A live attenuated SARS-CoV-2 and a vaccine made thereof

The present invention relates to a codon-pair deoptimized polynucleotide encoding a respiratory syndrome coronavirus 2 (SARS-CoV-2) protein, to a live attenuated SARS-CoV-2 comprising such polynucleotide, to a pharmaceutical composition comprising such a live attenuated SARS-CoV-2, to a vaccination method for administering this pharmaceutical composition, to a vector comprising such polynucleotide, to a host cell comprising such polynucleotide, as well as to method of producing a virus.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) emerged in December 2019 as the causative agent of coronavirus disease 2019 (COVID-19) (Wu et al., 2020; Zhou et al., 2020b). The virus is highly transmissible among humans (Chan et al., 2020). It has spread rapidly around the world within a matter of weeks and the world is still battling with the ongoing COVID-19 pandemic.

SARS-CoV-2 primarily replicates in the upper respiratory tract (Zou et al., 2020). The infection with SARS-CoV-2 can cause a wide spectrum of clinical manifestations, ranging from asymptomatic to life-threatening disease conditions (Chen et al., 2020; Zhou et al., 2020a). Especially the elderly and patients with pre-existing conditions are at greater risk of developing more severe disease such as pneumonia, acute respiratory distress syndrome and multiple organ failure (Chen et al., 2020; Garg et al., 2020; Zhou et al., 2020a). The ongoing pandemic imposes an enormous health, psychological, economic, and social burden. To date (December 2021 ) more than 270 million people have been infected with the virus, of whom more than 5.3 million have died as a result of the infection (https://coronavirus.jhu.edu/map.html) (Dong et al., 2020).

The unprecedented scale and severity of the COVID-19 pandemic prompted the rapid development of novel diagnostic tests, therapeutics and vaccines which could be used to contain the spread of the virus and limit the pandemic. Globally, more than 90 vaccines are being tested in clinical trials, but only few have reached the final stages of testing (Zimmer et al., 2021 ). Almost all vaccines that have been or are being evaluated in clinical trials are based either on inactivated or subunit virus preparations (Ella et al., 2021 ; Gao et al., 2020; Wang et al., 2020; Zhang et al., 2021 ), replication-defective virus vectors (Emary et al., 2021 ; Logunov et al., 2021 ; Solforosi et al., 2021 ; Voysey et al., 2021 ; Zhu et al., 2020), or DNA/RNA molecules (Anderson et al., 2020; Baden et al., 2021 ; Corbett et al., 2020; Dagan et al., 2021 ;

Jackson et al., 2020; Mulligan et al., 2020; Polack et al., 2020; Sahin et al., 2020; Walsh et al., 2020).

SARS-CoV-2 is rapidly evolving (Tegally et al., 2021 ; Faria et al., 2021 ; Davies et al., 2021 ). Benefiting from its global presence, the virus continues to adapt to its new host and to infection-or vaccine- induced immunity. During the course of the pandemic, a number of genetic variants have emerged (Tegally et al., 2021 ; Faria et al., 2021 ; Davies et al., 2021 ). Variants that exhibit increased infectivity, cause greater morbidity and mortality, or have the ability to evade infection- or vaccine-induced immunity pose an increased threat to public health. The World Health Organization (WHO) and other national health agencies have independently established classification systems that categorize emerging variants as variants of interest (VOIs), variants under investigation (VUIs), or variants of concern (VOCs) based on their risk to public health (cf. Table 1 of Trimpert et al. “Live attenuated virus vaccine protects against SARS-CoV-2 variants of concern B.1.1.7 (Alpha) and B.1.351 (Beta)", Science Advances, Vol.

7, No. 49 (2021 )). In addition, to simplify communication with the public, the WHO recommends that VOIs and VOCs should also be labeled using the letters of the Greek alphabet. As of 12 August 2021 , viruses belonging to lineages B.1.1.7 (Alpha), B.1.351 (Beta), B.1.1.28.1 (Gamma), B.1.617.2 (Delta), and, most recently, B.1 .159.1 (Omicron) are classified by several health agencies as VOCs. In countries where they emerged, these variants rapidly supplanted the preexisting variants and started to spread globally.

The B.1.1.7 variant, first detected in the United Kingdom in December 2020, is 50 to 100% more transmissible and possibly also more lethal than earlier variants but shows no tendency to evade immunity induced by infection or vaccination (Davies et al., 2021 ; Volz et al., 2021 ; Abu-Raddad et al., 2021 ). The B.1.1.7 variant has been detected in 132 countries and rapidly became the dominant variant in Europe and the United States. The B.1.351 variant, first detected in South Africa in May 2020, is not only more transmissible but also capable of reinfecting individuals and of breaking through vaccine protection (Madhi et al., 2021 ; Johnson & Johnson; Naveca et al., 2021 ). The B.1 .1 .28.1 variant is similar to B.1 .351 in that both share some important mutations in the spike glycoprotein (E484K, K417N/T, and N501 Y). B.1 .1 .28.1 emerged in late 2020 in Manaus, Brazil (Faria et al., 2021 ). Similar to the B.1.351 variant, it can cause reinfection because it can bypass immunity developed after infection with other virus variants (Faria et al., 2021 ; Naveca et al., 2021 ). It is estimated that B.1.1.28.1 is 40 to 140% more transmissible, more pathogenic, and 10 to 80% more lethal than other variants (Faria et al., 2021 ). On 7 May 2021 , the WHO reclassified the B.1 .617.2 variant, first detected in India, as a VOC due to its high transmissibility (WHO). As of August 2021 , B.1.617.2 has largely outcompeted B.1.1.7 and became the predominant variant in Europe and the United

States. According to the WHO, B.1.617.2 is the most dangerous strain worldwide, and it has attracted considerable attention for its ability to evade infection- and vaccine-mediated protection (Dyer et al., 2021 ). Variant B.1 .1 .529 of SARS-CoV-2 was first detected in Botswana in November 2021 . It has become the predominant variant in circulation around the world. After the original BA.1 variant, several subvariants of Omicron have evolved: BA.2, BA.3, BA.4, and BA.5, with BA.5 dominating the world as of August 2022.

