The present invention relates to recombinant YjgF/YER057c/UKl 14 RID proteins (YjgF/YER057c/UKl 14) in stable polymer form for use in the treatment, diagnosis and prevention of malignant solid and systemic tumours.

Prior art

The family of proteins called YjgF/YER057c/UKl 14, which are highly conserved in various prokaryotic and eukaryotic organisms, has formed the subject of numerous studies designed to investigate the structural and functional characteristics of said proteins (Bartorelli et al., J. Tumor and Marker Oncology, 1994, 9, 37; Lambrecht JA et al., J.Biol. Chem., 2010, 285, 34401-34407; Lambrecht JA et al, J.Biol. Chem. 2012, 287, 3454- 3461; Mistiniene E et al, Bioconjugate Chem. 2003, 14, 1243-1252; Dhawan L et al, Mol. Cell. Biol, 2012, 32, 3768-3775; Flynn JM et al, Mol. Microbiol, 2013, 89(4)751- 759; Ernst DC et al, J. Bacteriology, 2014, 196 (18), 3335-3342; Niehaus TD et al, BMC Genomics, 2015, 16, 382). UK114 belongs to the YjgF/YER057c/UK 114 family. The common denominator of members of said family is their biological activity, especially the deaminase activity of 2-amino-acrylate (see Lambrect et al, 2012). Said activity defines all members of the YjgF/YER057c/UK 114 family with the acronym RID.

WO 9602567 describes the antitumoral activity of a perchloric acid soluble goat liver extract (called UK 101) containing protein UK 114. Said protein has a molecular weight of 14.2 kDa and the sequence of 137 amino acids as reported in Ceciliani F. et al, FEBS Letters, 1996, 393, 147-150.

The native protein UK 114, isolated from various species and/or proteins cross- reacting with it, is only present in the cytoplasm of normal cells, whereas various percentages thereof are also present in the cell membrane of malignant solid systemic tumour cells (Bartorelli A, et al, Int J Oncol. 1996 Mar;8(3):543-8.; Bussolati G, et al, Int J Oncol. 1997 Apr; 10(4):779-85). In adenocarcinomas, it is expressed in over 80% of tumours. E. coli expression of protein UK 114 in recombinant form was described in WO 0063368 and in Colombo I et al.( Biochim. Biophys. Acta, 1998, 1442, 49-59) and Burman JD et al.,(BMC Structural Biology, 2007, 7, 30), but, unlike WO 0063368, which reported a dimeric structure, Deriu D et al.,( Acta Cryst., 2003, D59, 1676-1678), reported that protein UK 114 in crystalline form has a trimeric configuration,. It is known from the literature that protein UK114 with a monomeric structure (Oka et al, 1995 J.Biol. Chem. 270:30060-30067) can present in an unstable dimeric or trimeric form. Experimentally, various physicochemical conditions, especially various ion and protein denaturation concentrations, can cause the dimeric or trimeric protein UK114 to take on a monomeric conformation, as demonstrated by chromatographic and mass spectrometry analysis. According to studies by Mistiniene et al, recombinant protein UK114 (obtained from E. coli) appears in monomeric form under conditions of slight denaturation induced by KC1 3M or 3M guanidine hydrochloride, whereas in 0.1M NaCl solution, it is in trimeric form. This variability of results is due to the tendency of the trimer to dissociate into a monomer under various conditions of instability.

In any event, no immunogenic activity of the form considered trimeric has been described, while studies conducted by the Applicant demonstrate that the recombinant protein obtained from monomeric E. coli does not produce any antibody response when injected into xenogeneic mammals. The recombinant proteins disclosed in Colombo I. et al.,( Biochim. Biophys. Acta, 1998, 1442, 49-59) and in WO 00/63368, described as dimers and/or trimers, actually dissociate rapidly into monomers under both reducing and non-reducing conditions, and are therefore unable to induce an antibody response.

WO 00/63368 describes a recombinant form of UK114 which differs from the natural form by a single amino acid. The absence of tags in the form claimed implies various purification steps, as described in the examples of WO 00/63368. The protein claimed is declared to be in dimeric form, but in SDS totally dissociates to a monomer, whereas under non-reducing conditions, only 90%> thereof dissociates. The form described by Colombo I. et al. in Biochimica et Biophysica Acta presents 3 more amino acids (VPM) deriving from cloning at the N-terminal end, and also dissociates into a monomer. When injected into xenogeneic mammals, neither of said proteins produces a cytotoxic immune response. WO 00/63368 and Colombo et al. also state that the protein is immunologically recognised by anti-UKlOl extraction antibodies, which is obvious, because the monomeric protein is almost a hapten (molecular weight 14.2 kD), which is unable to induce a good immune response apart from a weak IgG response, but is recognised by antibodies targeted against its antigenic determinants, because it binds in Western Blot to the anti-native UK114 (extraction) antibodies produced by immunising with extract UKlOl which (in addition to molecule UK114, most of which dissociates into a monomer) also contains a carrier (a 50 kD glycoprotein). However, when these recombinant proteins are injected into mammals, they do not produce a cytotoxic immune response, because although they are considered to be a dimer and/or trimer, they are rapidly converted into a monomer, which is a hapten, and therefore not immunogenic alone.

Description of the invention

It has now been discovered that if recombinant YjgF/YERO 57c7/UK 114 RID protein in stable polymer form is injected into xenogeneic animals, it elicits a strong, long-lasting, cytotoxic antibody response against identical and/or cross-reacting antigens expressed on malignant solid and systemic tumour cells.

