Far western blot as a tool for identification of interaction between ribosomal proteins and translation elongation factor 1A

We have used Far Western blot technique to identify several proteins of 60S ribosomal subunit from rabbit liver which are capable of interacting with homologous translation elongation factor 1A. They belong to phosphoprotein Ð0.Ð1/Ð2 complex situated near A site region of 80S ribosome. It was also shown that eEF1A interacts with several proteins of small subunit of 80S ribosomes. It is important since none of 30S subunit proteins binds elongation factor EF-Tu. Identification of the above-mentioned proteins is the aim of future investigations.

In tro duc tion. Sys tem atic in ves ti ga tion of prokaryotic ri bo somes in the course of re cent decades re sulted in the ac cu mu la tion of an enor mous mas sive of knowl edge on the struc ture and func tions of such mo lec u lar con vey ing ma chines. Par tic u larly, the mech a nisms of de cod ing [1] and peptidyl transferase re ac tion [2][3][4] were stud ied for 70S ri bosomes, in clud ing the de tailed study on the atomic level.
The main ef forts in re gards of eukaryotic 80S ribo some are di rected to the in ves ti ga tion of trans la -tion ini ti a tion -a more com plex pro cess, com pared to sim i lar stage of biosynthesis of pro tein in bac te ria (over 30 polypeptides par tic i pat ing in eukaryotic initi a tion vs. 3 bac te rial fac tors have been dis cov ered). Un like the case of ini ti a tion, gen eral scheme of transla tion elon ga tion cy cle is very sim i lar in pro-and eukaryotes. The ex is tence of three tRNA bind ing sites was proven for ri bo somes of Esch e richia coli [5], yeasts [6], and rab bit liver [7]. Be sides, allosteric link be tween A and E ri bo somal sites, ini tially discov ered in bac te rial sys tem [5], was dis cov ered in yeast ri bo somes as well [6]. Note wor thy is the fact that our ideas on struc ture and func tions of

STRUCTURE AND FUNCTIONS OF BIOPOLYMERS
eukaryotic ri bo some are mainly founded on the exper i men tal data on bac te rial and archaea ri bo somes. At the same time eukaryotic ri bo some is 50% larger than its bac te rial an a logue, it con tains ad di tional proteins and nu mer ous in ser tions into rRNA se quence. Pre lim i nary re sults of in ves ti ga tion on the struc ture of 80S ri bo some by the method of cryo-elec tron micros copy [8] re veal dif fer en tial conformational dynam ics of pro-and eukaryotic ri bo somes. Also, the conformational changes in 80S ri bo some dur ing trans la tion elon ga tion are dis tinct, un like the case of 70S ri bo some. The study on the lo ca tion of ma trix in side hu man pla centa ri bo some, car ried out us ing the method of chem i cal cross-link ing [9], re vealed the con tent of mRNA-con tain ing chan nel to dif fer in hu man and bac te rial ri bo somes (in the for mer case mRNA-con tain ing chan nel con sisted of pro teins and rRNA, in the lat ter one -of rRNA only).
Along with func tions, con nected with as so ci a tion of ri bo some and trans la tion of mRNA, many ri bo -somal pro teins are spe cific for func tions not as so ciated with ri bo somes. Be ing out side of ri bo some, they take part in var i ous cell pro cesses, for in stance, splic ing reg u la tion of its own pre-mRNA (L30 of Saccharomyces cerevisiae, hu man S13 and S26) [10] or the de vel op ment of vi ral in fec tion (pro teins L18 and L24 bound with transactivator of cau li flower mo saic vi rus, i.e. pro tein, reg u lat ing trans la tion re-ini ti a tion on IRES of polycistronic mRNA) [11].
Tu mour cells are known for the changed level of gene ex pres sion of var i ous ri bo somal pro teins, e.g. large in tes tine can cer cells are spe cific for the increased gene ex pres sion of a num ber of ri bo somal pro teins, namely S3, S6, S8, S12, S19, P0, L5, L18, L31, and L37; dur ing liver can cer the sim i lar pic ture is ob served for gene of ri bo somal pro tein P0; in lymphoid tis sue tu mours these are S3a, S4, S17, L41; dur ing pros tate can cer -genes en cod ing pro teins L7a and L37 [12]. The ex act rea sons as well as cell mecha nisms in the ba sis of in creased gene ex pres sion in ribo somal pro teins in the course of carcinogenesis have not been de ter mined.
