Isoform of the TGF-beta receptor II

Abstract

An isoform of the TGF beta receptor II comprising a sequence of about of 80 amino acids and lacking a transmembrane domain; wherein the isoform is a TGF-1 agonist. The isoform comprises the amino acid sequence set forth in SEQ ID No. 12. The isoform may have the amino acid sequence set forth in SEQ ID No. 2 or sequences having at least 85% sequence identity to the sequence set forth in SEQ ID No. 2.

Claims

1. A polynucleotide comprising: a nucleotide sequence set forth in SEQ ID No. 1, wherein said polynucleotide is a cDNA, and said cDNA encodes the TRII-SE isoform consisting of the amino acid sequence set forth in SEQ. ID No. 2.

2. A cell transduced, comprising the polynucleotide of claim 1.

3. The cell transduced according to claim 2, wherein the cell expresses the TRII-SE isoform set forth in SEQ ID No 2.

4. The cell transduced according to claim 2, wherein the cell is a mammalian cell.

5. The cell transduced according to claim 4, wherein the mammalian cells are selected from the group consisting of cell lines and primary culture.

6. The cell transduced according to claim 2, wherein the cell is a bacterial cell.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 shows a schematic diagram of the TRII receptor indicating the extracellular (ECD), transmembrane (TMD) and intracellular (ICD) domains. FP and RP boxes indicate the forward and reverse primers used to amplify the TRII cDNA by RT-PCR;

(2) FIG. 2 shows a gel with the results of recombinant plasmid digestion containing the two already described human TRII (A and B) isoforms and the newly described by the present inventors, TRII-SE, obtained by RT-PCR from human lymphocytes;

(3) FIG. 3 shows the alignment of partial cDNA sequences of the two known TRII (A and B, SEQ ID No 14 and 13, respectively) isoforms, and the one disclosed in the present application (TRII-SE, SEQ ID No 1); cDNA sequences include the start codon (ATG) and the last nucleotide encoding the transmembrane domain (TMD); the dark grey bar indicates an additional deletion found in exons II and III of TRII-SE;

(4) FIG. 4 shows alignments of partial predicted protein sequences belonging to the human TRII isoforms A and B (SEQ ID No 16 and SEQ ID No 15, respectively), and the TRII-SE (SEQ ID No 2); light grey boxes show residues involved in disulfide bridges critical for receptor-ligand bonding (C54-C71, C61-C67); dark grey boxes show residues which are fundamental for interaction with TGF- (D55, I76, E142);

(5) FIG. 5 shows the results of detection by RT-PCR of the different TRII isoforms (A, B and SE) in different human cell types; HT1080 (fibrosarcoma), A549 (pulmonary adenocarcinoma), CaCo-2 (colorectal adenocarcinoma), Hep3B (hepatic carcinoma), Jurkat (acute T-cell leukemia), 293T (epithelial cells from embryonic kidney immortalized with the SV40 virus large T-antigen), HEK-293 (epithelial cells from embryonic kidney immortalized with adenovirus), EBV-LCL (lymphoblastoid cell line immortalized with the Epstein-Barr virus), and hASC (stromal mesenchymal cells from human adipose tissue);

(6) FIG. 6 shows the results obtained by flow cytometry plots showing cell purity of monocytes (CD14+), B-cells (CD19+), and T-cells (CD3+) separated by immune purification;

(7) FIG. 7 shows TRII splicing variant mRNA profiles in human leukocyte subsets, such as granulocytes, T-lymphocytes (CD3+), B lymphocytes (CD19+), and monocytes (CD14+);

(8) FIG. 8 shows lentiviral vectors encoding the newly described hTRII-SE variant and a dominant negative (DN) mutant of the TRII-A receptor under the action of the CMV promoter; as a control, a lentiviral vector encoding eGFP under the CMV promoter was used. The complete names of the vectors are indicated at the left side of the diagram. The abbreviated names are shown on top of each vector;

(9) FIG. 9 shows overexpression of TRII-SE in A549 cells. A): results of a flow cytometry analysis showing the percentage of eGFP expressing A549 cells transduced with a lentiviral vector encoding TRII-SE (Lt-TRII-SE) and control vectors; B): results of a RT-PCR showing overexpression of TRII-SE at the mRNA level; C): results of a demonstration of the presence of TRII-SE only in the supernatant of cells transduced with Lt-TRII-SE as detected by Western blot with a TRII specific antibody recognizing the extracellular domain;

(10) FIG. 10 shows the results of a proliferative MTT assay. A): A549 cells untransduced (UT) and transduced with Lt-TRII-SE, Lt-TRIIA-DN, and Lt-eGFP, treated with 0.4 nM TGF-1 and untreated. B): TGF-1 curve in A549 cells transduced with a lentiviral vector encoding TRII-SE and untransduced (UT). *p<0.05; **p<0.01, ***p<0.001;

(11) FIG. 11 shows: A) results of a flow cytometry analysis of hASC transduced with lentiviral vectors encoding TRII-SE, TRIIA-DN, and eGFP; and untransduced (UT), and B) representative histogram showing percentage of purity after cell sorting;

(12) FIG. 12 shows the results of a RT-PCR analysis of hASC cells showing overexpression of TRIIA-DN and TRII-SE; GAPDH was used as reference gene;

(13) FIG. 13 shows relative mRNA levels of TRII receptors (TRII-A, TRII-B and TRII-SE) in untransduced hASCs (UT) and transduced with Lt.TRII-SE.

