RNA molecules with modified 3' terminal sequences and application thereof

Abstract

The subject of the invention is modified 3' terminal sequences of the mRNA 3' end, DNA sequence encoding it, and their applications as an improvement of the mRNA molecular tool for a wide range of medical therapies based on mRNA molecules.

Description

[0001] RNA molecules with modified 3’ terminal sequences and application thereof

[0002] Field of the invention

[0003] The presented invention relates to RNA molecules, methods of producing RNA molecules, methods of producing a protein or peptide in vitro, in cells, or living organisms, and methods of treating or preventing a disease.

[0004] STATE OF THE ART

[0005] The efficacy of utilizing RNA (ribonucleic acid) molecules for curative purposes hinges upon the structural characteristics of the RNA itself, which impact both its stability and the metabolic processes governing RNA processing within living cells. A typical mRNA (messenger RNA) molecule comprises a gene-of-interest encoding sequence, flanked by untranslated regions (UTRs) that regulate both transcription and translation. Additional structures at both ends serve primarily to stabilize and shield the mRNA from degradation. The 5'-end features the m7G cap structure, while the 3'-end terminates with a poly(A) tail. This poly(A) tail, found at the 3' end of most eukaryotic mRNA molecules, is composed of a series of adenine (A) nucleotides connected by 3',5'-phosphodiester bonds (Hocine et al. 2010). While the precise length of the poly(A) tail may vary, it typically consists of 100 to 250 adenine residues. This repetitive adenine sequence lacks coding information but plays a vital role in several cellular functions. It serves to regulate mRNA stability by safeguarding the mRNA from enzymatic degradation, thereby preserving the genetic information it carries (Wang and Kiledjian 2000). Additionally, the poly(A) tail plays a pivotal role in mediating the export of mRNA from the cell nucleus to the cytoplasm and is involved in the initiation of translation. It facilitates ribosome binding for protein synthesis (Passmore and Coller 2022). During translation initiation, the small ribosomal subunit, along with various initiation factors, binds to the mRNA. The presence of a poly(A) tail enhances this binding process, promoting efficient ribosome recruitment and, consequently, increased protein production (Bradrick et al. 2007).

[0006] The length of the poly(A) tail can be dynamically regulated, and this regulation significantly impacts mRNA stability and turnover. In some instances, shortening of the poly(A) tail can target the mRNA for degradation, limiting its lifespan and reducing protein production (Jalkanen et al. 2014). Conversely, elongating the poly(A) tail can stabilize the mRNA, prolonging its presence in the cytoplasm and promoting sustained protein synthesis. This dynamic control of the poly(A) tail allows cells to finely tune gene expression in response to various internal and external signals (Zarnack et al. 2020). The poly(A) tail facilitates specific interactions between mRNA molecules and proteins by engaging with poly(A) binding proteins (PABP) (Idler and Yan 2012). PABP is a highly conserved, 636-residue protein, featuring four tandem RNA recognition motifs (RRMs), an unstructured linker region, and a C-terminal PABC domain (Kozlov et al. 2001 ; Melo et al. 2003). All four RRMs directly participate in binding to the adenine nucleotides of the poly(A) tail with a dissociation constant (KD) of approximately 4 nM (Baer and Kornberg 1983; Kuhn and Pieler 1996). The linker and C-terminal domain, on the other hand, play a crucial role in the intermolecular interactions between multimerized PABP molecules bound to mRNA (Sawazaki et al. 2018). Studies have shown that a single PABP protein requires about 27 adenine bases for binding to the poly(A) tail (Baer and Kornberg 1980). This high affinity of PABP for the poly(A) tail structure is vital for the translation process. During translation initiation, PABP simultaneously binds to the poly(A) tail and interacts with other factors, including the eukaryotic translation initiation factor 4G (elF4G) and the poly(A) binding protein-interacting protein 1 (PAIP1 ) (Martineau et al. 2008). These interactions promote the circularization of the mRNA, bringing the 5' and 3' ends of the mRNA molecule into proximity. This circularized mRNA structure is believed to enhance translation efficiency. PABP's interaction with elF4G and other translation initiation factors fosters the assembly of the translation initiation complex, which includes the small ribosomal subunit (40S), initiator tRNA, and various initiation factors. PABP aids in recruiting the 40S subunit to the mRNA 5' terminus, ensuring that translation initiates at the correct site. PABP's role in translation initiation extends to facilitating the recruitment of the large ribosomal subunit (60S) to the mRNA (Jackson et al. 2010), allowing the ribosome to scan along the mRNA until it encounters the start codon (AUG). PABP helps position the ribosome correctly for start codon recognition. By promoting mRNA circularization, stabilization, and aiding in translation initiation, PABP significantly contributes to the overall efficiency and accuracy of the translation process.

[0007] Furthermore, PABP proteins, upon binding to poly(A), coordinate the process of deadenylation of the poly(A) tail, exhibiting both protective and degradation-promoting functions, which are influenced by numerous factors (Wigington et al. 2014). In the majority of cases, PABPC protects mRNA molecules against degradation by temporarily inhibiting the activity of the PAN2-PAN3 (PAN2 / 3) and CCR4-NOT (CNOT) complexes (Deo et al. 1999). The PAN2 / 3 complex selectively trims long tails (>150A), and then CAF1 and CCR4 (which serve as catalytic subunits of CNOT) continue to cooperatively degrade the poly(A) until a short fragment of approximately 25 nucleotides remains (Schwede et al. 2009). As reported in a recent study, CAF1 cleaves naked poly(A) fragments, and its activity is blocked when it encounters PABP protein bound to a poly(A) strand. The CCR4 enzyme can displace the PABP protein from the poly(A) tail and continue deadenylating, resulting in a shorter poly(A) tail and, consequently, a shorter mRNA half-life (Webster et al. 2018). The cooperative actions of CAF1 and CCR4 enzymes can be observed through ~27 nucleotide periodic PABP footprints during poly(A) tail shortening (Nicholson and Pasquinelli 2019). Overall, various studies support the prevailing consensus that longer poly(A) tails stabilize mRNA and enhance its translational activity. Therefore, the presence of extended poly(A) sequences is also desirable in laboratory-produced mRNAs designed for gene delivery applications.

[0008] Laboratory and industrial methods for RNA synthesis, characterized by high yields and product quality, are primarily based on In Vitro transcription (IVT) reactions. The successful execution of an IVT reaction necessitates the presence of several key components: an appropriate DNA template determining the RNA sequence, enzymes that play a crucial role in the reaction's performance, such as polymerase and pyrophosphatase, and small molecules essential for constructing the growing nucleotide chain, specifically NTPs and a 5' cap analog. DNA templates are typically added in quantities of tens of nanograms per microliter to the IVT reaction and can be prepared and amplified either through Polymerase Chain Reaction (PCR) or by amplifying a plasmid vector within a bacterial system. The use of plasmid DNA templates is recommended due to the greater control and reproducibility it offers in the amplification process, which is particularly critical when scaling up production under good manufacturing practices (GMP). Plasmid vectors are circular DNA molecules designed to amplify in bacterial systems, primarily based on Escherichia coli (E coli). These plasmids comprise essential sequence elements for bacterial amplification, including the origin of replication (ori) and antibiotic resistance genes. They also contain the entire gene coding sequence (CDS), flanked by 5' and 3' untranslated regions, and elements necessary to initiate efficient transcription, including a promoter.

[0009] However, amplifying plasmid DNA templates in bacterial systems despite the advantages, such as high controllability of the process and lower cost of large-scale production compared to PCR has its drawbacks. The bacterial transcription machinery in E. coli tends to inaccurately amplify "challenging" sequences, such as those containing repeat elements and unstable inserts. Amplifying long homosequences, exceeding 80-100 nucleotides, like the poly(A) tail sequence, is particularly problematic. This identical base sequence causes bacterial polymerase to glide through the sequence without accurately replicating the full length of the DNA fragment, resulting in the subsequent shortening of the poly(A) tail sequence during amplification, reducing it to about 30 adenine bases. This reduction in the length of the poly(A) tail increases the heterogeneity of mRNA and decreases its relative biological activity. mRNAs with shortened poly(A) tails are more susceptible to degradation and have shorter half-lives in living cells. A shortened poly(A) tail can also lead to lower levels of encoded protein expression due to a less efficient translation initiation process. Furthermore, the shortening of poly(A) tracts can introduce bias during the plasmid amplification process, where some DNA species may be more prone to tail shortening than others, resulting in an uneven representation of full- length and shortened poly(A) tails in corresponding transcripts. This can lead to misinterpretations of the true biological effects. Obtaining a homogeneous, full-length template containing a poly(A) tail of at least 100 adenines is particularly challenging.

[0010] Previous studies have identified a critical 30-50 base fragment near the 5' end of the poly(A) tail in the amplification process in bacteria. They proposed the use of a single 10-nucleotide linker in the poly(A) sequence as a stabilizing element in the replication process of plasmid DNA templates in bacteria. This approach allowed access to templates encoding a poly(A) tail of up to 110 adenine bases in length without affecting the biological activity of the encoded mRNA (W02016005004A1 ). Other research has shown that introducing nucleotide linkers within the poly(A) sequence stabilizes the DNA constructs during amplification. The 3'-terminal sequence of 120 adenines was punctuated using a mono- or hexa-nucleotide linker to create a segmented (A60Jinker_A60) tail system (Trepotec et al. 2019). A version that split the entire poly(A) tail into three fragments of 40 adenines forming an (A40_linker_A40_linker_A40) arrangement was also tested. These segmented poly(A) tails reduced the occurrence of recombination events in plasmid DNA. In most of the studied systems, including in vivo, the A60Jinker_A60 tail did not negatively impact mRNA translation or stability. Another study investigated the impact of non-adenosine (non-A) nucleotides near the 3' end of mRNA tails on protein production. A series of EGFP mRNAs with 40-nt poly(A) tails containing different single nucleotide substitutions were synthesized, revealing that cytidine (C) substitutions near the end of the tail increased protein production (Li et al. 2022) (WO2022028559A1 ). The effect of C substitution was consistent across various human cell lines, regardless of their basal protein expression levels. The study delved into the mechanisms underlying these effects and identified the CCR4-NOT (CNOT) complex, involved in mRNA degradation, as a key player. Knockdown of specific CNOT proteins reduced the protein production enhancement effect of C-containing tails, indicating their role in extending mRNA half-life. The impact of incorporating these modifications on DNA construct stability has not been investigated. The above- mentioned invention (WO2022028559A1 ) discloses sequences containing between 30 and 150 adenines, with at least one adenine substituted by a cytosine in the last one-third portion closest to the 3' end. The length of the sequence can vary between 18 and 129 adenines, with the last nucleotide typically not being a cytosine and may contain up to 40% cytosines, with a preference for cytosines in the last one-third portion with at least one cytosine substitution in the specified region.

