Enzymatic Hydrolosis of 16s Ribosomal RNA and 30s Ribosomal Subunits

!:..~ 30s ribosomal subunits and protein-free 168 RNA have been mildly hydrolyzed with pancreatic ribonuclease and the RNA fragments analyzed by polyacrylamide gel electrophoresis. The protein-free RNA gives nine discrete fragments and the 308 subunits give six discrete fragments. A comparison of electrophoretic mobilities, indicates that at least three fragments from 168 RNA are distinct from the fragments from 308. The kinetics of the hydrolysis reaction is pseudo first-order for the protein-free 168 RNA and pseudo secondorder for the 308 ribosomes. The rate of hydrolysis of the proteinfree 168 RNA is much faster than that of the 308 subunit. These data suggest that in the protein-free 168 RNA there are certain regions exposed to the ribonuclease which are not exposed in the ribosome, and that this is due to either some shielding by the specific proteins or to a different conformation of the RNA in the ribosome.


RESULTS
Mild hydroly· sis of protein-free 16S rRNA generates nine di.screte fragments Mild hyrolysis of 30S subunits generates six discrete RNA fragments Notable differences between figure l and 2 Pseudo first-order rate constant for 16S rRNA Pseudo second-order rate constants for 30S subunits Resistance of 30S subunits to hydrolysis

