The effect of phosphocreatine on the distribution of different subunits of translation elongation factor 1 in gently permeabilized human fibroblasts

Highly effective permeabilizationltranslation system developed recently (Negrutskii et al (1994) Proc. Nat Acad. Sci. USA 91, 964—968} was used to study the distribution of FIT С-labeled a and д subunits of translation elongation factor 1 in human fibroblasts. Somewhat granular fluorescent picture observed for the both proteins distribution during active protein synthesis may be indicative of putative protein synthesis compartments. Exogenous phosphocreatine, an absolutely required agent to support the protein synthesis in permeabilized cells, is shown to influence the EF-la distribution, while not affecting the EF-16. Most of EF-la under «low energy» condition is found in nucleus. Thus, the possibility of the functional change in EF-la distribution is demonstrated supporting the idea about transient character of EF-la involvement into «heavy» EF-1 complex.

Introduction.The protein synthesis is thought to be the most complex and highly organized metabolic process in mammalian cell.Recently a number of evidence have been presented supporting the existe nce In vivo of structural and functional compartmentalization of the translation machinery.The chan neling of aminoacyl-tRNA during protein synthetic process in vivo [1,2] and sequestering of cellular aminoacyl-tRNA [3] are the functional signs of the compartmentalization.The line of structural evidence involves detection of high-molecular aminoacyl-tRNA synthetases [4,5] and elongation factors [6,7] complexes, as well as the cytoskeletal association of translation factors [8], ribosomes [9,10], mRNA [11 ] and aminoacyl-tRNA synthetases [12,13].
The various protein synthesis components have been found recently to be co-localized in granules inside highly specialized glial cells [14].The de tergent-extracted oligodendrocytes were used to visu alize protein (aminoacyl-tRNA synthetase and trans lation elongation factor EF-la) and nucleic acid (rRNA and mRNA) components by immunofluo rescence and by fluorescent in situ hybridization correspondingly.Such a co-localization has been in terpreted as a potential demonstration of the protein © B.

