Heparin-binding of the human chitinase-like protein YKL-40 is allosterically modified by chitin oligosaccharides

Abstract

The chitinase-like protein YKL-40 (CHI3L1) has been implicated in the pathophysiology of inflammation and cancer. Recent studies highlight the growing interest in targeting and blocking the activity of YKL-40 to treat cancer. Some of those targeting-strategies have been developed to directly block the heparin-affinity of YKL-40 with promising results. This study explores how short chain chitooligosaccharides (ChOS) affect the heparin-binding affinity of YKL-40. Our findings reveal that ChOS act as allosteric effectors, decreasing the heparin-binding affinity of YKL-40 in a size- and dose-dependent manner. Our results provide insights into the heparin affinity of YKL-40 and how ChOS can be used to target the heparin activity of YKL-40 in diseases. Since ChOS has many beneficial properties, such as being non-toxic and biodegradable, these results provide intriguing opportunities for applying them as allosteric effectors of the heparin-binding affinity of YKL-40.

Highlights

• •Chitin oligosaccharides A₄, A₅, and A₆ bind to YKL-40 with micromolar affinity.
• •Chitin oligosaccharides A₄, A₅, A₆, and chitooligosaccharides (ChOS) allosterically decrease the heparin affinity of YKL-40.
• •Rotation of the Trp99 sidechain out of the chitin-binding site is the crucial factor for decreased heparin affinity of YKL-40.

1. Introduction

The human YKL-40, also known as chitinase 3-like 1 protein (CHI3L1), is a secreted 40 kDa glycosylated protein and a member of the glycoside hydrolase 18 (GH18) family. Although YKL-40 lacks chitinase activity, due to a mutation of Glu to Leu in the active site [1], it still binds chitin and chitooligosaccharides (ChOS) with micromolar affinity [2]. Additionally, YKL-40 binds heparin/heparan sulfate [[2], [3], [4], [5]], and the heparin-carrying receptors syndecan-1 [6], syndecan-4 [7], and CD44v3 [8], as well as other receptors, such as IL-13Rα2 [9,10], and TMEM219 [11]. A broad range of studies has collectively revealed YKL-40′s activity in innate and adaptive immunity through its modulation of the inflammatory response, in tissue remodeling, and cancer progression [[12], [13], [14]]. However, the full extent of its activity and the mechanism by which it mediates its function is still not fully understood and remains the subject of ongoing investigations. YKL-40 plasma levels are low in healthy young adults but increase with age, and before the onset or during pathological conditions [[15], [16], [17]]. In diseases, upregulated YKL-40 expression is observed in highly metabolically active cells, such as inflammatory- and cancer cells [18,19]. YKL-40 is overexpressed in several cancers, including breast, glioblastoma, colorectal, and lung cancers. For instance, in glioblastoma, YKL-40 expression is associated with tumor aggressiveness, resistance to therapy, and poor prognosis [6,20]. Similarly, in breast cancer, YKL-40 is linked to enhanced tumor cell proliferation, migration, and invasion, as well as the suppression of anti-tumor immune responses [21,22]. In colorectal cancer, high plasma levels of YKL-40 have been correlated with advanced disease stages and reduced survival rates [[23], [24], [25], [26]]. In inflammatory conditions, YKL-40 is a modulator of immune cell recruitment, activation, and tissue repair [27,28]. Elevated serum levels of YKL-40 have been reported in chronic inflammatory diseases such as asthma, where it exacerbates airway remodeling and fibrosis, and liver fibrosis, where it serves as a marker of disease severity and progression [13,[27], [28], [29], [30]]. Due to the increased YKL-40 expression in various pathological conditions, YKL-40 has been proposed as a biomarker for inflammation and cancer [18,31]. Furthermore, YKL-40 has been suggested as a therapeutic target in some of these disorders [32]. Animal models utilizing knockout constructs of YKL-40 to elucidate its role as a therapeutic target in inflammation and cancer have shown promising results of decreased inflammation and tumor progression [22,25,[33], [34], [35], [36]].