Eukaryotic trans la tion elon ga tion fac tor 1A (eEF1A) be longs to a large fam ily of GTP-bind ing pro teins, the rep re sen ta tives of which take part in the reg u la tion of cru cial cell pro cesses and are spe cific for high con ser va tism of pri mary amino acid sequences [13]. eEF1A ca tal y ses GTP-bind ing of aminoacyl-tRNA (aa-tRNA) to ri bo somal ac cep tor (A) site.
Ter nary com plex of eEF1A*GTP*aa-tRNA binds to GTPase as so ci ated cen tre of ri bo some, which in cludes pentameric com plex of L10.L7/L12 (eukaryotic homo logues -phosphorylated pro teins P0.P1/P2), L11 (eukaryotic homologue -eL12), and con ser va tive sarcin/ricin rRNA loop [14] (Fig.1). Pro duc tive codon-anticodon rec og ni tion ac cel er ates GTP hy dro ly sis by eEF1A mol e cule and dis so ci a tion of eEF1A in GDP-bound form leads to ac com mo dation of tRNA in A site. The views on the se quence of events in the course of aa-tRNA bind ing as a part of ter nary com plex with eEF1A*GTP as well as on the list of par tic i pants of this pro cess are rather con fine at the ab sence of struc tural data on eukaryotic de cod ing com plex as these ideas re side on the sim i lar i ties on de cod ing com plex of E. coli [15]. There fore, the lack of in for ma tion on the or gani sa tion and func tion ing of eukaryotic ap pa ra tus of trans la tion elon ga tion is ev ident.
Due to the fact that ob tain ing of in for ma tion on struc tural or gani sa tion of 80S ri bo some of higher eukaryotes re mains the hard est ex per i men tal task, the cen tral role in the study on eukaryotic ri bo somes is given to bio chem i cal meth ods of in ves ti ga tion. The lat ter sup ply to the bulk of data, com ing to be help ful later for in ter pre ta tion of re sults, ob tained using struc tural meth ods. The aim of cur rent work is to iden tify eukaryotic ri bo somal pro teins, ca pa ble of inter act ing with ho mol o gous trans la tion elon ga tion fac tor eEF1A us ing Far West ern blot tech nique. The data of this method pre sup pose that the main 60S proteins of ri bo somal sub units, in ter act ing with eEF1A, are phos pho pro teins P.0P1/P2, com pris ing GTPase as so ci ated cen tre of ri bo some. More over, our data tes tify that 4 pro teins of mi nor sub unit are ca pa ble of par tic i pat ing with eEF1A.
Fur ther iden ti fi ca tion of 40S sub unit pro teins by mass-spec trom e try may pos si bly al low de ter min ing the role of these in ter ac tions dur ing trans la tion or other cell pro cesses, in volv ing both eEF1A [16] and ri bo somal pro teins.
Ma te ri als and Meth ods. Iso la tion of ri bo somal sub units and eEF1A. Adult rab bit liv ers were used as the source of prep a ra tions of 60S and 40S sub units. The method of ob tain ing ri bo somal sub units, free from en dog e nous tRNA and mRNA, is based on scheme [17] with some mod i fi ca tions. In par tic u lar, to elim i nate endogenic RNAses from the polysomal prep a ra tions at the early stage of pu ri fi ca tion, post-mi to chon drial supernatant was cen tri fuged through 1 M su crose cush ion, con tain ing 0.5 M of KCl. Con cen tra tions of sub units were de ter mined on the ba sis of the fol low ing cor re la tions: 60 pM/1 op tic unit of A 260 for 40S and 30 pM/1 op tic unit of A 260 for 60 S. eEF1A prep a ra tion was ob tained com bin ing gel-fil tra tion and ion-ex change chro ma tog ra phy, as ear lier de scribed in [18]. Pro tein con cen tra tion was de ter mined ac cord ing to Brad ford's method [19].
Far West ern blot tech nique. The method is based on the abil ity of pro tein part ners to in ter act on the sur face of nitrocellulose mem brane [20]. Af ter electro pho re sis at the denaturating con di tions, gel was in -cu bated in re-naturating buffer (50 mM tris-HCl, pH 7.4, 20% glyc erol) for 1 hour at room tem per a ture to elim i nate SDS and con se quently for par tial re-naturation of ri bo somal pro teins [21]. Electrotransfer of pro teins from gel to nitrocellulose mem brane was car ried out in car bon ate buffer (10 mM NaHCO 3 , 7 mM Na 2 CO 3 , pH 9.9, 20% eth a nol) at 30-40 Wt in a cold room (4°C) in the course of a night. Coomassie bril liant blue and Ponceau S were ap plied on gel and mem brane re spec tively in or der to con trol the trans fer. The next stage was the in cu bation of mem brane with eEF1A in the con cen tra tion of 25-50 nM in PBS buffer (1.7 mM KH 2 PO 4 , 5.2 mM Na 2 HPO 4 , 150 mM NaCl), 3% BSA or 5% of dried milk, 15-20% of glyc erol at room tem per a ture for 1 hour con stantly stir ring. The use of high ionic force buffer so lu tion al lowed avoid ing pos si ble non-specific elec tro static pro tein con tacts on mem brane. Pro teins, ca pa ble of in ter act ing with eEF1A, were de ter mined by monoclonal anti-eEF1A an ti bod ies (Up state Group Inc.) (1:1000).
