METHODS AND MEANS FOR ENHANCING RNA PRODUCTION

Abstract

The present invention relates to a method for synthesizing an RNA molecule of a given sequence, comprising the step of determining the fraction (1) for each of the four nucleotides G, A, C and U in said RNA molecule, and the step of synthesizing said RNA molecule by in vitro transcription in a sequence-optimized reaction mix, wherein said sequence-optimized reaction mix comprises the four ribonucleoside triphosphates GTP, ATP, CTP and UTP, wherein the fraction (2) of each of the four ribonucleoside triphosphates in the sequence-optimized reaction mix corresponds to the fraction (1) of the respective nucleotide in said RNA molecule, a buffer, a DNA template, and an RNA polymerase.

Further, the present invention relates to a bioreactor (1) for synthesizing RNA molecules of a given sequence, the bioreactor (1) having a reaction module (2) for carrying out in vitro RNA transcription reactions in a sequence-optimized reaction mix, a capture module (3) for temporarily capturing the transcribed RNA molecules, and a control module (4) for controlling the infeed of components of the sequence-optimized reaction mix into the reaction module (2), wherein the reaction module (2) comprises a filtration membrane (21) for separating nucleotides from the reaction mix, and the control of the infeed of components of the sequence-optimized reaction mix by the control module (4) is based on a measured concentration of separated nucleotides.

Claims

1. A method for synthesizing an RNA molecule of a given sequence, comprising the following steps: a) determining the fraction (1) for each of the four nucleotides G, A, C and U in said RNA molecule, and b) synthesizing said RNA molecule by in vitro transcription in a sequence-optimized reaction mix, wherein said sequence-optimized reaction mix comprises the four ribonucleoside triphosphates (NTPs) GTP, ATP, CTP and UTP, wherein the fraction (2) of each of the four ribonucleoside triphosphates in the sequence-optimized reaction mix corresponds to the fraction (1) of the respective nucleotide in said RNA molecule, a buffer, a DNA template, and an RNA polymerase.

2. The method of claim 1, wherein step b) comprises the steps of b1) preparing a sequence-optimized ribonucleoside triphosphate (NTP) mix comprising the four ribonucleoside triphosphates (NTPs) GTP, ATP, CTP and UTP, wherein the fraction (2) of each of the four ribonucleoside triphosphates in the sequence-optimized ribonucleoside triphosphate (NTP) mix corresponds to the fraction (1) of the respective nucleotide in said RNA molecule, and b2) synthesizing said RNA molecule by in vitro transcription in the sequence-optimized reaction mix comprising the NTP mix of step (b1), a buffer, a DNA template, and an RNA polymerase.

3. The method of claim 1 or 2, wherein before the start of the in vitro transcription a start nucleotide is added to the sequence-optimized reaction mix which corresponds to the first nucleotide of said RNA molecule.

4. The method of claim 3, wherein said start nucleotide is a nucleoside monophosphate, a nucleoside diphosphate, a nucleoside triphosphate or a di-nucleoside triphosphate.

5. The method of claim 3, wherein said start nucleotide is a cap analog.

6. The method of any of claims 3 to 5, wherein said start nucleotide is added in excess compared to the fraction of that nucleotide in said RNA molecule which is found at the first position of said RNA molecule.

7. The method of any of claims 1 to 6, wherein for the nucleotides which do not correspond to the first nucleotide of the RNA molecule fraction (1) and fraction (2) differ by at most 10%.

8. The method of any of claims 1 to 7, wherein a part or all of at least one ribonucleoside triphosphate is replaced by a modified nucleoside triphosphate.

9. The method of claim 8, wherein said modified nucleoside triphosphate is selected from the group consisting of pseudouridine-5-triphosphate, 1-methylpseudouridine-5 -triphosphate, 2-thiouridine-5-triphosphate, 4-thiouridine-5-triphosphate and 5-methylcytidine-5-triphosphate.

10. The method of any of claims 1 to 9, wherein in the course of the in vitro transcription the sequence-optimized reaction mix is supplemented with the sequence-optimized ribonucleoside triphosphate (NTP) mix as defined in claim 2 b1).

11. The method of any of claims 1 to 10, wherein said RNA molecule is selected from the group consisting of non-coding and coding RNA molecules.

12. The method of any of claims 1 to 11, wherein said RNA molecule is an mRNA.

13. The method of any of claims 1 to 12, wherein said RNA molecule is longer than 100 nucleotides.

14. The method of any of claims 1 to 13, wherein said synthesizing of an RNA molecule of a given sequence is performed as a large scale synthesis.

15. The method of any of claims 1 to 14, wherein the NTP counter ion is tris(hydroxymethyl)-aminomethane (Tris).

16. The method of any of claims 1 to 15, wherein the synthesizing of said RNA molecule by in vitro transcription is followed by separating and quantifying the unincorporated NTPs.

17. The method of any of claims 1 to 16, wherein the synthesizing of said RNA molecule by in vitro transcription is carried out in a bioreactor (1).

18. The method of any of claim 17, wherein said bioreactor (1) comprises a DNA template immobilized on a solid support.

19. The method of any of claims 17 to 18, wherein said bioreactor (1) comprises a filtration membrane (21) for separating nucleotides from the sequence-optimized reaction mix.

20. The method of claim 19, wherein said filtration membrane (21) is selected from the group of regenerated cellulose, modified cellulose, poly sulfone (PSU), polyacrylonitrile (PAN), polymethylmethacrylate (PMMA), polyvinyl alcohol (PVA) and polyarylethersulfone (PAES).

21. The method of any of claims 19 to 20, wherein said filtration membrane (21) has a molecular weight cut-off in a range of approximately 10 to 50 kDa.

22. The method of any of claims 17 to 21, wherein said bioreactor (1) comprises a sensor unit (41) for the real-time measurement of the nucleotide concentration during the reaction.

23. The method of claim 22, wherein said sensor unit (41) measures the nucleotide concentration by photometric analysis.

24. The method of any of claims 17 to 23, wherein said bioreactor comprises a control module (4) which controls the addition of the sequence-optimized ribonucleoside triphosphate (NTP) mix as defined in claim 2 b1).

25. The method of any of claims 17 to 24, wherein said bioreactor (1) comprises an actuator (43) which adds the sequence-optimized ribonucleoside triphosphate (NTP) mix as defined in claim 2 b1).

26. The method of any of claims 17 to 25, wherein said bioreactor (1) comprises a resin to capture the RNA molecules and to separate the RNA molecules from the other components of the transcription reaction mix.

27. The method of any of claims 17 to 26, wherein said bioreactor (1) operates in a semi-batch mode or in a continuous mode.

28. The method of any of claims 17 to 27, wherein said bioreactor (1) comprises at least one ion-selective electrode,

29. The method of claim 28, wherein the at least one ion-selective electrode is used for measuring the concentration of one or more types of ions in a liquid comprised in at least one compartment of the bioreactor (1).

30. The method of claim 29, wherein the ion is selected from the group consisting of H.sup.+, Na.sup.+, K.sup.+, Mg.sup.2+, Ca.sup.2+, Cl.sup. and PO.sub.4.sup.3.

31. An RNA molecule obtainable by the method according to any one of claims 1 to 30.

32. Use of a sequence-optimized ribonucleoside triphosphate (NTP) mix optimized for an RNA molecule of a given sequence for the synthesis of said RNA molecule.

33. The use of claim 32, wherein the sequence-optimized NTP mix has been optimized by a method comprising the steps of: a) determining the fraction (1) of each of the four nucleotides G, A, C and U in said RNA molecule, and b) preparing the sequence-optimized ribonucleoside triphosphate (NTP) mix comprising the four ribonucleoside triphosphates GTP, ATP, CTP and UTP, wherein the fraction (2) of each of the four ribonucleoside triphosphates in the sequence-optimized ribonucleoside triphosphate (NTP) mix corresponds to the fraction (1) of the respective nucleotide in said RNA molecule.

34. The use of any of claims 32 to 33, wherein a part or all of at least one ribonucleoside triphosphate is replaced by a modified nucleoside triphosphate.

35. The use of any of claims 32 to 34, wherein said modified nucleoside triphosphate is selected from the group consisting of pseudouridine-5-triphosphate, 1-methylpseudouridine-5-triphosphate, 2-thiouridine-5-triphosphate, 4-thiouridine-5-triphosphate and 5-methylcytidine-5-triphosphate.

