Identification and characterization of naturally occurring splice variants of SAMHD1
© Welbourn et al.; licensee BioMed Central Ltd. 2012
Received: 31 August 2012
Accepted: 24 September 2012
Published: 23 October 2012
Skip to main content
© Welbourn et al.; licensee BioMed Central Ltd. 2012
Received: 31 August 2012
Accepted: 24 September 2012
Published: 23 October 2012
Sterile Alpha Motif and HD domain-containing protein 1 (SAMHD1) is a recently identified host factor that restricts HIV-1 replication in dendritic and myeloid cells. SAMHD1 is a dNTPase that presumably reduces the cellular dNTP levels to levels too low for retroviral reverse transcription to occur. However, HIV-2 and SIV encoded Vpx counteracts the antiviral effects of SAMHD1 by targeting the protein for proteasomal degradation. SAMHD1 is encoded by a multiply spliced mRNA and consists of 16 coding exons.
Here, we identified two naturally occurring splice variants lacking exons 8–9 and 14, respectively. Like wildtype SAMHD1, both splice variants localize primarily to the nucleus, interact with Vpx, and retain some sensitivity to Vpx-dependent degradation. However, the splice variants differ from full-length SAMHD1 in their metabolic stability and catalytic activity. While full-length SAMHD1 is metabolically stable in uninfected cells, both splice variants were inherently metabolically unstable and were rapidly degraded even in the absence of Vpx. Vpx strongly increased the rate of degradation of full-length SAMHD1 and further accelerated the degradation of the splice variants. However, the effect of Vpx on the splice variants was more modest due to the inherent instability of these proteins. Analysis of dNTPase activity indicates that neither splice variant is catalytically active.
The identification of SAMHD1 splice variants exposes a potential regulatory mechanism that could enable the cell to control its dNTPase activity on a post-transcriptional level.
HIV-2 and many SIV isolates encode an accessory protein, Vpx, that is required for replication in myeloid cells [1–4]. HIV-1 does not encode a vpx gene; interestingly, however, the presence of Vpx enhances replication of HIV-1 in monocyte-derived macrophages, dendritic cells, and the differentiated THP-1 cell line [1, 5–8] suggesting the presence of a Vpx-sensitive host restriction factor. Also, Vpx was shown to rescue HIV-1 but not HIV-2 or SIV from an interferon-induced antiviral state . A Vpx-sensitive restriction factor was recently identified as Sterile Alpha Motif and HD domain-containing protein 1 (SAMHD1), which was found to be targeted for proteasomal degradation by Vpx [10–13]. Mutations in the SAMHD1 gene have been implicated with Aicardi-Goutieres Syndrome (AGS), a disease that is associated with increased production of interferon-alpha and thus mimics congenital virus infections [14, 15]. This suggests SAMHD1, along with other AGS-associated proteins (e.g. TREX1 and RNASEH2), may be involved in the regulation of the innate immune response . Moreover, SAMHD1 was shown to contribute to the restriction of HIV-1 in resting CD4+ T cells [17, 18]. SAMHD1 has recently been shown to possess dGTP-dependent dNTPase activity [19, 20], and the current hypothesis is that SAMHD1 is able to deplete the intracellular pool of dNTPs in susceptible cell types to levels below that required for reverse transcription, thus resulting in restriction of HIV-1 replication [19–22].
Here, we describe the identification and characterization of two SAMHD1 splice variants that are expressed naturally together with full-length SAMHD1 in a variety of cell types. The splice variants identified either lack exons 8–9 (Δ8-9), eliminating a C-terminal portion of the HD domain, or exon 14 (Δ14). Like wildtype SAMHD1, the Δ8-9 and Δ14 splice variants exhibit nuclear localization, and they interact with Vpx. However, unlike wildtype SAMHD1, which is stable in the absence of Vpx, both splice variants are inherently unstable and are rapidly degraded even in the absence of Vpx. Pulse/chase analyses confirmed that Vpx had a strong impact on the degradation of full-length SAMHD1. Addition of Vpx also increased the turnover of the splice variants. However, this effect was more modest because of the inherent instability of these proteins in the absence of Vpx. Neither splice variant exhibited dNTPase activity in an in vitro assay, suggesting they also lack antiviral activity. The identification of SAMHD1 splice variants exposes a potential regulatory mechanism that could enable a cell to control its dNTPase activity at a post-transcriptional level.