The proteins according to the invention are therefore useful in the treatment, diagnosis and prevention of malignant solid and systemic tumours.

Analysis of the identity of UK114 present in various species demonstrated that identity with Homo sapiens (taken as 100%) ranges from 99.3% in Macaca mulatta to 45.6% in Heterocephalus glaber.

Identity values lower than 86% and higher than 60% are suitable for the purpose of the invention, as they are sufficiently xenogeneic to stimulate the antibody response but have a good identity able to produce a polyclonal antibody that recognises the determinants expressed on the tumour cell membrane. TABLE 1 shows the identity percentages of said UK 114 proteins compared with human UK114. Analysis of the identity of UK114 present in various species demonstrated that identity with Homo sapiens (taken as 100%) ranges from 99.3% in Macaca mulatta to 45.6%> in Heterocephalus glaber.

TABLE 1 BLAST SEARCH

Entry Protein names Identity

P 52758 Ribonuclease UK114 (Homo sapiens) 100.0%

Heat-responsive protein 12, isoform CRA a (Homo

A0A024R9H2

sapiens) 100.0%

H9FNK5 Ribonuclease UK114 (Macaca mulatta) 99.3%

H2QWH5 Heat-responsive protein 12 (Pan troglodytes) 99.3%

H2PQW3 Uncharacterised protein (Pongo abelii) 99.3%

G7PCA5 Ribonuclease UK114 (Macaca fascicularis) 99.3%

G3QFT9 Uncharacterised protein (Gorilla gorilla gorilla) 99.3%

G1QQ41 Uncharacterised protein (Nomascus leucogenys) 99.3%

G7MZU3 Ribonuclease UK114 (Macaca mulatta) 97.8%

U3FHY9 Ribonuclease UK114 (Callithrix jacchus) 95.6%

G7PHN9 Ribonuclease UK114 (Macaca fascicularis) 96.3%

F6VKH4 Uncharacterised protein (Macaca mulatta) 94.9%

M3VY13 Uncharacterised protein (Felis catus) 92.0%

G1L170 Uncharacterised protein (Ailuropoda melanoleuca) 91.2%

F7A6J6 Ribonuclease UK114-like protein (Equus caballus) 92.0%

E2QS94 Uncharacterised protein (Canis familiaris) 92.0%

M3YUY4 Uncharacterised protein (Mustela putorius furo) 89.8%

F6VML1 Uncharacterised protein (Callithrix jacchus) 92.7%

G9K4S5 Heat-responsive protein 12 (Mustela putorius furo) 89.7%

H0WIN4 Uncharacterised protein (Otolemur garnettii) 93.3%

Putative ribonuclease ukll4 isoform 1 (Desmodus

K9IWF2

rotundus) 89.1%

L8I1S8 Ribonuclease UK114 (Bos mutus) 86.9%

Q3T114 Ribonuclease UK114 (Bos taurus) 86.1%

F1S0M2 Uncharacterised protein (Sus scrofa) 87.6%

P80601 Ribonucase UK114 (Capra hircus) 85.4%

P 52759 Ribonuclease UK114 (Rattus norvegicus) 87.6%

W5P8I3 Uncharacterised protein (Ovis aries) 84.7%

G1Q9C4 Uncharacterised protein (Myotis lucifugus) 85.9%

F7B859 Uncharacterised protein (Monodelphis domestica) 88.5%

G3SPS4 Uncharacterised protein (Loxodonta africana) 83.2%

G1U8R2 Uncharacterised protein (Oryctolagus cuniculus) 82.5%

H 3 AW 97 Uncharacterised protein (Latimeria chalumnae) 80.1%

F7H790 Uncharacterised protein (Macaca mulatta) 99.1% Entry Protein names Identity

H0YB34 Ribonuclease UK114 (Homo sapiens) 100.0%

F6ZTJ8 Uncharacterised protein (Ornithorhynchus anatinus) 86.7%

P 52760 Ribonuclease UK114 (Mus musculus) 81.5%

G5B8Y4 Ribonuclease UK114 (Heterocephalus glaber) 80.7%

Q6NTT0 MGC82310 protein (Xenopus laevis) 77.2%

Putative ribonuclease ukll4 isoform 1 (Desmodus

K9IFZ3

rotundus) 81.0%

Q28GJ3 Heat-responsive protein 12 (Xenopus tropicalis) 77.2%

L8HUS2 Uncharacterised protein (Bos mutus) 80.0%

W5MUF4 Uncharacterised protein (Lepisosteus oculatus) 76.5%

Q5M8W7 LOC496671 protein (Xenopus tropicalis) 77.6%

H0ZP41 Uncharacterised protein (Taeniopygia guttata) 77.0%

U3JT39 Uncharacterised protein (Ficedula albicollis) 75.0%

G5E3D7 Putative uncharacterised protein (Hymenochirus curtipes) 74.4%

W5KVI7 Uncharacterised protein (Astyanax mexicanus) 76.5%

I3KHW6 Uncharacterised protein (Oreochromis niloticus) 74.3%

E3TEU7 Ribonuclease ukll4 (Ictalurus punctatus) 75.7%

K4G4Q6 Heat-responsive protein 12 (Callorhynchus milii) 71.5%

S9XJZ3 Ribonuclease UK114 (Camelus ferus) 88.9%

A0A098LZJ9 Heat-responsive protein 12 (Hypsiglena sp. JMG-2014) 74.3%

I3MHC8 Uncharacterised protein (Spermophilus tridecemlineatus) 89.7%

H3D6Z5 Uncharacterised protein (Tetraodon nigroviridis) 74.8%

Chromosome 8 SCAF14759, whole genome shotgun

Q4S2X6 74.8% sequence (Tetraodon nigroviridis)