36. A sequence-optimized ribonucleoside triphosphate (NTP) mix for the synthesis of an RNA molecule of a given sequence comprising the four nucleoside triphosphates GTP, ATP, CTP and UTP, wherein the fraction (2) of each of the four nucleoside triphosphates in the sequence-optimized ribonucleoside triphosphate (NTP) mix corresponds to the fraction (1) of the respective nucleotide in said RNA molecule.

37. A kit comprising a sequence-optimized ribonucleoside triphosphate (NTP) mix optimized for an RNA molecule of a given sequence as defined in claim 32 or the components thereof.

38. A bioreactor (1) for synthesizing RNA molecules of a given sequence, comprising: i) a reaction module (2) for carrying out in vitro transcription reactions in a sequence-optimized reaction mix as defined in claim 1 (b); ii) a capture module (3) for temporarily capturing the transcribed RNA molecules; and iii) a control module (4) for controlling the infeed of components of the sequence-optimized reaction mix into the reaction module (2), wherein the reaction module (2) comprises a filtration membrane (21) for separating nucleotides from the sequence-optimized reaction mix, and wherein the control of the infeed of components of the sequence-optimized reaction mix by the control module (4) is based on a measured concentration of separated nucleotides.

39. The bioreactor (1) of claim 38, wherein the filtration membrane (21) is an ultrafiltration membrane (21) for separation of high molecular weight components from low molecular weight components, preferably wherein said filtration membrane (21) has a molecular weight cut-off in a range from 10 to 100 kDa, 10 to 75 kDa, 10 to 50 kDa, 10 to 25 kDA or 10 to 15 kDa, further preferably the filtration membrane has a molecular weight cut-off value in a range of approximately 10 to 50 kDa.

40. The bioreactor (1) of claim 38 or 39, wherein the filtration membrane (21) is selected from the group of regenerated cellulose, modified cellulose, polysulfone (PSU), polyacrylonitrile (PAN), polymethylmethacrylate (PMMA), polyvinyl alcohol (PVA) and polyarylethersulfone (PAES).

41. The bioreactor (1) of any one of claims 38 to 40, wherein said reaction module (2) comprises a DNA template immobilized on a solid support as basis for the RNA transcription reaction.

42. The bioreactor (1) of any one of claims 38 to 41, wherein said capture module (3) comprises a resin to capture the transcribed RNA molecules and to separate the transcribed RNA molecules from other soluble components of the transcription reaction mix.

43. The bioreactor (1) of any one of claims 38 to 42, wherein said capture module (3) comprises means (31) for purifying the captured transcribed RNA molecules.

44. The bioreactor (1) of any one of claims 38 to 43, wherein said capture module (3) comprises means (31) for eluting the captured transcribed RNA molecules, preferably by means of an elution buffer.

45. The bioreactor (1) of any one of claims 38 to 44, wherein the bioreactor (1) further comprises a reflux module (5) for returning the residual filtrated reaction mix to the reaction module (2) from the capture module (3) after capturing the transcribed RNA molecules, preferably wherein the means (51) for returning the residual filtrated reaction mix is a pump (51).

46. The bioreactor (1) of claim 45, wherein the reflux module (5) comprises immobilized enzymes or resin to capture disruptive components.

47. The bioreactor (1) of any one of claims 38 to 46, wherein said control module (4) comprises a sensor unit (41) for the real-time measurement of the concentration of separated nucleotides during the reaction.

48. The bioreactor (1) of claim 47, wherein said sensor unit (41) measures, as a transcription reaction parameter, the concentration of separated nucleotides by photometric analysis.

49. The bioreactor (1) of claim 48, wherein said sensor unit (41) measures further transcription reaction parameters in the filtrated reaction mix, preferably wherein the further transcription reaction parameters are pH-value and/or salinity.

50. The bioreactor (1) of any one of claims 38 to 49, wherein said control module (4) controls an addition of the sequence-optimized ribonucleoside triphosphate (NTP) mix of claim 32 to the sequence-optimized reaction mix, preferably wherein said bioreactor (1) comprises an actuator (43) for addition of the sequence-optimized ribonucleoside triphosphate (NTP) mix to the sequence-optimized reaction mix.

51. The bioreactor (1) of any one of claims 38 to 49, wherein said bioreactor (1) operates in a semi-batch mode or in a continuous mode

52. The bioreactor (1) of any one of claims 38 to 49, wherein said bioreactor (1) is adapted to carry out the method of any one of claims 1 to 30.

Description

BRIEF DESCRIPTION OF THE FIGURES

[0179] The figures shown in the following are merely illustrative and shall describe the present invention in a further way. These figures shall not be construed to limit the present invention thereto.

[0180] FIG. 1: G/C optimized mRNA sequence of R1871 coding for Homo sapiens prostate stem cell antigen (HsPSCA), which corresponds to SEQ ID NO: 1.

[0181] FIG. 2: G/C optimized mRNA sequence of R2988 coding for Photinus pyralis luciferase (PpLuc), which corresponds to SEQ ID NO: 2.

[0182] FIG. 3: G/C optimized mRNA sequence of R1626 coding for Homo sapiens Mucin-1 signal peptide/epidermal growth factor receptor/Mucin-1 fusion protein (EGFR/Mucin-1), which corresponds to SEQ ID NO: 3.

[0183] FIG. 4: Non-coding immunostimulatory RNA sequence of R2025, which corresponds to SEQ ID NO: 4.

[0184] FIG. 5: RNA yields over time for mRNA encoding HsPSCA (R1871) and EGFR/Mucin-1 (R1626). mRNAs were synthesized by in vitro transcription as shown in Example 1. [0185] (A) The RNA yield of standard transcription reactions reaches after about 30 minutes a plateau of approx. 1.4 mg/ml RNA for the 5337 nucleotide long RNA encoding EGFR/Mucin-1 (R1626) and of approx. 1.8 mg/ml RNA for the 589 nucleotide long RNA encoding HsPSCA (R1871). [0186] (B) The RNA yield of sequence-optimized transcription reactions is significantly higher compared to standard transcription reactions. After 60 minutes (R1626) and 120 minutes (R1871) both RNAs reach a similar plateau of approximately 3.9 mg/ml RNA. Mean and standard deviation of triplicates are shown.

[0187] FIG. 6: Comparison of RNA yields obtained from using standard and sequence-optimized nucleotide (CapNTP) mixes for mRNA encoding HsPSCA (R1871), Luciferase (PpLuc, R2988) and EGFR/Mucin-1 (R1626). The experiment was performed as described in Example 1 (reaction time 5 hours). [0188] The RNA yields for the three different RNA molecules of different lengths are roughly the same for each type of transcription reaction. However, different yields are obtained depending on the nucleotide mix used for the in vitro transcription. Standard transcription (equal NTP concentration for each NTP) yields about 1.5 mg/ml RNA, transcription with a twofold concentrated Cap-NTP mix (2CapNTP) about 3.0 mg/ml RNA, sequence-optimized (seq-opt) transcription about 3.9 mg/ml RNA and sequence-optimized transcription with NTP feed about 6.75 mg/ml RNA. Mean and standard deviation of triplicates are shown.

[0189] FIG. 7: Analysis of capping efficiency achieved by standard and sequence-optimized in vitro transcription of Photinus pyralis Luciferase (PpLuc) mRNA. [0190] RNAs were cleaved with the hammerhead ribozyme HHNUH2d as described in example 2 and the resulting RNA fragments were separated by denaturing polyacrylamide gel electrophoresis (dPAGE). Non-capped (no cap) and enzymatically capped (E-cap) RNAs served as controls. [0191] Comparable capping efficiencies were achieved when using standard and sequence-optimized NTP mixes for the synthesis of mRNAs encoding Photinus pyralis Luciferase (PpLuc).

[0192] FIG. 8: Comparison of RNA yields using UTP and pseudo-UTP in sequence-optimized CapNTP mixes for Mucin-1 signal peptide/epidermal growth factor receptor/Mucin-1 fusion protein (EGFR/Mucin-1) mRNA (R1626) and prostate stem cell antigen (HsPSCA) mRNA (R1871). The experiments were performed as described in Example 3. In the mixes UTP was replaced with 0%, 10%, or 100% pseudo-UTP (psU) as indicated. The mean and standard deviation of triplicates are shown.

