Utilization of host SR protein kinases and RNA-splicing machinery during viral replication Takeshi Fukuhara*†, Takamitsu Hosoya‡§, Saki Shimizu¶ʈ, Kengo Sumi‡, Takako Oshiro*†, Yoshiyuki Yoshinaka**, Masaaki Suzuki‡, Naoki Yamamoto¶ʈ, Leonore A. Herzenberg††, Leonard A. Herzenberg††‡‡, and Masatoshi Hagiwara*†‡‡ *Laboratory of Gene Expression, School of Biomedical Science, †Department of Functional Genomics, Medical Research Institute, ¶Molecular Virology, Graduate School, and **Human Gene Sciences Center, Tokyo Medical and Dental University, Tokyo 113-8510, Japan; ‡Division of Regeneration and Advanced Medical Science, Graduate School of Medicine, Gifu University, Gifu 501-1193, Japan; ʈInfectious Disease Surveillance Center, National Institute of Infectious Diseases, Tokyo 162-8640, Japan; and ††Department of Genetics, Stanford University School of Medicine, Stanford, CA 94305-5318 Contributed by Leonard A. Herzenberg, June 12, 2006 Although the viral genome is often quite small, it encodes a broad residues in the RS domain occurs in all SR proteins (9, 10), which series of proteins. The virus takes advantage of the host-RNA- exist predominantly in a highly phosphorylated state in vivo (11). processing machinery to provide the alternative splicing capability The precise physiological role of this phosphorylation is still necessary for the expression of this proteomic diversity. Serine– unknown. However, it is reasonable to expect that phosphory- arginine-rich (SR) proteins and the kinases that activate them are lation of SR proteins affects their protein–protein and protein– central to this alternative splicing machinery. In studies reported RNA interactions, intracellular localization and trafficking, and here, we use the HIV genome as a model. We show that HIV alternative splicing of precursor mRNA (12). Consistent with expression decreases overall SR protein͞activity. However, we also this idea, SR protein kinase (SRPK)-dependent herpes simplex show that HIV expression is significantly increased (20-fold) when virus splicing and SRPK-mediated phosphorylation of the hep- one of the SR proteins, SRp75 is phosphorylated by SR protein atitis B virus core protein were reported in these viral diseases kinase (SRPK)2. Thus, inhibitors of SRPK2 and perhaps of function- (13–15). Therefore, SR proteins and the kinases that phosphor- ally related kinases, such as SRPK1, could be useful antiviral agents. ylate them could be practical targets for therapeutic modulation of alternative splicing in diseases and viral infections where Here, we develop this hypothesis and show that HIV expression alternative mRNA splicing is important (16). down-regulates SR proteins in Flp-In293 cells, resulting in only By extensively screening a chemical library for inhibitors of low-level HIV expression in these cells. However, increasing SRPK2 SRPKs, we initially found a benzothiazol compound, TG003, function up-regulates HIV expression. In addition, we introduce SR that has an inhibitory effect on the activity of Clk1 (17). TG003 protein phosphorylation inhibitor 340 (SRPIN340), which preferen- suppresses SR protein phosphorylation, blocks dissociation of tially inhibits SRPK1 and SRPK2 and down-regulates SRp75. Al- nuclear speckles, and inhibits Clk1-dependent alternative splic- though an isonicotinamide compound, SPRIN340 (or its derivatives) ing in mammalian cells. Other research (18) has shown that remain to be optimized for better specificity and lower cytotoxic- tricyclic quinoxaline derivatives inhibit SRPK1 kinase activity ity, we show here that SRPIN340 suppresses propagation of Sindbis but that these small molecules also inhibit a broad range of other virus in plaque assay and variably suppresses HIV production. Thus, kinases. Here, we define the relationship between RNA pro- we show that SRPK, a well known kinase in the cellular RNA- cessing and viral replication and introduce an inhibitor of processing machinery, is used by at least some viruses for propa- SRPK1. gation and hence suggest that SRPIN340 or its derivatives may be We show here that (i) HIV infection down-regulates SR useful for curbing viral diseases. proteins in Flp-In293 cells and results in low level HIV expres- sion; (ii) increasing SRp75 and the SRPK kinases that stabilize HIV ͉ kinase inhibitor ͉ SR protein phosphorylation inhibitor 340 and activate SRp75 up-regulates HIV expression; (iii) an isoni- cotinamide compound, SR protein phosphorylation inhibitor (SRPIN)340, which we identified by extensively screening a IV-1 precursor RNA transcribed from proviral DNA inte- chemical library, specifically inhibits SRPK and down-regulates Hgrated in the host cell genome contains all of the transcribed SRp75; and (iv) SRPIN340 variably inhibits HIV expression but viral reading frames (1). Alternative splicing is essential for regularly inhibits expression of another RNA virus (Sindbis). producing mRNAs encoding various viral proteins from the Finally, we present evidence suggesting that SRPIN340 may be limited size of a single precursor mRNA (2). In the early phase most effective for inhibiting acutely