[0011] The presented invention discloses a new solution to the problem of instability of long homosequences in DNA templates encoding mRNA and independently discloses a new way of enhancing translational efficiency of in vitro transcribed mRNAs. We found that poly(A) tail homosequences in mRNA can be replaced with properly designed heterosequences consisting of tracks of 10-35 adenines spaced by mono- or oligonucleotide spacers consisting of other nucleobases. The disclosed 3’ terminal heterosequences do not destabilize DNA constructs encoding mRNA (in contrast to poly(A) homosequences) during the amplification process. Importantly, these heterosequences, despite quite dense placing of the heterobases, do not disturb the translational activity of mRNA, suggesting that the disclosed patterns of modification do not disturb the interaction of the 3’-terminal fragment of mRNA with PABP. In contrast, in some embodiments of the invention, the heterosequences unexpectedly confer increased translational activity to mRNA molecules compared to mRNAs carrying homosequences. The state-of the-art solution to the problem of plasmid DNA instability are practically limited to poly(A) tail analogs reaching 110-120 or up to 150 nt in length. In contrast to the state-of-the-art solutions, the solution proposed here can be applied to construing 3’- terminal sequence virtually unlimited in length. Hence, the mRNA molecules with 3’-terminal sequences modified according to the invention combine two desirable features: (i) they show increased translational activity compared to template-encoded mRNAs containing unmodified 3’-terminal poly(A) tails and (ii) increase relative stability during amplifications of DNA plasmids encoding these molecules, and, additionally, enable preparation of DNA templates encoding mRNAs with 3’-terminal sequences functionally replacing poly(A) tails of lengths exceeding the current limitations (110-150 nt).

[0012] Summary of the invention

[0013] Subject of the invention has been defined in attached claims.

[0014] Unless defined otherwise, all terms used herein have the same meaning as are commonly understood by one of skill in the art to which this invention belongs.

[0015] The disclosed 3’-terminal sequences of mRNA molecules are polymers composed of monomers (ribonucleotides) made of pentose (sugar residue) linked to a nitrogenous base and linked to the sugar residue of an neighboring monomers through a 3',5'-phosphodiester bond. The 3’-terminal sequence comprises, from the 5' end of the consecutive adenine (A) oligonucleotide motif (M) defined as a number (k) in the range (from 10 to 35 A), followed by a heteronucleotide linker (L1) defined in the range from 1 to 6 nucleotides comprising independently natural, modified, or unnatural nucleoside bases, followed by another adenine oligonucleotide motif (M’) defined as a number (k) in the range (from 10 to 35 A) followed by a heteronucleotide linker (L2) defined in the range from 1 to 6 nucleotides representing independently a natural, modified, or unnatural nucleoside base, followed by another adenine oligonucleotide motif (M”) defined as a number (k) in the range (from 10 to 35 A) followed by a heteronucleotide linker (L3) defined in the range from 1 to 6 nucleotides representing independently a natural, modified, or unnatural nucleoside base, followed by n additional similarly designed segments where (n) is a number in the range of 0 to 15 and defines the number of adenine oligonucleotide motifs (Mn) containing k consecutive adenines, wherein k is in the range from 10 to 35), followed by a heteronucleotide linker (Ln) ranging from 1 to 6 nucleotides in length comprising independently natural, modified, or unnatural nucleoside bases (claimed Formula 1 , Formula 2).

[0016] M-L1-M’-L2-M”-L3-(Mn.Ln)n

[0017] (Formula 1)

[0018] To obtain mRNAs with modified 3’-terminal sequences, DNA templates encoding mRNAs were prepared from plasmid DNA. The synthetic double-stranded oligonucleotides (Table 1A and Table 1 B) and DNA templates encoding the Firefly luciferase (FLuc) and mKate2-PEST (mKate2) genes were utilized to demonstrate the invention (Table 2). Both components were cloned using the blunt-end cloning method, ligated, and the resulting ligation mix was used to transform stable chemo-competent bacteria. This transformation yielded single bacterial colonies, which were subsequently used to purify the target plasmids containing inserts that encode the disclosed mRNA molecules, as detailed in Example 1. The accuracy of the DNA sequence was confirmed through Sanger sequencing. The DNA sequences encoding the disclosed mRNA molecules were constructed according to the method described in Example 1. Each 3’-terminal sequence of the mRNA molecule consisted of 10 up to 35 adenine nucleotides, separated by heteronucleotide linkers composed of one to six nucleotides. This arrangement formed segmented 3’-terminal sequences comprising three to eighteen such fragments (as illustrated in Figure 1 ). In one embodiment of the invention (variant A2< Figure 1 ) the 3’-terminal sequence A2comprises a fragment of 30 consecutive adenines, followed by guanine mononucleotide linker (G), 30 consecutive adenines, second guanine mononucleotide linker (G), 30 consecutive adenines, third guanine mononucleotide linker (G) ending with 30 consecutive adenines. In another embodiment of the invention (variant A3< Figure 1 ) the 3’-terminal sequence A3comprises a fragment of 30 consecutive adenines, followed by heteronucleotide linker (GCATAT), 30 consecutive adenines, second heteronucleotide linker (GCATAT), 30 consecutive adenines, third heteronucleotide linker (GCATAT), 30 consecutive adenines, fourth heteronucleotide linker (GCATAT) ending with 30 consecutive adenines. Variant A4is analogous to A3but with one segment of consecutive adenines and one heteronucleotides linker removed (Figurel ). In another embodiment of the invention (variant A6< Figure 1 ) the 3’-terminal sequence A6comprises a fragment of 30 consecutive adenines, followed by cytidine mononucleotide linker (C), 30 consecutive adenines, second cytidine mononucleotide linker (C), 30 consecutive adenines, third cytidine mononucleotide linker (C) ending with 30 consecutive adenines. In another embodiment of the invention (variant A8< Figure 1 ) the 3’-terminal sequence A8comprises a fragment of 30 consecutive adenines, followed by guanine mononucleotide linker (G), 15 consecutive adenines, second guanine mononucleotide linker (G), 15 consecutive adenines, third guanine mononucleotide linker (G), 15 consecutive adenines, fourth guanine mononucleotide linker (G), 15 consecutive adenines, fifth guanine mononucleotide linker (G), 15 consecutive adenines, sixth guanine mononucleotide linker (G) ending with 15 consecutive adenines. Variants A9and A10were created analogous to the above description with a change in the number of consecutive adenines in each motif and the number of guanine mononucleotide linkers (Figurel ). In another embodiment of the invention (variant A11; Figure 1 ) the 3’-terminal sequence A11comprises a fragment of 30 consecutive adenines, followed by cytidine mononucleotide linker (C), 15 consecutive adenines, second cytidine mononucleotide linker (C), 15 consecutive adenines, third cytidine mononucleotide linker (C), 15 consecutive adenines, fourth cytidine mononucleotide linker (C), 15 consecutive adenines, fifth cytidine mononucleotide linker (C), 15 consecutive adenines, sixth cytidine mononucleotide linker (C) ending with 15 consecutive adenines. In another embodiment of the invention (variant A13; Figure 1 ) the 3’-terminal sequence A13comprises a fragment of 10 consecutive adenines, followed by cytidine mononucleotide linker (C), 10 consecutive adenines, second cytidine mononucleotide linker (C), 10 consecutive adenines, third cytidine mononucleotide linker (C), 10 consecutive adenines, fourth cytidine mononucleotide linker (C), 10 consecutive adenines, fifth cytidine mononucleotide linker (C), 10 consecutive adenines, sixth cytidine mononucleotide linker (C), 10 consecutive adenines, seventh cytidine mononucleotide linker (C), 10 consecutive adenines, eighth cytidine mononucleotide linker (C), 10 consecutive adenines, nineth cytidine mononucleotide linker (C), 10 consecutive adenines, tenth cytidine mononucleotide linker (C) ending with 10 consecutive adenines. In another embodiment of the invention (variant A15< Figure 1 ) the 3’-terminal sequence A15comprises a fragment of 15 consecutive adenines, followed by cytidine mononucleotide linker (C), 15 consecutive adenines, second cytidine mononucleotide linker (C), 15 consecutive adenines, third cytidine mononucleotide linker (C), 15 consecutive adenines, fourth cytidine mononucleotide linker (C), 15 consecutive adenines, fifth cytidine mononucleotide linker (C), 15 consecutive adenines, sixth cytidine mononucleotide linker (C), 15 consecutive adenines, seventh cytidine mononucleotide linker (C) ending with 15 consecutive adenines. Variants A14, A16, A17, A18, A19and A20were created analogous to the above description with a change in the number of consecutive adenines in each motif and the number of cytidine mononucleotide linkers (Figure 1 ). In another embodiment of the invention, L, L’, L”, L’”, and each of L1to Lnis independently selected among of nucleotides G or C or oligonucleotides: GCAUAU, GCAUAUGACU, UGGGGUUUGGGGUUUGGGGUUUGGGGU or UCUAG or is abandoned.

