyrosine phosphatase that controls the over wt controls at steady state levels (Fig. 3C, 3D). These results duration of pre-TCR or TCR signaling by removing the inhibitory suggest that hnRNP L controls TCR signaling intensity through phosphate group at tyrosine (Y) 505 in the Lck kinase that is the alternative splicing of CD45. associated with the cytoplasmic domains of the TCR complex (22, 36). The removal of the Y505 phosphate enables Lck to phos- Deletion of hnRNP L in thymocytes results in aberrant cell phorylate its substrate, the ZAP70 kinase (Fig. 3A). It has been proliferation shown that larger CD45 isoforms are more efficient phosphatases Next, we assessed whether the aberrant expression of CD45 iso- than the smaller variants, such as CD45RO, probably owing to forms and the ensuing deregulation of Lck phosphorylation had any their different potential to form homodimers (27). Because the effect on cell proliferation or cell death, which could explain the The Journal of Immunology 5

FIGURE 2. hnRNP L regulates the alternative splicing of CD45. Flow cytometric histograms and bar graphs represent the mean fluores- A cence of CD45RB ( ) and CD45RA Downloaded from (B) on different population of thy- mocytes from hnRNP L-deficient (black) or control mice (gray). RNA was extracted from sorted DN4, DP, CD4SP, and CD8SP from wt and LckCre+hnRNP Lfl/fl and RT-PCR

detection of the different isoforms of http://www.jimmunol.org/ CD45 mRNA and GAPDH control on an agarose gel (C), or (D) radio- active RT-PCR was performed with a gene-specific primer and a 32P-la- beled forward primer. Products were separated on a denaturing PAGE and quantified by Phosphoimager anal- ysis. Ratio was calculated between the isoform containing both exons (CD45RB) and the isoform contain- by guest on October 2, 2021 ing neither (CD45R0DE7) and com- pared between the different thymic subpopulations (n = 3). All error bars are means 6 SEM (*p , 0.01, **p , 0.001) from at least three independent experiments.

lower thymic cellularity in hnRNP L2/2 mice. Staining with apoptotic, but have shortened cell cycle phases resulting in higher annexin V for apoptotic cells did not show any differences between proliferation in stages following the b-selection checkpoint. hnRNP L-deficient pre-T cell subsets compared with wt controls Deletion of hnRNP L results in loss of peripheral T cells (Fig. 3E). However, in vivo BrdU pulse labeling clearly indicated We next investigated whether hnRNP L deletion and the differential that DN4 and CD4SP cells from hnRNP L mice progressed faster expression of CD45 isoforms affected CD4 or CD8 SP cells. We through S phase than their wt counterparts (Fig. 3F). This result noted that the frequency of peripheral CD3-positive cells was indicated that thymocytes from LckCre+hnRNP Lfl/fl mice are not considerably reduced in blood and spleen of LckCre+hnRNP Lfl/fl 6 hnRNP L IN T CELL DEVELOPMENT AND MIGRATION Downloaded from http://www.jimmunol.org/ by guest on October 2, 2021

FIGURE 3. Loss of hnRNP L in thymus does not affect cell death, but causes aberrant TCR activation. (A) Schematic representation of CD45 phos- phatase activity on Lck following TCR activation. (B) Phosphorylation levels of Lck in total thymus from the indicated mice were assessed by Western blot or (C) by flow cytometry on gated DN4 cells using Abs specifically recognizing LckY505 or LckY394 at basal level and following 2-min anti-CD3 stimulation (gray = isotype control, black = nonstimulated, dotted line = 2-min CD3 activation). (D) Basal phosphorylation of ERK and Akt in wt and hnRNP L-deficient DN4-gated cells by flow cytometry analysis. Data are representative of three independent experiments. (E) Flow cytometric analysis of cell death by annexin V staining on different thymocyte subpopulations and bar graph representing the percentage of annexin V-positive cells. (F) Pro- liferation was assessed by incorporation of BrdU in all thymocyte subsets after 16 h postadministration and quantified by flow cytometry. All graphs represent the mean 6 SEM from n = 6 mice (*p , 0.01). mice compared with wt animals and that this reduced frequency counterselected for complete excision, or that CD4SP or CD8SP corresponded to a significantly reduced absolute number of pe- thymocytes that have completely excised hnRNP L are not mi- ripheral CD3+ T cells (Fig. 4A, 4B). Importantly, the loss of pe- grating to the periphery. To investigate this hypothesis further, we ripheral T cells affected both CD4 and CD8 cells equally (Fig. quantified recent thymic emigrants by injecting FITC into the 4B). A genotyping PCR analysis showed the presence of both thymus of wt or LckCre+hnRNP Lfl/fl mice. Four days after in- floxed and excised hnRNP L alleles in sorted peripheral CD3+ jection, the mice were sacrificed and blood and spleen were an- T cells of LckCre+hnRNP Lfl/fl mice, although only fully excised alyzed for the presence of FITC-labeled CD4 or CD8 T cells by alleles were found in CD4SP and CD8SP subsets in the thymus FACS. A quantification showed clearly that FITC-labeled cells (Supplemental Fig. 1D). This suggests that peripheral T cells are were significantly reduced in the periphery of hnRNP L-deficient The Journal of Immunology 7

FIGURE 4. hnRNP L-deficient mice show a loss of peripheral T cells. (A) Flow cytometric analysis of blood and splenic T cells assessed by CD3 and CD4/CD8 staining from both control and hnRNP L-deficient mice. (B) Absolute cell numbers of splenic CD4 and CD8 T cell subsets. (C) Excision Downloaded from of floxed hnRNP L alleles visualized by PCR on sorted CD3-positive cells from blood and spleen from LckCre+hnRNP Lfl/fl and control mice. (D) Recent thymic emigrant assay. Briefly, 10 mg FITC was injected intrathymically, and mice were sac- rificed 4 d later. Blood and spleen were harvested http://www.jimmunol.org/ and analyzed by flow cytometry for side/forward light scatter and FITC expression on gated CD4 and CD8 cells. Graphs show the numbers of CD4 and CD8 FITC-positive cells in blood and spleen. (E) ROSAfltomatofl-GFP mice were crossed with LckCre+hnRNP Lwt/wt or LckCre+hnRNP Lfl/fl. Cells expressing the LckCre transgene delete the floxed tomato allele and switch the expression from tomato (red) to GFP (green), allowing to track cells with active Cre deletion. The numbers by guest on October 2, 2021 of GFP-positive cells were assessed in blood and spleen by flow cytometry after staining for CD4 and CD8 on CD3-gated cells. Absolute cell num- bers are shown in the bar graphs. All error bars are means 6 SEM (*p , 0.01, **p , 0.001), and a minimum of six mice was used in all experi- ments.