replicating viruses. Thus, we of HIV expression, eight splice acceptor sites compete for the show that kinases involved in RNA processing are important for splicing machinery to produce the vif, vpu, vpr, nef, env, tat, and viral replication and propose that targeting these kinases may be rev mRNAs (3). In the late phase of the virus life cycle, singly a useful strategy to control viral diseases. spliced longer RNA is translated to a polyprotein and then cleaved by HIV protease to generate gag and pol proteins. Results Several reports show that regulation of the complex splicing HIV Infection Alters the Phosphorylation State and Localization of SR pattern can dramatically affect HIV-1 infectivity and pathogen- Proteins. The mAb 1H4 detects phosphoepitopes in the RS esis (4–6). However, little is known about the molecular mech- domain of SR proteins (19). 1H4 can be used in confocal anism that links this alternative splicing regulation and the dynamics of virus propagation. Alternative splicing depends on the alternative utilization of Conflict of interest statement: No conflicts declared. MEDICAL SCIENCES four 5Ј splice sites and eight 3Ј splice sites (3). The combination Abbreviations: ASF, alternate splicing factor; DIC, differential interference contrast; of these splice sites are regulated by cis-regulatory elements, hnRNP, heterogeneous nucleoproteins; SF2, splicing factor 2; SR, serine–arginine-rich; RS, arginine–serine-rich; SRPK, SR protein kinase; mSRPK, mouse SRPK; SRPIN340, SR protein which bind cellular heterogeneous nucleoproteins (hnRNPs) of phosphorylation inhibitor 340. the A, B, and H groups and serine–arginine-rich (SR) proteins §Present address: Graduate School of Bioscience and Biotechnology, Tokyo Institute of (7). SR proteins are highly conserved in eukaryotes and are Technology, Yokohama 226-8501, Japan. characterized by having one or two RNA-recognition motifs at ‡‡To whom correspondence may be addressed. E-mail: [email protected] or the amino terminus and an arginine–serine-rich (RS) domain at [email protected]. the carboxyl terminus (7, 8). Extensive phosphorylation of serine © 2006 by The National Academy of Sciences of the USA www.pnas.org͞cgi͞doi͞10.1073͞pnas.0604616103 PNAS ͉ July 25, 2006 ͉ vol. 103 ͉ no. 30 ͉ 11329–11333 Downloaded by guest on September 24, 2021 Fig. 1. HIV alters the phosphorylation of SR proteins and the localization of SC35. The first image in each row is a single image of GFP. The second image in each row is for Alexa Fluor 555. The third image is a merged image, and the fourth image is DIC microscopy. (Top) Prepared GFP-tagged HIV virus infected HEK293 cells, and phosphorylated SR proteins were stained with an anti- phospho-SR-specific Ab 1H4 (Zymed). HIV-1 NL4-3-infected cells expressing GFP were not stained or were weakly stained by 1H4 Ab. (Middle and Bottom) Fig. 2. HIV-1 induced dephosphorylation of SR proteins. (A) Western blot Flp-In293 were transfected with GFP-expressing (Middle) or GFP-Tat- analysis of SR proteins in HIV-1 pNL4-3-transfected cells. Phosphorylation state expressing (Bottom) vector and further incubated for 72 h. Cells were fixed of SR proteins (SRp75, SRp55, SRp40, SC35, and SRp20) and expression of and stained with an anti-SC35 Ab as described in Materials and Methods.In SF2͞ASF and SC35 were examined with anti-phospho-SR (104, lanes 1–4; 1H4, GFP-expressing cells, nuclear speckle structure (multiple dots) of SC35 was lanes 5–8), anti-SC35 (lanes 9–12), anti-SF2 (lanes 13–16), anti-hnRNPA1 (lanes observed. In GFP-Tat-expressing cells, a strong GFP signal was observed in 17–20), and anti-Pin1 (lanes 21–24) Abs for the whole-cell extracts prepared nucleolus, and the SC35 signal was weaker in the nucleus and was seen as two from the mock-overexpressing (Venus) or SRPK2-overexpressing (SRPK2-2) dots near the nucleolus. The fluorescence images of SC35 and GFP and the DIC Flp-In293 cells with (ϩ) or without (Ϫ) HIV-1 pNL4-3 transfection. (B) Effects of images were taken with a confocal microscope (Olympus FV1000). (Scale bars, SR proteins on HIV-1 production. One of the expression vectors of HA-tagged 20 ␮m.) SR proteins (SRp75, SRp55, and SRp40) or hnRNPA1 was introduced into the Flp-In293 cells with HIV-1 pNL4-3. After 48 h of incubation, culture superna- tants were collected from each sample. (C) The SRPK2–SRp75 pathway en- fluorescence microscopy studies to compare the levels of phos- hances HIV-1 production. Culture supernatants were collected from the mock phorylated SR proteins in noninfected and HIV-infected cell and SRPK2-2 cells at 72 h after transfection of HA-SRp75 and HIV-1 pNL4-3. To lines. As Fig. 1 shows, phosphorylated SR proteins are readily estimate the virus production level, the amount of HIV-1 p24 antigen in the culture supernatant was quantified with LUMIPULSE (Fujirebio, Tokyo, Japan) detectable in uninfected Flp-In293 cells but decrease dramati- as described in ref. 28. Each sample was examined in triplicate. Values mea- cally (almost to undetectable) when these cells are infected with sured by LUMIPULSE are shown in the graph with standard deviations.