Generation of templates for RNA-protein fusions

Rehi Liu


General Considerations. The puromycin containing linkeroligonucleotide was ligated to the 3' end of an mRNA using bacteriophageT4 DNA ligase in the presence of a complementary DNA splint. SinceT4 DNA ligase prefers precise base-pairing near the ligation junctionand run-off transcription products with T7, T3, or SP6 RNA polymeraseare usually heterogeneous at their 3' end, only those RNAs containingthe correct 3'-terminal nucleotide can be efficiently ligated.When a standard DNA splint was used, approximately 40% of runofftranscription products were ligated to the puromycin oligo. Theamount of ligation product can be increased by using excess RNA,but not excess puromycin oligo. Apparently, the limiting factorfor ligation is the amount of RNA which is fully complementaryto the corresponding region of the DNA splint.

To allow ligation of those transcripts ending with an extra non-templatednucleotide at the 3' terminus [N+1 products], a mixture of thestandard DNA splint with a new DNA splint containing an additionalrandom base at the ligation junction was used. The ligation efficiencyis much better for myc-RNA in the presence of such a mixed DNAsplint (Figure 1)

Other considerations: 1) The 3' end of the mRNA shouldnot have any stable secondary structure that would interfere withannealing to the splint oligonucleotide. 2) A high concentrationof salt can cause failure of the ligation reaction. This can beprevented by desalting the oligonucleotides thoroughly using NAP-25columns (Pharmacia). 3) The ligation reaction is fast and is usuallyfinished within 40 minutes. Excessively long time incubation atroom temperature will result in unnecessary degradation of theRNA.

Enzymatic Synthesis of mRNA-Puromycin Conjugates. Ligationof the myc RNA sequence to the puromycin containing oligonucleotidewas performed using a standard DNA splint (5'-TTTTTTTTTTAGCGCAAGA)and/or a splint containing a random base (N) at the ligation junction(e.g., 5'-TTTTTTTTTTNAGCGCAAGA). The sequence of myc-RNA was


(RNA124) and the sequence of the phosphorylated puromycin containinglinker was dA21(C9)3dAdCdCP(C9: Spacer 9, Glen Research, represents HO(CH2CH2O)3PO2-; P represents puromycin). The reaction consisted of mRNA,DNA splint, and puromycin oligonucleotide in a molar ratio of1.0 : 1.5-2.0 : 1.0. First, a mixture of the three oligos wasfirst heated in water at 94°C for 1minute and then cooled on ice for 15 minutes. Ligation reactionwas performed for one hour at room temperature in 50 mM Tris-HCl(pH7.5), 10 mM MgCl2, 10 mM DTT, 1 mM ATP,25 mg/ml BSA, 15 mMpuromycin oligo, 15 mM mRNA, 22.5-30mM DNA splint, RNasin inhibitor (Promega)at 1 U/ml, and 1.6 units of T4 DNAligase (New England Biolabs) per picomol of puromycin oligo. Afterthe incubation, EDTA was added to a final concentration of 30mM and the reaction mixture was extracted with phenol/chloroform.Full length conjugate was purified on denaturing PAGE and isolatedby electroelution. The yield was approximately 70% for myc-RNAwhen a mixed DNA splint was used.


General conditions: Fusion reaction can be divided intotwo steps which are carried out under different conditions. Thefirst step is the translation of RNA template in rabbit reticulocytelysate at 30°C for 30 to 90 minutesin the presence of low concentration of Mg2+and K+. The second step is the post-translationalincubation which allows the translated RNA to fuse to its nascentpeptide via puromycin at its 3' end. This step is usually performedin the presence of elevated Mg2+ or K+(or both) at lower temperature (e.g. -20 °C,0°C, or room temperature) for an appropriatetime. To visualize the radiolabeled fusion products, 2mlof translation reaction mixtures were loaded onto an 8% tricineSDS-PAGE (for 35S-Met labeled proteins) ora 6% glycine SDS-PAGE (for 32P labeled RNAtemplates). The fusion products can also be isolated using dT25streptavidin agarose (oligo dT cellulose works better as describedby Glen) or thiopropyl sepharose or both as described previously.

Quality of Puromycin Oligo. The quality of the puromycinoligo is very important for the efficient generation of fusionproducts. The fusion efficiency might be very low if poor qualitypuromycin oligo was used. The coupling of 5'-DMT, 2'-succinyl,N-trifluoroacetyl puromycin with CPG is not as efficient as thecoupling of the standard nucleotides. The coupling reaction mustbe carefully monitored to avoid the formation of CPG with toolow a concentration of coupled puromycin, and unreacted aminogroups on the CPG must be fully quenched to avoid subsequent synthesisof oligonucleotides lacking a 3'-terminal puromycin. It is essentialthat CPG containing very fine mesh particles should not be usedto avoid problems with valve clogging during subsequent automatedoligo synthesis. Before used for oligo synthesis, the self-packedcolumn should be flushed with argon.

