The Human Cytomegalovirus Protein UL37 Exon 1 Associates with Internal Lipid Rafts

Cells and viruses.HeLa cells and HFFs (Viromed) were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS; HyClone), 100 U/ml penicillin, and 0.1 mg/ml streptomycin (1). The HCMV bacterial artificial chromosome strain AD169 (BAD) wild type (wt) (BADwt) was grown as previously described (72). Life-extended HFFs (LE-HFFs) were uninfected or infected with BADwt at a multiplicity of infection (MOI) of 3.

Mutagenesis.Site-specific mutations were introduced into the UL37 open reading frame (ORF) by site-directed mutagenesis using QuikChange kits (Stratagene) according to the manufacturer's instructions. To generate the CBD I (Y19S) mutation in the UL37x1 leader (CBD I_1-36, p1481) and the full-length pUL37x1 ORF (CBD I_1-163, p1515), forward primer 583 (5′-CTGCTGGCCTTTTGGTCCTTTAGCTACCGCTGG-3′) and reverse primer 584 (5′-CCAGCGGTAGCTAAAGGACCAAAAGGCCAGCAG-3′) were used. To generate CBD II (Y19S and R23A) in the UL37x1 leader sequence (CBD II_1-36, p1483) and in the full-length UL37x1 ORF (CBD II_1-163, p1521 and p1522), forward primer 587 (5′-CCTTTTGGTCCTTTAGCTACGCCTGGATCCAGCGAAAAC-3′) and reverse primer 588 (5′-GTTTTCGCTGGATCCAGGCGTAGCTAAAGGACCAAAAGG) were used to amplify the corresponding CBD I mutant plasmids (p1481 or p1515).

Transient transfection.For the isolation of ER/MAM and mitochondria, HeLa cells (∼6 × 10⁶ to 8 × 10⁶ cells) were transfected by using Lipofectamine 2000 (Invitrogen) suspended in Optimem (Gibco) according to the manufacturer's protocols, as previously described (9, 47). Briefly, DNA (μg)-to-lipid (μl) ratios for transfection were 1:1.75, with ∼20 μg DNA and 35 μl Lipofectamine 2000 used per T175 flask. Twenty-four hours after transfection, adherent cells were harvested and either stored on ice for fractionation or stored at −80°C for subsequent fractionation as previously described (8, 9).

Subcellular fractionations.Protocols for subcellular fractionations were previously described (8, 9). Briefly, for ER/MAM and mitochondrial fractionations, pellets of transfected cells were resuspended in 2 ml of MTE buffer (270 mM mannitol, 10 mM Tris-HCl, 0.1 mM EDTA [pH 7.4]) supplemented with complete protease inhibitor cocktail (Roche) and 1 mM phenylmethylsulfonyl fluoride (PMSF; Sigma) and lysed by sonication (an output power of 3.5). Cellular debris and intact cells were removed by centrifugation at 700 × g for 10 min at 4°C. Crude mitochondria were pelleted by centrifugation at 15,000 × g for 10 min at 4°C, and the postmitochondrial supernatant was used for the purification of ER/MAM fractions. Purified mitochondria were obtained by banding in discontinuous sucrose gradients consisting of 1.0 M and 1.7 M sucrose steps in 10 mM Tris-HCl (pH 7.6) at 40,000 × g for 22 min at 4°C. Postmitochondrial supernatants were layered onto discontinuous sucrose gradients (1.3 M, 1.5 M, and 2.0 M sucrose in 10 mM Tris-HCl [pH 7]) and banded by centrifugation at 100,000 × g for 45 min at 4°C. The purified ER/MAM and mitochondrial pellets were resuspended in 1× phosphate-buffered saline (PBS) and stored at −20°C until use.

Protein concentration determination.Protein concentrations of isolated subcellular fractions were determined by using a BCA assay kit (Pierce) according to the manufacturer's recommendations.