WO 2021/154828 A1 and WO 2023/283106 A1 describe modified SARS-CoV-2 coronaviruses. These viruses have been recoded, for example, codon deoptimized or codon pair bias deoptimized and are useful for reducing the likelihood or severity of a SARS-CoV-2 coronavirus infection, preventing a SARS-CoV-2 coronavirus infection, eliciting and immune response, or treating a SARS-CoV-2 coronavirus infection.

Y. Wang et al., 2021 , describes a live attenuated severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) vaccine named COVI-VAC that was developed by recoding a segment of the viral spike protein with synonymous suboptimal codon pairs (codon-pair deoptimization), thereby introducing 283 silent (point) mutations. In addition, the furin cleavage site within the spike protein was deleted from the viral genome for added safety of the vaccine strain.

CN 1 12175913 A describes a SARS-CoV-2 attenuated strain and its application in the prevention and/or treatment of a novel coronavirus pneumonia.

It is an object of the present invention to provide novel SARS-CoV-2 vaccine candidates.

This object is achieved with a specific polynucleotide encoding a) severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike protein; and optionally b) at least one non-structural SARS-CoV-2 protein selected from the group consisting of non-structural protein 7, non-structural protein 8, non-structural protein 9, non-structural protein 10, non-structural protein 1 1 , non-structural protein 12, an endoribonuclease (also referred to as non-structural protein 15), and a 2'-O-methyltransferase (also referred to as non-structural protein 16). In this context, the polynucleotide comprises or consists of at least one sequence part comprising codon-pair deoptimizations in comparison to the corresponding SARS-CoV-2 genome part. Furthermore, the polynucleotide comprises a furin cleavage site modification resulting in a functional loss of a furin cleavage site being naturally present in the SARS-CoV-2 genome. Thus, the polynucleotide is lacking some genetic information being present in naturally occurring SARS-CoV-2.

The term “furin cleavage site modification”, as used herein, refers to a site in the nucleotide sequence of the polynucleotide that corresponds to a site in the SARS-CoV-2 genome encoding for a furin cleavage site but having a modification such that it results in a substantial reduction or loss of (furin) cleavage susceptibility of the encoded cleavage site. Examples of such a modification are demonstrated herein below, e.g., in the examples. The furin cleavage site modification described herein can be any modification that alters the cleavage susceptibility, for example a (partial or full) deletion, an insertion, or a replacement of nucleotides or sequence parts in comparison to the SARS-CoV-2 genome. In some embodiments, the furin cleavage site modification described herein embodies not more than 2, not more than 3, not more than 4, not more than 5, not more than 6, not more than 7, not more than 8, not more than 9, not more than 10, not more than 11 , not more than 12, not more than 13, not more than 14, not more than 15, not more than 16, not more than 17, not more than 18, not more than 19, not more than 20, not more than 21 , not more than 22, not more than 23, or not more than 24 nucleotide modifications compared to the sequence part of the SARS-CoV-2 encoding the furin cleavage site.

The term “polynucleotide”, as used herein, refers to a molecule containing multiple nucleotides (e.g. mRNA, RNA, cRNA, cDNA or DNA). The term typically refers to oligonucleotides greater than 200, preferably greater than 300, preferably greater than 400, preferably greater than 500, preferably greater than 600, preferably greater than 700, preferably greater than 800, preferably greater than 900, preferably greater than 998 nucleotide residues in length. The polynucleotide of the present invention either essentially consists of the nucleic acid sequences described herein or comprises the aforementioned nucleic acid sequences. Thus, it may contain further nucleic acid sequences as well. The term polynucleotide encompasses single stranded as well as double stranded polynucleotides. Moreover, encompassed herein are also modified polynucleotides including chemically modified polynucleotides, artificial modified polynucleotides, or naturally occurring modified polynucleotides such as glycosylated or methylated polynucleotides.

The term “SARS-CoV-2” or “severe acute respiratory syndrome coronavirus 2”, as used herein, refers to any variant that is classified as SARS-CoV-2. In some embodiments, the SARS-CoV-2 described herein is at least one SARS-CoV-2 variant selected from the group consisting of Alpha, Beta, Gamma, Delta or Omicron variant. In some embodiments, the SARS-CoV-2 Omicron variant is at least one SARS-CoV-2 Omicron sub-lineage such as BA.1 , BA.2, BA.3, BA.4, or BA.5. In some embodiments, the SARS-CoV-2 refers to a SARS-CoV-2 Spike variant comprising at least one mutation selected from the group consisting of del 69-70,

RSYLTPGD246-253N, N440K, G446V, L452R, Y453F, S477G/N, E484Q, E484K, F490S, N501Y, N501 S, D614G, Q677P/H, P681 H, P681 R, and A701 V. In some embodiments, the SARS-CoV-2 refers to a SARS-CoV-2 variant comprising at least one mutation selected from the group consisting of G142D, G339D, S373P, S375F, K417N, N440K, S477N, T478K, E484A, Q493R, Q498R, N501Y, Y505H, D614G, H655Y, N679K, P681 H, N764K, D796Y, Q954H, N969K. In some embodiments, the SARS-CoV-2 refers to a SARS-CoV-2 variant comprising at least one mutation selected from the group consisting of L452R, F486V, and R493Q. Any combination of these mutations is possible and herewith disclosed. Specifically, all mutations listed in either of the precedingly mentioned groups of mutations can be present at the same time within a SARS-CoV-2 variant. Therefore, “SARS-CoV-2 protein” and “SARS-CoV-2 genome” may also be understood as the protein and genome of a SARS-CoV-2 variant, respectively.

The term “codon”, as used herein, refers to any group of three consecutive nucleotides in a coding part of a polynucleotide such as a messenger RNA molecule, or coding strand of DNA that specifies a particular amino acid, a starting or stopping signal for translation. Typically, codons are specific for one amino acid, however cases of a codon sharing at least one nucleotide with another codon are known for SARS-CoV-2.

The term “codon pair”, as used herein refers to two consecutive codons.

The term “codon-pair deoptimization” (CPD), as used herein, refers to a “reformulation” of codons or an exchange of codons by other codons encoding the same amino acid such that the encoded protein is the same, but suboptimal codon pairs and/or CpG dinucleotides emerge. Methods for codon-pair deoptimizations are known in the art (see, e.g., Coleman et al., 2008, Mueller et al., 2010). In some embodiments, the codon-pair deoptimization described herein comprises increasing the number of underrepresented or suboptimal codon pairs and CpG dinucleotides in recoded genomes. In some embodiments, the codon-pair deoptimization(s) described herein result(s) in increased mRNA decay and/or reduced translation efficiency. In some embodiments, the codon-pair deoptimization(s) described herein result(s) in less protein, less virus, reduced virulence, and/or a live-attenuated virus.

