WO2013080184A1

WO2013080184A1 - Chimeric protein comprising proinsulin fused to thioredoxin

Info

Publication number: WO2013080184A1
Application number: PCT/IB2012/056879
Authority: WO
Inventors: Aldana TRABUCCHI; Edgardo Poskus; Silvina N. VALDEZ
Original assignee: Consejo Nacional De Investigaciones Cientificas Y Tecnicas (Conicet); Inis Biotech Llc
Priority date: 2011-12-01
Filing date: 2012-11-30
Publication date: 2013-06-06
Also published as: AR089065A1

Abstract

The present invention relates to novel fusion proteins. More specifically, the present invention relates to a fusion protein comprising proinsulin fused via a peptide bond to thioredoxin, useful in the detection of anti-proinsulin (PAA) and/or anti-insulin (IAA). The present invention also relates to a method for producing the fusion protein, to DNA sequences encoding the fusion protein of Thioredoxin-Proinsulin, to expression vectors containing such DNA sequences, to cells transformed with said expression vector, to a process for producing proinsulin, to a process for producing insulin, and to methods for detecting and characterizing anti-proinsulin autoantibodies (PAA) and/or anti-insulin autoantibodies (IAA).

Description

CHIMERIC PROTEIN COMPRISING PROINSULIN FUSED TO THIOREDOXIN

The present invention relates to new fusion proteins. More specifically, the present invention relates to a fusion protein comprising proinsulin fused by a peptide bond to thioredoxin, useful in the detection of anti-proinsulin (PAA) and / or anti-insulin (IAA) autoantibodies. The present invention also relates to a method for producing the fusion protein, to DNA sequences encoding the Thioredoxin-Proinsulin fusion protein, to expression vectors containing said DNA sequence, to cells transformed with said expression vector , to a procedure to produce proinsulin, to a procedure to produce insulin, and to methods to detect and characterize anti-proinsulin autoantibodies (PAA) and / or anti-insulin (IAA).

Field of the Invention

Diabetes Mellitus (DM) and its complications constitute the third leading cause of death worldwide in industrialized countries, after cardiovascular diseases and cancer. Around 6-8% of the world's population suffers from some form of the disease. Although the disease can often be controlled with physical exercise, diet and oral antidiabetics, before forms or stages with acute symptoms, patients need replacement insulin therapy, and currently about 0.7% of the world population suffers from insulin DM. dependent with a number of constantly growing patients. Despite these therapeutic approaches, complications can occur that in advanced stages significantly reduce life expectancy. This pathology is the leading cause of kidney disease, blindness and amputation, and one of the major causes of disease.

Cardiovascular and premature death in developed countries such as the United States (Jahromi and Eisenbarth, 2007).

DM is not a single disease, but is a heterogeneous group of pathologies that, due to numerous causes, leads to the alteration of carbohydrate metabolism, and is characterized mainly by hyperglycemia. The The American Diabetes Association (ADA) currently defines 4 main forms of Diabetes Mellitus: type 1, type 2, diabetes due to other known causes and gestational diabetes. The first two, type 1 DM and type 2 DM, are considered the main types and are based on clinical phenotypes. In the latter, the first symptoms typically manifest in adult patients. This type of diabetes, which is the most prevalent, is a consequence of the resistance of peripheral tissues to the action of insulin, associated in varying degrees with secretory dysfunction.

Type 1 DM in turn is classified as type 1A (autoimmune) or type 1 B diabetes

(idiopathic, without demonstrable autoimmunity components). Type 1A DM is the most frequent, typically appearing in childhood-youth and is characterized by the destruction of beta-pancreatic cells, leading to complete insulin deficiency. Since the last decades of the twentieth century, numerous specific autoantigens have been identified, such as insulin, proinsulin, insulin peptide B9-23, in addition to membrane-associated proteins such as tyrosine phosphatase IA-2 and its protein homologous to fogrin ( ΙΑ-2β).

As autoimmune cell aggression (insulitis) is not easy to demonstrate directly, for the diagnosis of type 1 A DM, the search for the associated humoral immune response, evidenced by the detection of specific circulating autoantibodies for beta-pancreatic cells, is used and its products The associated humoral autoimmunity develops over a period ranging from months to years as a result of an environmental trigger that acts on individuals with genetic susceptibility (Graves et al., 2003; Norris et al., 2003), and the onset of autoantibodies is the first detectable sign of emerging autoimmunity developed in the islets of the pancreas, autoantibodies related to type 1 DM can be detected several years before the clinical debut of the disease (Juhl and Hutton, 2004). Unlike other human autoimmune diseases, such as Myasthenia Gravis or Graves' Disease, in autoimmune DM autoantibodies are not pathogenic themselves but arise as an epiphenomenon and therefore serve as indirect markers of the cellular damage suffered by the endocrine pancreas during the prodromal period of the underlying insulitis (Eskola et al., 2003).

Currently, several major autoantibodies related to the disease are recognized that serve to predict the clinical debut of type 1 DM (Knip, 2002). further Of the classic anti-islet autoantibodies (ICA), which actually represent a complex set of several markers, the main autoantibodies include GADA, IA-2A and ΙΑ-2βΑ.

More than three decades ago, the monitoring of autoantibodies associated with immunity-mediated diabetes was limited to the determination of anti-islet cell antibodies (ICAs) by indirect immunofluorescence. After a period of identification and sequencing of the main autoantigens recognized by the humoral immune response (insulin / proinsulin, glutamate decarboxylase and tyrosine phosphatase IA-2), the recombinant technology allowed the development of quasi-quantitative techniques (Mire-Sluis et al. ., 2004) for the valuation of the respective

specific autoantibodies. Currently, anti-insulin and anti-proinsulin autoantibodies (IAA / PAA) remain highly useful for diagnostic support purposes, as long as the sample has been taken prior to the start of insulin therapy. The early detection of the IAA / PAA is of interest to identify individuals at risk of contracting type 1 DM. In addition, the determination of the affinity of the IAA / PAA leads to a better stratification of said risk which in the future would allow to mitigate, or even prevent in these individuals the development of the disease.

Insulin is a small molecule of only 51 amino acid residues distributed in two chains, A and B, linked by disulfide bridges. Therefore, it is not surprising that the molecule directly bound to a solid phase is not fully useful for the detection of IAA / PAA present in sera of prediabetic individuals or patients with recent debut of type 1A DM. On the other hand, there is evidence describing that all patients positive for IAA react with Pl in radiometric fluid phase trials (Castaño et al., 1993).

Proinsulin (Pl) is considered an autoantigen with early breakdown of tolerance in the development of type 1 DM, both for NOD mice and for man, and could be considered as the first autoantigen triggering the immune response in pancreatic islets. (Ott et al., 2004). Pl is expressed in beta-pancreatic cells. The initial insulin messenger RNA transcript encodes preproinsulin, a precursor of 10 amino acids (preproinsulin) with a leading sequence of 24 amino acids. After the signal sequence that directs the protein to the endoplasmic reticulum is clivated, it becomes an 86 amino acid (Pl) polypeptide. The Pl is packaged within the secretory granules for export from the endocrine pancreas. Within the secretory granules, Pl is cleaved in dibasic residues with prohormone convertase, these being removed by carboxypeptidase H, and thus the Pl is converted into insulin and C-peptide (Figure 1). Approximately 5% of the Pl remains intact, or partially processed and is co-secreted by the pancreatic beta cell along with insulin and C-peptide.

Preproinsulin is a single chain polypeptide, with which insulin chains A and B will be linked by peptide C (Mackin, 1998). During in vivo synthesis, native folding and concomitant disulfide bridge formation are achieved. Then the Pl is converted to insulin after the enzymatic processing in which the C-peptide is clivated.

Mature insulin cannot be efficiently produced in its native conformation using recombinant techniques in prokaryotic expression systems, essentially because the formation of disulfide bridges is only favored in the normal context of beta cells during the biosynthesis of Pl.

E. coli is the best characterized and most used prokaryotic system for the expression of recombinant proteins (Baneyx 1999; Pines and Inouye, 1999). The E. coli expression system implies a rapid generation of biomass and very low cost culture conditions, added to its versatility due to the high number of cloning vectors available and different strains used as hosts. However, many of the proteins used in therapy and diagnosis, including Pl / insulin, have essential disulfide bridges for proper conformation, and often form inclusion bodies when expressed in the cytoplasm of E. coli. This can be seen as a benefit since it can facilitate the purification of the product when in vitro refolding is feasible and leads to the correct formation of the disulfide bridges.

Alternatively, the protein of interest can be directed to the periplasm where its native conformation will be achieved thanks to the oxidizing conditions of the

compartment. The major drawback of this last strategy is the limited capacity that E. coli has to secrete the protein to the periplasm and the limited space of the compartment.

E. coli is not able to secrete high amounts of protein naturally (Francetic et al., 2000); In addition, transport to the periplasm or culture medium is a particularly complex process (Economou 1999; Pugsley et al., 2004). Translocation of proteins through cytoplasmic membranes requires a signal peptide, but the mere presence of said signal sequence does not always guarantee efficient protein translocation (Kajava et al., 2000; Khokhlova and

Nesmeianova, 2003). However, different strategies for producing Pl in soluble form or in the form of inclusion bodies (Cl) have been described in Escherichia coli (Sung et al., 1986; Kang and Yoon, 1991; Tang and Hu, 1993; Winter et al. ., 2000; Mergulháo et al., 2004). Although high-performance Pl molecules such as Cl have been produced, the recovery processes needed to achieve the correct formation of the disulfide bridges during refolding, and the subsequent purification steps, are critical factors that make the process is extremely complex and difficult to achieve. Alternatively, Pl has been produced as a secretion protein in E. coli by recombinant techniques in which the protein is directed to the periplasmic space, using the appropriate signal sequence. When secreted to the periplasm, the half-life of Pl increases from 2 to 20 minutes (Talmadge and Gilbert, 1982). The oxidizing environment of the periplasm also contributes to the correct formation of Pl disulfide bridges. However, in this case, the yield achieved for correctly folded Pl was very low, compared to that obtained by intracellular production (Chan et al., 1981; Talmadge et al., 1981; Winter et al., 2000; Malik et al., 2007).

The first work on the efficient secretory expression in E. coli of a modified Pl was published by Kang and Yoon (1994). The authors generated a construction called ZZ-PI in which the C-peptide was either completely deleted or drastically shortened (subtracting only 1-1 1 residues), which led to a significantly increased expression yield. A similar approach was recently described using unmodified Pl fused to none, one or two Z domains, and thus the potential bottleneck that is generated in protein secretion could be circumvented (Mergulháo et al., 2004). Another strategy used was to fuse the Pl to a periplasmic oxide-reductase disulfide enzyme (DsbA) (Winter et al., 2000). On the other hand, alternative expression systems such as Bacillus subtilis, Streptomyces lividans and Saccharomyces cerevisiae have been used to achieve the synthesis and secretion of Pl to the culture medium (Thim et al., 1986; Koller et al., 1989; Novikov et al. ., 1990; Kjeldsen, 2000; Olmos-Soto and Contreras-Flores, 2003), but the yields achieved were not significantly higher than those obtained in E. coli.

The disadvantages found for the expression of human recombinant proteins in prokaryotic systems led to the use of eukaryotic systems, which generally produce the proteins actively, but are inefficient in large-scale production with high yield and low cost. Therefore, the production of Pl in eukaryotes, and in particular in E. coli, remains an appropriate but challenging route.

From all of the above it is evident that there is a need to have methods for the efficient production of correctly folded proinsulin, which can be used for example for analytical purposes such as the detection of IAA / PAA in the population at risk of type DM or for the production of insulin useful in insulin replacement therapy.

Brief Description of the Invention

The problems set forth above, and others that will be apparent from the description that follows, are resolved in accordance with the present invention by a thioredoxin-proinsulin fusion protein. Said fusion protein can be expressed efficiently and with a correct folding of the proinsulin portion in a prokaryotic system, retains a specific affinity for IAA / PAA comparable to that of proinsulin, and can be used as a starting material in the production of both proinsulin and insulin.

Therefore, an object of the present invention is a fusion protein characterized in that it comprises proinsulin fused by a peptide bond to thioredoxin. In a preferred embodiment, the fusion protein has the sequence as shown in SEQ ID NO .: 1.

Another object of the present invention is a polynucleotide, preferably a DNA molecule, which encodes the thioredoxin-proinsulin fusion protein of the invention, preferably a DNA molecule comprising the sequence SEQ ID NO .: 2. Another object of the present invention is a method of producing the thioredoxin-proinsulin fusion protein comprising (a) transforming a prokaryotic cell with an expression vector comprising a DNA sequence encoding the thioredoxin-proinsulin fusion protein, ( b) culturing said cell under conditions that allow the expression of the fusion protein, and (c) recovering the fusion protein. According to a particular preferred embodiment, the prokaryotic cell is Escherichia coli. According to the process for producing thioredoxin-proinsulin of the present invention, the thioredoxin-proinsulin fusion protein is recovered from the soluble intracellular phase and / or from the inclusion bodies. In a preferred embodiment, the fusion protein recovered from the inclusion bodies is subjected to an in vitro refolding step,

preferably an oxidative refolding by dialysis at 4 ° C against refolding or refolding buffer.

Another object of the present invention is an expression vector comprising a DNA sequence encoding the thioredoxin-proinsulin fusion protein of the invention. In a preferred embodiment, the expression vector comprising a DNA sequence encoding the thioredoxin-proinsulin fusion protein of the invention operably linked to the P _L promoter of bacteriophage λ.

Another object of the present is a transgenic cell characterized in that it is a prokaryotic cell that contains a DNA sequence encoding the thioredoxin-proinsulin fusion protein of the invention. In a preferred embodiment of the invention, the transgenic cell is an Escherichia coli cell. Preferably, the transgenic cell comprises an expression vector comprising a DNA sequence encoding the thioredoxin-proinsulin fusion protein of the invention operably linked to the P _L promoter of bacteriophage λ.

Still another object of the present invention is a method for producing proinsulin from the thioredoxin-proinsulin fusion protein of the invention, comprising the steps of (a) culturing a microorganism that expresses thioredoxin-proinsulin, and (b) cloning the Proinsulin thioredoxin, preferably through the use of enterokinase. Another object of the present invention is a method for producing insulin from the thioredoxin-proinsulin fusion protein of the invention comprising the steps of (a) culturing a microorganism that expresses the fusion protein of claim 1 under conditions that allow the expression of the fusion protein, (b) cloning the thioredoxin to obtain proinsulin, and (c) cloning the C-peptide of the proinsulin to obtain insulin; wherein the clivating of step (c) is preferably performed by the use of trypsin and / or carboxypeptidase B.