(14) FIG. 14 shows mRNA levels of TRII receptors in hASCs cells incubated with and without exogenous TGF-1;

(15) FIG. 15 shows mRNA levels of isoforms TRII-A and TRII-B in hASCs cells transduced with lentiviral vectors (Lt) encoding TRII-SE and control vectors incubated with and without TGF-1;

(16) FIG. 16 shows X-ray images of rats treated with ciprofloxacin (CPFX) and intra-articularly injected in the knees with Lt.coTRII-SE, Lt.eGFP, and culture medium (vehicle). White arrows indicate radiolucent images;

(17) FIG. 17 shows a graphic of serum level measurements for aspartate transaminase (AST), in the same animals;

(18) FIG. 18 shows a cDNA alignment to compare changes made to the recombinant TRII-SE (SEQ ID No 1). To obtain coTRII-SE/Fc (SEQ ID No 7) (underlined sequence), a Kozak sequence (light gray box) was included in the TRII-SE cDNA, to make translation initiation more efficient. Additionally, some nucleotides have been changed (black boxes with white letters) for codon optimization, to make translation more efficient in human cells. To allow fusion in frame of cDNA with the human IgG-Fc domain cDNA, the stop codon of TRII-SE was removed (italics) and replaced by a BglII recognition sequence in the new construct. Primers used for PCR-amplification of human IgG1 Fc coding sequences are shown in dark gray boxes (SEQ ID No 18 and SEQ ID No 19);

(19) FIG. 19 shows protein alignment to compare changes made to the recombinant TRII-SE (SEQ ID No 2). coTRII-Se was fused in frame to the human IgG1 Fc domain. Asterisk: Stop Codon; Black Box: linker aminoacids; Grey box: Fc domain (SEQ ID No 6);

(20) FIG. 20 shows a schematic diagram of the self-inactivating (SIN) bicistronic lentiviral vector encoding the fusion cassette coTRII-SE/Fc together with ires eGFP, under the control of an internal CMV promote.

(21) FIG. 21 shows flow cytometry dot plots demonstrating the efficiency of vector transduction of Lt.coTRII-SE/Fc.ires eGFP and the control vector Lt. eGFP;

(22) FIG. 22 shows the results of an agarose gel electrophoresis with RT-PCR products, using primers for amplifying IgG1 Fc, from RNAm of Mock, Lt.eGFP, and Lt. coTRII-SE/Fc transduced A549 cells; and

(23) FIG. 23 shows the results of a Western blot of cell lysates (CL) and supernatants (SN) from proteins of Mock, Lt.eGFP and Lt. coTRII-SE/Fc transduced A549 cells.

DETAILED DESCRIPTION OF THE INVENTION

(24) A variant or isoform of the TGF beta receptor II is disclosed, which is expressed in human cells referred to herein as endogenous soluble TRII (TRII-SE) and that contrarily to other isoforms acts like a TGF-1 agonist.

(25) By using specific primers, a region of the human TRII mRNA from T-lymphocytes only encoding the extracellular (ECD) and the transmembrane (TMD) domains and excluding the intracellular domain (ICD) was initially amplified by RT-PCR, (FIG. 1).

(26) After the PCR reaction, DNA products were cloned into the pGEM-T Easy plasmid. Plasmids were digested with AgeI and SalI and revealed in an agarose gel the presence of clones with inserts of three different sizes (FIG. 2). Clone 2 contained an insert of 650 bp. In clones 3, 7, 8, 11, and 12 the insert size was of 580 bp and in clone 10 the size reflected the presence of an insert of 430 bp.

(27) DNA sequencing and BLAST alignment (NCBI) of all clones indicated that clones 3, 7, 8, 11, and 12 (582 bp) were identical to human TGF receptor II variant A (TRII-A). Additionally, clone 2 (657 bp) showed 100% identity with the isoform TRII-B. Clone 10 (433 bp) was similar to the TRII-A sequence but with an additional 149 bp deletion. In this clone, the last 62 bp encoded by exon II and the first 88 bp encoded by exon III were absent, TRII-SE (SEQ ID No. 1) (FIG. 3).

(28) The alignment of the predicted amino acid sequence of all three isoforms (FIG. 4) indicated that the deletion found in clone 10 generates a frameshift starting at amino acid 68, which adds a stop codon 13 amino acids after the deletion generating a prematurely terminated 80 amino acids long isoform lacking the transmembrane domain and this is the new isoform TRII-SE (SEQ ID No. 2).

(29) This isoform differs in 12 amino acids at the carboxyl end compared to the membrane bound variants of TRII (isoforms A and B). Due to this, and according to the predicted amino acid sequence, the TRII-SE isoform of clone 10 lacks pivotal sites for the productive action of TGF- such as amino acid 176 of SEQ ID No. 3 that contributes to the ligand-receptor binding through hydrophobic contact; amino acid E142 of SEQ ID No. 3 which forms hydrogen bonds with R25 of TGF- increased affinity and determined binding specificity and amino acid C71 of SEQ ID No. 3 which forms a disulfide bridge with C54 of the same receptor (see FIG. 4) necessary both for binding to the ligand and for signaling (reference, Alain Guimond, et. al. FEBS Letters 515: 13-19, 2002). Thus, the TRII-SE isoform might not be able to bind TGF-1 with the same affinity than that of known isoforms. Additionally, due to the premature termination, the TRII-SE isoform lacks the amino acid sequence belonging to the transmembrane domain (TMD), showing the presence of a new endogenously secreted soluble TRII isoform in human T-lymphocytes.

(30) As previously mentioned, the new isoform is referred to as TRII Soluble Endogenous (TRII-SE). The TRII-SE isoform is different from the secretable genetically engineered TRII isoform. The latter is an artificial TRII receptor with a truncated TRII-A fused to the Fc region of human IgM and blocks the effects of TGF-, thus acting as an antagonist (reference, R. J Akhurst. J. Clin. Invest. 109: 1533-3610, 2002).

