GFP Fusion Proteins: A Solution or a Problem?

Y-box binding protein1 (YB-1) is a member of the evolutionarily conserved cold shock domain (CSD) proteins and was first identified as a DNA/RNA binding protein [1] involved in the control of gene expression at both transcriptional and translational level [2-4]. Given its multiple cellular functions, YB-1 is involved in the control of severalbiological processes including cell proliferation and migration. To properly perform its functions, YB-1 subcellular localization has to be finely regulated. Specific nuclear export (NES), nuclear localization (NLS) and cytoplasmic retention signals (CRS) contribute and direct the multifunctional tasking of YB-1 [5]. In normal resting cells, YB-1 localizes to cytoplasm where it is a majorcomponent of P-bodies and messenger ribonucleoproteins (mRNPs) [6]. Recent studies link YB-1 tothe cellular response tooxidative stress and DNA repair mechanisms. Indeed, following acute oxidative stress, YB-1 localizes to cytoplasmic Stress Granules (SGs), organelle-likestructures devoid of membranes engagedin mRNA sorting and pro-survival translational reprogramming [7,8]. In particular, YB-1 is recruited in TIA-containing Stress Granules (SGs) where it functions as a component of translationally inactive mRNPs todirectly block translational initiation of highly expressed [9]. Following DNA damage, YB-1 translocates to the nucleus andassociates with DNA repair protein complexes. Here we report the characterization of a stable pool of HEK293T cellsconstitutively ex pressing YB-1 as a GFP fusion protein enabling sensitiveanalysis of YB-subcellular localization by confocal microscopy.


Introduction
Y-box binding protein1 (YB-1) is a member of the evolutionarily conserved cold shock domain (CSD) proteins and was first identified as a DNA/RNA binding protein [1] involved in the control of gene expression at both transcriptional and translational level [2][3][4]. Given its multiple cellular functions, YB-1 is involved in the control of severalbiological processes including cell proliferation and migration. To properly perform its functions, YB-1 subcellular localization has to be finely regulated. Specific nuclear export (NES), nuclear localization (NLS) and cytoplasmic retention signals (CRS) contribute and direct the multifunctional tasking of YB-1 [5]. In normal resting cells, YB-1 localizes to cytoplasm where it is a majorcomponent of P-bodies and messenger ribonucleoproteins (mRNPs) [6]. Recent studies link YB-1 tothe cellular response tooxidative stress and DNA repair mechanisms. Indeed, following acute oxidative stress, YB-1 localizes to cytoplasmic Stress Granules (SGs), organelle-likestructures devoid of membranes engagedin mRNA sorting and pro-survival translational reprogramming [7,8].
In particular, YB-1 is recruited in TIA-containing Stress Granules (SGs) where it functions as a component of translationally inactive mRNPs todirectly block translational initiation of highly expressed [9]. Following DNA damage, YB-1 translocates to the nucleus andassociates with DNA repair protein complexes. Here we report the characterization of a stable pool of HEK293T cellsconstitutively ex pressing YB-1 as a GFP fusion protein enabling sensitiveanalysis of YB-subcellular localization by confocal microscopy.