The polynucleotide of the invention may be used in virus production or in the context of a vaccine. As such, the polynucleotide of the invention emits a decreased risk of uncontrolled replication during production, transport, storage, processing and/or administration compared to the wild type sequence.

Typically, this polynucleotide forms part of a live attenuated SARS-CoV-2. In comparison to a wild type virus, a live attenuated virus provokes less and/or less severe or even no symptoms in a host organism after the host organism has been confronted (infected) with the attenuated virus. At the same time, the live attenuated virus induces an immune response of the host to the attenuated virus that is at least partially protective against a wild type virus infection and/or at least one symptom thereof.

In contrast to most of the vaccines under development, the inventors generated attenuated but replicating SARS-CoV-2 vaccine candidates by genetic modification of the SARS-CoV-2 genome via codon pair deoptimization (Coleman et al., 2008). CPD is a virus attenuation strategy which has enabled rapid and highly efficient attenuation of a wide variety of viruses (Broadbent et al., 2016; Coleman et al., 2008; Eschke et al., 2018; Groenke et al., 2020; Khedkar et al., 2018; Kunec and Osterrieder, 2016; Le Nouen et al., 2014; Mueller et al., 2010; Shen et al., 2015; Trimpert et al., 2021 a; Trimpert et al., 2021 b). CPD rearranges the positions of existing synonymous codons in one or more viral genes, without changing the codon bias or amino acid composition of the encoded protein (Eschke et al., 2018; Groenke et al., 2020; Khedkar et al., 2018; Kunec and Osterrieder, 2016; Osterrieder and Kunec, 2018; Trimpert et al., 2021 a; Trimpert et al., 2021 b). Naturally underrepresented codons can become overrepresented by CPD. Since the effect of CPD is highly dependent on the genome sequence to be deoptimized, no exact instructions can be given on how the position of codons in the recoded sequence should be changed during deoptimization. However, the person skilled in the art is aware how to identify or estimate codon pairs that can be replaced by naturally underrepresented codon pairs at the target site (e.g. in a target species or a target tissue) at which the polynucleotide is intended to be translated. The inventors provide herein examples which codon pairs to recode to achieve overrepresentation of naturally underrepresented codon pairs and codon-pair deoptimization of the polynucleotide. The skilled person could therefore - at least from the codon pairs at the target site and the means and methods provided herein - arrive at other polynucleotides according to the invention.

A polynucleotide is to be considered as codon-pair deoptimized if at least one codon pair is deoptimized with respect to the corresponding natural sequence. Recoded viruses typically do not produce proteins from the recoded genes as efficiently as their parents, and can show defects in reproductive fitness, which enables the host to control wild-type virus infection by innate and adaptive immune responses (Eschke et al., 2018; Groenke et al., 2020; Khedkar et al., 2018; Kunec and Osterrieder, 2016; Mueller et al., 2010; Trimpert et al., 2021 b; Wimmer et al., 2009). The conserved antigenic identity and replicative potential enable recoded

attenuated viruses to fully engage the immune system of the host and provoke strong immune responses.

Thus, by the codon-pair deoptimization, the resulting proteins are not altered. Rather, even though the genomic sequence of SARS-CoV-2 is altered, the resulting proteins remain the same. However, typically the efficiency of translation is reduced so that the virus replication is also reduced. This leads to an immune response when the live attenuated SARS-CoV-2 is used as vaccine without the risk of a pathologic virus replication in a patient having received the vaccine. Another possible effect of codon-pair deoptimization is a CpG mediated immune response leading to virus attenuation. The present disclosure is not limited to a specific one of these effects.

The live attenuated SARS-CoV-2 lacking the furin cleavage site (FCS) was severely attenuated but elicited a strong humoral immune response and maintained a similar level of protection in a heterologous SARS-CoV-2 challenge as the live attenuated virus variants with intact FCS and the parental wild-type SARS-CoV-2. Most importantly, however, removal of the FCS completely abolished transmission of vaccine virus between co-housed hamsters. These results indicate that removal of the FCS from live attenuated SARS-CoV-2 is a promising strategy to further increase vaccine safety and prevent vaccine transmission without compromising vaccine efficacy.

It is a very surprising finding that the vaccine efficacy is indeed not compromised by the modification, in particular deletion, of the FCS and the combined attenuating, but independent attenuating mechanisms of codon-pair deoptimization and modification, in particular deletion, of the FCS. In many cases, one can observe “over-attenuating” by combining different attenuating effects, leading to insufficient virus growth or insufficient immune response. Interestingly, such negative over-attenuating effects were not observed for the presently claimed subject matter. In contrast, the modification, in particular deletion, of the FCS results in an increased virus growth in vitro and in an increased genetic stability of the life attenuated SARS-CoV-2. The latter is particularly important with respect to regulatory requirements, since the higher the stability of the life attenuated SARS-CoV-2, the easier it is to obtain a marketing authorization for a vaccine comprising the life attenuated SARS-CoV-2 and the higher is the medical safety of such vaccine.

Unintentional spread of vaccine viruses from vaccinated to unvaccinated individuals is a factor that can complicate the use of transmissible live attenuated vaccines (LAV) (Bull et al., 2018; Layman et al., 2021 ; Nuismer et al., 2018; Pons-Salort et al., 2016). While self-dissemination is desirable in some scenarios, specifically when herd immunity is sought in wildlife (Smithson et al., 2019), uncontrolled circulation of vaccine viruses potentially increases the likelihood for reversion to virulence (Bull et al., 2018; Layman et al., 2021 ; Nuismer et al., 2018). Recombination between different vaccine viruses or vaccine and field viruses is especially problematic because it can give rise to recombinants with increased virulence, transmissibility or immune evasion capabilities (Burns et al., 2014; Combelas et al., 201 1 ; Lee et al., 2012; Ming et al., 2020). The rapid evolution of SARS-CoV-2 urges extra caution in the use of LAVs with respect to their potentially irrevocable circulation. Moreover, transmission of attenuated viruses to immunocompromised individuals is a danger that accompanies the use of transmissible virus vaccines (Kamboj and Sepkowitz, 2007).

