The Localization of Protein L19 on the Surface of 50 S Subunits of Escherichia coli Aided by the Use of Mutants Lacking Protein L l *

Three independently isolated mutants of Escherichia coli which apparently lacked protein L19 on their ri- bosomes, as judged by two-dimensional gels, were ana-lyzed by a range of immunological tests to determine if the protein was indeed lacking. In two of the three, all the tests indicated that protein L19 was absent from both ribosome and supernatant. In the third, a drasti-cally altered form of proteinL19 was present on the ribosome. Electron micrographs of ribosomes obtained from the mutants were indistinguishable from those of wild type strains. The location of ribosomal protein L19 on the surface of the large subunit was determined. It was situated at the base of the 50 S particle facing the small subunit, on the side where the rod like appendage originates. The such

The ribosome of a procaryote such as Escherichia coli comprises over 50 different proteins, in addition to three RNA moieties. The conservation of these proteins in evolution, even when considering such diverse organisms as Bacillus subtilis and E. coli, suggests that they all have an important function in the organelle. However, the role that most of these proteins might play in the structure, function, and assembly of the ribosome has not yet been accurately determined. One way is to investigate and characterize mutants with a particular protein missing from the ribosome. A considerable number of such mutants have now been described (Dabbs, 1979;Dabbs et al., 1983). The absence of a protein was initially determined on the basis of two-dimensional polyacrylamide gels of ribosomal proteins. There was the possibility in mutants that a protein was altered so that its spot coincided with that of another protein or that a much smaller remnant of the protein was present but not detected. It was also possible that the protein did not firmly assemble into the ribosome and was thus present in the ribosome-free supernatant, as has been found for a mutant whose ribosomes lacked protein S1 (Dabbs et aL, 1983).
Since it was necessary to be certain a protein really was missing from the mutant cell, a battery of immunological techniques has been employed; these have been most comprehensive in the case of mutants lacking ribosomal protein L1 and L11 (Stoffler et al., 1980;Dabbs et al., 1981). Furthermore, ribosomes from mutants lacking a single ribosomal protein are ideal controls for the demonstration of antibody specificity * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "adoertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ Recipient of financial support from the Deutsche Forschungsgemeinschaft (Sf% 9) and the Fonds der Deutschen Chemischen Industrie.
__. ____ in immunoelectron microscopic experiments (Dabbs et al., 1981). Most of the mutants have been isolated as spontaneous antibiotic-independent revertants from strains dependent for growth in the presence of the antibiotic erythromycin in the medium. We here describe for the first time two mutants with large ribosomal subunit protein L19 lacking from the ribosome. They were detected among further revertants of erythromycin-dependent strain AM.'

MATERIALS AND METHODS AND RESULTS'
Immunoelectron Microscopic Localization of Protein LI9-The location of the epitopes of protein L19 on the 50 S subunit could now be determined by immunoelectron microscopy. Fig. 7a shows a general field of 50 S particles reacted with anti-119, as observed in the electron microscope. About two-thirds of the subunits were present in dimeric immunocomplexes connected by either one or by a pair of IgG molecules; 12% of the 50 S particles had no IgG molecule bound.
Altogether, 211 50 S. IgG and 616 50 S. IgG '50 S complexes were evaluated. The IgG molecule was best seen in monomeric immunocomplexes and thus allowed also an accurate positioning of the binding site on the ribosome (Fig. 7b). From the dimeric complexes an accurate positioning was only possible in about one-half of them. This might result from the frequent occurrence of dimeric immunocomplexes which are simultaneously connected by antibody pairs. This result was, however, important since it meant that at least two x 1 9 epitopes existed on the large subunit surface.
In the immunocomplexes, the 50 S subunits were seen in both the crown and the kidney projection ( Fig. 7 , b-f). Kidneys were observed relatively more frequently in anti-119 immunocomplexes than in standard 50 S preparations. In the crown projection, the antibody bound to the base of the subunit. Close inspection of the immunocomplexes revealed that the antibody attachment site was between the central point of the base and the region from where the stalk originates (see schematic drawing in Fig. 7 4 . In the crown projections, the Fab arm was not visible in its whole length. This was especially obvious with monomeric immunocomplexes (Fig. 76).
Inspection of these allowed us to conclude that the antibody-
binding site did not lay on the contour line when the subunit was in the crown projection.
When the 50 S subunit was seen in the kidney projection, antibody binding was observed to the notched side of the subunit (Fig. 7e). From complexes in which a crown form was connected with a kidney projection ( Fig. 70 we concluded that it was the same epitope we were pinpointing on both projectional forms. Only 2% of the immunocomplexes showed aberrant antibody binding. The position of protein L19 in the three-dimensional model of the 50 S subunit is shown in Fig.   8.
The localization of protein L19 was additionally undertaken using 50 S particles of mutant AM149. 50 S particles of mutant AM149 were indistinguishable in electron micrographs from particles of wild type strains. After reconstitution of ribosomes from AM149 with protein L19, the protein could be localized at the same site as in wild type ribosomes.
Accessibility of the Epitope of Protein L19 in the 70 S Ribosome-Electron microscopic investigations led to the conclusion that protein L19 was on the surface of the large subunit facing the small subunit (Fig. 8). The three-dimensional model of the ribosome (Kastner et al., 1982) indicated that the L19 epitope is not exposed in the 70 S particle (Fig.  9). Therefore, L19-specific antibodies should not bind to 70 S ribosomes. Various anti-119 IgG preparations were tested for their reactivity with 70 S particles, and in no case 70 S.1g-G. 70 S complexes were observed (data not shown). Therefore, the L19 epitope is covered when the small subunit associates with the large subunit.