mice over wt controls (Fig. 4D), suggesting a reduction of thymic LckCre+hnRNP Lfl/fl and ROSAfltomatofl-GFP LckCre+hnRNP Lwt/wt egress caused by the deletion of hnRNP L. mice had 95–98% GFP-positive thymocytes (data not shown). We Next, LckCre+hnRNP Lfl/fl and LckCre+hnRNP Lwt/wt were found 73–85% GFP-positive T cells (CD8+ or CD4+) in blood or crossed with Gt(ROSA)26Sortm4(ACTB-tdTomato,-EGFP)Luo mice spleen in ROSAfltomatofl-GFP LckCre+hnRNP Lwt/wt mice, indicat- (hereafter ROSAfltomatofl-GFP). These animals allow the tracking of ing efficient deletion by the Lck promoter-driven Cre recom- cells by measuring red fluorescence (tomato) and Cre activity by binase (Fig. 4E). However, T cells from blood or spleen of measuring green fluorescence (GFP). Both ROSAfltomatofl-GFP ROSAfltomatofl-GFP LckCre+hnRNP Lfl/fl mice were mostly tomato 8 hnRNP L IN T CELL DEVELOPMENT AND MIGRATION positive (81–93%), and only very few (0.2–2.5%) were GFP hnRNP L regulates the migration of thymocytes positive (Fig. 4E), indicating that cells that have deleted the tomato It is known that the motility of pre-T cells within the thymus and marker in the thymus, and by inference also the floxed hnRNP L their egress into the bloodstream are tightly regulated by different + alleles and are thus GFP , do not leave the thymus or fail to mi- chemokines such as CCL21, CXCL12, or S1P (37, 38). Because grate into the blood and settle in the peripheral lymphoid organs. the previous experiments suggested an impaired motility or egress Downloaded from http://www.jimmunol.org/ by guest on October 2, 2021

FIGURE 5. hnRNP L-deleted thymocytes show defects in chemotaxis. (A) Cell surface expression of chemokine receptors CXCR4, CCR7, and S1P1 on CD4- and CD8-gated thymocytes by flow cytometry. Histograms show cells from wt (gray) or LckCre+hnRNP Lfl/fl (black) mice, and dashed line the isotype control. (B) Transwell migration assay: chemotactic response of CD4SP and CD8SP to CXCL12, CCL21, and S1P after 3-h incubation depicted as the percentage 6 SEM of cells that migrated into the lower chamber normalized to the corresponding input cells. At least three different mice were used for each measurement. (C) hnRNP L-deficient CD4SP and CD8SP cells have reduced actin polymerization compared with wt control cells. To detect F-actin, cells were stained with phalloidin-Alexa488 and with Abs for CD4 or CD8 after 30-s stimulation with different concentrations of CXCL12, CCL21, or S1P. Cells were analyzed by flow cytometry. Relative F-actin content was calculated as the mean fluorescence intensity of phalloidin-Alexa488 after stimulation over nonstimulated cells. Graphs show means 6 SEM of at least three different mice. The Journal of Immunology 9 of hnRNP L-deficient thymic T cells, we tested the chemotactic signals from the chemokine receptor to actin polymerization, response of hnRNP L2/2 thymocytes toward these chemokines. which is required for cell migration (Fig. 5C). The expression of CXCR4, CCR7, and S1P receptor 1 (S1P1), which are the receptors for CXCL12, CCL21, and S1P, respec- hnRNP L regulates alternative splicing of other gene targets tively, was not altered in hnRNP L-deficient thymocytes, exclud- The aberrant splicing of CD45 in hnRNP L-deficient thymocytes ing a role of hnRNP L in the regulation of chemokine receptor may not account for all phenotypes observed in this mouse mutant. expression (Fig. 5A). However, a Transwell assay, which allows To gain more insight into the global effects of hnRNP L deletion, a quantification of thymocyte migration toward chemokine- we undertook a genome-wide analysis of all splicing events enriched media, demonstrated that hnRNP L2/2 CD4SP and through next-generation RNA sequencing of thymus from wt and CD8SP thymocytes migrated poorly toward the chemokines LckCre+hnRNP Lfl/fl mice. The RNA-Seq analysis did not reveal compared with wt cells (Fig. 5B). In addition, hnRNP L-deleted any major changes in the expression level of genes between cells were deficient compared with wt controls to polymerize actin hnRNP L-deficient thymocytes and controls (Supplemental Fig. upon chemokine treatment, suggesting a defect in the relay of 3A, 3B); hence, the differences between the two samples are thus Downloaded from http://www.jimmunol.org/ by guest on October 2, 2021