(31) To determine the theoretical molecular weight of the TRII-SE isoform, post-translational modifications (PTM) predicted from the amino acid sequence (SEQ ID No. 2) were established by using different computer programs (Table 1). In this analysis, three glycation sites at K46, K52 and K78 (NetGlycate program) (Johansen, M. B.; Glycobiology 16: 844-853, 2006); three phosphorylation sites at S31, S59 and Y73 (NetPhos program) (Blom, N.; Journal of Molecular Biology 294: 1351-1362, 1999) and one site for sumoylation in K46 (SUMOplot program, ABGENT, CA, USA) were identified. On the other hand, sites for sulfonation, C-mannosylation, O-GalNAC glycosilation, O-glycosilation, N-glycosilation, myristoylation, and palmitoylation were not found in TRII-SE. In this study it was estimated that the molecular weight of the mature TRII-SE isoform was of about 18.4 kDa.

(32) TABLE-US-00001 TABLE 1 In silico analysis of the TRII-SE amino acid sequence showing predicted post-translational modifications and molecular weight with and without modifications. Predicted pl/ 9.64/9161.72 theoretical Mw pl/Mw without a 9.05/6532.51 6,532.51 kDa signal peptide Secretion 0.960 SignalP probability of (first 12 aa) Program the signal peptide Clivage site Between pos. SignalP 23 and 24 Program C-mannosylation No sites GalNAc No sites O-glycosylation Glycations 3 sites NetGlycate 0.558 kDa (Lys 46, Program (0.186 kDa each) 52, and 78) N-glycosylations No sites NetNGlyc Program O-Glycosylations No sites (OGPT Program) O-(beta)-GlcNAc No sites Myristoylation No sites Palmitoylation No sites Phosphorylation 3 sites NetPhos 0.285 kDa (Ser 31 and Program (0.095 Da each) 59, Tyr 79) Sulfonations No sites Addition of 1 site SUMOplot 11 kDa SUMO protein (Lys 46) program Final Mw with 18.4 kDa modifications

(33) To confirm whether TRII-SE mRNA was also present in human cells other than lymphocytes, we amplified by RT-PCR using the same set of primers various human cell lines and primary cultures (FIG. 5). It may be observed that human solid tumor derived cell lines, for example, HT1080 (fibrosarcoma), A549 (lung adenocarcinoma), CaCo-2 (colon cancer) and Hep 3B (hepatocellular carcinoma) only showed the presence of mRNA of variants A and B, but not TRII-SE. Additionally, in Jurkat cells (acute lymphoid leukemia), 293T cells (embryonic kidney cells immortalized with the SV40 T-antigen), HEK-293 cells (embryonic kidney cells immortalized with the adenovirus E1A protein, EBV-LCL (Lymphoblastoid Cell Line immortalized with the Epstein Barr Virus) and ASC (human adipose derived mesenchymal stem cells) passage 6 primary culture, mRNA encoding for TRII-SE was present in all cases (FIG. 5). The presence of the TRII-SE isoform was further confirmed by DNA sequencing.

(34) To check whether TRII-SE is also present in leukocytes different from T-lymphocytes, granulocytes, monocytes, B-cells and T-cells were purified from human peripheral blood by density gradient and subsequent magnetic immune-purification with specific monoclonal antibodies, to high purity (FIG. 6). RT-PCR analysis showed that TRII-SE is present in all leukocyte subsets but with different expression levels (FIG. 7).

(35) To determine whether TRII-SE may be secreted to the extra cellular medium, TRII-SE cDNA was cloned downstream from the ubiquitous promoter CMV in a self-inactivating (SIN) bicistronic lentiviral vector also expressing eGFP, as described in the examples, to generate the Lt-TRII-SE vector. As a control, two lentiviral vectors were used: one bicistronic encoding a dominant negative TRII mutant together with eGFP (Lt-TRIIA-DN) and another encoding eGFP alone (Lt-eGFP), also under the action of the CMV promoter (FIG. 8).

(36) With these lentiviral vectors, shown in FIG. 8, A549 cells were transduced, at an MOI of 50. Seventy two hours after transduction, cell supernatants were frozen for further experiments and the percentage of eGFP expressing cells was measured by flow cytometry (FIG. 9A). In cells transduced with Lt-TRII-SE and Lt-eGFP, 68.63% and 65.27% of the cells, respectively, showed integration of the lentiviral vector as demonstrated by eGFP expression. RT-PCR of Lt-TRII-SE transduced cells revealed the presence of a 433 bp band, indicating overexpression at the mRNA level of the TRII-SE isoform (FIG. 9B). Cell supernatants were thawed, and Western blotted as described in the examples (FIG. 9C). Only TRII-SE was detected in the supernatant of Lt-TRII-SE transduced A549 cells cultured in the presence of protease inhibitors.

(37) The molecular weight of TRII-SE detected by Western blot is in agreement with the predicted molecular weight, after the addition of post-translational modifications (18 kDa) (Table 1). This is the first evidence ever that there exists a new secretable TRII receptor variant or isoform in human cells.

(38) To show the function of the TRII-SE isoform, functional assays were carried out wherein untransduced, expressing nearly undetectable levels of TRII-SE, transduced with lentiviral vectors encoding eGFP alone, or bicistronics together with either TRII-SE or the dominant negative (DN) mutant of the TRIIA variant known to work as a TGF-1 antagonist, A549 cells were used.