While wildtype SAMHD1 localizes almost exclusively to the nucleus , mutants associated with AGS cause the protein to be at least partially localized to the cytoplasm , suggesting a possible correlation between SAMHD1 nuclear localization and function. However, recent evidence has suggested that while SAMHD1 nuclear localization may not be strictly required for HIV-1 restriction, it may be important for the initiation of Vpx-dependent degradation of SAMHD1 . Immunocytochemistry of untagged wildtype SAMHD1 transfected into HeLa cells confirmed the nuclear localization of SAMHD1 (Figure 2, panel B). Similarly, SAMHD1Δ14 and Δ8-9 revealed exclusively nuclear localization in the vast majority of transfected cells (Figure 2, panels C and D). This indicates that sequences encoded by exons 8–9 and 14 are not critical for determining the subcellular distribution of SAMHD1. Indeed, a KRPR nuclear localization sequence has recently been mapped to the N-terminus of SAMHD1 .
To investigate whether the less pronounced effect of Vpx on the expression of SAMHD1 splice variants was due to a lack of protein-protein interaction, co-immunoprecipitation studies were performed employing N-terminally HA-tagged SAMHD1 variants, which were cotransfected into HeLa cells together with Vpx (Figure 3B). To minimize effects due to SAMHD1 degradation, cells were treated with the proteasome inhibitor MG132 for 4 h prior to cell lysis. Consistent with previous reports [10–12], wildtype SAMHD1 interacted with Vpx as indicated by the co-immunoprecipitation of Vpx in the SAMHD1 pulldowns (Figure 3B, bottom panel, lane 1). Interestingly, both SAMHD1 splice variants also interacted with Vpx as indicated by the efficient co-immunoprecipitation of Vpx (Figure 3B, lanes 2–3). Vpx was not pulled down by HA antibodies in the absence of SAMHD1 (Figure 3B, lane 4) attesting to the specificity of the assay. These results indicate that regions in SAMHD1 missing in the splice variants are not strictly required for Vpx binding. This is consistent with recent data suggesting a C-terminal region of SAMHD1 is involved in the interaction with Vpx [26, 27].
To assess the ability of our splice variants to restrict HIV-1 infection, we stably transduced U937 cells using a lentiviral vector system containing the indicated splicing variants ( Additional file 1: Figure S1A). Wild type SAMHD1 clearly restricted HIV-1 infection of U937 cells ( Additional file 1: Figure S1B). In contrast, neither one of the two splice variants inhibited HIV-1 infection ( Additional file 1: Figure S1B). Of note, expression of the splice variants was poor when compared to the wild type protein ( Additional file 1: Figure S1A). When wildtype SAMHD1 expression was reduced to levels comparable to those observed for the splice variants by titrating the transducing vector, no restriction of HIV-1 was observed even with wild type SAMHD1 (data not shown). Therefore, the results from this experiment do not conclusively prove or disprove restriction capacity of the SAMHD1 splice variants. However, since dNTPase activity of SAMHD1 appears to be required for its antiviral activity , our results would suggest that the Δ14 and Δ8-9 splice variants lack restricting activity.
Nearly 90% of human gene sequences consist of introns that must be removed through splicing (for review see ). Alternative splicing allows the production of different proteins from a single gene. Indeed, about 80% of genetic variability introduced through alternative splicing falls within open reading frames . Recent genome-wide analyses of alternative splicing indicate that up to 70% of human genes may have alternative splice forms, suggesting that alternative splicing together with various posttranslational modifications plays a major role in expanding proteome complexity . Proteins produced by alternative splicing can differ from the full-length product in their biophysical and biological properties at the level of protein stability, protein modification, intracellular localization, or surface expression . Thus, alternative splicing is a fundamental aspect of post-transcriptional gene regulation in eukaryotic cells with significant functional implications.
The 3.2 kb SAMHD1 mRNA (NM_015474) consists of 16 exons that have to be post-transcriptionally spliced from the 62 kb pre-mRNA. Recently, a G to A mutation in the SAMHD1 gene was found in patients with cerebral vasculopathy . The mutation localized to the exon 12 splice-acceptor site and resulted in the skipping of exon 13. Cell lines carrying this defect in SAMHD1 failed to show SAMHD1 expression suggesting that the exon 13-deficient splice variant was highly unstable . While the splicing defect observed in cerebral vasculopathy patients was not caused by alternative splicing but, instead, was induced by a genetic mutation, it nevertheless exemplifies the dramatic effects variation in splicing can have on the biophysical and biomedical properties of the resulting protein.