FIGURE 6. RNA-Seq analysis of splicing events affected by hnRNP L ablation. (A) Scatterplot of all junction-spanning sequence reads from either normal control thymocytes (WT) or hnRNP L-deficient thymocytes (KO), in which the sum of one or the other condition (or both) is at least 1. Green lines indicate a 62-fold difference between WT and KO samples, and blue dotted lines at 5 represent a minimum of 32 (25) reads spanning either WT or KO junctions. Differentially spliced junctions with low absolute counts (below the blue lines) may be valid, but were not considered in the analysis to avoid inclusion of junctions resulting from sampling noise. (B) For differential splice junction usage, two different approaches were assessed. First, splice junctions were filtered based on those that had at least 10 reads in both WT and KO samples and were then sorted based on the ratio of reads in KO versus WT. Second, the threshold of reads in either KO or WT samples was set to 32, and genes were then sorted based on a KO versus WT ratio. The Venn diagram shows the ∼80% overlap between both approaches, indicating robust detection of differentially spliced junctions. For gene ontology (GO) pathway analysis, gene lists were generated from data using former threshold (32 reads). (C) Gene ontology categories enriched in a set of 250 genes with differential exon junction usage between WT and hnRNP L-deficient thymocytes. (D) Quantitative PCR validation of the alternate exon usage. Two sets of primers were designed for each gene, one corresponding to a constant exon junction and a second specific for an exon junction that showed alternate usage between WT and hnRNP L- deficient thymocytes in the RNA-Seq analysis. Quantification was calculated using the DD method with the constant junction used as the normalizer and is presented as the fold increase relative to the DN subset from WT cells (set as 1-fold). Bar graphs are representative of three independent experiments. 10 hnRNP L IN T CELL DEVELOPMENT AND MIGRATION due to differential splicing. To survey the differences in splicing total thymocytes will be similar to the exon usage measured in the between control and hnRNP L-deficient thymocytes, sequence DP subset. Therefore, the differential splicing of all four tested reads were mapped against all possible annotated splice junction genes could be confirmed in the DPs (Fig. 6D). Interestingly, both combinations within a gene locus, regardless of prior evidence of Diap1 and Wasf2 also showed alternative splicing in the absence their usage in vivo. The total number of spliced reads in the sum of of hnRNP L in the CD4SP and CD8SP subpopulation, which the two biological replicates was scaled using all mapped se- supports the notion that these genes are involved in the migration quence reads to account for differences in sequencing depth. phenotype observed in hnRNP L-deficient cells (Fig. 6D). Scatterplots comparing exon junction usage between wt and hnRNP L2/2 cells were generated with evidence of junction usage Discussion in at least one cell type as a threshold for inclusion. Green diag- In this study, we provide evidence for an important role for the onal lines indicate a 62-fold difference between wt and hnRNP splicing factor hnRNP L in the development, intrathymic migra- L2/2, and blue dotted lines on either axis represent a minimum of tion, and thymic egress of pre-T cells. In particular, our study shows 32 (25) reads spanning junction for either wt or hnRNP L-deleted that hnRNP L regulates the alternative splicing of CD45 in vivo, as cells (Fig. 6A). To avoid evidence from low confidence junctions, was suspected from previous in vitro studies, but also indicates that gene lists were generated from data over the blue dotted lines hnRNP L controls the alternative splicing of hundreds of other (e.g., $32 junction-spanning reads) for gene ontology (GO)/An- genes implicated in early T cell differentiation and migration. notation pathway analysis (Supplemental Table I). Alternative CD45 expression is crucial for T cell development as well as methodologies of filtering splice junctions yielded similar results in controlling the threshold of Ag-mediated activation in periph- (Fig. 6B), but were not used. eral T lymphocytes (22, 29, 39, 40). Larger CD45 isoforms (e.g., The gene lists were analyzed for enrichment of genes of specific CD45RABC, -RAB, -RBC, or -RB) facilitate the activation Downloaded from functions using the DAVID tool (http://david.abcc.ncifcrf.gov/ through the TCR, whereas T cells expressing CD45RO need summary.jsp). The most significant annotation cluster generated a stronger signal to obtain the same level of activation (27, 41). for hnRNP L-deficient cells was for genes involved in cyto- Our observation that both the basal and TCR-induced inhibitory skeleton/microtubule organization, and no elements of this an- phosphorylation of Lck at tyrosine Y505 is reduced in LckCre+ notation cluster were contained in analysis of the wt cells (Fig. hnRNP Lfl/fl DN4 pre-T cells is in agreement with this, because we 6C). This cytoskeleton/microtubule category includes genes such show that hnRNP L-deficient cells express higher CD45RB and http://www.jimmunol.org/ as Wasf2, Diap1, and Arhgef2 (Table I). Another GO category that RA levels than wt cells. This is also consistent with the increased is significantly enriched in hnRNP L2/2 thymocytes concerns basal-activating phosphorylation of Lck at Y394 in hnRNP L- genes involved in G protein-mediated signaling (Fig. 6C). This deficient cells compared with wt controls. It is known that Lck suggests that hnRNP L may affect thymocyte egress by regulating mediates activation of T cells through recruitment and phos- chemokine receptors that are implicated in cell migration and bind phorylation of its substrates, such as ZAP70, but also activates GTPases and GTPase regulators like Dock9, Arhgef2, and Arh- PI3K (42–45). Hence, the higher basal Akt and ERK phosphor- gap17 (Fig. 6C, Table I). Most importantly, the analysis also ylation and related increased proliferation rate observed in hnRNP identified the gene for CD45, Ptprc, as alternatively spliced with L-deficient DN4 cells are likely to be a consequence of the ab- by guest on October 2, 2021 high significance. This confirms and validates our results from RT- errant Lck signaling in these cells. We conclude that alternative PCR and semiquantitative radioactive PCR analysis of sorted DN4 splicing of CD45 is regulated by hnRNP L and represents a cells (Fig. 2, Table I). In addition, to determine whether the exon mechanism to control pre-TCR signaling in DN4 cells, and thus, usage results obtained with total thymocytes are also true for each the early steps of pre-T cell differentiation at the DN to DP thymic subset, we validated the exon usage of four genes (Diap1, transition. However, in absence of hnRNP L, thymocytes prolif- Wasf2, Dock9, and Arhgef2) that were identified to be alternatively erate in a higher rate than wt cells, but their absolute number is spliced in absence of hnRNP L by RNA-Seq (Supplemental Fig. still reduced without showing an increased apoptosis or necrosis. 3C). This validation was done by quantitative PCR on FACS- Further experimentation is needed to clarify this phenomenon. In sorted thymocyte subsets (i.e., DN, DP, CD8SP, and CD4SP). this study, we can only suggest that either the hnRNP L-deficient Alternate exon usage was measured by comparing the expression thymocytes are dying from an unidentified mechanism, the dying of the variable exon junction with the expression of a C region of cells are not detectable by either propidium iodide or annexin V the same transcript. Because the DP subset represents 85% of all staining, or the increased BrdU incoporation seen in absence of thymocytes, it was predictable that the exon usage identified with hnRNP L is due to an increased rate of DNA repair. Recently, new studies suggested that instead of hnRNP L, the related protein hnRNP L-like (hnRNP LL) plays a major role in Table I. Selected genes that contain exon(s) showing preferential controlling the alternative splicing of CD45 (46, 47). Both ENU- splicing in hnRNP L-deleted thymocytes induced mutations in the hnRNP LL gene and shRNA knockdown showed the same shift in CD45 isoform expression as our hnRNP Genes Gene Description L-deficient model. It is possible that hnRNP L and hnRNP LL Ptprc Protein tyrosine phosphatase, receptor type C (CD45) may have overlapping functions, but with different mRNA- Diap1 Diaphanous homolog 1, related formin binding requirements. Alternatively, hnRNP L may be mediating Wasf2 WAS protein family member 2 Arhgef2 rho/rac guanine nucleotide exchange factor 2 the basal CD45 alternative splicing, whereas hnRNP LL may be Dock9 Dedicator of cytokinesis 9 necessary in cells that receive a TCR signal to induce exon Cd8b1 CD8 Ag b-chain 1 skipping (46–48). Further investigation has to clarify the shared or Akt3 Thymoma viral proto-oncogene 3 specific role of both proteins in TCR signaling and CD45 alter- Tln1 Talin 1 native splicing. Arhgap17 Rho GTPase-activating protein 17 Bcl10 B cell leukemia/lymphoma 10 Our observation that hnRNP L deficiency correlates with a loss Hdac4 Histone deacetylase 4 of peripheral T cells was unexpected and suggested a role of al- Tbk1 TANK-binding kinase 1 ternative splicing in thymic egress or the intrathymic migration of Fn1 Fibronectin 1 pre-T cells. Indeed, a number of experiments support this view. The Journal of Immunology 11