The inventors already found that sCPD9 confers robust immunity against several SARS-CoV-2 variants in COVID-19 hamster models (Trimpert et al., 2021 a; Trimpert et al., 2021 b). Most importantly, sCPD9 outperformed intra-muscularly administered adenovirus vector and mRNA vaccines in its ability to induce systemic and mucosal immunity (Nouailles et al., 2022). It was a very surprising finding that this robust immunity against several SARS-CoV-2 variants is not compromised by additionally introducing the modification, in particular deletion, of the FCS into the polynucleotide encoding at least some proteins of the live attenuated SARS-CoV-2. Rather, the immune protection conferred by a vaccination with a live attenuated SARS-CoV-2 comprising the novel polynucleotide construct was equally good as the immune protection conferred by a vaccination with the live attenuated SARS-CoV-2 previously described (Trimpert et al., 2021 a; Trimpert et al., 2021 b), even though the biologic safety of the novel construct is significantly higher.

The entry of SARS-CoV-2 into host cells is mediated by its major surface protein, the spike protein. The spike protein initiates infection by binding to its cellular receptor, angiotensinconverting enzyme 2 (ACE2), and enables the actual cell entry through fusion between the viral envelope with host cell membranes. To enable infection, the spike protein must be activated by cellular proteases. Activation involves proteolytic cleavage of the spike protein at approximately the midpoint of the protein at the S1/S2 site, resulting in two subunits S1 and S2 that are held together by non-covalent interactions. Cleavage of the spike protein induces conformational changes that enables the S1 subunit to bind to ACE2 via its receptor binding domain and triggers the fusogenic activity of the membrane-anchored S2 subunit.

Unlike other closely related viruses, SARS-CoV-2 contains a unique polybasic cleavage motif (PRRAj.) at the S1/S2 site, termed furin cleavage site (FCS). While several enzymes can cleave the FCS, it is most efficiently cleaved by the cellular transmembrane protease serine 2 (TMPRSS2), a cell surface trypsin-like protease (Hoffmann and Pohlmann, 2021 ). The TMPRSS2 protease determines the entry pathway of the virus (Koch et al., 2021 ). When the host cell expresses TMPRSS2, the virus is activated at the cell surface and rapidly enters the cells via cell fusion in a pH-independent manner. In contrast, if TMPRSS2 is absent, the virus is endocytosed and the virus entry is mediated by cathepsin L, an endosomal/lysosomal cysteine protease.

To prevent transmission of the vaccine virus sCPD9, the inventors deleted the FCS from its spike protein. It has been shown in preclinical studies that removal of the FCS renders mutant viruses non-transmissible and strongly attenuated (Johnson et al., 2021 ; Lau et al., 2020; Peacock et al., 2021 ; Sasaki et al., 2021 a). However, since removal of the FCS can enhance viral attenuation, combining it with a different attenuation mutation may result in an overly attenuated virus with limited ability to induce strong immunity. Therefore, it is important to compare transmissibility and protective efficacy of LAV candidates that lack the FCS with those that have the intact FCS. On the other hand, if removal of the FCS does not compromise the protective effect of the LAV candidates, then removal of the FCS is desirable because it increases the safety of the LAV candidates by introducing a second and independent attenuating mutation into the viral genome.

In addition, aside from eliminating transmission and increasing vaccine safety, removal of the FCS has a potentially important practical advantage for the production of SARS-CoV-2 LAVs. During propagation in cells that do not express TMPRSS2 such as Vero cells, which are commonly used by vaccine manufacturers, SARS-CoV-2 variants lacking a functional FCS become rapidly dominant because they outcompete variants with an intact FCS (Davidson et al., 2020; Klimstra et al., 2020; Lau et al., 2020; Liu et al., 2020; Ogando et al., 2020; Sasaki et al., 2021 b; Wong et al., 2021 ). Consistent with these reports, the inventors found that sCPD9 also rapidly loses its FCS when propagated on cells that do not express TMPRSS2. In contrast, removal of the FCS increases the genetic stability of vaccine viruses during production, and also increases viral titers on TMPRSS2-deficient cell lines.

In an embodiment, the polynucleotide encodes the non-structural protein 7. In an embodiment, the polynucleotide encodes the non-structural protein 8. In an embodiment, the polynucleotide encodes the non-structural protein 9. In an embodiment, the polynucleotide encodes the non-structural protein 10. In an embodiment, the polynucleotide encodes the non-structural protein 11. In an embodiment, the polynucleotide encodes the non-structural protein 12. In an embodiment, the polynucleotide encodes the nsp15, the endonuclease. In an embodiment, the

polynucleotide encodes the nsp16, the 2’-O-methyltransferase. In an embodiment, the polynucleotide encodes the spike protein (sometimes also referred to as spike glycoprotein).

In an embodiment, the polynucleotide encodes the spike protein and at least one of the non-structural proteins.

In an embodiment, the polynucleotide encodes at least two of the non-structural proteins. To give an example, the polynucleotide encodes, in an embodiment, the endoribonuclease and the 2’-O-methyltransferase. To give another example, the polynucleotide encodes, in an embodiment, non-structural protein 7, non-structural protein 8, non-structural protein 9, non-structural protein 10, and non-structural protein 1 1.

In an embodiment, the furin cleavage site modification is a partial or full deletion of a furin cleavage site being naturally present in the SARS-CoV-2 genome.

In an embodiment, the furin cleavage site modification is a loss-of-function mutation of a furin cleavage site being naturally present in the SARS-CoV-2 genome.

In an embodiment, the furin cleavage site modification is an at least partial substitution of a furin cleavage site being naturally present in the SARS-CoV-2 genome.

