Conferring extracellular matrix affinity enhances local therapeutic efficacy of anti-TNF-α antibody in a murine model of rheumatoid arthritis

Background Although disease in a majority of rheumatoid arthritis (RA) patients is often initially limited to one or a few joints, currently approved medications including anti-tumor necrosis factor-α antibody (α-TNF) are injected systemically. Given that α-TNF systemic injection typically does not cure RA and involves risk of treatment-related adverse events, one possible approach to enhance therapeutic efficacy and reduce α-TNF systemic exposure is to retain the antibodies in arthritic joints after local administration. The aim of this study was to evaluate the approach of conferring extracellular matrix (ECM) binding affinity to α-TNF antibodies in a RA model. Methods α-TNF was chemically conjugated with a promiscuous ECM-binding peptide derived from placenta growth factor 2 (PlGF-2123-144). The binding activity of PlGF-2123-144-conjugated α-TNF (PlGF-2123-144-α-TNF) against ECM proteins was assessed by ELISA and by immunostaining on human cartilage specimens. The effect of conjugation on antibody function was assessed as a neutralizing activity against osteoclast differentiation. Retention at the injection site and therapeutic efficacy of PlGF-2123-144-α-TNF were tested in a collagen antibody-induced arthritis (CAIA) model in the mouse. Results PlGF-2123-144 peptide conjugation conferred α-TNF with affinity to ECM proteins without impairment of antigen recognition. PlGF-2123-144-α-TNF locally injected at a paw in the CAIA model was retained for at least 96 h at the injection site, whereas unmodified α-TNF was dispersed rapidly after injection. Local treatment with unmodified α-TNF did not suppress the arthritis score relative to isotype controls. By contrast, local administration of PlGF-2123-144-α-TNF suppressed arthritis development almost completely in the treated paw even at a 1000× lower dose. Conclusion These data demonstrate that retention of α-TNF in arthritic joints can suppress arthritis development and enhance therapeutic efficacy. This simple bioengineering approach of ECM-binding peptide conjugation offers a powerful and clinically translational approach to treat RA.


Background
Rheumatoid arthritis (RA) is a chronic, progressive autoimmune disease characterized by the formation of hyperplastic synovial pannus tissue, which causes joint destruction [1,2]. Biological therapies to block cytokine signals in the human body, such as anti-TNFα (α-TNF) and anti-IL-6 receptor antibodies, are powerful approaches for treating RA [3][4][5][6]. However, currently approved medications show incomplete response in terms of joint damage by persistent synovitis even in patients in clinical remission, and they have the possibility of significant side effects by suppressing systemic immunity [7][8][9][10][11]. Arthritis is classified as monoarthritis, oligoarthritis, and polyarthritis, by the numbers of joints in which symptoms appear. Many polyarthritic disorders initially present as a monoarthritis. In patients with recent-onset arthritis, monoarticular involvement (38.3%) is more common than oligoarticular (34.1%) or polyarticular (24.6%) disease [12]. Considering the worse prognosis of polyarticular arthritis, early intervention is critical to maximize outcomes [12,13].
Direct intra-articular administration of drugs is a reasonable approach to enhance their local efficacy and reduce systemic adverse effects and is relevant for the treatment of monoarthritis and oligoarthritis. However, drugs injected into the joint are rapidly cleared via lymphatics or sub-synovial capillaries in the absence of a retention strategy [14]. In addition, the clearance rate of macromolecules from inflamed joints is increased in patients with RA due to enhanced drainage from the joint space by greater synovial lymph flow [15]. Although intraarticular α-TNF injections have been shown to be well tolerated and improve clinical signs in RA patients in small clinical trials, the therapeutic effect was limited and transient [16,17]. These results indicate that intra-articular injection is a promising approach for effective therapy, but requires an approach for local tissue retention.
Heparin-binding domains (HBDs) derived from growth factors (GFs) bind to HBDs derived from a variety of extracellular matrix (ECM) proteins [18][19][20][21][22]. By screening the binding of GFs to a variety of ECM proteins, we have discovered that the HBD of placenta growth factor-2 (PlGF-2 123-144 ) has an exceptionally high affinity for multiple ECM proteins [23]. Chemical conjugation of the PlGF-2 123-144 peptide to immune checkpoint blockade and immune agonist antibodies enhanced antibody retention within tumors when injected near the tumor, improving therapeutic efficacy and safety by limiting systemic exposure [24,25].
Here, we have hypothesized that utilizing ECMbinding technology for prolonged tissue retention would improve the ability of an anti-inflammatory cytokine antibody to effectively suppress RA development. We engineered α-TNF conjugated with PlGF-2 123-144 peptide for RA therapy. We tested whether the enhanced retention of α-TNF in tissue surrounding the joints would improve its local efficacy against arthritis development.