Generation of YB1-GFP Stable
Clones: HEK293T cells were tested for their sensitivity to blasticidin in order to determine the best concentration for a correct selection. Concentrations of 2,5, 8, 10 and 12mg/ml were tested. Controls start to die from the day after the addition of the higher blasticidin concentration up to 3 days for the lower antibiotic concentration. The concentration chosen for selection was 5mg/ml. HEK293T cells were seeded in 100mm dishes (Corning) and let grow to reach 80% confluence. Cells were transfected by lipofectamine with pcDNA6/V5-HISA encoding human YB-1-GFP bearing the gene for blasticidin resistance or with pcDNA3.1/His C control plasmid. After 48 hours from transfection, blasticidin was added to plates at a final concentration of 5mg/ml. Aspreviously observed, pcDNA3.1/His C transfected cells start to die after 3 daysof selection. After one week of selection cells were tripsynized and replated and after two weeks of selection stable clones were obtained.
Immunofluorescence Microscopy: 3x10 4 HEK293T, HEK293T GFP and HEK293T YB-1-GFP cells were seeded onpoly-D-lysine pre-treated coverslips (12mm thickness); the day after, cells were treated or not with H 2 O 2 (Sigma-Aldrich, St Louis, MO). At the end of treatment cells were gently rinsed with 1X PBS before adding PFA at 3,4% for 10 minutes to fix cells; after 3 washes with 1X PBS cells were permeabilized by adding 0.5% Triton X-100-PBS for five minutes; then, after 3 washes with 1X PBS, cells were blocked to avoid unspecific binding of antibodies with a 3% BSA solution for 20 minutes. Primary antibodies were added for 1hour (dilution range1:200); after incubation, three washes with 0.05% Tween PBS were performed; then Alexa Fluor 488 (Life Technologies) and Cy3 (Jackson Immuno Research, USA) secondary antibodies were added for 45 minutes in dark; after 3 washes in 0.05% Tween PBS cover-slips were immersed in a glycerol PBS solution and fixed. Direct eGFP signal was acquired without the addiction of fluorescent dyes. Images were acquired using a Carl Zeiss LSM700 equipped with an axio Observer Z1 or with a Nikon TE Eclipse 2000. Image processing and analysis were performed with Fiji (ImageJ) software.
Cell Viability Assay: Cell viability was determined by the MTT 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide assay (Sigma-Aldrich, St Louis, MO) as previously described [10]. Cells were seeded in 96-well plate at the cell density of 2 x10 3 ; the day after cells were treated or not with H 2 O 2 . After treatment, cell medium was substituted with DMEM without phenol-red (GIBCO) and without supplements; MTT solution was added to cells to produce formazan crystals. After 2hrs MTT solution was substituted by acidic isopropanol to solubilize theformazan crystals. The optical absorbancewas determined at 570nm using an iMark microplate reader (Bio-Rad, USA). The experiments were carriedout intriplicate for each experimental point and reported as percentage of the untreated control set at 1.0.
Trypan Blue Assay: Cells were seeded on poly-D-lysine pre-treated 12 well plates at 2,5 x10 5 . The day after plating, cells were treated or not; after treatment cells weregently rinsed with 1X PBS, trypsinized andcollected. An aliquot was diluted 1:1 with trypan blue (Sigma-Aldrich, St Louis, MO). The experiments were carried out in triplicate for each experimental point.
Cell Proliferation Analysis: Cells were seeded on poly-D-lysine pre-treated 12-well plates at 2,5 x10 5 . Every 24 hours cells were gently rinsed with 1X PBS, trypsinized and counted in a Burker chamber. The count was confirmed with analysis by a Scepter 2.0(Millipore). The experiments were carried out in triplicate for each experimental point.
Immunoblot Analysis: Cells were seeded at 60% confluence (1.5 × 10 6 ) in 100mm plates; the day after, plates were gently rinsed with 1X PBS, and collected in Lysis Buffer (50mM Tris-HCl pH 7.5, 5mM EDTA, 150mMNaCl, 1% NP-40, 1Mm phenylmethylsulphonyl fluoride, 0.5% sodium deoxycholate, and protease inhibitors). Cells were then left on ice for 30 minutes, and crude extracts were obtained following centrifugation at 13200rpm for 30 minutes at 4°C. Theamount of protein in the samples was determined by the Bio-Rad protein assay (Bio-Rad, Milan, Italy). After the addition of Laemmli buffer (Sigma Chemical Co, St. Louis, MO, USA) the samples were boiled at 100°C for 5min and resolved by SDS-polyacrylamide gel electrophoresis (SDS-PAGE). The proteins werethen transferred to a polyvinylidene difluoride membrane (PDVF, Millipore) using a Mini trans-blot apparatus (Bio-Rad, Milan, Italy) according to the manufacturer's instructions. The PVDF membrane was blocked in 5%w/v milkbuffer (5%w/v non-fat dried milk, 50mMTris, 200mM NaCl, 0.2% Tween 20) and incubated with primary antibodies diluted in 5% w/v milk or bovine serum albumin (BSA) buffer overnight at 4°C. The blots, washed three timeswith TTBS (Tris-bufferedsaline, 0.1%Tween), were incubated for 1 hour atRT with HRP-conjugated secondary antibodies (SigmaAldrich). Proteins were visualized by an enhanced chemi luminescence method (ECL, GE-Healthcare) and analyzed by Quantity One W software of Chem-iDoc TM XRS system (Bio-Rad, Milan, Italy).

Conclusion
Stable cell lines represent a permanent resource that can be stored under cryogenic conditions for long periods of time, retrieved, and cultured to provide a consistent and reliable level of sustained gene expression. To explore the role of YB-1 localization upon stress stimuli we sought to generate a stable pool of cells constitutively expressing YB-1as a GFP fusion protein. Preliminarily, we transiently transfected HEK293T cells with YB-1-GFP cDNA cloned in pcDNA6/V5-His, a vector derived from pcDNATM3.1(+) (Invitrogen). 5.0x105 cells were seeded in order to reach 80% confluence the day after; lipofection was performed using pcDNA6/V5-HisA-YB-1GFP and peGFP-C1 empty vector as control [10]. 48hours after transfection cells were analyzed by immuno blot to evaluate the level of YB-1-GFP fusion GFP exogenous expression. Transfected YB-1-GFP was highly expressed inHEK293T cells (Figure 1). Direct fluorescence, acquired by Nikon Eclipse 800 microscope, showed a prevalent cytoplasmic localization for YB1-GFP ( Figure  2). Moreover, viability of YB-1-GFP cells, as assessed by MTT assay, was undistinguishable from HEK293T parental cells and cells expressing GFP alone (data not shown). Next, we produced pools of clones of HEK293T stably expressing YB-1-GFP or GFP alone. YB-1-GFP expressing cells were selected by blasticid in while G418 was used for selection of peGFP-C1 stable pool. After 4 weeks, 91% of selected cells were GFP positive. Next, we analyzed the proliferation rate of YB-1-GFP expressing pool. Compared to the parental cell line, YB-1-GFP expressing cells grew slightly faster (Figure 3). To examine YB-1 protein localization under oxidative stress we treated HEK293T YB-1-GFP cells with hydrogen peroxide (H2O2). Byimmunofluorescence, we could observe typical stress granules (SGs) in control and GFP expressing cells, using a canonical SGs marker, PABP-1 Figure 4  Remarkably, although YB-1-GFP cells grew faster than control cells, their sensitivity to oxidative stress was increased as demonstrated by MTT and try pan blue assays ( Figure 6). Taken together, our data show that, even though the GFP fusion protein behaved like the endogenous one and is localized in the right sub cellular compart-ment, the presence of the huge GFP tag causes a lack in functionality that can be observed only in particular experimental conditions. Therefore, insights obtained so far with YB-1-GFP construct should be critically revised and, more in general, we should beware and be careful in interpreting data obtained with functionally uncharacterized GFP fusion proteins.    Table 1).   Table 1).  Table 1).