Western analyses.Twenty micrograms of total lysate or subcellular fractions was separated by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) in 10% Bis-Tris NuPage gels (Invitrogen) and transferred onto nitrocellulose membranes in Tris-glycine buffer with 20% methanol (50 V for 60 min) as described previously (9, 47). The parental control, wt control, and the vector negative control were included in all experiments. Membranes were blocked for 2 h at room temperature (RT) in 1× PBS containing 5% milk powder (Carnation) and 0.1% Tween 20 (Sigma) and then probed with primary antibody for 1 h at RT in 1× PBS containing 1% bovine serum albumin (BSA; Sigma). Primary antibodies used for these studies include rabbit anti-pUL37x1/vMIA (DC35; 1:2,500) (47), mouse anti-green fluorescent protein (GFP) (1:200; Santa Cruz Biotechnology), mouse anti-GRP75 (1:1,000, Stressgen); rabbit anti-fatty acid coenzyme A (CoA) ligase 4 (FACL4) (1:250; Abgent), goat anti-dolichyl phosphate mannose synthase 1 (DPM1) (1:100; Santa Cruz), mouse anti-Sig-1R (1:200; Santa Cruz), rabbit anti-prohibitin, (1:200; GeneTex), rabbit anti-cholera toxin subunit B (CTB) (1:200; Molecular Probes), rabbit anti-DsRed (1:2,000; Clontech), rabbit anti-erlin 2 (1:500; a gift from Stephen Robbins), and mouse anti-Lyn (1:250; BD Biosciences) antibodies.

Horseradish peroxidase (HRP)-conjugated secondary antibodies (1:2,500; Santa Cruz) were incubated for 1 h at RT in 1× PBS with 1% (wt/vol) BSA. Protein bands were detected by using an ECL detection kit (Pierce). Each blot was then stripped in 62.5 mM Tris buffer containing 2% SDS and 100 mM 2-mercaptoethanol for 1 h at 50°C with shaking. Stripped blots were washed seven times with 1× PBS, blocked, and reprobed for the detection of the specific protein organelle markers. Digital images were generated by using Scan Wizard Pro, version 1.21, and processed with Adobe Photoshop CS3.

Indirect immunofluorescence assay (IFA).HFFs were seeded at 20 to 30% confluence (∼4 × 10⁵ cells/cm²) on sterile coverslips in six-well plates (9.72 cm² per well). Twenty-four hours later, cells were transiently transfected by using Lipofectamine 2000 suspended in Optimem according to the manufacturer's protocols. DNA (μg)-to-lipid (μl) ratios for transfection were 1:1.75, with approximately 0.5 μg total DNA used per cm² of available plating surface area. Cells were fixed with ice-cold 100% methanol for 10 min. Fixed cells were washed with 1× PBS. Cells were then sequentially probed with primary antibodies as described previously (9, 13, 47).

Staining of lipid droplets.HFFs were transfected with expression vectors encoding Sig-1R-yellow fluorescent protein (YFP) or pUL37x1 wt_1-163-YFP and fixed 24 h later with 4% paraformaldehyde in 1× PBS for 30 min. Fixed cells were washed 5 times with 1× PBS, permeabilized with 1× PBS plus 0.75% Triton X-100, and blocked for 45 min using 1× PBS plus 4% BSA. Cells were then rinsed twice for 5 min each in 100% propylene glycol, stained for 7 min in prewarmed 100% propylene glycol containing Oil Red O (0.5%) in a 60°C oven, and rinsed again once for 3 min in 85% propylene glycol. Cells were then washed twice with distilled water and imaged by using confocal microscopy.

Confocal microscopy.Images were acquired with a Zeiss LSM 510 confocal microscope (Intellectual and Developmental Disabilities Research Center) as described previously (9, 47, 87, 88). Excitation wavelengths for the Zeiss LSM 510 microscope were 488 nm (for GFP), 514 nm (for YFP; emission collected through a 530- to 600-nm filter), and 561 nm (for DsRed1-mito). Images were acquired by sequential excitation through a 63× (numerical aperture [NA], 1.4) objective. Postacquisition processing was performed by using Adobe Photoshop CS3.

MβCD extraction of pUL37x1/vMIA from intracellular LRs.HeLa cells were transfected with either pUL37x1 wt_1-163-YFP or Sig-1R-YFP. Twenty-four hours later, transfected cells were incubated in 5% CO₂ at 37°C for 2 h in HFF medium alone (control) or HFF medium supplemented with 20 mM methyl-β-cyclodextrin (MβCD) (Sigma). Cells were then harvested and fractionated to purify ER/MAM and mitochondrial fractions (8, 88). Alternatively, comparable sets of transfected cells were grown in HFF medium for 24 h, harvested, and then resuspended in MTE buffer alone (control) or MTE buffer supplemented with 20 mM MβCD. Cells were sonicated in the above-described buffer and incubated at 37°C for 1 h with shaking. Subsequently, all cell lysate suspensions were separately pelleted at 700 × g for 10 min at 4°C and resuspended in fresh MTE buffer lacking MβCD. Pellets were fractionated to isolate purified ER/MAM and mitochondrial fractions as previously described (8, 9). Finally, comparable sets of transfections were performed and allowed to grow in HFF medium alone for 24 h and then harvested and fractionated to isolate purified ER/MAM and mitochondrial fractions. These purified fractions were resuspended in 10 ml of 1× PBS alone (control) or 1× PBS supplemented with 20 mM MβCD and incubated at 37°C for 1 h with shaking. Fraction suspensions were pelleted (15,000 × g for 10 min at 4°C for mitochondrial fractions and 100,000 × g for 45 min at 4°C for ER/MAM fractions) and analyzed by Western blotting.