[0019] In another embodiment of the invention, the 3’-terminal sequence comprises more than 150 nt (as demonstrated by variants A18-A20; Fig. 1 B). In particular embodiment of the invention, the 3’-terminal sequence has been selected among of Seq Id No: 81-96, preferably among of Seq Id No: 88, 95 and 96.

[0020] In addition to the mRNA molecules designed within the scope of the claimed invention, reference mRNA molecules were also prepared to represent existing state-of-the-art poly(A) tail modifications. These include an mRNA containing a poly(A) tail of 90 adenine nucleotides (R1), which is the longest unmodified poly(A) sequence that could be reproducibly amplified in plasmid DNA in our hands. Another reference represents a modification constructed with a sequence of 30 consecutive adenines followed by a 10 nucleotide linker (GCATATGACT), followed by a sequence of 70 consecutive adenines (R2), which has been applied in Comirnaty Pfizer-BioNTech BNT162b2 anti-COVID mRNA vaccine (Sahin et al. 2021 ). The third reference (R3) is a sequence of 90 consecutive adenines terminated by a pentanucleotide motif (TCTAG), which has been applied in Moderna’s mRNA-1273 anti-COVID mRNA vaccine (Corbett et al. 2020). Two reference modifications containing a single linker in the middle of the poly(A) tail were also prepared. Variant R4(Figure 1 ) consists of 60 consecutive adenines followed by guanine mononucleotide linker (G) followed by 60 consecutive adenines, while Variant R5(Figure 1 ) consists of 60 consecutive adenines followed by cytidine mononucleotide linker (C) followed by 60 consecutive adenines as described in (Trepotec et al. 2019). An additional variant (variant R6<Figure 1 ) represents an alternative modification containing the intra-molecular structure of G-quadruplex. Variant R6comprises a sequence of 30 consecutive adenines, followed by guanine mononucleotide linker, 30 consecutive adenines and G- quadruplex structure with the sequence: TGGGGTTTGGGGTTTGGGGTTTGGGGT linked to 15 consecutive adenines at the 3’ end. The use of such a variant made it possible to compare the biological effect of modified 3’-terminal sequences with a different, higher-order arrangement of the poly(A) tail (Sagi 2014).

[0021] The stability of DNA plasmids encoding mRNAs with the disclosed 3’-terminal sequences and reference poly(A) tails during plasmid amplification was assessed through DNA sequencing of individual clones, as outlined in Example 2. The data revealed a significantly lower percentage of clones that exhibited loss of the full sequence length during amplification for the modified 3'-terminal sequence variants compared to reference variants R4 and R5and a 130 nt poly(A) tail, as depicted in Figure 2. The stability of the modified sequences was similar to other reference sequences.

[0022] Using the corresponding DNA templates, mRNA molecules with the modified 3’- terminal sequences and reference mRNAs encoding Firefly luciferase or mKate2-PEST (as specified in Table 3) were synthesized through In Vitro transcription reactions and purified as detailed in Example 3. The quality of all obtained mRNAs post-purification was found to be satisfactory, as demonstrated in Figure 3.

[0023] These resulting mRNA molecules were employed in functional studies to characterize the differences between the various modified 3’-terminal sequences in living cells, as described in Example 4. The modified 3'-terminal sequence variants exhibited higher levels of protein expression in all tested cell lines compared to most of the reference modifications. The use of consecutive adenine motifs linked by heteronucleotide linkers increased reporter protein expression in comparison to the classical 3'-terminal mRNA sequence (R1) as well as other poly(A) tail modifications proposed in existing literature (R2and R3). The results of in vitro experiments comparing all tested mRNA variants are presented in Figure 4, Figure 5, and Figure 6. Additionally, RNA sequencing (DRS) was performed to confirm the correct sequences of all modified 3’ terminal sequences at the mRNA Fluc / mKate2-PEST molecule level, as outlined in Example 8. A comparison of individual sequences for selected modifications is provided in Figure 7. Qualitative Western Blot analysis of FLuc and mKate2 protein expression in cells was conducted using specific antibodies and immune cells, as described in Example 7. Both Firefly luciferase and mKate2 proteins were specifically detected in all tested modifications, except for the R6variant, which contains a G-quadruplex structure that negatively affects translation. The chemiluminescence signals were at similar levels for the other variants, and the results of this assay for selected 3’-terminal sequences are depicted in Figure 8. Finally, the most promising 3' terminal modified sequences were tested in vivo. First, total body bioluminescence post administration of firefly luciferse (Flue) mRNA formulated using SM-102 was evaluated in time (Figure 9). Next, human erythropoietin (hEPO) concentration was measured in the sera of mice administered with SM-102-formulated hEPO- mRNA (Figure 10).

[0024] DETAILED DESCRIPTION OF THE INVENTION

[0025] We and others have encountered challenges when preparing DNA templates with a long poly(A) tail (> 90 A) from plasmid DNA, which poses a high risk of losing the original length of the 3' terminal poly(A) sequence and increasing the final heterogeneity of the DNA template during plasmid amplification (Trepotec et al. 2019 and Sahin et al. 2021 ). This issue is attributed to the characteristics of bacterial systems used for plasmid DNA vector amplification, particularly bacterial polymerases that tend to skip sequence fragments containing repeat elements during replication. Consequently, this often leads to the generation of modified copies of plasmids with shortened poly(A) tails, reducing their original length to approximately 30 adenine nucleotides. The mRNA molecules synthesized from such modified DNA templates with short poly(A) tails have shorter half-lives, are degraded more rapidly, and exhibit significantly lower levels of protein translation in living cells. This outcome is undesirable, particularly in therapeutic applications. In our experiments, we found that a 90 adenine tail was the maximum length of a poly(A) sequence that could be maintained during the amplification process in bacteria, therefore mRNA comprising a 90 nt long poly(A) tail was used as reference R1. To prepare the construct with poly(A) A90 tail (R1), reference tails (R2-R5), and all constructs with modified 3’-terminal sequences disclosed here (Figure 1 ), the blunt end cloning method was employed for the plasmid vector encoding Firefly luciferase and mKate2 proteins as described in Example 1 using annealed oligonucleotides listed in Table 1A and Table 1 B. The applied method was characterized by low efficiency of the cloning process associated with the formation of inversely embedded poly(T) tails, which were eliminated on the basis of DNA sequencing.

[0026] The stability of the modified 3’-terminal sequences at the DNA level (plasmid vectors) was assessed in two different bacterial systems, as detailed in Example 2. The results were presented as the percentage of colonies that did not contain the original sequence encoding the 3’-terminal fragment of mRNA, serving as a measure of instability, and are illustrated in Figure 2. As expected, the plasmid with a poly(A) tail length close to 130 adenines showed the largest number of colonies characterized by altered length or sequence relative to the original one, making it impossible to obtain a stable construct of this length. In contrast, the classic poly(A) tail represented by reference R1, as well as references R2and R3, exhibited very similar levels of stability, with the correct sequence of the poly(A) tail maintained in a significantly larger proportion of clones compared to the plasmid containing 130 adenines in the poly(A) tail. On the other hand, references R4and R5, which represented poly(A) tails containing a single heteronucleotide in the middle of the 3' terminal sequence, displayed increased instability compared to references R1-R3. The single heteronucleotide inserted after 60 consecutive adenines in the poly(A) tail, as found in the R4and R5variants, reduced the stability of the construct, indicated by the number of clones with altered / shortened poly(A) sequences (Figure 2). Amplification of plasmid DNA comprising sequences encoding mRNA with modified 3’-terminal sequences according to this invention (A2, A3, A4, A6, A8, A9, A10, A11, A13, A14, A15, A16, A17, A18, A19and A20) occurred with relatively high fidelity, indicating high plasmid stability, which was notably higher than references R4, R5and comparable to references R1, R2and R3(Figure 2). The problem of plasmid instability was still observed to some extent in the standard E.coli Top10 strain (Figure 2A). However, the use of commercially available NEB® Stable Competent E. coli (High Efficiency) bacteria, led to almost complete elimination of the phenomenon of losing the full length of the poly(A) tail during DNA amplification in bacteria for most variants (Figure 2B).

[0027] Following the amplification of plasmids, they were subsequently purified and linearized, and mRNA molecules encoding FLuc / mKate2 were synthesized through In Vitro transcription (IVT), as detailed in Example 3. The sequences of the modified 3’ terminal sequences and the other modifications did not affect the efficiency of the IVT transcription reaction; similar amounts of crude mRNA were obtained for all tested variants. Similarly, the modifications did not affect the RNA purification process using oligo(dt)25 resin and HPLC as described in Example 3. This process produced high-quality mRNA preparations (as illustrated in Figure 3) characterized by high purity and devoid of double-stranded RNA (dsRNA) molecules. This was confirmed through chromatograms generated during HPLC purification, agarose gel analysis, and dsRNA-dot blot analysis for FLuc mRNAs with modified 3’-terminal sequences and the other selected modifications.