In an embodiment, the SARS-CoV-2 genome is a genome section extending from position 11 ,000 to position 27,000 of the genome of SARS-CoV-2. For position numbering and definition of the terms “genome of SARS-CoV-2” and “wild type SARS-CoV-2 “, reference is made to the gene bank accession number MT108784.1 (freely accessible via the website https://www.ncbi.nlm.nih.gov/genbank/) that comprises 29,891 bases or nucleotides. The first of these bases or nucleotides (at the 5’ terminus) is positioned at position 1 . The last of these bases or nucleotides (at the 3’ terminus) is positioned at position 29,891 . The skilled person is aware of how to adjust the numbering of the referenced sequence to embodiments, wherein the SARS-CoV-2 genome understood as a sequence from a different SARS-CoV-2 variant. In some embodiments, the polynucleotide of the invention is a codon-pair deoptimized sequence of a sequence comprised in the SARS CoV-2 genome section from position 1 1 ,000 to position 24,000. In an embodiment, the genome section extends from position 11 ,500 to position 26,000, in particular from position 1 1 ,900 to position 25,500, in particular from position 1 1 ,950 to position 25,350, in particular from position 12,000 to position 24,000. In an embodiment, the genome section extends from position 11 ,950 to position 14,400. In an embodiment, the genome section extends from position 11 ,900 to position 13,500. In an embodiment, the

genome section extends from position 13,900 to position 14,400. In an embodiment, the genome section extends from position 20,300 to position 21 ,600. In an embodiment, the genome section extends from position 24,300 to position 25,400. These embodiments can be combined in any desired way.

In an embodiment, the at least one sequence part comprising codon-pair deoptimizations has a length lying in a range of from 750 nucleotides to 2500 nucleotides, in particular of from 800 nucleotides to 2400 nucleotides, in particular of from 900 nucleotides to 2300 nucleotides, in particular of from 999 nucleotides to 2200 nucleotides, in particular of from 1000 nucleotides to 2100 nucleotides, in particular of from 1100 nucleotides to 2000 nucleotides, in particular of from 1146 nucleotides to 1900 nucleotides, in particular of from 1200 nucleotides to 1836 nucleotides, in particular of from 1300 nucleotides to 1800 nucleotides, in particular of from 1400 nucleotides to 1700 nucleotides, in particular of from 1500 nucleotides to 1600 nucleotides.

In an embodiment, between 15 % and 40 %, in particular between 20 % and 35 %, in particular between 25 % and 30 % of the nucleotides of the at least one sequence part comprising codonpair deoptimizations are different from the nucleotides of a corresponding (wild type) SARS-CoV-2 genome. Such a wild-type SARS-CoV-2 genome is the genomic sequence of a non-artificially modified virus variant or lineage, such as lineages B.1.1.7 (Alpha), B.1.351 (Beta), B.1.1.28.1 (Gamma), B.1.617.2 (Delta), or B.1.159.1 (Omicron), including any sub-variants such as Omicron sub-variants BA.1 , BA.2, BA.3, BA.4, BA.5. It can also be denoted as authentic SARS-CoV-2 genome or authentic SARS-CoV-2 genomic sequence.

In an embodiment, between 200 and 500 nucleotides, in particular between 250 and 450 nucleotides, in particular between 300 and 400 nucleotides of the at least one sequence part comprising codon-pair deoptimizations are different from the (in particular identically positioned) nucleotides of a corresponding SARS-CoV-2 genome.

In an embodiment, between 40 % and 70 %, in particular between 45 % and 65 %, in particular between 50 % and 60 %, in particular between 55 % and 62 % of the codons of the at least one sequence part comprising codon-pair deoptimizations are different from the respective codons of a corresponding SARS-CoV-2 genome.

In an embodiment, between 150 and 400 codons, in particular between 200 and 350 codons, in particular between 250 and 300 codons of the at least one sequence part comprising codon- pair deoptimizations are different from the (in particular identically positioned) codons of a corresponding SARS-CoV-2 genome.

In an embodiment, the at least one sequence part comprising codon-pair deoptimizations comprises a first deoptimized sequence part and a second deoptimized sequence part. Both deoptimized sequence parts are separated from each other by a non-deoptimized sequence section comprising at least 300 nucleotides, e.g., 300 to 1000 nucleotides, in particular 400 to 900 nucleotides, in particular 500 to 800 nucleotides, in particular 600 to 700 nucleotides. By conserving a specific part of the RNA sequence and by deoptimizing flanking parts upstream and downstream of the conserved RNA sequence, a particularly high efficacy in attenuating the SARS-CoV-2 is achieved while maintaining its ability to replicate.

In an embodiment, the first deoptimized sequence part has a length lying in a range of from 1300 nucleotides to 1600 nucleotides, in particular of from 1400 nucleotides to 1500 nucleotides, in particular of from 1450 nucleotides to 1490 nucleotides. At the same time, the second deoptimized sequence part has a length lying in a range of from 100 nucleotides to 400 nucleotides, in particular of from 200 nucleotides to 300 nucleotides, in particular of from 350 nucleotides to 400 nucleotides. The length of the first deoptimized sequence part and of the second deoptimized sequence part is chosen such that other applicable restrictions (such as an overall length of the at least one sequence part comprising codon-pair deoptimizations of not more than 2000 nucleotides) are fulfilled, if desired. If the length of the at least one sequence part comprising codon-pair deoptimizations shall not exceed 2000 nucleotides, it is immediately apparent that only the lower threshold of 1300 nucleotides can be combined with the upper threshold of 400 nucleotides for the first and second deoptimized sequence parts to fulfil the restriction of the maximum length of the at least one sequence part comprising codonpair deoptimizations, considering that the first deoptimized sequence part and the second deoptimized sequence part are separated by at least 300 nucleotides of the authentic SARS-CoV-2 genome. At the same time, the upper threshold of 1600 nucleotides for the first deoptimized sequence part can be combined with the lower threshold of 100 nucleotides for the second deoptimized sequence part to fulfil a maximum length of 2000 nucleotides, considering the intermediate 300 non-recoded nucleotides.

In an embodiment, the first deoptimized sequence part is at least 95 %, in particular at least 96 %, in particular at least 97 %, in particular at least 98 %, in particular at least 99 %, in particular 100 % identical to SEQ ID NO. 2. At the same time, the second deoptimized sequence part is at least 95 %, in particular at least 96 %, in particular at least 97 %, in particular at least 98 %, in particular at least 99 %, in particular 100 % identical to SEQ ID NO.

4.

In an embodiment, the polynucleotide is at least 95 %, in particular at least 96 %, in particular at least 97 %, in particular at least 98 %, in particular at least 99 %, in particular 100 % identical to SEQ ID NO. 6.

