Field of the invention

The present invention is related to bioengineered cyclic peptides containing citrul- line, based on SFTI (sunflower trypsin inhibitor) and cyclotides. These novel sequences have an effect in autoimmune diseases, e.g. citrullinated fibrinogen sequences that are grafted into the SFTI scaffold have been shown to block autoantibodies in rheumatoid arthritis and inhibit inflammation and pain. Further anti-citrullinated protein/peptide antibodies (ACPA) blocking agents including any ACPS active sequence listed in Table 2 and Table 4 can be used as diagnostic tools, i.e. for detection of subspecific antibodies as biomarkers, and for the isolation of subspecific antibodies.

Background

RA and ACPA: Rheumatoid arthritis (RA) is an autoimmune disease of incompletely known etiology that leads to chronic inflammation and destruction of the joints. Presence of anti-citrullinated protein/peptide antibodies (ACPA) in 60-70% of RA patients ² '3 is one of the specific serological diagnostic markers of the disease. It is believed that ACPAs emerge as an immune response towards proteins containing citrul- line. Within proteins/peptides, arginines may be converted into citrulline by means of specific peptidylarginine deiminases, a process known as citrullination.

ACPA levels in patients are determined using enzyme-linked immunosorbent assay (ELISA), which employs citrullinated peptides that are cyclized via a disulfide bond.

ACPAs appear in early stages of disease^ and are strongly associated with the genetic risk factor human leukocyte antigen-DRBi (HLADRBi) and polymorphisms in the protein tyrosine phosphatase N22 (PTPN22) gene ⁶ -?. Different models have been proposed to explain the appearance of ACPA including trauma, infections, genetic risk and several environmental factors ⁸ . ACPAs associate with an erosive disease course^ suggesting a direct pathogenic involvement in disease initiation and progression. It has been demonstrated that these autoantibodies can activate both the classical and the alternative complement pathways in a dose-dependent manner, in vitro ¹⁰ _' ¹¹ , they predict for the development of RA when present in undifferentiated arthritis and arthralgia ¹² , and induce production of TNF-alpha^ ¹ ^ Recently, induction of osteoclastogenesis and bone loss both in vitro and in vivo by vimentin- targeted ACPA has been proven^.

The stability of the peptide blocker is a major concern in the design. We have tackled this problem by using peptide scaffolds and chemistry developed at the University of Uppsala. Scaffolds include the circular plant proteins known as cyclotides ¹⁸ and sunflower trypsin inhibitors (e.g. SFTI-i). The circular peptide backbones of these peptides and the presence of disulfide bond(s) render them extreme stability. Their use as drug scaffolds have been demonstrated in a series of recent studies. Proof-of- concept studies have been conducted for both cyclotides and SFTI-i for their ability to accommodate bioactive sequences within their three-dimensional framework while maintaining the overall peptide stability. These studies have produced analogues to inhibit/activate vascular endothelial growth Factor ¹ ^ bradykinin receptors ²⁰ , mela- nocortin receptors ²¹ and to induce angiogenesis ²² . So far, these scaffolds appear safe. This includes possible immunogenic responses. To our knowledge no such effect has been demonstrated in in vivo studies (in rats, mouse). Here, cyclotide and SFTI-i frameworks have been used to stabilize the linear citrullinated peptides identified as potential inhibitors of ACPAs as indicated in Figure l.

The SFTI scaffold comprises peptides with a head-to-tail circular backbone of 13 to 18 amino acids, including two cysteines connected by a disulfide bond. The sequences between the two cysteines form two loops. The cyclotide scaffold comprises head to tail circular peptides of between 27 and 38 amino acids, including six cysteines connected by three disulfide bonds. The disulfide bonds are arranged in a so called cystine knot, i.e. Cys I is connected to Cys IV, Cys II to Cys V and Cys III to Cys VI. Cyclo- tides contain six loops, i.e. sequences between Cys residues. Both scaffolds are characterized by their extraordinary biological and chemical stability, conferred by their cyclic amide peptide backbone combined with the presence of disulfide(s). SFTI and cyclotide scaffolds being used originate from Asteraceae, Fabaceae, Violaceae, Rubia- ceae or Curcurbitaceae plant families.

The strategy of blocking autoantibodies using antigens grafted into these frameworks may also be used in other types of autoimmune diseases. They are also suitable for the normalization of altered bone metabolism treatment and treatment of fatigue.

Prior art

To our knowledge there is only one example of targeting autoantibodies using peptides in the literature: attempts are being exploited to neutralize autoantibodies against the cardiac βι-adrenenergic receptor. It has been demonstrated that targeting these pathogenic autoantibodies using cyclic peptides that mimic the real epitope structure (i.e. COR-i) could prevent autoantibody mediated myocardial damage in an experimental model of immune cardiomyopathy.(i6) In 2012, the results of a phase I clinical trial with 50 human volunteers demonstrated the safety of using the cyclic peptide COR-i ¹ ?. This peptide does not employ any scaffold structure such as SFTI or cyclotide.

Summary of the invention

The inventors have now developed potent ACPA blocking molecules and provided proof of concept that affinity purified ACPA can be neutralized in vitro and in vivo by stable molecules based on the amino acid primary structure of citrullinated fibrinogen peptides that have been previously identified in human arthritic tissue.

A lead compound and a diagnostic tool have now been developed. Detailed description of the invention

ACPA blocking agents

ACPA blocking agents include any peptidic compound or derivative that combine any ACPA active sequence epitope(s) and scaffold (s) below, in which the ACPA active sequence, or parts thereof, have been grafted into any loop of or loops of scaffold sequences.

ACPA binding agents for diagnostic purposes

ACPA binding agents for diagnostic purposes include ACPA blocking agents described above, which have been modified in a secondary loop with the purpose to facilitate binding to a column, ELISA-plate, or any other device.