[0028] The Flue mRNA molecules containing the modified 3'-terminal sequences as well as all reference variants were subjected to a comprehensive range of biological effect studies using an in vitro cell model. These experiments were conducted in four distinct cell lines (A549, HEK293T, HepG2, and Jaws II), encompassing both mouse and human cells, cancer cells, and immune cells. The experimental protocol involved introducing mRNA molecules into the cells, followed by measuring chemiluminescence of the Flue reporter protein at 4 h, 16 h, 24 h and 48 h, as detailed in Example 4. Measurements of protein expression kinetics, as well as the total amount of protein expressed for each mRNA are shown in Figure 4, Figure 5 and Figure 6. In general, the dynamics of protein production were consistent across all cell lines, with the highest levels of reporter protein typically observed at 16 and 24 hours posttransfection. However, it's worth noting that JAWSII cells exhibited different kinetics, with the highest readings consistently observed at the earliest time point, i.e., 4 hours after transfection. The mRNA variants featuring the modified 3'-terminal sequences according to this invention consistently exhibited significantly higher levels of protein expression when compared to the reference variant R1across all tested cell lines and time points. A similar effect was observed for the reference R2, wherein most of the modified poly(A) tails displayed a notable advantage in terms of their biological effects. Furthermore, the reference mRNAs containing a single heteronucleotide linker after 60 consecutive adenines (designated as R4and R5) demonstrated higher protein expression levels when compared to references R1and R2. The only exception was the R6variant, which contained a G-quadruplex structure that likely had a negative impact on translation efficiency. Comparing references R4and R5with tested variants according to this invention, in most cases modified 3' terminal sequence yielded a significantly higher biological effect, however, there have been exceptions. In particular, in some of the cell lines the disclosed mRNAs showed expression levels compared to R5. Interestingly, a significant advantage of RNA with modified 3’-terminal sequences according to this invention compared to R1, R2, R4and R5in expression levels became apparent at later time points after transfection (Figure 4). The R1and R2variants at 16, 24 and 48 hours after transfection showed significantly lower levels of Flue protein compared to the vast majority of modified 3' terminal sequences. The mRNAs with modified 3' terminal sequences allowed for stable protein production over extended periods, with a gradual decline in production at later measurement points. Notably, at later time points, the protein production for references R4and R5was lower compared to the best-performing modified 3' terminal sequence variants (A3, A4, A6, A8, A9, A10, A11). This data indicates the superiority of the disclosed invention over state-of-the-art solutions in terms of the prolonged time of protein production (Figure 4). The comparison of the total amount of protein produced from different mRNA variants in each cell line also underscores the advantages of the modified 3'-terminal sequences. The analysis of total protein expression revealed no significant differences in the amount of reporter protein produced for R1and R2, whereas R4yielded 1 .5- to 2-fold higher total Flue protein levels compared to R1and R2. R5displayed higher values than R4, as previously shown in a separate study (Li et al. 2022). The modified 3'-terminal variants exhibited 2-8 fold higher total protein amounts compared to R1and R2. The degree of advantage in biological effect varied from twofold to eightfold that of the classical poly(A) A90 tail (R1). The variants displaying the most favourable biological properties were modifications A3, A6, A8, A9, and A11, which had protein expression values exceeding the most efficient reference, R5(Figure 5, Figure 6). Statistical analysis of the performed experiments highlighted the significance of the observed differences in relation to the R1reference. Variants A13, A14, A15, A16and A17characterized by an identical number of adenines in each motif (without the key fragment of 30 consecutive adenines from the 5' end of the poly(A) tail) yielded protein expression levels similar or slightly higher than the classical Ago tail (Figure 6). Compared to the most similar variants having a key motif of 30 consecutive adenines (A6and A11), these variants showed significantly lower levels of reporter protein, although the biological activity for these modifications was still relatively high (Figure 6). The finding that even heavily modified 3’-terminal sequences, such as A13and A14containing a heterobase every 10 or 13 adenines maintain high translational activity is unexpected based on the current knowledge of the interaction of PABP with poly(A). Recent studies, have shown that modifications of oligoadenylate fragments with heterobases such as guanine, reduce the binding affinity to PABP, suggesting that such modifications of the poly(A) tail could negatively impact translational activity of mRNA.(Perzanowska et al. 2022) Consequently, most of the technologies developed for poly(A) tail modification focused on modifying the very 3’-end of poly(A) tail (Li et al. 2022) or inserting a single heterosequence within the poly(A) tail (Comirnaty Pfizer-BioNTech BNT162b2 anti-COVID mRNA vaccine) (Trepotec et al. 2019; Sahin et al. 2021 ). Here, we demonstrated that even dense placement of heterobases, not only does not decrease translational activity of mRNA, but surprisingly, even increases it. Such heavily modified 3’-terminal sequences according to our invention combine two features desirable in the context of therapeutic mRNA - increased stability of corresponding plasmid DNA encoding these molecules and high translational activity of mRNA.

[0029] The observation that the highest levels of protein expression in cells were recorded for modifications containing a motif with a key fragment of 30 consecutive adenines at the beginning of the 3’-terminal sequence (A2, A6, A11) is noteworthy. In contrast, variants of similar length but lacking this uninterrupted fragment (A13-A17) exhibited notably lower protein outputs (Figure 6). The almost twofold lower luciferase activities for mRNA variants lacking the uninterrupted adenine fragment (A13-A17) compared to the most similar modification carrying this motif (A11) suggests that the 5’ terminal part of the poly(A) tail plays a critical role in efficient translation (Figure 6). This may indicate that binding the first PABPC protein to the poly(A) tail of mRNA is particularly important for translation initiation or mRNA stability. The frequency of linkers in the further part of the 3’-terminal sequence after the key fragment of 30 adenines at the 5' end does not significantly affect the translational activity of mRNA, allowing for greater flexibility in selecting the length of subsequent adenine fragments. The effect of using a key fragment of 30 adenines at the beginning of the 3’-terminal sequence on the translational properties of mRNA is unexpected and, to the best of our knowledge, has not been previously disclosed. The best biological effect can be achieved by combining the key fragment of 30 adenines at the beginning of the poly(A) tail with optimal consecutive motifs in the poly(A) tail linked by single-nucleotide linkers.

[0030] Nanopore DRS sequencing was performed for all variants of FLuc and mKate2 mRNA with modified 3' terminal sequences as described in Example 7, which enabled a direct and accurate measurement of the composition of 3’-terminal sequences in each mRNA variant (Figure 7). The sequencing confirmed the presence of a long homologous adenine chain in the mRNA with Ago tail (R1), a tail consisting of two segments in R2, and the presence of single nucleotide linkers in A2, A6, A8, A9, and A10within the 3’-terminal mRNA sequence, as well as the proper incorporation of the hexanucleotide linkers in A3and A4. The current-voltage changes observed during sequencing provided precise information about the type and length of the linkers in each sequence, and the results were a perfect match with the initial DNA templates (Figure 7). On the basis of the DRS sequencing reads, a 100% match of the mRNA sequences of the modified 3'-terminal sequences was confirmed, compared to the initial DNA templates. Additionally, mouse macrophage (BMDM) cells were also transfected with all modified 3' terminal sequence mRNAs and after 24 hours of incubation, qualitative analysis of the presence of the Flue and mKate2 reporter protein was performed using Western Blot technique, as described in Example 8. Three independent biological replicates were performed and indicated the presence of FLuc in all modifications tested except for the R6- containing a G-quadruplex structure (Figure 8). The intensity of the chemiluminescence signal in all replicates was significantly higher for most of the modified 3' terminal sequences compared to the references. At the same time, no clear change was observed in the control analysis detecting a-tubulin, indicating a very similar condition of all cells tested after transfection.

[0031] Functional in vivo studies were conducted for Flue mRNA containing the poly(A) variants of the R1and R5references, as well as two modified 3’-terminal sequences with the highest biological activity in vitro, namely A6and A11. The mRNA molecules were synthesized and purified using standard procedures, followed by formulation with a commercially available lipid mixture as described in Example 9. Lipid nanoparticles for the tested variants were then administered to mice (Example 10). Each group consisted of 4 animals. Survival bioluminescence imaging was performed at time points of 4, 8, 12, and 24 hours post intravenous mRNA administration. The obtained results showed clear bioluminescence signals as shown in Figure 9. mRNA molecules with modified 3’-terminal sequences exhibited significantly higher levels of Flue protein expression compared to the reference R1. In particular, the A11variant displayed a very high level of bioluminescence at all tested time points, obtaining a significant advantage in the total protein level for this modification compared to both references (Fig. 9). The A6variant and the reference R5yielded results comparable to each other, maintaining a significant advantage in Flue protein levels compared to the classical poly(A)-tail R1, albeit less pronounced than for A11. Statistical analysis confirmed the significance of differences in the total bioluminescence signal obtained for the modified 3’- terminal sequences relative to the reference R1.

[0032] To verify the obtained results on another reporter protein, mRNA molecules with selected modified 3’-terminal sequences encoding human erythropoietin (hEPO) were prepared. In this experiment, reference R1, previously tested variants A6and A11, as well as variant A17, which exhibited the highest protein expression levels in vitro among the modifications in the A13-A17set, were selected. mRNA molecules with the mentioned modifications were prepared using the same methodology as described above. The LNPs containing hEPO mRNA with modified 3'-terminal sequences (A6, A11, A17) and the R1reference were intravenously administered to mice, with each group consisting of 5 animals (Example 11 ). Blood was collected from the mice after 4 and 24 hours and serum was prepared. This serum was then used for the quantitative determination of human erythropoietin (hEPO) levels using the ELISA immunoassay, as shown in Figure 10. The analysis revealed the highest level of hEPO protein after 4 hours for the A11variant. Protein expression for the A17modification was slightly higher compared to the R1reference, and the difference between A6and R1was not statistically significant. After 24 hours, the A11modification still showed a statistically significant advantage in protein production compared to R1. Interestingly, at the later time point, the A17variant also exhibited significantly higher hEPO expression levels compared to R1, as shown in Figure 10. These results indicated that mRNA with modified 3’- terminal sequences, particularly A11and A17, enhanced hEPO production levels in the mouse model compared to the classical poly(A)-tail sequence R1. This effect was further strengthened at the later time point, demonstrating prolonged expression for the modified mRNA. The experiment also confirmed that the observed effects are universal for different proteins. Regardless of the reporter protein used, mRNA containing appropriately modified 3’-terminal sequences yielded significantly higher protein output, which is of great significance in therapeutic applications, where the produced levels of a specific protein plays a crucial role in treatment success.