Matrix-assisted laser desorption/ionization-time-of-flight mass spectrometry
Antibodies were analyzed by Matrix-assisted laser desorption/ionization-time-of-flight mass spectrometry (MALDI-TOF MS) using a Bruker Ultraflextreme MALDI TOF/TOF instrument. All spectra were collected with acquisition software Bruker flexControl™ and processed with analysis software Bruker flexAnalysis™. First, a saturated solution of the matrix, α-cyano-4hydroxycinnamic acid (MilliporeSigma, Billerica, MA, USA), was prepared in 50:50 (v/v) acetonitrile:(1% trifluoroacetic acid in water) as a solvent. The analyte in phosphate-buffered saline (PBS; 5 μL, 0.1 mg/mL) and the matrix solution (25 μL) were then mixed, and 1 μL of that mixture was deposited on the MTP 384 ground steel target plate. The drop was allowed to dry in a nitrogen gas flow, which resulted in the formation of uniform sample/matrix co-precipitate. All samples were analyzed using the high mass linear positive mode method with 5000 laser shots at a laser intensity of 75%.
The measurements were externally calibrated at three points with a mix of carbonic anhydrase, phosphorylase B, and bovine serum albumin.

Immunohistochemistry
Frozen sections of cartilage from an osteoarthritis patient were purchased from OriGene Technologies (Rockville, MD, USA). The sections were blocked with 2% FBS in PBS overnight at room temperature. The sections were incubated with primary antibodies for 3 h at room temperature. Either 50 μg/mL of rat α-TNF or equimolar PlGF-2 123-144 -α-TNF, 5 μg/mL of rabbit anti-human collagen II antibody (Abcam), or 25 μg/mL of mouse antihuman decorin antibody (R&D Systems, Minneapolis, MN, USA) were used as primary antibodies. After incubating with the fluorescently tagged secondary antibodies, slides were covered with ProLong Gold Antifade Mountant with DAPI (Thermo Fisher Scientific). The images were scanned with a Pannoramic digital slide scanner (3DHistech, Budapest, Hungary) and analyzed using a Pannoramic Viewer software (3DHistech).
Osteoclast differentiation from RAW264.7 cells RAW264.7 cells were seeded on 60-mm dishes at a density of 3 × 10 6 cells per well and cultured for 7 days in the presence of 50 ng/mL of recombinant mouse TNFα (Insight Biotechnology, Middlesex, UK), 10 ng/mL of recombinant mouse receptor activator of nuclear factor κΒ ligand (RANKL; PeproTech), and 50 to 1000 ng/ mL of test antibody. The cells were then washed with PBS and stained using Acid Phosphatase, Leukocyte (TRAP) kit (MilliporeSigma) according to the manufacture's instruction. TRAP-positive multinucleated cells having three or more nuclei were considered as osteoclasts, and their number was counted in each well. Percent inhibition was calculated from a ratio of the number of osteoclasts in each treatment to that in the untreated control. The 50% inhibitory concentration (IC 50 ) values were obtained by nonlinear regression analysis (GraphPad Prism Software, LaJolla, CA, USA).

Mice
Balb/c and athymic nude mice at 8 to 9 weeks of age were obtained from the Jackson Laboratory (Bar Harbor, ME, USA). Experiments were performed with approval from the Institutional Animal Care and Use Committee of the University of Chicago (Chicago, IL, USA).