S. NEGRUTSKII, 1997
synthesis compartments in eukaryotic cell though the functional role of such granules has not yet been verified [14].
In this work we have used somewhat different experimental approach to look at possible functional importance of the structural organization of putative protein synthesis compartments in cell.The fluorescently labeled and functionally active components of the protein synthetic machinery (namely a and д subunits of translation elongation factor EF-1) were used as markers of supposed compartments.The proteins were deposited into the gently permeabilized human fibroblasts which were able to continue mRNA translation with the high efficiency.The cells were incubated under the conditions supporting or exclu ding the effective protein synthesis.After fixation the distribution of labeled EF-1 subunits in cell under both conditions w T as viewed using fluorescent micro scopy.
Materials and Methods.Materials.Human EFla and EF-1<5 were kindly provided by J. Kriek (Leiden University).
Buffer NPS: buffer S containing 5 mM ATP, 5 mM glucose and 6 mM magnesium acetate.
Cell culturing.Primary human fibroblasts VH25 from human foreskin (15-18 passages) were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10 % fetal calf serum.After trypsinization they were plated at density of 10-20 % on sterilized glass coverslips in the culture dishes (3 coverslips per dish).Cells were allowed to settle overnight and then were either used for expe riment immediately or incubated for five more days in DMEM containing 0.5 % fetal calf serum to enter G 0 phase of cell cycle.
Preparation of FITC-proteins.Fluorescein isothiocyanate (FITC) was dissolved in DMSO at con centration 1 mg/mi.a or 6 subunits of EF-1 were dialyzed against FITC buffer (final concentration 2 mg/ml).14 jul of dye were mixed with 140 /Л of protein solution and were incubated two hours at room temperature.Then 2 M NH 4 C1 (final con centration 50 mM) was added to quench the reaction and the proteins were separated from free dye by microcolumn Sephadex G-25 chromatography.To test the absence of free FITC in FITC-EF-1 preparations the portions of G-25-purified FITC-EF-la and FITC-EF-1 б were subjected to SDS-PAGE.No fluorescence was found in low-molecular region.
Cell permeabilization.All procedures were con ducted at 37 °С Cells grown on coverslips were washed two times with 2 ml phosphate-buffered saline and once with 2 ml S buffer per dish. 1 ml of saponin in buffer S (final concentration 125/^g/ml) was added to each dish and cells were incubated 6 min.Then saponin was aspirated and cells were carefully washed twice with 2 ml PS or NPS buffer depending on further incubation.PS buffer was designed to effec tively support the protein synthesis in permeabilized cells [2], while NPS buffer was the same buffer without creatine phosphate.The latter is show y n to be the absolutely required component of the protein synthesis mixture and its withholding leads to the complete stop of the mRNA translation [2 ].
FlTC-protein labeling of the cells.10 ml of FITC-protein (~ 0.5 mg/ml) were mixed with 1.2 /Л of PSW buffer, 12 pi of ovalbumin in PS or NPS buffer (final concentration 2 mg/ml) and 27.6 /Л of PS or NPS buffer.40 {4-І of mixture were gently applied on the coverslip with permeabilized cells.The incubation was continued at 37 °С for 7 min unless indicated otherwise.The coverslips were washed with PBS and 0.5 ml of 3.8 % formaldehyde was carefully applied per coverslip.Cells were left for 15 min at room temperature.Fixed cell were washed thoroughly with PBS, and excess of liquid was carefully aspirated by filter paper.5 fil 2.3 % DABCO (1,4-di-azabicyclo-(2,2,2)-octane) anti-bleaching agent were ad ded per coverslip.The coverslips were mounted at object glasses with nail polish.Slides were observed with a microscope equipped for fluorescence micro scopy.The level of accumulation of FITC-EF-la in the nucleus was quantitated as described earlier [15].
[ 3 H]GDP binding test.Increased amounts of EF-la or FITC-EF-la were incubated in the buffer A containing 4 {4-М [H 3 ]GDP (total volume 20 /Л) for 5 min at 37 °С.Then samples were put on ice, diluted with 100 {Л of the ice-cold same buffer without BSA.EF-la • [H 3 ]GDP complexes were absorbed on nitro cellulose filters (pore diameter 0.4 {im) and washed twice with 200 /лі of the same buffer without glycerol.
Protein synthesis in permeabilized cells.To test cell protein-synthetic activity in situ the permeabilized G 0 -fibroblasts grown on the coverslips were incubated in miniature Petri dishes (one coverslip per dish) in 75 ,ul PS buffer containing 0.4 mM GTP, 30 /tg/ml creatine phosphokinase, 240/^M [ 35 S ]methionine and the 19 amino acids mixture (methionine omitted), 240 /лМ of each, at 37 °С.At the end of incubation 1 ml 10 % hot TCA was added to each dish to stop the reaction.Then TCA-insoluble suspension was eva cuated into glass tubes, boiled 15 min and absorbed on Whatman GF/C filters.The coverslips were trans ferred to plastic scintillation vials and boiled for 15 min in 0.2 ml of 10 % TCA.Samples were treated according to the procedure for filter or liquid/liquid radioactivity counting correspondingly.Data for each time point, corrected for zero point radioactivity, represent the summation of the corresponding coverslip and filter radioactivities.
Results.Adjustments of the permeabilization-protein synthesis system.Three main prerequisites to study the functional distribution of fluorescently labe led EF-1 in permeabilized cells are: i) the deve lopment of very gentle procedure to permeabilize the cell membrane without perturbing the internal struc ture of cell ii) high protein synthetic activity of the permeabilized cell Hi) low-level labeling of EF-1 subunits with fluorescent agent.AH three precon ditions were shown to be fulfilled in this study.
First, since the gentle permeabilization method has been developed for CHO cells in suspension [2j it was necessary to find the conditions rendering the human fibroblasts completely permeable in situ.Wide range of the saponin concentration (30-500 jug/ml) was tested.125 //g/ml concentration of the detergent is determined to make 100 % cells accessible to trypan blue without visible destruction of cell mem branes.
Second prerequisite to study the functional EF-1 localization in cell was to prove the high protein synthetic activity of the peraieabilized fibroblasts.Linear incorporation of [ 35 S ]methionme into hot TCA-insoluble product for at least 15 min was demonstrated (Fig. 1), Thus, the EF-1 localization study (incubation time 8 min) was performed during active protein synthesis in cell.
Third, the conditions of FITC-labeling of the EF-1 subunits were adapted to preserve the factor activity in the protein synthesis.Namely, pH of the labeling buffer was decreased to 8,2 and incubation time at room temperature was reduced to 2,5 hours to modify the conventional procedure [15,16].Since EF-1 a is the most labile protein in the EF-1 complex and contains much greater amount of lysines per molecule in comparison with other subunits of EF-1 the effect of FITC labeling on the EF-1 a activity was tested.No marked difference was found in the activity of FITC-EF-la and non-labeled EF-la in [ 3 H]GDP binding (data not shown).
EF-la and EF-ld localization in human fibro blasts.Since the ATP-regenerating system was shown to be an absolute requirement for the protein syn thesis in permeabilized cells [2 ] it was of interest to find out whether distribution of different subunits of EF-1 in permeabilized cells depends on the energy supply.Fig. 2 demonstrates the FITC-EF-la distri bution in sapoixm-permeabilized G 0 ~human fibroblasts in the presence of creatine phosphate (A) or without it (B).The major portion of EF-la is observed in the nucleus in the absence of creatine phosphate (Fig. 2,  B).Fig. 3 demonstrates time-course of the FITC-EFla entry into the nucleus of permeabilized cell.Much less amount of the factor is detected in the nucleus if ATP-regenerating system functions actively (Fig. 2,  A).Contrary to EF-la, the EF-ld localization in permeabilized G 0 human fibroblasts was not influ enced by the presence of creatine phosphate in the incubation mixture (Fig. 4, A and B), Somewhat granular distribution throughout the cytoplasm is found for both proteins being is more evident in the case of EF-1<3.
The controls.To decrease the unspecific absorp tion of the FITC-labeled EF-1 subunits inside per meabilized cells nearly 5-fold excess of ovalbumin over EF-1 was included in the permeabilization me dium.Ovalbumin has similar to EF-la molecular weight (43 kDa) allowing to penetrate the nucleic membrane.FITC-labeled ovalbumin is found disper sed evenly throughout the permeabilized fibroblasts ( [16], our unpublished observation).It should be noted that addition up to 6 mg/ml ovalbumin and albumin into the incubation mixture had virtually no effect at the overall picture of FITC-EF-1 subunits distribution (data not shown).To test the functional importance of the EF-1 localization observed the inactivated FITC-labeled protein was used as a con trol.The incubation of EF-la without glycerol is known to cause fast and unrecoverable loss of GDPbinding activity of the factor [18].Therefore, FITC-EF-la was inactivated by 6 hours incubation in glycerol-free buffer (40 mM HEPES, pH 7.5) at room temperature.The inactivation of EF-la caused the less retention of the protein in cell (Fig. 2, C).The distribution of inactive FITC-EF-la in cell incubated with the complete energetic mixture was shown to be strictly nuclear (Fig. 2, C), resembling to some extent that of active FITC-EF-la under the «low energy» conditions (Fig. 2, B).So, certain specificity of the energy-dependent re-distribution of FITC-EF-la in human fibroblasts may be suggested.
Discussion.The study of the intracellular loca lization of a protein may be escorted by the variety of artifacts.The severe methods of cell permeabilization used in conventional procedures may lead to the re-distribution of proteins into uncharacteristic cel lular compartments [16 j.Therefore, one should take care to avoid as much as possible the destruction of  ryotic cells [2] suits well to this purpose.On the other hand, the permeabilized ceil system in which the protein synthesis persists with the activity indis tinguishable from that of intact cells 12] gives a good opportunity to study the location of the protein synthesis apparatus in situ in the conditions matching its native surrounding in cell.
The G 0 -arrested human fibroblasts were used for the experiments.The protein synthesis in the quies cent cells is slowed down significantly while the pattern of proteins synthesized are quite similar to that in the growing cells [17].Thus, G 0 cells may be taken as a reference point to portray low functional level of the translation machinery.The protein syn thetic process was shown to proceed linearly in permeabilized G 0 cells for at least 15 min after supplementing with the necessary components (Fig.