DISCUSSION
Protein L19 is located on a unique region of the 50s subunit which is, in the 70s ribosome, at the subunit interface (

9). Results indicating that protein
L19 was located at the interface region have previously been reported by Zeichardt (1976). Furthermore, it has been found (Noll et al., 1976) that binding of L19-specific antibodies to 50 S particles prevents those particles associating with 30 S subunits. Of all the protein-specific antisera which do not react with the 70 S monosome (see Fig. 9 and Stoffler and Stoffler-Meilicke, 1983), this same effect has only been found for anti-S11 (Noll et al., 1976). Since 70 S particles derived from L19-lacking mutant AM149 were stable, it was, however, likely that protein L19 played no essential role in subunit association. However, it could not be excluded that other compensatory mutations were present in the mutant which countered an adverse effect of absence of protein L19 on subunit association. A more detailed study of the various parameters involved in dissociation and reassociation of ribosomal particles should be interesting.
No other ribosomal proteins have so far been mapped in vicinity of the protein L19 epitopes. Protein L19 has also not been cross-linked to any other large subunit ribosomal protein (Traut et al., 1979). Since protein L19 is located at the ribosomal interface, cross-linking data between components of the two ribosomal subunits have to be considered. Among the many protein-protein cross-links between 30 and 50 S proteins, there are two cross-linked pairs with protein L19, uiz. S4-Ll9 and S13-Ll9 (Cover et al., 1981). The large region of the 30-S subunit on which epitopes of protein S4 have been mapped could definitely allow a cross-link between these two proteins (see Fig. 9). The cross-link between protein S13 and L19 is, however, incompatible with any of the four proposed 70-S models (Lake, 1976;Boublik et al., 1977;Kastner et al., 1982;Vasiliev et al., 1983).
Two research groups have identified proteins cross-linked to the heterologous RNA (Baumert et al., 1978;Skold, 1981;Chiam and Wagner, 1983); none of them has found protein L19 cross-linked to 16 S RNA. It seems in general that most of the "interface cross-links" are found between components which are around the edges of the interface and not at that part of the subunit interface where there is the most intimate contact. Cross-linking molecules may not have access to this area of most intimate contact, which probably is also the region where protein L19 is located. We have found no evidence for any assembly defect of 50 S subunits lacking protein L19. This is in agreement with the 50 S assembly map, in which protein L19 has only weak interactions with protein L4 and L17 (Nierhaus, 1982). We did not find any evidence that these proteins are present in reduced amounts on 50 S ribosomes nor could elevated levels be detected in the supernatant f r a~t i o n .~ However, the mutants grow slowly and ribosomes from the mutants employed in an in uitro system programmed with poly(U) synthesized polyphenylalanine very s10wly.~ The experiments performed to demonstrate the specificity of the antibody reacting with ribosomal subunits for its cognate antigen protein are of general importance (see Stoffler and Stoffler-Meilicke, 1983). It is well established that a tetrapeptide is the minimal size for an antigenic determinant (Sela, 1969). Ribosomal proteins share common tetrapeptides with a high frequency, and common penta-and hexapeptides have also been found. Protein L19 has pentapeptides in common with proteins L15 and L16 and tetrapeptides with proteins L1, L3, L6, L10, L12, and L24 (Wittman-Liebold, 1980). The absorption experiments with TP70 from the mutant lacking protein L19 and the observation that anti-119 does not react with mutant ribosomes unless they are reconstituted with protein L19 from wild type exclude cross-reactivity of L19-specific antibodies with any of the other ribosomal proteins. Since mixtures of TP70 missing a single protein can be artificially made, this kind of experiment is generally applicable. It is evident that such control experiments are also necessary when monoclonal antibodies are being used.