Collagen antibody-induced arthritis model
Arthritis was induced in Balb/c mice by intraperitoneal injection of anti-collagen antibody cocktail at 1.5 mg (Chondrex, Redmond, Washington, USA) on day − 3, followed by intraperitoneal injection of LPS at 50 μg (Chondrex) on day 0. On the day of LPS injection, mice were injected at the left hind footpad with PBS, 15 μg of PlGF-2 123-144 peptide (theoretical equivalent amount of peptides within 100 μg of PlGF-2 123-144 -α-TNF), 100 μg of control IgG (rat IgG1 isotype control, Bio X Cell), 100 μg of unmodified α-TNF, or 0.1 or 1 μg of PlGF-2 123-144 -α-TNF. Joint swelling in the left and right hind paws was scored daily as described elsewhere [26]. The left hind paws of the collagen antibody-induced arthritis (CAIA) mice were fixed in 10% neutral formalin (Milli-poreSigma), decalcified in Decalcifer II (Leica Microsystems Inc., Buffalo Grove, IL, USA), and then embedded in paraffin. The paraffin-embedded paws were sliced at 5 μm thickness and stained with hematoxylin and eosin (H&E). The images were scanned with a Pannoramic digital slide scanner and analyzed using a Pannoramic Viewer software.
In vivo bio-distribution study α-TNF was fluorescently labeled using sulfo-Cyanine 7 (Cy7) NHS ester (Lumiprobe, Broward, FL, USA) according to the manufacture's instruction. PlGF-2 123-144 peptide that was labeled with Cy7 in its N-terminus was chemically synthesized with > 90% purity by Thermo Fisher Scientific. Cy7-labeled PlGF-2 123-144 was conjugated to α-TNF as described above. The Cy7 labeled antibody at a dose of 20 μg was injected at the left hind footpad of CAIA mice or athymic nude mice. The fluorescence level was measured using the Xenogen IVIS Imaging System 100 (Xenogen, Alameda, CA, USA) under the following conditions: f/stop: 2; optical filter excitation 745 nm; excitation 800 nm; exposure time: 5 s; small binning. Retention rates of the injected antibody immediately after the injection and that at 168 h after the injection were defined as 100% and 0%, respectively.

Measurement of plasma concentration of the injected antibodies
CAIA mice received 1 μg of PlGF-2 123-144 -α-TNF or 100 μg of unmodified α-TNF. Blood was collected from the tail vein with a heparinized capillary at 24, 48, 72, and 96 h after the injection, and plasma was obtained by centrifugation. The injected antibodies in plasma were measured by ELISA as described above.

Statistical analysis
Statistical analyses were performed using GraphPad Prism software, and P < 0.05 was considered statistically significant. Binding activities against ECM proteins were analyzed using Student's t test for comparisons between PlGF-2 123-144 -α-TNF and unmodified α-TNF. The retention effect of PlGF-2 123-144 -α-TNF was analyzed in the area under the percent of retention at the injection sitetime curve from 0 to 96 h (AUC 0-96h ) compared with unmodified α-TNF using Student's t-test. The AUC 0-96h of plasma concentration was analyzed using Student's ttest for comparisons between PlGF-2 123-144 -α-TNF and unmodified α-TNF. To compare the efficacy of PlGF-2 123-144 -α-TNF with unmodified α-TNF, the data on day 6 were analyzed using Tukey's multiple comparison test.

Results
PlGF-2 123-144 peptide is covalently conjugated to α-TNF The PlGF-2 123-144 peptide was covalently conjugated with α-TNF using a crosslinker. SDS-PAGE revealed that the molecular weights of both the light and heavy chains of α-TNF were increased (Fig. 1a). Under the stoichiometric conditions used, the α-TNF bound approximately 4.2 PlGF-2 123-144 peptides per antibody (average 13.4 kDa shift) as measured by MALDI-TOF MS (Fig. 1b). We have previously reported that multiple PlGF-2 123-144 peptides are conjugated to an IgG under the same reaction conditions [24], suggesting that this reaction is unaffected by antibody clones.

PlGF-2 123-144 -α-TNF retention is prolonged at the injected paw
Retention of PlGF-2 123-144 -α-TNF at the injection site was determined by in vivo bio-distribution analysis in CAIA mice. Fluorescently labeled α-TNF or PlGF-2 123-144 -α-TNF was injected at the left hind footpad of CAIA mice on the day following arthritis induction by LPS. Whole-body fluorescence was measured immediately after the injection, and 0.5, 1, 2, 3, 24, 48, 96, and 168 h after the injection (Fig. 4). The fluorescence level was detected at the injected paw, and retention of α-TNF at the injected paw was reduced immediately after the injection. To the contrary, the fluorescence of PlGF-2 123-144 -α-TNF was maintained at a higher level at the injected paw and was detectable until 96 h after the injection. We also investigated the bio-distribution using athymic nude mice. The fluorescence of PlGF-2 123-144 -α-TNF was detected only at the injected paw until 96 h after the injection and the retention of PlGF-2 123-144 -α-TNF was significantly greater than unmodified α-TNF (Additional file 1: Figure S1). These data suggest that Fig. 3 Effect of PlGF-2 123-144 conjugation on the recognition of α-TNF against the target antigen. a Blocking affinities of unmodified α-TNF and PlGF-2 123-144 -α-TNF to TNFα were analyzed by ELISA (n = 3, mean + SE). b Osteoclast precursor RAW264.7 cells were incubated with TNFα and either unmodified α-TNF or PlGF-2 123-144 -α-TNF for 7 days, and then differentiated osteoclasts were stained by TRAP kit. Percent inhibition of osteoclast differentiation calculated from the ratio of the number of osteoclasts in each treatment to that in the untreated control (n = 6, mean ± SE). c Representative images of osteoclast differentiation study taken at 5-fold magnification PlGF-2 123-144 -α-TNF binds to ECM proteins and is retained at the injection site.