The puromycin oligo synthesized should be tested before largescale use. The presence of puromycin at the 3' end is crucial.No fusion was detected if puromycin was substituted with a deoxyadenosinecontaining a primary amino group at the 3' end. To test for thepresence of 3' hydroxyl groups (i.e. the undesired synthesis ofoligos lacking a 3'-terminal puromycin), the puromycin oligo maybe first radiolabeled (e.g. by 5'-phosphorylation) and then usedas a primer for extension with terminal deoxynucleotidyl transferase:no extension product should be observed.

Time Course of Translation and Post-translational Incubation.The translation reaction is relatively rapid and is usually finishedwithin 25 minutes at 30°C. The fusionreaction under translation conditions, however, is slower. Whena standard linker (dA27dCdCP) is used at 30 °C,fusion synthesis has reached its maximum level in an additional45 minutes.

The post-translational incubation may be carried out at lowertemperatures: room temperature, 0 °C,or -20_C. Less degradation of the mRNAtemplate was observed at -20 °C andthe best fusion results were obtained after incubation at -20 °Cfor 2 days.

The effect of high concentration of Mg2+ andmonocations. A high concentration of Mg2+in the post-translational incubation greatly stimulated fusionformation. For the myc-RNA template, there was a 3-4 fold stimulationif the standard linker (dA27dCdCP) was usedwhen 50 mM Mg2+ (MgCl2or Mg(OAc)2) was added during the 16 hr incubationat -20 °C.

Addition of high concentration of monocation (e.g. 0.5 M KCl orNH4Cl) during post-translational incubationalso stimulated fusion formation. The best results were obtainedwhen post-translational incubation was performed at room temperaturein the presence of both elevated Mg2+ andK+.

Linker Length and Sequence. The dependence of the fusionreaction on the length of the linker was also studied. In therange between 21 and 30 nucleotides (n=18-27), little change wasseen in the efficiency of the fusion reaction as reported previously(for E. Coli lysate translation system). Longer linkers (e.g.,45 or 54 nucleotides in length) or shorter linkers (e.g., 13 nucleotidesin length) resulted in much lower fusion efficiency, suggestingthat the linker may pass directly from the decoding site throughthe site normally occupied by the A site tRNA to the peptidyltransferase center. It is very likely that the optimal lengthof linker is in the range of 20 to 40 nucleotides.

Substitution of deoxyribonucleotide residues near the 3' end withribonucleotide residues did not change the fusion efficiency verymuch. The dCdCP (or rCrCP) sequence at the 3' end of the linkeris, however, essential. Substitution of dCdCP with dUdUP reducedthe efficiency of fusion formation significantly.

Flexibility of the Linker. The dependence of the fusionreaction on the flexibility of the linker was also tested. Thefusion efficiency was very low if the linker was made more rigidby annealing with a complementary oligonucleotide near the 3'end.

When a more flexible linker (e.g., dA21C9C9C9dAdCdCP;where C9 (Glen Research) represents HO(CH2CH2O)3PO2)was used, however, the fusion efficiency was significantly improved.Compared to the standard linker (dA27dCdCP),using the flexible linker (dA21C9C9C9dAdCdCP)improved the fusion efficiency for myc-RNA (RNA124) more than4-fold. In contrast to the template with the standard linker whosepost-translational fusion proceeds poorly in the absence of ahigh concentration of Mg2+, the template withthe flexible linker does not require elevated Mg2+to get a good yield of fusion, as long as an extended post-translationalincubation at -20 °C is used. This linker,therefore, is very useful if post-translational addition of highconcentration of Mg2+ is not desired. However,even with the flexible linker, better fusion yields are obtainedin the presence of elevated Mg2+.

Quantitation of Fusion Efficiency. Fusion efficiency maybe expressed as either the fraction of translated peptide convertedto fusion product, or the fraction of input template convertedto fusion product.

To determine the fraction of translated peptide converted to fusionproduct, 35S-Met can be used to label thetranslated peptide. To determine the percentage of the input templateconverted to fusion product, the translations can be performedusing 32P labeled mRNA-linker template. Thebest results were achieved when lysate from Novagen, Amersham,or Ambion was used.

The mobility difference between mRNA and mRNA-peptide fusion onSDS-PAGE may be very small if the mRNA template is long. In suchcases, the template may be labeled at the 5' end of the linkerwith 32P. The long RNA portion may then bedigested with RNase H in the presence of a complementary DNA splintafter translation/incubation and the fusion efficiency determinedby quantitation of the ratio of unmodified linker to linker-peptidefusion. Compared to RNase A digestion which produces 3'-P and5'-OH, this approach has the advantage that the 32Pat the 5' end of the linker is not removed.