In an embodiment, the polynucleotide is at least 95 %, in particular at least 96 %, in particular at least 97 %, in particular at least 98 %, in particular at least 99 %, in particular 100 % identical to SEQ ID NO. 8.

In an embodiment, the polynucleotide is at least 95 %, in particular at least 96 %, in particular at least 97 %, in particular at least 98 %, in particular at least 99 %, in particular 100 % identical to SEQ ID NO. 10.

In an embodiment, the polynucleotide is at least 95 %, in particular at least 96 %, in particular at least 97 %, in particular at least 98 %, in particular at least 99 %, in particular 100 % identical to SEQ ID NO. 15.

In an embodiment, the polynucleotide is at least 95 %, in particular at least 96 %, in particular at least 97 %, in particular at least 98 %, in particular at least 99 %, in particular 100 % identical to SEQ ID NO. 16.

In an embodiment, the polynucleotide is at least 95 %, in particular at least 96 %, in particular at least 97 %, in particular at least 98 %, in particular at least 99 %, in particular 100 % identical to SEQ ID NO. 17.

Even though an FCS deletion in the SARS-CoV-2 genome was described in prior art, one could not have assumed that such a deletion could have the effects observed by the present inventors when combining such deletion (or at least an FCS modification) with the specific codon-pair deoptimizations present in the above-mentioned sequences. It is rather surprising to learn that the immune protective properties of these codon-pair deoptimized sequences is not compromised by combining the codon-pair deoptimization with a modification, in particular deletion, of the FCS.

In an embodiment, the furin cleavage site modification comprises a deletion of the nucleotides encoding an amino acid sequence XRRA (i.e., a furin cleavage site), wherein X denotes P, R

or H. The following Table 1 lists specific embodiments of the presently claimed subject matter relating to different amino acid sequences missing in the expressed protein due to the deletion of the furin cleavage site. It should be noted that it is less relevant which nucleotides are excised from the SARS-CoV-2 genome by the deletion of the furin cleavage site, as long as the furin cleavage site is no longer present in the resulting protein.

Table 1 : Embodiments of missing amino acids due to the deletion of the furin cleavage site.

In an embodiment, the furin cleavage site modification, in particular deletion, effects that an expression of the polynucleotide results in a protein, in particular a spike protein, in which at least or exactly 5, in particular at least or exactly 6, in particular at least or exactly 7, in particular at least or exactly 8, in particular at least or exactly 9, in particular at least or exactly 10, in particular 5 to 10, in particular 6 to 9, in particular 7 to 8, consecutive amino acids of the naturally expressed protein are replaced by a single amino acid. This can be achieved by an off-reading-frame deletion of nucleotides. By such an off-reading-frame deletion, a single nucleotide of a first triplet is combined with two nucleotides of a second triplet to form a novel triplet that is not present in the natural SARS-CoV-2 genome at this genome position.

In an embodiment, the single amino acid that replaces the naturally expressed consecutive amino acids is an isoleucine.

In an embodiment, the furin cleavage site modification consists of or comprises a deletion of nucleic acid sequence as defined by SEQ ID NO. 18 or a nucleic acid sequence having at least 95 % sequence identity to SEQ ID NO. 18.

In an aspect, the present invention relates to a live attenuated SARS-CoV-2. This live attenuated SARS-CoV-2 comprises a partially recoded genomic RNA sequence, i.e., a partially recoded genomic viral sequence. The partially recoded genomic RNA sequence is a codonpair deoptimized sequence coding for a spike protein and/or a specific non-structural protein (nsp). The non-structural protein is chosen from the group consisting of non-structural protein 7, non-structural protein 8, non-structural protein 9, non-structural protein 10, non-structural protein 11 (a small, only 13 amino acids long protein), non-structural protein 12 (also referred to as RNA-dependent RNA polymerase), non-structural protein 15 (an endoribonuclease), and non-structural protein 16 (a 2’-O-methyltransferase) of the live attenuated SARS-CoV-2. The live attenuated SARS-CoV-2 further comprises a furin cleavage site modification resulting in a functional loss of a furin cleavage site. This furin cleavage site is naturally present in the SARS-CoV-2 genome. As a result, the spike protein of the live attenuated SARS-CoV-2 does not comprise a functional furin cleavage site, whereas non-engineered SARS-CoV-2 does comprise such furin cleavage site.

In an embodiment, the partially recoded genomic RNA sequence codes for the non-structural protein 12. In an embodiment, the partially recoded genomic sequence codes for the spike protein (sometimes also referred to as spike glycoprotein).

In an embodiment, the partially recoded genomic RNA sequence comprises at least two of the non-structural proteins. To give an example, the partially recoded genomic RNA sequence codes, in an embodiment, for the endoribonuclease and the 2’-O-methyltransferase. To give another example, the partially recoded genomic RNA sequence codes, in an embodiment, for non-structural protein 7, non-structural protein 8, non-structural protein 9, non-structural protein 10, and non-structural protein 1 1.

In an embodiment, the partially recoded genomic RNA sequence lies in a genome section extending from position 11 ,000 to position 27,000 of the genome of the live attenuated SARS-CoV-2. According to gene bank accession number MT108784.1 , the genome of the wild type SARS-CoV-2 comprises 29,891 bases or nucleotides. The genome of the live attenuated

SARS-CoV-2 has essentially a similar length. A difference is the length of a polyA tail at the 3’ terminus that was determined by sequencing to be eight adenine nucleotides longer than in case of the wild type SARS-CoV-2. It should be noted that some uncertainty remains by determining the length of a polyA tail by sequencing. Consequently, it is possible that the polyA tail in the wild type sequence is longer or shorter than indicated in the sequence according to gene bank accession number MT 108784.1 . Likewise, it is possible that the polyA tail in the live attenuated SARS-CoV-2 sequence is longer or shorter than presently determined. A further difference is that the live attenuated SARS-CoV-2 lacks at least 12 nucleotides that encode the furin cleavage site in the wild type SARS-CoV-2.