І).
Somewhat granular EF-1 localization is observed in permeabilized cells which continue to synthesize proteins (Fig. 2, A, 4, A), Most of fluorescence is found throughout the cytoplasm.It is noteworthy that granular appearance of EF-ld in human fibroblasts has been demonstrated also by immunofluorescent technique [19].These data appear in agreement with xent finding of the large granules containing .RNA, rRNA, EF-la and arginyl-tRNA synthetase L oligodendrocytes [14 J. it was suggested that these ~ains represent actual protein synthesis compartents [14 J.Here we.showed that the involvement of іе essential component of protein synthesis machiіту, EF-a, in supposed protein synthesis compartents may depend on the efficiency of translation rocess in cells (Fig. 2, A, B).To our knowledge, this the first demonstration of the energy-dependent ^-distribution of the translation elongation factors in ikaryotic cell.Pearlier, it was found that inactivation : EF-2 itself by diphtheria toxin or inhibition of dlular protein synthesis by ribosome inhibitor, ricin, id not change the distribution of EF-2 in human broblasts [8 j.
Certain correlation could be found between the istribution of EF-2 in G 0 and cycling cells measured у immunofluorescence [2 j and the distribution of F-la in permeabilized G 0 cells under low and high :vel of protein synthesis conditions (Fig. 2) corres-3ndingly.It cannot be excluded that such a reistribution of protein synthesis components may reflect some rearrangement of protein synthesis com partments under low functional level of the translation machinery.
Some non-specific fluorescent staining of the nucleus was observed for any FITC-labeled protein studied which molecular size fitted to the size of the nuclear membrane pore (data not shown).FITC-EF-1(5 (29 kDa) and FITC-ovalbumin (43 kDa) location in permeabilized cell demonstrate no creatine phos phate dependence while the presence of ATP-regenerating mixture which is of crucial importance for protein synthesis in permeabilized cells [2] caused the noticeable exclusion of EF-la from the nucleus (Fig. 2, A, B).This fact may be connected also with the functioning of ATP-dependent system of nuclear transport [5].
However, this system is known to implicate mainly large proteins which cannot pass through the nuclear membrane passively [15].The time-depen dent and reversible character of the FITC-EF-la presence in the nucleus of permeabilized cell (Fig. 3) suggests that the accumulation of EF-la does not involve strong intranuclear binding.The nuclear loca lization of EF-la has been demonstrated indepen dently by immunofluorescence and immunoelectron microscopy [14,19] and was considered to represent other than translational functions of EF-la.On the other side, it cannot be excluded that EF-la plays a role in transfer of newly synthesized tRNA from nucleoli to translational compartments.Discovered recently ability of EF-la • GDP to interact with uncharged tRNA [21 ] makes such a suggestion even more likely.
The difference observed in the phosphocreatine dependence of EF-la and EF-ld localization seems to be consistent with the commonly accepted view on the EF-1 complex [22].According to the model, EF-ld is the part of EF-1 Дуд complex associated with cellular structures while EF-la interacts with the complex transiently during active protein synthesis.Consis tently, similar distribution of a and д subunits of EF-1 was observed under active protein synthesis condition while significant re-arrangement of EF-la localization was found under the translational arrest (Fig. 2).EF-ld localization appeared alike under the both conditions (Fig. 4).
To substantiate the protein synthesis effect on EF-1 localization we attempted to modulate the trans lation restriction by other than the energy depletion means, namely, by addition of protein synthesis inhibitor, cycloheximide, to the perforated cells.Unfortunately, the addition of cycloheximide to per meabilization mixture was shown to be accompanied by noticeable destruction of perforated cells (data not shown) which impeded correct protein localization measurements.