For sec ond time hy bridi sa tion with an ti bod ies against C-ter mi nal do main of yeast ri bo somal proteins P0.P1/P2 (kindly pre sented by Pro fes sor J. Ballesta, Spain) the mem brane was ad di tion ally pu rified from ear lier bound an ti bod ies by 30 min in cu bation at 60°C in the buffer (2% SDS, 100 mM â-mercaptoethanol, 62.5 mM tris-HCl, pH 6.8). Then the mem brane was treated with pri mary anti-P0.P1/P2 an ti bod ies (1:3 di lu tion) and ex posed with sec ond ary anti-mouse HRP con ju gate.
Re sults and Dis cus sion. De ter mi na tion of ri bosomal sub unit 60S pro teins, ca pa ble of in ter act ing with eEF1A. Far West ern blot was men tioned above to be based on ob ser va tion of pro tein part ners in teract ing with each other on the sur face of nitrocellulose mem brane af ter elec tro pho re sis of pro tein mix ture at denaturating con di tions, renaturation of pro teins in gel and electrotransfer (see Ma te ri als and Meth ods). Thus, we per formed SDS-PAAG elec tro pho re sis of 60S sub unit pro teins (Fig.2, a). Then the pro teins were par tially renaturated and trans ferred to the mem brane, in cu bated with eEF1A mix ture, in or der to al low this pro tein to form a com plex with ri bosomal pro teins. Next the mem brane was thor oughly washed from un bound eEF1A. eEF1A mol e cules, which formed com plexes with ri bo somal pro teins and thus re mained on mem brane, were local ised using anti-eEF1A an ti bod ies. Pos i tive sig nal was received for 3 pro teins (Fig. 2, b). On the ba sis of lit era ture data it is pos si ble to sup pose that P0.P1/P2 com plex pro teins -eukaryotic homo logues of prokaryotic pro teins L10.L7/L12, as so ci ated with GTP as so ci ated cen tre on 50S sub unit -are the primary can di dates for the role of pro teins part ners for eEF1A. The men tioned cen tre is spe cific for the bind ing of ter nary com plex with ri bo some dur ing decod ing [15].
A se ries of Toshio Uchiumi's works [14,22] on in vi tro re con struc tion of a part of GTP as so ci ated cen -tre, per formed on bac te rial ri bo some, re vealed the change of bac te rial L7/L12 stem for ho mol o gous struc ture of rat liver to re sult in the loss of ca pa bil ity to re cog nise prokaryotic EF-Tu (bac te rial homologue of eEF1A) and EF-G (bac te rial homologue of eEF2) by chi me ric 70S ri bo somes, and their GTPase and translational ac tiv ity could be compared with ac tiv ity of na tive eukaryotic 80S ri bosomes [22].
The data of cross-link ing method [23], along with di rect visu ali sa tion of bac te rial de cod ing com plex by cryo-elec tron mi cros copy [15] show that pro teins L10 and L7/L12 are the pri mary and, most likely, the only pro teins of ma jor sub unit, in ter act ing with EF-Tu. In re gards of eukaryotic eEF1A, cross-linking method, ap plied to eukaryotic sys tem ri bosome-fac tor, did not re veal any def i nite re sults compared to the case of EF-Tu [24]. The au thors iden tified 3 pro teins of 60S sub unit (L12, L23, and L39) as those in ter act ing with eEF1A. How ever, it is worth men tion ing that all these pro teins are local ised on the in ter nal side of ma jor sub unit, di rected to wards minor sub unit, which is sup posed to re sult in sig nif i cant steric ob sta cles for in ter ac tions with eEF1A.