DISCUSSION
The resulting discrete fragments from the hydrolysis of protein- 13 14 15 15 free 16S rRNA suggests that it has a tertiary structure . 17 Possible mechanism for the onset of the hydrolysis reaction of protein-free 16S rRNA 17 RNA fragment pattern of 30S subunit suggests that there is a definite ribosomal conformation 18 Comparison between the fragment patterns of protein-free RNA and 30S subunits 18 Possible mechanism for the hydrolysis reaction of 30S subunits 19 Th. e 30S subunit is mostly protected from the action of the RNase 20 The conformation of protein-fr ee 16S rRNA is not necessarily the same of rRNA in the ribosome 21 FIGURES AND TABLES  individual nucleotides and paired polynucleotides, it has been indicated that yeast rRNA at 30°C contains about 60 percent of its bases participating in double helical regions (2).
No detectable difference has been found in the ability to bind Ca++ and Mg++ between protein-free RNA and ribosomes from ~ coli as measured by equilibrium dialysis (J). The authors have concluded that Mg++ binding only occurs at the phosphate groups, and that all of these groups are available as binding sites for small cations. This conclusion has lead the authors to suggest that the organization of the ribosome must be such that the phosphate groups are not directly involved in the protein-RNA interaction.
Powder patterns from X-ray diffraction studies on isolated rRNA and ribosomes from rat liver' yeast and ~ coli are similar and have been taken as an indication that the conformation of the rRNA may be similar in all cases (4 ,5}. Protein-free rRNA as well as ribosomes from the various sources investigated have been reported to exhibit equally low extinction coefficients at 2600 A (property known as hypochromism).
The fact that the hypochromicities of the isolated rRNA and ribo somes change identically as the RNA is being denatured by heat or by bydroloysis has been interpreted by the authors as evidence that the secondary structure of rRNA does not change significantly after removal of the proteins (6, 7 ,8} . It has also been reported that RNA in rabbit reticulocyte ribosomes give an ORD spectrum similar to that of the rRNA dissociated from the ribosomal proteins. This has lead the authors to suggest that the conformation of rRNA in the ribosome is similar to that in the protein-free state (9,10).
Pancreatic ribonuclease (pRNase} has been used as a probe to investigate the internal organization of ribosomes. It has been established that portions of the rRNA are exposed at the surface of the ribosome instead of being entirely coated by the protein (11) since the 708 ribosome is hydrolyzable by ribonucleas e but to a lesser extent than the i .solated rRNA. Using gel electrophoresis, it 3.
has been shown that specific cleavage is obtained from unfractionated yeast protein-free rRNA by the action of T 1 and pancreatic ribonucleas.es (12). Likewise, rRNA extracted from purified subunits of rabbit reticulocyte ribosomes and treated with pRNase has been reported to hydrolyze initially into a number of discrete fragments resolvable by gel electrophoresis (13}. The authors have suggested that discrete fragments are produced because certain sites in the RNA are preferentially attacked as a result of a specific folding of the polymer chain into a tertiary structure (12 , 13 ) .
From the data obtained in previous work, it appears that no conclusive answer can yet be given to the question of whether the conformation of rRNA changes upon removal of the proteins. The error of the i.on binding measurements is too large to clearly show that no hate bonds are directly involved in protein-RNA interactions. phOSP The ~uality of the X-ray diffraction data is not sufficiently high to 4. n d istinguish between heated-and-cooled TMV RNA and a ribosome (4), eve and therefore the data cannot be used to detect changes in the conformati.on of the rRNA. The level of hypochromicity is correlated to the mole fraction of paired bases in the structure and is not sensitive to the positions of the short base paired regions. The ORD curve for different conformation states of a given polymer is only a function of the average helical content and will not discriminate between two different secondary structures with the same percent of helical content. Therefore these spectroscopic methods, if used together, merely analyze the tota l helical content of rRNA.
The present work is a comparative study of the pRNase mediated hydrolysis of purified 30S subunits and of protein-free 168 RNA from E. coli. The factors underlying this enzymatic approach are first, the presence of sites of rRNA which are specifically hydrolyzed by pRNase, and second, the existence of dissimilar rates of hydrolysis depending on individual steric conditions of the substr ate. pRNase at l ow concentrati ons is known to exclu sivel y hydrolyze distal to the phosphate group attached to 3' carbon of the pyrimidine nucleosides in RNA. These residues are assumed to be distributed along the polymer chain as dictated by the primary structure.
H i .s widely accepted that pRNase princ ipally attacks the single regions of RNA. In addition to this selectivity, the access of pRNase to s ites in single s tranded regions can be hindered by a tertiary structure of the RNA and the specific or associated proteins in the ribosome. Therefore the method used here is a probe for single stranded regions of rRNA with pyrimidine residues which are not shielded by protein or by the conformation of the RNA. The 30S subunits were separated from the 508 subunits by zonal centrifugation using a sucrose gradient of the form developed by Eikenberry (14) , and approximated by means of a constant volume exponential gradient (Eikenberry, personal communication). The detailed procedure of the approximated exponential gradient adapted for the Beckman Ti-14 rotor is as follows. A Mariette flask with 900 ml of 50% Cw/w) sucrose (Fi s her Scientific Co.) feeds a mixing chamber initially containing 295 ml of 7% Cw/w) sucrose, and whos.e volume remains constant throughout the loading of the gradient. The gradient is set in the rotor at 3 ,000 rpm by pumping into i . t the output of the mixing chamber. When the rotor has been filled , some 300-500 mg of ribosomes previously placed iri a linear sucrose gradient After 4 hours of centrifugation at 45,000 rpm the 508 and 308 particles reached good separation . The speed was then reset at 3 ,000 rpm, and the gradient unloaded by injecting 600 ml of 50% sucrose and collected in 10 ml fractions. The absorbance of the fractions was measured at The resulting concentration profile facilitates the pooling of the fractions containing 30S and 508 respectively. Subunits thus obtained were dialyzed and concentrated in a single operation using ultrafiltration in an Amicon cell with a UM-20E membrane. Upon several additions of a buffer containing 0.001 M Mg(Acetate} 2 , 0.005 M Tris-Cl (pH 7. 4), the concentration of K+ was lowered by a factor of 500, the sucrose was lowered to less than 0.1% and the concentration of the 308 ribosomes was adjusted to an absorbance (2600 A) of thirty. Then the solution was aliquoted, frozen in liq_uid nitrogen, and stored at -15°C, ready to be used for enzymatic hydrolysis.