LR purification.LRs were isolated by purification on flotation sucrose gradients as previously described (12, 79). Briefly, transiently transfected HeLa cells (2 × 10⁷ to 5 × 10⁷ cells) or HFFs (6 × 10⁷ fibroblasts) were harvested by scraping or trypsinization, resuspended in MBS buffer (25 mM 4-morpholineethanesulfonic acid and 150 mM sodium chloride [pH 6.5] containing a protein inhibitor cocktail [Roche] at 1 tablet/50 ml or 1 mM PMSF), lysed by Dounce homogenization and sonication (three 30-s bursts; Polytron), and treated with 1% Triton X-100 or 1% Triton X-114 (MBST buffer) on ice for 30 min. Following detergent treatment, permeabilized cell lysates were mixed 1:1 with a 90% sucrose solution and overlaid with discontinuous layers of 35% and 5% sucrose. Gradients were resolved by centrifugation in a Beckman SW41 rotor at 160,030 × g (36,000 rpm) for 18 h or at 187,813 × g (39,000 rpm) for16 h at 4°C. Twelve 1-ml fractions were sequentially collected starting from the top to the bottom of the tube after centrifugation. Aliquots of each fraction (20 to 30 μl) were separated by SDS-PAGE and Western blot analyses with antibodies as described above.

pUL37x1/vMIA intercalates into lipid-rich ER/MAM domains.To determine if pUL37x1/vMIA localizes to lipid-rich domains within intracellular ER/MAMs, similarly to Sig-1R, lipids were extracted from transfected HeLa cells transiently expressing pUL37x1 wt_1-163-YFP using methanol (92) prior to subcellular fractionation (Fig. 1 A). Purified ER/MAM and mitochondrial organelles were banded in discontinuous sucrose gradients during fractionation and were subsequently examined for the presence of pUL37x1/vMIA by using Western analyses. The abundance of pUL37x1/vMIA was reduced substantially from ER/MAMs but not from mitochondrial membranes following methanol extraction. Controls included YFP-transfected cells and a known LR marker, Lyn, as well as MAM components, FACL4 and GRP75, a mitochondrial marker and a cytosolic component of the MAM junction complex, IP3R3-GRP75-VDAC (82). The extraction of pUL37x1/vMIA predominantly from the ER/MAM fraction suggests that pUL37x1/vMIA localizes to lipid-rich domains therein.

• Open in new tab
• Download powerpoint

FIG. 1.

(A) Methanol extraction of pUL37x1/vMIA from ER/MAMs but not mitochondrial membranes. HeLa cells were transfected with vectors expressing pUL37x1 wt_1-163-YFP or YFP and harvested 24 h later. Cells were either treated with ice-cold methanol (+) or left on ice (−) for 10 min, prior to subcellular fractionation to purify ER/MAM and mitochondrial fractions as described previously (8, 9). Purified fractions (20 μg) were separated by SDS-PAGE and analyzed by Western analysis using anti-UL37x1 (DC35) (1:2,500), anti-FACL4 (1:250), anti-Lyn (1:250), or anti-GRP75 (1:1,000) antibodies. (B) pUL37x1/vMIA does not detectably localize to cytosolic lipid droplets. HFFs were transfected with expression vectors encoding pUL37x1 wt_1-163-YFP or Sig-1R-YFP and harvested 24 h later. Fixed cells were permeabilized, stained with Oil Red O, and imaged for YFP (green) and Oil Red O (red) using confocal microscopy. The micrographs on the left and middle are grayscale, and that on the right is the color overlay. The boxes show enlargements of the regions of interest in the cells. (C) MβCD extraction of pUL37x1/vMIA from intracellular LRs. HeLa cells were transfected with either pUL37x1 wt_1-163-YFP or Sig-1R-YFP expression vectors and harvested 24 h later. Cells were sonicated (lysed) or left intact (whole cells [WC]) and treated with 20 mM MβCD for 1 h at 37°C (11). MβCD-treated cells were fractionated to obtain ER/MAM and mitochondrial fractions as described previously (8, 9). Purified fractions (20 μg) were resolved by SDS-PAGE and subjected to Western analysis using anti-GFP/YFP (1:200), anti-DPM1 (1:100), anti-FACL4 (1:250), or anti-GRP75 (1:1,000) antibodies.