G3NHG8 Uncharacterised protein (Gasterosteus aculeatus) 75.6%

C3KK60 Ribonuclease UK114 (Anoplopoma fimbria) 75.2%

E1C7E1 Uncharacterised protein (G alius gallus) 72.7%

C3KJM3 Ribonuclease UK114 (Anoplopoma fimbria) 72.1%

HOXUHl Uncharacterised protein (Otolemur garnettii) 81.5%

C3KH08 Ribonuclease UK114 (Anoplopoma fimbria) 70.6%

I3J3B1 Uncharacterised protein (Oreochromis niloticus) 73.7%

T1DKW2 Heat-responsive protein 12 (Crotalus horridus) 69.9%

A0A087Y9U8 Uncharacterised protein (Poecilia formosa) 70.7%

A0A087Y3T2 Uncharacterised protein (Poecilia formosa) 71.5%

Q6AXL2 Uncharacterised protein (Danio rerio) 72.1%

H2LCD2 Uncharacterised protein (Oryzias latipes) 72.1%

J3S8W6 Heat-responsive protein 12 (Crotalus adamanteus) 69.1%

B9EM65 Ribonuclease UK114 (Salmo salar) 70.6%

A0A091SB34 Ribonuclease UK114 (Mesitornis unicolor) 81.7%

A0A091Q1X3 Ribonuclease UK114 (Leptosomus discolor) 81.7%

U3J3U4 Uncharacterised protein (Anas platyrhynchos) 80.2%

A0A093P527 Ribonuclease UK114 (Pygoscelis adeliae) 80.0%

A0A091 WBZ6 Ribonuclease UK114 (Opisthocomus hoazin) 80.0%

A0A091 VG32 Ribonuclease UK114 (Nipponia nippon) 80.0%

A0A091J6T7 Ribonuclease UK114 (Egretta garzetta) 80.0%

A0A091FU34 Ribonuclease UK114 (Cuculus canorus) 80.0%

I3J3B0 Uncharacterised protein (Oreochromis niloticus) 70.1%

K4FU80 Ribonuclease (Callorhynchus milii) 67.9% Entry Protein names Identity

A0A087VAG6 Ribonuclease UK114 (Balearica regulorum gibberic.) 80.0%

A0A093G8Q8 Ribonuclease UK114 (Picoides pubescens) 79.8%

A0A093SE99 Ribonuclease UK114 (Manacus vitellinus) 78.3%

A0A091MX96 Ribonuclease UK114 (Cariama cristata) 80.0%

A0A091HE47 Ribonuclease UK114 (Buceros rhinoceros silvestris) 79.1%

A0A087QUR5 Ribonuclease UK114 (Aptenodytes forsteri) 79.1%

H0VRQ6 Uncharacterised protein (Cavia porcellus) 82.1%

E3TDF8 Ribonuclease ukll4 (Ictalurus furcatus) 70.6%

A0A0A0A VW1 Ribonuclease UK114 (Charadrius vociferus) 79.1%

A0A091KIM5 Ribonuclease UK114 (Chlamydotis macqueenii) 78.3%

K4GC04 Ribonuclease (Callorhynchus milii) 67.2%

A0A093C3S1 Ribonuclease UK114 (Tauraco erythrolophus) 79.5%

G3VD69 Uncharacterised protein (Sarcophilus harrisii) 86.1%

A0A091EJL2 Ribonuclease UK114 (Corvus brachyrhynchos) 77.4%

K4GBJ3 Ribonuclease (Callorhynchus milii) 67.2%

A0A096MBS1 Uncharacterised protein (Poecilia formosa) 68.4%

A0A093HH07 Ribonuclease UK114 (Struthio camelus australis) 79.3%

M3ZNB1 Uncharacterised protein (Xiphophorus maculatus) 68.4%

A0A094NNG1 Ribonuclease UK114 (Podiceps cristatus) 77.4%

F7GP86 Uncharacterised protein (Callithrix jacchus) 79.7%

C1BLC3 Ribonuclease UK114 (Osmerus mordax) 68.6%

G3SJQ3 Uncharacterised protein (Gorilla gorilla gorilla) 76.9%

A0A091IGZ5 Ribonuclease UK114 (Calypte anna) 76.4%

Putative uncharacterised protein (Ailuropoda

D2GY46 90.9% melanoleuca)

K7E4Y7 Uncharacterised protein (Monodelphis domestica) 73.3%

H2UMU8 Uncharacterised protein (Takifugu rubripes) 67.7%

F6USW0 Uncharacterised protein (Canis familiaris) 89.9%

U6E0W1 Ribonuclease UK114 (Neovison vison) 89.8%

R0JFK7 Ribonuclease UK114 (Anas platyrhynchos) 79.2%

Q3B7E4 Uncharacterised protein (Danio rerio) 66.4%

H2TRV4 Uncharacterised protein (Takifugu rubripes) 74.6%

Q6DGX0 Zgc: 92739 (Danio rerio) 66.4%

G1TDL7 Uncharacterised protein (Oryctolagus cuniculus) 68.2%

C0H8I4 Ribonuclease UK114 (Salmo salar) 67.5%

B5XA58 Ribonuclease UK114 (Salmo salar) 67.5%

I3N7V1 Uncharacterised protein (Spermophilus tridecemlineatus) 69.4%

A0A099ZWV1 Ribonuclease UK114 (Tinamus guttatus) 81.4%

Chromosome 21 SCAF14577, whole genome shotgun

Q4SJ46 58.6% sequence (Tetraodon nigroviridis)