First, our Transwell assays showed that hnRNP L-deficient CD4SP be shown how the ablation of hnRNP L leads to a loss or alteration and CD8SP cells failed to adequately respond to chemotactic of mDia1 function, but it can be speculated that hnRNP L does not signals provided by CCL21, CXCL12, or S1P, although they ex- necessarily need to affect the protein expression of the target gene press normal levels of the respective CCR7, CXCR4, or S1P1 candidates identified in this work. Altered mRNA isoforms gen- receptors. Second, hnRNP L-deficient mice show very low yields of erated by hnRNP L-controlled alternative splicing might either labeled T cells in the periphery after intrathymic injection of FITC. encode proteins that undergo differential posttranslational modi- Third, almost no peripheral T cells in LckCre+hnRNP Lfl/fl mice fication, lose their binding capacity to specific interaction partners, show complete excision, which have been verified by both geno- or produce truncated proteins with altered functions. typing PCR and GFP expression in the ROSAfltomatofl-GFP LckCre+ Our study not only reveals a new, critical role of the splicing hnRNP Lfl/fl mice. This suggests that thymic egress or intrathymic factor hnRNP L in the differentiation and migration or thymic migration selects against cells that lack hnRNP L. egress of pre-T cell, it also illustrates to which extent alternative Stage-dependent migration of thymocytes is important to obtain splicing serves as a regulatory mechanism of biological processes the right signal at the right time (37). It has been shown in Jurkat at a posttranscriptional level. The fine-tuning of CD45 isoform cell lines that CD45 and CXCR4 colocalize to mediate signal expression at differential stages of thymocyte development is one transduction following CXCL12 treatment (49). Furthermore, example of how alternative splicing regulates the generation of a recent study linked CD45 expression in DN1 cells with the a functional T cell-mediated immune response. The regulation of migration toward CXCL12 (50). Absence of CD45 results in a alternative splicing of genes involved in relaying signals from deficiency in CXCL12-induced migration of DN1 cells in the chemokine receptors to F-actin polymerization is another example. cortex, which is opposite of our model because CD45 is present on Our findings link alternative splicing controlled by hnRNP L to cell surface. Finally, it has been shown that ablation of SC35, define cellular processes in the acquired immune system, and thus, Downloaded from which also regulates alternative splicing of CD45, correlates with ascribe this factor important new biological roles. a reduced number of peripheral CD4 and CD8 T cells in the spleen. However, in contrast to our findings, the CD45RO isoform Acknowledgments was predominant and CD45RB was undetectable in SC35- We thank Me´lanie St.-Germain, Marie-Claude Lavalle´e, and Caroline 2/2 deficient mice (51). Although our hnRNP L -deficient SP Dube´ for the excellent animal care, Eric Massicotte, Martine Dupuis, cells show a similar migration defect as SC35-deficient animals, and Julie Lord for precious help in the flow cytometry core, and the http://www.jimmunol.org/ higher m.w. isoforms of CD45 are upregulated in the absence of genomics core facility at Institut de Recherches en Immunologie et Cancer hnRNP L, which differs from the situation seen in SC35-deficient for technical assistance with sequencing. We are grateful to Mathieu mice. These observations suggest that aberrant CD45 alternative Lapointe and Karina Savoie for technical assistance. splicing may not account for the migratory deficiency seen in hnRNP L-deficient mice. Disclosures One of the first steps in the chemotactic response after binding The authors have no financial conflicts of interest. of the chemokine to its receptor is the activation of GTPase- dependent events and the polymerization of F-actin. Rho-GTPases by guest on October 2, 2021 such as cdc42 and Rac are converted upon receptor activation References from the inactive form to the active GTP-bound form to control F- 1. Benz, C., V. C. Martins, F. Radtke, and C. C. Bleul. 2008. The stream of pre- cursors that colonizes the thymus proceeds selectively through the early T lin- actin polymerization through interaction with Arp2/3, WASP, and eage precursor stage of T cell development. J. Exp. Med. 205: 1187–1199. the Wave complex (52, 53). Our evidence suggests that hnRNP L 2. Allman, D., A. Sambandam, S. Kim, J. P. Miller, A. Pagan, D. Well, A. Meraz, and A. Bhandoola. 2003. Thymopoiesis independent of common lymphoid controls the steps between chemokine receptor engagement and progenitors. Nat. Immunol. 4: 168–174. F-actin polymerization, and thereby affects cell motility, leading 3. Godfrey, D. I., J. Kennedy, T. Suda, and A. Zlotnik. 1993. A developmental to the phenotype observed in hnRNP L-deficient mice. Two sets pathway involving four phenotypically and functionally distinct subsets of CD32CD42CD82 triple-negative adult mouse thymocytes defined by CD44 of data support this notion, as follows: first, we show that F-actin and CD25 expression. J. Immunol. 150: 4244–4252. polymerization is indeed disturbed by the ablation of hnRNP L 4. Baldwin, T. A., M. M. Sandau, S. C. Jameson, and K. A. Hogquist. 2005. The in CD4SP and CD8SP thymic cells. Second our next-generation timing of TCR alpha expression critically influences T cell development and selection. J. Exp. Med. 202: 111–121. RNA sequencing data identified different alternatively spliced 5. Dudley, E. C., H. T. Petrie, L. M. Shah, M. J. Owen, and A. C. Hayday. 1994. RNAs belonging to GO categories, such as GTPase binding and T cell receptor beta chain gene rearrangement and selection during thymocyte development in adult mice. Immunity 1: 83–93. regulation and cytoskeletal proteins, among them the genes Arh- 6. Gibbons, D., N. C. Douglas, D. F. Barber, Q. Liu, R. Sullo, L. Geng, gap17, Arhgef1, GEF-H1, Wave2, and Diap1. H. J. Fehling, H. von Boehmer, and A. C. Hayday. 2001. The biological activity Previous studies have already shown the importance of these of natural and mutant pTalpha alleles. J. Exp. Med. 194: 695–703. 7. Hayday, A. C., and D. J. Pennington. 2007. Key factors in the organized chaos of molecules in immune cell chemotaxis (54–56). For instance, early T cell development. Nat. Immunol. 8: 137–144. macrophage migration toward CSF-1 requires the Wave2 complex 8. Hoffman, E. S., L. Passoni, T. Crompton, T. M. Leu, D. G. Schatz, A. Koff, and its phosphorylation by MAPKs because the reduction of its M. J. Owen, and A. C. Hayday. 1996. Productive T-cell receptor beta-chain gene rearrangement: coincident regulation of cell cycle and clonality during devel- expression through iRNA abrogated F-actin–rich membrane pro- opment in vivo. Genes Dev. 10: 948–962. trusions (54). In addition, GEF-H1 plays a pivotal role in uropod 9. Mombaerts, P., A. R. Clarke, M. A. Rudnicki, J. Iacomini, S. Itohara, J. J. Lafaille, L. Wang, Y. Ichikawa, R. Jaenisch, M. L. Hooper, et al. 1992. formation during transendothelial migration of T cells (57). Mutations in T-cell antigen receptor genes alpha and beta block thymocyte de- Among these potential hnRNP L targets, the Diap1 gene, which is velopment at different stages. Nature 360: 225–231. involved in actin nucleation and polymerization, might be the key 10. Cheng, A. M., I. Negishi, S. J. Anderson, A. C. Chan, J. Bolen, D. Y. Loh, and T. Pawson. 1997. The Syk and ZAP-70 SH2-containing tyrosine kinases are im- regulator of lymphocyte migration in the thymus (56, 58). Simi- plicated in pre-T cell receptor signaling. Proc. Natl. Acad. Sci. USA 94: 9797–9801. larly to the results presented in this study, it has been reported that 11. Ciofani, M., and J. C. Zu´n˜iga-Pflu¨cker. 2005. Notch promotes survival of pre- mDia12/2 mice show reduced numbers of peripheral T cells and T cells at the beta-selection checkpoint by regulating cellular metabolism. Nat. Immunol. 6: 881–888. an impaired chemotactic response to CXCL12 and CCL21. 12. Germain, R. N. 2002. T-cell development and the CD42CD8 lineage decision. Moreover, the block in migration in mDia12/2 mice was due to Nat. Rev. Immunol. 2: 309–322. 13. Groves, T., P. Smiley, M. P. Cooke, K. Forbush, R. M. Perlmutter, and a suppressed production of F-actin (58), which is very similar to C. J. Guidos. 1996. Fyn can partially substitute for Lck in T lymphocyte de- the results we obtained with hnRNP L-deficient cells. It remains to velopment. Immunity 5: 417–428. 12 hnRNP L IN T CELL DEVELOPMENT AND MIGRATION