The fact that our SAMHD1 splice variants are catalytically inactive but retain the ability to bind Vpx, raises the possibility that these proteins might act as decoys for full-length SAMHD1 and interfere with the ability of Vpx to target the functional protein. Interference studies are currently in progress to experimentally test this possibility. However, given the inherent instability of the two splice variants identified in our study, it seems more plausible to speculate that alternative splicing of the SAMHD1 pre-mRNA is a regulatory mechanism employed by the cells to modulate expression and function of catalytically active SAMHD1 at a post-transcriptional level. This view is supported by our finding that both splice variants can be identified in multiple cell types. Because of its inherent dNTPase activity, SAMHD1 is likely to be involved in regulating the cellular dNTP levels. In non-dividing cells, such as terminally differentiated macrophages or dendritic cells, there is no active host DNA replication and SAMHD1 activity contributes to keeping the cellular dNTP level low. This might in fact explain, at least in part, the resistance of non-dividing cells to infection by HIV-1 [21, 22, 33]. In contrast, in actively dividing cells, e.g. immortalized cell lines such as 293 T or HeLa or in undifferentiated THP-1 cells, expression of an active dNTPase might be counterproductive. Still, we noted that the levels of SAMHD1 in differentiated (non-dividing) and undifferentiated (dividing) THP-1 cells are indistinguishable (data not shown) suggesting that dNTP levels in these cells may be regulated by other yet to be identified mechanisms.
While the SAMHD1 splice variants identified here are less stable than the full-length protein, it is possible that these or other yet to be identified forms of SAMHD1 may have functional roles in specific cell types or under specific conditions. In any case, characterization of different SAMHD1 splice variants, like those described here, not only yields important insight into potential functional differences of these proteins but also helps characterize the regions in SAMHD1 involved in dNTPase activity, HIV-1 restriction, and Vpx-dependent inactivation.
THP-1 cells were maintained in RPMI 1640 media containing 10% fetal bovine serum. For differentiation, 5 million cells were treated with 100 nM phorbol 12-myristate 13-acetate (PMA, Sigma-Aldrich, Inc., St. Louis MO) in a 6-well plate. Cells were washed after 24 h and left an additional 24 h in culture prior to harvest. HeLa cells were propagated in Dulbecco’s modified Eagles medium (DMEM) containing 10% fetal bovine serum. For transfection, HeLa cells were grown in 25 cm2 flasks to about 80% confluency (~3x 106 cells). Cells were transfected using LipofectAMINE PLUSTM (Invitrogen Corp, Carlsbad CA) following the manufacturer’s recommendations. A total of 5–6 μg of plasmid DNA per 25 cm2 flask was used. Where appropriate, empty vector DNA was used to adjust total DNA amounts. Samples were harvested 24 h post-transfection.
Polyclonal antibodies to human SAMHD1 and SIVmac239 Vpx were generated in rabbits immunized with purified recombinant fusion proteins expressed in E. coli. A MS2-SAMHD1 fusion protein containing amino acids 1–91 of the MS2 replicase followed by approximately 100 amino acids of a SAMHD1 fragment beginning at amino acid 416 was used to generate the SAMHD1 antiserum (SAM416), whereas the MS2 replicase was fused to the full length Vpx protein for generation of Vpx antiserum. A polyclonal antibody to actin (Sigma-Aldrich, Inc., St. Louis MO; Cat# A-5060) was used as a loading control. A mouse monoclonal antibody to human lamin B (Accurate Chemical & Scientific Corp., Westbury NY; Cat# BMDV3002) was used to stain the nuclear lamina. Fluorescently conjugated secondary antibodies used for immunofluorescence were purchased from Jackson ImmunoResearch Laboratories Inc. (West Grove PA).