[0033] Modified 3'-terminal sequences exceeding 150 nucleotides (A19and A20variants, with approximately 200 nucleotides of the poly(A) tail) were also prepared and subjected to in vitro studies on three cell lines (A549, HEK293T, and JAWSII; Example 12). mRNA encoding human erythropoietin (EPO) for the A19and A20variants, as well as for the A11variant, which demonstrated the highest therapeutic potential among the previously studied modifications, was synthesized. The cells were transfected with EPO-encoding mRNA, and protein levels were assessed using an ELISA immunoassay at 4, 24, and 48 hours. The results (Fig. 11 ) are presented as summed signals for each variant at all time points (AUC). The A19and A20variants showed comparable levels of EPO production in JAWSII cells and slightly higher levels in A549 cells compared to A11. In contrast, the highest protein production level in HEK293T cells was observed for the A11variant. These data highlight the high translational potential of mRNAs with modified 3'-terminal sequences exceeding 150 nucleotides. These modifications demonstrated stable EPO expression, with the overall protein production levels indicating comparable, or even more efficient, translation of mRNA strands with long modified poly(A) tail sequences in the selected cell lines.

[0034] We conclude that the modified 3' terminal sequences according to our invention represent a significant development in the field of mRNA as a molecular tool and a therapeutic. The use of unique motifs consisting of heteronucleotide linkers at specific sites and consecutive adenines fragments appropriate length resulted in stable plasmid constructs at the DNA level, which does not shorten during amplification, and at the same time augments mRNA translational properties and stability. This enables longer and more efficient translation of the protein in comparison to mRNAs with unmodified poly(A) tails that can be reliably encoded in a plasmid-derived DNA template. This is especially important given the potential use of such modified mRNAs in the therapeutic mRNA field. In this case, it is desirable that the expression of a given protein occurs not only with higher efficiency, but also over a longer period of time, which, depending on the application, may affect not only the therapeutic effect but also the reduced response of the host immune system.

[0035] SHORT DESCRIPTION OF THE FIGURES

[0036] In order to facilitate understanding of the nature of the invention, the individual steps are presented in the examples section as well as in the form of attached graphics and tables in which:

[0037] Fig.1 A shows the general scheme of mRNA tail structures, in which the labels on the left (R1, R2, R3, R4,R5and R6) are the names of the used reference mRNAs with various poly(A) structures, the following labels A2, A3, A4, A6, A8, A9, and A10, A11, A13, A14, A15, A16, A17are the names of the mRNAs with 3' terminal sequences modified according to this invention. Inside each rectangle of the construct: A is adenine nucleotide, G is guanine nucleotide, C is cytosine nucleotide. The sign below the A indicates the number of adenine nucleotides in a particular fragment.

[0038] Fig. 1 B shows the general scheme of additional mRNA tail structures in which the labels on the left A18, A19and A20are the names of the mRNAs with 3' terminal sequences modified according to this invention. Inside each rectangle of the construct: A is adenine nucleotide, G is guanine nucleotide, C is cytosine nucleotide. The sign below the A indicates the number of adenine nucleotides in a particular fragment.

[0039] Fig. 2 shows the effect of different 3’-terminal sequences in the mRNA molecules encoded in plasmid DNA tail on the percentage of bacterial transformants undergoing sequence loss or shortening during amplification. The graph contains the number of bacterial colonies from which purified plasmids after amplification showed a shortened poly(A) tail compared to the original version. The group tested contained about 30 colonies for each modification. (A) Bacterial strain E.coli TOP10. (B) Commercially available NEB® Stable Competent E. coli (High Efficiency) strain.

[0040] Fig.3 shows quality control after transcription and purification of mRNA molecules with modified 3' terminal sequences. A, B) Chromatograms obtained during mRNA purification using HPLC for both references (R1, R2, R4, R5, R6) and all modifications, including mRNAs with modified 3' terminal sequences (A2-A10). C) The electrophoretic analysis of mRNAs demonstrating their homogeneity. An image of 1 % agarose gel prepared after complete purification for all variants of mRNA with modified 3' terminal sequences (R1, R2, R4, R5, R6, A2-A10). An aliquot containing -100 ng of each mRNA was analysed. D) Dot blot analysis to determine the double-stranded RNA (dsRNA) contents. An image of the chemiluminescence signal on a nylon membrane for serially diluted positive control dsRNA (left) and all samples of mRNA variants with modified 3’ terminal sequences in an amount of 25 and 250 ng (middle and right).

[0041] Fig.4 shows relative activity (chemiluminescence) of Firefly luciferase (which is proportional to luciferase protein expression levels) in cells following transfection with different mRNA variants, including all references (R1-R6) as well as mRNA with modified 3'-terminal sequences (A2-A10). A, B) A549 cell line, relative activity of Firefly luciferase at four time points posttransfection (4 h, 16 h, 24 h, and 48 h). C, D) HEK293T cell line, analogous to above. E, F) HepG2 cell line analogous to above. G, H) JAWS II cell line, analogous to above.

[0042] Fig.5 shows the total protein expression levels considering statistical significance (P < 0.05).

[0043] A) The total protein expression in A549 cell line with statistic significance relative to R1(A90).

[0044] B) The total protein expression in HEK293T cell line with statistic significance relative to R1(A90). C) The total protein expression in HepG2 cell line with statistic significance relative to R1(A90). D) The total protein expression in JAWS II cell line with statistic significance relative to R1(A90). The one-way analysis of variance (ANOVA), (ns,* - p< 0.1 ,**- p<0.01 , ***- p<0.001 , p<0.0001 ).

[0045] Fig.6 shows protein expression levels (Flue) for mRNA modifications of the poly(A) tail containing modules of identical length throughout the poly(A) tail (variants: A13, A14, A15, A16, A17) compared with selected mRNA modifications with the arrangement of the key first module of 30 adenines (variants: A6, A11). Values are expressed by the parameter AUC (Area Under the Curve) calculated, as the summed value of the areas under the graphs obtained at each time point. The data indicate a significant effect of the first segment containing 30 adenines on the expression level of the reporter protein in all cell lines tested. Variants having a shorter first segment than 30 adenines had significantly lower FLuc protein expression compared to variants containing a segment at the 5' end of the polyA tail containing an uninterrupted sequence of 30 adenines. This demonstrates the non-obvious advantage of the key requirement of the first 30 adenines in the modified 3’ terminal sequences over other solutions.

[0046] Fig.7 shows the results of sequencing selected variants (R1-R6, A2-A10) of mRNA with modified 3’ terminal sequences expressed as a graph of changes in current [pA] from the length of sequencing time [s]. The ranges of the graphs shown include only the poly(A) tail or corresponding 3’-terminal sequences of the variant. The data obtained confirm the presence (at the RNA level) of modifications designed and introduced by cloning techniques at the DNA level. The graphs very clearly show individual linkers both single-nucleotide, e.g. A2and A6, as well as longer six-nucleotide linkers, e.g. A3and A4, and higher-order structures, e.g. G- quadruplex (R6).

[0047] Fig. 8 shows the results of Western blot analysis for selected mRNA with modified 3' terminal sequences for the Flue reporter protein (top) and mKate2 protein (bottom) performed on mouse macrophages one day after cell transfection. Specific antibodies recognizing both reporter proteins (Anti-FLuc and Anti-mKate2) were used to prepare the analyses. The results show very clear expression of Flue and mKate2 proteins for all tested mRNA variants with modified 3’ terminal sequences except the modification with G-quadruplex, for which chemiluminescence measurements also indicated low production of the reporter protein.

[0048] Fig. 9 shows the production of Firefly Luciferase from mRNA in a mouse model. A) Images of mouse bioluminescence after administration of Flue mRNA in SM-102 formulation for references (R1, R5) and 3' terminal modified sequences A6and A11after 4 h. B) Graph shows total bioluminescence after administration of Flue mRNA in SM-102 formulation for references (R1, R5) and 3' terminal modified sequences A6and A11in timepoints (4h, 8h, 12h, 24h). C) Statistical significance analysis of the total bioluminescence signal for R5, A6and A11variants compare to the reference

[0049] Fig. 10 shows the production of hEPO from hEPO mRNA in a mouse model. A) General scheme of the experiment: hEPO mRNA with unmodified poly(A) (R1) or one of three poly(A) variants (A6, A11, A17) formulated with SM-102 lipid was administered intravenously, followed by blood collection after 4 and 24 hours. Protein levels in blood serum were determined by ELISA immunotest. B) Comparison of hEPO protein levels in vivo in blood serum at 4 and 24 hours post-administration expressed as mIU / ml. The one-way analysis of variance (ANOVA), (ns,* - p< 0.1 ,**- p<0.01 , ***- p<0.001 , ****- p<0.0001 ). \

[0050] Fig. 11 shows the production of hEPO from hEPO mRNA in cell lines A549, HEK293T and JAWSII for modified 3’-terminal variants with poly(A) longer than 150 nt (A19, A20) compared to A11modification. Values are expressed by the parameter AUC (Area Under the Curve) calculated, as the summed value of the areas under the graphs obtained at each time point (4 h, 24 h, 48 h). The data for all three cell lines confirm higher or comparable EPO protein expression levels for 3' terminal modified poly(A) tail sequences exceeding 150 nucleotides (A19, A20) relative to variant A11.

[0051] Table 1A shows nucleotide sequences of complementary DNA oligonucleotides used to clone and prepare plasmid DNA templates for modified 3’ terminal sequences R1, R2, R3, R4, R5, R6, Table 1 B shows nucleotide sequences of complementary DNA oligonucleotides used to clone and prepare plasmid DNA templates for modified 3’ terminal sequences A18, A19, A20.

[0052] Table 2 shows the coding DNA sequence for the Firefly luciferase, mKate2_PEST and human erythropoietin proteins used as a reporter genes for mRNA studies with modified 3’ terminal sequences.

[0053] Table 3 shows the samples of modified terminal 3’ sequences constructs coding the Firefly luciferase variant A11(Fluc_A11) and mKate2_PEST variant A2(mKate2_A2) used as reporter genes for mRNA studies with modified 3’ terminal sequences.