H0YBX3 Ribonuclease UK114 (Homo sapiens) 100.0%

G1KS26 Uncharacterised protein (Anolis carolinensis) 67.0%

G3RRV5 Uncharacterised protein (Gorilla gorilla gorilla) 68.6%

K7G3G2 Uncharacterised protein (Pelodiscus sinensis) 70.1%

S4R576 Uncharacterised protein (Petromyzon marinus) 63.2%

G3UJF8 Uncharacterised protein (Loxodonta africana) 71.7%

A0A093HP53 Ribonuclease UK114 (Gavia stellata) 79.8%

G3VGS7 Uncharacterised protein (Sarcophilus harrisii) 58.5% Entry Protein names Identity

B9EQ43 Ribonuclease UK114 (Salmo salar) 62.4%

M3ZJS5 Uncharacterised protein (Xiphophorus maculatus) 72.4%

A0A091SVV5 Ribonuclease UK114 (Nestor notabilis) 63.2%>

V8P2W9 Ribonuclease (Ophiophagus hannah) 50.7%

F6VVM7 Uncharacterised protein (Ornithorhynchus anatinus) 77.9%

Oncorhynchus mykiss genomic scaffold, scaffold_20632

A0A060YX21 71.9%

(Oncorhynchus mykiss)

G5AL28 Ribonuclease UK114 (Heterocephalus glaber) 45.6%

G1NHU5 Uncharacterised protein (Meleagris gallopavo) 73.2%

H2UH67 Uncharacterised protein (Takifugu rubripes) 34.2%

V9L685 ΑΊΡ -binding domain 4 (Callorhynchus milii) 30.0%

W5N319 Uncharacterised protein (Lepisosteus oculatus) 27.9%

W5N314 Uncharacterised protein (Lepisosteus oculatus) 27.9%

U3JT08 Uncharacterised protein (Ficedula albicollis) 34.5%

ATP-binding domain-containing protein 4 (Callorhynchus

V9KG05 30.0% milii)

G3P8F9 Uncharacterised protein (Gasterosteus aculeatus) 30.3%

W5L5R2 Uncharacterised protein (Astyanax mexicanus) 26.0%

F1P550 Uncharacterised protein (G alius gallus) 36.8%

A0A096LQQ1 Uncharacterised protein (Poecilia formosa) 29.4%

P83145 Ribonuclease UK114 (Gallus gallus) 53.7%

R4G926 Uncharacterised protein (Anolis carolinensis) 26.6%

I3IYS6 Uncharacterised protein (Oreochromis niloticus) 25.4%

Q6GPL5 MGC83562 protein (Xenopus laevis) 47.3%

A0A087YAW6 Uncharacterised protein (Poecilia formosa) 26.2%

F6XPS3 Uncharacterised protein (Xenopus tropicalis) 47.3%

M4A570 Uncharacterised protein (Xiphophorus maculatus) 26.7%

E5RIP8 Ribonuclease UK114 (Homo sapiens) 100.0%

H3BDI6 Uncharacterised protein (Latimeria chalumnae) 25.7%

E7F9Q1 Diphthine— ammonia ligase (Danio rerio) 25.0%

Oncorhynchus mykiss genomic scaffold, scaffold_38082

A0A060ZCJ4 24.4%

(Oncorhynchus mykiss)

R0KDP7 Uncharacterised protein (Anas platyrhynchos) 40.5%

G1NI69 Uncharacterised protein (Meleagris gallopavo) 24.0%

F7FWD1 Uncharacterised protein (Monodelphis domestica) 22.4%

The stable polymeric recombinant YjgF/YER057c/UK 114 RID proteins according to the invention (hereinafter called PRP14) comprise three or more monomer units each having a sequence of 137 amino acids, and therefore have a molecular weight of about 42.6 or more.

A preferred sequence of the monomer unit is the sequence SEQ ID 1 :

GSHM S SL VRRII ST AKAP AAIGP YS Q AVL VDRTI YI S GQLGMDP AS GQL VP GGVVEEAKQALTNIGEILKAAGCDFTNVVKATVLLADINDFSAVNDVYKQYFQS SFPARAAYQVAALPKGGRVEIEAIAVQGPLTTASL.

Said sequence, shown here as an example, deriving from Capra hircus and expressed by suitable vectors in Escherichia coli as described below, is characterised by four more amino acids at the N-terminal end than the above-mentioned sequences. Said additional amino acids are G-S-H-M (glycine-serine-histidine-methionine). It has been found that due to the presence of said four additional amino acids at the N-terminal end, the polymeric forms consisting of clusters of more than one 14.2 kDA monomer unit are more stable than the dimeric or trimeric form of the native or recombinant YjgF/YERO 57c7/UK 114 proteins previously described. In fact, while the known proteins, in aqueous solution, give rise to unstable dimeric or trimeric forms that dissociate rapidly into the monomer, the sequence SEQ ID 1 gives rise to polymeric (dimeric and trimeric) forms which do not dissociate, thus maintaining their immunogenic capacity.

In fact, only a stable polymeric form leads to a 3D conformation responsible for the RID functional activity proper of the proteins of the YjgF/YER057c/UKl 14 family (Lambrecht JA et al, J.Biol. Chem. 2012, 287, 3454-3461; Mistiniene E et al, Bioconjugate Chem. 2003, 14, 1243-1252; Dhawan L et al, Mol. Cell. Biol, 2012, 32, 3768-3775; Flynn JM et al, Mol. Microbiol, 2013, 89(4)751-759; Ernst DC et al, J. Bacteriology, 2014, 196 (18), 3335-3342; Niehaus TD et al, BMC Genomics, 2015, 16, 382).