14. Juntilla, M. M., J. A. Wofford, M. J. Birnbaum, J. C. Rathmell, and 36. Falahati, R., and D. Leitenberg. 2008. Selective regulation of TCR signaling G. A. Koretzky. 2007. Akt1 and Akt2 are required for alphabeta thymocyte pathways by the CD45 protein tyrosine phosphatase during thymocyte devel- survival and differentiation. Proc. Natl. Acad. Sci. USA 104: 12105–12110. opment. J. Immunol. 181: 6082–6091. 15. Saint-Ruf, C., M. Panigada, O. Azogui, P. Debey, H. von Boehmer, and F. Grassi. 37. Bunting, M. D., I. Comerford, and S. R. McColl. 2011. Finding their niche: 2000. Different initiation of pre-TCR and gammadeltaTCR signalling. Nature chemokines directing cell migration in the thymus. Immunol. Cell Biol. 89: 185– 406: 524–527. 196. 16. Yamasaki, S., and T. Saito. 2007. Molecular basis for pre-TCR-mediated au- 38. Misslitz, A., O. Pabst, G. Hintzen, L. Ohl, E. Kremmer, H. T. Petrie, and tonomous signaling. Trends Immunol. 28: 39–43. R. Fo¨rster. 2004. Thymic T cell development and progenitor localization depend 17. Hermiston, M. L., Z. Xu, and A. Weiss. 2003. CD45: a critical regulator of on CCR7. J. Exp. Med. 200: 481–491. signaling thresholds in immune cells. Annu. Rev. Immunol. 21: 107–137. 39. Falahati, R., and D. Leitenberg. 2007. Changes in the role of the CD45 protein 18. Hermiston, M. L., J. Zikherman, and J. W. Zhu. 2009. CD45, CD148, and Lyp/ tyrosine phosphatase in regulating Lck tyrosine phosphorylation during thymic Pep: critical phosphatases regulating Src family kinase signaling networks in development. J. Immunol. 178: 2056–2064. immune cells. Immunol. Rev. 228: 288–311. 40. Wallace, V. A., J. M. Penninger, K. Kishihara, E. Timms, A. Shahinian, 19. Mustelin, T., K. M. Coggeshall, and A. Altman. 1989. Rapid activation of the T- H. Pircher, T. M. Ku¨ndig, P. S. Ohashi, and T. W. Mak. 1997. Alterations in the cell tyrosine protein kinase pp56lck by the CD45 phosphotyrosine phosphatase. level of CD45 surface expression affect the outcome of thymic selection. J. Proc. Natl. Acad. Sci. USA 86: 6302–6306. Immunol. 158: 3205–3214. 20. Mustelin, T., T. Pessa-Morikawa, M. Autero, M. Gassmann, L. C. Andersson, 41. Takeda, A., A. Matsuda, R. M. Paul, and N. R. Yaseen. 2004. CD45-associated C. G. Gahmberg, and P. Burn. 1992. Regulation of the p59fyn protein tyrosine protein inhibits CD45 dimerization and up-regulates its protein tyrosine phos- kinase by the CD45 phosphotyrosine phosphatase. Eur. J. Immunol. 22: 1173– phatase activity. Blood 103: 3440–3447. 1178. 42. Juntilla, M. M., and G. A. Koretzky. 2008. Critical roles of the PI3K/Akt sig- 21. Ostergaard, H. L., D. A. Shackelford, T. R. Hurley, P. Johnson, R. Hyman, naling pathway in T cell development. Immunol. Lett. 116: 104–110. B. M. Sefton, and I. S. Trowbridge. 1989. Expression of CD45 alters phos- 43. Palacios, E. H., and A. Weiss. 2004. Function of the Src-family kinases, Lck and phorylation of the lck-encoded tyrosine protein kinase in murine lymphoma T- Fyn, in T-cell development and activation. Oncogene 23: 7990–8000. cell lines. Proc. Natl. Acad. Sci. USA 86: 8959–8963. 44. Salmond, R. J., L. McNeill, N. Holmes, and D. R. Alexander. 2008. CD4+ T cell 22. Zikherman, J., C. Jenne, S. Watson, K. Doan, W. Raschke, C. C. Goodnow, and hyper-responsiveness in CD45 transgenic mice is independent of isoform. Int. A. Weiss. 2010. CD45-Csk phosphatase-kinase titration uncouples basal and Immunol. 20: 819–827. inducible T cell receptor signaling during thymic development. Immunity 32: 45. Shiroki, F., S. Matsuda, T. Doi, M. Fujiwara, Y. Mochizuki, T. Kadowaki, Downloaded from 342–354. H. Suzuki, and S. Koyasu. 2007. The p85alpha regulatory subunit of class IA 23. Tong, A., J. Nguyen, and K. W. Lynch. 2005. Differential expression of CD45 phosphoinositide 3-kinase regulates beta-selection in thymocyte development. J. isoforms is controlled by the combined activity of basal and inducible splicing- Immunol. 178: 1349–1356. regulatory elements in each of the variable exons. J. Biol. Chem. 280: 38297– 46. Oberdoerffer, S., L. F. Moita, D. Neems, R. P. Freitas, N. Hacohen, and A. Rao. 38304. 2008. Regulation of CD45 alternative splicing by heterogeneous ribonucleo- 24. Fujii, Y., M. Okumura, K. Inada, K. Nakahara, and H. Matsuda. 1992. CD45 protein, hnRNPLL. Science 321: 686–691. isoform expression during T cell development in the thymus. Eur. J. Immunol. 47. Topp, J. D., J. Jackson, A. A. Melton, and K. W. Lynch. 2008. A cell-based