The first of the 29,891 bases or nucleotides of the genome of the wild type SARS-CoV-2 (at the 5’ terminus) is positioned at position 1 . The last of these bases or nucleotides (at the 3’ terminus) is positioned at position 29,891. In an embodiment, the genome section extends from position 1 1 ,500 to position 26,000, in particular from position 11 ,900 to position 25,500, in particular from position 1 1 ,950 to position 25,350, in particular from position 12,000 to position 24,000. In an embodiment, the genome section extends from position 1 1 ,950 to position 14,400. In an embodiment, the genome section extends from position 11 ,900 to position 13,500. In an embodiment, the genome section extends from position 13,900 to position 14,400. In an embodiment, the genome section extends from position 20,300 to position 21 ,600. In an embodiment, the genome section extends from position 24,300 to position 25,400. These embodiments can be combined in any desired way.

In an embodiment, the partially recoded genomic RNA sequence has a length lying in a range of from 750 nucleotides to 2500 nucleotides, in particular of from 800 nucleotides to 2400 nucleotides, in particular of from 900 nucleotides to 2300 nucleotides, in particular of from 999 nucleotides to 2200 nucleotides, in particular of from 1000 nucleotides to 2100 nucleotides, in particular of from 1 100 nucleotides to 2000 nucleotides, in particular of from 1146 nucleotides to 1900 nucleotides, in particular of from 1200 nucleotides to 1836 nucleotides, in particular of from 1300 nucleotides to 1800 nucleotides, in particular of from 1400 nucleotides to 1700 nucleotides, in particular of from 1500 nucleotides to 1600 nucleotides.

In an embodiment, between 15 % and 40 %, in particular between 20 % and 35 %, in particular between 25 % and 30 % of the nucleotides of the partially recoded genomic RNA sequence are different from the nucleotides of a corresponding wild-type genomic RNA sequence. Such a wild-type genomic RNA sequence is the genomic viral sequence of a non-artif icially modified virus variant or lineage, such as lineages B.1 .1 .7 (Alpha), B.1 .351 (Beta), B.1 .1 .28.1 (Gamma), B.1.617.2 (Delta), or B.1.159.1 (Omicron). It can also be denoted as authentic SARS-CoV-2 genomic RNA sequence.

In an embodiment, between 200 and 500 nucleotides, in particular between 250 and 450 nucleotides, in particular between 300 and 400 nucleotides of the partially recoded genomic RNA sequence are different from the identically positioned nucleotides of a corresponding wildtype virus genomic RNA sequence.

In an embodiment, between 40 % and 70 %, in particular between 45 % and 65 %, in particular between 50 % and 60 %, in particular between 55 % and 62 % of the codons (i.e., three nucleotides in each case that code for a specific amino acid) of the partially recoded genomic RNA sequence are different from the respective codons of a corresponding wild-type virus genomic RNA sequence.

In an embodiment, between 150 and 400 codons, in particular between 200 and 350 codons, in particular between 250 and 300 codons of the partially recoded genomic RNA sequence are different from the identically positioned codons of a corresponding wild-type virus genomic RNA sequence.

In an embodiment, the partially recoded genomic RNA sequence comprises a first recoded part and a second recoded part. Both recoded parts are separated from each other by a nonrecoded genome section comprising at least 300 nucleotides, e.g., 300 to 1000 nucleotides, in particular 400 to 900 nucleotides, in particular 500 to 800 nucleotides, in particular 600 to 700 nucleotides. By conserving a specific part of the RNA sequence and by recoding flanking parts upstream and downstream of the conserved RNA sequence, a particularly high efficacy in attenuating the SARS-CoV-2 while maintaining its general viability is achieved.

In an embodiment, the first recoded part has a length lying in a range of from 1300 nucleotides to 1600 nucleotides, in particular of from 1400 nucleotides to 1500 nucleotides, in particular of from 1450 nucleotides to 1490 nucleotides. At the same time, the second recoded part has a length lying in a range of from 100 nucleotides to 400 nucleotides, in particular of from 200 nucleotides to 300 nucleotides, in particular of from 350 nucleotides to 400 nucleotides. The length of the first recoded part and of the second recoded part is chosen such that other applicable restrictions (such as an overall length of the recoded genomic RNA sequence of not more than 2000 nucleotides) are fulfilled, if desired. If the length of the partially recoded genomic RNA sequence shall not exceed 2000 nucleotides, it is immediately apparent that only the lower threshold of 1300 nucleotides can be combined with the upper threshold of 400 nucleotides for the first and second recoded parts to fulfil the restriction of the maximum length of the partially recoded genomic RNA sequence, considering that the first recoded part and the second recoded part are separated by at least 300 nucleotides of the authentic SARS-CoV-2 genomic sequence. At the same time, the upper threshold of 1600 nucleotides for the first recoded part can be combined with the lower threshold of 100 nucleotides for the second recoded part to fulfil a maximum length of 2000 nucleotides, considering the intermediate 300 non-recoded nucleotides.

In an embodiment, the first recoded part is at least 95 %, in particular at least 96 %, in particular at least 97 %, in particular at least 98 %, in particular at least 99 %, in particular 100 % identical to SEQ ID NO. 2. At the same time, the second recoded part is at least 95 %, in particular at least 96 %, in particular at least 97 %, in particular at least 98 %, in particular at least 99 %, in particular 100 % identical to SEQ ID NO. 4.

In an embodiment, the partially recoded genomic RNA sequence is at least 95 %, in particular at least 96 %, in particular at least 97 %, in particular at least 98 %, in particular at least 99 %, in particular 100 % identical to SEQ ID NO. 6.

In an embodiment, the partially recoded genomic RNA sequence is at least 95 %, in particular at least 96 %, in particular at least 97 %, in particular at least 98 %, in particular at least 99 %, in particular 100 % identical to SEQ ID NO. 8.

In an embodiment, the partially recoded genomic RNA sequence is at least 95 %, in particular at least 96 %, in particular at least 97 %, in particular at least 98 %, in particular at least 99 %, in particular 100 % identical to SEQ ID NO. 10.

The phrase “being % identical” or “having percent (%) sequence identity” with respect to a reference sequence is defined as the percentage of nucleotides or amino acid residues in a candidate sequence that are identical with the nucleotides or amino acid residues in the reference sequence, after aligning the sequences and introducing gaps, if necessary, to achieve the maximum percent sequence identity, and not considering any conservative substitutions as part of the sequence identity. Alignment for purposes of determining percent amino acid sequence identity can be achieved in various ways that are within the skill in the art, for instance, using publicly available computer software such as BLAST, BLAST-2, ALIGN or Megalign (DNASTAR) software. Those skilled in the art can determine appropriate parameters for aligning sequences, including any algorithms needed to achieve maximal alignment over the full length of the sequences being compared.