It is still another object of the present invention a method for detecting in vitro anti-proinsulin (PAA) and / or anti-insulin (IAA) autoantibodies in a biological sample, which comprises bringing the biological sample into contact with the thioredoxin fusion protein. proinsulin of the invention.

An object of the present invention is also a method for determining in vitro the concentration and affinity of anti-proinsulin (PAA) and / or antiinsulin (IAA) autoantibodies present in a biological sample, characterized in that it comprises contacting a biological sample with the thioredoxin-proinsulin fusion protein of the invention. Preferably, the method for detecting autoantibodies is a Surface Plasma Resonance (SPR) method. More preferably, the thioredoxin-proinsulin fusion protein of the invention is immobilized on a chip having a carboxymethyl-dextran matrix.

Description of the figures

Figure 1. Scheme of synthesis of insulin and C-peptide from pre-proinsulin.

Figure 2. Schematic representation of the TrxPI fusion protein. The residuals in white represent those corresponding to Trx, the residuals in gray represent those of Pl.

Figure 3. Schematic development of the ELISA for titration of HPI-1 and HPI-2 polyclonal sera. Figure 4. Titration of rabbit polyclonal sera against proinsulin (A) and insulin (B) by ELISA. For each dilution value, the left bar is HPI-1, the center bar is HPI-2, and the right bar is preimmune serum.

Figure 5. WB made with the Pl and insulin (Ins) antigens run on SDS-PAGE with (A) and without βΜΕ (B). The sera used after the transfer were: rabbit preimmune serum (CNS) and the same HPI-1 and HPI-2 sera applied in the ELISAs of Figure 24. It was revealed with a rabbit anti-lg peroxidase conjugate.

Streets 1: Molecular weight markers. In (A) 3 and 4 a band compatible with the expected molecular weight for Pl (9 kDa) is observed in the presence of βΜΕ, while in the absence of this, in (B) 3 and 4, two bands are observed, one 9 kDa and another close to 20 kDa that could be a consequence of a dimerization of Pl.

Figure 6. Dose-response curves obtained by RIA using rabbit polyclonal serum HPI-2 and standard Pl. (A) Graph from which KO was calculated. (B) Graph from which qO was calculated.

Figure 7. Immobilization of standard Pl on the surface of the CM5 sensor chip.

Figure 8. Sensorgrams representative of the kinetic adjustment of the interaction of the specific antibodies of the HPI-2 serum against standard Pl. From this graph, the respective kinetic constants of association, dissociation and affinity constant Ka-PI were calculated.

Figure 9. Schematic development of the pGEM-3Zf-PPI construction.

Figure 10. Schematic development of the pGEM-3Zf-PI construction.

Figure 11. Schematic development of the pTrx-PI construction.

Figure 12. Analysis of the expression of TrxPI in E. coli by SDS-PAGE and WB of the LT without inducing and induced at different times. (A) TrxPI expression in strain GI698, (B) TrxPI expression in strain GI724. The arrow indicates the expected molecular weight band for the designed protein (~ 22 kDa).

Streets: 1- Molecular weight markers, 2- LT without induction, 3- LT 1 h post induction, 4- LT 2h post induction, 5- LT 3h post induction, 6- LT induced overnight. Figure 13. Analysis of the TrxPI fractions by SDS-PAGE (A) and WB (B) of the different purification steps from the soluble intracellular fraction. The arrow indicates the band of interest.

Streets: 1- Molecular weight markers, 2- LT without induction, 3- LT 3h post induction, 4- Total FIS, 5- FIS not retained, 6- washing with βΜΕ 1 mM, 7- FIS purified with βΜΕ 100mM.

Figure 14a. Analysis by SDS-PAGE (A) and UV (B) spectral scanning of TrxPI recovered and purified from Cl.

Dotted line: Raw spectrum data, Dotted and striped line: data corrected by subtraction of the dispersion.

Figure 14b Analysis of recombinant TrxPI after treatment with enterokinase. WB revealed with HPI-2 or anti-Trx serum.

Streets: 1- and 3- TrxPI with no treatment with EK; 2- and 4- TrxPI treated with EK. The red arrow indicates the band compatible with the expected molecular weight for TrxPI (22 kDa), the blue arrow indicates the band compatible with the expected molecular weight for Trx (12 kDa), the green arrow indicates the band compatible with the expected molecular weight for Pl (9 kDa).

Figure 15. SDS-PAGE (A) and WB (B) of samples from uninduced and induced LT, FIS and Cl corresponding to strain GI724-TrxPI (the arrow indicates the ~ 22 KDa band compatible with the presence of TrxPI).

Streets: 1- Molecular weight markers, 2- LT without induction, 3- LT 3h post induction, 4- FIS 3h post induction, 5- Cl 3h post induction.

Figure 16. TrxPI analysis by mass spectrometry. The circle indicates the major peak compatible with the expected molecular mass for TrxPI.

Figure 17. Scheme of the peptides resulting from the digestion of TrxPI with V8 protease. The residuals in white represent those corresponding to Trx, the residuals in gray represent those corresponding to Pl. The arrows indicate the cleavage sites of V8. Numbers 1 through 12 indicate the peptides generated by digestion. Figure 18. Chromatographic profile of digestions with V8 obtained with RP-HPLC. The upper panel corresponds to standard Pl, the lower panel to TrxPI. The full line represents the protein treated with V8, the dotted line represents the protein treated with V8 and reduced with DTT. The arrows indicate the peptides evaluated by MALDI-TOF

Figure 19. Schematic development of the preincubation ELISA.

Figure 20. Pl displacement curve in the presence of increasing amounts of Pl (o) or TrxPI (·) for rabbit polyclonal serum HPI-2.

Figure 21. Displacement curve of [35S] PI in the presence of increasing amounts of Pl (o) or TrxPI (·) for rabbit polyclonal serum (A) and a pool of sera from PAA + patients (B).

Figure 22. Schematic development of radioimmunoassay.

Figure 23. RBA performed in 30 normal control sera (A), in 30 PAA + sera (B), in 30 PAA + sera in the presence of micromolar concentrations of TrxPI from FIS (C) and in 30 PAA + sera in the presence of micromolar concentrations of TrxPI recovered from Cl (D). The results are expressed in SD scores. The dotted line represents the cut-off value of the test.

Figure 24. Schematic development of the trial using SPR.

Figure 25. Correlation between the values obtained from K _a using standard TrxPI or Pl as immobilized antigens on the surface of the sensor chip. The regression slope was 0.086 ± 0.07, and the correlation coefficient (r ² ) was 0.80.

Figure 26. TrxPI RP-HPLC after enterokinase digestion. (to)

Standard Pl chromatogram, (b) undigested TrxPI chromatogram and (c)

TrxPI chromatogram digested with enterokinase. The arrow indicates the elution peak of the digest fraction at the same time as retention of standard Pl. The insert shows a WB analysis with polyclonal rabbit anti-PI serum from the beak marked with the arrow (lane 1), standard Pl (lane 2), and TrxPI (lane 3).

Figure 27. Schematic development of the anti-Trx chemiluminescence assay. Figure 28. Results of the detection of anti-Trx antibodies in the 51 sera of diabetic patients by chemoiluminescence assay. The results were expressed as SD score (SDs). The 51 patients were negative. The dotted line shows the cut-off value in the test.

Figure 29. (A) Sensorgram and (B) Resonance Units Calibration Curve (RU) vs. Concentration of anti-PI antibodies generated with the final bleeding of HPI-2.

Figure 30. Sensorgram obtained for the different bleeding of rabbit polyclonal serum HPI-2, from which it was possible to calculate the values of q by interpolation in the calibration curve.

Figure 31. Representative sensorgram of the kinetic adjustment of HPI-2 final bleeding, from which the kinetic constants of association, dissociation and equilibrium constant K _a .

Figure 32. Schematic representation of the determination of kinetic (k-ι, k.-i) and affinity (K _a ) parameters by SPR.

Figure 33. Distribution of the signals obtained in 28 children and youth patients and 23 adult patients. (A) Values obtained by SDs RBA (B) q values obtained by SPR and (C) K values obtained by SPR.

Figure 34. Representative sensorgrams obtained for 3 dilutions of the same sample of an infant-juvenile patient (A) and an adult patient (B).

Figure 35. Schematic representation of an Amplification Assay for the determination of (sub) isotypes of specific anti-PI antibodies.

Figure 36. Isotyping of patient samples 1 and 2 by SPR. A) Sensorgram generated by the injection of human anti-lgG. B) Sensorgram generated by the injection of human anti-lgM. C) Sensorgram generated by the injection of anti-human IgA.

Figure 37. Characterization of isotypes (A) and subisotypes (B) of anti-insulin / proinsulin antibodies in patients 1 and 2. Figure 38. Subisotyping of serum samples of patients 1 and 2 by SPR. A) Sensorgram generated by the injection of human anti-lgG1. B) Sensorgram generated by the injection of human anti-lgG2. C) Sensorgram generated by the injection of human anti-lgG3. D) Sensorgram generated by the injection of human anti-glgG4.

Figure 39. Vector map constructed for the expression of TrxPI in E. coli.

Detailed description of the invention

The main problem that is solved with the present invention is that of efficiently expressing recombinant human proinsulin (Pl) in prokaryotes, so as to have said autoantigen as a reagent and as a starting material for obtaining human insulin. This and other problems are solved in accordance with the present invention by creating a thioredoxin-proinsulin fusion protein. This strategy provides a solution to the main problems that have hindered the expression of recombinant Pl in E. coli. By fusing the Pl with Trx, the solubility of the product was greatly increased and a significant fraction of the protein entered the native folding path. In addition to this, TrxPI was less likely than Pl to undergo proteolysis in the bacterial cytoplasm. In addition, the in vitro folding of the chimera was favored by the presence of the Trx motif.

In the context of the present description, "fusion protein" or "chimeric protein" means an artificial protein, that is, it is not produced naturally by any organism, in which at least part of two or more different proteins are bound together by a peptide bond. In the case of the present invention, the fusion protein comprises proinsulin fused by a peptide bond, preferably at its N-terminal end, to thioredoxin.

The expression of Pl as a Trx fusion protein was selected with the purpose of improving the production of correctly folded protein in the soluble intracellular fraction (FIS), protecting it from bacterial proteases. The construction of this type of chimera was based on previous reports that indicated a seemingly solubilizing effect of the Trx motif (LaVallie et al., 1993). Although the Exact mechanism by which this happens, it has been shown that Trx facilitates the soluble expression of various cytokines, peptides and proteins (Chopra et al., 1994; Wilkinson et al., 1995). Specifically, from previous experiences with another of the DM-associated autoantigens, glutamate decarboxylase (GAD) (Papouchado et al., 1997), it was known that this system was suitable for the majority expression of the product in the soluble intracellular fraction, helping to the correct folding and the purification of the same, reason why it was expected that the use of a fusion partner that can direct an efficient translocation of the Pl and in this way increase the chances of the correct folding. In addition to achieving this objective, the expression of Pl as a Trx fusion protein resulted, unexpectedly, in facilitating the in vitro refolding of TrxPI obtained from the inclusion bodies (Cl), apparently thanks to the activity of chaperone of Trx on the refolding of Pl. Chaperones are molecules that temporarily stabilize the deployed or partially folded domains of proteins, preventing the formation of inappropriate intermolecular or intramolecular interactions. They also catalyze isomerization steps that limit the rate of protein folding. For example, the protein disulfide isomerase (PDI) helps the folding of Pl by acting as a chaperone and as an isomerase (Winter et al., 201 1). On the other hand, Burkart et al. (2010) described that Pl, and not insulin, interacts with the chaperone hsp70 DnaK which facilitates its correct folding. However, this mechanism does not provide a universal solution to the problem of the formation of inclusion bodies, since proteins follow different folding pathways and some cases are refractory to this approach.

It is therefore an object of the present invention a fusion protein characterized in that it comprises proinsulin fused by a peptide bond to thioredoxin. The fusion of the Pl to the Trx allows the purification of the protein present in the soluble intracellular fraction (FIS) in a single step by affinity chromatography. In addition, TrxPI can be recovered from inclusion bodies, subjected to in vitro refolding and purified by anion exchange chromatography with high TrxPI yields and high purity. In a preferred embodiment, the thioredoxin portion is of bacterioan origin, and the proinsulin portion in the fusion protein is human proinsulin, or proinsulin with a similarity percentage of at least 90%, more preferably at least 95%, and still more preferably at least 98% with respect to the amino acid sequence of human proinsulin. In its form of Most preferred embodiment, the fusion protein of the present invention has an amino acid sequence as shown in Figure 2 and in SEQ ID NO .: 1.

The fusion protein of the present invention can be obtained as a heterologous protein in a recombinant system. Therefore, another object of the present invention is a polynucleotide encoding the thioredoxin-proinsulin fusion protein. Any polynucleotide encoding the fusion protein of the invention is within the scope of the present invention and its sequence can be easily derived from the amino acid sequence of the fusion protein. Due to the redundancy of the genetic code, more than one polynucleotide can code for a protein with a given amino acid sequence, all of which are within the scope of the present invention. When more than one codon codes for a given amino acid of the protein of the present invention, the most frequent codon for said amino acid is ideally selected in the expression system that is desired to be used for protein expression.

Preferably, the polynucleotide of the invention is a DNA molecule comprising a sequence encoding the fusion protein of the invention in which the proinsulin portion is human proinsulin, and more preferably a DNA molecule comprising a sequence encoding the fusion protein of SEQ ID NO .: 1. More preferably, the polynucleotide is a DNA molecule comprising the sequence SEQ ID NO .: 2.

Another object of the present invention is a process for producing the fusion protein of the invention. Said method comprises transforming a prokaryotic cell with an expression vector comprising a DNA sequence encoding the thioredoxin-proinsulin fusion protein, culturing said cell under conditions that allow the expression of the fusion protein, and recovering the fusion protein. .

Methods for transforming prokaryotic cells with expression vectors are well known in the art, and comprise, for example, cooling in the presence of divalent cations and subsequent thermal shock, and electroporation. The person skilled in the art will know which method of transformation should be used depending on the type of genetic material to be introduced into the cell, although electroporation is preferred when it is desired to introduce large plasmidial DNA into the cell. As explained above, E. coli is the best characterized and most used prokaryotic system for the expression of recombinant proteins, and allows a rapid generation of biomass and very low cost culture conditions. It is also a very versatile system due to the high number of cloning vectors available and different strains used as hosts. Accordingly, in the process for producing the fusion protein of the invention, the prokaryotic cell is

preferably an Escherichia coli cell.