(39) Initially, MTT ((3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide; thiazolyl blue) assays were performed to evaluate if overexpression of TRII-SE inhibits or not cell proliferation in the presence of 0.4 nM TGF-1 (FIG. 10A). As may be noted, in the presence of TGF-1, TRII-SE-transduced cells proliferate significantly less than the same cells not treated with TGF-1 and at levels found in control untransduced cells (UT) and Lt.eGFP-transduced cells. These results indicated that TRII-SE is not a TGF-1 antagonist.

(40) Additionally, to check whether TRII-SE acts as a TGF-1 agonist, A459 cells either overexpressing TRII-SE or not (untransduced cells or UT) were incubated in the presence of increasing concentrations of TGF-1 (FIG. 10B). These results show that in UT cells, proliferation started to decrease in the presence of 0.2 nM TGF-1 compared to the values obtained in the absence of TGF-1. However, in cells overexpressing TRII-SE, proliferation started to decrease at a TGF-1 concentration of 0.1 nM compared to the same cell line without the addition of TGF-1. These results indicate that in cells overexpressing TRII-SE, TGF-1 achieved the same effect than in UT cells but at half concentration, which would suggest that the TRII-SE isoform may act as an agonist.

(41) To further assess the agonistic role of the TRII-SE isoform, hASCs were transduced with Lt-TRII-SE, Lt-TRIIA-DN, and Lt.eGFP, at an MOI of 150 as described in the examples. Seventy two hours after transduction the percentage of eGFP expressing cells was measured by flow cytometry (FIG. 11A). For further experiments with pure cell populations, transduced cells were expanded and cell sorted in a FACSAriaII Cell Sorter (Becton Dickinson, San Jose, Calif.) to a purity of eGFP-expressing cells of more than 90% (FIG. 11B), indicating that most cells overexpress the new isoform.

(42) RT-PCR performed on poly A+ mRNA from either transduced or untransduced hASC cells showed the pattern of TRII isoforms expression depicted in FIG. 12. Cells overexpressing TRII-SE showed a strong band of 433 bp and a weak band of 582 bp reflecting the fact that overexpression of TRII-SE downregulates TRII isoform A expression. Similarly, when TRIIA-DN was overexpressed in hASC cells, TRII-SE expression (433 bp) could not be detected. Finally, in hASC cells transduced with the lentivector encoding only the eGFP marker gene, a weak band representing expression of TRII-A was detected, suggesting that viral transduction per se downregulates TRII expression.

(43) mRNA levels of all three isoforms of Type II TGF- receptor were also quantified by qRT-PCR (FIG. 13). It was found that in untransduced cells (UT), membrane bound TBRII-A and B variants were the main molecules to be expressed and TRII-SE expression was minimal, as expected. Contrarily, when the new isoform expression was increased in hASC cells, both TRII-A and B variants decreased dramatically, due to a compensation effect which shows the agonistic effect of the TRII-SE isoform.

(44) This compensation effect was also verified by addition of exogenous TGF-1 and analysis of mRNA levels of the TRII variants in hASCs cells (FIG. 14). It was found that upon addition of TGF-1, TRII-A increased and TRII-SE decreased compared to untreated cells, suggesting once again that the TRII-SE isoform acts as a TGF-1 agonist.

(45) According to this, it was also found that mRNA of both TRII-A and TRII-B are highly upregulated (40- and 50-fold increase, respectively) in cells overexpressing Lt-TRII-SE in the presence of physiological concentrations of TGF-1 compared to levels of mRNA produced in the absence of exogenous TGF-1, further confirming the role of TRII-SE acting as a TGF-1 agonist by increasing the expression of membrane-bound receptors TRII and TRII-B (FIG. 15).

(46) Furthermore, the effect of TRII-SE recombinant isoform was measured on a panel of 80 cytokines secreted by hASCs cells (FIG. 16). Cells were transduced with either control Lt-GFP, the TGF-1 inhibitor Lt. TRII-DN, or Lt-TRII-SE and incubated in the presence or absence of exogenous TGF-1. Collected supernatants were used to analyze the cytokines in a Cytokine Array G5 (Raybiotech, Inc. Norcross, USA).