Thus, the re li abil ity of re sults ob tained is rather ques tion able. It is hard to com pre hend why sim i lar method us ing the same bi-func tional cross-link ing agent 2-iminotiolane in the case of 70S ri bo somes and EF-Tu pro vided def i nite re sult, cor re lated with the data, ob tained us ing other meth ods [23], whereas in the case of 80S/eEF1A the con clu sion pub lished [24] is most likely to be an ar ti fact. It is pos si ble that the same cross-link ing agent can not be used for proand eukaryotic ri bo somes due to dif fer ent na ture of avail able cross-links of 70S and 80S ri bo somes and elon ga tion fac tors.
It was pos si ble that the pro tein eEF1A part ners, dis cov ered by us on the ma jor sub unit of eukaryotic ri bo somes (Fig.2, b) could be the an a logues of prokaryotic ri bo somal pro teins L10.L7/L12, in teract ing with EF-Tu. Eukaryotic an a logues of these pro teins are acidic phosphorylated pro teins P0.P1/P2. It is im por tant to men tion also that mo lecu lar weights of eEF1A pro tein part ners cor re spond to the weights of these pro teins (34, 12, and 11 kDa). C-ter mi nal do main of men tioned 3 pro teins is known Fig.2 Analysis of large subunit proteins, interacting with rabbit liver eEF1A: a -stained with Coomassie brilliant blue; b -Far Western blot (exposition by antibodies against eEF1A); c -Western blot with antibodies against phospho-proteins P0.P1/P2. to be highly con ser va tive and the use of an ti bod ies against this do main makes it pos si ble to iden tify all three pro teins. The use of these an ti bod ies in Western blot tech nique con firmed that the sig nal ob tained was in deed re ferred to P0.P1/P2 pro teins (Fig. 2, c). One and the same nitrocellulose mem brane was used for iden ti fi ca tion of eEF1A and P0.P1/P2, which is the clear ev i dence of the in ter ac tion of the same proteins with both an ti bod ies against eEF1A and (upon wash ing out of the lat ter) an ti bod ies against P0.P1/P2. Thus, the pos si bil ity of di rect con tact be tween ribo somal pro teins P0.P1/P2 with eEF1A has been deter mined. As P0.P1/P2 and L10.L7/L12 pro teins are known to be func tional an a logues, the in ter ac tion described in the cur rent pa per tes ti fies to the con ser vatism of de cod ing in pro-and eukaryotes.
In trigu ing is the fact of the pres ence of sev eral mi nor sig nals on the mem brane, which may sug gest the pres ence of ad di tional eEF1A bind ing sites on 80S ri bo some, e.g. in the re gion of E site [16].
De ter mi na tion of 40S sub unit pro teins, ca pa ble of in ter act ing with eEF1A. The pro ce dure, anal o gous to the afore men tioned, was also used to ob tain small sub unit pro teins, ca pa ble of con tact ing with eEF1A (Fig.3). This kind of in ter ac tions was not shown for bac te rial ri bo somes. Re gard ing eukaryotic ri bosomes, the au thors of al ready men tioned work [24] used 2-iminotiolane as a bi-func tional cross-link ing agent and pro posed pro teins S23/S24, as well as S26, for eEF1A part ners. How ever, as it has been mentioned be fore, it is pos si ble that these in ter ac tions are the ar ti facts of the pro ce dure of chem i cal cross-linking in the eukaryotic sys tem. Re cently, eEF1A, bound with small sub unit, was dis cov ered to play the role of as so ci a tion fac tor of yeast ri bo somes [25]. Un for tu nately, the data on where ex actly the site of eEF1A bind ing on the sur face of 40S sub unit is and which ri bo somal pro tein may par tic i pate in such inter ac tion have not been pro vided.
We have dem on strated the in ter ac tion of eEF1A with four pro teins of small ri bo somal sub unit us ing the method of Far West ern blot tech nique (Fig.3). Mo lec u lar weight of these pro teins was shown to be in the range of 37-25 and 20-15 kDa. As pro tein part ners of EF-Tu on small sub unit of bac te rial ri bo -some re main un known, and more over, it is quite possi ble that con tacts be tween EF-Tu and 30S sub unit are com pletely ab sent, the method of anal o gies, used for iden ti fi ca tion of large sub unit pro teins, can not be ap plied. Our fur ther in ves ti ga tion will be aimed at iden ti fi ca tion of dis cov ered pro teins us ing the methods of mass-spec trom e try. It is pos si ble to sup pose that the in ter ac tions, ob served be tween eEF1A and 40S sub unit pro teins re flect the ca pa bil ity of these pro teins to in ter act while per form ing func tions, not di rectly con nected with syn the sis of pro teins.