According to a preferred embodiment of the present invention, the fusion protein can be recovered directly from the soluble intracellular fraction, for example by affinity chromatography. The protein isolated in this way has a high degree of purity (90-95%), and is correctly folded. Obtaining from the FIS also has the advantage of allowing purification in a single step. According to another preferred embodiment, in the process for producing the fusion protein of the invention the fusion protein is recovered in addition or exclusively from the inclusion bodies (Cl). Inclusion bodies are aggregates of substances, usually proteins, in the cytoplasm or nucleus of cells. Most of the Trx-PI produced according to the present invention is contained in the Cl. Although traditionally it is considered that the proteins contained in the Cl are incorrectly folded and are therefore of little use (or require refolding processes costly and inefficient in vitro), the fusion of the Trx to the Pl helps in the correct folding of the Pl obtained from the inclusion bodies, allowing the recovery of Trx-PI with a high yield and a degree of purity of 90-95% An in vitro refolding process applicable to Trx-PI obtained from the inclusion bodies is for example described in Valdez et al. (2004), according to which an oxidative refolding is carried out by dialysis at 4 ° C against refolding buffer. The chimera thus refolded can then be purified by ion exchange chromatography and subsequently concentrated by centrifugal filtration (Centricon ®).

The expression vectors used for the transformation of prokaryotic cells and expression of the fusion protein of the present invention also constitute an object of the present invention. The expression vectors of the present invention are for example plasmids, in which a DNA sequence encoding the thioredoxin-proinsulin fusion protein of the invention, preferably the DNA sequence of SEQ ID N0.:2, is operably linked to a promoter.

Preferably, the promoter is regulated. Thus, in a preferred embodiment, protein expression is directed by the bacteriophage λ PL promoter, which is strongly regulated by the bacteriophage repressor. The expression of the repressor is also regulated, since according to this preferred embodiment the vector is propagated in strains of E. coli, where the gene of said repressor is under the control of the trp promoter (for example E. coli GI724 or GI698). When the transformed bacteria grow in a tryptophan-free medium (Trp), the repressor gene is transcribed and this repressor binds to the PL promoter preventing transcription. The expression of the protein of interest is achieved by adding Trp to the medium. Therefore, in a particularly preferred embodiment of the invention, the DNA sequence encoding the thioredoxin-proinsulin fusion protein of the invention is operably linked to the P _L promoter of bacteriophage λ.

Also part of the present invention are cells containing a DNA sequence encoding the thioredoxin-proinsulin fusion protein of the invention. Obviously, because the protein of the invention is an artificial chimeric protein, the cells that contain DNA encoding this protein are transgenic cells, that is, in which the DNA encoding the thioredoxin-proinsulin fusion protein of the invention has been incorporated artificially. Cells that descend from such transformed cells are also within the scope of the present invention, even if the DNA encoding the fusion protein is present in them as a result of inheritance and not necessarily because it has been introduced by genetic engineering techniques. in those particular cells.

The cells of the invention are prokaryotic cells, preferably Escherichia coli cells. Douillard et al. generated (201 1) two expression vectors to overexpress Trx fusion proteins in Lactococcus lactis and demonstrated that it was possible to obtain high levels of heterologous protein expression in soluble form, such as Tuc2009 ORF40, Bbr_0140 and Tuc2009 BppU / BppL which they obtained in insoluble form using L. lactis vectors, so it is also possible that the cells of the invention are from a prokaryotic other than Escherichia coli, for example Lactococcus cells. In a particularly preferred embodiment of the invention, the cells of the invention are Escherichia coli cells comprising a vector in which the DNA sequence encoding the thioredoxin-proinsulin fusion protein is operably linked to the P _L promoter of bacteriophage λ.

There are numerous strains of Escherichia coli available to the public, with E. coli GI724 and E. coli GI698 being two of the preferred strains for expressing the chimeric protein of the present invention.

The thioredoxin-proinsulin of the invention can be used in the production of proinsulin by olive oil of the thioredoxin fraction. This is achieved, for example, by the use of a specific enzyme that is capable of selectively cutting the peptide bond between the two constituents of the fusion protein. Specifically, the fusion protein has a cutting site for enterokinase (EK). Said enzyme recognizes the sequence - (Asp) ₄ Lys (DDDDK) and produces the cleavage after the Lys residue, so that only two amino acids that come from the multiple cloning site remain attached to the Pl. Thus it is possible to obtain a 50% yield on removal of the Trx domain. The PL thus produced can be isolated from the other products of digestion by HPLC using a reverse phase column. It is therefore another object of the present invention a method for producing proinsulin from the thioredoxin-proinsulin fusion protein comprising the steps of (a) culturing a microorganism that expresses thioredoxin-proinsulin, and (b) cloning the thioredoxin of proinsulin preferably by using enterokinase.

The proinsulin obtained by thioredoxin cleavage can be used for analytical and diagnostic purposes, or it can also serve as a starting point for insulin production. Therefore, according to another of its aspects, the present invention comprises a method for producing insulin from the thioredoxin-proinsulin of the present invention, which comprises the steps of (a) culturing a microorganism that expresses the thioredoxin-proinsulin of the present invention under conditions that allow the expression of the fusion protein, (b) cloning the thioredoxin to obtain proinsulin, and (c) cloning the C-peptide of the proinsulin to obtain insulin. One method for cloning proinsulin C peptide and obtaining insulin is described by Gardner et al. (2012). This working group describes the expression of human proinsulin containing a methionine at the N-terminal end (M-proinsulin) in E. co // from which they obtain M-insulin. To clivate Pl, they use trypsin, which cuts at residues R (31) R (32) -E (33) and K (64) R (65) -G (66) which are the binding sites of the chains B / C and C / A of insulin and peptide C, respectively. In vitro Pl cleavage can also be done with trypsin and carboxypeptidase B (Kemmler et al., 1971).

Surprisingly and unexpectedly, the thioredoxin-proinsulin chimera of the present invention exhibits immunochemical properties similar to the unbound proinsulin. The specific recognition between an antibody and its antigen is determined both by the three-dimensional conformation of the bonds formed by the complementarity determining regions (CDR's) in the variable domain of the antibody (which in turn depends mainly on the amino acid sequence in the CDR ^' s and its interaction with certain amino acids in the structure region), such as (in the case of conformational epitopes) by the particular three-dimensional "landscape" in the surface region of the antigen constituting the epitope. Most PAAs present in sera of diabetic patients only recognize epitopes

Constitutional discontinuations of Pl, and given that the topographic structure of the antigenic determinants usually reflects very sensitive interactions not only near but also steric implications, the person skilled in the art would expect that the fusion of a 12 kDa protein such as thioredoxin to a protein of 9kDa as proinsulin will result in the loss at least in a noticeable reduction in

immunoreactivity of proinsulin. The fact that thioredoxin does not exert direct or indirect adverse effects on the immunochemical properties of proinsulin is thus contrary to intuition and constitutes yet another inventive aspect of the present invention.

In summary, the chemical structure of the antigenic determinants is stipulated by particular functional groups of each protein and by their spatial arrangement. If these functional groups are formed by amino acid residues that are continuous in the primary structure of the macromolecule or are residues that are far apart in the primary sequence, but that are grouped in the three-dimensional folding of the protein, the spatial relationship, and in consequently the chemical structure of the antigenic determinant is highly dependent on the folding and conformation of the entire macromolecule. Based on these concepts, it is possible to evaluate the cross-reactivity between standard Pl and the TrxPI chimera against the same population of specific antibodies and to perform a quality control of the recombinant protein generated in the laboratory. When comparing the immunoreactive properties of standard Pl vs. TrxPI (for example by means of an ELISA, RIA or SPR test) it is seen that there is an immunochemical identity between said molecules and, therefore, conserved folds at the level of the epitope or epitopes involved. Moreover, in the case of thioredoxin-proinsulin, there is cross-reactivity of type 1, also called "true cross-reactivity", between both ligands (standard Pl and TrxPI) against the same antibody, that is, the antibodies bind to it. binding site in both antigens, but the affinity of the binding can vary if there are differences in the conformation of the peptide. With this type of cross-reactivity a sufficiently high concentration of the competitor can completely displace the homologous ligand that is in traces of its binding with the antibody (Berzofsky and Schechter, 1981).

From all of the above it is evident that the expression of TrxPI in E. coli is an excellent tool for obtaining the antigen with the correct conformation. In turn, TrxPI is recognized by PAA + sera, indicating its potential use in immunochemical assays and its subsequent application in the detection of the PAA marker in patient sera.

Since TrxPI is immunoreactive against the sera of diabetic patients presenting with PAA, it can be concluded that advantageously and unexpectedly, removal of Trx from the chimera is unnecessary, at least for use as an analytical reagent in the detection of IAA / PAA. Moreover, the presence of Trx helps the orientation in the immobilization of the antigen for example in the cells of the sensor chip for SPR studies when it is based on the covalent attachment of the carboxyl groups on the surface of the sensor chip (by For example, the CM5 sensor chip, which has a carboxymethyl-dextran matrix covalently attached to a gold plate) with the amino groups of the proteins. In this way, the Trx domain provides a greater number of amino groups available for said binding, thus improving the orientation of immobilized Pl and ensuring that all its epitopes are exposed to the

antibodies

Consequently, according to another of its aspects, the present invention comprises a method for detecting in vitro the presence of anti-proinsulin (PAA) and / or anti-insulin (IAA) autoantibodies in a biological sample, which comprises contacting a sample biological with the thioredoxin-proinsulin fusion protein. For "sample" A derivative of at least one body fluid of the individual should be understood. Body fluids include but are not limited to blood, urine, milk, cerebrospinal fluid, and the like. Preferred samples are compositions comprising blood, plasma or serum obtained or derived from an individual, preferably a human individual, that have preferably been processed to be in a condition suitable for the method of the invention. Said method can be, for example, an ELISA, RIA, etc. intended to diagnose DM from the presence of anti-insulin / proinsulin autoantibodies. According to yet another aspect, the present invention comprises a method for determining in vitro the concentration and affinity of anti-proinsulin (PAA) and / or anti-insulin (IAA) autoantibodies present in a biological sample, which comprises contacting a biological sample with the thioredoxin-proinsulin fusion protein of the invention, for example a Resonance method

Surface Plasma (SPR). In this case, the fusion protein is

preferably immobilized on a sensor chip that has a carboxymethyl-dextran matrix.

The different aspects of the present invention will be better understood by the following embodiments. The scope of the invention, however, is not limited to the exemplified, since the purpose of the examples is not to limit the invention but to illustrate particular embodiments of the present invention from which alternative embodiments will be evident to the art expert.

Examples

In general, in the following examples standard protocols of Molecular Biology were used as described in Sambrook et al. (1989), detailing only those that might require a more complete description.

Example 1

Obtaining polyclonal rabbit anti-PI sera Anti-PI antibodies (HPI-1 and HPI-2) were generated by a conventional immunization plan. Briefly, two New Zealand White rabbits were immunized with 0.1 mg of standard Pl emulsified in complete Freund's adjuvant. The initial injection was followed by booster injections with 0.1 mg of Pl in incomplete Freund's adjuvant at 3-week intervals. The rabbits were bled 15 days after booster shots.

Example 2

Characterization of polyclonal rabbit anti-PI sera by ELISA, SDS-PAGE and WB.

To evaluate immunoreactivity and primary interaction parameters

(concentration and affinity constant) of the polyclonal antibodies HPI-1 and HPI-2 by enzyme immunoassay (ELISA), 96-well polystyrene microplates (Maxisorp, NUNC, Roskilde, Denmark) were used which were sensitized with 50 μΙ of one 1 g / ml dilution of Pl or insulin (Elli Lilis) in PBS buffer (0.14 M NaCI; 2.7 mM KCI; 1.5 mM KPO ₄ H _2, 1 mM Na ₂ PO ₄ H 8; pH 7 , 4) overnight at 4 ° C. After 3 washes with PBS, nonspecific reagent sites were blocked with 200 µΙ per well of a 3% skimmed milk powder (LPD) solution in PBS for 2 h at room temperature. After this time, 5 washes were performed with 0.05% PBS-Tween (PBS-T) and incubated with 50 μΙ of serial dilutions 1/10 of the HPI-1 and HPI-2 sera in 3% LPD in PBS- T (1/10 ² to 1/10 ⁷ ) in duplicate, for 1 h at room temperature. After 5 washes with PBS-T, 50 μΙ of a peroxidase-conjugated rabbit anti-gamma globulin antibody (Jackson ImmunoResearch Laboratories, Inc., West Grove, Pennsylvania, USA) diluted 1/500 in LPD 3 was added % in PBS-T, incubating 1 h at room temperature. Finally, the formation of immunocomplexes was revealed colorimetrically with 3, 3 ', 5,5'-Tetramethylbenzidine (TMB; GE Healthcare, Uppsala, Sweden) (Figure 3). The specific antibody titer was calculated as the inverse of the last dilution that gave a positive colorimetric reaction. Preimmune rabbit serum was used as a negative control in the assay. The anti-PI antibody titers of two polyclonal sera (HPI-1 and HPI-2) were greater than 10 ⁶ . In contrast, when ELISA plates were sensitized with insulin, the titre of both sera was less than 10 ² . This indicates that both HPI-1 and HPI-2 were highly specific for Pl (Figure 4), exhibiting marginal levels of cross-immunoreactivity towards insulin and / or denouncing the vestigial presence of subpopulations of genuinely anti-insulin antibodies.

The immunoreactivity analysis of the HPI-1 and HPI-2 polyclonal sera towards the denatured forms of Pl and insulin was performed by SDS-PAGE (Schagger and von Jagow, 1987) and WB. To do this, 4 μg of Pl or insulin diluted in sample buffer (50 mM Tris-HCI, 12% glycerol, 0.005% bromophenol blue, 4% SDS, pH 6.8) with and without β-mercaptoethanol (βΜΕ) were seeded in a 16% crosslinking gel. After the electrophoretic run, the proteins were transferred to a membrane of

nitrocellulose (Tras-Blot Transfer Medium, BioRad Laboratories, Richmond, CA, USA). Free sites were blocked with 3% LPD in TBS (0.05 M Tris-HCI, pH 7.5, 0.15 M NaCl) overnight at 4 ° C. After 3 washes with TBS, the membrane was incubated overnight at 4 ° C with the rabbit polyclonal serum HPI-1, HPI-2 or the preimmune serum, in a 1/100 dilution in 3% LPD in TBS- Tween 20 0.05% (TBS-T). The next day the membrane was washed 5 times with TBS-T, and incubated for 2 h with a rabbit anti-IgG antibody conjugated to Peroxidase (Jackson ImmunoResearch Laboratories, Inc., West Grove, Pennsylvania, USA) in a 1/2000 dilution in 3% LPD in TBS-T. Finally, the development was performed with 0.5 mg / ml chloronaphthol (Sigma-Aldrich, Inc., St Louis, MO, USA) in 17% methanol and H ₂ 0 ₂ 0.005% V / V dissolved in TBS . The titers of both sera were greater than 10 ³ using Pl as an antigen and showed no reactivity towards insulin as an antigen. These results were correlated with those obtained by ELISA (Figure 5).