ACPA active sequence epitopes

ACPA active peptides include, but are not limited to, sequences in Table 2 and Table 4. ACPA active sequences also include

i) truncated sequences of peptides in Table 2 and Table 4, which show binding to AC-

PAs

ii) citrullinated protein fragments other than those in Table 2 from alpha-enolase, vimentin, collagen type II, filaggrin, and fibrinogen, which show binding to ACPAs, iii) citrullinated protein fragments from other proteins than those listed in ii) showing binding to ACPAs

Scaf d sequences

Scaffold sequences include sunflower trypsin inhibitory (SI I/SI ) peptides, which comprise a circular peptide backbone of 13 to 18 amino acids, a single disulfide bond between cysteines, and two loops defined as the sequences between cysteines. Examples include,

SFTI-1 CFPDGRCTKSIPPI

Loop 1 2

Cys I II and in addition, but not limited to the further sequences listed below and in Table 3

SFT-L1 CFPDGCIEGSPV

PDP-16 CYPDGRCTRSNPPI

PDP-3 CYPDGRCTKSIPPI

PDP-12 CFPDGRCTKSIPPV

PDP-15 CFPDGRCTRSIPPI PDP-4 CFRRDGSCFGAF

PDP- 5 CYMDGRYRRC I PGMFRAY

PDP- 13 CYMDGRYRRCI PGMFRSY

PDP- 14 CYMDGRCRAGMFRSY

PDP- 7 GCLRDHCI P SGPI

PDP- 8 CFRLPDGGRLCVPPG

PDP- 17 C PDGDCHWI PAPPFFM

PDP- 9 CTPDGDCYWTSTPPFFT

PDP" 6 CFSDGHCIQVPPMATEI

PDP- 18 CTPDGRCYPVPYPPFYT

Scaffold sequences comprise cyclotides, e.g. from the Mobius (e.g. kalata Bi, kalata S, kalata B2) and bracelet subfamilies (e.g. cycloviolacin Oi - O20) and hybrids thereof, and the cyclotide like cyclic cystine knotted peptides from squash (Cucurbitacee; Momordica cochinchinensis Trypsin Inhibitor}' peptides MCoTI-I to MCoTI-VHI). Cyclotide scaffolds sequences above include a cyclic backbone and disulfide bonds of CysI-CysIV, CysII-CysV and CysIII-CysVI, arranged in a cystine knot, and their sequences are of the following pattern:

X(2,7)-C-X(o,4)-C-X(2,9)-C-X(3,8)-C-X-C-X(2,8)-C-X(i,7)

Loop 1 2 3 4 5 6

Cysteine 1 II III IV V VI

In which X denotes the number/range of numbers of amino acids between adjacent cysteines.

Mobius cyclotides include, but are not limited to, cyclotides containing the following sequence pattern between Cysl and Cys III:

C-[GA]-E-[ST]-C-[FTV]-[GLTI]-G-[TSK]-C

Bracelet sequences include, but are not limited to, cyclotides containing the following sequence pattern between Cys I and Cys III:

C-x(o,i)-[ES]-S-C-[AV]-[MFYW]-I-[PS]-x(o,i)-C Any known cyclotide sequence may be used. The ones explicitly mentioned are the preferred. Peptide scaffolds being used originate from Asteraceae, Fabaceae, Violace- ae/Rubiaceace or Cucurbitaecae.

A schematic representation of the grafting approach followed in the design of ACPA neutralizer peptides.

Selected bioactive (antigenic) epitopes are grafted into the scaffolds of sunflower trypsin inhibitor l (SFTI-i) and cyclotide Momordica conchinensis trypsin inhibitor II (MCoTI-II) in the design of stable ACPA neutralizer peptides. These compounds represent ultra-stable motifs that withstand enzymatic, chemical and thermal degradation. SFTI-i at the top is a 14 residue cyclic peptide isolated from the seeds of sunflower plants ² 3. It is stabilized by a single disulfide bond and a network of intramolecular hydrogen bonds. MCoTI-II contains a cyclic cystine knot (CCK), in which three conserved disulfide bonds are arranged such that one disulfide penetrates an embedded ring formed by the two other disulfides and their interconnecting backbone ² ^

In Figure 1 there are listed two classes of cyclic peptides that are used as the basic scaffold for preparing the new immunoblocking variants. SFTI-I, sunflower trypsin inhibitor I contains, as described above 14 amino acids in a cyclic peptide backbone, and one disulfide bond. The disulfide bond divides the peptide into two halves, which we call loops. The natural function of SFTI is as a trypsin inhibitor - one of the loops contain a potent inhibitory sequence and loop 2 maintains the structure. There is more than one SFTI molecule known today.

The lower three structures in Figure 1 are all examples of cyclotides. Peptides of this family of plant proteins are between 27 and 38 amino acids long and contain three disulfide bonds within their cyclic backbone. As such they contain six loops, i.e. sequences between cysteines. That gives the opportunity to insert more than one functional sequence: e.g. inserting one immunoblocking sequence and one albumin binding sequence to prolong half life in circulation.

Almost 300 different cyclotides have been characterized in plants. All such cyclotides may be used according to the present invention.

The peptides have been explored to identify lead compounds, to isolate antigen specific ACPAs and to develop a tool for research and diagnostics.

The cyclic backbones (also illustrated in Figure 1) in the peptide scaffolds in combination with disulfides give an extraordinary thermal, chemical and enzymatic stability.

The concept of grafting

This is being described in Figure 2

The bioactive sequence, in the example in Figure 2, namely the citrullinated fibrinogen peptide is inserted by replacing one of the loops in the peptides. In the case of SFTI the trypsin inhibitory loop is exchanged and the structural loop is kept. Process of preparing the peptides

The process for preparing SFTI (sunflower trypsin inhibitor) and cyclotide based peptides, wherein SFTI and cyclotide peptides incorporate citrullinated sequence epitopes from fibrinogen peptides will be described with reference to the compounds listed below.