[0054] EXAMPLES

[0055] The following examples are provided only to illustrate the invention and to explain its particular aspects. The following examples do not to limit the invention, and should not be considered as its entire scope, that is defined in the appended claims. The following examples used standard materials and methods or followed manufacturer’s recommendations for specific materials and methods unless otherwise indicated.

[0056] Example 1. Preparation of DNA plasmid vectors and inserts for the designed modified 3’ terminal sequences (R1-R6, A2- A20)

[0057] Plasmid DNA vectors with modified 3’ terminal sequences were prepared by designing proper insertions as double-stranded DNA oligonucleotides (Table 1A and Table 1 B) and adding each modified 3’ terminal variant to the pJet plasmid vector that encodes Firefly luciferase (a short-lived version of the FLuc protein with a half-life of approximately 4 h) and mKate2-PEST gene. For this process blunt end cloning method was used. The doublestranded DNA insert for R4was prepared as follows: Solutions of two DNA oligonucleotides (Genomed) with sequence: A28GA60 (coding strand); T60CT28 (template strand) were mixed in a 1 :1 ratio (final 60 pM of each DNA strand) and an enzyme reaction was set up to phosphorylate the 5' ends of the oligonucleotides by adding T4 Polynucleotide Kinase (NEB) and 10x bufferforT4 Polynucleotide Kinase (NEB). The reaction was carried out for 30 minutes at 37 °C. The strands were then hybridized by heating to 95 °C and slowly cooling to 25 °C for 2 h, step gradient ~2 °C / ~ 3 min. The insert DNA was purified using a commercial DNA purification kit (Macherey-Nagel) according to the protocol.

[0058] The DNA insert for A2was prepared as follows: Solutions of two DNA oligonucleotides (Genomed) with the sequences: A2GA30GA30GA30 (coding strand); T30CT30CT30CT2 (template strand) were mixed 1 :1 (final 60 pM of each DNA strand) and continued as for variant R4. The DNA insert for A3was prepared as follows: Solutions of two DNA oligonucleotides (Genomed) with the sequences: A2GCATATA30GCATATA30GCATATA30GCATATA30 (coding strand); T30ATATGCT30ATATGCT30ATATGCT30ATATGCT2 (template strand) were mixed 1 :1 (final 60 pM of each DNA strand) and continued as for variant R4.

[0059] The DNA insert for A4was prepared as follows: Solutions of two DNA oligonucleotides (Genomed) with the sequences: A2GCATATA30GCATATA30GCATATA30 (coding strand); T30ATATGCT30ATATGCT30ATATGCT2 (template strand) were mixed 1 :1 (final 60 pM of each DNA strand) and continued as for variant R4.

[0060] The DNA insert for R5was prepared as follows: Solutions of two DNA oligonucleotides (Genomed) with the sequences: A28CA60 (coding strand); T60GT28 (template strand) were mixed 1 :1 (final 60 pM of each DNA strand) and continued as for variant R4.

[0061] The DNA insert for A6was prepared as follows: Solutions of two DNA oligonucleotides (Genomed) with the sequences: A2CA30CA30CA30 (coding strand); T30GT30GT30GT2 (template strand) were mixed 1 :1 (final 60 pM of each DNA strand) and continued as for variant R4.

[0062] The DNA insert for R6was prepared as follows: Solutions of two DNA oligonucleotides (Genomed) with the sequences: A2GA30TGGGGTTTGGGGTTTGGGGTTTGGGGTA15 (coding strand); T15ACCCCAAACCCCAAACCCCAAACCCCT30T2 (template strand) were mixed 1 :1 (final 60 pM of each DNA strand) and continued as for variant R4.

[0063] The DNA insert for A8was prepared as follows: Solutions of two DNA oligonucleotides (Genomed) with the sequences: A2GA15GA15GA15GA15GA15GA15 (coding strand); T15CT15CT15CT15CT15CT15CT2 (template strand) were mixed 1 :1 (final 60 pM of each DNA strand) and continued as for variant R4.

[0064] The DNA insert for A9was prepared as follows: Solutions of two DNA oligonucleotides (Genomed) with the sequences: A2GA20GA20GA20GA20GA20 (coding strand); T20CT20CT20CT20CT20CT2 (template strand) were mixed 1 :1 (final 60 pM of each DNA strand) and continued as for variant R4.

[0065] The DNA insert for A10was prepared as follows: Solutions of two DNA oligonucleotides (Genomed) with the sequences: A2GA45GA45 (coding strand); T45CT45CT2 (template strand) were mixed 1 :1 (final 60 pM of each DNA strand) and continued as for variant R4.

[0066] The DNA inserts for the remaining variants (A11, A12, A13, A14, A15, A16, A17, A18, A19and A20) were prepared in an analogous manner (Table 1A and Table 1 B).

[0067] The circular plasmid that encodes Firefly luciferase or mKate2-PEST (5 pg) was linearized (16h, 37 °C) with Aarl restriction enzyme (Thermo), 10x Aarl buffer (Thermo) and 50x Oligo (Thermo) and purified using a commercial DNA purification kit (Macherey-Nagel). The enzymatic reaction was then prepared with DNA Polymerase I, Large Fragment (Klenow), 10x buffer 3 (NEB), and 10 mM NTP (Thermo) to remove 3' overhangs and filling in 5' overhangs to form blunt ends (15 min, 25 °C) and again purified the vector using a commercial DNA purification kit (Macherey-Nagel). Finally, dephosphorylation reaction was performed using FastAP Thermosensitive Alkaline Phosphatase (Thermo) and 10 x Fast AP buffer (Thermo) by incubating (10 min, 37 °C) and purifying the final vector (V1 ) using a commercial DNA purification kit (Macherey-Nagel). The final (V1 ) solution including 100 ng of DNA was mixed with the inserts for all variants (A2-A20) at a molar ratio of 1 :3 (vector- insert), added 10x Ligase buffer (Thermo), PEG 2000 (Thermo), ATP (Thermo) and T4 DNA Ligase (Thermo) and incubated (1 h, 25 °C) and then reaction mix (20 pl) cooled to 4 °C and incubated an additional 1 h. Half of the ligation mixture (10 pl) was mixed with pre-melted on ice commercially available chemo competent bacteria (50 pl) Stable Competent E. coli (NEB) and incubated on ice for 30 min, then the mixture was transferred to 42 °C (heat shock) and cooled on ice for 2 min. Then 500 pl of NEB Stable outgrowth medium (NEB) was added and incubated at 30 °C for 1 h with shaking (300 RPM). The transformation mixture (200 pl) was spread on LB-agar plate (Roth) with 100 mg / ml Ampicillin (Roth) and incubated (30 °C, 16 h). Then, single colonies for each variant (A2-A20) were selected and inoculated with liquid LB medium (Roth) in a volume of 5 ml and supplemented with Ampicillin (Roth) at a concentration of 100 mg / ml and the cultures were incubated (30 °C, 16 h). The bacterial cultures were then centrifuged (4000g, 10 min) and the plasmids of all variants (A2-A120) were purified using the commercial GeneJET Plasmid Miniprep Kit (Thermo). Concentration was measured using a Nanodrop 2000c spectrophotometer (Thermo) and plasmids were sent for DNA sequencing using the Sanger method (Genomed). Based on the sequencing results confirming the incorporation of the correct sequences encoding modified 3’ terminal sequences (A2-A20), plasmid solutions were used to transform fresh chemo competent bacteria (analogous to the previous procedure), then a single colony was selected from an LB agar plate (Roth) and inoculated with 200 ml of LB medium (Roth) with Ampicillin (Roth) at a concentration of 100 mg / ml and incubated (30 °C, 16 h). The bacteria were then centrifuged (4000g, 10 min) and the plasmids were purified on a large scale using the commercial E.Z.N.A.® FastFilter Plasmid DNA Maxi Kit (Omega) for all variants (A2-A20). The plasmid vectors encoding the Firefly lucyferase or mKate2-PEST with different modified 3’ terminal sequences (A2-A20) were linearized by mixing their solutions containing 30 pg DNA with Aarl enzyme (Thermo), 10x Aarl buffer (Thermo), 50x Oligo (Thermo) and incubated (16 h, 37 °C). DNA templets for in vitro transcription of RNA were then purified using a commercial DNA purification kit (Macherey-Nagel). Furthermore, to control template homogeneity, a 1 % agarose DNA gel (VWR) was made by applying -100 ng of DNA templates for each variant (A2-A20) and 6x load dye (Thermo). Electrophoresis was carried out under conditions (1x TAE, 140 V, 25 min).

[0068] Example 2. Instability analysis of poly(A) tails in plasmid DNA vectors

[0069] Chemocompetent E. coli Top10 bacteria and NEB NEB® Stable Competent E.coli (High Efficiency) bacteria were transformed as described in Example 1 with plasmids containing all modifications of poly(A) tails (A2-A20) and plasmids containing reference poly(A) tails (R1-R6). Approximately 30 individual colonies for each modification and references were selected from bacterial dishes, liquid bacterial cultures were established and plasmids were purified, analogous to the procedure described in Example 1. Purified plasmids sets were then sequenced and evaluated the length and sequence validity of poly(A) tails after amplification. Based on the results, the number of colonies that showed deviations from the initial sequence parameters (length, homogeneity) was counted and presented as a percentage of bacterial clones in which the poly(A) tail sequence was shortened during plasmid amplification.