Fig. 2 .
Fig. 2. The distribution of FITC-EF-la in permeabilized human fibroblasts: A cells were incubated in complete protein synthesis mixture; В ~~ in the same but phosphocreatine mixture; С -dis tribution of inactivated FITC-EF-la in cells studied in complete protein synthesis mixture

Fig. 3 .
Fig. 3. Time-course of FITC-EF-la entry into the nuclei of permeabilized human fibroblasts incubated in complete protein synthesis mixture without phosphocreatine.Cells were treated as described in Materials and Methods, fixed at times indicated, washed and photographed.The intensivity of fluorescence at different time-points was calculated by scanning densitometry of the nuclei.No noticeable changes in the intensivity of cytoplasmic fluorescence was found in the interval between one to thirty minutes

g. 4 .
The distribution of FITC-EF-1<5 in permeabilized fibroblasts complete protein synthesis mixture (A) or in the same mixture thout phosphocreatine (B)

. Негруцкий Влияние фосфокреатина на распределение различных субъединиц фактора элонгации трансляции 1 в пермеабилизованных фибробластах человека Резюме Фосфокреатин абсолютно необходим для обеспечения белково го синтеза в бесклеточных системах и в пермеабилизованных клетках. Показано, что локализация EF-la в пермеабилизо ванных клетках высших эукариот зависит от присутствия фосфокреатина. в среде. Распределение EF-la в цитоплазме
The attempt to get a definite answers whether phosphocreatine effects EF-la distribution through inhibition of protein synthesis and what is the main reason of EF-la nuclear localization will be a subject of further investigation.I wish to thank Prof. Wim Moller and Dr. Jan Dijk for invaluable help in performing the inves tigation.I am indebted to Dr. Romas Stapulionis for sharing recent achievements in permeabilization tech nique.The support of short-term FEBS fellowship program is appreciated.Partial support of the Minis try of Science and Technology of Ukraine (grant 5.2/130 and 5.4/73) is acknowledged.