Example 3

Determination of primary interaction parameters (q ₀ and K ₀ ) of serum

Polyclonal rabbit anti-PI using Radioimmunoassay (RIA).

A radioimmunoassay (RIA) was performed to determine the primary interaction parameters (PIP) of the HPI-2 anti-PI polyclonal serum, which was used as the standard serum for subsequent examples because it was the one with the highest ELISA titre.

Dialing [ ³⁵ S] -PI The recombinant protein [S] Cys-proinsulin ([S] -PI) was produced by transcription and translation in vitro using a rabbit reticulocyte lysate (Promega, Wisconsin, USA) in the presence of [ ³⁵ S] cysteine ( specific activity 1075 Ci / mmol, New England Nuclear, Massachusetts, USA) for 1.5 h at 30 ° C. The translation products were diluted in RBA buffer (50 mM sodium phosphate, 100 mM NaCl, pH 7, 0.1% Aprotinin and 0.1% bovine serum albumin) and seeded on a PD10 column (Amersham Biosciences, New Jersey, USA). ) previously equilibrated with RBA buffer to remove [ ³⁵ S] free cysteine. The percentage of incorporation of [ ³⁵ S] cysteine in the protein was 3-9%.

In order for the disulfide bridges of Pl to form correctly, the radioactive tracer was subjected to a refolding in vitro overnight, favored by a reduction-reoxidation procedure, as previously described (Valdez et al., 2004) . For this, [ ³⁵ S] -PI was diluted to the medium in 100 mM Tris-HCI buffer, pH 8.5, containing 0.2 M dithiothreitol (DTT). The reduction of the protein was carried out for 2 h at room temperature . A buffer change was then made for 50 mM Tris-HCI, pH 8.5, using a PD10 column and 0.1% aprotinin was added. The refolding of the completely reduced protein was carried out by dialysis overnight at 4 ° C against refolding buffer (0.5 M L-Arginine, 50 mM Tris-HCI, 5 mM EDTA, 5 mM GSH and 0 GSSG, 5 mM, pH 9.5) (Valdez et al., 2003).

Finally the [ ³⁵ S] -PI was isolated by RP-HPLC using a 4.6 x 250 mm C18 column (Vydac, California, USA). An aqueous solution of trifluoroacetic acid (TFA) 0.07% was used as solvent A and as solvent B a solution of 90% acetonitrile containing 0.07% TFA. A linear gradient of solvent A with 22% to 33% of solvent B in 20 minutes and 33% to 67% of solvent B in 40 minutes was used for the run, using a flow of 1 ml / min. Fractions of 0.5 ml were collected, which were monitored by liquid scintillation.

HPI-2 titration

To adjust the appropriate concentration of specific interacting antibodies in the displacement assay, 1/10 dilutions were evaluated in RBA buffer of the HPI-2 polyclonal serum (1/10 to 1/10 ⁵ ). That dilution that will cause was selected

approximately 50% non-radioactive antigen binding (Yalow, 1973a, b). The titration test was carried out by incubating 30 μΙ of the diluted serum for 7 days at 4 ° C with 1000 cpm of the tracer [ ³⁵ S] -PI in 90 μΙ of RBA buffer. The immunocomplexes formed were separated with G-Sepharose 4B FF protein (Amersham Biosciences, New Jersey, USA), the precipitates were washed four times with RBA buffer and resuspended in 1% SDS. The radioactivity of the supernatants was measured in an automatic beta counter (Liquid Scintillation Analyzer Model 1600TR, Packard, Canberra, Australia).

The HPI-2 rabbit polyclonal serum presented the highest titre for ELISA. Displacement Test

Said assay was carried out by incubating 30 μΙ of a 1/100 dilution of the polyclonal serum for 7 days at 4 ° C with 1000 cpm of the tracer [ ³⁵ S] -PI in the presence of 90 μΙ of serial concentrations of standard Pl (2.57 pM at 1, 00 μΜ) in RBA buffer. The immunocomplexes formed were separated with G-Sepharose 4B FF protein, the precipitates were washed four times with RBA buffer and resuspended in 1% SDS. The radioactivity of the supernatants was measured in an automatic beta counter. PIPs expressed as medians for concentration (q ₀ ) and affinity constant (K ₀ ) of the HPI-2 polyclonal serum were derived from the graphs B / F = f (Log dose or Log F), where B / F represents the molar ratio of bound antigen and free antigen and F represents the molar fraction of free antigen (Berzofsky and Schechter, 1981).

The parameter K ₀ obtained was 3.34 x10 ⁸ M ^-1 , which was calculated in the graph B / F = f (log F) by interpolation on the axis of the abscissa of the value of free antigen (F) corresponding to the maximum signal, (B / F) _0/2 , and that is equal to 1 / K ₀ . The parameter q ₀ obtained was 5.92 x10 ^-7 M, which was calculated from the RIA graph B / F = f (log dose), where the value of the interpolation in the abscissa (B / F) ₀ / 2 is equal to 1 / K ₀ + q _0/2 (Berzofsky et al., 1993) (Figure 6).

Example 4

Plasma Surface Resonance (SPR)

Standard Pl Immobilization Standard Pl immobilization and subsequent interactions were developed according to the BIAcore T100 biosensor user manual (BIAcore, GE Healthcare, Uppsala, Sweden). The protein was immobilized on the carboxymethyldextran surface of a CM5 sensor chip (Figure 7) using conventional carbodiimide binding chemistry (Johnsson et al., 1991). The carboxyl groups found on the surface of the chip were activated for 7 minutes with 0.2 M N-ethyl-N- (3- diethylmiopropyl) carbodiimide (EDC) and 0.05 M N-hydroxysuccinimide (NHS). Standard pl at a concentration of 30 μg / ml in 10 mM acetate buffer (pH 4.5), was injected into the immobilization surface. Chips with immobilization levels of 2000 or 500 resonance units (RU) were prepared for concentration or affinity tests respectively. Finally, the remaining reactive groups on the surfaces were blocked by injection for 7 minutes of 1 M ethanolamine hydrochloride pH 8.5.

Determination of the affinity constant K _g

To determine the K _a of HPI-2, the serum was diluted 1/10, 1/20 and 1/40 in PBS-T. Each sample was injected for 300 seconds, and the level of binding in was measured after 300 seconds. The tests were carried out at 20 ° C with a flow of 10 μΙ / min. The kinetic association constant (k-ι), the kinetic dissociation constant (k.-i) and the association equilibrium constant (K _a ) were calculated from the analysis of the corresponding sensorgrams using the BIA-evaluation software.

Using this procedure, for the measurement of anti-Pl antibody concentration (here definable as q), the amount of immobilized standard Pl was the

corresponding to 1600 RU. For the affinity measurement (here definable as the association equilibrium affinity constant, K _a , calculable by the BIAcoreH OO processor from the parameters k-ι and k.-i of the sensorgram), the surface of the sensor chip it was prepared with smaller amounts of immobilized antigen (300 RU).

Knowing the parameters q ₀ and K ₀ of the polyclonal HPI-2 serum, determined by RIA, the kinetic tests were carried out to determine the K _a by SPR. The average affinity of HPI-2 towards standard Pl was evaluated in parallel by a conventional fluid phase test (RIA), with which the values obtained were K ₀ . _P i = 3.34 x10 ⁸ M \ and using SPR whose results were K _a . _P | = 2.48 x10 ⁸ M ^" (Figure 8). The affinity measured by the recommended procedures, and with the original nomenclature, for both techniques it was shown in a reasonable agreement, either using standard Pl or the alternative Thioredoxin-Proinsulin chimera, described below, as an immobilized antigen. These results justify, and to a large extent, validate the use of the SPR technique to determine the primary interaction parameters of the PAA present in sera of diabetic patients.

Example 5

Generation of the expression vector for the thioredoxin-proinsulin fusion protein (Thioredoxin-Proinsulin). E. coli transformation and Thioredoxin-Proinsulin expression.

In general, standard protocols of Molecular Biology (Sambrook et al., 1989) were used, detailing only those that require a more complete description.

From the 4.9 kb plasmid pBR328, a 510 bp fragment containing the coding region for human pre-proinsulin was isolated. The restriction enzyme EcoR I (Promega, Wisconsin, USA) was used for this. Said fragment was separated from the rest of the digested vector, or partially digested, by electrophoresis on an agarose gel. Agarose gels were prepared at 0.7% in TBE buffer (45 mM Tris-borate pH 8.0, 1 mM EDTA) containing 0.5 μg / ml ethidium bromide. The electrophoretic run was performed in a horizontal tank (Sigma-Aldrich Techware, Missouri, USA) with a potential gradient of 1-5 V / cm using, TBE as a run buffer (Sambrook et al., 1989). The 510 bp fragment resolved in the agarose gel was recovered by using a silica gel resin (Concert ™ Matrix, Gel Extraction System, Gibco BRL, USA).

The 3.2 kb pGEM ^® -3Zf vector (Promega, Wisconsin, USA) was digested with the same restriction enzyme, EcoR I, and purified in a manner similar to that explained above. Next, the vector was incubated with the DNA fragment

containing the coding sequence for pre-proinsulin (vector: insert 1: 1 ratio), in the presence of T4 DNA ligase (Promega, Wisconsin, USA) to bind the cohesive ends generated by the restriction enzyme. This ligation mixture was incubated for 2 h at 22 ° C and was used to transform competent E. coli strain JM109 (ampicillin sensitive) bacteria by electroporation technique (Micropulser ^® Bio-Rad Laboratories, California, USA) (Sambrook et al., 1989). Then you

selected colonies of transformed bacteria capable of growing in solid medium LB (10 g / l NaCI, 10 g / l tryptone, 5 g / l yeast extract, 1.5% agar), supplemented with 100 μg / μΙ ampicillin (Sigma, Missouri , USA), because the pGEM ^® -3Zf vector confers resistance to said antibiotic. Plasmid DNA was then extracted from a colony developed in LB liquid medium, (Sambrook et al., 1989). The DNA was analyzed by 0.7% agarose gels, after digestion with the restriction enzyme EcoR I. In this way, the pGEM ^® -3Zf vector was isolated with the insert for human pre-proinsulin. Said vector was called pGEM-3Zf-PPI (Figure 9).

The human Pl gene was amplified by PCR from the pGEM-3Zf-PPI vector using the following oligonucleotides as primers.

5 'CCCAGCCATGGCCTTTGTGAACCAACACCTGT 3' (SEQ ID NO .: 3) and

5 'TTTATTC G AG CTCTCTCGGTG C AG G AG G C G 3' (SEQ ID NO .: 4), which include the cutting sites of the restriction enzymes Neo I and Sac I at the 3 'and 5' ends, respectively. For amplification, Pfu polymerase (Promega, Wisconsin, USA) at 94.61 and 72 ° C was used for denaturation, annealing and synthesis, respectively, in 30 cycles. The amplification product was flanked by a Neo I site at the 5 'end and by a Sac I site at the 3' end. The synthesis product, approximately 300 bp, was purified, digested with the restriction enzymes Neo I and Sac I, again purified and ligated into the respective sites of the vector pGem3Zf. The resulting vector, coding for the amino acid sequence of proinsulin, was named pGem3Zf-PI (Figure 10). The identity of the new DNA molecule coding for Pl was corroborated by sequencing at the Cancer Research Center, Chicago University (Chicago, Illinois, USA).

The vector pGem3Zf-PI was digested with restriction enzymes Neo I and EcoR I and treated with S1 nuclease. The fragment containing the Pl gene with 5 'and 3' blunt ends was isolated and ligated to the pTrxFus vector (Invitrogen, Carlsbad, California, USA). The latter was linearized with Sma I and purified prior to ligation. The resulting construct coding for the thioredoxin-proinsulin fusion protein (Thioredoxin-Proinsulin) was called pTrx-PI (Figure 1 1). Example 6

E. coli transformation and Thioredoxin-Proinsulin expression.

In order to obtain a high-performance source of human Pl a prokaryotic expression system was used. The competent bacteria Escherichia coli GI724 and E. coli GI698 were transformed by electroporation with the pTrx-PI construct. The expression of the cloned protein in the pTrxFus vector is directed by the bacteriophage λ PL promoter, which is strongly regulated by the bacteriophage repressor. The expression of the repressor is also regulated, since the pTrxFus vector is propagated in strains of E. coli GI724 or GI698, where the gene of said repressor is under the control of the trp promoter. When bacteria grow in a tryptophan-free medium (Trp), the repressor gene is transcribed and this repressor binds to the PL promoter preventing transcription. The expression of the protein of interest is achieved by adding Trp to the medium.

The competent bacteria E. coli GI724 and E. coli GI698 were transformed by electroporation with the pTrx-PI construct. The bacteria were grown at 30 ° C in induction medium (0.2% casamino acids, 0.5% glucose, 1 mM MgCl2 , and ampicillin 100 μg / ml) up to OD600 = 0.5. Induction was carried out with Trp 100 μg / ml at 37 ° C for strain GI724 and at 20 ° C for strain GI698, for 1 h, 2 h, 3 h and overnight. Recombinant human Pl was expressed as a thioredoxin (Trx) fusion protein bound to the N-terminal (Figure 2).

To obtain the total lysate (LT), the transformed bacteria from 0.5 ml of culture were collected by centrifugation for 15 minutes at 13000 rpm, and resuspended in 0.2 ml of sample buffer for SDS-PAGE. Total lysates (LT) of strains GI724-thioredoxin-proinsulin and GI698-thioredoxin-proinsulin without induction and induced at different times were analyzed by SDS-PAGE and WB revealed with rabbit polyclonal serum HPI-2. Efficient thioredoxin-Proinsulin expression was achieved after 3 h induction at 37 ° C. The decrease in induction temperature at 20 ° C led to the expression of lower levels of protein in strain GI698. The increase in induction time to 16 h decreased total expression with a significant effect on the relative yields for the biosynthesis of the chimera (Figure 12).