Discussion
Interest in clinical translation of intra-articularly administered therapies is growing rapidly [28]. Systemic administration of bio-therapeutics can result in only a fraction of the drug reaching the targeted tissues, with the majority of the drug being distributed to tissues irrelevant to the drug's site of action. Localized therapy may allow for greater accumulation at target sites, better efficacy, and improved safety. Therefore, local injection is a reasonable approach for patients with monoarthritis or oligoarthritis who do not necessarily need to be exposed to systemic α-TNF treatment. In many cases, polyarthritic disorders initially present as a monoarthritis, which is potentially locally treatable. In this study, we demonstrate that PlGF-2 123-144 peptide conjugation enhanced the local efficacy of α-TNF by at least 1000× based on dose-comparison and that the therapeutic efficacy was robust. This demonstrates that PlGF-2 123-144 peptide conjugation enables α-TNF to be retained at joints to block TNFα signaling thoroughly and enhance its potency. Significant reduction of the effective dose and corresponding systemic exposure of locally-administered PlGF-2 123-144 -α-TNF suggests that this technology may help avoid unnecessary exposure of α-TNF to the whole body, with a corresponding reduction in adverse events.
PlGF-2 123-144 -α-TNF showed high efficacy when injected at the paw. Since the mouse paw contains many joints and disease scoring in a single joint over time is difficult, in this study we injected drugs nearby the joint space. PlGF-2 123-144 -α-TNF binds strongly to ECM proteins including collagen II, i.e., major components of cartilage and synovium, and was retained at the injection site. Therefore, these data support the concept that that  PlGF-2 123-144 -α-TNF could bind to ECM in cartilage and the synovium before draining from lymphatic vessels even when PlGF-2 123-144 -α-TNF is injected intra-articularly into the joint, although further evaluation of intra-articular injection using larger animals is warranted.
Various technologies have been proposed to achieve prolonged retention of locally injected drugs, such as particle-or liposome-based drug delivery systems and modification by polypeptide conjugation [28]. Although these technologies prolong the intra-articular residence time of biologic drugs [29][30][31], these approaches only slightly reduced systemic circulation and there is no report to show clearly their enhanced therapeutic efficacy against disease symptoms. Therefore, our data suggest a novel approach of tissue retention of α-TNF to efficiently and locally treat RA and reduce systemic exposure due to dose reduction.
For clinical translation of ECM-binding antiinflammatory bio-therapeutics, an advantage also lies in the use of the PlGF-2 123-144 peptide, which naturally exists in the human body as a domain on PlGF, limiting the possibility of recognition by the immune system. It was confirmed that the PlGF-2 123-144 peptide-conjugated antibody was capable of binding with human ECM proteins and human specimens. Also, the PlGF-2 123-144 peptide can be conjugated to antibodies with a simple chemical reaction. The advantage of this feature is in the simplicity in production, in that it is possible to work with antibodies for which production has already been optimized. The PlGF-2 123-144 peptide conjugation synthesis reaction for antibodies can be accomplished using chemistry that is analogous to PEGylation of proteins. Indeed, the same chemical reaction is used in antibody-drug conjugates, such as in the production of trastuzumab emtansine. These features may facilitate the development of ECM binding-drug therapy to overcome the barriers to clinical translation.

Conclusions
We have found that installing an ECM-binding affinity to α-TNF enables it to be retained at the injection site. Moreover, ECM-binding α-TNF showed robust therapeutic efficacy in a mouse model of RA at a very low dose. This simple approach of generating an engineered ECM-binding α-TNF bio-therapeutic may hold potential for clinical translation as a local anti-inflammatory therapeutic in RA.