[0070] Example 3. In vitro transcription and purification of the mRNA encoding Firefly lucyferase with different modified 3’ terminal sequences

[0071] In vitro transcription mix (100 pl) for each modified 3’ terminal sequences variant was prepared by mixing: 4 pg linearized DNA template for each variant (A2-A20), transcription buffer (x1 , Thermo), GTP (4. 0 mM, Thermo), CTP (5.0 mM, Thermo), ATP (5.0 mM, Thermo), me1- UTP (5.0 mM, Jena Bioscience), rn7GpppAmpG- cap1 (10 mM), MgCh (15 mM), Ribolock (1 U / pL, Thermo), Inorganic Pyrophosphatase (0.002U / pL, Thermo), T7 RNA Polymerase (0.125 mg / mL). After incubation for 1 h at 37°C, DNase I (6 pL, 30 min, Thermo) was added and incubation continued for another 30 min. The mixture was then diluted twice with water and an EDTA solution (9 pl, 500 mM, VWR) was added to inhibit the reaction. The crude RNA for all variants (A2-A20) was purified after IVT using POROS™ Oligo (dT)25 Affinity Resin (Thermo) by applying the entire reaction volume to a previously conditioned resin (2 ml) of Oligo (dT)25(20mM Tris pH=7.4, 800 mM NaCI) and incubated (10 min, 25°C). The resin was then washed with equilibration buffer (5 volumes, 20 mM T ris pH=7.4, 800 mM NaCI) and wash buffer (5 volumes, 20 mM Tris pH=7.4, 300 mM NaCI). The elution was performed by incubation resin with water (2 times 3 volumes each, 5 min, 65 °C). All RNA (A2-A20) solutions were sequentially filtered through a sterile 0.22 pM filter and concentrated (Amicon 100 kDa, 15 min, 5000g, 4 °C) to a final volume of 100 pl. The final purification step was separation by HPLC using the RNASep™ Prep - RNA Purification Column (ADC Biotec). Buffers A: 100 mM TEAA, B: 100 mM TEAA, 100 mM MeCN. Program: 20% B for 5 min, 20-29% B in 20 min, 29- 100% B in 1 min, 100% B for 4 min, flow 5.0 mL / min, 55°C (RT ~ 35 min). The collected fractions of RNA for all variants (A2-A20) were precipitated (0.1 volume 3M NaOAc, 0.7 volume Isopropanol, 30 min, -80 °C), centrifuged (20 min, 15 000g, 4 °C), resuspended in 3 ml of 80% EtOH, centrifuged (10 min, 15 000g, 4 °C) and redissolved in water (120 pL). As a quality control, RNA grade agarose gel (1 %, Thermo) was made and -100 ng of each RNA variant (A2-A17) was applied along with 2x loading RNA dye (Thermo). Electrophoresis was performed under the following conditions: (1x TBE, 140V, 20 min). Additionally, dot blot analysis was performed to check the presence of double-stranded RNA (ds-RNA) molecules. Solutions of 25 and 250 ng of each RNA variant with modified 3’ terminal sequences (A2- A20) were applied to positively charged nylon membrane (Sigma) using a Bio-Dot Apparatus (Bio-Rad), UV crosslinked, blocked with 5% nonfat dried milk in PBST buffer and incubated with mouse monoclonal J2 antibody (16h, 4 °C, SCICONS) and then with secondary anti-mouse horseradish peroxidase conjugated antisera (1 h, 25 °C, Thermo). As a positive control, a commercial dsRNA mix ladder (25 ng, NEB) was used. Signals were detected with Immobilon Western Chemiluminescent HRP Substrate (Merck Millipore) on Amersham Imager 600 (Citiva).

[0072] Example 4. Functional in vitro characterization by protein expression level of different modified 3’ terminal mRNA’s

[0073] A549 cells (human epithelial lung carcinoma, ATCC CCL-185), HEK293T (human epithelial kidney, ATCC CRL-3216) and HepG2 (human epithelial hepatocellular carcinoma, ATCC HB-8065) were grown in DMEM medium (Gibco) supplemented with 10% FBS (Sigma), GlutaMAX (Gibco) and 1 % penicillin / streptomycin (Gibco) at 5% CO2 and 37 °C. JAWS II (mouse immature dendritic cells, ATCC CRL-11904) were grown in RPMI 1640 (Gibco) supplemented with 10% FBS, sodium pyruvate (Gibco), 1 % penicillin / streptomycin and 5 ng / ml GM-CSF (PeproTech) at 5% CO2 and 37 °C. One day before transfection cells were seeded in a 96-well plate (10A4 cells per well). The next day, cells were transfected with RNA constructs containing different variants of poly(A) tails. In each cell line, each RNA variant was transfected in triplicate. Transfection was performed with Lipofectamine Messenger MAX (Invitrogen, Waltham, MA, USA), according to the manufacturer's instructions. Briefly, for each reaction 0,15 pL of Lipofectamine was first diluted (Mix 1 ) in 5 pL of Opti-MEM (GIBCO, Grand Island, NY, USA) medium and incubated at room temperature for 10 minutes. In another tube, 50 ng of RNA was diluted in 5 pL of Opti-MEM medium (Mix 2). Next, both Mix 1 and Mix 2 were mixed and incubated for 5 minutes at room temperature. Afterward, the mixture was transferred to the appropriate well of a 96-well plate containing previously seeded cells. After the transfection, the cells were incubated in standard culture conditions (37 °C, 5% CO2, saturating humidity). At appropriate time points (4h, 16h, 24h, 48h, 72h) the luminescence was measured utilizing Bright-Glo Luciferase Assay System (Promega, Madison, Wl, USA), according to the manufacturer's instructions. The luminescence was read in white, opaque plates. Measurements were made using an EnVision plate reader (Perkin Elmer, Waltham, MA, USA).

[0074] Example 5. Murine bone marrow-derived macrophage (BMDM) cell cultures

[0075] The primary BMDM cell cultures were established from the bone marrow monocytes isolated from wild-type adult mice (12-25 weeks old). Animals were sacrificed by cervical dislocation, then femurs and tibias were isolated and bone marrow was harvested by centrifugation-based protocol. Bone marrow cells were plated in IMDM medium (Thermo Fisher Scientific; 21980065) supplemented with 10% FBS (Gibco), 100 U / ml penicillin / 0.1 mg / ml streptomycin solution (Sigma-Aldrich), and 10 ng / ml macrophage colony-stimulating factor (M-CSF, Preprotech; 315-02) and cultured at 37°C in 5% CO2. Cells were used for experiments after 14 days of differentiation.

[0076] Example 6. In vitro BMDM cells transfections for FLuc modified 3’ terminal mRNA

[0077] The 0.5 M of cells were seeded the day before on a 6-well plate in media as described above. Cells were transfected with 1 pg of IVT mRNA coding FLuc. All transfections were carried out Lipofectamine™ MessengerMAX™ Transfection Reagent (Invitrogen, LMRNA001 ) according to manufacturer instructions. Cells were harvested 24h after transfections for subsequent analyses.

[0078] Example 7. Western blots for FLuc modified 3’ terminal mRNA

[0079] An equal amount of cells were lysed in PBS supplemented with 0.1 % NP40, protease inhibitors and viscolase (final concentration 0.1 U / ml; A&A Biotechnology, 1010-100) for 30 min at 37 °C with shaking 1200 rpm, then 3x SDS Sample buffer (187.5 mM Tris-HCI pH 6.8, 6% SDS, 150 mM DTT, 0.02% Bromophenol blue, 30% glycerol, 3% 2-Mercaptoethanol) was added and samples were boiled for 10 min. Samples were resolved on 12-15% SDS-PAGE gels and then proteins were wet transferred to Protran nitrocellulose membranes (GE Healthcare) at 400 mA at 4°C for 1 .5 h in 1x Transfer buffer (25 mM Tris base, 192 mM glycine, 20% methanol (v / v)). Next, the proteins were visualized by staining with 0.3% w / v Ponceau S in 3% v / v acetic acid and digitalized. Membranes were blocked by incubation in 5% milk in TBST buffer for 1 h followed by overnight incubation with specific primary diluted 1 :3000 (mKate2, Firefly luciferase), 1 :5000 (actin, tubulin) in 5% milk in TBST buffer. Membranes were washed three times in TBST buffer, 10 min each, incubated with HRP-conjugated secondary antibodies: anti-mouse (Millipore, 401215) diluted 1 :5000 and anti-rabbit (Millipore, 401393) diluted 1 :5000, for 2 h at RT. Membranes were washed three times in TBST buffer and proteins were visualized using ChemiDoc System. Example 8. NanoPore DRS sequencing of modified 3’ terminal sequences

[0080] Direct RNA sequencing was performed as described by Bilska (Bilska et al. 2020). The 50-1 OOng of IVT mRNA was used for library preparation with a Direct RNA Sequencing Kit (catalog no. SQK-RNA002, Oxford Nanopore Technologies) according to the manufacturer’s instructions. Sequencing was performed using R9.4 flow cells on a MinlON device (ONT). Raw data were base called using Guppy(ONT). Raw sequencing data (fast5 files) were deposited at the European Nucleotide Archive (ENA, accession numbers to be provided).

[0081] Example 9. mRNA formulation into LNPs

[0082] The formulation process was conducted using the Ignite device (Precision Nano Systems). mRNA molecules for all tested variants were combined with SM-102 lipid mixtures through the single-use mixing cartridges of the Ignite system, utilizing a fourfold molar excess of mRNA relative to lipids. The flow rate of the mixed solutions was set at 12 ml / min, and the calculated N / P ratio (the ratio of positively charged amine groups of the polymer [N - nitrogen] to negatively charged phosphate groups of nucleic acids [P]) based on the gene sequence was 6. Subsequently, the mRNA concentration of the obtained LNPs was measured using the RiboGreen assay. The formulated mRNA underwent quality control to assess the size of the nanoparticles (nm) and the homogeneity of the samples, determined via the polydispersity index (PI). The evaluation of the LNPs was performed using the dynamic light scattering (DLS) method. All LNP-mRNA samples demonstrated similar nanoparticle sizes in the range of approximately 70-90 nm. The PI values for all LNP-mRNA samples did not exceed 0.1 , indicating high purity and homogeneity of the obtained LNPs.