Example 7

Recovery of thioredoxin-proinsulin from the FIS of transformed E. coli cultures

To obtain the soluble intracellular fraction (FIS), bacteria from 200 ml of culture were collected by centrifugation for 15 minutes at 7000 rpm, and resuspended in 4.0 ml of lysis buffer I (Na ₃ PO ₄ 50 mM , 100 mM NaCI, pH 7). Then, the cell suspension was subjected to 4 sonication pulses of 30 seconds each, in the presence of protease inhibitors (0.1% aprotinin, 2 mM PMSF, 1 mM EDTA). Next, Triton X-100 was added in a final concentration of 0.1%, and the mixture was incubated at 0 ° C for 10 minutes. The soluble intracellular fraction (FIS) was separated by centrifugation at 10,000 rpm for 15 minutes at 4 ° C.

The thioredoxin-proinsulin fusion protein was purified from the FIS by affinity chromatography against the Trx motif following the protocol of Hoffman and Lañe (1992). The affinity column consisted of a resin on an agarose support containing coupled groups of high affinity for the active site of the Trx (Figure 16). For its preparation, 14.64 mg (8 moles) of 4-amino-phenylarsine oxide (PAO) synthesized in the Chair of Organic Chemistry (School of Pharmacy and

Biochemistry, UBA) from phenylarsonic acid (Sigma-Aldrich, Inc., St Louis, MO, USA) in 800 μΙ of 0.2 M βΜΕ in dimethylsulfoxide and 7 μΙ of concentrated HCI. The solution was rapidly neutralized with 80 μΙ of 1 M NaOH in methanol.

Four hundred microliters of the PAO preparation was incubated with 1 ml of Affi-Gel® 10 resin (BioRad Laboratories, Richmond, CA, USA) for 24 h at room temperature under constant stirring. Resin free sites were blocked with 8 μΙ of ethanolamine for 1 h at room temperature under stirring. Finally, the resin was washed 3 times with 6 volumes of 1M HCI and 3 times with 8 volumes of distilled water. For conservation, it was stored at 4 ° C in 50% ethanol until it was used.

For the purification of thioredoxin-proinsulin, 4 ml of the FIS were incubated with 2 ml of resin previously equilibrated in lysis buffer I with 20 mM βΜΕ for 1.5 ha 4 ° C in constant agitation. The resin was then placed on a column (0.7 cm x 9.0 cm) and washed sequentially with 6 volumes of lysis buffer I containing 1 mM βΜΕ and 3 volumes of lysis buffer I containing 5 mM βΜΕ. The bound proteins were eluted with 2 volumes of lysis buffer I containing 100 mM β.. This single purification step allowed obtaining 1.5 mg of thioredoxin-proinsulin per liter of culture medium, with a purity of 90-95% (Figure 13).

Example 8

Recovery of thioredoxin-proinsulin from inclusion bodies

Thioredoxin-proinsulin was recovered from Cl by means of protocols

solubilization and refolding described previously (Valdez et al., 2004). The inclusion bodies (Cl) from 200 ml of culture were washed with lysis buffer II (50 mM Tris-HCI pH 8, 10 mM EDTA, 100 mM NaCI, 0.5% Triton X-100) and

subsequently with 2 M urea in 0.1 M Tris pH 8.5. The Cls were then solubilized overnight at 4 ° C with 5 M urea in 0.1 M Tris pH 8.5. The fraction recovered from the Cls was separated from their debris by centrifugation at 13,000 rpm for 15 minutes at 4 ° C. The supernatant was subjected to oxidative refolding by dialysis at 4 ° C against refolding buffer (Valdez et al., 2003).

After in vitro refolding of the solubilized Cl, the thioredoxin-proinsulin was dialyzed overnight at 4 ° C against buffer A (20 mM Tris-HCI, pH 8.5) and purified by FPLC by ion exchange chromatography on a Q-column. Sepharose (GE Healthcare, Sweden) (1.5 x 5 cm), using a flow of 1 ml / min, and following detection at 280 nm and 215 nm. Fractions eluting at 30% buffer B (20 mM Tris-HCI, 1M NaCl, pH 8.5) and 40% buffer B were collected. These fractions were then concentrated 10 times by Centricon® (Millipore Corporation, Billerica , Massachusetts, USA).

The refolding chimera was purified by ion exchange chromatography, and concentrated by Centricón®. Its purity and concentration were determined by UV spectral scanning, yielding 10 mg of thioredoxin-proinsulin per liter of culture medium with a purity of 90-95%. The gel and UV spectrum of the protein recovered from Cl are shown in Figure 14.

Example 9

Analysis of the expression and purification of thioredoxin-proinsulin by SDS-PAGE and WB

The proteins obtained from the E. coli LT were analyzed by SDS-PAGE (Schagger and von Jagow, 1987) and WB. Using the strain GI724, induced for 3 h at 37 ° C, the presence of thioredoxin-proinsulin in the FIS, in the periplasmic space (EP) and in the Cl. Was analyzed by SDS-PAGE and WB. To make the comparison feasible, Proteins recovered from the same amount of cells were seeded on all streets of the gel.

Once the electrophoretic run was completed, the protein bands were transferred to a nitrocellulose membrane (Tras-Blot Transfer Medium, BioRad Laboratories, Richmond, CA, USA). Free sites were blocked with 3% LPD in TBS by incubating the membrane overnight at 4 ° C. After 3 washes with TBS-T (0.05 M Tris-HCI, pH 7.5, 0.15 M NaCI, 0.05% Tween-20), the membrane was incubated overnight at 4 ° C with serum Rabbit polyclonal HPI-2 in a 1/100 dilution in 3% LPD in TBS-T. Subsequently, the membrane was washed 5 times with TBS-T and incubated for 2 h at room temperature with a rabbit anti-IgG antibody conjugated to Peroxidase in a 1/2000 dilution in 3% LPD in TBS-T. Finally, the development was carried out with ocloronaphthol 0.5 mg / ml in 17% methanol and H ₂ 0 ₂ 0.005% V / V dissolved in TBS.

The majority of thioredoxin-proinsulin fusion protein was expressed as insoluble aggregates present in Cl (Figure 15), a smaller proportion in FIS and could not be recovered from EP. An approximate estimate based on the intensity of the bands indicated that about 5-10% of thioredoxin-proinsulin was soluble. Because the majority of the fusion protein was found in Cl, and it was feasible to recover it with a high degree of purity, this fraction was selected to perform the biochemical and immunochemical characterization studies. Example 10

Biochemical characterization of thioredoxin-proinsulin

For the analysis of the total thioredoxin-Proinsulin mass, the thioredoxin-proinsulin chimera was biochemically controlled by mass spectrometry

(National Laboratory for Research and Services in Peptides and Proteins, -LANAIS- CONICET-UBA). Briefly, the protein was analyzed by RP-HPLC-MS using a Vydac C8 column of 1 mm x 30 mm, operating at a flow of 40 μΙ / min, connected to Surveyor HPLC System with an LCQ Duo mass spectrometer (ESlion trap) (Thermo Fisher, San José, CA, USA). The protein was then eluted using a 15 minute gradient, from 10 to 100% of solvent B (solvent A: 2% acetic acid, 2% acetonitrile; solvent B: 2% acetic acid, 96% acetonitrile). Protein molecular mass was determined using Full Sean 300-2000 amu and a deconvolution program (ProMass Deconvolution Program). The approximate error of the trial was 0.06%. By means of the respective theoretical forecasts, the probable theoretical masses of the 5 alternatives for the fusion protein could be predicted depending on the amount of disulfide bridges formed. These were:

• 4 S-S bridges: 22174.2 Da

• 3 S-S bridges: 22176.2 Da

• 2 S-S bridges: 22178.2 Da

• 1 S-S bridge: 22180.2 Da

• No S-S Bridge: 22182.2 Da

Experimental analysis demonstrated the presence of a majority peak of PM 22173.5 Da, compatible with the theoretical PM calculated for thioredoxin-proinsulin (22174.17 Da) when all disulfide bridges are formed (Figure 16 and Table 1). Therefore, the result obtained indicates univocally that when it meets all its disulfide bridges formed, the fusion protein has 3 bridges adjudicable to Pl and 1 corresponding to the active Trx site.

Table 1: Mass spectrometry analysis of recombinant TrxPI and peptides from digestion with protease V8 and with chymotrypsin. ^a Eluted at 22 min (Figure 17), Eluted at 46 min (Figure 17)

To determine the proportion of properly formed disulfide bridges, enzymatic digestion was carried out using the V8 endoprotease of Staphylococcus aureus (Sigma-Aldrich, Inc., St Louis, MO, USA). The S. aureus V8 enzyme is a serine endopeptidase that specifically digests peptide bonds at the carboxyl end of aspartic acid (D) or glutamic (E) residues. To do this, 30 μg of standard Pl and thioredoxin-proinsulin were incubated overnight with 5 μg of V8 at room temperature. In parallel, a control was performed in which there was no added protein. Another 30 μg of thioredoxin-proinsulin were treated with 0.2 M DTT after digestion with V8. The reactions stopped at -20 ° C. The samples were then analyzed by RP-HPLC using a Vydac C ¹⁸ column. The material bound to the column was eluted with a linear gradient in which the concentration of acetonitrile was increased to a maximum of 90%. The chromatograms obtained, for both the reduced samples with DTT and those corresponding to the non-reduced samples, were compared and the relevant peaks were selected to be examined by Matrix-Assisted Laser Desorption / lonization-Time Of FIight Analysis (MALDI-TOF spectrometer, Ultraflexil, Bruker; Center for Chemical and Biological Studies by Mass Spectrometry -CEQUIBIEM- CONICET-UBA, Argentina). When treating thioredoxin-proinsulin with the V8 protease 12 peptides are generated (Figure 17), two of which are carriers of the disulfide bridges (peptides 6 and 7 as indicated in Figure 17).

The products of the digestion with V8 were separated into two equal parts, one of which was treated with DTT for the reduction of the disulfide bridges. Peptides resulting from digestion were subjected to RP-HPLC with comparable chromatograms observed between thioredoxin-proinsulin and standard Pl (Figure 18). The peaks suspected of containing the disulfide bridge-carrying peptides (peptides 6 and 7) that disappeared when treated with DTT, were collected and analyzed by MALDI-TOF (CEQUIBIEM-CONICET-UBA). Peptide 7 whose theoretical mass is 1378 Da was successfully characterized, being consistent with the expected disulfide bridge formation between Cys140 and Cys206 (Table 1). However, it was not possible to

characterization of peptide 6 (theoretical mass 4845 Da, disulfide bridges Cys128-Cys193 and Cys192-Cys197), even using a special sinapine acid matrix for high molecular weight peptides.

To further characterize the peptide structures involved in the formation of disulfide bridges, the reduced and non-reduced thioredoxin-proinsulin, and subsequently blocked with iodoacetamide, were subjected to a chymotrypsin digestion. Chymotrypsin is an enzyme with proteolytic activity that catalyzes the hydrolytic breakdown of peptide bonds adjacent to aromatic amino acid residues. The main substrates of the enzyme, which are hydrolyzed in the terminal carboxyl, include tryptophan (W), tyrosine (Y), phenylalanine (F) and methionine (M). A proportion of thioredoxin-proinsulin was in principle reduced with 0.2 M DTT in 6 M urea, pH 8.2. After 1 h of incubation at room temperature, 5 μΙ of 1 M iodacetamide was added and incubated in the dark for 1 h at room temperature. The reaction was stopped with 5 μΙ of TFA. The reduced and non-reduced chimera was then digested with chymotrypsin in a 1: 25 ratio chymotrypsin: Thioredoxin-Proinsulin. Said reaction was carried out at room temperature overnight. Digestion products were analyzed by MALDI-TOF

(CEQUIBIEM-CONICET-UBA). This procedure again confirmed the formation of the Cys140-Cys206 bridge (Table 1). In turn, the presence of a peak of 3869 Da (Table 1) that was not observed when thioredoxin-proinsulin was reduced and

carbamidomethylated, was indicative of the presence of a peptide whose structure contains the disulfide bridges between Cys128-Cys193 and Cys192-Cys197. Also, the peptide 123-137 possessing Cys128 was verified by fragmentation, in the sample reduced with DTT and blocked with iodoacetamide.

Example 11

Immunochemical characterization of TrxPI

In order to evaluate the potential use of TrxPI in immunochemical assays and its application in the detection and characterization of the PAA marker in patient sera, immunoreactivity studies of the chimeric protein were performed. For this, ELISA techniques for preincubation, radioimmunoassay (RIA) and surface plasma resonance technology (SPR) were implemented, comparing the

TrxPI behavior with a standard of human Pl, against rabbit polyclonal serum HPI-2 and sera of IAA / PAA positive patients.

Serotheques

Thirty sera from patients with type 1A DM with a wide range of reactivity for PAA were used. These samples corresponded to children and adolescents with an average age of 8.31 ± 4.20 years at the time of diagnosis. Serum samples were collected before or within 72 hours of the start of insulin treatment. The diagnosis was made according to the criteria of the WHO (Diabetes mellitus. Report of a WHO Study Group, 1985).

Normal controls included sera from 30 healthy individuals with no personal or family history of DM, free of autoimmune diseases, none of whom had received insulin, or experienced episodes of hyperglycemia.

The sera were evaluated by RBA (Valdez et al., 2003). The SDs of the B% signals was 1 1, 64 ± 8.35 (mean ± SD, range 3.58 to 35.85). The cut-off value for positivity was considered as SDs ≥ 3.00.

Polyclonal rabbit anti-PI serum

For the immunochemical characterization of TrxPI, the rabbit polyclonal serum HPI-2 was used, since it had a higher anti-PI antibody titer than the HPI-1 serum. Statistic analysis

To obtain the cut-off level of the RBA, the distribution of the data was evaluated

corresponding to normal control subjects. For this, the Kolmogorov-Smirnoff Normality Test was performed, using the computer program

GraphPad Prism 5. Since the results of the control sera followed a normal distribution, it was considered as a cut-off average value plus 3 standard deviations.

To analyze the parallelism between the slopes of the two regression lines corresponding to the RIA dose-response curves, the ANOVA test was applied. When the P values obtained were greater than 0.05, it was concluded that there was no significant difference between the slopes (Royal Spanish Pharmacopoeia, 1997).

To assess whether there was a significant difference between EC50 derived from the dose-response curves of the RIA, the Mann-Whitney U Test was applied. P values less than 0.1 were considered significant.

To evaluate the correlation between the values of K _a obtained with standard Pl or TrxPI immobilized in the sensor chip cell, a standard linear regression test was used.