Preparation of 16S RNA
A 308 ri. bosome solution of concentration not greater than 2 mg/ml was deproteni zed with phenol at room tempera ture according 8.
to the following procedure. One volume of water-saturated phenol (24°C) was shaken for 10 minutes with one volume of the ribosome suspensi on, a nd the aqueous. phase separated after cent rifugation. The ioni.c strength of the aqueous phase was then increased with 1/10 Volume of 20% Na :Acetate and the RNA precipitated at -15°C wi th 3 volumes of ethan ol. After the white precipitate was sedime nted at 2 2,000g for 20 minutes, the pellet was redissolved in about one Volume of 0. 005 M Tris-Cl (pK 7 .4)., and the RNA was again precipitated Vi.th ethanol after the addit i on of 20% Na Aceta t e _ . The precipitation step was done three times . This RNA , free of phenol, was di ssolved in M tr is to an ab-s.orb-ance (2600 A) of 30 and stored at -15°C in 0.005 small ali q_uot s .
By extracting unfractionated RNA with phenol as a function of the concentra tion of the RNA, it was possible to show that the rati o 9. of l6S to 238 was abnormally low for input concentrations greater than 1 ,5 mg/ml. Best recoveries of 168 RNA were obtained working at 1 mg/ml or less, and using one volume of phenol at room temperat ure. I found that the method of Bishop (16) for the preparation of polyacrylamide gels for the analysis of RNA can be modified for the sake of simplicity. His method calls for the recrystallization of the acryla.mide and the N ,N' -methylenebi sacrylamide , and for pre soaking the gels for several days. Gels of recry stallized materials were made and presoaked for about ten days . Gels of un-recrystallized materials were also made and allowed t o pre.soak for 8 and 24 hours . Then the three kinds of gels were scanned versus distance along gel i .n the ultraviolet between 2600-2900A , before and aft er passing a 5-mAmp CUrrent t hrough them in the electrophoresis buffer (prerun). No . ·ficant difference was found in the back.ground level of all gels s1gn1 .

The Ribonuclease
after 20 minutes of prerun. All of the scans had a smooth, low and 11. flat background, except for the gels with long presoaki ng which probably exhib-ited some light scattering from dust.

Electrophoresis of the RNA
The g els to be used i n an electrophoretic run were cut t o 1 0 . 5 cm in lengt h , l eaving t h e ends f l at a n d strai ght . Before l oadi ng the sample , a 5 mAmp / gel cu r r e nt was passed throu gh the gel for 30 minutes (pre-run) in E-SDS (0. 2% SDS ) buffer. The t otal rea c t ed v olume (littl e over 60 ) was loaded ont o a g el, a nd t hen el ectrophoresed with a current of 5 mAmp/ gel for 3 or 3 . 5 h ours usin g a Canal co d i scelectrophore sis instrument.

. Reading the Gel s
The gel s were scann ed for absorbance at 2600 A in a DU-Gilford spectrophotomet e r wi t h mechanical gel scanner a nd r ecor d e r . Qu a ntitat ive · of the products of the reaction was done by both cutting the ana.lysi.s

12.
tracing of the recorder and weighing the areas, and by direct measurement of the heights of the bands. To supplement the results from the UV-scan, gels in 0. 2% SDS were first washed in distilled water in a test tube for 24 hours (with one change of water) and then stained with one-tenth volume of 0. 2% methylene blue in 0. 4 M acetate buffer (plf 4. 7}. The staining could be stopped anywhere between 4 and 8 h ours ~y placi,ng th. e g el in a test tube wi th distilled water, with no need of further washing.

III. RESULTS
Protein-free 16S rRNA was mildly hydrolyzed with 0.02 µg /ml of pRNase, and the products of the reaction were analyzed in a polyacrylamide gel to give the pattern shown in figure 1. Nine product bands even at the earliest kinetic points were observable by either staining the gel with methylene blue or by scanning the gel at 2600 A.
These nine bands correspond to nine discrete RNA fragments whose molecular weights have been calculated from their individual mobilities (_16} and reported in table 1.
In order to follow a brief history of the b ands of fi gure 1, I have arbitrarily divided the course of mild hydrolysis into two stages according to the remaining amount of input RNA namely, when intact 16S RNA remaining is 25% or more, and when it is less than 25%. Toward the middle of the first stage most of the new bands were observed to attain maximum neight and best resolution. In the second stage, band 16A lost considerable height, the general background was raised, and the other bands lost their original shape . Bands 16A and 16G remained particularly prominent until the onset of the second stage (figure lb, le). 30S ri. bosomal subunits , i.n i . oni.c conditions ident ical to those used for the protein-free rRNA, were midly hydrolyzed with 0.4· µ g/ml of PRNase. The products of the reaction were deproteinized with SDS, and then analyzed in a polyacrylamide gel to give the pattern shown in figure 2 . Six small bands were observable by either staining the gel l(ith methylene blue or br scanning the gel at 2600A. Bands 30A and 30 B appeared almost fused into one broad and flat band, but were well resolved bands in the stained gels. Notice the 30C and 30D are the highest bands among the products. The molecular weights of the RNA rra.gments range between 0. 5 and 0. 02 times that of the 168, and their values are given in table 2. 14. By comparing the gel patterns of the RNA fragments from proteinfree rRNA and 30S subunits, two maj or observations were made. a) Upon reaction with the ribonuclease, the protein-free RNA generated more fragments than the RNA in the ribosome. b) The bands that exhibited greater absorbance did not correspond to the larger molecular weight, but rather to some intermediate molecular weight fragments.
Th. e kinetic study of the hydrolysis of RNA was done by measuring the height of the 168 band of the gel pattern as a function of time since the height of the band is dir ectly proportional to the concentrati on of unhydrolyzed 16S RNA. In order to learn the order of the kinetics of the reactions, the data from both protein-free 168 RNA and 308 ribosomes were treated as a zeroth-order , first-order, and second-order reaction with respect to the substrate. The rate of decomposition of a reactant that follows zeroth-order kinetics obeys the rate eq_uation dC/dt = -k , whose integrated form, 0 =kt 0 Eq_uation 1 .
represents a straight line of slope k 0 • In turn, the rate of decomposition of a reactant that follows first-order kinetics obeys the rate eq_uation dC/dt = -k 1 C, whose integrated form ,