(47) TABLE-US-00002 TABLE 2 Autocrine TGF-1 DN SE DN SE Hematopoietic cytokines Insulin Like Growth Factor Superfamily G-CSF IGF-1 M-CSF UC IGFBP-1 UC GM-CSF IGFBP-3 (15,60) IL-6* UC UC IGFBP-4 abs abs IL-7 (2,02) Tumor Necrosis Factor Superfamily LIF UC UC TNF- (7,77) FLT3-L UC TNF- UC (1,85) SCF abs abs LIGHT abs abs IL-3 UC Fibroblast Growth Factor Family Oncostatin M UC FGF-7 UC Angiogenic cytokines FGF-9 UC VEGF (0,65) (1,85) Neurotrophins Angiogenin UC UC BDNF UC HGF (1,81) (7,65) NT-3 UC EGF abs abs NT-4 UC UC PIGF UC Tissue Inhibitor of Metalloproteinases Family Chemokines TIMP-1 UC UC cxcl GRO UC UC TIMP-2 UC UC CXCL1 (GRO) UC Macrophage Activating Factors CXCL5 (ENA-78) UC (1,62) INF- UC CXCL6 (GCP-2) UC UC MIF UC (1,97) CXCL8 (IL-8) UC (1,67) IL-2 UC CXCL9 (MIG) UC UC Bone Remodeling Cytokines CXCL10 (IP-10) UC UC Osteopontin abs abs CXCL12 (SDF-1) UC Osteoprotegerin UC UC CXCL13 (BLC) abs abs Hormones ccl CCL1 (I-309) UC Leptin (1,79) CCL2 (MCP-1) UC UC GDNF Family CCL4 (MIP1b) UC GDNF UC UC CCL5 (RANTES) (1,89) (7,85) Anti-inflammatory Interleukins CCL7 (MCP-3) UC UC IL-10 UC CCL8 (MCP-2) (3,60) IL-13 (5,47) CCL11 (Eotaxin) UC (2,39) Pro-inflammatory Interleukins* CCL17 (TARC) UC UC IL-1 (3,11) CCL18 (PARC) (3,46) IL-1 abs abs CCL20 (MIP3a) abs abs IL-5 UC (1,87) CCL24 (Eotaxin-2) IL-12 p70 UC CCL26 (Eotaxin-3) abs abs IL-15 abs abs TGF- Family TGF-1 (2,57) (4,94) TGF-2 (1,61) (1,55) Paracrine TGF-1 (3 pg/l) DN SE DN SE Hematopoietic cytokines Insulin Like Growth Factor Superfamily G-CSF IGF-1 abs Abs M-CSF abs abs IGFBP-1 abs Abs GM-CSF (4,89) IGFBP-3 IL-6* UC UC IGFBP-4 UC IL-7 UC Tumor Necrosis Factor Superfamily LIF (2,43) UC TNF- UC FLT3-L UC UC TNF- SCF UC UC LIGHT IL-3 Fibroblast Growth Factor Family Onc M abs abs FGF-7 abs Abs Angiogenic cytokines FGF-9 UC UC VEGF (2,35) UC Neurotrophins Angiogenin (1,59) UC BDNF abs Abs HGF (4,16) NT-3 EGF NT-4 abs Abs PIGF Tissue Inhibitor of Metalloproteinases Family Chemokines TIMP-1 (2,26) UC cxcl GRO UC UC TIMP-2 (2,07) (1,52) CXCL1 (GRO) abs abs Macrophage Activating Factors CXCL5 (ENA-78) (1,64) UC INF- CXCL6 (GCP-2) (2,45) MIF (1,76) UC CXCL8 (IL-8) UC (1,57) IL-2 abs Abs CXCL9 (MIG) Bone Remodeling Cytokines CXCL10 (IP-10) Osteopontin CXCL12 (SDF-1) abs abs Osteoprotegerin UC (3,32) CXCL13 (BLC) Hormones ccl CCL1 (I-309) abs abs Leptin (1,82) (16,58) CCL2 (MCP-1) UC UC GDNF Family CCL4 (MIP1b) abs abs GDNF abs Abs CCL5 (RANTES) (3,33) (4,20) Anti-inflammatory Interleukins CCL7 (MCP-3) UC (1,78) IL-10 (5,36) CCL8 (MCP-2) UC IL-13 CCL11 (Eotaxin) UC UC Pro-inflammatory Interleukins* CCL17 (TARC) UC UC IL-1 CCL18 (PARC) (3,38) IL-1 UC UC CCL20 (MIP3a) UC UC IL-5 (5,61) CCL24 (Eotaxin-2) abs abs IL-12 p70 CCL26 (Eotaxin-3) abs abs IL-15 TGF- Family TGF-1 UC UC TGF-2 UC (2,27)

(48) The results obtained with cytokine arrays are shown in Table 2. Increase or decrease of cytokines levels are referred to the levels secreted by cells transduced with the control vector Lt.eGFP either in the presence (paracrine) or absence (autocrine) of exogenous TGF-1. UC: unchanged levels with respect to cells transduced with the control vector Lt.eGFP. Abs: absent in mock transducer cells control.

(49) It is shown that in ASC cells overexpressing TRII-DN with a high TGF-1 concentration, OPG secretion remains unchanged with respect to the values obtained in Lt.eGFP-transduced control cells, making cells insensitive to TGF-1.

(50) On the other hand, high TGF-1 concentrations caused a dramatic drop of OPG secretion in TRII-SE overexpressing cells compared to control cells (Lt.eGFP-transduced). The TRII-SE isoform acts oppositely to the TGF-1 inhibitor (TRII-DN) and seems to favor osteoclastogenesis.

(51) Table 3 summarizes the results obtained by other authors, and those compared to the results disclosed in the present application regarding the cytokine array and the relationship with osteoarthritis (OA).

(52) TABLE-US-00003 Bone/cartilage Results of the MSC/Osteoblasts Disease remodeling Invention High TGF-1 OA Bone loss/increase Lower OPG of osteoclastic TGF-1 resorption agonist Increased PTG content Higher HGF High angiogenesis TGF-1 Osteophyte outgrowth agonist TGF-1 inhibition OA-like Decreased osteoclastic Higher OPG (TRII-DN) resorption Decreased TGF-1 PTG content/increased antagonist cartilage loss No HGF Angiogenesis TGF-1 Decreased osteophyte antagonist formation

(53) In is shown that in cells overexpressing TRII-SE HGF secretion is highly upregulated both in the presence (4.16 times) or absence (7.65 times) of exogenous TGF-1, whereas in cells overexpressing the dominant negative mutant TRII-DN, HGF secretion decreases 1.81 times or is absent, in the absence and presence of exogenous TGF-1, respectively. These results show that the TRII-SE isoform is involved in the positive regulation of HGF.

(54) Increased TGF-1 acts differently in animals depending on whether injections were applied in normal or osteoarthritic models. In normal animals, either TGF-1 protein or adenovirus TGF-1 injection generates increased synthesis and content of proteoglycan and osteophyte formation. On the other hand, in osteoarthritis (OA)-induced models, increases in the TGF pathway help to decrease cartilage damage, proteoglycan and osteophyte formation. Thus, the effect of the TII-SE isoform was analyzed either in CPFX-treated juvenile rats (24 days old) or untreated rats, by intra-articular injections of lentiviral vectors encoding a recombinant protein of the codon-optimized (co) TRII-SE fused to the constant fragment (Fc) of the human immunoglobulin 1 (IgG1) (Lt.coTRII-SE/Fc) or the enhanced green fluorescent protein (Lt.eGFP).