[0193] FIG. 9: Comparison of theoretical and actual RNA yields for the noncoding immunostimulatory RNA R2025 using standard (equal) and sequence-optimized (seqopt) NTP mixes in the presence of additional nucleotides (13.45 mM total NTP concentration; 13.45 mM cap or GTP for equal mixes (4-fold excess over GTP); 11.2 mM cap or GTP for sequence-optimized NTP mixes (4-fold excess over GTP)). White bars: actual yields. Black bars: theoretical yields. The experiments were performed as described in Example 4. Mean and standard deviation of triplicates are shown.

[0194] FIG. 10: Comparison of theoretical and actual RNA yields for the mRNA encoding Homo sapiens prostate stem cell antigen (HsPSCA) (R1871) using standard (equal) and sequence-optimized NTP ratios in the presence of additional nucleotides (13.45 mM total NTP concentration; 13.45 mM cap or GTP for equal mixes (4-fold excess over GTP); 13.7 mM cap or GTP for sequence-optimized NTP mixes (4-fold excess over GTP)). White bars: actual yields. Black bars: theoretical yields. The experiments were performed as described in Example 4.

[0195] FIG. 11: Transcription efficiencies using sequence-optimized NTP mixes of Na-NTPs or Tris-NTPs, in the presence of respective added salts (NaCl; Tris/HCl, pH7.5). The experiment was performed as described in Example 5.

[0196] FIG. 12: Monitoring of the progress of the sequence-optimized transcription reaction by measuring the amount of produced RNA and the consumption of nucleotide mix. The experiment was performed as described in Example 6.

[0197] FIG. 13: RNA yields for mRNA encoding Photinus pyralis luciferase (PpLuc) (R2988) depending on cap concentration. The mRNA was synthesized by in vitro transcription using total NTP concentrations of 4 mM and 13.45 mM of the PpLuc sequence-optimized NTP mix in the presence of varying concentrations of cap analog. The experiment was performed as described in Example 7. (A) Actual RNA [mg/ml]. (B) Relative RNA yield [%].

[0198] FIG. 14: RNA yields for mRNA encoding Homo sapiens prostate stem cell antigen (HsPSCA) (R1871) depending on cap concentration. The mRNA was synthesized by in vitro transcription using total NTP concentrations of 2 mM, 4 mM and 13.45 mM of the HsPSCA sequence-optimized NTP mix in the presence of varying concentrations of cap analog. The experiment was performed as described in Example 7. (A) Actual RNA yield [mg/ml]. (B) Relative RNA yield [%].

[0199] FIG. 15: RNA yields for mRNA encoding Homo sapiens prostate stem cell antigen (HsPSCA) (R1871) depending on GTP start nucleotide concentration. The mRNA was synthesized by in vitro transcription using total NTP concentrations of 13.45 mM of the HsPSCA sequence-optimized NTP mix to which GTP start nucleotide was added up to a concentration of 20 mM. The experiment was performed as described in Example 8. (A) Actual RNA yield [mg/ml]. (B) Relative RNA yield [%].

[0200] FIG. 16: RNA yields for mRNA encoding Photinus pyralis luciferase (PpLuc) (R2988) depending on GTP start nucleotide concentration. The mRNA was synthesized by in vitro transcription using total NTP concentrations of 13.45 mM of the PpLuc sequence-optimized NTP mix to which GTP start nucleotide was added up to a concentration of 20 mM. The experiment was performed as described in Example 8. (A) Actual RNA yield [mg/ml]. (B) Relative RNA yield [%].

[0201] FIG. 17: RNA yields for mRNA encoding EGFR/Mucin-1 (R1626) depending on GTP start nucleotide concentration. The mRNA was synthesized by in vitro transcription using total NTP concentrations of 13.45 mM of the EGFR/Mucin-1 sequence-optimized NTP mix to which GTP start nucleotide was added up to a concentration of 20 mM. The experiment was performed as described in Example 8. (A) Actual RNA yield [mg/ml]. (B) Relative RNA yield [%].

[0202] FIG. 18: Bioreactor in a schematic illustration, including modules for continuous or semi-batch process, with resin-immobilized linear DNA as template for the transcription reaction.

[0203] FIG. 19: Reduced immunostimulation by RNA synthesized with a sequence-optimized NTP mix compared to a standard equimolar NTP mix. Cytokine and chemokine levels in cell supernatants were measured as described in Example 10.

[0204] FIG. 20: G/C optimized mRNA sequence encoding HA from Influenza A H1N1 (Netherlands 2009), which corresponds to SEQ ID NO: 6 (Example 11).

[0205] FIG. 21: RNA yields over time for mRNA encoding HA (Example 11). The RNA yield at different time points is shown for RNA obtained by in vitro transcription in a bioreactor using a standard NTP mix (TS(1)), by sequence-optimized transcription in a bioreactor without feed (TS(2)), by sequence-optimized transcription in a bioreactor with feed (TS(3)), or by sequence-optimized transcription in a bioreactor with reduced T7 RNA polymerase concentration and reduced template concentration (TS(4)), respectively.

[0206] FIG. 22: Surface expression of the HA protein as determined by using flow cytometric analysis (Example 11). The geometric mean of fluorescence intensity (GMFI) was determined for cells transfected with RNA obtained by in vitro transcription in a bioreactor using a standard NTP mix (TS(1)), by sequence-optimized transcription in a bioreactor without feed (TS(2)), by sequence-optimized transcription in a bioreactor with feed (TS(3)), or by sequence-optimized transcription in a bioreactor with reduced T7 RNA polymerase concentration and reduced template concentration (TS(4)), respectively.

[0207] FIG. 23: Immunostimulation by RNA synthesized by in vitro transcription in a bioreactor using a standard NTP mix (TS(1)), by sequence-optimized transcription in a bioreactor without feed (TS(2)), by sequence-optimized transcription in a bioreactor with feed (TS(3)), or by sequence-optimized transcription in a bioreactor with reduced T7 RNA polymerase concentration and reduced template concentration (TS(4)), respectively. Cytokine and chemokine levels in cell supernatants were measured as described in Example 11.

EXAMPLES

[0208] The examples shown in the following are merely illustrative and shall describe the present invention in a further way. These examples shall not be construed to limit the present invention thereto.

Example 1: Preparation of the mRNA

[0209] 1. Preparation of DNA and mRNA Constructs

[0210] For the present example DNA sequences encoding Homo sapiens prostate stem cell antigen (HsPSCA) mRNA (R1871), Photinus pyralis Luciferase (PpLuc) mRNA (R2988) and Mucin-1 signal peptide/epidermal growth factor receptor/Mucin-1 fusion protein (EGFR/Mucin-1) (R1626) were prepared and used for subsequent in vitro transcription reactions.

[0211] According to a first preparation, a vector for in vitro transcription was constructed containing a T7 promoter followed by a sequence coding for the above mentioned proteins. The constructs were prepared by modifying the wild type coding sequence by introducing a GC-optimized sequence for stabilization, followed by a stabilizing sequence derived from the alpha-globin-3-UTR (muag (mutated alpha-globin-3-UTR)), a stretch of 64 adenosines (poly-A-sequence), a stretch of 30 cytosines (poly-C-sequence), and a histone stem loop.

[0212] In addition, a vector for in vitro transcription was constructed containing a T7 promoter followed by the sequence encoding an immunostimulatory RNA (R2025), which does not encode a protein.

[0213] The RNA constructs and their nucleotide compositions are listed in Table 1 and Table 2, respectively.

TABLE-US-00003 TABLE 1 RNAs Identifier SEQ Description (R number) Sequence ID No. HsPSCA R1871 FIG. 1 1 mRNA PpLuc mRNA R2988 FIG. 2 2 EGFR/Mucin-1 mRNA R1626 FIG. 3 3 Non-coding RNA R2025 FIG. 4 4

TABLE-US-00004 TABLE 2 Nucleotide composition of RNAs Length RNA (nt) G C A U HsPSCA 589 150 205 154 80 (25.5%) (34.8%) (26.1%) (13.6%) PpLuc 1870 571 604 428 267 (30.5%) (32.3%) (22.9%) (14.3%) EGFR/ 5337 1630 1967 1086 654 Mucin-1 (30.5%) (36.9%) (20.3%) (12.3%) Non-coding 547 114 111 112 210 RNA (20.8%) (20.2%) (20.5%) (38.4%)

2. In Vitro Transcription

[0214] The respective DNA plasmids prepared according to paragraph 1 were transcribed in vitro using T7 polymerase. Subsequently the mRNA was purified using PureMessenger (CureVac, Tubingen, Germany; WO2008/077592A1).