cDNA was generated from total RNA from PMA-differentiated THP-1 cells using a Superscript III kit and an oligo dT primer (Invitrogen). The SAMHD1 specific sequence was then isolated by PCR from the cDNA pool using forward primer 5′ AAT AAG CTA GCG AAA CCA TGC AGC GAG CCG ATT CCG AG 3′ and reverse primer 5′ TGC TCT AGA TTA CAT TGG GTC ATC TTT AAA AAG CTG GAC 3′. The PCR primers include Bmt1 and XbaI restriction sites for insertion into the Bmt1/Xba1 site of pcDNA3.1-APOBEC3G-HA . The reverse primer also contained a stop codon to prevent expression of the C-terminal vector HA tag. This strategy led to the generation of pcDNA-SAMHD1 WT encoding the full-length protein as well as pcDNA-SAMHD1Δ14 and pcDNA-SAMHD1Δ8-9, encoding splice variants lacking exons 14 and 8–9, respectively. N-terminally HA-tagged SAMHD1 constructs were generated from the untagged constructs using similar Bmt1/Xba1 primers with an N-terminal HA tag included in the forward primer (5′ AA CTA GCT AGC GAA ACC ATG TAT CCA TAT GAC GTT CCA GAT TAC GCT TGT ACA CAG CGA GCC GAT TCC GAG CAG 3′). An active site mutant (SAMHD1 AA) was created by introducing a H206A/D207A double alanine mutation into wildtype SAMHD1 using Quikchange mutagenesis (Stratagene, Agilent Technologies). Plasmid pCMV-Vpxmac239 was constructed by replacing the SalI/NotI vif insert in pVif-HA  with the vpx gene from SIVmac239. For that purpose, vpx from SIVmac239 was PCR amplified using primers 5′ A CGC GTC GAC ACC ATG TCA GAT CCC AGG GAG AG 3′ (forward) and 5′ ATA GTT TAG CGG CCG CGG GTT ATG CTA GTC CTG GAG G 3′ (reverse). These primers include the Vpx start and stop codons and create SalI and NotI restriction sites for subcloning into pVif-HA. The resulting construct expresses untagged Vpx protein. In all cases, fidelity of DNA sequence and insertion of mutants was verified by sequencing.
For immunoblot analysis of cell-associated proteins, whole cell lysates were prepared as follows: Cells were washed once with PBS, suspended in PBS (~200 μl/5x 106 cells) and mixed with an equal volume of sample buffer (4% sodium dodecyl sulfate, 125 mM Tris–HCl, pH 6.8, 10% 2-mercaptoethanol, 10% glycerol, and 0.002% bromophenol blue). Proteins were solubilized by heating 10 to 15 minutes at 95°C with occasional vortexing. Cell lysates were subjected to SDS-PAGE; proteins were transferred to PVDF membranes and reacted with appropriate antibodies as described in the text. Membranes were then incubated with horseradish peroxidase-conjugated secondary antibodies (GE Healthcare, Piscataway NJ) and proteins were visualized by enhanced chemiluminescence (ECL, GE Healthcare, Piscataway NJ).
HeLa cells were transfected with expression vectors for HA-SAMHD1 and Vpx as indicated in the text. Cells were harvested 24 h post transfection, washed with cold PBS and lysed in lysis buffer A (50 mM Tris–HCl [pH 7.5], 150 mM NaCl, 1% [v/v] Triton X-100, 10% glycerol) at 4°C for 20 min, then clarified by centrifugation at 10,000 x g for 10 min at 4°C. Ten to fifteen percent of the lysate was used as an input control. The remaining lysate was used for immunoprecipitation of HA-tagged antigens. Cleared cell lysates were mixed with anti-HA antibody-conjugated agarose beads (Sigma-Aldrich, Inc., St. Louis MO) and incubated at 4°C for 4 h. Samples were then washed four times with lysis buffer A. Proteins were eluted by boiling beads in sample buffer and subjected to immunoblot analysis with antibodies to SAMHD1 and Vpx.
For pulse/chase analyses, transfected cells were harvested by scraping, pelleting, and suspending in 5 ml labeling media lacking methionine and cysteine (MP Biomedicals, Solon OH). Cells were incubated for 15 minutes at 37°C to deplete the endogenous methionine and cysteine pool. Cells were then pelleted, suspended in 600 μl of labeling medium containing 350 μCi of Expres35S35S protein labeling mix (Perkin Elmer, Shelton CT) and pulse-labeled at 37°C for 15 minutes. Cells were pelleted, suspended in complete RPMI, and equal volumes were distributed into separate tubes (1 tube per time point) containing 1 ml each of pre-warmed complete RPMI. After indicated chase times, cells were pelleted, supernatants were discarded, and cell pellets were frozen on dry ice. Cells were lysed in 200 μl of lysis buffer B (50 mM Tris–HCl [pH 7.5], 150 mM NaCl, 1% [v/v] Triton X-100) and incubated on ice for 5 minutes. Cell lysates were pelleted at 13,000 × g for 2 minutes to remove insoluble material. Cleared lysates were incubated for 1 h at 4°C with antibody-conjugated protein A-Sepharose. Beads were washed twice with wash buffer (50 mM Tris pH 7.4, 300 mM NaCl, 0.1% Triton X-100). Bound proteins were eluted by heating in sample buffer for 10 min at 95°C, separated by SDS-PAGE, and visualized by fluorography.