Besides HCMV pUL37x1/vMIA, the hepatitis C virus (HCV) core protein is the only other viral protein known to target MAM subdomains (71). The HCV core protein, however, also accumulates in ER-tethered, cytosolic lipid droplets (4, 55, 89). It was suggested previously that cytosolic lipid droplets are derived from Sig-1R-circumscribed LRs of the MAM by the budding of the cytoplasmic leaflet (30, 32, 84). To test if pUL37x1/vMIA, like the HCV core protein, buds into cytoplasmic lipid droplets, we examined the colocalization of wt pUL37x1/vMIA tagged with yellow fluorescent protein (pUL37x1 wt_1-163-YFP) with a cytosolic lipid droplet marker, Oil Red O dye (7, 38). pUL37x1 wt_1-163-YFP did not detectably colocalize with the cytosolic lipid droplet marker (Fig. 1B). Analogously, Sig-1R, serving as a MAM control, did not detectably colocalize with Oil Red O.

Selective MβCD extraction of pUL37x1/vMIA from internal lipid rafts.Cholesterol is a critical scaffolding component of LR domains that is necessary to maintain the integrity of the LRs as a cohesive unit (35, 64). Its removal from membranes with drugs like MβCD leads to the dissolution of LRs. To identify the methanol-sensitive lipid domains containing pUL37x1/vMIA as authentic lipid rafts, we sought to reproduce the specific depletion of pUL37x1/vMIA from internal membranes using cholesterol extraction. To that end, HeLa cells were transiently transfected with pUL37x1 wt_1-163-YFP and treated with MβCD either before or directly following cell lysis by sonication. Because MβCD does not permeate cellular membranes, treatment of whole cells dissolves only surface-exposed LRs on the plasma membrane (PM). After lysis, cells were subjected to subcellular fractionation and Western analyses to compare the abundances of pUL37x1 wt_1-163-YFP in their ER/MAM and mitochondrial fractions. MβCD treatment dramatically reduced the pUL37x1 wt_1-163-YFP abundance by 74% specifically from ER/MAMs when treatment occurred after cell lysis (compared to prelysis treatment control), consistent with the presence of pUL37x1/vMIA proteins in internal LRs of ER/MAM origin (Fig. 1C). Conversely, MβCD treatment did not detectably reduce pUL37x1 wt_1-163-YFP levels (increased ∼62%) in mitochondrial membranes. These data concur with the results obtained by methanol extraction. MβCD treatment of lysed cells analogously reduced the Sig-1R-YFP abundance (decreased 55%), consistent with its known internal LR localization (28, 30, 32). In contrast, other MAM (FACL4 [decreased 9% to 11%]) and mitochondrial/MAM (GRP75 [increased 30%]) markers were minimally affected by MβCD treatment. Together, these data establish that pUL37x1/vMIA localizes to intracellular LRs, predominantly in ER/MAMs.

pUL37x1/vMIA NH₂-terminal leader sequences mediate LR associations.To independently verify its association with LRs, we physically purified LRs from transiently transfected HeLa cells using flotation gradients. The tight packing of LRs enables them to strongly resist nonionic detergent solubilization at low temperatures, compared to surrounding nonraft membranes, which are readily solubilized (20, 68). HeLa cells expressing full-length, untagged pUL37x1/vMIA were treated with 1% Triton X-100 on ice and then layered at the bottom of a discontinuous sucrose gradient. Intact LRs, whose buoyant density caused them to float up to less dense sucrose layers from their original position, were purified in the upper fractions of the gradient. A portion of pUL37x1/vMIA was detected in DRMs (fraction 4), while the remaining pUL37x1/vMIA was detected in detergent-soluble fractions (non-DRMs), predictably localized in nonraft portions of the OMM or bulk ER, or in Triton X-100-solubilized ER LRs. The cholera toxin B subunit (CTB) was used to detect ganglioside GM1, a major component of plasma membrane LRs (26), in these whole-cell-derived DRMs. Also, Lyn, a cytosolic protein known to associate with LRs (41), was also partially detected in our collected DRM fractions. These results, while not directly differentiating the origin of observed LRs, resoundingly substantiate that pUL37x1/vMIA can associate with LRs during transient transfection.