22: 1843–1850. screen for splicing regulators identifies hnRNP LL as a distinct signal-induced http://www.jimmunol.org/ 25. Fukuhara, K., M. Okumura, H. Shiono, M. Inoue, Y. Kadota, S. Miyoshi, and repressor of CD45 variable exon 4. RNA 14: 2038–2049. H. Matsuda. 2002. A study on CD45 isoform expression during T-cell devel- 48. Wu, Z., X. Jia, L. de la Cruz, X. C. Su, B. Marzolf, P. Troisch, D. Zak, opment and selection events in the human thymus. Hum. Immunol. 63: 394– A. Hamilton, B. Whittle, D. Yu, et al. 2008. Memory T cell RNA rearrangement 404. programmed by heterogeneous nuclear ribonucleoprotein hnRNPLL. Immunity 26. McNeill, L., R. L. Cassady, S. Sarkardei, J. C. Cooper, G. Morgan, and 29: 863–875. D. R. Alexander. 2004. CD45 isoforms in T cell signalling and development. 49. Fernandis, A. Z., R. P. Cherla, and R. K. Ganju. 2003. Differential regulation of Immunol. Lett. 92: 125–134. CXCR4-mediated T-cell chemotaxis and mitogen-activated protein kinase acti- 27. Xu, Z., and A. Weiss. 2002. Negative regulation of CD45 by differential vation by the membrane tyrosine phosphatase, CD45. J. Biol. Chem. 278: 9536– homodimerization of the alternatively spliced isoforms. Nat. Immunol. 3: 764– 9543. 771. 50. Lai, J. C., M. Wlodarska, D. J. Liu, N. Abraham, and P. Johnson. 2010. CD45 28. Byth, K. F., L. A. Conroy, S. Howlett, A. J. Smith, J. May, D. R. Alexander, and regulates migration, proliferation, and progression of double negative 1 thy-

N. Holmes. 1996. CD45-null transgenic mice reveal a positive regulatory role for mocytes. J. Immunol. 185: 2059–2070. by guest on October 2, 2021 CD45 in early thymocyte development, in the selection of CD4+CD8+ thymo- 51. Wang, H. Y., X. Xu, J. H. Ding, J. R. Bermingham, Jr., and X. D. Fu. 2001. SC35 cytes, and B cell maturation. J. Exp. Med. 183: 1707–1718. plays a role in T cell development and alternative splicing of CD45. Mol. Cell 7: 29. Kishihara, K., J. Penninger, V. A. Wallace, T. M. Ku¨ndig, K. Kawai, 331–342. A. Wakeham, E. Timms, K. Pfeffer, P. S. Ohashi, M. L. Thomas, et al. 1993. 52. Park, H., M. M. Chan, and B. M. Iritani. 2010. Hem-1: putting the “WAVE” into Normal B lymphocyte development but impaired T cell maturation in CD45- actin polymerization during an immune response. FEBS Lett. 584: 4923–4932. exon6 protein tyrosine phosphatase-deficient mice. Cell 74: 143–156. 53. Parsons, J. T., A. R. Horwitz, and M. A. Schwartz. 2010. Cell adhesion: inte- 30. Mee, P. J., M. Turner, M. A. Basson, P. S. Costello, R. Zamoyska, and grating cytoskeletal dynamics and cellular tension. Nat. Rev. Mol. Cell Biol. 11: V. L. Tybulewicz. 1999. Greatly reduced efficiency of both positive and negative 633–643. selection of thymocytes in CD45 tyrosine phosphatase-deficient mice. Eur. J. 54. Kheir, W. A., J. C. Gevrey, H. Yamaguchi, B. Isaac, and D. Cox. 2005. A Immunol. 29: 2923–2933. WAVE2-Abi1 complex mediates CSF-1-induced F-actin-rich membrane pro- 31. Rothrock, C. R., A. E. House, and K. W. Lynch. 2005. HnRNP L represses exon trusions and migration in macrophages. J. Cell Sci. 118: 5369–5379. splicing via a regulated exonic splicing silencer. EMBO J. 24: 2792–2802. 55. Lebensohn, A. M., and M. W. Kirschner. 2009. Activation of the WAVE complex 32. Heyd, F., G. ten Dam, and T. Mo¨ro¨y. 2006. Auxiliary splice factor U2AF26 and by coincident signals controls actin assembly. Mol. Cell 36: 512–524. transcription factor Gfi1 cooperate directly in regulating CD45 alternative 56. Tanizaki, H., G. Egawa, K. Inaba, T. Honda, S. Nakajima, C. S. Moniaga, splicing. Nat. Immunol. 7: 859–867. A. Otsuka, T. Ishizaki, M. Tomura, T. Watanabe, et al. 2010. Rho-mDia1 33. Heyd, F., and K. W. Lynch. 2010. Phosphorylation-dependent regulation of PSF pathway is required for adhesion, migration, and T-cell stimulation in den- by GSK3 controls CD45 alternative splicing. Mol. Cell 40: 126–137. dritic cells. Blood 116: 5875–5884. 34. Kosan, C., I. Saba, M. Godmann, S. Herold, B. Herkert, M. Eilers, and T. Mo¨ro¨y. 57. Heasman, S. J., L. M. Carlin, S. Cox, T. Ng, and A. J. Ridley. 2010. Coordinated 2010. Transcription factor miz-1 is required to regulate interleukin-7 receptor RhoA signaling at the leading edge and uropod is required for T cell trans- signaling at early commitment stages of B cell differentiation. Immunity 33: endothelial migration. J. Cell Biol. 190: 553–563. 917–928. 58. Sakata, D., H. Taniguchi, S. Yasuda, A. Adachi-Morishima, Y. Hamazaki, 35. Lucas, B., and R. N. Germain. 1996. Unexpectedly complex regulation of CD4/ R. Nakayama, T. Miki, N. Minato, and S. Narumiya. 2007. Impaired CD8 coreceptor expression supports a revised model for CD4+CD8+ thymocyte T lymphocyte trafficking in mice deficient in an actin-nucleating protein, mDia1. differentiation. Immunity 5: 461–477. J. Exp. Med. 204: 2031–2038. A 3‘probe 5‘probe Stu1 ATG ATG ATG ATG TAA