[0215] The standard transcription reaction volume was 20 l. For subsequent HPLC purification of mRNAs, e.g. for cap analysis, 1 ml reactions were set up.

[0216] Linearized DNA plasmid templates (50 g/ml) were transcribed at 37 C. for three hours (or as indicated) in 80 mM HEPES/KOH, pH 7.5, 24 mM MgCl.sub.2, 2 mM spermidine, 40 mM DTT, 5 U/ml pyrophosphatase (Thermo Fisher Scientific), 200 U/ml Ribolock RNase inhibitor (Thermo Fisher Scientific), 5000 U/ml T7 RNA polymerase (Thermo Fisher Scientific). Ribonucleoside triphosphates (NTPs) were added according to sections 3 to 7 below, respectively. Following transcription, the DNA template was removed by DNaseI digestion (Roche) (100 U/ml, 1 mM CaCl.sub.2), 30 minutes at 37 C.).

[0217] RNAs were precipitated in 2.86 M LiCl in a 3.45-fold reaction volume for 16 hours at 20 C., followed by centrifugation (30 minutes, 16.000 g, 4 C.). Pellets were washed in five transcription reaction volumes of 75% ethanol (invert tubes, centrifuge 5 minutes, 16.000 g, 4 C.), dried and re-dissolved in 2.5 transcription reaction volumes H.sub.2O.

[0218] RNA yields were determined via absorbance measurement at 260 nm using a NanoDrop Spectrophotometer. One absorbance unit at 260 nm corresponds to 40 ng/l of RNA (1 A260=40 ng/l RNA).

[0219] To determine the number of incorporated nucleotides, the total amount of RNA produced was converted to the number of molecules produced by dividing by the molecular mass. Multiplying by the number of the respective nucleotide present in the sequence yielded the incorporated nucleotides. To determine the remaining nucleotides (in %) at the end of the transcription reaction, this number was divided by the number of nucleotides available, according to:

[00001]$\begin{array}{cc}\mathrm{NTP}\left(\mathrm{remaining}\right).Math.[.Math.\%]=\left[1-\frac{\mathrm{RNA}.Math..Math.\mathrm{yield}*\mathrm{number}.Math..Math.\mathrm{of}.Math..Math.\mathrm{NTP}.Math..Math.\mathrm{in}.Math..Math.\mathrm{mRNA}}{\left[\mathrm{NTP}.Math..Math.\left(\mathrm{start}\right)\right]*\mathrm{reaction}.Math..Math.\mathrm{volume}}\right]*100& \mathrm{Equation}.Math..Math.\left(1\right)\end{array}$

[0220] RNA yield indicates the number of molecules produced per reaction (nmol). The NTP starting concentration [NTP (start)] is indicated in mM, the reaction volume in l.

[0221] To calculate the remaining concentration of the respective nucleotides, NTPs available at the beginning of the reaction were multiplied by the percentage of remaining NTPs at the end of a transcription reaction (see above) according to:

NTP(remaining)[mM]=NTP(start[mM]*NTP(remaining[%]Equation (2):

3. Standard In Vitro Transcription in the Presence of Cap Analog

[0222] For the production of 5-capped RNAs using cap analog, standard transcription was carried out with 5.8 mM m7G(5)ppp(5)G cap analog, 4 mM ATP, 4 mM CTP, 4 mM UTP, and 1.45 mM GTP (all Thermo Fisher Scientific) (see Table 3). The cap analog and GTP were used at a ratio of 4:1.

TABLE-US-00005 TABLE 3 Nucleotide concentrations (mM) for standard in vitro transcription reactions RNA CAP G C A U HsPSCA 5.8 1.45 4 4 4 PpLuc 5.8 1.45 4 4 4 EGFR/ 5.8 1.45 4 4 4 Mucin-1

TABLE-US-00006 Amount of nucleotides remaining at the end of standard transcription reactions (after 2.5 hours, in percent of nucleotides at the start of the reaction) RNA CAP G C A U HsPSCA 99.86 17.35 59.05 69.24 84.02 PpLuc 99.96 16.39 67.94 77.28 85.83 EGFR/ 99.99 16.37 63.42 79.80 87.84 Mucin-1

TABLE-US-00007 TABLE 5 Nucleotide concentrations (mM) remaining at the end of a standard in vitro transcription reaction (after 2.5 hours) RNA CAP G C A U HsPSCA 5.79 0.25 2.36 2.77 3.36 PpLuc 5.80 0.24 2.72 3.09 3.43 EGFR/ 5.80 0.24 2.54 3.19 3.51 Mucin-1

[0223] The typical yield of RNA transcripts in a standard transcription is about 1.5 mg/ml reaction.

4. In Vitro Transcription in the Presence of Cap Analog Using Double Concentrations of Cap Analog and NTPs (2CapNTP)

[0224] Cap analog and NTP concentrations were doubled compared to standard transcription conditions, so that reactions were carried out in 11.6 mM m7G(5)ppp(5)G cap analog, 8 mM ATP, 8 mM CTP, 8 mM UTP, and 2.9 mM GTP (all Thermo Fisher Scientific) (see Table 3). The cap analog and GTP were used at a ratio of 4:1.

TABLE-US-00008 TABLE 6 Nucleotide concentrations (mM) for 2xCapNTP in vitro transcription reactions RNA CAP G C A U HsPSCA 11.6 2.9 8 8 8 PpLuc 11.6 2.9 8 8 8 EGFR/ 11.6 2.9 8 8 8 Mucin-1

TABLE-US-00009 TABLE 7 Amount of nucleotides remaining at the end of 2xCapNTP transcription reactions (after 2.5 hours, in percent of nucleotides at the start of the reaction) RNA CAP G C A U HsPSCA 99.87 23.45 62.08 71.51 85.20 PpLuc 99.96 17.93 68.53 77.70 86.09 EGFR/ 99.99 20.15 65.07 80.72 88.39 Mucin-1

[0225] The typical yield of a transcription using double concentrations of cap analog and NTPs is about 3 mg/ml reaction.

5. Sequence-Optimized In Vitro Transcription in the Presence of Cap Analog

[0226] For sequence-optimized in vitro transcription reactions the concentration of ribonucleoside triphosphates (NTPs) was calculated for each individual sequence according to the nucleotide composition of the sequence (Table 2) so that the total concentration of all for NTPs was 13.45 mM as in standard transcription reactions. The concentration of the cap analog was four times higher than the calculated concentration for GTP so that a cap/GTP ratio of 4:1 was obtained.

TABLE-US-00010 TABLE 8 Nucleotide concentrations (mM) for sequence-optimized in vitro transcription RNA CAP G C A U HsPSCA 13.6 3.4 4.7 3.5 1.8 PpLuc 16.4 4.1 4.3 3.1 1.9 EGFR/ 16.4 4.1 5.0 2.7 1.7 Mucin-1

TABLE-US-00011 TABLE 9 Amount of nucleotides remaining at the end of sequence- optimized transcription (after 2.5 hours, in percent of nucleotides at the start of the reaction) RNA CAP G C A U HsPSCA 99.86 14.83 14.71 14.57 14.83 PpLuc 99.96 14.62 14.72 14.76 14.85 EGFR/ 99.99 14.60 14.82 14.51 15.04 Mucin-1

TABLE-US-00012 TABLE 10 Nucleotide concentrations (mM) remaining at the end of a sequence-optimized in vitro transcription reaction (after 2.5 hours) RNA CAP G C A U HsPSCA 13.58 0.51 0.69 0.51 0.27 PpLuc 16.39 0.60 0.64 0.45 0.29 EGFR/ 16.40 0.60 0.74 0.40 0.25 Mucin-1

[0227] The typical RNA yield of a transcription using sequence-optimized cap analog and NTPs is about 3.9 mg/ml reaction.

6. Sequence-Optimized In Vitro Transcription in the Presence of Cap Analog with NTP Feed

[0228] For sequence-optimized in vitro transcription reactions the concentration of ribonucleoside triphosphates (NTPs) was calculated for each individual sequence according to the nucleotide composition of the sequence (Table 2) so that the total concentration of all for NTPs was 13.45 mM as in standard transcriptions. The concentration of the cap analog was four times higher than the calculated concentration for GTP so that a cap/GTP ratio of 4:1 was obtained (see Table 7).

[0229] For the NTP feed, 13.45 mM NTPs without cap analog were added (in a volume of 2.69 l) to the reaction mix after 2.5 hours. As at this time point >99% of cap analog was still present in the transcription reaction, the 4:1 Cap/GTP ratio could be retained.