FIB 573 CIT HHP GIA EFP ScitG KSS SYS KQF

All sequences that were made for ACPA inhibition based on the peptide we identified in fibrinogen are listed above. The initial ACPA binding citrullinated peptide at the top is called FIB 573 CIT. The next two sequences are truncated variants of that peptide; these were made in an attempt to facilitate synthesis and find a loop of a size that is more suitable to put into one of the scaffolds. The longest loops that have been shown to work in this grafting strategy is ca 25 amino acids long, but the shorter the better.

The first truncated peptide, with the suffix (lini), is as effective as the original one. As such we chose FIB 573 CIT (lini) as the starting peptide to insert into our scaffolds. All other peptides above, except the first three are macrocyclic, i.e. the N and C termini of the peptides are joined by an amide (peptide) bond. New compounds according to the present invention include cyclic versions if FIB 573 CIT and FIB 573 CIT (lini), with Cit and Arg (as a control); and FIB 573 CIT (lini) inserted into that scaffold (with Cit and Arg). The scaffold SFTI itself is also shown. We have also tested two different cyclotide scaffolds, kalata Bi and MCoTI-II, into the latter we have also inserted the shortest FIB 573 CIT peptide; (lin2).

CCP-i (cyclic citrullinated peptide I) is a published binder sequence of ACPA that is used as a positive control (note that this peptide is not cyclic N-C termini as SFTI, and cyclotides, it is called "cyclic" because it has one disulfide bond joining the two ends of the peptide).

Of the compounds listed above SFTI FIB 573 CIT(lini) (cyclo-CFPDGR CGIA EFP ScitG KSS SYS) has the best activity in the ACPA binding assay. Because of that it has been subject of most of the follow up studies.

Examples

Peptide design and syntheses:

Citrullinated peptides that have been identified as ACPA antigens, and truncated peptides containing both citmlline and arginine are synthetized in order to identify a minimal motif that can be grafted into the scaffolds. First, the 573 a-fibrinogen peptide was targeted, and used in the proof of concept study described in detail below.

First peptide sequence epitopes will be subjected to truncation studies. For example, full-length fib-a 573 is 21 residues long, but was truncated down to 15 and 11 residues, which may be regarded as more suitable lengths of sequence epitopes for grafting. These peptides were tested in the ACPA neutralization assay and their IC50 determined. Peptides with the highest IC50 and lowest corresponding arginine control response were selected for optimization regarding stability and IC50. Then, these most potent linear analogues are i) head-to tail cyclized and ii) grafted onto cyclic peptide scaffolds to improve stability and IC50. Cyclic compounds are then tested in complex matrices like serum and blood for stability and sustained ACPA neutralization activity.

After fib-a 573 , fib-a 591 are currently targeted together with the following peptides that have been identified as the most ACPA reactive epitopes^ (X=Cit): a-enolase (5 KIHAXEIFDSXGNPTVE 21), vimentin (1 STXSVSSSSYXXMFGG 16; 59 VYATXSSA- VXLXSSVP 74), collagen type II (59 AXGLTGXPGDA 369), and filaggrin (306 SHQESTXGXSRGRSGRSG 324). These peptides will be subjected to limited truncation and mutation studies, capitalizing on the knowledge from the truncations of fib- α 573. Hence, the total number of peptides will not exceed 35 (including 10 linear Cit peptides with their Arg controls; 10 scaffold peptides, 5 large scale syntheses). Table I: Sequences that are being synthesized as first generation peptide leads

Peptides Estimated number of peptides

Truncation studies 32

Cyclization linear analogues (Cit, Arg) 4

Grafting into scaffolds (Cit,Arg/SFTI, 8

McoTI)) 2 (>ioo mg)

Large scale synthesis of lead peptide(s)

Cyclic peptides are synthesized using FMOC chemistry, combined with native chemical ligation for cyclization of peptide backbones. In short, the starting point of synthesis is chosen so that the final peptide contains a N-terminal Cys, and the linker chosen so that a C-terminal thioester is generated upon cleavage (ref 26). The thus obtained thioesters can then be cyclized via native chemical ligation, forming an amide bond between the N and C termini. Cyclisation and folding (oxidation of disulfide bonds) are done in an established cyclization buffer (ref 26), Linear peptides are made using standard FMOC chemistry. Synthesized peptides are analyzed by NMR to ensure that folded products are obtained (as judged from the dispersion of amide protons in Ή NMR). Three-dimensional structure (3D) of active peptides will be determined using two dimensional NMR, including NOESY and TOCSY experiments.

Increased stability

When using the sequence FIB 573 CIT (lini) the effect increases when the peptide is inserted into the scaffold. The effect is greater than both cyclic and linear versions of FIB 573 CIT (lini)

The gained stability is further shown in Figures 3 and 4

Figure 3 shows that stability is gained by inserting FIB 573 CIT (lini) into the scaffold when incubated in human serum. The four lines at the top represent peptide controls incubated in physiological salt solution.

Note that grafted peptides, i.e. peptides with the ACPA binding epitope inserted into the scaffold, has superior stability compared to both linear and head-to tail cyclic versions of FIB 573 CIT (lini).

Figure 4 shows that the same trend is seen in blood (human): the FIB 573 CIT (lini) shows superior stability compared to the other variants.

