Role of SUMO-1 and SUMO interacting motifs in rhesus TRIM5α-mediated restriction
© Lukic et al.; licensee BioMed Central Ltd. 2013
Received: 16 September 2012
Accepted: 16 January 2013
Published: 1 February 2013
Skip to main content
© Lukic et al.; licensee BioMed Central Ltd. 2013
Received: 16 September 2012
Accepted: 16 January 2013
Published: 1 February 2013
TRIM5α is a member of the tripartite motif family of proteins that restricts retroviral infection in a species-specific manner. The restriction requires an interaction between the viral capsid lattice and the B30.2/SPRY domain of TRIM5α. Previously, we determined that two SUMO interacting motifs (SIMs) present in the B30.2/SPRY domain of human TRIM5α (huTRIM5α) were important for the restriction of N-tropic Murine Leukemia Virus. Here, we examined whether SUMO expression and the SIM1 and SIM2 motifs in rhesus monkey TRIM5α (rhTRIM5α) are similarly important for Human Immunodeficiency Type 1 (HIV-) restriction.
We found that mutation of SIM1 and SIM2 of rhTRIM5α abolished the restriction of HIV-1 virus. Further, knockdown of SUMO-1 in rhTRIM5α expressing cells abolished restriction of HIV-1. These results may be due, in part, to the ability of SUMO-1 to stabilize rhTRIM5α protein expression, as SUMO-1 knockdown increased rhTRIM5α turnover and the mutations in SIM1 and SIM2 led to more rapid degradation than the wild type protein. The NF-κB signaling ability of rhTRIM5α was also attenuated by SUMO-1 knockdown. Finally, upon inhibition of CRM1-dependent nuclear export with Leptomycin B (LMB), wild type rhTRIM5α localized to SUMO-1 bodies in the nucleus, while the SIM1 and SIM2 mutants did not localize to SUMO-1.
Our results suggest that the rhTRIM5α B30.2/SPRY domain is not only important for the recognition of the HIV-1 CA, but it is also important for its association with SUMO-1 or SUMO-1 modified proteins. These interactions help to maintain TRIM5α protein levels and its nuclear localization into specific nuclear bodies.
TRIM5α is a member of the TRIpartite Motif (TRIM) family of proteins, characterized as having three domains: a RING domain, either one or two B-box domains, and a coiled-coil domain (RBCC) . TRIM5α is capable of restricting retroviral infection in a species-specific manner. This restriction requires an interaction between the retroviral capsid (CA) lattice and the B30.2/SPRY domain of TRIM5α [2–6]. The B30.2/SPRY domain, located at the C-terminal of TRIM5α, confers the restriction spectrum of TRIM5α proteins . Human TRIM5α (huTRIM5α) potently restricts N-tropic murine leukemia virus (N-MLV) but it does not restrict B-tropic or NB-tropic MLV (B-MLV, NB-MLV respectively) [3, 7]. On the other hand, rhesus macaque TRIM5α (rhTRIM5α) restricts N-MLV and human immunodeficiency type 1 (HIV-1) [6, 8].
Small ubiquitin-related modifier (SUMO) proteins are conjugated to cellular substrates and regulate diverse cellular processes (for review see [9, 10]). SUMO proteins are transferred to lysine residues within the UBC9 binding site of the target protein. This binding site has a consensus sequence ΨKXE (where Ψ is a hydrophobic residue, K is the lysine to which SUMO-1 is conjugated, X is any amino acid and E is glutamic acid) [11, 12]. Conjugation of SUMO proteins to a substrate mediates distinct protein-protein interactions in vivo. These non-covalent interactions with SUMO modified proteins are mediated by SUMO interacting motifs (SIMs) [11, 13, 14]. The best-characterized SIMs have the consensus sequence V/I/L-x-V/I/L-V/I/L or V/I/L-V/I/L-x-V/I/L (where x is any amino acid) [14, 15].