TABLE-US-00013 TABLE 11 Amount of nucleotides remaining at the end of sequence-optimized transcription with NTP feed (after 5 h, in percent of nucleotides at the start of the reaction) RNA CAP G C A U HsPSCA 99.75 26.3 26.2 26.1 26.3 PpLuc 99.94 26.1 26.2 26.2 26.3 EGFR/ 99.98 26.1 26.3 26.0 26.5 Mucin-1

[0230] The typical RNA yield of a transcription using sequence-optimized cap analog and NTPs followed by NTP feed is around 6.75 mg/ml reaction.

7. Standard In Vitro Transcription of Non-Capped RNAs

[0231] For the production of non-capped, 5 triphosphate RNAs, transcription was carried out in the presence of 4 mM of each ATP, GTP, CTP and UTP (all Thermo Fisher Scientific). Non-capped RNAs were used as control in the capping analysis assay (FIG. 7).

8. Enzymatic Capping of mRNA

[0232] Enyzmatic capping was performed using the ScriptCap m.sup.7G Capping System (Cellscript, Madison, Wis., USA) according to the manufacturer's instructions. In brief, per reaction, 60 g of non-capped RNAs were heat-denatured (10 minutes, 65 C.) in a volume of 68.5 l and immediately cooled on ice (5 minutes). Following addition of reaction components (lx ScriptCap Capping buffer, 1 mM GTP, 0.1 mM SAM, 1000 U/ml ScripGuard RNase Inhibitor, 400 U/ml ScriptCap Capping Enzyme) to a final volume of 100 l, reactions were incubated for 1 hour at 37 C. RNAs were precipitated in 2.86 M LiCl in a 3.45-fold reaction volume for 16 hours at 20 C., followed by centrifugation (30 minutes, 16.000 g, 4 C.). Pellets were washed in 0.5 reaction volumes 75% ethanol (invert, centrifuge 5 minutes, 16000 g, 4 C.), dried and re-dissolved in H.sub.2O. Enzymatically capped RNAs were used as control in the capping analysis assay (FIG. 7).

9. Results

[0233] The RNA yield of standard and sequence-optimized in vitro transcription reactions was determined at defined time points for up to two hours as described above (paragraph 2).

[0234] As can be seen from FIG. 5A, after about 30 minutes the RNA yield of standard transcription reactions reaches a plateau of about 1.4 mg/ml for the 5337 nucleotide long RNA encoding EGFR/Mucin-1 (R1626) and of about 1.8 mg/ml for the 589 nucleotide long RNA encoding HsPSCA (R1871).

[0235] As can be seen from FIG. 5B, the RNA yield of sequence-optimized transcription reactions is significantly higher compared to standard transcription reactions. After 60 minutes (R1626) and 120 minutes (R1626), respectively, both RNAs reach a similar plateau of approximately 3.9 mg/ml.

[0236] As can be seen from FIG. 6, the RNA yields for the three different RNA molecules of different length are roughly the same for each type of transcription reaction after five hours. Standard transcription (equal NTP concentration) yields about 1.5 mg/ml RNA, transcription with a twofold concentrated Cap-NTP mix (2CapNTP) about 3.0 mg/ml RNA, sequence-optimized transcription about 3.9 mg/ml RNA and sequence-optimized transcription with NTP feed about 6.75 mg/ml RNA.

[0237] Thus, the sequence-optimized transcription reaction results in an about threefold increase in RNA yield compared to standard transcription reactions. This yield can be further increased by about twofold by supplementing the reaction with NTP (NTP feed).

Example 2: CAP Analysis Assay

1. Principle of the Assay

[0238] The hammerhead ribozyme HHNUH2d (5-GCAUGGCUGAUGAGGCCUCGACCGAUAGGUCGAGGCCGAAAAGCUUUCUCCC-3) (SEQ ID NO: 5) was incubated with the in vitro transcribed RNAs of example 1 and the cleavage products were separated by denaturing polyacrylamide-gel-electrophoresis (dPAGE).

2. Ribozyme Cleavage Reaction

[0239] Per reaction, 10 pmol of HHNUH2d and 10 pmol of the respective generation 4 RNA were annealed in 0.625 mM EDTA in a total volume of 6 l (2 minutes at 95 C., 0.1 C./seconds to 25 C., 10 minutes at 25 C.). After addition of 4 l of 100 mM MgCl.sub.2, 125 mM Tris/HCl, pH 7.5 (final concentration 40 mM MgCl.sub.2, 50 mM Tris/HCl), the reaction was incubated at 25 C. for one hour. For analysis via PAGE, the 1 reaction was stopped with 30 l 95% formamide, 20 mM EDTA.

3. Gel Separation, Quantification of Cleavage Products and Calculation of Capping Degree

[0240] Stopped reactions were heat-denatured (heated to 80 C. for 2 minutes, immediately put on ice for 5 minutes) and separated on a 10 cm8 cm1.0 mm 20% denaturing polyacrylamide gel (8 M urea (AppliChem), 20% acrylamid:bisacrylamid 19:1 (AppliChem), lx TBE, 1% APS (AppliChem), 0.1% TEMED (AppliChem); 180 V, 2 hours, Mini-PROTEAN Tetra Cell (BioRad)). Gels were stained for 10 minutes in 1:10,000 SYBR Gold (Invitrogen) in TBE and documented on a E-BOX VX2 gel documentation system with 312 nm-UV Transilluminator (Peqlab) (excitation maximum for SYBR Gold: 300 nm, emission: 537 nm).

[0241] To determine the capped proportion in the mRNA preparations, bands of the respective 13-mer (derived from the non-capped fraction) or 14-mer (derived from the capped fraction) cleavage products were quantified using Quantity One 1-D Analysis Software (BioRad). The degrees of capped and non-capped RNA, respectively, were calculated according to:

[00002]$\begin{array}{cc}\mathrm{capped}.Math..Math.\mathrm{RNA}.Math..Math.\left(\%\right)=\frac{\mathrm{signal}.Math..Math.\mathrm{intensity}.Math..Math.14.Math.\mathrm{mer}}{.Math.\mathrm{signal}.Math..Math.\mathrm{intensities}.Math..Math.\left(13.Math.\mathrm{mer}+14.Math.\mathrm{mer}\right)}100& \mathrm{Equation}.Math..Math.\left(4\right)\\ \mathrm{non}.Math.\text{-}.Math.\mathrm{capped}.Math..Math.\mathrm{RNA}.Math..Math.\left(\%\right)=\frac{\mathrm{signal}.Math..Math.\mathrm{intensity}.Math..Math.13.Math.\mathrm{mer}}{.Math.\mathrm{signal}.Math..Math.\mathrm{intensities}.Math..Math.\left(13.Math.\mathrm{mer}+14.Math.\mathrm{mer}\right)}100& \mathrm{Equation}.Math..Math.\left(5\right)\end{array}$

4. Results

[0242] As can be seen in FIG. 7, comparable capping efficiencies were achieved for standard and sequence-optimized NTP mixes for Photinus pyralis Luciferase (PpLuc) mRNA.

Example 3: Comparison of RNA Yields Using UTP and Pseudo-UTP in Sequence-Optimized Nucleotide Mixes

[0243] In vitro transcription reactions can be performed by replacing one or more of the four nucleotides ATP, GTP, CTP and UTP by nucleotide analogs. Examples of such modified NTPs are pseudouridine (psU or ) triphosphate and 5-methylcytidine (5mC) triphosphate. The percentage of the modified nucleotide in the mix can be varied from 0% to 100% of the natural nucleotide that it replaces.

[0244] To test whether it is possible to use modified nucleotides such as pseudouridine (psU) triphosphate in sequence-optimized nucleotide mixes, UTP was replaced by 10%, and 100% pseudouridine triphosphate. In a control reaction, 100% UTP was used.

Sequence-Optimized In Vitro Transcription in the Presence of Cap Analog

[0245] For sequence-optimized in vitro transcription reactions the concentration of ribonucleoside triphosphates (NTPs) was calculated for each individual sequence according to the nucleotide composition of the sequence (Table 2) so that the total concentration of all for NTPs was 13.45 mM as in standard transcription reactions. The concentration of the cap analog was four times higher than the calculated concentration for GTP so that a CAP/GTP ratio of 4:1 was obtained.