Preincubation ELISA

75 μΙ of the rabbit polyclonal serum HPI2 1 / 5,000 was incubated in PBS-T-Albumin 3% with 75 μΙ of standard Pl or TrxPI in serial concentrations of 1 pM to 1 μΜ.

In parallel, a 96-well polystyrene microplate was sensitized with 50 μ 50 of a 1-μg / ml dilution of standard Pl in PBS overnight at 4 ° C. After 3 washes with PBS, the nonspecific reactive sites of the plates were blocked with PBS-3% Albumin (200 μΙ per well) for 2 h at room temperature. After this time, 5 washes were carried out with PBS-T and 50 μΙ of the pre-incubated were poured into each well and incubated for 1 h at room temperature. After 5 washes with PBS-T, the plate was incubated with 50 μΙ of a rabbit anti-gamma antibody globulin conjugated with Peroxidase diluted 1/3000 in PBS-T-Albumin 3% for 1 h at room temperature, and developed colorimetrically with TMB. At higher concentrations of standard Pl or TrxPI antigen present in the pre-incubated, the lower the optical density (OD) signal obtained in the test (Figure 19).

The displacement curves of standard Pl and TrxPI against the HPI-2 serum were superimposed, showing immunochemical identity between the two (Figure 20). To objectively confirm the degree of overlapping of the curves, the EC50 values were compared and the existence of parallelism was analyzed (Table 2). The 50% maximum signal binding is dependent on the avidity of the antibodies towards the standard Pl or TrxPI.

Radioimmunoensavo (RIA)

To deepen the comparative study between TrxPI and standard Pl, quantitative competence radioimmunometric tests were performed using HPI-2 in a final dilution 1/100 or a pool of sera from diabetic PAA + patients (Figure 21). Thirty microliters of the sera were incubated for 7 days at 4 ° C with 1000 cpm of [ ³⁵ S] -PI in the presence of 90 μΙ of serial concentrations (10 pM to 1 μΜ) of standard Pl or purified TrxPI from both FIS or from Cl. Los

Immunocomplexes formed were isolated with 50% G-Sepharose 4B FF protein (Amersham Biosciences, New Jersey, USA) in RBA buffer. The suspension was then incubated 2 h at room temperature under continuous stirring. The samples were centrifuged and the precipitates containing the immunocomplexes were washed four times with 200 µΙ of RBA buffer and resuspended in 1% SDS. The supernatants were transferred to vials to be counted in an automatic liquid scintillation counter (Figure 22). The dose-response curves were adjusted to the following mathematical equation:

where B% min and B% max correspond to the minimum and maximum response

respectively, and parameter EC50 represents the 50% doses of B% max.

The identity between the curves obtained with standard Pl and TrxPI were analyzed by comparing slopes and EC50 values. The analysis of the variances of the data was carried out using a model of parallel lines and design

"completely randomized". As can be seen in Figure 21 and Table 3, the dose-response curves with standard Pl and with TrxPI showed parallel and overlapping sigmoid curves for the HPI-2 serum, or with a difference of non-significant variances (p <0, 1) for the PAA + pool (Berzofsky and Schechter, 1981). Also, EC50 values showed no significant differences (p <0.05 and p <0.1,

respectively). This confirms that the TrxPI has a behavior

immunochemical indistinguishable from the standard of Pl.

Radiolysis binding inhibition assay

To assess the ability of TrxPI to react with PAA present in human sera, PAA present in sera of 30 infant-juvenile patients with type 1 A DM were detected by RBA as described by Valdez et al. (2003). The [S] -PI was obtained as explained above for the RIA. Aliquots of 30 μ of human sera were incubated for 7 days at 4 ° C with 1,000 cpm of [ ³⁵ S] -PI in a final volume of 120 μΙ in RBA buffer. Subsequently, 50 μΙ of 50% G-Sepharose 4B FF protein in RBA buffer was added and the suspension was incubated for 2 h at room temperature under continuous stirring. The samples were centrifuged and the precipitates containing the immunocomplexes were washed four times with 200 μΙ of RBA buffer. Then they were resuspended in 100 μΙ of 1% SDS,

All sera analyzed were PAA positive for RBA (Valdez et al., 2003) with a value of SD score 11, 64 ± 8.35 (mean ± SD, range 3.58 to 35.85, cut-off value for SD positivity score≥ 3.00). The assay was conducted in the absence and in the presence of TrxPI (1 μΜ) from both the soluble intracellular fraction and the inclusion bodies. Most sera were negativized in the presence of micromolar concentrations of TrxPI, with a value of SD score 1, 01 ± 1, 22 (mean ± SD, range 0.77 to 3.52) (Figure 23).

Plasma Surface Resonance Assay (SPR)

The immobilization of standard Pl and TrxPI was carried out following the guidelines described in example 4. For this, TrxPI was used at a concentration of 30 μg / ml in 10 mM acetate buffer (pH 4.0). The immobilization levels reached were 2000 and 500 RU for concentration and affinity tests, respectively.

The HPI-2 rabbit polyclonal serum was used as a control of specific binding to the surface immobilized with the antigens.

To confirm the correct folding of TrxPI, the specific antigen-antibody interaction was analyzed by surface plasma resonance technology (BIAcore, GE Healthcare, Uppsala, Sweden). For this assay, the standard Pl and TrxPI proteins (300 RU and 466 RU, respectively) were immobilized in cells separated from the surface of a CM5 sensor chip. Each patient serum was used pure or diluted ½, ¼ and 1/8 in PBS-T buffer. In each case, the starting samples were diluted ½ with 1 mg / ml carboxymethyldextran and 0.35 M NaCl, in order to eliminate nonspecific binding. Each sample was injected for 300 seconds, and the run buffer buffer range, PBS-T, was measured after 300 seconds of interaction. The Assays were carried out at 20 ° C, with a flow rate of 10 μΙ / min (Figure 24). The association rate constant (k ^, the dissociation rate constant (k-1) and the equilibrium affinity constant (association constant K _a ) were calculated from the sensorgrams using the BIAevaluation software of the same instrument The correlation of affinities of the sera under study obtained with standard Pl or TrxPI was analyzed.The correlation analysis showed no significant differences

being the

correlation (r ² ) between these results of 0.80 (Figure 25).

Example 12

Thioredoxin-proinsulin digestion with enterokinase

The Trx domain of the chimera was removed by digestion with enterokinase (EK). For this, 100 to 200 μg of thioredoxin-proinsulin were incubated with a range of 6 to 10 units of EK (Invitrogen, Carlsbad, California, USA) in 200 μΙ of 50 mM Tris / HCI buffer, pH 8.0, CaCI2 0 , 1 mM and Tween-20 0.1%, for 7h at 4 ° C. Next, the Pl was isolated from the other digestion products by HPLC using a Vydac C ₁₈ reverse phase column (4.6 x 250 mm). A linear gradient similar to that described previously was used for the radioimmunoassay of Example 3. The recovered fractions were analyzed by SDS-PAGE and WB.

Under these conditions, Trx was removed by approximately 50% (Figure 14). The peak eluted at 30.8 min (similar to the retention time obtained by running standard Pl) proved to be a 9 kDa protein compatible with the presence of Pl (Figure 26). Likewise, said fraction managed to displace the [ ³⁵ S] -PI binding to the HPI-2 polyclonal serum from 19.8% to 3.3%.

Example 13

Evaluation of the possible presence of anti-Trx antibodies in patient samples In order to rule out the presence of anti-Trx antibodies in the sera under study, which, if present, could react with the TrxPI chimera and interfere with false positovas, a high sensitivity chemiluminescence assay was developed.

Reagents

PBS was used as a sensitization buffer, 3% LPD in PBS was used for blocking, and the dilution buffer used was 3% LPD in PBS-T. The chemiluminescence substrate was Luminigen® PPD (DPC®, Los Angeles, CA, USA) diluted 1/10 in PBS. The second antibody used was rabbit anti-lgG / human anti-lgG conjugated to alkaline phosphatase (Jackson ImmunoResearch Laboratories, Inc., West Grave, PA, USA).

The normal control sera used to establish the cut-off value came from healthy subjects.

As a positive control, a polyclonal rabbit anti-Trx serum generated by immunization of a New Zealand White rabbit with 1 mg of recombinant Trx emulsified in complete Freund's adjuvant was used. The initial injection was followed by booster injections with 1 mg of Trx in incomplete Freund's adjuvant at 4-week intervals. The rabbit was bleeding 15 days after booster shots.

Recombinant Trx protein was expressed with the pTrx vector (Invitrogen, San Diego, CA, USA) in E. coli, and purified by osmotic shock according to the manufacturer's instructions. The product was dialyzed against PBS and then lyophilized.

Chemiluminescence Assay

All samples and blanks were evaluated in duplicate and incubations were performed at room temperature.

Opaque white 96-well plates (Corning Costar Corporation, Cambridge, MA, Great Britain) were used which were sensitized with 0.5 μg / well Trx overnight at 4 ° C, washed 5 times with PBS and blocked with 300 μΙ of the corresponding buffer for 1.5 h. After 5 washes with PBS-T, they were incubated for 1 h with 50 μΙ of normal human sera (n = 40) and patient samples diabetics (n = 50) (the same ones that were later used in SPR tests) diluted 1/100 in the dilution buffer containing normal rabbit serum in a final dilution of 1 / 10,000. The positive control was diluted 1 / 10,000. To measure the nonspecific signal, the dilution buffer where the serum aggregate was omitted was duplicated. After 5 washes with PBS-T, the plates were incubated with 50 μΙ of the 1 / 8,000 conjugate for 1 h. Finally, after another 5 washes with PBS-T, the chemiluminescence substrate (50 μΙ / well) was added and incubated for 10 min in the dark. The reaction was measured with the 1420 Wallac Multilabel Counter VICTOR3 ™ instrument (PerkinElmer Inc., USA) (Figure 27).

The specific signal (SE) was calculated as the mean of each sample minus the mean of the nonspecific control. Data were processed and expressed as SDs = (SE - SEc) / SDc, where SEc is the mean of the SE of normal sera and SDc is its standard deviation. A result was considered positive if the SDs> 3. This test was able to detect an anti-Trx rabbit serum in dilutions greater than 1 / 100,000. When evaluating the 51 sera of diabetic patients, all were negative for anti-Trx antibodies (Figure 28).

Example 14

Study of concentration and affinities of IAA / PAA by SPR

Used libraries

For the determination of concentration (q) and affinity (K _a ) of IAA / PAA in patient samples by SPR tests, HPI-2 was used as a standard serum, previously characterized by RIA (Example 3). For the tuning and optimization of this methodology, sera from the different exploratory bleeding and the final bleeding of HPI-2 were used, with which the evolution of the immune response was evaluated in terms of the parameters q and K _a from the preimmune state until the final bleeding (bleeding 1 to 6).

Patients with type 1 diabetes: Serum samples from 28 children and youth patients with type 1 DM were used with a varied range of PAA values determined by RBA, with an average age of 8.31 ± 4.20 years at diagnosis. He Type 1A diagnosis of DM was made according to the WHO criteria (Diabetes mellitus. Report of a WHO Study Group, 1985). The initial group of patients included 71 samples of mainly Caucasian children and adolescents. Samples were collected before or within 72 h of the start of insulin treatment.

All these patients were analyzed in parallel for the different humoral markers of DM, PAA, GADA, IA-2A and ZnT8A. Clinical and laboratory data for these patients are shown in Table 4.

Diabetic patients diagnosed in adulthood: Serum samples from 23 adult subjects with diabetes diagnosis at age ≥ 65 years and body mass index (BMI) <30 were used. The diagnosis of DM was made according to the American Diabetes Association (Expert Committee on the Diagnosis and

Classification of Diabetes Mellitus, 1997). The criteria for establishing oral hypoglycemic therapy (HGO) was performed according to the UK Prospective Diabetes Study (Matthews et al., 1998).

This collection of sera was selected from a starting population of 68 patients. All these diabetic patients with clinical debut in adulthood were analyzed in parallel for the previously mentioned humoral markers. These patients had not been treated with insulin prior to the sample collection for immunochemical analyzes.

Clinical and laboratory data are shown in Table 4. Blood samples were collected after a fast of at least 8 h, and the serum was stored at -20 ° C until the tests were performed.

Normal control individuals: 30 normal control sera obtained from subjects without a personal or family history of DM, free of autoimmune diseases, none of whom had received insulin, or experienced episodes of hyperglycemia were used.

For the execution of the SPR tests, the CM5 chips immobilized with standard Pl and TrxPI were used as described in Examples 4 and 1 1. For the measurement of concentration (q) of PAA, the amount of standard Pl immobilized on the surface of the CM5 sensor chip was corresponding to 1600 RU. For the affinity measurement (K _a ) of PAA, the surface of the sensor chip was prepared with smaller amounts of immobilized antigens (300 RU and 466 RU, for Pl and TrxPI, respectively), in this way it was possible to evaluate the interaction of the antibodies depending on their binding strength.

Commissioning of the SPR test for the determination of q and K _a with HPI-2 anti-PI polyclonal serum

For the optimization of the trial, preimmune bleeding, exploratory bleeding 1 to 5 and final bleeding were used in dilutions 1/8, 1/16 and 1/32 in PBS-T.

To evaluate q, the samples were injected for 120 seconds so that the association with the standard Pl immobilized on the surface of the sensor chip occurs. The cell with an immobilization level of around 2000 RU was used. Binding levels were measured after another 120 seconds of PBS-T injection. After each antigen-antibody interaction, a regeneration step was carried out for complete dissociation of antibodies using 10 mM glycine-HCI pH 1.5. All sensorgrams were corrected by subtracting the signal produced in the reference cell. For the determination of q of the different HPI-2 indentations, a calibration curve was constructed (RU vs. HPI-2 concentration determined by RIA). The Nanomolar concentrations of HPI-2 used were: 75.00, 37.50, 18.75, 9.38, 6.69, 2.34, 1, 17 and 0.59 (Figure 29). From this curve, the q of the different exploratory indentations were determined. For each sample, it was determined by interpolation in the standard curve by means of the BIA-evaluation software program.

For the evaluation of K _a of each of the analyzed bloodletting, the samples were injected in parallel for 300 seconds on the surface of the sensor chip with standard Pl or immobilized TrxPI. The level of binding was measured after 300 seconds, and after each antigen-antibody interaction, a step of

Regeneration for complete dissociation of antibodies using 10 mM glycine-HCI pH 1.5. The kinetic association constant (k-ι), the kinetic dissociation constant (k.-i) and the association equilibrium constant (K _a ) were calculated from the analysis of the corresponding sensorgrams using the BIA-evaluation software . As immunizations were repeated in the rabbit, both the q and K _a values of the anti-PI antibodies were progressively increased (Figure 30, Figure 31 and Table 5). In this way, the evolution of the specific humoral immune response was evidenced in the experimental model.