15.
Equation 2 thus plotting ln C 0 /Ct versus t gives a straight line of slope k 1 .
If the decomposition of a reactant follows second-order kinetics, t h en the rate equation is dC/dt = k 2 C 2 , whose integrated f or m, Equ at i on 3 thus plotting c~l -~o 1 versus t gives a straight line of slope k 2 .
Pseudo first-order and pseudo second-order rate constant s (k{, k2) which are proportional t o the enzyme concentrati on, are c ommonly u s ed for enzyme-catalyzed reactions.
The kinetic data for the hydrolysis of protein-free 16S RNA best fit e qu a t ion 2 for t he f ir s t stage of the react ion (figure 3b ) .
For an enzyme conc e ntration close t o 0.01 µ g /ml, the pseudo fi rst-ord e r rate constant Ck{} was estimated from the slope of fi gure 3b to be equal to 2 . 6 x io-2 min -1 , a nd the cor responding half -life for t h e protein-free 16S RNA was approximately 25 mi nut es . When the same set of data was treated acc ording to the zeroth-order a nd second-or der rate eq_uati on s , curved lines were observed t o fit the point s (figure 3a).
It was ob served that the order of t h e reaction cou ld cha nge if the extent of hydrolysis pa ssed the first stag e.
The kinetic s f or the hydrolysi s of 30S ribosome s with 0 . 4 and 0.8 µg /ml pRNase was found to f i t a strai ght line according to equat ion 3 ( figure 4c ) . Starting with a n inpu t of 0. 5 mg /ml of the substrate , the pseudo second-order rate consta nts were 8 . 6 x 10-2 and 17. 2 x io-2 ml / mg ·min, and the half-life times were 24 a nd 12 m· inutes for 0.4 a nd 0 . 8 µ g / ml of pRNase respect ively. Attempts to fit these data to equations 1 and 2 for zeroth-order and first-order kinetics respectively, gave the curves shown in figures 4a, b.

16.
When 308 ribosomes and 168 RNA were treated at RNase concentrations which largely hydrolyzed 168 RNA, little or no hydrolysis of the 30 s ribosomes was observed (figure 3b). It was therefore necessary to employ for the ribosomes RNase in concentrations 20 to 40 times that needed for the protein-free RNA.