(55) Seven days after injecting the vector into rats treated with ciprofloxacin (CPFX), only articulations overexpressing the fusion peptide or a fused coTRII-SE/Fc isoform showed radiolucent images with irregular borders in the femoral condyle, consistent with intraosteal geodes (FIG. 16). It is shown that coTRII.SE/Fc could cause osteolytic damage by bone resorption.

(56) When compared to serum levels of urea, creatinine, total proteins, albumin, alkaline phosphatase, alanine transaminase (ALT), and aspartate transaminase (AST), a statistically significant difference was only found for the latter. An increase in aspartate transaminase (AST) was only observed in serum of rats treated with CPFX and intra-articularly injected with Lt.coTRII-SE (FIG. 17). Mitochondrial and cytoplasmic forms of AST are found in all cells, so the increase of AST which was only observed in rats injected with Lt.coTRII-SE/Fc in combination with CPFX show that coTRII-SE enhance the effect of CPFX on tissue damage in muscle, tendons or other tissues.

(57) In the present application, the generation of a new recombinant TRII-SE protein expressed in human cells is shown. It is known that in nature, the concentration of soluble receptors is very low, thus, to increase the levels of the recombinant TRII-SE protein, the original coding sequence was codon optimized, and a Kozak sequence was included (Epoch Biolabs Inc., Texas, USA) referred to herein as coTRII-SE (SEQ ID No. 4) and encoded by SEQ ID No. 5 (FIG. 18). Additionally, to make the protein more stable in vivo, and for a more effective purification, the human IgG1 Fc region was cloned in frame downstream of the coding sequence of coTRII-SE to obtain the fusion peptide coTRII-Se/Fc, as previously mentioned (SEQ ID No. 6), encoded by SEQ ID No. 7 (FIGS. 18 and 19).

(58) As can be observed, FIG. 18 shows a cDNA alignment to compare changes made to the recombinant TRII-SE. To obtain the coTRII-SE/Fc (underlined sequence), a Kozak sequence (light gray box) was included in the TRII-SE cDNA, to make the initiation of translation more efficient. Additionally, some nucleotides have been changed (black boxes and white letters) for codon optimization, in order to make translation more efficient. To allow fusion in frame of cDNA with the human IgG-Fc domain cDNA, the stop codon of TRII-SE was removed (italics) and replaced by a BglII recognition sequence in the new construct. Primers used for PCR-amplification of human IgG1 Fc coding sequences are shown in dark gray boxes.

(59) As can be observed, FIG. 19 shows a protein alignment and allows for comparing changes made to the recombinant TRII-SE. coTRII-Se was fused in frame to the human IgG1 Fc domain. Asterisk: Stop Codon; Black Box: linker aminoacids; Grey box: Fc domain.

(60) Subsequently, the recombinant coTRII-SE/Fc cDNA was inserted between the AgeI and EcoRV sites of a SIN lentiviral vector (FIG. 20).

(61) To check recombinant protein production, A549 cells were transduced at an MOI=300 either with the control vector Lt.eGFP (93% of eGFP expressing cells) or Lt.coTRII.SE/Fc (47.53% of eGFP expressing cells) and Mock transduced (FIG. 21).

(62) To verify the presence of human IgG1 mRNA in Lt.coTRII-SE/Fc transduced cells, total mRNA of Mock transduced (vehicle), Lt.eGFP transduced and Lt.coTRII-SE/Fc transduced cells was extracted and RT-PCR assays were performed using specific primers for human IgG1-Fc (FIG. 22). As expected, human IgG1 Fc domain mRNA was only detected in Lt.coTRII-SE/Fc transduced A549 cells.

(63) Additionally, to verify the presence of the TRII-SE/Fc protein both in cell lysates and supernatants, total proteins from Mock, Lt.eGFP and Lt.coTRII-SE/Fc transduced cells lysates and supernatants were western blotted (FIG. 23) using a monoclonal antibody, capable of specifically detecting TRII-SE. In this way, a predicted protein of circa 50 kD could be detected, which included 18 kD of TRII-SE plus 35 kD of the human IgG1 Fc domain, both in cell supernatants and lysates of Lt.coTRII-SE/Fc-transduced cells only.

(64) This invention is better illustrated in the following examples, which should not be construed as limiting the scope thereof. On the contrary, it should be clearly understood that other embodiments, modifications and equivalents thereof may be possible after reading the present description, which may be suggested to a person of skill without departing from the spirit of the present invention and/or the scope of the appended claims.

EXAMPLES

Example 1: Isolation, Cloning and Sequencing of the TRII-SE Isoform

(65) Human adipose derived mesenchymal stromal cells (hASC) were obtained from 20 g subcutaneous fat following the protocol described by Zuk et al. (Zuk P A, et al. Mol Biol Cell 13: 4279-95, 2002) and cultured in the presence of DMEM supplemented with 10% human serum and 1% L-glutamine. Epstein Barr Virus immortalized lymphoblastoid cells were generated from peripheral blood mononuclear cells as described (Protocols in Immunology) and cultured with RPMI medium. Human A459 (lung adenocarcinoma), HT1080 (fibrosarcoma), Caco-2 (colorectal carcinoma), Hep 3B (hepatocellular carcinoma), Jurkat (acute lymphoblastoid leukemia), HEK293 (human embryonic kydney), and 293T cell lines were cultured in DMEM supplemented with 10% FCS and 1% penicillin/streptomycin. The cells were cultured in a humidified 5% CO.sub.2 incubator at 37 C.