We have previously identified that SUMO conjugation system is involved in huTRIM5α- mediated restriction of N-MLV . Three SIMs were identified in the huTRIM5α B30.2/SPRY domain. Two of them, SIM1 and SIM2, were responsible for the enhanced restriction of N-MLV in human cells observed upon SUMO-1 overexpression. Specifically, mutation of SIM1 and SIM2 resulted in a dramatic decrease in the ability of huTRIM5α to restrict N-MLV infection. The SIMs are conserved among TRIM5α orthologs, as the same SIM mutations also affected the rhTRIM5α restriction of N-MLV . As such, we hypothesized that rhTRIM5α restriction of HIV-1 was similarly dependent on SUMO-1 expression and the SIM1 and SIM2 motifs of rhTRIM5α.
Here we report that SIMs are important for HIV-1 restriction, and that knockdown of SUMO-1 in rhTRIM5α expressing cells drastically reduces HIV-1 restriction. We hypothesize that the presence of SUMO-1 stabilizes proteins levels of rhTRIM5α, as knockdown of SUMO-1 decreases steady state levels of rhTRIM5α and NF-κB activation. Additionally, mutation of the SIMs abolishes the co-localization of rhTRIM5α with promyelocytic leukemia protein (PML, also known as TRIM19)/SUMO-1 in the nucleus upon LMB treatment. Our results suggest that the rhTRIM5α B30.2/SPRY domain is not only important for the recognition of the HIV-1 CA, but that it is also important for its association with SUMO-1 or with SUMO-1 modified proteins. This association helps maintain TRIM5α protein levels, as well as its ability to mediate NF-κB, and nuclear localization into specific nuclear bodies.
We next asked how SUMO-1 depletion affected the ability of wild type rhTRIM5α to induce NF-κB activation. We co-transfected 293T cells with wild type HA-rhTRIM5α, SUMO-1 or control siRNA and an NF-κB driven luciferase reporter. We measured NF-κB activation by rhTRIM5α in SUMO-1 siRNA treated cells compared to rhTRIM5α cells treated with control siRNA. As shown in Figure 4C knocking down SUMO-1 had little effect on NF-κB activation when transfected with EV (~10%, black bar). However, depletion of SUMO-1 significantly reduced (~60%, p<0.004, Student’s T-test) NF-κB activation by wild type rhTRIM5α (Figure 4C, grey bar). The reduction of NF-κB activation in the presence of SUMO-1 siRNA was not due to reduced rhTRIM5α protein levels (Figure 4C inset) as SUMO-1 knockdown does in cells stably expressing rhTRIM5α (Figure 3). Conversely, overexpression of SUMO-1 increased NF-κB activation following transfection with empty vector or vector expressing wild type rhTRIM5α (Figure 4D). This increase was more pronounced when rhTRIM5α was present, although this result was not statistically significant (p=0.187, Student T-test). These experiments demonstrate that the ability to associate with SUMO-1 or SUMOylated proteins is relevant to rhTRIM5α-mediated NF-κB signaling.
Collectively, the data presented here demonstrate that SIMs in rhTRIM5α and SUMO-1 are important for rhTRIM5α-mediated retroviral restriction. rhTRIM5α is sensitive to changes in SUMO-1 expression as SUMO-1 knockdown relieves rhTRIM5α restriction (Figure 2) and increases rhTRIM5α protein turnover in stable cells (Figure 3). Additionally, SUMO-1 modulates rhTRIM5α-dependent activation of NF-κB (Figure 4). Upon SUMO-1 knockdown, NF-κB activation is decreased in a rhTRIM5α-dependent manner (Figure 4C). We also observed decreased localization of SIM1 and SIM2 to SUMO-1 puncta following LMB treatment (Figure 6).