Development of grafted peptides for isolation of specific antibodies and into a diagnostic tool:

Currently, the used ACPA antibodies isolated are not fibrinogen specific, but are active against all citrullinated peptides including for example vimentin and alpha eno- lase as outlined above. With grafted and specific ACPA blockers in hand, we will use them to i) purify specific subspecificities of ACPA and ii) develop specific Elisa assays. To that end, we will exploit the secondary loop(s) of the scaffold to introduce a handle for immobilization on columns and on microtiterplates. For example, a lysine will be introduced into the secondary loop of SFTI-FIB 573 CIT (lini), which will be used to immobilize the peptide to a NHS activated Sepharose-based column to isolate ACPA specific towards fib-a 573. Specific antibody/peptide pairs will then be used to set up Elisa assays. These assays will represent better models to monitor specific interactions, but will also be developed into a multiscreen Elisa platform that can identify which variant of ACPA that is present in individual patient samples (Fig 5). Thus, we will develop a diagnostic tool that will make possible patient tailored treatment using anti-citrullinated peptides in future therapeutic interventions.

Figure 5 shows the development of specific Elisa assays. Plates with our best candidate cyclic peptides will be used to develop a multiscreen Elisa. Streptavidin coated plates and a biotinylated linker will be used. (Note that we will replace any Lysine that is already present in the binding loop to prevent any interaction with the antibody binding loop).

Significance

The relevance of the results in this project will contribute to the development of new drug candidates for additional and targeted treatment of a subset of RA with poorer prognosis. If we are successful one may speculate if a compound that neutralizes ACPA could work as standalone therapy or as we believe work in combination with other anti-rheumatic drugs. An interesting combination would be anti-B cells therapy plus neutralization of autoantibodies. It has also been reported that in most cases, bone erosions continue although symptoms are relieved by different anti rheumatic drugs. Thus, removal of ACPA by neutralization may be a way for stopping further subclinical inflammation and progression of joint damages. If ACPA blocking will help against the often debilitating pain these patients sense this would be sensational. Moreover, chronic pain is very common in this disease despite disease activity is low. If it becomes possible to block the ACPA binding domain in its active conformation within the stable scaffold peptides, then the outcome will be a more stable and potent binder for ACPA due to improved bioavailability of the epitope. Such molecules will represent drug-leads for the development of new drugs with a new principle of mechanism of action. Finally, the methodology will allow for personalized medicine by using the same epitopes for diagnostics and subsequent treatment. The methodology will also allow for an efficient way of purifying autoantibodies present in RA but also in other autoimmune diseases allowing for their characterizations and contribution to pathology.

Preliminary results.

Identification of citrullinated fibrinogen peptides and characterization as autoanti- gens

We have previously identified endogenously citrullinated residues at positions R573Cit and R59iCit within the alpha fibrinogen chain as well as R72Cit and R74Cit in the corresponding beta chain ²⁸ . In a cohort consisting of 936 RA patients and 461 matched controls, we have demonstrated that >6s% of patients display a positive antibody response to at least one of these citrullinated peptides. More than 90% of patients display a response to a combination of two of these peptides ² ^

Purification of ACPA from patients and ACPA neutralization assays: We have purified ACPA from synovial fluid (SF) and plasma of RA patientss ⁰ .. The method allows purification of mg amounts of human ACPA and pools of ACPA are prepared by combin- ing material from 6-38 individuals. ACPA binding is measured using commercial ELISA assays. Our characterized fibrinogen peptides were individually or in combinations incubated with different ACPA pools and their blocking efficiency was expressed as percentage of inhibition and IC50. At physiological relevant concentration (deduced from concentrations in patient sera), the purified ACPA pool was treated with peptides containing the specific citrulline residues. The four peptides used are referred to as fibrinogen a chain peptide fragments "Cit573", "Cit59i" as well as fibrinogen β chain peptide fragments "Cit72", "Cit74", respectively. The unmodified corresponding arginine peptides were used as controls.

Results from the ACPA neutralization assay: Figure 6 shows dose-response curves for ACPA neutralizing peptides. Cit573 resulted in the highest degree of ACPA neutralization, 84% (Figure 6.a) with an average calculated IC50 of 59 μΜ. Cit59i resulted in a similar dose response curve as peptide Cit573, although not as efficient. A maximum of 63% ACPA neutralization was recorded for Cit59i (Figure 6.b) and the IC50 calculated was 190 μΜ. The non-citrullinated version only showed marginal inhibition (Figure 6.a). When combined, the citrullinated peptides from the alpha chain displayed a synergistic effect reaching 91% inhibition (Figure 7). Thus, these in vitro studies confirm that endogenously citrullinated fibrinogen peptides efficiently bind to ACPA.

The dose-response curves for ACPA neutralization with fibrinogen peptides (Arg/Cit 573 and 591) in Figure 6 shows the following: Citrullinated 573 peptide (Figure 6a) resulted in the highest degree of inhibition in the Elisa assay, 84%. Citrullinated 591 resulted in a similar dose response curve as peptide 573 (Figure 6b). X-axis show peptide concentrations (μΜ) and Y-axis show the percentage of ACPA neutralization (%). Circles represent means of 2 to 7 experiments per ACPA pool and error bars represent SEM. ^* P<o.05 and ^** P<o.ooi, significantly different from values found with arginine peptides.

Dose-response curves for ACPA neutralization with a combination of the two fibrino- gen-a chain peptides (Arg/Cit 573 and 591) in Figure 7 shows that two citrullinated peptides mixed together in equal amounts inhibit 91% of ACPA. X-axis shows peptide concentration (μΜ) and Y-axis shows the percentage of ACPA neutralization (%).