IV. DISCUSSION
The observation that mild hydrolysis of protein-free 16S rRNA results in a number of discrete fragments suggests the RNA has a tertiary structure. The appearance of discrete fragments of rRNA upon mild hydrolysis has been already reported for yeast rRNA and unfractionated ~ coli rRNA (12) , rabbit reticulocyte 19S rRNA (13) , and for 23S rRNA from ~ coli (17). Evidence is given that a tertiary structure assumed by the RNA is selectively exposing the few sites that are primarily hydrolyzed while there exist many other sites which remain temporarily "guarded" by the tertiary structure of the RNA. It has been reported (13). that during the course of mild hydrolysi s of rRNA, the fragment pattern changes as the large fragments are hydrolyzed into smaller fragments which~ in turn, originate new bands. Furthermore, small fragments can be produced by the action of pRNase (about 2S) of rRNA which are known to contain both helical and single stranded reg i ons (19). Therefore, the presence of many hidden sites strongly suggests that there is a tertiary structure in the protein-free 16S RNA determining, over the dictates of the primary and secondary structures , which s it es are most likely to be hydrolyzed.
By examination of the molecular weights in table 1, it is possible to suggest a mechanism for the hydrolysis of protein-free 16S rRNA.
It is likely that fragment 1 6A (M. W. 440 ,000) proc e eds from the 16S 18. by a one-break react ion s.ince that 16A has a very high molecular weight, and the enzyme-to-substrate molar ratio in this hydrolysis is very low (about 1/1000). Making also the consideration that the absorbance of RNA fragments of molecular weight less than 100 ,000 is quite low, it is then reasonable to expect from the reaction an RNA component with a molecular weight in the neighborhood of 110 ,000. Fragment 16G (M.W. 130,000} fits best the desired condition and, with its molecular weight reas signed a s 110 ,000, it could be used as a marker together with i6s to correct the weights; of the intermediate fragments. Taking  narrow site on the RNA might be strategically exposed after the removal of the proteins so that fragments· 1 6A and 16G could be produced. It is possible that if three out of the nine fragments in figure 1 do not coincide with any of those in figure 2 at least 3 and perhaps 6, or even all 9 , of them are characteristic of the protein-free rRNA.
Since no two RNA fragments from the hydrolysis of the subunit add up to 550 , 000 daltons, it is concluded that no single-break reaction could be detected in the present conditions. At the high enzyme concentration used for the ribos.omes, multiple-break reactions are more probable and they tend to obscure the mechanism of the total reacti on .
However, since the sum of the corrected molecular weig hts (table 2} of fragments 30A through 30F eg_uals 583 ,000 daltons, it is possible to suggest at least one mechanism for the totai reaction. The mechanism would reg_uir e the existence of n neighboring sites which originate n equally probable reactions on the ribosome. A model of the 308 ribosome that makes the mechanism feasible could be one having a larg e surface protected by protein, and a "cavity" by which the enzyme could reach unpr otected reg ions of the RNA . This interpretation would i mply that the reaction for the entire s e t of sites would occur by random collisions while n -1 of the sites would not react independently, as if the enzyme remained "trapped " i _ n a cavity after the first reaction .

20.
The hydrolysis of protein-free 168 RNA follows first-order kinetics and the hydrolys:is of 30S ribosomes is found to fit a secondorder rate equation when the half-lives of both substrates are comparable.
Th.ere is not a clear cut differentiation between the first, second found to follow first-order kinetics (17).
The hydrolysis rate of the protein-free 168 rRNA is substanti.ally greater than that of the 308 ribosomes. In order to make the ha lf-lives about 25 minutes, the subunits required a 40-fold increase of the ribonuclease. This striking difference between the 308 subunit and 1 68 RNA contrasts with the f act that the 508 s ubunit h as a f irst-order rate constant equal to twice that of the 238 RNA (17) .
The RNA in the 308 ribosomal subunit is mostly protected from the action of the ribonuclease. The significantly smal ler hydrolysis rate of the ribosomes could be partly a result of shiel ding by the 2) The hydrolysis rate of the 308 subunit is low enough to suspect that the RNA may undergo conformational change upon removal of the proteins. : : ..  Fragment pattern from the hydrolysis of 30S ribosomal subunits as analyzed in a 4% polyacrylarnide gel. The product bands display maximum height attained during mild hydrolysis.
The reaction was carried out with 0.4-0.8 µg/ml of RNase. The RNA from input ribosomes gave a single band undistinguishable from that in figure la.      I t has been shown in section TU of the thesis that mild hydrolysis of 30S ribosomes in the absence of of K+ yields fragments which do not accumulate. The preliminary data presented in this appendix argue t hat significant changes are introduc ed in the fragment pattern of the 30S subunit by varying the concentration of KCl . An innnediate conclusion is that perhaps the ribos.ome undergoes conformational transiti on s , + and even expos.es different sites to the RNase, at the measure that K is being added or subtracted. It has been suggested that a particular concentration of monovalent cations, such as K+ and NH+ 4 may be responsible for the conformation adopted by the 50S subunit (18 ). A possible explanation of this phenomenon could be found in the ability of monovalent cations to exchange with the divalent cations that cross-link the RNA chain.
Appendix: Figure 1 Fragment pattern of 308 ribosomal subunits digested in 0. 005 Molar KCl. Deproteinization carried out with. 0.2% 8DS and RNA fragments analyzed in a 3% polyacrylamide gel. Input ribosomes gave a single band in the position of the 168 RNA.