(66) Purification of Different Leukocyte Subpopulations

(67) Granulocytes, lymphocytes and monocytes were isolated from heparinized peripheral blood by Ficoll-Paque PLUS (GE Healthcare Bio-Sciences AB) gradient centrifugation. After centrifugation two fractions were obtained, one containing granulocytes/erythrocytes and another with peripheral blood mononuclear cells (PBMC). To obtain granulocytes, erythrocytes were lysed with KCl 0.6 M. PBMCs were labelled with anti CD3.sup.+, CD14.sup.+, and CD19.sup.+ monoclonal antibodies conjugated with magnetic microbeads (Miltenyi Biotech) and separated using MS columns (Miltenyi Biotech) in a MiniMACS magnet (Miltenyi Biotech). Viable cells were determined by Trypan blue dye exclusion and counted in an hemocytometer. The purity of B- and T-lymphocyte and monocyte sub-populations was determined by flow cytometric analysis using a FACSCalibur flow cytometer (BD Biosciences). Cell sub-populations homogenized in RNA Lysis Buffer (SV Total RNA Isolation System, Promega) were stored at 80 C. until RNA extraction.

(68) Cloning and Sequencing of PCR Fragments

(69) TRII PCR fragments were cloned by insertion into the pGEM-T Easy plasmid (Promega Corporation WI, USA) under the conditions established by the manufacturers and E. coli transformation. TRII PCR fragments were sequenced by using M13 forward and direct primers in a DNA sequencer ABI 3130 (Applied Biosystems Inc, CA, USA).

Example 2: Cloning of the Codon Optimized (Co) TRII-SE/Fc Isoform Fusion Construct

(70) The TRII-SE coding sequence containing an AgeI site was codon optimized, the stop codon was deleted and a Kozak sequence included (Epoch Biolabs Inc. Texas, USA). The human IgG1 Fc coding sequence was obtained by RT-PCR from total blood mRNA using specific oligonucleotides as primers (forward: 5AGA TCT GAC AAA ACT CAC ACA TGC 3 (SEQ ID No. 8) and reverse: 5 GAT ATC TTT ACC CGG AGA CAG G 3 (SEQ ID No. 9)), containing a BglII site (forward primer) and EcoRV (reverse primer), to allow in frame fusion to TRII-SE and to the lentiviral vector, respectively. The fusion construct (coTRII-SE/Fc) of 951 bp AgeI/EcoRV comprises 258 bp of the coTRII-SE fused in frame with 693 bp of the human IgG1-Fc.

Example 3: Lentiviral Vectors

(71) The cDNA encoding the three human TRII isoforms were cloned into the pRRLsin18.cPPT.WPRE lentiviral vector, generating the transfer vectors pRRLsin18.cPPT.CMV-TRII-SE.ireseGFP.WPRE, pRRLsin18.cPPT.CMV-TRII-DN.ireseGFP.WPRE, and pRRLsin18.cPPT.CMV-coTRII-SE/Fc.ireseGFP.WPRE. Vesicular Stomatitis Virus G protein-pseudotyped lentiviruses (VSV-G) were generated by transient transfection of the transfer vectors together with the envelope plasmid (pCMV-VSVG), the packaging plasmid (pMDLg/pRRE) and Rev plasmid (pRSV-REV), into the 293T cell line, as previously described (R. A. Dewey, et al. Experimental Hematology 34: 1163-1171, 2006). The supernatant was harvested once every 12 hours for 48 hours and frozen in aliquots. Viral titers were determined by transducing A549 cells (yielding 10.sup.7 infectious particles per milliliter). The pRRLsin18.cPPT.CMV-eGFP.WPRE lentiviral vector was used as control.

Example 4: RT-PCR and RT-qPCR

(72) Total RNA from different primary cultures and cell lines was isolated using the Absolutely RNA kit (Stratagene, La Jolla, Calif., USA). First-strand cDNA was synthesized by mixing 1 g of DNA free total RNA, 50 pmol primer p(DT)15 (Roche Diagnostics GmbH, Mannheim, Germany), 0.5 mM deoxyribonucleotide triphosphate, 5 mM dithiothreitol, and 1 U Expand Reverse Transcriptase (Roche Diagnostics GmbH). cDNA corresponding to different isoforms of TRII receptor was detected by PCR amplification in the presence of Expand High Fidelity polymerase (Roche Diagnostics GmbH), 0.2 mM dNTPS, and 0.5 OA of each primer (forward: 5ACCGGTATGGGTCGGGGGCTGCTC3 (SEQ ID No. 10) and reverse: 5GTCGACTCAGTAG CAGTAGAAGATG3 (SEQ ID No. 11) for 35 cycles using the following PCR conditions: 1 min. at 95 C., 1 min. at 55 C., and 1 min. at 95 C.

(73) Quantitative RT-PCR was performed on diluted cDNA samples with FastStart Universal SYBR Green Master (Rox) (Roche Applied Science) using the Mx3005P Real-Time PCR Systems (Stratagene) under universal cycling conditions (95 C. for 10 min; 40 cycles of 95 C. for 15 s; then 60 C. for 1 min). All results were normalized to GAPDH mRNA levels and further the results were analyzed using the MxPro QPCR computer program and Infostat statistical computer program (Di Rienzo J. A., et al. InfoStet versin 2010. Grupo InfoStet, FCA, National University of Cordoba, Argentina. URL, http://www.infostat.com.ar).

Example 5: In Vitro Bioassay for the TRII-SE Isoform and Other Isoforms Using the MTT Proliferation Assay

(74) A549 cells were transduced with lentiviral vectors at a multiplicity of infection (MOI) of 50 in the presence of 8 g/ml polybrene. Percentage of eGFP positive cells was measured in a FACscalibur (Becton Dikinson) cytometer.