Initial syntheses and optimization of ACPA blocking efficiency: To determine whether cyclization improves ACPA binding activity a cyclized version of the full-length Cit 573 peptide was synthesized (cyclic FIB 573 CIT, shortened to CIT573Cyc in the figure) and tested in the ACPA binding assay. Interestingly, cyclic FIB 573 CIT gave 92% ACPA inhibition with an IC50 of 28 μΜ (Figure 8). Thus, cyclization appears to have resulted in significantly improved affinity to ACPA in comparison to the linear counterpart (CIT573Cyc in Figure 8). Moreover, two truncated peptides of 15 residues and 11 residues {FIB 573 CIT (lini) and FIB 573 CIT (lin2)}, originating from the full length Cit573 were tested and found to provide 75% (IC50 51 μΜ) and 69% (IC50 123 μΜ) ACPA inhibition respectively (Cit573Lini and Cit573Lin 2 in the figure). These results also suggest that the full residue span within Cit573 may not be necessary for antibody binding and could be dissected down to a fewer residues by designing truncated peptide variants of FIB 573 CIT. Together with the concept of cyclization, it is likely that a minimalized version of FIB 573 CIT that efficiently binds to ACPA can be designed. It is also possible that introduction of non-mammalian amino acids or chemically modified residues may enhance the affinity further. The insertion of the truncated linear l peptide into one of the loops of SFTI-i {SFTI FIB 573 CIT (lini)} displayed high ACPA inhibition percentage reaching up to 86% ACPA inhibition (IC50 19 μΜ). We can conclude that the use of the SFTI-i scaffold improved the IC50 in comparison to the original linear peptides and, is comparative with the effect of the cyclic version of the original peptide and its truncated fragment {cyclic FIB 573 CIT/ FIB 573 CIT (lini)}. Initial structural studies, using NMR, demonstrate that the SFTI scaffold is intact, i.e. the secondary loop (including the disulfide bond) maintains the structure.

In addition, we have shown that adding a Lysine in the structural loop, to facilitate immobilization of peptide on e.g. resins for purification and for the development of the diagnostic tool, does not influence activity. Sequences CFPKDGRCGIAEFPScit- GRSSSYS (SEQ ID NO 92) and CFPDGKCGIAEFPScitGRSSSYS (SEQ ID NO 93) have been synthesized.

Effects of grafted peptide on ACPA induced MO migration. Incubation of ACPA (^g) with SFTI FIB 573 CIT (lini) (86 nmol) for one hour at 37°C prior to dermal fibroblast treatment rendered a significant (0.5 fold) decrease in comparison with ACPA treatment alone (Figure 9).

Figure 9 shows the effect of SFTI FIB 573 CIT (lini) on ACPA induced fibroblasts migration. SFTI FIB 573 CIT (lini) reduced the fibroblasts migration rate in comparison to the treatment of the cells with ACPA alone. X-axis shows the different conditions tested with the dermal fibroblasts and Y-axis shows migration rate fold.

In vivo and in vitro models of pain: As stated in the introduction, the role of ACPA in RA pathogenesis is unclear. We have made the striking observation that ACPA induces long-lasting pain-like behavior in mice, without generating any visual or histological signs of joint inflammation or sickness). Intravenous injection of ACPA (0.125-1 mg) to mice induces long-lasting (> 7 days) pain-like behavior in form of mechanical hypersensitivity and reduced locomotion, assessed by Von Frey filments and the comprehensive laboratory animal monitoring system (CLAMS), respectively Thus, we propose that certain antibodies directly activate pain nerves and function as pain- inducing molecules. For the first time, we can now test the hypothesis that blocking of these autoantibodies may relief the symptoms of pain. Potentially, this opens a new avenue for treatment, not only of RA-induced pain, but also pain in other autoimmune diseases. Pain behavior in mice treated with ACPA in the presence or absence of the described novel ACPA blocking compounds will be assessed by measuring changes in different pain modalities (sensitivity to mechanical and cold stimulation) and changes in normal behavior (locomotion, burrowing and gait). In addition to these in vivo studies, we will be able to address mechanistically the mechanism of action of the new compounds by analyzing their ability to prevent ACPA induced i) release of neurotransmitters like CGRP and glutamate induced by activation of peripheral sensory nerves, ii) Ca ²⁺ -flux, live imaging of Ca ²⁺ increase in neurons provides an indirect measure of action-potential generation within individual neurons, and iii) measurement of action potential conduction by patch-clamp electrophysiolo- gy-

In vitro model of bone and cartilage destruction: CD14+ monocytes are isolated from peripheral blood of healthy individuals and/or ACPA positive RA patients and cul- tured in the presence of GM-CSF and IL-4 to generate dendritic cells (DC) or with GM-CSF to generate macrophages (ΜΦ). Following 6 day incubation, DCs and ΜΦ are incubated in the presence of RANKL and M-CSF, with or without ACPA IgG or flow through IgG (at a final concentration of 100 ng/ml). Osteoclasts differentiation is evaluated by total number of multinucleated TRAP+ cells. In parallel, cultures are grown on artificial bone surfaces and amount of resorbed areas are evaluated by computer assisted image analysis as percentages of total area. Our preliminary results demonstrated that ACPA isolated from the synovial fluid of RA patients have the ability to mediate osteoclastogenesis from monocytes of ACPA-positive RA individuals, in both a classic macrophage-driven osteoclastogenesis assay as well as in a dendritic cell-derived osteoclastogenesis. These results are in agreement with our previous published data ² ? and suggest that ACPA isolated from patients mediate the same cellular events as specific antibodies against citrullinated vimentin. We will study the optimized ACPA blockers in these assays in order to achieve proof of concept for the beneficial role of ACPA neutralization in bone and cartilage homeostasis in rheumatoid arthritis.

Administration

The peptides are preferably administered intraarticularly, intraveneously or subcuta- neously dissolved in physiological saline solution. SFTI FIB 573 CIT (lini) is given at a molar ratio of administered ACPA antibody/peptide of between 1/1000-10000; e.g. for an administered dose of 50 ug i.a. in the mouse in vivo model of pain-like behavior, 700 ug of SFTI FIB 573 CIT (lini) is given.