HeLa cells were transfected as indicated in the text. Transfected cells were trypsinized 3 h later and single-cell suspensions were distributed into 12 well plates containing 0.13 mm cover slips. Cells were grown for overnight at 37°C in DMEM containing 10% FBS. Cells were fixed in 10% methanol for 10 min at −20°C. Cells were blocked with 1% BSA in PBS for 30 minutes. For antibody staining, coverslips were incubated with appropriate primary antibodies in 1% BSA in PBS for 45 minutes at 37°C. Cells were washed with PBS and incubated with appropriate secondary antibodies for 45 minutes at 37°C. Cells were then washed twice with PBS and mounted onto microscope slides with glycerol gelatin (Sigma-Aldrich Inc., St. Louis MO) containing 0.1 M N-propyl gallate (Sigma) to prevent photo bleaching. Samples were analyzed on a Zeiss LSM410 inverted laser scanning microscope equipped with a krypton/argon mixed-gas laser. Images were acquired with a Plan-Apochromat 63x/1.4 oil immersion objective (Zeiss).
cDNA pools were generated from THP-1 cells treated with 100 nM PMA for 24 h and used as template for a PCR reaction using Taq polymerase (Roche) and the following primer sets: Δ14 primers consist of a forward primer (F14) located in exon 10 (5′ GAA GTT GGA AAT CTG TAT GAC ATG TTC CAC 3′) and a reverse primer (R14) spanning the junction of exons 13 and 15 (5′ CTC TGG CAG AAG TTG TGA AAC ATC 3′). Δ8-9 primers consist of a forward primer (F8/9) spanning the junction of exons 7 and 10 (5′ CTT GAA TCA CCT GTC GAA GAT TCA TTG GAA G 3′) and a reverse primer R8/9 in exon 15 (5′ CTG CGG CTT GGT GAA ATT TCT GTC TG 3′). Plasmid controls consisted of 1 ng of pcDNA-SAMHD1 WT, Δ14, or Δ8-9 used as template in the PCR reaction.
HA-SAMHD1 proteins were isolated from HeLa cells and used in a dNTPase assay using the conditions described in Lahouassa et al. . In brief, HeLa cells transfected with HA-SAMHD1 constructs were lysed in 50 mM Tris–HCl pH 8.0, 150 mM NaCl, 1% Triton at 4°C for 20 minutes, then clarified by centrifugation at 10,000 x g for 10 minutes at 4°C. Cleared cell lysates were mixed with anti-HA antibody-conjugated agarose beads (Sigma-Aldrich, Inc., St. Louis MO) and incubated at 4°C for 1.5-2 h. Samples were then washed three times with 50 mM Tris–HCl pH 8.0, 150 mM NaCl, 0.1% Triton and once with assay buffer A (50 mM Tris–HCl pH 8.0, 50 mM KCl, 5 mM MgCl2, 0.1% Triton). An equal volume of assay buffer A was then added to the bead pellet, 15% of the bead slurry reserved as an input control and the rest was mixed in a 1:1 ratio with assay buffer B (50 mM Tris–HCl pH8.0, 50 mM KCl, 5 mM MgCl2, 0.1% Triton, 400 μM dTTP, 0.5 μCi α-32P]-dTTP, +/− 400 μM dGTP). Reactions were incubated at 37°C for 3 h with occasional mixing and stopped by heating to 70°C for 5 minutes. A portion of the reaction products was separated out on polyethyleneimine (PEI)-cellulose thin layer chromatography plates (Sigma) using 0.8 M LiCl as the mobile phase.
We thank Amy Andrew, Takeshi Yoshida, Sandra Kao, and Haruka Yoshii-Kamiyama for helpful discussions and critical reading of the manuscript. This work was supported in part by the Intramural Research Program of the NIH, NIAID. SW was supported by a postdoctoral fellowship from the Canadian Institutes of Health Research (C.I.H.R) and an Intramural AIDS Research Fellowship from NIH. TEW and FD-G were funded by a NIH R01 AI087390 to FD-G.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.