pUL37x1/vMIA NH₂-terminal MTS residues (aa 1 to 34), including a low-hydropathy leader, ²¹SY, and PRD, are necessary and sufficient to dictate its sequential trafficking through the ER and MAM to mitochondria (27, 45, 88). To determine if this leader sequence is also sufficient for LR associations, LRs were isolated from HeLa cells expressing pUL37x1/vMIA MTS fused to YFP (Fig. 2 B). pUL37x1 wt_1-36-YFP was detected in DRM fractions of a sucrose flotation gradient. The presence of GM1 in these DRMs (fractions 3 and 4) was verified by the reactivity of CTB. Conversely, a nonraft inner mitochondrial membrane (IMM) marker, DsRed1-mito, was completely solubilized and not detected in the DRM fractions. These results establish that the pUL37x1/vMIA MTS, which dictates the protein's sequential trafficking, suffices for its LR association. Roughly half of the residues within the MTS are predicted to be imbedded within the lipid bilayer of ER, MAM, or mitochondrial membranes (47, 87, 88).

• Open in new tab
• Download powerpoint

FIG. 2.

(A) Detection of HCMV pUL37x1/vMIA in purified LRs. HeLa cells were transfected with a vector (p327) expressing untagged pUL37x1 (14) and harvested 36 h after transfection. LRs were isolated by extraction with 1% Triton X-100 on ice and flotation in discontinuous sucrose gradients as previously described (12, 79). Twelve fractions (1 ml) were collected from the top of the gradient, and aliquots (30 μl) of each fraction were subjected to SDS-PAGE and Western analysis with anti-UL37x1 (1:2,500) and anti-Lyn (1:250) antibodies. For GM1 detection, aliquots of each fraction (2 μl) were spotted directly onto the nitrocellulose membrane and reacted with horseradish peroxidase-labeled CTB protein (Sigma), followed by ECL detection. (B) The UL37x1 MTS permits its LR association. HeLa cells were transfected with expression vectors encoding pUL37x1 wt_1-36-YFP or DsRed1-mito. Twenty four hours later, whole cells were either treated with 1 μg/ml CTB (Molecular Probes) or left untreated prior to harvesting. LRs were isolated by Triton X-100 extraction on ice and flotation gradients as described above (A). Twenty microliters of each fraction was then separated by SDS-PAGE and subjected to Western analyses using anti-UL37x1 (1:2,500), anti-CTB (1:200), or anti-DsRed1-mito (1:2,000) antibodies.

Recently, detergent-specific disparities in LR extractions have been identified, dependent upon the identity of the subcellular membrane in which the LR is incident (44). As a result, LRs from intracellular membranes are more resistant to solubilization by Triton X-114 detergent than Triton X-100 extraction, to which LRs from the plasma membrane are more resistant (44). For experimental rigor, we repeated our LR extractions using 1% Triton X-114 solubilization of HeLa cells transiently transfected with pUL37x1/vMIA (Fig. 3 A). Indeed, pUL37x1/vMIA was detected at higher levels in DRMs from Triton X-114-treated cells than in those from Triton X-100-treated cells (Fig. 3B). These results are consistent with the association of pUL37x1/vMIA with internal LRs.

• Open in new tab
• Download powerpoint

FIG. 3.

(A) Detection of pUL37x1/vMIA in DRM following Triton X-114 extraction. HeLa cells were transfected with a vector expressing pUL37x1 wt_1-163-YFP or YFP, extracted with 1% Triton X-114 on ice, and isolated by flotation gradients. Twenty microliters of each gradient fraction was separated by SDS-PAGE and analyzed by Western analysis using anti-UL37x1 (1:2,500) antiserum or anti-GFP/YFP (1:200) antibody. (B) Increased detection of HCMV pUL37x1/vMIA in DRMs in Triton X-114-extracted compared to Triton X-100-extracted DRMs. HeLa cells were transfected as described above (A) and extracted in parallel with 1% Triton X-100. Equal aliquots of the gradient fractions (fractions 1 to 4) containing DRMs following extraction with 1% Triton X-100 (left) and 1% Triton X-114 (right) were resolved by SDS-PAGE in parallel and analyzed for the presence of pUL37x1/vMIA as described above (A).