ACPA binding agents include ACPA blocking agents and may be used as diagnostic tools for purification and detection of ACPA. This includes different subpopulations of ACPA, ie antibodies directed to protein fragments, such as

a-enolase 5 KIHAXEIFDSXGNPTVE 21,

Vimentin 59 VYATXSSAVXLXSSVP 74,

1 STXSVSSSSYXXMFGG 16

Coallagen type II 59 AXGLTGXPGDA 369,

Filaggrin 306 SHQESTXGXRGRSGRG 324

Fibrinogen/fibrin a-chain 36 GPXWEXHQSACKDS

b-chain 60 XPAPPPISGGGYXAX 74,

which all are known targets of citrullination. The second loop of SFTI will then be used to anchor the peptide to a solid support, e.g a column for separation or a plate for development of Elisas.

A method for developing compounds against ACPA induced pain in vivo.

The experimental animal model consists of an animal that is injected with antibodies that recognize citrullinated proteins (ACPA). The animal starts to develop pain that can be measured and quantified by conventional methods by anyone skilled to the art, like Von Frey filaments, hot plate, Hargreaves thermal test etc. The effect of the inhibitor can be measured by comparing animals treated with ACPA with animals treated with the compound and ACPA.

References

Cerqueira, C, Klareskog, L. & Jakobsson, P.J. Neutralization of anti-citrullinated protein antibodies (ACPA) in rheumatoid arthritis - A way to go? Basic & Clinical Pharmacology & Toxicology 114, 17-17 (2014).

. Avouac, J., Gossec, L. & Dougados, M. Diagnostic and predictive value of anti-cyclic citrullinated protein antibodies in rheumatoid arthritis: a systematic literature review. Ann Rheum Dis 65, 845-851 (2006).

. Schellekens, G.A., et al. The diagnostic properties of rheumatoid arthritis antibodies recognizing a cyclic citrullinated peptide. Arthritis Rheum 43, 155-163 (2000).

. Agrawal, S., Misra, R. & Aggarwal, A. Autoantibodies in rheumatoid arthritis: association with severity of disease in established RA. Clin Rheumatol 26, 201-204

(2007) .

Johansson, M., Arlestig, L., Hallmans, G. & Rantapaa-Dahlqvist, S. PTPN22 polymorphism and anti-cyclic citrullinated peptide antibodies in combination strongly predicts future onset of rheumatoid arthritis and has a specificity of 100% for the disease. Arthritis Res Ther 8, R19 (2006).

. Kallberg, H., et al. Gene-gene and gene-environment interactions involving HLA- DRBi, PTPN22, and smoking in two subsets of rheumatoid arthritis. Am J Hum Genet 80, 867-875 (2007).

Plenge, R.M., et al. Replication of putative candidate-gene associations with rheumatoid arthritis in >4,ooo samples from North America and Sweden: association of susceptibility with PTPN22, CTLA4, and PADI4. Am J Hum Genet 77, 1044-1060 (2005).

. Klareskog, L., Ronnelid, J., Lundberg, K., Padyukov, L. & Alfredsson, L. Immunity to citrullinated proteins in rheumatoid arthritis. Annu. Rev. Immunol. 26, 651-675

(2008) .

. van der Helm-van Mil, A.H., Verpoort, K.N., Breedveld, F.C., Toes, R.E. & Huizinga, T.W. Antibodies to citrullinated proteins and differences in clinical progression of rheumatoid arthritis. Arthritis Res Ther 7, R949-958 (2005).

. Trouw, L.A., et al. Anti-cyclic citrullinated peptide antibodies from rheumatoid arthritis patients activate complement via both the classical and alternative pathways. Arthritis Rheum 60, 1923-1931 (2009).

. Banda, N.K., et al. Alternative complement pathway activation is essential for inflammation and joint destruction in the passive transfer model of collagen- induced arthritis. J Immunol 177, 1904-1912 (2006).

. Bos, W.H., et al. Arthritis development in patients with arthralgia is strongly associated with anti-citrullinated protein antibody status: a prospective cohort study. Ann Rheum Dis 69, 490-494 (2010).

. Sokolove, J., Zhao, X., Chandra, P.E. & Robinson, W.H. Immune complexes containing citrullinated fibrinogen costimulate macrophages via Toll-like receptor 4 and Fcgamma receptor. Arthritis Rheum 63, 53-62 (2011).

. Clavel, C, et al. Induction of macrophage secretion of tumor necrosis factor alpha through Fcgamma receptor Ila engagement by rheumatoid arthritis- specific autoantibodies to citrullinated proteins complexed with fibrinogen. Arthritis Rheum 58, 678-688 (2008).

Harre, U., et al. Induction of osteoclastogenesis and bone loss by human autoantibodies against citrullinated vimentin. J Clin Invest 122, 1791-1802 (2012).. Jahns, R., Schlipp, A., Boivin, V. & Lohse, M.J. Targeting receptor antibodies in immune cardiomyopathy. Semin Thromb Hemost 36, 212-218 (2010).

Munch, G., et al. Administration of the cyclic peptide COR-i in humans (phase I study): ex vivo measurements of anti-betai-adrenergic receptor antibody neutralization and of immune parameters. Eur J Heart Fail 14, 1230-1239 (2012).. Craik, D.J., Daly, N.L., Bond, T. & Waine, C. Plant cyclotides: A unique family of cyclic and knotted proteins that defines the cyclic cystine knot structural motif. JMol Biol 294, 1327-1336 (1999).

. Gunasekera, S., et al. Engineering stabilized vascular endothelial growth factor-A antagonists: synthesis, structural characterization, and bioactivity of grafted analogues of cyclotides. J Med Chem 51, 7697-7704 (2008).