According to a preferred aspect of the invention, the protein, which is xenogeneic for humans, is expressed in Escherichia coli cells transformed with an expression vector that expresses the protein with a polyhistidine segment at the N-terminal end, allowing efficient purification with a single chromatography stage on nickel resin.

The resulting protein presents bands in WB corresponding to polymeric, dimeric and trimeric forms of the 14.2 kDa monomer unit (Figure 1).

The trimeric form of the PR 14 protein according to the invention constitutes the correct physiological structure of the protein empowered with RID function, as well as with antigenic and cytotoxic activity. It can dissociate, but only partly, into dimeric and monomeric form, unlike the UK 114 proteins described in WO 0063368 and Colombo et al. and protein UK 1 14 expressed in yeast, which dissociate totally in monomeric form, as can be seen from Western Blotting, obtained by blotting with rabbit antibody RICl, immunised with said protein PRP14, shown in Figure 2/A, wherein track 1 refers to the standard of molecular weights; track 2 refers to recombinant protein UK 114 described in Colombo I. et al; track 3 refers to recombinant protein UK 114 expressed in yeast, and track 4 refers to the recombinant protein described in WO 00/63368.

The sequences of said proteins are reported in Table 2 below.

TABLE 2

Recombinant UK114 of the invention (SEP ID 1)

GSHM S SL VRRII ST AKAP AAIGP YS Q AVL VDRTI YI S GQLGMDP AS GQL VP GGVVEEAKQALTNIGEILKAAGCDFTNVVKATVLLADINDFSAVNDVYKQYFQS SFPARAAYQVAALPKGGRVEIEAIAVQGPLTTASL

Recombinant UK114 from E. coli (Colombo I. et al., Biochim. Biophys. Acta, 1998, 1442, 49-59) (SEP ID 6)

VPMSSLVRRIISTAKAPAAIGPYSQAVLVDRTIYISGQLGMDPASGQLVPG GVVEEAKQALTNIGEILKAAGCDFTNVVKATVLLADINDFSAVNDVYKQYFQSS FPARAAYQVAALPKGGRVEIEAIAVQGPLTTASL

Recombinant UK114 from yeast (SEP ID 7)

MSSLVRRIISTAKAPAAIGPYSQAVLVDRTIYISGQLGMD

PASGQLVPGGVVEEAKQALTNIGEILKAAGCDFTNVVKATVLLADINDFSAVND VYKQYFQSSFPARAAYQVAALPKGGR VEIEAIAVQGPLTTASL

Recombinant UK114 from E. coli (WO 0063368) (SEP ID 8)

VPMSSLVRRIISTAKAPAAIGPYSQAVLVDRTIYISGQLGMD PASGQLVPGGVVEEAKQALTNIGEILKAAGCDFTNVVKATVLLADINDFS AVNDVYKQYFQSSFPARAAYQVAALPKGGRVEIEAIAVQGPLTTASV

Figure 2/b shows the comparison with stable polymer PRP14. It has been demonstrated that 90% of the protein described in WO 00/63368, produced by recombinant techniques, dissociates to a monomer even under non-reducing conditions.

It has also been found that further polymer forms PRP 14 can be obtained by treating the proteins described above with strong acids such as perchloric acid. Perchloric acid also has an oxidising action, and the formation of polymers of sequence SEQ ID1 can also be obtained by using an oxidant such as hydrogen peroxide.

It has also been observed that the formation of polymers of sequence SEQ ID1 can be obtained by using crosslinking agents such as glutaraldehyde.

Under said conditions, the proteins cluster, giving rise to the formation of the forms according to the invention, which have high molecular weights and are therefore more immunogenic (Figure 3: track 1 : dissociated protein, track 2: trimeric PRP 14, track 3: protein PRP 14 treated with HC10 ₄ ).

Said trimeric and polymeric forms express IgM-competent antigenic determinants. Moreover, acid treatment of the proteins allows the identification and/or modification of antigenic determinants that further enhance the antigen capacity, especially in terms of the IgM response, which is mainly responsible for cytotoxicity, but without inducing modifications in the primary structure or destroying the quaternary structure of the protein.

The IgMs thus obtained are cytotoxic towards tumour cells, and their presence in hyperimmune sera persists for much longer than the normal IgM response to common antigens. Said polymers produce a marked increase in the immunological capacity of the protein, and therefore a corresponding increase in the immune response, which has very marked cytotoxic, not cytostatic characteristics (as in the case of native UK114, most of which is reduced to monomeric form, as demonstrated by mass analysis), due not only to the class G immunoglobulins, but mainly to the class M immunoglobulins which, contrary to current immunological theory, remain in circulation for months (Figure 4).

Unlike UK101 and native UK 114, the proteins according to the invention induce an antibody response without the aid of adjuvants or carriers which, however, can be used to further strengthen the immune response. The polymeric recombinant proteins according to the invention, preferably administered parenterally, in particular subcutaneously or intramuscularly, are particularly useful for the treatment of malignant tumours, preventive immunisation against recurrence in patients previously treated for tumours, adjuvant treatment, and vaccination of individuals at risk of onset or recurrence of said tumours in general.

For the recommended therapeutic or preventive uses the protein will be administered subcutaneously or intramuscularly in the form of solutions or suspensions in sterile aqueous carriers, optionally conjugated with adjuvants to further strengthen the immune response.