Stu1 Stu1

neo Targeting vector/ modified locus Stu1 probe

Final locus

Inactivated locus

B C

wt

flox

D E

DN4 DP CD4SP CD8SP Thymus LckCre LckCre LckCre LckCre LckCre α fl/fl fl/fl fl/fl fl/fl fl/fl fl/fl fl/fl fl/fl fl/fl fl/fl
hnRNP L

deleted αactin floxed

Gaudreau et al., Supplemental Figure 1 hnRNP Lfl/fl A 5.78 74.4 11.7 25.1 B

8.8 5.81 3.76 16 0.08 63.1 1.01 95.7 26.5 16.5 76.1 3.07 CD8

2.96 0.34 1.1 55.8 CD4

LckCre+ fl/fl hnRNP L 6.92 51.9 7.47 18.3

1.45 3.54 4.34 36.8 0.09 74.1 1.22 90.1 19.5 20.4 CD69 90.3 1.24 CD8

6.59 2.13 2.32 57.8 TCRβ CD4

C                        

  Counts CD8 CD8  

      CD4 CD4 CD69 TCRβ

D                        

  Counts CD8 CD8  

      

CD4 CD4 CD69 TCRβ

Gaudreau et al., Supplemental Figure 2 A WT replicates-expression KO replicates-expression 10 10

5 5 R^2=0.973 R^2=0.964 0 0

-5 -5

-10 -10 -10 -5 0 5 10 -10 -5 0 5 10

B 350

300 300 250

200 200 150 Frequency Frequency 100 100

50

-10 -5 0 5 10 15 -10 -5 0 5 10 Log2 of WT RPKM value Log2 of KO RPKM value

C Alternative splicing pattern Ptprc NM_011210 Alternative splicing pattern Diap1 NM_007858 WT splicing WT splicing

KO splicing KO splicing

Alternative splicing pattern Wasf2 NM_153423 Alternative splicing pattern Arghgef2 NM_001198913 WT splicing WT splicing

KO splicing KO splicing

Alternative splicing pattern Dock9 NM_001128308 Alternative splicing pattern Arhgap17 NM_001122640 WT splicing WT splicing

KO splicing KO splicing

Gaudreau et al., Supplemental Figure 3 Supplemental Figure 1. A) Schematic representation of the murine hnRNP L locus and the targeting strategy to generate the conditional deletion of hnRNP L allele. Exons 2 and 6 are flanked by loxP sites. Upon Cre recombinase activity, the exons 2 to 6 will be excised, inactivating the locus and therefore the expression of the hnRNP L protein. B) Southern blot of Southern blot of ES cells clones partially recombined with the neo cassette removed leaving the endogenous locus with only 2 loxP sites inserted were identified. C) Relative expression of hnRNP L on different sorted T cell subset by qRT-PCR. D) Genotyping PCR to detect floxed and deleted hnRNP L alleles performed on sorted thymocyte subpopulations or complete thymus from control or LckCre+hnRNPLfl/fl mice. E) Western blot analysis of expression of hnRNP L protein in thymus from wt or LckCre+hnRNPLfl/fl mice and the loading control actin.

Supplemental Figure 2. A) Positive selection was assessed by flow cytometry on thymus from control wt or hnRNP L deleted based on surface expression of CD4 and CD8 on gated population TCRloCD69-, TCRloCD69+, TCRhiCD69+ and TCRhiCD69-. B) Absolute cell number of each positively selected population. Graphs show mean ± SEM (n=3). Wild-type and LckCre+hnRNP Lfl/fl mice were crossed with OTI TCR (C) and OTII (D) transgenic mice. Thymus from the indicated compound mutant animals were harvested and analyzed by flow cytometry for CD69 and TCRβ expression level on the following gated populations : CD8SP and CD4SP for OTI TCR and OTII transgenic mice respectively. Histograms show cells from wt (gray) or LckCre+hnRNP Lfl/fl (blue) mice in either OTI TCR or OTII TCR transgenic background. Plots are representative of three independent experiments for each OT transgene.

Supplemental Figure 3. A) Plot of the gene expression profile of each duplicate from wt and mutant thymus. The data indicate that both biological replicates are highly similar with regard to the global gene expression levels. B) Gene expression profiles (histograms) from wt and hnRNP L deleted thymocytes. Shown is the log2 RPKM (Reads per kilobase of exon model per million mapped reads) values against the number of genes. C) Graphical depiction of splice junction usage of the genes shown in Table I in wt or hnRNP L-/- (KO) thymocytes. The exons of the genes are shown in the middle (green filled boxes). The thickness of the lines above and below the exons represent the number of reads connecting individual exons in wt (blue) or KO (red) cells. Individual splice junctions that show an enrichment in use of more than 2-fold are marked by an asterisk (*). Supplementary Table I: List of genes which contain exon(s) showing preferential splicing in wt and hnRNP L deleted thymocytes