The mechanism by which SUMO-1 stabilizes rhTRIM5α expression or affects retroviral restriction is unclear. It is possible that rhTRIM5α interacts with a protein that stabilizes its expression in a SUMO-dependent manner. It is also possible that the expression of another cellular protein that is stabilized by SUMOylation, stabilizes rhTRIM5α expression. Consistent with this, it is known that SUMOylation can protect some proteins from ubiquitin dependent degradation . Our observation that the SIM1 and SIM2 mutations, which abrogate restriction, are also turned over more rapidly than wild type rhTRIM5α are consistent with all of these possibilities. Interestingly, a recent study by Brandariz-Nunez et al. suggests that these residues are not present on the surface of the SPRY domain , as modeled using an NMR structure recently described for the PRY/SPRY domain . It may be that conformational changes induced by the binding of other cellular factors conditionally expose the SIM1 and SIM2 motifs. These authors also observe that SIM1 and SIM2 mutants oligomerize normally , suggesting these proteins are not misfolded. While we cannot exclude the possibility that SIM1 and SIM2 mutations prevent proper folding of the B30.2/SPRY domain, our own results demonstrate that alterations in SUMO-1 expression can influence the stability of wild type rhTRIM5α.
The data presented here may begin to explain the mechanistic of rhTRIM5α nuclear trafficking. It is possible that the recruitment of TRIM5α to SUMO-1/PML bodies has direct relevance to the restriction process, as both the SIM1 and SIM2 mutants that do not localize to SUMO-1/PML bodies are completely unable to restrict retroviral infection. Considering that PML protein undergoes SUMOylation on three lysine residues [28–30] and rhTRIM5α has a rapid turnover , we can speculate that the steady state levels of rhTRIM5α protein are maintained by its interaction via SIMs with SUMOylated-PML in the nucleus. When SIMs or SUMO-1 are not present rhTRIM5α is less retained in the nucleus and therefore is degraded faster in the cytosol, reducing the amount of protein available to restrict incoming viruses.
Interestingly, owl monkey TRIM-Cyp, which contains a C-terminal cyclophilin A domain instead of a B302/SPRY domain (the B30.2/SPRY domain contains all three putative SIMs) , does not traffic to the nucleus . Therefore, given the strong homology between the RBCC domains of owl monkey TRIM-Cyp and rhTRIM5α, it is possible that the SPRY domain contains the determinants that govern nuclear trafficking. Consistent with this hypothesis, our data demonstrate that wild type rhTRIM5α, but not the SIM1 and SIM2 mutants, localizes to nuclear SUMO-1 bodies (Figure 5) support this idea. Although the results we have obtained using the SIM1 and SIM2 mutants were consistent with the effect of SUMO-1 knockdown on wild type rhTRIM5α activity, the results obtained with these mutants should be cautiously interpreted .
Future studies are needed to determine if this association is merely correlative or the association of rhTRIM5α with nuclear SUMO-1/PML bodies potentiates specific steps in the restriction process.
Our results demonstrate that rhTRIM5α restriction and expression is governed by SUMO-1 expression. SUMO-1 depletion, or mutations that are expected to disrupt SUMO-1 binding, abrogate restriction and increase the rate of rhTRIM5α turnover. The ability of rhTRIM5α to activate innate immune signaling pathways is also sensitive to SUMO-1 expression. Additionally, the ability to traffic to nuclear SUMO-1 bodies is abrogated by mutation of the SIM1 or SIM2 motifs present in rhTRIM5α. As rhTRIM5α does not appear to itself be SUMOylated, future studies are needed to identify the cellular factor or factors that are likely regulating rhTRIM5α in a SUMO-1 dependent fashion.
Human embryonic fibroblast 293T, 293A, human medulloblastoma cell line TE671, HeLa and Crandall feline kidney (CRFK) fibroblast were maintained in Dulbecco’s modified Eagle medium supplemented with 10% fetal bovine serum, 100 UI/ml penicillin and 100 mg/ml streptomycin. All cells were cultured at 37°C in 5% CO2.
The rhTRIM5α constructs encoding the FLAG tagged wild type and mutant versions were identified and cloned as previously described . YFP tagged wild type and mutant versions were cloned into a retroviral vector previously described . RIG-I construct was kindly donated to us by Dr. Susan Baker (Loyola University Chicago).