The protein in polymer form can also be administered by other routes, such as sublingually or topically. The doses for therapeutic, prophylactic and vaccinal applications can range from 5 to 40 mg per administration. A typical vaccination protocol involves 4 administrations, one every fortnight. 20/30 days after the last dose, a booster dose 4 times higher is given. Immunotherapy is monitored with laboratory tests to evaluate the antibody titre and cytotoxicity. The treatment can continue until a satisfactory clinical result is achieved (reduction or disappearance of the tumour mass), with boosters given at intervals, depending on the antibody titre monitored with the laboratory tests, to prevent a fall in the antibody titre or overimmunisation, leading to the risk of immune tolerance.

The polymeric recombinant protein can also be used to prepare monoclonal antibodies (anti-PRP14 Mabs) which are useful for intramuscular or intravenous passive immunotherapy, and passive seroprophylaxis of immunodepressed patients and/or those who fail to respond to specific immunotherapy.

For example, human monoclonal antibodies can be obtained by fusing a human- mouse myeloma K6H6/B5 with lymphocytes transformed with Epstein-Barr virus from patients pre-treated with the protein. Unlike the common practice for passive immunotherapy, which uses IgG monoclonal antibodies, it is preferable to use IgM- secreting clones which, in the presence of complement, are cytotoxic to tumour cells. Alternatively, a mixture of both types of monoclonal antibody (IgG and IgM) can be used.

Humanised murine IgM monoclonal antibodies or human IgM-secreting monoclonal antibodies produced in transgenic animals can also be used.

Immunocytochemistry has demonstrated that human tumours exhibit over 80% positivity to polymeric hyperimmune anti-protein sera (83%) and to polymeric anti- protein Mabs (93%>), as shown in Table 3 below.

Table 3

IMMUNOCYTOLOGICAL LOCALISATION IN CANCER TISSUES

The positivity is far more evident in percentage terms in adenocarcinomas (neoplasms of glandular origin which constitute 65% of malignant human tumours), and is not equally present in all cells of the same tumour. Some tumours, such as breast, pancreas, kidney, liver and prostate tumours, are therefore intensely positive, while others alternate intensely positive with negative or low-positivity areas (lung, colon) due to the presence of mucinous cells which do not express the antigen on their membrane.

In systemic tumours, leukaemias, lymphomas, myelomas and sarcomas, the positivity is 60%>. The invention is described in detail in the examples set out below.

EXAMPLE 1: Preparation of recombinant protein UK 114 from E. coli in stably trimeric form: PRP14

Expression vector pET-15b is used for the expression of recombinant protein UKl 14 from Escherichia coli. To transfer the nucleotide sequence encoding UKl 14 from a plasmid preceding the vector pET-15b, the sequence of UK114 is amplified by PCR, using the following pair of primers:

UK-Nde-FOR (SEQ ID 2):

5 ' AGCATATTCGACTGACATATGTCGTCTTTGGTCAGAAGGAT-3 ' UK- Xho-REV (SEQ ID 3):

5 'ATCGTCGGGCTCACTCGAGCTAGAGTGATGCTGTCGTGAGA-3 ' The product of PCR is digested with restriction enzymes Ndel and Xhol, and the resulting fragment is eluted from the gel.

The resulting fragment of DNA encoding UK114 is cloned in the vector which expresses the protein sequence fused to a polyhistidine tag and to a cleavage site recognised by a specific protease positioned at the N-terminal end.

The plasmid created, called pET15b-UKl 14, is used to transform Escherichia coli cells (DH5).

The transformers were screened. The cloned nucleotide sequence of plasmid pET15b-UKl 14 proved to be correctly fused to the vector and error- free.

Nucleotide sequence SEQ ID 4 (the start codon of UK114 is underlined and bolded):

ATGGGCAGCAGCCATCATCATCATCATCACAGCAGCGGCCTGGTGCCG CGCGGCAGCCATATGTCGTCTTTGGTCAGAAGGATAATCAGCACGGCGAAAG CCCCCGCGGCCATTGGTCCCTACAGTCAGGCTGTGTTAGTCGACAGGACCATT TACATTTCAGGACAGCTAGGTATGGACCCTGCAAGTGGACAGCTTGTGCCAG GAGGGGTGGTAGAAGAGGCTAAACAGGCTCTTACAAACATAGGTGAAATTCT GAAAGCAGCAGGCTGTGACTTCACGAATGTGGTAAAAGCAACGGTTTTGCTG GCTGACATAAATGACTTCAGTGCTGTCAATGATGTCTACAAACAATATTTCCA GAGTAGTTTTCCGGCGAGAGCTGCTTACCAGGTTGCTGCTTTGCCCAAAGGAG GCCGTGTTGAGATCGAAGCAATAGCTGTGCAAGGACCTCTCACGACAGCATC ACTCTAA

The amino-acid sequence SEQ ID 5 of the fusion protein His-tag-UKl 14 is shown below (the segment containing His-tag up to the thrombin cleavage site is underlined and bolded):

MGSSHHHHHHSSGLVPRGSHMSSLVRRIISTAKAPAAIGPYSQAVLVDRT IYISGQLGMDPASGQLVPGGVVEEAKQALTNIGEILKAAGCDFTNVVKATVLLAD INDFSAVNDVYKQYFQSSFPARAAYQVAALPKGGRVEIEAIAVQGPLTTASL

The amino-acid sequence after thrombin treatment is shown below (SEQ ID 1):

GSHMSSLVRRIISTAKAPAAIGPYSQAVLVDRTIYISGQLGMDPASGQLVP GGVVEEAKQALTNIGEILKAAGCDFTNVVKATVLLADINDFSAVNDVYKQYFQS SFPARAAYQVAALPKGGPvVEIEAIAVQGPLTTASL.