Results As can be seen from FIG. 8, using UTP and pseudo-UTP in sequence-optimized nucleotide mixes with cap analog (CapNTP mixes) results in comparable RNA yields independent of the pseudo-UTP percentage in the sequence-optimized nucleotide mix. This was demonstrated for two different mRNAs encoding Mucin-1 signal peptide/epidermal growth factor receptor/Mucin-1 fusion protein (EGFR/Mucin-1) (R1626) and prostate stem cell antigen (HsPSCA) mRNA (R1871), respectively.

Example 4: Comparison of Theoretical and Actual RNA Yields Using Standard and Sequence-Optimized Nucleotide Mixes

[0246] Transcription reactions were assembled as described in Example 1, Section 2. The NTPs were either equally distributed (equimolar) or distributed according to the sequence of the produced RNA as described in Example 1, Section 5. For some reactions, an additional nucleotide (GTP or cap analog) was added at a ratio of 4:1 over GTP.

Results

[0247] As can be seen from FIG. 9, the actual RNA yield for R2025 can be increased for sequence-optimized NTP mixes compared to standard NTP mixes (equal NTP mix).

[0248] As can be seen from FIG. 10, the actual RNA yield for the mRNA encoding Homo sapiens prostate stem cell antigen (HsPSCA; R1871) can be increased for sequence-optimized NTP mixes compared to standard NTP mixes (equal NTP mix).

Example 5: Influence of NTP Counter Ions on RNA Yields

[0249] The impact of NTP counter ions on RNA yield was investigated using mRNA encoding Homo sapiens Mucin-1 signal peptide/epidermal growth factor receptor/Mucin-1 fusion protein (EGFR/Mucin-1, R1626) as example. Transcription reactions were assembled as described in Example 1, Section 2, using sequence-optimized NTP ratios and a total NTP concentration of 13.45 mM. NTPs contained either Na.sup.+ or Tris.sup.+ (both Thermo Scientific) as counter ions. In addition, Na-NTP reactions were supplemented with different concentrations of NaCl, Tris-NTP reactions with Tris/HCl. After 2.5 hours of reaction time, the RNAs were purified and their concentration was determined as described in Example 1, Section 2.

Results

[0250] As can be seen from FIG. 11, the RNA yield for Homo sapiens Mucin-1 signal peptide/epidermal growth factor receptor/Mucin-1 fusion protein (EGFR/Mucin-1, R1626) using a sequence-optimized NTP mix remained roughly the same up to a concentration of 150 mM Tris-HCl. By contrast, the RNA yield started to decline at NaCl concentrations above 75 mM.

[0251] The negative impact of high NaCl concentrations on RNA yields has been described (e.g. Kern et al., 1997. Biotechnol. Prog., 13, 747-756; U.S. Pat. No. 6,586,218 B2). High concentrations of Na-NTPs, especially as consequence when pursuing a NTP feeding strategy, could therefore result in decreased RNA yields. This limitation should be circumvented with Tris-NTPs, because the polymerase activity is not affected by high Tris/HCl concentrations.

Example 6: Monitoring of the Progress of the Transcription Reaction

[0252] Larger-scale transcription reactions (350 l) of Homo sapiens prostate stem cell antigen (HsPSCA; R1871) were assembled as described in Example 1 Section 2, using sequence-optimized NTP ratios and a total NTP concentration of 13.45 mM Tris-NTPs. Cap analog was present in a 4:1 excess over GTP. At defined time points (15/30/60/90/120 minutes after reaction start), a 20 l sample was taken, the RNA purified and its absorbance at 260 nm determined as described in Example 1, Section 2. A second sample of 40 l was taken at the same time point and was filtered through a Microcon YM10 device (Merck Millipore, Darmstadt, Germany) (16000*g, 5 minutes, 17 C.). The absorbance of the flow-through at 260 nm, corresponding to unincorporated cap analog and NTPs, was determined using a NanoDrop Spectrophotometer according to the instructions of the manufacturer (T009-Technical Bulletin NanoDrop 1000% 8000; Thermo Fisher Scientific, Wilmington, Del., USA).

Results

[0253] As can be seen from FIG. 12, the use of a sequence-optimized ribonucleotide mix allows measuring the progress of the in vitro transcription reaction by determining the remaining total nucleotide concentration at defined time points. The decrease in total NTP concentration directly correlates with the amount of synthesized RNA.

[0254] Thus, the progress of the transcription reaction can be accurately determined as a function of measured total NTP concentration at a given time point and calculating the moles of NTPs consumed. Based on this information it becomes possible to calculate the amount of synthesized RNA.

[0255] This procedure is especially useful to continually monitor the progress of a transcription reaction, for example in a transcription reactor. This would not be possible when a standard NTP mix is used because the consumption of NTPs would not as easily reflect the amount of synthesized RNA.

Example 7: RNA Yields for Sequence-Optimized Nucleotide Mixes as a Function of Cap Concentration

[0256] Transcription reactions were assembled as described in Example 1, section 2, and were carried out at total NTP concentrations of 2 mM, 4 mM, and 13.45 mM NTPs as indicated in FIGS. 14 and 15. The NTPs were distributed according to the sequence of the produced RNA as described in Example 1, section 5 (sequence-optimized ribonucleotide mix for PpLuc and HsPSCA). The reactions were performed at various concentrations (0, 0.25, 2.0, 10, 16 and 20 mM) of the CAP analog (m7G(5)ppp(5)G) as indicated in FIGS. 14 and 15.

Results

[0257] As can be seen from FIG. 13A, the actual RNA yield for PpLuc mRNA increases with higher cap analog concentrations. The actual RNA yield is higher for the total NTP concentration of 13.45 mM compared to 4 mM. FIG. 13B shows that the relative RNA yield for PpLuc mRNA increases up to a cap analog concentration of approximately 16 mM. The increase in relative RNA yield is stronger for the low NTP concentration (4 mM) than for the high NTP concentration (13.45 mM).

[0258] As can be seen from FIG. 14A, the actual RNA yield for HsPSCA mRNA increases with higher Cap analog concentrations. The actual RNA yield is higher for the total NTP concentration of 13.45 mM compared to 4 mM and 2 mM. FIG. 14B shows that the relative RNA yield for HsPSCA mRNA increases up to a cap analog concentration of approximately 16 mM. The strongest increase of the relative RNA yield is observed for the lowest NTP concentration tested (2 mM).

[0259] These results demonstrate that the use of a sequence-optimized ribonucleotide mix leads to an increased efficiency of capped RNA synthesis even at low initial total nucleotide concentrations (e.g. at 2 mM). By contrast, it has previous been suggested that for an increased RNA yield high concentrations of total nucleotides, in the order of 12 mM to 40 mM, are necessary (U.S. Pat. No. 6,586,218).

[0260] Comparison of PpLuc mRNA (1870 nucleotides) and HsPSCA mRNA (589 nucleotides) shows that the relative RNA yields are independent of the RNA lengths for a defined total NTP concentration.

Example 8: RNA Yields for Sequence-Optimized Nucleotide Mixes as a Function of GTP Start Nucleotide Concentration

[0261] Transcription reactions were assembled as described in Example 1, section 2, and were carried out at a total NTP concentration of the sequence-optimized nucleotide mix of 13.45 mM for P625, P1040 and P532.

[0262] The NTPs were distributed according to the sequence of the produced RNA as described in Example 1, section 5 (sequence-optimized ribonucleotide mix for PpLuc, HsPSCA and EGFR/Mucin-1). The reactions were performed by adding defined concentrations (0, 0.25, 2.0, 10, 16 and 20 mM) of GTP start nucleotide to the sequence-optimized NTP mix as indicated in FIGS. 16 to 18.

Results

[0263] As can be seen from FIGS. 15A and 15B, the actual and relative RNA yield for HsPSCA mRNA increases up to a GTP start nucleotide concentration of approximately 10 mM and declines at higher GTP concentrations.

[0264] As can be seen from FIGS. 16A and 16B, the actual and relative RNA yield for PpLuc mRNA slightly increases up to a GTP start nucleotide concentration of approximately 10 mM and then declines at higher GTP concentrations.

[0265] As can be seen from FIGS. 17A and 17B, the actual and relative RNA yield for EGFR/Mucin-mRNA increases up to a GTP start nucleotide concentration of approximately 10 mM and declines at higher GTP concentrations.

[0266] These results demonstrate that the use of a sequence-optimized ribonucleotide mix and an additional amount of the start nucleotide GTP leads to an increased efficiency of RNA synthesis up to a GTP start nucleotide concentration of approximately 10 mM.