. Wong, C.T., et al. Orally active peptidic bradykinin Bi receptor antagonists engineered from a cyclotide scaffold for inflammatory pain treatment. Angew Chem IntEd Engl 51, 5620-5624.

. Eliasen, R., et al. Design, synthesis, structural and functional characterization of novel melanocortin agonists based on the cyclotide kalata Bi. J Biol Chem 287, 40493-40501.

. Chan, L.Y., et al. Engineering pro-angiogenic peptides using stable, disulfide-rich cyclic scaffolds. Blood 118, 6709-6717.

. Luckett, S., et al. High-resolution structure of a potent, cyclic proteinase inhibitor from sunflower seeds. JMol Biol 290, 525-533 (1999).

. Goransson, U., Burman, R., Gunasekera, S., Stromstedt, A.A. & Rosengren, K.J. Circular proteins from plants and fungi. J Biol Chem 287, 27001- 27006.

. Montgomery, A.B., Venables, P.J. & Fisher, B.A. The case for measuring antibodies to specific citrullinated antigens. Expert review of clinical immunology 9, 1185-1192 (2013).

. Gunasekera, S., Aboye, T.L., Madian, W.A., El-Seedi, H.R. & Goransson, U. Making Ends Meet: Microwave-Accelerated Synthesis of Cyclic and Disulfide Rich Proteins Via In Situ Thioesterifi cation and Native Chemical Ligation. Int JPept Res Ther 19, 43-54 (2013).

. Harre, U., et al. Induction of osteoclastogenesis and bone loss by human autoantibodies against citrullinated vimentin. The Journal of clinical investigation 122, 1791-1802 (2012).

. Hermansson, M., et al. MS analysis of rheumatoid arthritic synovial tissue identifies specific citrullination sites on fibrinogen. Proteomics 4, 511-518 (2010).

. Ossipova, E., et al. ACPA response against fibrinogen epitopes citrullinated in vivo in the synovial membrane in RA patients detected with an autoantibody microarray. Annals of the Rheumatic Diseases 71(2012).

. Ossipova, E., et al. Affinity purification and characterisation of human ACPAs. Manuscript (2014). Name SEQ ID Sequence Organism

NO

FIB 573 CIT 1 HHP GIA EFP SXG KSS SYS KQF Human

FIB 573 CIT (lin 1) 2 GIA EFP SXG KSS SYS Human

FIB 573 CIT (lin 2) 3 A EFP SXG KSS S Human cyclic FIB 573 CIT 4 CHHP GIA EFP SXG KSS SYS KQF Human derivative cyclic FIB 573 CIT (lin 1) 5 CGIA EFP SXG KSS SYS Human derivative cyclic FIB 573 ARG (lin 1 ) 6 CGIA EFP S R G KSS SYS Human derivative cyclic FIB 573 ALA (lin 1) 7 CGIA EFP S A G KSS SYS Human derivative

SFTI-1 8 CFPDGR CTKSIPPI Asteraceae

SFTI FIB 573 CIT (lin 1) 9 CFPDGR CGIA EFP SXG KSS SYS Human and

Asteraceae

SFTI FIB 573 ARG (lin 1) 10 CFPDGR CGIA EFP S R G KSS SYS Human and

Asteraceae kalata B l 1 1 CGETCVGGTCNTPGCTCSWPVCTRN Rubiaceae

GLPV

CCP-1 12 HQCHQESTXGRSRGRCGRSGS Artificial

MCoTI-II 13 CPKILKKCRRDSDCPGACICRGN- Cucurbitaceae

GYCGSGSDGGV

MCoTI-II (lin 2) 14 CPKILKKCRRDSDCPGACICRGN- Human and

GYCGIAEFPSXGKSSSYS Cucurbitaceae

TABLE 1: Sequences made for ACPA-inhibition (X is CitruUine).

Name SEQ ID Sequence Organism

NO

a-enolase 15 5 KIHAXEIFD SXGNPTVE 21 Human

Vimentin fragment 1 16 1 STXSVSSSSYXXMFGG 16 Human

Vimentin fragment 2 17 59 VYATXSSAVXLXSSVP 74 Human

Collagen 18 59 AXGLTGXPGDA 69 Human

Filaggrin 19 306 SHQESTXGXSRGRSGRSG 324 Human Fibrinogen a-chain fragment 20 36 GPXWEXHQSACKDS 50 Human 1

Fibrinogen a-chain fragment 559 ESSSHHPGIAEFPSXGK 575 Human 2

Fibrinogen a-chain fragment 563 HHPGIAEFPSXGKSSSYSKQF 583 Human 3

Fibrinogen a-chain fragment 580 SKQFTSSTSYNXGDSTFESKS 600 Human 4

Fibrinogen b-chain fragment 60 XPAPPPISGGGYXAX 74 Human 1

Fibrinogen b-chain fragment 52 KXEEAPSLXPAPPPISGGGYXAX Human 2 PAK 77

Fibrinogen b-chain fragment 60 APPPISGGGYXARPAKAAAT 81 Human 3

Fibrinogen b-chain fragment 60 APPPISGGGYRAXPAKAAAT 81 Human 4

TABLE 2: ACPA reactive peptide fragments (X is Citrulline)