To express protein UK114, the plasmid is transferred from strain DH5a to the expression strain (Rosetta DE3).

The transformed cells are grown in a suitable medium (LB+antibiotics), and expression of the protein is induced with IPTG for 4 hours at 37°C. This induction time and the temperature of 37°C are ideal for the stable trimeric protein UK114 (PRP14).

Electrophoresis on polyacrylamide gel containing SDS (SDS-PAGE), and staining of the protein with Coomassie Blue, is conducted to verify the expression of the protein.

UK114-His is analysed by Western Blotting using the primary anti-poly-His antibody.

The clones tested express the recombinant protein UK114 at a high level. The protein is present in the soluble fraction of the cell extract.

The protein can be purified by exploiting the affinity of the polyhistidine tag for nickel resin. The soluble fraction obtained from cells collected 4 hours after induction at 37°C is incubated with the resin to allow the Histag of the recombinant protein to bind to the nickel conjugated with the resin. The flow-through (everything not bonded to the resin) is then collected, and after various washes the protein is eluted with a high concentration of imidazole which competes with histidine for the nickel bond.

The sample, obtained by pooling the eluates with a significant protein concentration, is dialysed in 50 mM NaH ₂ PC>4, 300 mM NaCl, pH 7.5. The sample is then incubated with thrombin, a serine protease which specifically recognises the cleavage site at the N-terminal end between the protein sequence and the Histag sequence, allowing the tag to be removed.

A portion of the digested sample is used to conduct analytical gel filtration with a SUPERDEX 75 column; by exploiting the known molecular weight standards, the molecular weight of recombinant protein UK114 can be obtained (MW=42.6 KDa). The purified protein, if compared with the molecular weight of the monomer, is a stable trimer, indicated by the code PRP14.

As demonstrated by Western Blot analysis (Figure 1), under these reducing conditions the protein does not dissociate, but remains stably monomeric, dimeric and trimeric. The protein therefore presents good immunogenic activity.

EXAMPLE 2: Manufacture of polymers by treating the dimeric recombinant protein with 2M perchloric acid, hydrogen peroxide and glutaraldehyde

A) treatment with 2M perchloric acid

5 ml of a protein solution obtained according to example 1 (3 mg/ml) is mixed at

4°C with 5 ml of 2 M perchloric acid. Sodium acetate is added to adjust the pH to 4.1, and the resulting solution is incubated overnight at 4°C. Any precipitates are removed by centrifugation at 10,000 rpm for 1 hour.

The solution is dialysed under stirring at 4°C against PBS, using a dialysis tube with a cut-off of 1000 Da, and filtered through an 0.2 micron filter. Under these conditions, as demonstrated by Western Blotting (Figure 3) performed with a rabbit antibody obtained by immunisation with the trimeric protein of example 1, different polymers of the same protein are formed, with covalent (and therefore not dissociable) bonds, with the same antigenic determinants, and having higher molecular weights up to about 70 kD. In Figure 3, track 1 refers to the recombinant protein described in WO0063368; track 2 refers to protein PRP14 of example 1, and track 3 refers to the polymer forms obtained after treatment with perchloric acid. B) treatment with hydrogen peroxide

5 ml of a protein solution obtained according to example 1 (3 mg/ml) is mixed at room temperature with 5 ml of 30% hydrogen peroxide. The resulting solution is incubated for 2 hours at room temperature.

The solution is dialysed under stirring at 4°C against PBS, using a dialysis tube with a cut-off of 1000 Da, and filtered through an 0.2 micron filter. Under said conditions, different polymers of the same protein are formed, with covalent bonds, which therefore cannot be dissociated. The resulting polymers have a molecular weight corresponding to those obtained using perchloric acid.

C) treatment with glutaraldehyde

5 ml of a protein solution obtained according to example 1 (3 mg/ml) is mixed at 4°C with 1 ml of an 0.01 M solution of glutaraldehyde in phosphate buffer pH 7.2, 0.1 M. The resulting solution is incubated overnight at 4°C. Any precipitates are removed by centrifugation at 10,000 rpm for 1 hour.

The solution is dialysed under stirring at 4°C against PBS, using a dialysis tube with a cut-off of 1000 Da, and filtered through an 0.2 micron filter. Under said conditions, different polymers of the same protein are formed, with covalent bonds, which therefore cannot be dissociated, with the same antigenic determinants, and having higher molecular weights up to about 70 kD. The resulting polymers have a molecular weight corresponding to those obtained using perchloric acid or hydrogen peroxide.

EXAMPLE 3: In vitro binding of anti-protein PRP14 antibodies to tumour cells

The binding to antigens present on the cell membrane and complement-mediated cytotoxic activity on numerous human tumour cell lines (KATO 3°, HT29, TSA, MCF7, JURKAT), and the absence of cytotoxicity on non-tumoral cell lines (NCTC, MCF10, HUVEC), was demonstrated using hyperimmune anti-PRP14 sera produced in mammals and the corresponding Mabs.

A) FACS analysis using human tumour cells HT29 as target. Cells not fixed and not permeabilised were tested with rabbit RICl antibody obtained by immunising with the protein of example 1 (PRP 14)

Flow cytometry analysis (FACS) demonstrates the antibody bond with the antigen present on the cell membrane (Figure 5): 1 control serum; 2 anti-protein serum of example 1, and the bond of the IgG and IgM antibodies present in the rabbit serum immunised with the trimeric protein of example 1 (PRP 14, Figure 6).