pUL37x1/vMIA is associated with ER/MAM LRs.To unambiguously verify whether pUL37x1/vMIA is associated with internal LRs in the ER/MAM or mitochondria, these organelles were first purified from HeLa cells expressing pUL37x1/vMIA. DRMs were then isolated from the purified fractions, and the presence of pUL37x1/vMIA in the DRMs was examined (Fig. 4). pUL37x1/vMIA was detected in DRMs from ER/MAM fractions but not in DRMs from mitochondrial fractions. The MAM LR marker Sig-1R showed a similar association with DRM from ER/MAM but not with DRM from mitochondria. Erlin 2, a known ER LR marker (11), served to verify the ER/MAM gradient fractions containing DRMs. Conversely, mitochondrial prohibitin (53) was detected in the DRM fractions obtained from purified mitochondria. These results establish that HCMV pUL37x1/vMIA is partially associated with LRs within the ER/MAM but not in mitochondria. Furthermore, they are consistent with the results described above showing methanol and MβCD extraction of pUL37x1/vMIA from ER/MAM but not mitochondrial fractions.

• Open in new tab
• Download powerpoint

FIG. 4.

pUL37x1 associates with internal LRs. HeLa cells were transfected with a vector expressing untagged pUL37x1 wt_1-163 (p327). Thirty-six hours later, the ER/MAM and mitochondria were isolated as previously described (8, 9). LRs were then isolated from the purified subcellular fractions by cold 1% Triton X-100 extraction and flotation in sucrose gradients as described in the legend of Fig. 2A and as described previously (12, 79). Each gradient fraction (20 μl) was resolved by SDS-PAGE and analyzed by Western blotting using anti-UL37x1 (1:2,500), anti-Sig-1R (1:200), anti-erlin 2 (1:500), or anti-prohibitin (1:200) antibody.

pUL37x1/vMIA is LR associated throughout HCMV infection.To determine if pUL37x1/vMIA associates with LRs during infection, DRMs from HCMV-infected HFFs were examined at all temporal phases for pUL37x1/vMIA (Fig. 5 A). pUL37x1/vMIA was detected in DRMs isolated from HCMV-infected HFFs but not uninfected HFFs at IE (12 h postinfection [hpi]), early (24 hpi), and late (48 hpi and 72 hpi) times of infection. At IE and early times of infection, the buoyant density of DRMs with which pUL37x1/vMIA is associated is similar to that of transfected cells. At later times of infection (48 hpi and 72 hpi), the buoyant densities of pUL37x1/vMIA DRMs appear to change and are more broadly distributed. The presence of Lyn, a known LR marker protein, verified the identities of the DRM fractions from uninfected and HCMV-infected HFFs (Fig. 5B). These results suggest that HCMV pUL37x1/vMIA associates with LRs throughout infection and validate its LR association in transfected cells. These results further suggest that HCMV infection progressively alters the proteome composition of internal LRs and thereby alters their buoyant densities.

• Open in new tab
• Download powerpoint

FIG. 5.

pUL37x1/vMIA is LR associated throughout HCMV infection. HFFs were uninfected (left) or HCMV infected (BADwt at an MOI of 3) (right) and harvested at the indicated times after infection. LRs were isolated as described in the legend of Fig. 2. Fifteen microliters of each gradient fraction was resolved by SDS-PAGE and subjected to Western analysis using anti-UL37x1 antiserum (1:2,500) (A) or anti-Lyn antibody (1:250) (B).