Cis- vs. Trans- Fusion During Post-translational Incubation."Cis-fusion" is defined as the fusion reaction betweenan mRNA and its encoded peptide or protein, whereas "trans-fusion"is defined as the fusion reaction between an mRNA and a peptideor protein which is not encoded by it. The central idea of themRNA-protein fusion is to recover the sequence information inthe peptide or protein portion via its attached mRNA. This requiresthat the fusion reaction be cis. I tested whether the fusion reactionthat occurred at -20 °C or room temperaturein the presence of elevated Mg2+ or K+(or both) was cis or trans. Free linker (dA27dCdCPor dA21C9C9C9dAdCdCP;C9: -O(CH2CH2O)3PO2-)was coincubated with a template containing a DNA linker, but withoutpuromycin at the 3' end, under the translation and post-translationalincubation conditions described above. No detectable amount (lessthan 2% of the normal level) of 35S-Met wasincorporated into linker-peptide product, suggesting that post-translationalfusion occurs primarily between the nascent peptide and mRNA boundto the same ribosome.

Reticulocyte Translation Reactions. Translation reactionswere performed in reticulocyte lysate from different commercialsources (Novagen, Amersham, or Ambion). A typical reaction mixture(25 ml final volume) consisted of 20mM HEPES pH7.6, 2 mM DTT, 8 mM creatine phosphate, 100 mM KCl,0.75 mM Mg(OAc)2, 1 mM ATP, 0.2 mM GTP, 25mM of each amino acid (0.7 mMof methionine if 35S-Met (NEN) was used),RNasin (Promega) at 1U/ml, and 60%v/v of lysate. The final concentration of template was in therange of 50 nM to 800 nM. For each incubation, all the componentsexcept lysate were mixed carefully on ice and the frozen lysatewas thawed immediately before use. After addition of lysate, thereaction mixture was mixed thoroughly by gentle pipetting andincubated at 30°C to start translation.The optimal concentrations of Mg2+ and K+vary with different mRNAs and therefore should be determinedin preliminary experiments. For a poorly translated mRNA, it isalso worth optimizing the concentrations of hemin, creatine phosphate,tRNA, and amino acids. Potassium chloride is usually better thanpotassium acetate for fusion reactions, but sometimes a mixtureof KCl and KOAc leads to better results. After translation at30°C for 30 to 90 minutes, the reactionwas cooled on ice for 40 minutes and Mg2+or K+ (or both) was added. The final concentrationof Mg2+ or K+ added atthis step should also be optimized for different mRNA templates,usually in the range of 50 mM to 100 mM for Mg2+and 0.3 M to 0.6 M for K+. The resulting mixturewas incubated at -20°C for 16 hoursto 48 hours (or 30 minutes to 1 hour if both Mg2+and K+ were used). After incubation, 2 mlof reaction mixture was mixed with 4 mlloading buffer and heated at 75°C for3 minutes. The resulting mixture was then loaded onto a 6% glycineSDS-PAGE or an 8% tricine SDS-PAGE.

To remove the RNA portion of the RNA-linker-puromycin-peptideconjugate for subsequent analysis by SDS-PAGE, an appropriateamount of EDTA was added after post-translational incubation andthe reaction mixture was desalted using a microcon-10 (or microcon-30)column (Pharmacia). 2 ml of the resultingmixture (approximately 25 ml total)was mixed with 18 ml of RNase H buffer(30 mM Tris-HCl, pH7.8, 30 mM (NH4)2SO4,8 mM MgCl2, 1.5 mM b-mercaptoethanol,and an appropriate amount of complementary DNA splint) and incubatedat 4°C for 45 minutes. RNase H (Promega)was then added and digestion was performed at 37°for 20 minutes.

Translation of myc-RNA (RNA124-dA21C9C9C9dAdCdCP).Translation was performed in reticulocyte lysate (Novagen) in25 ml total volume consisted of 400nM myc RNA template (RNA124-dA21C9C9C9dAdCdCP)body labeled with 32P, 0.5 X translation mixwithout methionine, 0.5 X translation mix without leucine, 100mM KCl, 0.75 mM Mg(OAc)2, RNasin (Promega)at 1U/ml, and 60% v/v of lysate. Aftertranslation at 30°C for 90 minutes,the reaction was cooled on ice for 40 minutes and Mg(OAc)2was added to 50 mM. After 2 days incubation at -20°C,2 ml of reaction mixture was mixedwith 4 ml loading buffer, heated at75°C for 3 minutes, and then loadedonto a 6% glycine SDS-PAGE. It was found that approximately 30-40%of the input RNA was converted to fusion product.