[0083] Example 10. Determination of Firefly Luciferase activity in vivo after administration of 3’-modified mRNA

[0084] The experiments were carried out in 10-12-week-old (~25 g) female BALB / c mice under the protocol approved by the II Local Ethical Committee for Experiments on Animals in Warsaw, Poland (WAW2 / 126 / 2021 ). All experiments were conducted in accordance with the Directive of the European Parliament and Council No. 2010 / 63 / EU on the protection of animals used for scientific purposes. Mice were obtained from the Breeding Facility of the Mossakowski Institute, Polish Academy of Science, Warsaw. All animals were maintained in specific pathogen-free (SPF) environment in the individually ventilated cages (IVC) under the conditions of a 12-h day / night cycle with unrestricted access to food and drinking water. The mice were randomly assigned to experimental groups (n=5 female BALB / c mice per group). Each mouse received 100 pL of SM-102-encapsulated FLuc mRNA (10 pg per mouse) intravenously. Bioluminescence imaging was performed using the In Vivo Imaging System (MS, PerkinElmer, Waltham, MA, USA) at 4-, 8-, 12-, and 24-hours post-injection. Mice were anesthetized with isoflurane, and D-luciferin (150 mg / kg, D-Luciferin Potassium Salt, Syd Labs, Inc., # MB000102-R70170) was injected intraperitoneally 5 minutes prior to imaging. Bioluminescence was quantified as total photon flux (photons / second) using Living Image software (PerkinElmer).

[0085] Example 11. Determination of hEPO levels in blood after administration of 3’-modified mRNA

[0086] The experiments were carried out in 10-12-week-old (~25 g) female C57BL / 6 mice under the protocol approved by the II Local Ethical Committee for Experiments on Animals in Warsaw, Poland (WAW2 / 085 / 2023). All experiments were conducted in accordance with the Directive of the European Parliament and Council No. 2010 / 63 / EU on the protection of animals used for scientific purposes. Mice were obtained from the Breeding Facility of the Mossakowski Institute, Polish Academy of Science, Warsaw. All animals were maintained in specific pathogen-free (SPF) environment in the individually ventilated cages (IVC) under the conditions of a 12-h day / night cycle with unrestricted access to food and drinking water. The mice were randomly assigned to experimental groups (n=5 female C57BL / 6 mice per group). Each mouse received 100 pL of SM-102-encapsulated human erythropoietin (hEPO) mRNA (1 pg per mouse) intravenously. Blood samples were collected from submandibular vein (cheek pouch) of each mouse at 4- and 24-hours post-injection. The blood was allowed to clot at room temperature, and serum was separated by centrifugation at 2000 x g for 15 minutes. Human EPO levels in serum were quantified using an ELISA kit (Invitrogen, Human EPO ELISA Kit, # BMS2035-2) according to the manufacturer’s instructions.

[0087] Example 12. Determination of hEPO levels in cell lines of 3’-modified mRNA with length exceeding 150 nt

[0088] One day prior to transfection, cells (HEK 293T, A549, or JAWS II) were seeded in a 96-well plate at a density of 104cells per well in 200 pL of culture medium. The following day, cells were transfected with mRNA. Each mRNA variant was transfected in triplicate for each cell line. Transfection was conducted using Lipofectamine Messenger MAX (Invitrogen, Waltham, MA, USA) according to the manufacturer's instructions. Briefly, for each reaction, 0.15 pL of Lipofectamine was first diluted (Mix 1 ) in 5 pL of Opti-MEM (GIBCO, Grand Island, NY, USA) medium and incubated at room temperature for 10 minutes. In a separate tube, 200 ng of mRNA was diluted in 5 pL of Opti-MEM medium (Mix 2). The two mixes were then combined, gently mixed, and incubated for 5 minutes at room temperature. Subsequently, 10 pL of the mixture was transferred to the appropriate well of the 96-well plate containing the previously seeded cells. After transfection, the cells were incubated under standard culture conditions (37°C, 5% CO2, and saturating humidity). At designated time points (6, 24, and 48 h post- transfection), 50 pL of medium was collected from each well for protein quantification, and the remaining medium was replaced with fresh culture medium. The concentration of erythropoietin (EPO) secreted into the medium was measured using the Human EPO ELISA Kit (Invitrogen, Waltham, MA, USA) according to the manufacturer’s protocol. For analysis, medium samples were diluted 5000-fold. Absorbance was measured at 450 nm using an EnVision plate reader (Perkin Elmer, Waltham, MA, USA), and hEPO concentrations were determined based on a standard curve generated using recombinant human EPO standards.

[0089] BIBLIOGRAPHY

[0090] Bilska A, Kusio-Kobialka M, Krawczyk PS, Gewartowska 0, Tarkowski B, Kobylecki K, Nowis D, Golab J, Gruchota J, Borsuk E et al. 2020. Immunoglobulin expression and the humoral immune response is regulated by the non-canonical poly(A) polymerase TENT5C. Nature Communications 11.

[0091] Corbett KS, Edwards DK, Leist SR, Abiona OM, Boyoglu-Barnum S, Gillespie RA, Himansu S, Schafer A, Ziwawo CT, DiPiazza AT et al. 2020. SARS-CoV-2 mRNA vaccine design enabled by prototype pathogen preparedness. Nature 586: 567- +.

[0092] Li CY, Liang ZH, Hu YX, Zhang HX, Setiasabda KD, Li JW, Ma SH, Xia XJ, Kuang Y. 2022. Cytidine-containing tails robustly enhance and prolong protein production of synthetic mRNA in cell and in vivo. Molecular Therapy-Nucleic Acids 30: 300-310.

[0093] Perzanowska O, Smietanski M, Jemielity J, Kowalska J. 2022. Chemically Modified Poly(A) Analogs Targeting PABP: Structure Activity Relationship and Translation Inhibitory Properties. Chemistry-a European Journal 28.

[0094] Sagi J. 2014. G-quadruplexes incorporating modified constituents: a review. Journal of Biomolecular Structure & Dynamics 32: 477-511.

[0095] Sahin U, Muik A, Vogler I, Derhovanessian E, Kranz LM, Vormehr M, Quandt J, Bidmon N, Ulges A, Baum A et al. 2021. BNT162b2 vaccine induces neutralizing antibodies and poly-specific T cells in humans. Nature 595: 572-+.

[0096] Trepotec Z, Geiger J, Plank C, Aneja MK, Rudolph C. 2019. Segmented poly(A) tails significantly reduce recombination of plasmid DNA without affecting mRNA translation efficiency or half-life. Rna 25: 507-518.

[0097] (W02016005004A1)

[0098] (W02022028559A1)

Claims

CLAIMS1. An mRNA molecule comprising at least an open reading frame encoding a protein and a 3’ terminal sequence of nucleotides downstream of the open reading frame, wherein the 3’-terminal sequence is an oligonucleotide according to formula 1 :M-L-M’-L’-M”-L”-(Mn.Ln)n,Formula 1 wherein: each n represents an integer chosen independently from the range 0 to 15,M, M’, M” and each of M1to Mnis an oligonucleotide sequence chosen independently from the oligonucleotides according to formula 2,Formula 2 wherein each k is an integer chosen independently from the range 10 to 35;L, L’, L” and each of L1to Lnis: a nucleotide including a natural, modified, or unnatural nucleoside base different than adenine or adenine analogue, oran oligonucleotide sequence according to formula 3:Formula 3 wherein:B, Z and each of B1through B4is independently a natural, modified, or unnatural nucleoside base, each of m1, m2, m3, m4, m5is independently 0 or 1 with proviso that B and Z is not adenine or adenine analogue.

2. An mRNA molecule according to claim 1 , wherein m1= m2= m3= m4= m5= 0.

3. An mRNA molecule according to claim 1 , wherein M, M’, M” and each of Mi to Mnis identical.

4. An mRNA molecule according to claim 1 , wherein M is a structure according to formula 2, wherein k is from 27 to 33.

5. An mRNA molecule according to claim 1 , wherein L, L’, L” and each of L1to Lnis independently selected among of nucleotides G or C or oligonucleotides: GCAUAU,GCAUAUGACU, UGGGGUUUGGGGUUUGGGGUUUGGGGU or UCUAG or is abandoned.

6. An mRNA molecule according to claim 1 , wherein the 3’-terminal sequence is a structure according to following formulas 4-8:Formula 4 (A2)Formula 5 (A6)Formula 6 (A8)Formula 7 (A11)Formula 8 (A15)7. An mRNA molecule according to claim 1 , wherein the natural, modified, or unnatural nucleoside base is independently selected from the group consisting of:A, G, C, U, T, N7-methylguanosine, N6-methyladenosine, N2-methyladenosine, dihydrouracil, 5-methylcytosine, 5-hydroxymethylcytosine, pseudouridine, methylpseudouridine, as well as 2-aminopurine, 2,6-diaminopurine, C8-substituted adenine, C8-substituted guanine, C7-substituted 7-deazaadenine, C7-substituted 7- deazaguanine, C5-substituted uridine, C5-substituted cytosine, di- and tricyclic nitrogenous bases.

8. An mRNA molecule according to claim 1 , wherein the 3’-terminal sequence comprises more than 150 nucleotides.

9. An mRNA molecule according to claim 1 , wherein the 3’-terminal sequence has been selected among of Seq Id No: 81-96, preferably among of Seq Id No: 88, 95 and 96.

10. A nucleic acid molecule encoding an mRNA molecule according to claim 1-9.

11. The nucleic acid molecule according to claim 10, which is suitable, in particular after linearization, for in vitro transcription of RNA, in particular mRNA.

12. An mRNA molecule according to any of claims 1-9 or nucleic acid molecule according to any of claims 10-11 for use in medicine, in particular as antiviral vaccines, cancer vaccines, preventive vaccines, oral and epidermal pharmaceuticals or in personalized mRNA gene therapies.

Images