Mapping of pUL37x1/vMIA MTS residues required for its LR association.The HCMV UL37x1 NH₂-terminal MTS is sufficient to associate proteins with LRs (Fig. 2). The pUL37x1/vMIA MTS is a bipartite leader with a weakly hydrophobic leader (MTSα) needed for ER translocation and mitochondrial import as well as proximal downstream sequences (MTSβ), which regulate mitochondrial import (27, 45, 88). To determine which MTS residues are required for LR associations, we examined pUL37x1/vMIA MTS mutants for the ability to associate with LRs (Fig. 6 A). The pUL37x1 HH mutant has a greatly increased MTSα hydropathy score and is preferentially retained in the secretory apparatus and blocked in OMM import (88). Despite this trafficking defect, both the full-length (HH_1-163-YFP) and leader (HH_1-36-YFP) fusion proteins partially associated with DRMs (Fig. 6B). Similarly, both the full-length (LH_1-163-YFP) and leader (LH_1-36-YFP) fusion proteins of mutants with marginal hydropathy scores (which are known to retain the ability for sequential trafficking to the OMM) also associated with DRMs (Fig. 6C). A pUL37x1-Tom20 leader chimera, which has the Tom20 leader in place of the pUL37x1/vMIA leader, retained the ability to associate with DRMs, as did the parental pUL37x1 wt_1-36-YFP control (Fig. 6D). Finally, a pUL37x1/vMIA MTSβ proline-rich-domain mutant (PRD_1-36-YFP), which traffics preferentially to the OMM (88), associated with DRMs. Controls included the binding of CTB to verify the presence of ganglioside GM1 in the LR fractions (21) and the IMM marker DsRed1-mito, which was solubilized and retained in the non-DRM fractions. Some variation in the total abundances of wt and MTS mutants was observed for these and previous (88) experiments, suggesting differences in overall protein stability. To quantify this variability, total cell lysates (40 μg) from HeLa cells transfected with either pUL37x1 wt_1-36-YFP or the MTS mutants were compared side by side using densitometry of a Western blot probed with mouse anti-GFP/YFP antibodies (data not shown). By calculating the MTS mutant abundances as percentages of the pUL37x1 wt_1-36-YFP total protein, they could be arranged from highest to lowest in total protein amounts as follows: HH_1-36-YFP (87%), HH_1-163-YFP (70%), PRD_1-36-YFP (37%), pUL37x1-Tom20 leader chimera (24%), LH_1-163-YFP (13%), and LH_1-36-YFP (9%). Importantly, while total protein abundances of the wt and MTS mutants differed, the efficiency in the lipid raft targeting of detected proteins appeared to be similar. These results establish that pUL37x1/vMIA domains which explicitly guide sequential trafficking to mitochondrial membranes are not required for the pUL37x1/vMIA LR association.

• Open in new tab
• Download powerpoint

FIG. 6.

pUL37x1 trafficking mutants do not block its association with LRs. (A) UL37x1 wt and mutant MTS sequences. Shown on the top are wt MTSα (aa 4 to 22) and MTSβ (aa 23 to 36). The hydrophobic leader is enclosed by the box. The mutant sequences are indicated by the arrowheads and boldface type. In UL37x1-Tom20 B, the Tom20 leader replaces UL37x1 MTSα. The predicted hydrophobicity scores (grand average of hydropathicity [GRAVY], Kyte-Doolittle scale) were calculated for the boxed residues of the wt and the LH mutant using the ProtParam application of the ExPASy proteomics server. (B to D) HeLa cells were transfected with expression vectors for HH_1-163-YFP, HH_1-36-YFP, wt_1-36-YFP, or YFP (B); LH_1-163-YFP, LH_1-36-YFP, or YFP (C); or wt_1-36-YFP, UL37x1-Tom20 B_1-36-YFP, PRD_1-36-YFP, YFP, or DsRed1-mito (D). Triton X-100-resistant LRs were extracted on ice from HeLa cells and enriched by flotation in discontinuous sucrose gradients as described in the legend of Fig. 2A and as previously described (12, 79). Twelve fractions (1 ml) were collected from the top to the bottom of the gradients. Equal volumes (30 μl) were resolved by PAGE and examined by Western analyses using anti-UL37x1, anti-Lyn, anti-GFP/YFP, anti-CTB, or anti-DsRed1-Mito antibodies.

The CBD motif in the HCMV UL37x1 MTS affects its LR association.Cholesterol is a major component of LRs, and the MAM is a cholesterol-rich subdomain of the ER (28). Consensus cholesterol binding domains (CBDs), defined as L/V-X_1-5-Y-X_1-5-K/R, in the benzodiazepine receptor (36) and Sig-1R (59) were recently characterized. We found a consensus CBD in the HCMV pUL37x1/vMIA MTS, spanning the predicted membrane-cytoplasm interface (Fig. 7). Moreover, the CBD is retained in primate CMV pUL37x1/vMIA leaders, suggesting its importance for evolutionarily conserved pUL37x1/vMIA function. These conserved domains are all positioned at the boundary of predicted α-helical regions, near the membrane interface, as required for functional folding (36). Tellingly, the chimpanzee CMV (CCMV) pUL37x1/vMIA homologue contains two overlapping CBDs.

• Open in new tab
• Download powerpoint

FIG. 7.