Example 9: Bioreactor

[0267] FIG. 18 shows a preferred embodiment of a bioreactor 1 in accordance with the present invention in a schematic illustration. From FIG. 18, the modular structure of the bioreactor 1 becomes evident. Here, the bioreactor 1 consists of several bioreactor modules 2, 3, 4, 5. Reaction module 1 is a reaction vessel used for a continuous or semi-batch process for synthesizing RNA molecules of a given sequence. The reaction module 2 contains resin-immobilized DNA used as a template for the RNA transcription reaction. Here, the immobilization of the DNA allows a repeated usage of the template and reduces the contamination of the desired RNA product by any kind of residual DNA. In addition, the immobilization of the DNA template supersedes the use of the enzyme DNAse for terminal DNA digestion. After transcription, the produced RNA molecules can be released batch by batch or continuously into the capture module 3. The capture module 3 contains a resin/solid phase to capture the RNA molecules and to separate the RNA from other soluble components of the transcription reaction. Thereafter, the RNA molecules can be dispensed from the bioreactor 1 by means of an exit line or the like (not shown) to a receiving unit or the like, in which further RNA elution and purification can be carried out. A washing fluid and/or elution buffer can be provided to the capture module 3 by means of a respective wash and buffer tank 31 connected to the transfer area between reaction module 2 and capture module 3.

[0268] In order to be able to monitor and control the transcription process in the reaction module 2, an ultrafiltration membrane 21 for separation of high molecular weight components, such as proteins and polynucleotides, from low molecular weight components, such as nucleotides, is provided in the reaction module 2. The membrane separates a reaction core 22, in which the RNA transcription reaction is carried out, from a filtration compartment 23, in which the filtered reaction mix is received. Based on the nucleotide concentration in the filtrated reaction mix in filtration compartment 23 of the reaction module 2, used as critical process parameter, the feed of nucleotides, buffer components and/or enzymes into reaction module 2 from a feed tank 24 can be controlled and regulated by means of a feed pump 43, which allows performing the RNA transcription reaction in an optimal steady-state condition yielding high transcriptional performance. As a measuring means, a sensor unit 41 is provided for measuring reaction parameters in the reaction mix. Here, the sensor unit 41 at least comprises a sensor for photometric analysis, such as an UV flow cell for UV 260/280 nm, in the filtrated fluid containing the low molecular weight components, which filtrated fluid is extracted from the filtration compartment 23, circulated by a recirculation pump 25 and returned into the filtration compartment 23. In the circulation line, the sensor of the sensor unit 41 is provided in order to achieve real-time monitoring of the filtrated fluid inside the filtration compartment 23. The application of a sequence-optimized ribonucleotide mix in the bioreactor 1 enables a real-time measurement of the nucleotide concentration in the filtration compartment 23 during the RNA transcription reaction in the reaction core 22 of reaction module 2. The sensor unit 41 is part of control module 4, which further comprises a controller 42 and an actuator in the form of feed pump 43. The sensor unit 41 and the feed pump 43 are connected to the controller 42 in order to provide measurement signals to and receive instruction signals from the controller 42. Furthermore, other critical process parameters, such as a pH-value of the filtrated fluid, or a conductivity of the filtrated fluid can be analyzed by further suitable sensors of the sensor unit 41. Data collection and analyses by the controller 42, usually in the form of a computer based system or the like, allows the control of the feed pump 43 as an actuator for repeated feeds of nucleotides, buffer components and/or enzymes into the reaction module 2, as well as the control of further pumps in the bioreactor 1 in order to adjust key process parameters in optimal steady-state reaction conditions.

[0269] In order to prevent waste, the bioreactor 1 of the preferred embodiment further comprises a reflux module 5 connected to the capture module 3, which reflux module 5 collects unused raw materials, such as nucleotides and enzymes, and recirculates the same back into the reaction module 2 by means of a reflux pump 51. The reflux module 5 contains immobilized enzymes, such as pyrophosphatase, or resin to capture disruptive components, such as phosphate or the like.

[0270] The above described embodiments of the present invention and the accompanying drawings are merely intended to be illustrative and should not be considered as limiting, since modifications of the described invention can be made within the scope of the accompanying claims without departing from the scope of the same.

Example 10: Immunostimulatory Activity of RNA Molecules

[0271] In this example the immunostimulatory properties of RNA molecules synthesized with a sequence-optimized NTP mix and a standard equimolar NTP mix were compared. Immunostimulation was determined by measuring cytokine and chemokine levels in the supernatants of cells transfected with mRNA.

[0272] Standard and sequence-optimized in vitro transcription reactions for Luciferase mRNA (pPluc) were performed as described in Example 1.

[0273] Subsequently the mRNA was purified by LiCl precipitation.

Immunostimulation Assay

[0274] HeLa cells were seeded at a density of 410.sup.5 cells per well in a 6-well plate in 2 ml HeLa cell culture medium consisting of Gibco RPMI 1640 medium supplemented with 25 mM HEPES, 2 mM L-Glutamine and 100 IU/ml penicillin/streptomycin (all Lonza, Basel, Switzerland) and 10% fetal calf serum (Perbio Science, Bonn, Germany). On the next day the cells were transfected with 2 g of RNA or water-for-injection (WFI) as negative control using Lipofectamine 2000 (Life Technologies, Darmstadt, Germany, catalog no. 11668-027). Briefly, Lipofectamine reagent and RNA were each diluted in Opti-MEM medium (Life Technologies), combined in a ratio of RNA: Lipofectamine of 1:1.5 and incubated for 20 minutes at room temperature. The negative control contained WFI instead of RNA mixed with Lipofectamine. In the meantime the cells were washed once with 2 ml Gibco RPMI 1640 medium supplemented with 25 mM HEPES and 2 mM L-Glutamine (serum free and penicillin/streptomycin free medium) and 2 ml of the serum free and penicillin/streptomycin free medium was added to the cells followed by the addition of 0.5 ml RNA: Lipofectamine transfection mix. After incubation for 4 hours at 37 C. and 5% CO2, the medium containing the transfection mix was replaced by 2 m1 of HeLa cell culture medium.

[0275] After 24 hours, cell-free supernatants were collected and the concentrations of IL-6, CXCL10 and CCL5 were measured by Cytometric Bead Array (CBA) according to the manufacturer's instructions (BD Biosciences) using the following kits: Human Soluble Protein Master Buffer Kit (catalog no. 558264), Assay Diluent (catalog no. 560104), Human IL-6 Flex Set (catalog no. 558276), Human CXCL10 Flex Set (catalog no. 558280) and Human CCL5 Flex Set (catalog no. 558324) (all kits from BD Biosciences). The data was analyzed using the FCAP Array v3.0 software (BD Biosciences).

Results

[0276] As can be seen from FIG. 19, the levels of secreted IL-6, CXCL10 and CCL5 were lower for the RNA synthesized with the sequence-optimized NTP mix compared to the same RNA synthesized with a standard equimolar NTP mix indicating a lower immunostimulatory activity of the RNA resulting from the sequence-optimized NTP mix.

Example 11: In Vitro Transcription in a Bioreactor

Preparation of the DNA Used for In Vitro Transcription (P1140):

[0277] A DNA vector for in vitro transcription (P1140) was prepared by insertion of the following elements into a DNA vector (pCV32(KanR)):

5 UTR: 32L4 (Top-UTR)

[0278] ORF: HA from H1N1(Netherlands2009) (GC-enriched)

3 UTR: Albumin?

[0279] In vitro transcription of the obtained DNA vector results in an RNA molecule having a length of 2083 nt. The respective RNA sequence (SEQ ID NO: 6) is illustrated in FIG. 20.

[0280] The RNA construct is characterized by the following nucleotide composition:

G=540 (25,92%)

C=676 (32,45%)

A=541 (25,97%)

U=326 (15,65%)

G/C=58,37%

Linearization of the DNA Vector:

[0281] The plasmid P1140 was linearized using the following conditions:

0.5 g plasmid DNA
1.5 l 10reaction buffer

1 l EcoRI

[0282] ad 15 l WFI (water for injection)

[0283] The reaction was incubated for 3 h at 37 C. Subsequently a phenol/chloroform extraction and an isopropanol precipitation were performed.