IgM antibodies are mainly responsible for the cytotoxic effect.

B) ELISA assay, using HT29 cells adhering to the incubation well as target. Cells dried and fixed but not permeabilised.

Figure 7 shows the results of the test performed with different concentrations (0.1 to 5%) of the various antisera obtained: rabbit antibody obtained by immunising with the protein of example 1 (PRP 14); rabbit antibody obtained by immunising with the polymeric protein obtained by treating with perchloric acid (PRP a). Control: pre-immune rabbit serum.

C) Cytotoxic activity test, using the MTT assay. HT29 cells tested with rabbit antisera immunised with polymeric proteins UK114 + complement.

The cytotoxicity tests are reported in Figure 8: 1) Control: pre-immune rabbit serum; 2) anti-native UK 114 mAb (International Journal of Oncology 8: 543-548,1996); 3) rabbit antibody obtained by immunising with polymeric protein PRP14; 4) rabbit antibody obtained by immunising with polymeric protein PRP 14 treated with perchloric acid; 5) rabbit antibody obtained by immunising with polymeric protein PRP14 treated

EXAMPLE 4: Cytotoxicity vs. cytostaticity

ATP lite is a cell viability measurement test that does not distinguish between cytotoxic and cytostatic effects.

Using the Annexin V plus Propidium Iodide test, it was found that for anti-PRP14, of the 78% viability loss of the neoplastic cells detected with the ATP lite tests, 60% was due to a cytotoxic effect (apoptosis) and 60%> to a cytostatic effect, while for native antibodies against UK101/UK114, almost the entire loss of viability was due to a cytostatic, not a cytotoxic effect (Figure 9).

EXAMPLE 5: Treatment with PRP14 of Balb/c) mice carrying murine colon adenocarcinoma (C26)

C26 is a highly aggressive murine tumour which grows rapidly, reaching dimensions incompatible with the life of Balb/c mice (weight 18/20 grams) in 20 days.

4 days after tumour implantation (when the neoplastic mass is already appreciable), the treated mice received 10 micrograms of polymeric protein PRP14 on the 4th, 11th and 18th days, while the control animals received saline solution or the recombinant protein UKl 14 described in WO 00/63368. The test was conducted with the luciferase technique. As demonstrated in Figure 10, treatment of BALB/c mice with polymeric protein PRP14 gave rise to a significant (>50%) reduction in the growth and tumour mass of C26 carcinomas implanted subcutaneously. There was no significant reduction in the control animals or the mice treated with the recombinant protein UKl 14 described in WO 00/63368.

EXAMPLE 6): Inhibition of mammary carcinogenesis in BALB/c mice carrying the activated HER2 rat gene.

This is an experimental model, extensively illustrated in the literature (Calogero et al, Breast Cancer Research 2007, 9:21), wherein mammary carcinomas appear in female BALB/c mice that overexpress the HER2 neu gene. The onset of the carcinomas, which appear at well-established fixed times, can be monitored over time with this model.

20 animals were used, and the appearance of mammary carcinomas was monitored over time. The model is based on the finding that in these animals, atypical hyperplasia develops spontaneously in the mammary ducts and lobules in the 2nd week of life. In the 10th week the epithelial cells occlude and distend the ducts, which take on the appearance of carcinomas in situ. As the weeks pass, said pre -neoplastic lesions are transformed into invasive carcinomas. By the end of the experiment, in the 33rd week, palpable invasive carcinomas had appeared in all 10 mammary glands. To evaluate the efficacy of preventive immunisation with protein PRP14 in preventing or slowing the appearance of invasive carcinomas, the polymeric protein + Freund's adjuvant was administered to 10 animals (test animals) from the 8th week of life. A subcutaneous booster was then administered every 4 weeks. An equal number of animals (controls) were treated with saline solution.

The difference between the incidence and tumour appearance curves and the number of multiple tumours that appeared was evaluated in the test and control animals.

A significant reduction in said parameters was observed between the test and control animals. Only 10% of the animals treated with PRP14 presented tumour implants, and the size and number of said implants were smaller.

The result therefore confirms the ability of antitumoral vaccination with the polymeric protein to inhibit and/or prevent carcinomatous transformation and the formation of invasive carcinomas induced in mammary epithelial cells by the mitogenic activity of the activated HER2 gene.

EXAMPLE 7: Passive seroprophylaxis in nude mice carrying human tumour

HT29

HT29 human colon adenocarcinoma cells were implanted into nude mice to demonstrate the transmission of antitumoral immunity to the nude mouse carrying human tumours, using hyperimmune mammalian sera (immunised with protein PRP14) and sera from the same mammals before immunisation with said protein.

It was demonstrated that the hyperimmune serum of mammals immunised with protein PRP14 considerably reduces implantation of human tumours in the nude mouse. Only 20% of tumours took root in the animals that underwent seroprophylaxis (with hyperimmune mammalian serum), with a considerable increase in implantation latency and survival, compared with 100% of the controls injected with pre-immune control serum.

In the treated animals the growth time of the tumours, which were implanted in any event, was three times less than that of the controls, and the survival rate was at least twice as high as in the controls (TABLE 4).

Table 4. HT-29 cells in nude mice. Treatment with trimeric anti-protein and anti-PRP14 sera. Control with pre-immune serum.

Serum No. of Tumours Latency (D) Survival (D)

Pre-immune 12/12 9 30

Anti-PRP14 3/12 28 87