All experiments were carried out at 20 ° C with a flow of 10 μΙ / min.

Determination of K _a in samples of patients with DM Once the trials with the biosensor were completed, PIP studies (K _a and q) were started in sera of patients with different variants of Autoimmune Diabetes Mellitus.

Each patient serum was analyzed pure or diluted ½, ¼ and 1/8 in PBS-T. In turn, said samples were diluted ½ with carboxymethyl-dextran and NaCI in one

final concentration of 1 mg / ml and 0.35 M, respectively, in order to eliminate nonspecific reactions.

The conditions of the run were those described in the previous section for commissioning.

A representative sensorgram of the antigen-antibody interaction and the determination of Ka is shown in Figure 32.

Statistic analysis

corresponding to normal control subjects. For this, the Kolmogorov-Smirnoff Normality Test was performed using the GraphPad Prism 5 computer program. Since the results of the control sera followed a normal distribution, it was considered as an average cut-off value plus 3 standard deviations.

The values of q and K _a were expressed as median and range (minimum-maximum). To evaluate the existence of a significant difference between the groups of patients studied, the Mann-Whitney Test was applied, considering significant values of P <0.05.

Once the calibration curve (RU vs. antibody concentration) was constructed with the rabbit polyclonal serum HPI-2, it was used as a standard curve to interpolate the RU values of the samples studied. The range of q obtained in the 28 samples of children and youth patients was 24.08 x10 ^-9 M to 243.65 x10 ^-9 M, with a median of 67.12 x10 ^-9 M and in the 23 patients with debut in adulthood it was from 24.10 x10 ^-9 M to 318.4 x10 ^"9 M, with a median of 167.40 x10 ^-9 M (Figure. 33 B, Table 6).

Table 6: Comparison of the primary antigen-antibody interaction parameters: Standard Deviation (SDs) score signals obtained by RBA, concentration (q) and affinity (K _a ) of PAA calculated using BIAevaluation software (BIAcore ™).

For the affinity determination of the PAA present in the patient samples, a 1: 1 binding protocol was used as an adjustment model (Langmuir). In the infanto-juvenile patients studied, the resulting K _to K values ranged from 4.45 x10 ⁴ M ^-1 to 5.62 x10 ⁸ M ^-1 for Pl as an immobilized antigen

and from 5.90 x10 ⁴ M ^-1 to 5, 18 x10 ⁸ M ^-1 for TrxPI as immobilized antigen

On the other hand, for the PAA of the adults evaluated, the range of

it was 6.78 x10 ⁴ M ^-1 to 1, 36 x10 ⁸ M ^-1 and that of

Unexpectedly, the values of the medians of

were significantly higher in the child-youth population than in the group of patients with debut in adulthood (3.50 x10 ⁷ M ^-1 vs. 0.84 x10 ⁷ M ^-1 4.65 x10 ⁷ M ^-1 vs. 0.72 x10 ⁷ M ^-1 respectively; p <0.05) (Figure 33 C, Table 6), since the shorter evolution of the former's prodrome led to fewer somatic hypermutation events for affinity maturation, compared to seconds. Figure 34 shows representative sensorgrams obtained for 3 dilutions of the same sample of an infant-juvenile patient (A) and an adult patient (B).

Added to this, when injecting sera in cell immobilized either Pl or TrxPI standard, the correlation between the values of K obtained was satisfactory, with r ² = 0.80 (Figure 25).

Example 15

Application of SPR in the characterization of IAA / PAA in specific clinical cases Two particular cases of patients with high IAA / PAA or IA titre and with uncontrolled blood glucose levels were analyzed by SPR, since conventional methodologies have serious limitations to achieve the same performance.

PATIENT 1

Presumptive diagnosis: Benign monoclonal gammopathy or Autoimmune Insulin Syndrome (ASI) not Hirata. Age: 27 years His first clinical consultation was at 1 1 years where he stated that since 10 years he had recurrent symptoms of hypoglycemia. The determination of anti-insulin autoantibodies (IAA) was performed early by RBA, detecting extremely high levels of B% (62, 1%). The weighting of such response in absolute terms (RIA / Union Capacity -BC-) yielded IAA values always greater than the 30 international units of insulin (Ul lns) / L plasma over several years of follow-up.

He resorted again to the doctor at 15 years and 8 months after presenting uncontrolled glycemia, with alternating episodes of hyper and hypoglycemia during the day. He was prescribed treatment with Methylprednisone (6/1/2000) 40 mg / day. He then showed hyperglycemia symptoms, so the dose was reduced to 20 mg / day until gradually reduced to 2 mg / day. Before the recurrence of severe hypoglycemia, the previous dose of corticosteroid was returned. Finally, a year and a half later the corticosteroid was suspended. In addition, from June 2000 to date, he was prescribed treatment with 50 mg of Acarbose before each meal and with a diet without fast-absorbing carbohydrates. Uncontrolled blood glucose levels continued, so plasmapheresis was performed and treatment with the Rituximab immunomodulatory monoclonal antibody was started since March 2004.

In recent years, including the postplasmapheresis period, it was evaluated

longitudinally the presence of IAA / PAA by RBA and RIA / BC barely showing a decline after plasmapheresis. The levels of these antibodies were not modified by the use of corticosteroids or by the application of Rituximab, while the alternating profiles of hyperglycemia did not improve with the change in treatment used. Moreover, throughout the period under medical care with

Interconsultations were not detected endocrine abnormalities that were responsible for the hyper and hypoglycemic conditions. PATIENT 2

Diagnosis: Labile diabetes with lipodystrophy. Age: 12 years His first clinical consultation was at 4 years where the parents stated that the child was type 1 diabetic from the year of life. He had lipodystrophy at the insulin application sites (arms, gluteal area and mild thighs). The determination of anti-insulin (AI) antibodies performed by RBA detected B% values of 48.2% (cut off 3.28%). The treatment scheme given at that time was 16 units of pre-breakfast human NPH insulin and 3 pre-snack units and corrections with human current insulin when blood glucose was ≥ 350 mg / dl. He resorted to medical controls once a month, always reporting sustained nocturnal and morning hypoglycemia, and lipodystrophy in the arms and buttocks. He was indicated treatment with swine NPH 1, 2 U / kg / day, with which he improved lipodystrophy, but the blood glucose remained very labile. One year after the consultation, he is given insulin glargine (Lantus, long-acting insulin analogue) in the morning and Novorapid (fast-acting insulin analogue) before meals. The hypoglycemia episodes became mild and sporadic but symptomatic, and lipodystrophy persisted at the injection sites. Three years later the

administration of Lantus by Levemir, continuing with Novorapid before meals.

The patient's monthly checks were continued. Despite the changes in therapeutic behavior, the symptoms of morning hypoglycemias persisted and stages of postprandial hyperglycemia were added. Currently, the patient has 10 years of evolution of type 1 DM with regular metabolic control, lipodystrophy and high glycosylated hemoglobin. The last therapeutic plan established was 30 units of Lantus before dinner.

Given the clinical characteristics of patients 1 and 2 selected as "specific clinical cases", since they were not assimilable to typical cases of type 1A DM, we proceeded with the detailed characterization of the IAA / PAA and IA, respectively, by SPR. As far as we know, the proposed approach, SPR, to study the specific humoral immune response against insulin in "complex" patients at the PIP level, complemented by isotyping, has not been carried out by other authors. Study of q, K _a and Iq subisotypes by SPR

To determine the parameter q of the IAA / PAA and AI of the patients studied, the serum samples were diluted 1/10 and 1/30 in PBS-T. The test was performed following the guidelines previously described for commissioning in Example 14. All the sensorgrams were corrected by subtracting the signal produced in the reference cell, and the q was determined by interpolation in the standard RU vs. curve. HPI-2 concentration through the BIA-evaluation software program.

Known respective values of q of IAA / PAA and IA present in the samples of patients studied, K were determined by kinetic assays, as described for tuning in Example 14. For this, the serum samples ½, ¼ and 1/8 were diluted in PBS-T. In turn, said samples were diluted ½ with carboxymethyl-dextran and NaCl in a final concentration of 1 mg / ml and 0.35 M, respectively, in order to eliminate nonspecific reactions.

The q values of said antibodies were 352.25 x 10 ^-9 M and 250.55 x 10 ^-9 M for patients 1 and 2, respectively. As for their K _a , these turned out to be

for patients 1 and 2, respectively (Table 7).

For the determination of the IAA / PAA and IA isotypes, each patient serum was analyzed in 1/10 dilution in PBS-T, and injected for 300 seconds on the surface of a standard Pl sensor chip immobilized at a corresponding level to 2000 RU. The chip also had immobilized TrxPI. Thus, of the 4 cells of the sensor chip, cells 1 and 3 were used as reaction targets, cell 3 contained immobilized standard Pl and cell 4 contained immobilized TrxPI. In this way it was possible to compare the immunoreactivity of the IAA / PAA + sera to each of the proteins in the same SPR run

The anti-isotype-anti-lgG, anti-lgM and anti-lgA- antibodies (DAKO Denmark S / A, Glostrup, Denmark) were subsequently injected at a 1/50 dilution in PBS-T for 120 seconds, without prior regeneration of The surface of the sensor chip. The net response in RU was reported as the difference in RU of the amplification level from the RU of the basal amplification level (Figure 35).

As expected for a mature secondary response and of sufficient evolution, most of the IAA / PAA and IA were found to be IgG, with a value of 2534 RU and 961 RU for patient 1 and 2, respectively (Figure 36 A and Figure 37 A), while the other isotypes were undetectable (Figure 36 B and 36 C, and Figure 37 A).

A more exhaustive analysis to know the subisotypes (subclasses anti-lgG1, -lgG2, - lgG3, -lgG4) was carried out in the same way, being the dilutions of the anti-subisotype antibodies -anti-lgG1, anti-lgG2, anti- lgG3, anti-lgG4 (BD Biosciences Pharmingen, San Diego, CA, USA) - 1/10 in PBS-T. In patient 1, a response of type lgG3 (875 RU) predominated, coinciding with the fact that said serum did not precipitate with the use of A-Sepharose protein in radioimmunoassays. On the other hand, in patient 2 a lgG1 subtype response (83 RU) prevailed (Figure 37 B and Figure 38). The specificity of the (sub) isotyping reagents was demonstrated by injecting the reagents in the absence of serum where no binding was observed above background levels.

References

- Baneyx F (1999) Recombinant protein expression in Escheríchia coli. Curr Opin Biotechnol 10 (5): 411-421. - Berzofsky JA, Berkower IJ, Epstein SL (1993) Antigen-Antibody Interactions and Monoclonal Antibodies. In: Paul WE (ed) Fundamental Immunology. 3rd edn. Raven Press, New York, pp 421-465.

- Berzofsky JA, Schechter AN (1981) The concepts of crossreactivity and

specificity in immunology. Mol Immunol 18 (8): 751-763.

- Burkart V, Siegenthaler RK, Blasius E, Vandenbroeck K, Alloza I, Fingberg W, Schloot NC, Christen P, Kolb H (2010) High affinity binding of hydrophobic and autoantigenic regions of proinsulin to the 70 kDa chaperone DnaK. BMC Biochem 11:44.

- Chestnut L, Ziegler AG, Ziegler R, Shoelson S, Eisenbarth GS (1993)

Characterization of insulin autoantibodies in relatives of patients with type I diabetes. Diabetes 42 (8): 1202-1209.

- Chan SJ, Weiss J, Konrad M, White T, Bahl C, Yu SD, Marks D, Steiner DF (1981) Biosynthesis and periplasmic segregation of human proinsulin in Escherichia coli. Proc Nati Acad Sci USA 78 (9): 5401-5405.

- Chopra AK, Brasier AR, Das M, Xu XJ, Peterson JW (1994) Improved synthesis of Salmonella typhimurium enterotoxin using gene fusion expression systems. Gene 144 (1): 81-85.

- Mellitus diabetes. Report of a WHO Study Group (1985). World Health Organ Tech Rep Ser 727: 1-113.

- Douillard FP, O'Connell-Motherway M, Cambillau C, van Sinderen D (2011) Expanding the molecular toolbox for Lactococcus lactis: construction of an inducible thioredoxin gene fusion expression system. Microb Cell Fact 10:66.

- Economou A (1999) Following the leader: bacterial protein export through the Sec pathway. Trends Microbiol 7 (8): 315-320.

- Eskola V, Vahasalo P, Akerblom HK, Knip M (2003) Increased frequency of islet cell antibodies in unaffected brothers of children with type 1 diabetes. Horm Res 59 (4): 195-200.

- Expert Committee on the Diagnosis and Classification of Diabetes Mellitus

(1997) Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 20 (7): 1183-1 197.

- Francetic O, Belin D, Badaut C, Pugsley AP (2000) Expression of the

endogenous type II secretion pathway in Escherichia coli leads to chitinase secretion. EMBO J 19 (24): 6697-6703. - Gardner QT, et al., Studies on the regioselectivity and kinetics of the action of trypsin on proinsulin and its derivatives using mass spectrometry, Biochim. Biophys Acta (2012), http://dx.doi.Org/10.1016/j.bbapap.2012.09.004.

- Serious PM, Rotbart HA, Nix WA, Pallansch MA, Erlich HA, Norris JM, Hoffman M, Eisenbarth GS, Rewers M (2003) Prospective study of enteroviral infections and development of beta-cell autoimmunity. Diabetes autoimmunity study in the young (DAISY). Diabetes Res Clin Pract 59 (1): 51-61.

- Hoffman RD, Lañe MD (1992) lodophenylarsine oxide and arsenical affinity chromatography: new probes for dithiol proteins. Application to tubulins and to components of the insulin receptor-glucose transporter signal transduction pathway. J Biol Chem 267 (20): 14005-1401 1.

- Jahromi MM, Eisenbarth GS (2007) Cellular and molecular pathogenesis of type 1A diabetes. Cell Mol Life Sci 64 (7-8): 865-872.

- Juhl K, Hutton J (2004) Stimulus-secretion coupling in the pancreatic beta-cell.

Adv Exp Med Biol 552: 66-90.

- Kajava AV, Zolov SN, Kalinin AE, Nesmeyanova MA (2000) The net charge of the first 18 residues of the mature sequence affects protein translocation across the cytoplasmic membrane of gram-negative bacteria. J Bacteriol 182 (8): 2163-2169.