Genes with at least one junction Genes with at least one junction preferentially used by wt cells preferentially used by hnRNP L-/- cells Tbc1d17 Cep97 Kdm3b Baz1a Slc15a2 Ptprc Hps3 Cd8b1 Zbed4 Gm14085 Camk4 9030617O03Rik Ept1 2210404J11Rik Dntt Zfp143 Ofd1 Zfp703 Slc9a6 Dst Dcaf7 Exoc6b Tnrc6b Gstm4 Rpa2 Tnc Abcf3 Cntrob Sh3bp1 Mfn2 0610037L13Rik 9030624J02Rik Vmp1 Gprc5b Dpp8 3110009E18Rik Zcchc2 Ift52 Pfkl Cdk5rap1 Banp Ptch1 Sytl2 Agpat9 Folr4 Ik Inpp5e C030044B11Rik Ccdc164 Ccdc90b Nek3 Chd8 Tmie C630004H02Rik Ankrd16 Rasal1 Casd1 Prickle1 Serac1 2610301B20Rik Eif4g3 Birc6 Mccc1 Myo19 Tyk2 Usp37 Luc7l Gar1 Nf2 Slc6a19 Ppp6r2 Itpr1 Elf2 Ass1 Plcl1 Rrp12 Dzip1 Diap1 Lmbr1l Sept8 Ggps1 Alas2 Hsph1 St3gal4 Rbbp8 Erc1 Txndc11 Cntln P2rx1 Ylpm1 Ccdc88a 2700049A03Rik Fmnl2 Ints7 Wac Nop14 Whrn Rad54b Fancd2 Eif2a Eif4enif1 Snap23 Tspan5 Zfp90 Qser1 Zfyve26 Arfgap3 Rsrc1 Dock2 Nsf Fam116a Wnk1 Traf4 Ktn1 Rhot2 N6amt2 Dpm1 Mtrr Galnt11 Fam38a Nfkb1 Tmem135 Unc119 Brd4 Cept1 Fus Gramd3 Prdx6 Skap1 Phf16 Usp31 Ndufa13 Ccdc163 AI506816 Fam65a Krt8 Lsm14b Dock9 Ptk2 Pygl Tmx3 Traf3ip1 Jmjd7 Wdr11 Vezt Exoc1 Ttc3 Ift172 0610010F05Rik Slc28a2 Akt2 Trmt5 Epc2 4933407C03Rik Kif4 Alkbh4 Nup214 Zdhhc6 Ptpn9 Mpp7 Stat5a Polq Dlg3 Ubap2 Dmwd Eml3 Rxrb Parp4 Ttyh3 Rrp1 Chic2 Sbf1 Ube2t Psmd11 Atp9a Dars2 Dhps Uap1 Map3k10 Tbc1d25 Mtor Cwf19l2 Wdfy4 Dock8 Wnt5b Ccdc66 Tnfrsf9 2610008E11Rik Ulk2 1110032A03Rik Vps36 Odf2 Coq10b Ankle1 Hltf Tln1 Hivep2 Abca2 Gmip Baz2b 1600014C10Rik Pask Acad8 Rab6b Dgkg Polr3gl Ddit4 Madd Abcc4 Haus6 Arvcf Ccdc84 C80913 Nudt19 Elmod2 Aasdh Ralgapa1 As3mt Stxbp3a Rfc4 Wdr8 Lmf2 Htatip2 2310003H01Rik Ng23 Usp48 Edem3 Tatdn3 Cbx5 Parp2 Myo9a 1300001I01Rik Sel1l Engase Atp13a3 Akt3 St7 Ankrd39 Trdmt1 Slc22a3 C030046I01Rik Ube4a Ranbp9 Klhl17 Taf15 Smndc1 Lemd3 Fanca Gcn1l1 Sec16a Mapkapk5 Vps13b Rnf20 Hacl1 Xrn1 Mcam Ttc13 Wasf2 Pex6 Heatr2 Asb3 Dennd1a Clip2 St3gal1 Itch Tmem71 Plod2 Drg2 Prdx4 Gcc2 Rnft1 Tns1 Slc12a4 Gtf2h3 Smc5 Polr3a Tarbp1 Pacs1 Wibg Morc2a Vps13b Fam133b Dscr3 Nedd4l Tnk2 Bcas3 Rnps1 L3mbtl3 Ctnna1 Ano8 Pold2 Atxn2 Slu7 Stx17 Ppat Cox18 Prkci Fam46c Rfc2 Eya2 Setd4 Tbc1d22b Acap3 Arhgef1 Mknk2 Tmem175 Rrn3 Uba6 Senp6 Rbm33 BC024659 Rnf157 Siah2 Abca1 Tesk1 Rps6ka1 Pip4k2c Setdb1 Cutc Dzip3 Dnajc2 Pcnt Ubr1 4922501C03Rik Esyt2 Mcts1 Pign Nwd1 Snx14 Arhgap33 Fam134b Klc3 Aff4 Slc30a9 Nnt Xkr8 Tti1 Sclt1 Usp34 D19Bwg1357e Atm Arhgap26 Usp49 Fam203a B3galtl Ppwd1 Gnpda1 Hdac4 Spred1 C330018D20Rik Tlk1 Prss16 Tmtc4 Sept6 Alkbh1 Xpo6 Gpd1l Ipo11 Nfat5 Lrp8 Phf20 Ndufaf4 Tusc3 Polr3d Tstd2 Nf2 Tex10 Ncapd2 Hnrnpl Erlin1 Mfsd11 Pigq Pde1b Haus2 Dynll1 Dtl Cnot3 Btaf1 Dnajb2 C2cd3 Fbxw4 Bcl10 Galnt10 Brpf3 Itgav Ext1 Lrig2 Ddx19b Rcor2 Taf4a Tm2d3 Tbp Magi3 A230050P20Rik 4922501C03Rik Mcph1 Eno3 Tmub2 Ccm2 Smarcc2 Glod4 Chkb Gm5801 Rhobtb2 Itih5 Scml4 Mtfmt Gipc1 Ndor1 Smn1 Zfp719 Arhgef2 Pyroxd1 Fgfr1op2 Aco1 Slfn8 1300018J18Rik Acy1 Fbxl20 Sigmar1 H2afv 5730455O13Rik Mib2 Snx29 Tdrd5 Tctex1d2 Ints12 5133401N09Rik Macrod1 Myo1g Pla2g6 Sbno2 Usp45 Wdr78 Acap2 Tbk1 Osbpl8 Ankrd13d Pecam1 Sf1 Ccdc45 Tomm40l Fam160b2 Cep164 Lrba Arhgap17 Nme4 Pgam5 Nup160 Trrap Tgfbr3 Grasp Trio Fbf1 Slc43a2 Helz Araf Vat1 Mta3 Ccdc41 Fancc Pola1 Cramp1l Usp19 Arap2 Col15a1 Ndufb8 Ppp2r2a Gpi1 Slu7 Syt13 Sin3a Tasp1 Prcp 1300018I17Rik Ythdc1 AI316807 Prrc2a Cep170 Dvl2 Glipr2 Yif1b Pde5a BC032203 Rab28 Pja2 Pafah1b2 Xpo4 Sirt1 Tmem194 Fancl Rnpep Pik3r1 Vrk1 Zswim6 Chtf18 Wdr36 9130011E15Rik Arl5a Slc5a9 Cep192 Ccar1 Ephb6 Cenpc1 Hcfc2 Nin Mapk8ip3 Creb3 Terf2ip Scmh1 Anks1 Pogz Ahcyl2 Sh3glb2 Chchd6 St6galnac4 Ap1g2 Sufu Camk1d Haus1 D19Ertd386e Mettl7a1 Ddhd1 4931414P19Rik Kif1c Adssl1 Prps2 Fn1 Dennd1c Usp40 Fanca Eps8l1 Kdm5c Gtf2i Pbrm1 Inadl Itgb1bp1 Pomgnt1 Zcchc6 2310001A20Rik Dnajc24 Cstf3 Pitrm1 Elp4 Ttc14 Nphp1 1700021K19Rik Rbm17 Rps6ka5 Sec23b Vps13d Hells 1810020D17Rik Dym Lfng Kcnk5 Mtg1 Mybbp1a Uvrag 9430070O13Rik 0610007P08Rik Fam50a Avl9 Tpr Hsd11b1 Rngtt Mcm3ap Jmjd8 Rasa3 Arrb1 Stard3 Smyd1 Sytl3 Tbc1d15 Ogfrl1 Sox13 Phkg2 Pdcd11 Snx25 Gusb Map3k4 Crebbp Wdr78 Zfp692 Med6 Prkaa1 Erap1 P4ha1 Ehmt1 Atp11a Fam92a Slc7a1 B230118H07Rik 4930547N16Rik Btaf1 Atr Ccdc12 Wdr77 Nlrc3 Ndufv3 Rrnad1 Sass6 Sos2 Zfp317 Zfp395 Rpl7a Sike1 Fgfr1op2 Rbl2 Asah1 BC013529 Igf2r Mlxip Pkp4 Tbcb Tmem154 Brwd1 Ddah2