Name SEQ ID Sequence Organism

NO

SFT-L1 28 CFPDGCIEGSPV Asteraceae

PDP-16 29 CYPDGRCTRSNPPI Asteraceae

PDP-3 30 CYPDGRCTKSIPPI Asteraceae

PDP-12 31 CFPDGRCTKSIPPV Asteraceae

PDP-15 32 CFPDGRCTRSIPPI Asteraceae

PDP-4 33 CFRRDGSCFGAF Asteraceae

PDP-5 34 CYMDGRYRRCIPGMFRAY Asteraceae

PDP-13 35 CYMDGRYRRCIPGMFRSY Asteraceae

PDP-14 36 CYMDGRCRAGMFRSY Asteraceae

PDP-7 37 GCLRDHCIPTTSGPI Asteraceae

PDP-8 38 CFRLPDGGRLCVPPG Asteraceae

PDP-17 39 CTPDGDCHWIPAPPFFM Asteraceae

PDP-9 40 CTPDGDCYWTSTPPFFT Asteraceae

PDP-6 41 CFSDGHCIQVPPMATEI Asteraceae

PDP-18 42 CTPDGRCYPVPYPPFYT Asteraceae

Table 3: Scaffold sequences

Name SEQ Sequence Organis

ID NO

Fibrin a-chain fragment 1 43 36 GPXWEXHQSACKDS 50 Human

Fibrin a-chain fragment 2 44 171VDIDIKIXSCXGSCS 185 Human

Fibrin a-chain fragment 3 45 181 SCSXALAXEVDLKDY 197 Human

Fibrin a-chain fragment 4 46 246 PEWKALTDMPQMXME 260 Human

Fibrin a-chain fragment 5 47 259 MELEXPGGNEITXGG 273 Human

Fibrin a-chain fragment 6 48 366 EXGSAGHWTSESSVS 380 Human

Fibrin a-chain fragment 7 49 396 DSPGSGNAXPNNPDW 410 Human

Fibrin a-chain fragment 8 50 411 GTFEEVSGNVSPGTX 325 Human

Fibrin a-chain fragment 9 51 501 SGIGTLDGFXHXHPD 515 Human

Fibrin a-chain fragment 10 52 546 SXGSESGIFTNTKES 560 Human

Fibrin a-chain fragment 11 53 589 SYNXGDSTFESKSYK 602 Human

Vimentin fragment 3 54 71 LXSSVPGVR 79 Human

Vimentin fragment 4 55 263 PDLTAALRDVRQQYESVAAK 281 Human

Vimentin fragment 5 56 295 FADLSEAANRNNDALRQAK 313 Human

Vimentin fragment 6 57 345QMXEMEENFAVEAANYQDTIGR 364 Human

Vimentin fragment 7 58 404 LLEGEEXISLPLPNFSSLNLR 423 Human

Vimentin fragment 8 59 447 TVETX DGQVINETSQHHDDLE 465 Human

Fillagrin fragment 1 60 306 SHQESTXGRSRGRSGRSGS 324 Human

Collagen type II fragment 61 916 GDKGEAGEPGEXGLKGHXGFTGLQ 939 Rat

BiP fragment 1 62 273 RKDNRSVQKLXREVEKAKRA 298 Human

BiP fragment 2 63 295 AKRALSSQHQAXIEIESFFE 314 Human a-Enolase fragment 1 64 394 TGAPCXSEXLAK 405 Human a-Enolase fragment 2 65 422 FAGXNFXNPLAK 433 Human a-Enolase fragment 3 66 405 YNQLLXIEEELGSK 419 Human a-Enolase fragment 4 67 256 YDLDFKSPDDPSXYISPDQLADLYK 280 Human

Apolipoprotein E fragment 68 197 XLGPLVEQGX 207 Human

Myeloid nuclear differentiation anti69 121 KLTSEAXGRIPVAQK 135 Human gen, fragment

b-actin, fragment 70 190 MKILTEXGYSFTTTAEXEIVRDIKEK 216 Human alpha-2-macroglobulin, fragment 71 705 VGFYESDVMGXGHAR 719 Human alpha-2-macroglobulin, fragment 72 1169 SLNEEAVKKDNSVHWERPQKPK 1190 Human apolipoprotein B- 100, fragment 73 3211 NXNNALDFVTK 3221 Human alpha-2-macroglobulin 74 705 VGFYESDVMGXGHAR 719 Human complement C3 75 1365 VTIKPAPETEKXPQDAK 1381 Human fibronectin 76 977 VFAVSHGXESKPLTAQQTTK 986 Human fibronectin 77 1029 LTVGLTXXGQPR 1039 Human alpha- 1 -antitrypsin 78 158 FLENEDXR 165 Human histone HI .2 79 53 EXSGVSLAALKK 64 Human histone H3.lt 80 25 VAXKSAPATGGVK 37 Human histone HI .4 81 53 ERSGVSLAALK 63 Mouse

Putative ribosomal RNA methyltrans- 82 142 KLLPIEXAALKQK 154 Mouse ferase NOP2

Putative ribosomal RNA methyltrans- 83 80 KGAVQAXGKKRPA 92 Mouse ferase NOP3

Splicing factor 3B subunit 1 84 151 TYMDVMXEQHLTK 163 Mouse

THO complex subunit 4 85 134 AAVHYDXSGRSLG 146 Mouse

THO complex subunit 4 86 137 HYDRSGXSLGTAD 149 Mouse

Myb-binding protein 1A 87 1316 RLSLVSXSPSLLQ 1328 Mouse

60S ribosomal protein L10;60S ribo88 26 VPDAKIXIFDLGR 38 Mouse somal protein LlO-like

Transcription intermediary factor 1 - 89 464 HVSGMKXSXGEG 475 Mouse beta

Coiled-coil domain-containing protein 90 402 LRSIEKXDTLALL 141 Mouse 86

60S ribosomal protein L19;Ribosomal 91 31 IASRQQIXKL 40 Mouse protein LI 9

Lysine mutant 1 92 CFPKDGRCGIAEFPSXGRSSSYS Human and Aste- raceae

Lysine mutant 2 93 CFPDGKCGIAEFPSXGRSSSYS Human and Aste- raceae

Table 4: ACPA reactive peptide fragments (X is Citrulline)