Mutants lacking ribosomal protein L19 are also of interest, since the gene is not located in one of the large transcriptional units. The gene coding for this protein (rplS) was located a t minute 56 on the Escherichia coli K12 chromosome (Isono, 1978), in proximity of the gene of ribosomal protein S16 (rpsP). The nucleotide sequence of this DNA region has recently been determined and the operon encodes four polypeptides: ribosomal protein S16 (rpsP), a 21,000 polypeptide (unknown protein), tRNA m'G methyltransferase (trrnD), and ribosomal protein L19 (rplS) in that order (Bystrom et al., 1983). FI ore 4 : Sucrose gradlent profLles. 2 A of 50s subunlts of la1 wrldtype, -n t AM149 and I C 1 mutant AM149, rea88stltuted Wlth protein L19 were pretreated for 30 mln at 37'C and Incubated n t h 1.5 A 2 8 0 IgG from rabbrt 2 3 3 m a buffer containing 50 mM T r l s -H C 1 IpH 7 . 8 1 . 2 0 mn MgC12. 500 mM NH4Cl and 2 mH DTT 11.4 Dlthlothreitoll for 5 m l n at O°C. Dimerlc lmunOCOmpleXeS were separated from unreacted 50s subunits and 1gG by centrlfugation through NHqC1 and 6 mn 8-mercaptoethanol for 16 hrs at 19 000 rpm m an SW40 rotor.
S U C~O S~ gradients were run with 196 preparations purlfled from the sera from 3 rabbits and a sheep mmunired with E. ribosomal proteln L19. Dimer peaks were Obtained with each IgG preparation; the degree of dlmer formation of the different IgGs varied. The amounts Of dimer formation Obtained with the different IqG6 plotted against antlbody ConcentratLon is Shown in Fig.  5. POT all further experlnents IgG from rabblt 2 3 3 was used, Since it was the most reactive (see Fig. 51. Maximally 65% of the 50s Subunits reacted wlth I~G obtained from the latter antiserum. since the d m e r formation occurs under non-equilibrium conditions, it is llkely that more than 65% Of the SUbYnltS are reactive with anti-LlY and thus epitopes of protein L19 were exposed on the surface of the majority of the Subunits.
It was essential to demonstrate that the reactlve epltape I" 50s subunits wan provided by proteln 119. W e thus investigated the effect of prelncubatrng anribody with protein Ll9 upon Its abillty to brnd to 505 suhunlts. For thrs purpose an antibody Concentration was chosen whrch would dlmerlre half the number Of subunits that could be dlmerlsed. The result Of this experiment 1s shorn ~n Fig. 6a. The addltion Of 2 ug Of proteln L19 to anti-LlY IqG completely inhibited the formatLon of IgG.50S complexes ~n sucrose gradrents. Thls exoerrment showed that the reactlve e~l t o~e on 50s Sublmlts was contamed 2" the purified protein 119. However, t h k ekperlment dld not exclude that the comprised the Same antlgenic determlnant (see belovl.
reactlve determinant on the ribosome was present on another proteln vhlch I \ I \ '\ k . '. preparatlans would n o t react wlth total ribosomal proteln ITP701 extracted from It was also antlcrpated on the basls Of t h e s e e x p e r~m e n t s t h a t the v a r~o u s IgG antl-Ll9 IgG prelncubated vlth excess TPlO of mutant A1149 st111 reacted wlth t h e ribosomes of mutant Iv114Y. As shown I" F l q . 6b. i t was lndeed found that strain drd not react. S l m l l a r results could be abtalned, ~f a mixture of TP70 5 0 s subunlti Of vlldtype, whllst IgG prelncubated wrrh excess TP70 of a wlldtype l a c k r n g the respecrrve prater" was prepared from purlfled pratelns I E l g . 6 b l .