In Vitro Transcription:

Standard Cap/NTP Mix

[0284]

TABLE-US-00014 Final concentration Standard Cap/NTP-Mix 4 L [mM] Cap (100 mM) 1.16 5.8 ATP (100 mM) 0.8 4 CTP (100 mM) 0.8 4 UTP (100 mM) 0.8 4 GTP (100 mM) 0.29 1.45 WFI 0.15 (Final NTP concentration without Cap is 13.45 mM)

Calculation of NTPs and Cap:

[0285] The same total NTP concentration of 13.45 mM as used in the standard transcription reaction is used for sequence-optimized transcription. The fourfold amount of GTP is used for the Cap analog.

TABLE-US-00015 P1140 G C A U Cap total 2083 nt 540 676 541 326 2083 % 25.9 32.5 26.0 15.7 100 mM each 3.5 4.4 3.5 2.1 13.45 (total 13.45 mM NTPs) Cap analog (4x GTP) 13.9 13.9 total Cap/NTP conc. [mM] 27.4

Preparation of the Sequence-Optimized Cap/NTP Mix for P1140:

[0286]

TABLE-US-00016 Final H.sub.2O volume P1140 G C A U Cap (ad 7 l) [l] per reaction 0.70 0.87 0.70 0.42 2.79 1.52 7.00 (l 100 mM NTP)

5 Transcription Buffer:

400 mM HEPES

120 mM MgCl.SUB.2

[0287] 10 mM spermidine

200 mM DTT

[0288] 25 U/ml inorganic pyrophosphatase
4 Different Transcription Reactions were Tested in a Bioreactor:

[0289] As bioreactor, a DasBox Bioreaktor from Dasgip was used. The reaction was stirred at 50 rpm. At the indicated time points, samples of 20 l each were removed. The RNA concentration was measured by determining the absorption at 260 nm after LiCl precipitation.

[0290] Four different conditions were used for in vitro transcription:

[0291] 1. Transcription Using a Standard NTP Mix

TABLE-US-00017 Reagent ad 80000 L Linearized plasmid DNA (P1140) 8300 [0.48 g/L] (L) 5 x transcription buffer (L) 16000 standard Cap/NTP-Mix (L) 16000 RNAse inhibitor [40 U/L] (L) 400 T7 RNA Polymerase [200 U/L] (L) 2000 WFI (L) 37300 Final volume 80000

[0292] The transcription reaction was incubated for 3h at 37 C.

[0293] Subsequently, 6 l DNAse I (1 mg/ml) and 0.2 l CaCl.sub.2) solution (0.1 M)/g DNA template were added to the transcription reaction, and incubated for 2 h at 37 C.

[0294] 2. Sequence-Optimized Transcription (1.5 h without Feed)

TABLE-US-00018 Reagent ad 80000 Linearized plasmid DNA (P1140) 8300 [0.48 g/L] (l) 5 x transcription buffer (l) 16000 Sequence-optimized Cap/NTP- 28000 Mix (l) RNAse inhibitor [40 U/l] (l) 400 T7 RNA Polymerase [200 U/l] (l) 2000 WFI (l) 25300 Final volume 80000

[0295] The transcription reaction was incubated for 1.5 h at 37 C.

[0296] Subsequently, 6 l DNAse I (1 mg/ml) and 0.2 l CaCl.sub.2) solution (0.1 M)/g DNA template were added to the transcription reaction, and incubated for 2 h at 37 C.

[0297] 3. Sequence-Optimized Transcription with Feed

TABLE-US-00019 Reagent ad 80000 l Linearized plasmid DNA (P1140) 8300 [0.48 g/l] (l) 5 x transcription buffer (l) 16000 Sequence-optimized Cap/NTP- 28000 Mix (l) RNAse inhibitor [40 U/l] (l) 400 T7 RNA Polymerase [200 U/l] (l) 2000 WFI (l) 25300 Final volume 80000

[0298] The transcription reaction was incubated for 1.5 h at 37 C.

[0299] 12934.6 l sequence-optimized Cap/NTP-Mix and 5transcription buffer were added after 1.5 h. The transcription reaction was incubated for additional 1.5 h at 37 C.

[0300] Subsequently, 6 l DNAse I (1 mg/ml) and 0.2 l CaCl2 solution (0.1 M)/g DNA template were added to the transcription reaction, and incubated for 2 h at 37 C.

[0301] 4. Sequence-Optimized Transcription with Reduced T7 RNA Polymerase Concentration and Reduced Template Concentration

TABLE-US-00020 Reagent ad 80000 l Linearized plasmid DNA (P1140) 4200 [0.48 g/l] (l) 5 x transcription buffer (l) 16000 Sequence-optimized Cap/NTP- 28000 Mix (l) RNAse inhibitor [40 U/l] (l) 400 T7 RNA Polymerase [200 U/l] (l) 1000 WFI (l) 30400 Final volume 80000

Results:

[0302] Transcription in a sequence-optimized transcription mix results in higher concentrations of transcribed RNA compared to transcription under standard conditions (FIG. 21, TS(1)). An a addition feed with nucleotides and transcription buffer further increased the amount of transcribed RNA (FIG. 21, TS(3).

Yield:

[0303]

TABLE-US-00021 [RNA] Sample ID (mg) P1140-TS(1) 130.6 P1140-TS(2) 317.1 P1140-TS(3) 656.4 P1140-TS(4) 312.6

Expression and Immunostimulation:

[0304] HeLa cells were seeded at a density of 410.sup.5 per well in a 6-well plate in 2 ml HeLa cell culture medium consisting of Gibco RPMI 1640 medium supplemented with 25 mM HEPES, 2 mM L-Glutamine and 100 IU/ml penicillin/streptomycin (all Lonza, Basel, Switzerland) and 10% fetal calf serum (Perbio Science, Bonn, Germany). On the next day, the cells were transfected with different concentrations of 2 g RNA or water-for-injection (WFI) as negative control using Lipofectamine 2000 (Life Technologies, Darmstadt, Germany, catalog no. 11668-027). Briefly, Lipofectamine reagent and RNA were each diluted in Opti-MEM medium (Life Technologies), combined in a ratio of RNA: Lipofectamine of 1:1.5 and incubated for 20 min at room temperature. Negative control contained WFI instead of RNA mixed with Lipofectamine2000. In the meantime, the cells were washed once with 2 ml Gibco RPMI 1640 medium supplemented with 25 mM HEPES and 2 mM L-Glutamine (serum- and penicillin/streptomycin free medium), 2 ml of the serum- and penicillin/streptomycin-free medium was added to the cells following by the addition of 0.5 ml RNA: Lipofectamine transfection mix. Upon incubation for 4 h at 37 C. and 5% CO.sub.2, the medium containing the transfection mix was removed and 2 ml of the HeLa cell culture medium were added.

[0305] After 24 hours, supernatants and cells were collected.

Protein Expression:

[0306] Surface expression of the HA protein was determined using flow cytometric analysis. Adherent HeLa cells were washed once with lml PBS and harvested using trypsin-free detach buffer (40 mM Tris HCl pH 7,5; 150 mM NaCl, 1 mM EDTA). The cells were incubated with mouse monoclonal anti-HA (H1N1) antibody (Immune Technology, New York, USA) followed by a secondary anti-mouse FITC-conjugated antibody (Sigma-Aldrich, Taufkirchen, Germany). The cells were measured on a BD FACS Canto and analyzed using FlowJo Software Version 10.6. Statistical analysis was performed using Graph Pad Prism Software, Version 5.01.

Results:

[0307] RNA transcribed in a sequence-optimized reaction mix (FIG. 22, TS(2), TS(3), TS(4)) resulted in a higher expression of the encoded HA protein than RNA transcribed under standard conditions (FIG. 22, TS(1)).

Immunostimulation:

[0308] The concentrations of IL-6, CXCL10 and CCL5 were measured in cell-free supernatants by cytometric bead array (CBA) according to the manufacturer's instructions (BD Biosciences) using the following kits:

TABLE-US-00022 reagent catalog no. Human Soluble Protein Master Buffer 558264 Kit Assay Diluent 560104 Human IL-6 Flex Set 558276 Human CXCL10 Flex Set 558280 Human CCL5 Flex Set 558324

[0309] The data was analyzed using the FCAP Array v3.0 software (BD Biosciences). Statistical analysis was performed using Graph Pad Prism Software, Version 5.01.

Results:

[0310] RNA transcribed under standard conditions (FIG. 23, TS1) induced higher levels of the cytokines IL-6, CXCL10 and CCL5 in Hela cells compared to RNAs transcribed in a sequence-optimized reaction mix (FIG. 23, TS2, TS3, TS4).