Retroviruses for transduction were produced by transfection of 293T cells with 1 μg pMD.G, 1 μg pCMVI and 1.5 μg of either pQCXIN, pQCXIN-FLAG-rhTRIM5α wild-type or mutant versions, using FUGENE (Roche). Viruses were harvested 48 h after transfection, filtered (0.45 μm) and used to infect 5x105 cells in 100 mm dishes in the presence of 8 μg/ml polybrene. HeLa, TE671, 293A cells infected with vectors delivering the Neor gene were selected in 1.5 μg/ml G418. The CRFK cells overexpressing rhTRIM5α were previously described . Lentiviruses for transduction were produced by transfection of 293T cells with 1 μg pMD.G, 1 μg p8.91 and 1.5 μg of pGIPz (Open Biosystems) or pGIPzSUMO-1 DNAs containing shRNA (Open Biosystems). Viruses were harvested 48 h after infection, filtered (0.45 μm) and used to infect 5x104 TE671-rhTRIM5α cells in 35 mm dishes in the presence of 8 μg/ml polybrene. Cells were selected in 1.5 μg/ml G418 and 1.5 μg/ml puromycin.
Cells were lysed in 20 mMTris-HCl (pH 8.0), 137 mMKCl, 10% glycerol, 1% NP-40 and complete protease inhibitor (Roche) or Reporter lysis buffer (Promega). Samples were then boiled in 5x sodium dodecyl sulphate (SDS) loading buffer, and the proteins were resolved by SDS-polyacrylamide gel electrophoresis (PAGE). After transfer to nitrocellulose membranes, the blots were probed with mouse anti-β actin (Sigma), anti-HA (Covance), anti SUMO-1 (Abcam clone Y299), anti-FLAG (Sigma) or anti-GAPDH (Calbiochem).
HIV-1 luciferase reporter virus was produced by transfection of 293T cells with 2μg pNL4.3-env luciferase, and 1 μg pMD.G (per 100 mm plate) using FUGENE (Roche). Reporter virus stocks were harvested 48 h after transfection, then filtered (0.45 μm) and stored at −80°C. 293A (3x104 per well), HeLa (2.5x104 per well), TE671 (2.5x104 per well) and CRFK (3x104 per well) cells were seeded in 24-well plates and infected with MLV luc reporter viruses. Forty-eight hours post-infection cells were collected and assayed for firefly luciferase activity (Promega) in a luminometer.
In the case of the TE671 cells where SUMO-1 was stably knocked down, the cells were harvested and total RNA was extracted using TRIZOL reagent (Invitrogen). 2 μg of total RNA per cell line were used to produce cDNA using random hexamers and SuperScript III kit (Invitrogen). 2 μl of each cDNA were used for quantitative PCR analysis of SUMO-1 and GAPDH transcript levels. Fold change was calculated using the relative standard curve method.
For the transient knockdown of SUMO-1, HeLa cells stably expressing HA-rhTRIM5α were transfected with siRNA specifically targeting SUMO-1 or a non-targeting siRNA (Santa Cruz Biotechnology). Cells were transfected with Lipofectamine2000 (Invitrogen) using a 2-day transfection protocol followed by 24-hours or 48-hours of rest to gain maximum knockdown efficiency. Cells were collected following the resting period and analyzed by western blot using a monoclonal anti-SUMO-1 antibody (Abcam clone Y299). Densitometry analysis was performed on the western blots using ImageJ software.