The pUL37x1 association with LRs is cholesterol binding dependent. (A) Conservation of the UL37x1 CBD in primate CMVs. The consensus CBDs (L/V-X_1-5-Y-X_1-5-R/K) in HCMV, chimpanzee CMV (CCMV), rhesus monkey CMV (RhCMV), and African green monkey CMV (AgmCMV) UL37x1 MTSs are indicated by parentheses and boldface type. The mutant sequences in HCMV CBD I and CBD II are indicated by the arrowheads and boldface type. (B) The consensus cholesterol binding site supports lipid raft targeting of pUL37x1. HeLa cells were transfected with vectors expressing pUL37x1 wt_1-36-YFP, CBD I_1-36-YFP, CBD II_1-36-YFP, or YFP. Twenty-four hours later, cells were harvested and LRs were isolated by Triton X-100 extraction on ice and flotation on sucrose gradients as described in the legend of Fig. 2A. Twenty microliters of each fraction was separated by SDS-PAGE and analyzed by Western blotting using mouse anti-GFP/YFP (1:200). (C) Colocalization of CBD I_1-36-YFP and CBD II_1-36-YFP with DsRed1-mito. HFFs were transiently cotransfected with vectors expressing CBD I_1-36-YFP, CBD II_1-36-YFP, or wt_1-36-YFP (green) and DsRed1-mito (red). Cells were methanol fixed at 24 h after transfection and then imaged by confocal microscopy. The panels on the left and center are grayscale. The panels on the right are the color overlay. (D) The CBD I_1-36 mutant traffics to mitochondria. HeLa cells were transiently transfected with vectors expressing wt_1-36-YFP, CBD I_1-36-YFP, or YFP. Twenty-four hours after transfection, cells were harvested, and ER/MAM and mitochondrial fractions were isolated as previously described (8). Twenty micrograms of the purified fractions was resolved by SDS-PAGE and analyzed by Western blotting using anti-GFP/YFP (1:200), anti-DPM1 (1:100), and anti-GRP75 (1:1,000) antibodies.

The pUL37x1/vMIA mutants described above (Fig. 6) retained a consensus CBD in their NH₂-terminal α-helix sequences. The chimeric pUL37x1-Tom20 leader chimera serendipitously created a consensus CBD sequence from the Tom20 leader (16L and 20Y) and the pUL37x1/vMIA MTSβ (25R).

To determine if affinity for cholesterol plays a role in the pUL37x1/vMIA association with LRs, we mutated the key tyrosine residue to alanine (Y19A) within the pUL37x1/vMIA CBD (CBD I_1-36-YFP) and examined its ability to associate with LRs (Fig. 7B). CBD I_1-36-YFP was markedly reduced in its association with DRMs (decreased 93% compared to the wt), as detected by flotation gradients. An additional mutation in the pUL37x1/vMIA CBD site (CBD II_1-36-YFP) was further reduced (decreased 99% compared to the wt) in its LR association. From these results, we conclude that pUL37x1/vMIA CBD mediates, at least in part, the LR association.

LRs can regulate the trafficking of some proteins. To determine whether the pUL37x1 CBD I_1-36-YFP and CBD II_1-36-YFP mutants, which are defective in LR associations, are also defective in trafficking, we examined their colocalization with DsRed1-mito using confocal microscopy (Fig. 7C). Both CBD I_1-36-YFP and CBD II_1-36-YFP partially colocalized with DsRed1-mito (Clontech), suggesting their ability to traffic to mitochondria. To independently verify trafficking to mitochondria, HeLa cells expressing CBD I_1-36-YFP, control wt_1-36-YFP, or YFP alone were fractionated into ER/MAM and mitochondrial fractions (Fig. 7D). CBD I_1-36-YFP was detected in proportion in the ER/MAM and mitochondrial fractions, comparable to those observed for wt_1-36-YFP. The use of ER (DPM1) and MAM/mitochondrial (GRP75) markers verified the identity and purity of the fractions. Similarly, CBD II_1-36-YFP trafficking to mitochondria was comparable to that of the wt protein, as detected by subcellular fractionation (C. D. Williamson and A. M. Colberg-Poley, unpublished results). These results demonstrate the ability of CBD mutants to traffic to mitochondria despite their notable reduction in LR associations. This finding substantiates that the LR association does not dictate pUL37x1/vMIA sequential trafficking to the OMM and suggests another functional consequence for the LR association of these proteins.