- Kang Y, Yoon JW (1991) Development of a high-expression vector (Pyk 10-9) of human proinsulin gene. Biotechnol Lett 13: 755-760.

- Kang Y, Yoon JW (1994) Effect of modification of connecting peptide of

Proinsulin on its export. J Biotechnol 36 (1): 45-54.

- Kemmler W, Peterson JD, Steiner DF (1971) Studies on the conversion of

proinsulin to insulin. I. In vitro conversion with trypsin and carboxypeptidase B. J Biol Chem 246 (22): 6786-6791.

- Khokhlova OV, Nesmeianova MA (2003) Interaction of SecB and SecA with the N-terminal region of mature alkaline phosphatase on its secretion in Escherichia coli. Mol Biol (Mosk) 37 (4): 712-718.

- Knip M (2002) Can we predict type 1 diabetes in the general population?

Diabetes Care 25 (3): 623-625.

- Koller DP, Rieb G, Sauber K, Uhlmann E, Wallmeier H (1989) Recombinant Streptomyces lividans secretes a fusion protein of tendamistat and proinsulin. Biotechnology (NY) 7: 1055-1059. - LaVallie ER, DiBlasio EA, Kovacic S, Grant KL, Schendel PF, McCoy JM (1993) A thioredoxin gene fusion expression system that circumvents inclusion body formation in the E. coli cytoplasm. Biotechnology (NY) 1 1 (2): 187-193.

- Mackin RB (1998) Proinsulin: recent observations and controversies. Cell Mol Life Sci 54 (7): 696-702.

- Malik A, Jenzsch M, Lübbert A, Rudolph R, Sóhling B (2007) Periplasmic

production of native human proinsulin as a fusion to E. coli ecotin. Protein Expr Purif 55 (1): 100-11 1.

- Matthews DR, Culi CA, Stratton IM, Holman RR, Turner RC (1998) UKPDS 26:

Sulphonylurea failure in non-insulin-dependent diabetic patients over six years. UK Prospective Diabetes Study (UKPDS) Group. Diabet Med 15 (4): 297-303.

- Mergulhão FJ, Taipa MA, Cabral JM, Monteiro GA (2004) Evaluation of

bottlenecks in proinsulin secretion by Escherichia coli. J Biotechnol 109 (1-2): 31-43.

- Mire-Sluis AR, Barrett YC, Devanarayan V, Koren E, Liu H, Maia M, Parish T, Scott G, Shankar G, Shores E, Swanson SJ, Taniguchi G, Wierda D, Zuckerman LA (2004) Recommendations for the design and optimization of immunoassays used in the detection of host antibodies against biotechnology products. J Immunol Methods 289 (1-2): 1-16.

- Norris JM, Barriga K, Klingensmith G, Hoffman M, Eisenbarth GS, Erlich HA, Rewers M (2003) Timing of initial cereal exposure in infancy and risk of islet autoimmunity. JAMA 290 (13): 1713-1720.

- Novikov AA, Borukhov SI, Strongin AY (1990) Bacillus amyloliquefaciens alpha- amylase signal sequence fused in frame with human proinsulin is properly processed by Bacillus subtilis cells. Biochem Biophys Res Commun 169 (1): 297-301.

- Olmos-Soto J, Contreras-Flores R (2003) Genetic system constructed to

overproduce and secrete proinsulin in Bacillus subtilis. Appl Microbiol Biotechnol 62 (4): 369-373.

- Ott PA, Dittrich MT, Herzog BA, Guerkov R, Gottlieb PA, Putnam AL, Durinovic- Bello I, Boehm BO, Tary-Lehmann M, Lehmann PV (2004) T cells recognize multiple GAD65 and proinsulin epitopes in human type 1 diabetes , suggesting determinant spreading. J Clin Immunol 24 (4): 327-339. - Papouchado ML, Valdez SN, Ghiringhelli D, Poskus E, Ermacora MR (1997) Expression of properly folded human glutamate decarboxylase 65 as a fusion protein in Escherichia coli. Eur J Biochem 246 (2): 350-359.

- Pines O, Inouye M (1999) Expression and secretion of proteins in E. coli. Mol Biotechnol 12 (1): 25-34.

- Pugsley AP, Francetic O, Driessen AJ, by Lorenzo V (2004) Getting out: protein traffic in prokaryotes. Mol Microbiol 52 (1): 3- 1 1.

- Royal Spanish Pharmacopoeia (1997). Statistical Analysis, Second edn.

Ministry of Health and Consumer Affairs, Madrid.

- Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: a laboratory

manual. 2nd edn edn. Cold Spring Harbor Laboratory, Cold Spring Harbor NY.

- Schagger H, von_Jagow G (1987) Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal Biochem 166 (2): 368-379.

- Sung WL, Yao FL, Zahab DM, Narang SA (1986) Short synthetic

oligodeoxyribonucleotide leader sequences enhance accumulation of human proinsulin synthesized in Escherichia coli. Proc Nati Acad Sci USA 83 (3): 561-565.

- Talmadge K, Brosius J, Gilbert W (1981) An 'internal!' signal sequence directs secretion and processing or proinsulin in bacteria. Nature 294 (5837): 176-178.

- Talmadge K, Gilbert W (1982) Cellular location affects protein stability in

Escherichia coli. Proc Nati Acad Sci USA 79 (6): 1830-1833.

- Tang JG, Hu MH (1993) Production of human proinsulin in Escherichia coli in a non-fusion form. Biotechnol Lett 15: 661-666.

- Thim L, Hansen MT, Norris K, Hoegh I, Boel E, Forstrom J, Ammerer G, Fiil NP (1986) Secretion and processing of insulin precursors in yeast. Proc Nati Acad Sci USA 83 (18): 6766-6770.

- Valdez S, Villalba A, lacono R, Poskus E (2004) Preparation of recombinant human ³⁵ S-proinsulin and its application in a combined radioligand-binding assay. Usefulness of the novel method as first-line screening for predicting insulin requirement in adult onset diabetic patients. Acta Bioquim Clin Latinoam 38 (1): 3-15. - Wilkinson DL, Ma NT, Haught C, Harrison RG (1995) Purification by immobilized metal affinity chromatography of human atrial natriuretic peptide expressed in a novel thioredoxin fusion protein. Biotechnol Prog 1 1 (3): 265-269.

- Winter J, Gleiter S, Klappa P, Lilie H (201 1) Protein disulfide isomerase

isomerizes non-native disulfide bonds in human proinsulin independent of its peptide-binding activity. Protein Sci 20 (3): 588-596.

- Winter J, Neubauer P, Glockshuber R, Rudolph R (2000) Increased production of human proinsulin in the periplasmic.

- Yalow RS (1973a) Radioimmunoassay methodology: application to problems of heterogeneity of peptide hormones. Pharmacol Rev 25 (2): 161-178.

- Yalow RS (1973b) Radioimmunoassay. Practices and pitfalls. Circ Res 32: Suppl 1: 1 16-128.

ABBREVIATIONS

Ac Antibodies

ADA American Diabetes Association

Asp Aspartate

BB Bio-Breeding rat strain

BC Capacity of Union

βΜΕ β-mercaptoethanol

B% Binding percentage

BSA Bovine Seroalbumin

CDC Center for Disease Control

Cl Inclusion Bodies

CMH Major Histocompatibility Complex

CPAs Antigen Presenting Cells

Cpm Accounts per minute

CT Total Accounts

CTLA4 Protein 4 associated with cytotoxic T lymphocytes

CV Coefficient of variation

Cys Cysteine

DM Diabetes Mellitus

DO Optical Density DTT Dithiothreitol

EDC N-ethyl-N- (3-diethylmiopropyl) carbodiimide

EDTA Ethylenediaminetetraacetic acid

EK Enterokinase

ELISA Enzyme Linked Immunosorbent Assay

EP periplasmic space

FIS Soluble intracellular fraction

FPLC Fast Protein Liquid Cromatography

GABA Gamma Aminobutyric Acid

GAD Gultamate Decarboxylase

GADA Anti-Gultamate Decarboxylase Autoantibodies

GSH Glutathione Reduced

GSSG Oxidized Glutathione

HbA1 c Glycosylated hemoglobin

HGO Oral hypoglycemic agents

HLA Human Leucocyte Antigen

hPI Human Proinsulin

HPI Polyclonal rabbit anti-PI antibodies

HPLC High efficiency liquid chromatography

HRP Horseradish Peroxidase

Hsp Heat shock proteins

IA Anti-insulin antibodies

IA-2 Tyrosine phosphatase IA-2

ΙΑ-2β Fogrina

IAA Anti-insulin Autoantibodies

IAS Autoimmune Insulin Syndrome

IASP Islet Antibody Standarization Program

ICA Pancreatic islet anti-cell autoantibodies

IL Interleukin

BMI body mass index

INF-γ lnterferon-γ

INS Insulin Gene

Insulin Ins K ₀ Average antibody affinity constant determined by radiometry

k ₁ Association speed constant

k _-1 Dissociation rate constant

K _a Antibody affinity constant

K _ap Apparent affinity antibody constant

K _d Antibody dissociation equilibrium constant

LADA Latent autoimmune diabetes in adults

LB Lymphocyte B

LT Lymphocyte T

LT Total Listed

LPD Skim milk powder

MALDI-TOF Matrix-Assisted Laser Desorption / lonization-Time Of Flight Analysis

NHS N-hydroxysuccinimide

NOD Non-obese Diabetics strain of mice

PAA Anti-Proinsulin Autoantibodies

PAO 4-amino-phenylarsine oxide

PBS Saline phosphate buffer

PBS-T phosphate buffered saline with 0.05% Tween-20

Pl Proinsulin

PIP Primary Interaction Parameters

PMSF Phenylmethylsulfonyl Fluoride

PTP protein tyrosine phosphatase

q Antibody concentration

q ₀ Antibody concentration determined by radiometry

RBA radioligand binding assay

RIA Radioimmunoassay

rpm Revolutions per minute

RU Resonance Units

SDS Sodium dodecyl sulfate

SDs Standard Deviation Units

SDS-PAGE Polyacrylamide gel electrophoresis with sodium dodecyl sulfate

SMS Stiff-man syndrome

CNS Rabbit preimmune serum SpA Protein A from Stafiloccocus aureus

SPR Surface Plasma Resonance

SPS Stiff-person syndrome

TBE Buffer Tris-Borato-EDTA

TBS Buffer Tris Saline

TBS-T Tris saline buffer with Tween-20 0.05%

TFA Trifluoroacetic Acid

Th1 Auxiliary T lymphocytes 1

Th2 Auxiliary T lymphocytes 2

TMB 3,3 ', 5,5'-Tetramethylbenzidine

TNF-α Tumor necrosis factor-a

Tryptophan Trp

Trx Thioredoxin

TrxPI Thioredoxin-Proinsulin

Ul Ins International Insulin Units

VNTR Variable Number Tandem Repeats

WB Western blotting

ZnT8 Isoform 8 of the zinc conveyor

ZnT8A Anti-ZnT8 Autoantibodies

Claims

1. A fusion protein characterized in that it comprises proinsulin fused to thioredoxin by means of a peptide bond.

2. The fusion protein of claim 1 characterized in that the portion

proinsulin in the fusion protein is human proinsulin, or proinsulin with a similarity percentage of at least 90%, more preferably at least 95%, and even more preferably at least 98% with respect to the amino acid sequence of the proinsulin human

3. The fusion protein of claim 2, characterized in that it comprises SEQ ID NO .: 1.

4. A polynucleotide characterized in that it encodes the thioredoxin-proinsulin fusion protein of claim 1.

5. The polynucleotide of claim 4, characterized in that it is a DNA molecule comprising the sequence SEQ ID NO .: 2.

6. A process for producing the fusion protein of claim 1

characterized in that it comprises:

a) transforming a prokaryotic cell with an expression vector comprising a DNA sequence encoding the thioredoxin-proinsulin fusion protein,

b) Cultivate said cell under conditions that allow the expression of the fusion protein, and

c) Recover the fusion protein.

7. The method of claim 6, characterized in that the prokaryotic cell is Escherichia coli.

8. The method of claim 7, characterized in that the thioredoxin-proinsulin fusion protein is recovered from the soluble intracellular phase.

9. The method of claim 7, characterized in that the fusion protein is recovered from the inclusion bodies.

10. The method of claim 9, characterized in that the recovered fusion protein is subjected to an in vitro refolding step.

1 1. The method of claim 10, characterized in that the step of

In vitro refolding comprises an oxidative refolding by dialysis at 4 ° C against refolding buffer.

12. An expression vector characterized in that it comprises a DNA sequence encoding the thioredoxin-proinsulin fusion protein of claim 1.

13. The vector of claim 12 characterized in that the DNA sequence encoding the thioredoxin-proinsulin fusion protein of claim 1 is operably linked to the P _L promoter of bacteriophage λ.

14. A transgenic cell characterized in that it is a prokaryotic cell containing a DNA sequence encoding the thioredoxin-proinsulin fusion protein of claim 1.

15. The transgenic cell of claim 14, characterized in that it is an Escherichia coli cell.

16. The transgenic cell of claim 15, characterized in that it comprises the vector of claim 12.

17. A method of producing proinsulin from the fusion protein of the

claim 1, characterized in that it comprises the steps of:

a) cultivate a microorganism that expresses thioredoxin-proinsulin,

b) Clivate proinsulin thioredoxin

18. The method of claim 17, characterized in that the clivating of step b is performed by using enterokinase.

19. A method for producing insulin, characterized in that it comprises the steps of: a) culturing a microorganism that expresses the fusion protein of claim 1 under conditions that allow the expression of the fusion protein,

b) Clivate thioredoxin to obtain proinsulin

c) cloning proinsulin C peptide to obtain insulin

20. The method of claim 19, characterized in that the clivating of step c is performed by using trypsin.

21. The method of claim 19, characterized in that the clivating of step c

includes the use of carboxypeptidase B

22. A method for detecting in vitro the presence of anti-proinsulin (PAA) and / or anti-insulin (IAA) autoantibodies in a biological sample, characterized in that it comprises contacting a biological sample with the fusion protein of claim 1 .

23. A method for determining in vitro the concentration and affinity of

anti-proinsulin (PAA) and / or anti-insulin (IAA) autoantibodies present in a biological sample, characterized in that it comprises contacting a biological sample with the fusion protein of claim 1.

24. The method of claim 23, characterized in that it is a method of

Plasma Surface Resonance (SPR).

25. The method of claim 24, characterized in that the fusion protein of claim 1 is immobilized on a chip having a carboxymethyl-dextran matrix.