Dissecting limiting factors of the Protein synthesis Using Recombinant Elements (PURE) system

Abstract

Reconstituted cell-free protein synthesis systems such as the Protein synthesis Using Recombinant Elements (PURE) system give high-throughput and controlled access to in vitro protein synthesis. Here we show that compared with the commercial S30 crude extract based RTS 100 E. coli HY system, the PURE system has less mRNA degradation and produces up to ∼6-fold full-length proteins. However the majority of polypeptides PURE produces are partially translated or inactive since the signal from firefly luciferase (Fluc) translated in PURE is only ∼2/3^rd of that measured using the RTS 100 E. coli HY S30 system. Both of the 2 batch systems suffer from low ribosome recycling efficiency when translating proteins from 82 k_D to 224 k_D. A systematic fed-batch analysis of PURE shows replenishment of 6 small molecule substrates individually or in combination before energy depletion increased Fluc protein yield by ∼1.5 to ∼2-fold, while creatine phosphate and magnesium have synergistic effects when added to the PURE system. Additionally, while adding EF-P to PURE reduced full-length protein translated, it increased the fraction of functional protein and reduced partially translated protein probably by slowing down the translation process. Finally, ArfA, rather than YaeJ or PrfH, helped reduce ribosome stalling when translating Fluc and improved system productivity in a template-dependent fashion.

Keywords: cell-free protein synthesis; ribosome stalling; translation.

Figure 1.

Compare PURE system transcription and translation with RTS 100 E. coli HY S30 system. (A) RNA denaturing gel of PURE reaction transcribing TC-Fluc mRNA. PURE system with no plasmid template and plasmid encoding TC-Fluc were incubated at 37°C for 2 h. Total RNA were purified by phenol/chloroform extraction and isopropanol precipitation from equal volumes of the 2 reactions and then run on lane 2 and 3, respectively. (B) RNA denaturing gel of RTS 100 E. coli HY S30 system transcribing TC-Fluc mRNA. RTS reactions with no plasmid template and plasmid encoding TC-Fluc were incubated at 30°C for 6 h. Total RNAs were purified by phenol/chloroform extraction and isopropanol precipitation from equal volumes of the 2 reactions and then run on lane 2 and 3, respectively. (C) Assessment of functional TC-Fluc translated in PURE and RTS 100 E. coli HY S30 systems. Equal volume aliquots were taken for luciferase assay to measure the actual amount of functional TC-Fluc produced. Values represent averages and error bars are ± standard deviations, with n = 3. (D) Assessment of full-length TC-Fluc produced in PURE and RTS 100 E. coli HY S30 systems. TC-Fluc synthesized in (C)were incubated with FlAsH-EDT₂ biarsenical labeling reagent and analyzed on a SDS-PAGE. The gel was scanned by a typhoon scanner with filter set (508nmEx/528nmEm). Asterisk * indicates full-length product.

Figure 2.

Assessment of PURE and RTS 100 E. coli HY S30 system translation by production of HaloTag fusion proteins using linear DNA templates. HT-Ch-TolA, HT-EntE-Ch-TolA and HT-EntF-Ch-TolA with and without stop codon (82 k_D, 141 k_D, 224 k_D) were synthesized in the PURE system and S30 extract and then incubated with HaloTag TMR Ligand. Equal volume aliquots of samples from PURE system and S30 extract reactions were analyzed on a SDS-PAGE and then scanned by a typhoon scanner with filter set (555nmEx/580nmEm). Lane 1, 3, 5, 8, 10, 12 are HaloTag fusion proteins with stop codon. Lane 2, 4, 6, 9, 11, 13 are HaloTag fusion proteins without stop codon. Lane 7 and 14 are negative controls with no template. (HT: HaloTag, 34 k_D; Ch: mCherry, 29 k_D. TolA, 19 k_D, is a C-terminal 171-amino-acid α-helical spacer excised from E. coli TolA domain II. EntE, 59 k_D, and EntF, 142 k_D, are multidomain enzymes from E. coli enterobactin biosynthetic pathway.)

Figure 3.

Replenishing small molecule substrates in the PURE system protein synthesis at different time points. PURE system reactions synthesising TC-Fluc were set up and incubated at 37°C. Small molecule substrates as indicated were fed to the reaction at 15 min, 30 min, 45 min, 60 min or 75 min. Control reactions were fed equivalent volumes of water at 0 min. Overall functional TC-Fluc produced was assessed by luciferase assay at 90 min. Values represent averages and error bars are ± standard deviations, with n = 3. PURE system feed supplements of (A) 0.3 mM amino acids; (B) 27 OD₂₆₀/ml tRNAs; (C) 0.3 mM amino acids and 27 OD₂₆₀/ml tRNAs in combination; (D) 1 mM DTT; (E) 10 µg/ml 10-CHO-THF; (F) 2 mM magnesium acetate; G. 1.6 mM ATP, 1.6 mM GTP, 0.8 mM CTP and 0.8 mM UTP; (H) 4.8 mM magnesium acetate together with 1.6 mM ATP, 1.6 mM GTP, 0.8 mM CTP and 0.8 mM UTP; (I) 20 mM creatine; and (J) 20 mM creatine phosphate and 2 mM magnesium acetate in combination.

Figure 4.

Supplementing the PURE system with EF-P. (A) Time course study of functional TC-Fluc production in the PURE system with or without 4 µM EF-P. TC-Fluc activity was measured after incubation at 37°C for the indicated time lengths. Values represent averages and error bars are ± standard deviations, with n = 3. (*: p < 0.05; ***: p < 0.005; N.S: not significant) (B) Time course study of full-length TC-Fluc produced in the PURE system with or without 4 µM EF-P. PURE system reaction aliquots were harvested at different time points, were incubated with FIAsH-EDT₂, and separated by SDS-PAGE. Gel was scanned by a typhoon scanner with filter set (508nmEx/528nmEm). Arrow indicates full-length TC-Fluc.

Figure 5.

Supplementing the PURE system with ribosome rescuing systems. (A) Effect of ArfA presence to the PURE system TC-Fluc synthesis. PURE system was assembled with 9 nM different templates encoding TC-Fluc as indicated, with or without 8 µM ArfA protein. TC-Fluc activity was measured after 2 h incubation at 37°C. Values represent averages and error bars are ± standard deviations, with n = 3. (***: p < 0.005; N.S: not significant) (B) Effect of YaeJ presence to the PURE system TC-Fluc synthesis. PURE system was assembled with different templates encoding TC-Fluc as indicated, with or without 6 µM YaeJ protein. TC-Fluc activity was measured after 2 h incubation at 37°C. Values represent averages and error bars are ± standard deviations, with n = 3. (N.S: not significant).