(75) Cells were harvested, counted, and inoculated at the appropriate concentrations into 96-well plates using a multichannel pipette. After 24 hr, TGF-1 (10 ng/ml and 20 ng/ml; Sigma) was added to the culture wells, and cultures were incubated for 24 hr and 48 hr at 37 C., under an atmosphere of 5% CO.sub.2. MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) (Sigma) solution at a concentration of 5 mg/ml was added to the media and the cells were further incubated for 4 hr. After replacing 100 l of supernatant with 100 l of DMSO, the absorbance of each well was determined at 540 nm with a SEAC (Sirio S) photometer (Italy). The percentage of cell survival was defined as the relative absorbance of treated versus untreated cells.

Example 6: Transduction and Flow Cytometry

(76) A549 and hASC cells were transduced at an MOI of 50 and 200 respectively, with the different lentiviral constructs, in the presence of 8 g/ml polybrene (Sigma). Forty-eight hours after transduction, cells were harvested, washed in phosphate-buffered saline (PBS) supplemented with 10% fetal calf serum and the percentage of eGFP positive cells was analyzed by flow cytometry (FACscalibur, BD).

Example 7: Protein Immunoblot (Western-Blot)

(77) For Western blot analysis, both 20 l and 100 l of cell supernatant were loaded on 10% SDS-polyacrylamide gels, separated by electrophoresis and blotted onto Immovilon PVDF membranes (Millipore Corporation, Bedford, Mass., USA). The membrane was exposed to anti-TRII monoclonal primary antibody (clone C-4) (Santa Cruz, Biotechnology) diluted 1/200, or the monoclonal antibody IM 0577 (unprotected)], capable of specifically detecting TRII-SE, diluted 1/500. Horseradish peroxidase (HRP)-conjugated goat anti-mouse antibody (Becton Dickinson GmbH) diluted 1/10000 was used as secondary antibody. Protein detection was performed with the Amersham ECL Plus Western blotting detection reagents (Amersham Buchler GmbH, Germany) in a Typhoon 9410, Variable Mode Imager (GE Healthcare Bio-Sciences AB, Uppsala, Sweden).

Example 8: DNA and Protein Sequence Analysis

(78) cDNA sequences belonging to the different TRII isoforms were used and the predicted protein sequences and statistics were obtained using the EditSeq software (DNAstar, Inc. Madison, Wis., USA). Both the DNA and the predicted protein sequences belonging to the TRII-SE cDNA were aligned to known isoforms of the human TRII receptor (A and B) using the MegAlign software (DNASTAR, Inc. Madison, Wis., USA).

Example 9: Analysis of Cytokines and Chemokines Secreted by hASC Cells

(79) A cytokine/chemokine array kit G5 (Ray Biotech Inc., Norcross, Ga.) was used to detect a panel of 80 secreted cytokines as recommended by the manufacturer. hASCs P7 untransduced or transduced with lentiviral vectors were grown for 72 h in a medium supplemented with 0.1% BSA. Supernatants were collected, filtered and frozen after collection. For densitometry analysis of the arrays, Typhoon 9410 Variable mode Imager (GE Healthcare Life Sciences) was used, and signal intensity values were measured using the Image analysis software ImageQuant TL 7.0 (GE Healthcare Life Sciences). Microarray data were analyzed with RayBio Antibody Array Analysis Tool. Good data quality and adequate normalization were ensured using internal control normalization without background. Any 1.5-fold increase or 0.65-fold decrease in signal intensity for a single analyte between samples or groups may be considered a measurable and significant difference in expression, provided that both sets of signals are well above background (Mean background+3 standard deviations, accuracy99%).

Example 10: Generation of Monoclonal and Polyclonal Antibodies Raised Against Human TRII-SE

(80) Antibodies were generated by Rheabiotech, Campinas, Brazil. Immunization of both rabbit (polyclonal antibody) or mice (monoclonal antibody), was performed using a Multiple Antigene Peptide System (MAPS) with 8 identical copies of a peptide containing the 13 amino acids (FSKVHYEGKKKAW) (SEQ ID No. 12), which are only found in TRII-SE and not in the other splicing variants of the receptor. The monoclonal antibody IM-0577 was developed in mice and purified by protein G affinity chromatography. Antibodies specificity was assayed by indirect ELISA by sensitization with antigen at a concentration of 5 g/ml in Carbonate Buffer 0.2 M, blocked by PBS/BSA and detected with serial dilutions (1:1000-1:64000) of the specific antibody. The ELISA test was developed with a Horseradish peroxidase (HRP)-conjugated secondary antibody together with H.sub.2O.sub.2/OPD as chromogenic substrate, and detected by absorbance at 492 nM.

Example 11: In Vivo Study of Articular Cartilage Damage by Ciprofloxacin (CPFX) and the TRII-SE Isoform

(81) Male 24-day-old Wistar rats were housed under controlled conditions at 211 C. with 50%5% relative humidity and a constant light-dark schedule (light, 8 a.m. to 8 p.m.). Food and tap water was provided ad libitum. The rats received ciprofloxacin hydrochloride on day 24 by oral administration of 200 mg/kg of body weight during 10 days. The animals were examined for clinical abnormalities including motility alterations and weighted during the treatment period.

(82) On day 14 after ciprofloxacin treatment, 50 l viral vectors were injected intra-articularly with either Lt.coTBRII-SE/Fc (2.3510.sup.6 transducing Units, TU) or Lt.eGFP (610.sup.6 TU). Control animals without ciprofloxacin were treated in the same manner.