HeLa cells transfected with wild type YFP-rhTRIM5α and SIM mutants (Lipofectamine2000 protocol) were allowed to adhere to fibronectin-treated glass coverslips. Cells were treated with LMB for 4 hours and fixed with 3.7% formaldehyde (Polysciences) in 0.1 M PIPES, pH 6.8 [piperazine-N, N'-bis(2-ethanesulfonic acid)] (Sigma). Monoclonal rabbit anti-SUMO-1 (Abcam clone Y299) antibody was used to stain SUMO-1. Primary antibody was secondarily labeled with Cy5 fluorophore-conjugated donkey anti-mouse antibody (Jackson ImmunoResearch). Nucleus was stained using a DAPI stain (Jackson ImmunoResearch). Images were collected with a DeltaVision microscope (Applied Precision) equipped with a digital camera (CoolSNAP HQ; Photometrics), using a 1.4-numerical aperture 100× objective lens, and they were deconvolved with SoftWoRx deconvolution software (Applied Precision).
20 Z-stack images were acquired using identical acquisition parameters. Surfaces for cytoplasmic bodies in all samples analyzed were defined by using a fluorescence threshold for YFP-rhTRIM5α, and all YFP-rhTRIM5α bodies over a volume of 0.011 μm3 were used in the analysis. Deconvolved images were analyzed for SUMO-1 maximum fluorescence intensity (MFI) in cytoplasmic bodies using the Surface Finder function of the Imaris software package (Bitplane) and the data was plotted in Prism (Graphpad Software Inc) for statistical analysis.
293T cells seeded in a 96-well plate were transfected with empty vector, ΔRING/SPRY rhTRIM5α (ΔRS, negative control), wild type rhTRIM5α, SIM1 mut, SIM2 mut, SIM3 mut or RIG-1 (positive control) in triplicate. Transfection was carried out using polyethylenimine (PEI) protocol in which the constructs were added at a 9 (EV/rhTRIM5α/RIG-I): 3 (NF-kB-responsive firefly luciferase construct): 1 (Renilla luciferase construct for transfection efficiencies) ratio. Cells were lysed 48-hours post transfection with Passive lysis buffer (Promega) and the luciferase activity was measured using a Dual-Glo luciferase assay system (Promega) in a Veritas Microplate luminometer. Firefly luciferase data were normalized to Renilla luciferase readings in each well. Data were plotted by determining the fold increase over empty vector.
293T cells seeded in a 96-well plate were transfected with empty vector, and rhTRIM5α constructs in presence or absence of SUMO-1 in triplicate. Transfection was carried out using PEI protocol in which the constructs were added at a 5 (rhTRIM5α): 4 (SUMO-1 or EV): 3 (NF-κB-responsive firefly luciferase construct): 1 (Renilla luciferase construct for normalization of transfection efficiencies) ratio. Cells were lysed 48-hours post transfection with Passive lysis buffer (Promega) and the luciferase activity was measured using a Dual-Glo luciferase assay system (Promega) in a Veritas Microplate luminometer. Firefly luciferase data were normalized to Renilla luciferase readings in each well. Data were plotted by dividing SUMO-1 siRNA NF-κB activation by Control siRNA NF-κB activation x 100.
293T cells seeded in a 12-well plate were transfected with Control siRNA or SUMO-1 siRNA (Santa Cruz Biotechnology, Inc) following a Lipofectamine2000 (Invitrogen) protocol for two days. On the third day the cells were seeded in a 96-well plate in triplicate and transfected with empty vector or wild type rhTRIM5α using PEI in which constructs were added at a 9 (EV/rhTRIM5α): 3 (NF-kB-responsive firefly luciferase construct): 1 (Renilla luciferase construct for transfection efficiencies) ratio. Cells were lysed 48-hours post transfection with Passive lysis buffer (Promega) and the luciferase activity was measured using a Dual-Glo luciferase assay system (Promega) in a Veritas Microplate luminometer. Firefly luciferase data were normalized to Renilla luciferase readings in each well. Data were plotted by dividing SUMO-1 siRNA NF-κB activation by Control siRNA NF-κB activation x 100.
Small ubiquitin-related protein
Rhesus monkey TRIM5α
Human immunodeficiency virus
GA was a Pew Latin American Fellow. SPG is a Howard Hughes Medical Institute Investigator. EMC is supported by National Institutes of Health grant R01 AI093258.
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.