The skilled person is aware that sequences as exemplified herein can be altered to a certain percentage without or without substantially altering biologic functions such as attenuation or the encoded protein. Therefore, the skilled person using means and methods described herein can amend codon pairs according to the teaching of the invention. E.g., codon pairs can be replaced with synonymous versions that are similarly attenuating and/or naturally underrepresented as the deoptimized codon pairs of the sequences described herein. Thus, a similar degree of deoptimization may be achieved by replacing codon pairs in the position ranges described herein.

The invention is based at least in part on the finding that codon deoptimizations in the position ranges described herein are particularly useful for attenuating the SARS-CoV-2 virus while sufficiently maintaining immunogenicity and the ability of the virus to replicate. It is further based on the finding that a furin cleavage site modification, which results in a functional loss of a furin cleavage site being naturally present in the SARS-CoV-2 genome, further attenuates the SARS-CoV-2, while fully maintaining its immunogenicity.

The individual recoded sequences and their position in the SARS-CoV-2 genome are summarized in the following Table 2.

Table 2: Summary of recoded RNA sequences.

pp = polyprotein

nsp = non-structural protein

It should be noted that WT6A/CPD6A encodes only nine nucleotides (three amino acids) of nsp12 (RNA-dependent RNA polymerase, RdRp). The majority of this protein is encoded by WT6B/CPD6B.

The middle part of fragment WT6 was not recoded in fragment CPD6, because it contains a conserved regulatory RNA sequence that is essential for virus replication. Fragment WT6 encodes a -1 ribosomal frameshifting element, a so-called RNA pseudoknot structure. This structure promotes a ribosomal frameshifting, a process during which the reading frame of translation is changed at the junction between open reading frames (ORFs) 1 a and 1 b. During this process, a single nucleotide of the slippery sequence that is located downstream from the RNA pseudoknot structure is read twice by the translating ribosome, and the reading frame is shifted by -1 nucleotide (the ribosome slips one nucleotide backwards at the slippery sequence). Often, translation of ORF1 a terminates at the stop codon of ORF1 a. However, when -1 ribosomal frameshifting occurs, translation of ORF1 a continues directly to ORF1 b and polyprotein (pp) 1 ab is produced. Thus, the CPD6A sequence is translated into both polyproteins 1 a and 1 ab; but the CPD6B sequence is translated only into polyprotein 1 ab.

Claims

1 . A polynucleotide encoding

a) severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike protein; and

b) optionally at least one non-structural SARS-CoV-2 protein selected from the group consisting of non-structural protein 7, non-structural protein 8, non-structural protein 9, non-structural protein 10, non-structural protein 11 , non-structural protein 12, an endoribonuclease, and a 2'-O-methyltransferase,

wherein the polynucleotide comprises at least one sequence part comprising codon-pair deoptimizations in comparison to the SARS-CoV-2 genome,

wherein the polynucleotide further comprises a furin cleavage site modification, wherein the furin cleavage site modification results in a loss of a furin cleavage site being naturally present in the SARS-CoV-2 genome.

2. The polynucleotide of claim 1 , wherein the furin cleavage site modification is an at least partial deletion of a furin cleavage site being naturally present in the SARS-CoV-2 genome.

3. The polynucleotide of any one of claims 1 and 2, wherein the polynucleotide comprises a nucleic acid sequence as defined by SEQ ID NO. 6 or a nucleic acid sequence having at least 95 % sequence identity to SEQ ID NO. 6.

4. The polynucleotide of any one of claims 1 and 2, wherein the polynucleotide comprises a nucleic acid sequence as defined by SEQ ID NO. 8 or a nucleic acid sequence having at least 95 % sequence identity to SEQ ID NO. 8.

5. The polynucleotide of any one of claims 1 and 2, wherein the polynucleotide comprises a nucleic acid sequence as defined by SEQ ID NO. 10, a nucleic acid sequence having at least 95 % sequence identity to SEQ ID NO. 10, a nucleic acid sequence as defined by SEQ ID NO. 15, a nucleic acid sequence having at least 95 % sequence identity to SEQ ID NO. 15, a nucleic acid sequence as defined by SEQ ID NO. 16, a nucleic acid sequence having at least 95 % sequence identity to SEQ ID NO. 16, a nucleic acid sequence as defined by SEQ ID NO. 17, or a nucleic acid sequence having at least 95 % sequence identity to SEQ ID NO. 17.

6. The polynucleotide of any one of the preceding claims, wherein the furin cleavage site modification comprises a deletion of the nucleotides encoding an amino acid sequence XRRA, wherein X denotes P, R or H.

7. The polynucleotide of any one of the preceding claims, wherein the furin cleavage site modification effects that an expression of the polynucleotide results in a protein in which at least 5 consecutive amino acids of the naturally expressed protein are replaced by a single amino acid.

8. The polynucleotide of any one of the preceding claims, wherein the furin cleavage site modification consists of or comprises a deletion of a nucleic acid sequence as defined by SEQ ID NO. 18 or a nucleic acid sequence having at least 95 % sequence identity to SEQ ID NO. 18.

9. A live attenuated severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) comprising the polynucleotide according to any one of claims 1 to 8.

10. The live attenuated SARS-CoV-2 according to claim 9, wherein the SARS-CoV-2 has a nucleic acid sequence as defined by SEQ ID NO. 19, a nucleic acid sequence having at least 98 % sequence identity to SEQ ID NO. 19, a nucleic acid sequence as defined by SEQ ID NO. 20, or a nucleic acid sequence having at least 98 % sequence identity to SEQ ID NO. 20.

11 . A pharmaceutical composition comprising the live attenuated SARS-CoV-2 according to any one of claims 9 to 10.

12. Pharmaceutical composition according to claim 11 for use as vaccine.

13. A vector comprising the polynucleotide according to any one of claims 1 to 8.

14. A host cell comprising the polynucleotide according to any one of claims 1 to 8.

15. A method for production of a virus, the method comprising the steps of:

a) culturing a host cell according to claim 14; and

b) isolating a virus, wherein the virus is a live attenuated SARS-CoV-2.