The heparin-binding activity of YKL-40 has been proposed to be the functional binding site responsible for its effect in the tumor microenvironment [25,26,37]. Two separate locations within YKL-40 have been suggested as heparin-binding sites; one located in a surface loop at amino acid (aa) residues 143–149, GRRDKQH [2,5], and the other located in a KR-rich area at aa residues 334–345 at the C-terminus of YKL-40 [4,38]. Although the former site contains the typical heparin-binding motif BBXBXB (B is a basic amino acid, while X is any other amino acid) and has been suggested to bind heparin in silico [5], the latter, despite lacking this common motif, has been proposed as the functional binding site [4,38]. The heparin-binding site interacts with heparin/heparan sulfate proteoglycans (HSPGs) in the extracellular matrix (ECM) and on cell surface receptors. This binding is suggested to facilitate binding to cell surface receptors, leading to the activation of downstream signaling pathways involved in cancer cell proliferation, migration, invasion, and angiogenesis [25,37]. Several approaches hindering the heparin-binding of YKL-40 have shown promising results towards improved outcomes of inflammation and cancer progression [4,25,36,39]. Additionally, blocking YKL-40 binding to syndecan-1 and syndecan-4 has been found to result in decreased tumor angiogenesis [4,6,7,40]. Furthermore, experiments have shown YKL-40′s ability to modulate immune responses and contribute to immunosuppressive conditions in various context, including cancer, via heparin-binding to the heparin-carrying receptor CD44v3 [41]. These results make the heparin-binding site of YKL-40 a compelling target for therapeutic intervention aimed at disrupting its interactions within the tumor microenvironment to inhibit tumor growth and improve treatment outcomes.

ChOS is a mixture of hetero-oligosaccharides with repeating units of β-(1–4)-linked N-acetyl-glucosamine (GlcNAc) and d-glucosamine (GlcN). They are derivatives of chitin or chitosan and are typically produced through chemical and/or enzymatical treatments. Besides being water-soluble, non-toxic, and biodegradable, ChOS have gained attention due to their reported bioactivities, such as antioxidant action, anti-inflammatory effects, and tumor growth inhibition [[42], [43], [44], [45], [46], [47]]. Anti-tumor properties of ChOS have been investigated in cellular models in vitro where ChOS have been found to reduce matrix metalloproteinase expression, increase apoptosis, and decrease the proliferation of cancer cells [48,49]. Additionally, ChOS have been shown to reduce plasma levels of YKL-40 by reducing YKL-40 secretion by fibroblasts and myeloid cells in a syngeneic mouse lung cancer model [50].

There appears to be a significant overlap between the biological function of YKL-40 and ChOS. Since the negative effect of YKL-40 in cancer has been correlated with the heparin-binding of YKL-40, and ChOS have been shown to alleviate some of those effects, we hypothesized that their binding to YKL-40 allosterically induces conformational changes in YKL-40 that modulate its heparin-binding affinity. In this study we aimed to investigate if chitin oligosaccharides and/or ChOS affect the heparin-binding affinity of YKL-40 using a heparin-binding assay utilizing recombinant human YKL-40, ChOS blend, and chitin oligosaccharides of various lengths.

2. Materials and methods

2.1. Materials

Cell culture medium and supplements were obtained from Gibco® Life Technologies, USA. Nunc® cell culture flasks were from Thermo Scientific, USA. Protino Ni-NTA 5 ml column was from Macherey-Nagel, Germany. Native YKL-40 was bought from Quidel, USA. A human YKL-40 ELISA kit was purchased from Quidel, USA, and Cyclex Co, Japan. A Stericup ultrafiltration unit (0.45 μm) and an Amicon ultracentrifugation device (10 kDa) were from Millipore, USA. HiTrap Heparin HP column and Heparin Sepharose 6 Fast Flow resin was from GE Healthcare Life Sciences, USA. The 0.8 ml Pierce™ Micro Spin Columns were from Gibco® Life Technologies, USA. The chitin oligosaccharides tri-N-acetylchitotriose (GlcNAc_3, A₃); tetra-N-acetylchitotetraose (GlcNAc₄, A₄); penta-N-acetylchitopentaose (GlcNAc₅, A₅) and hexa-N-acetylchitohexaose (GlcNAc₆, A₆) were purchased from Isosep, Sweden. The ChOS blend T-ChOS™ was provided by Genis hf, Iceland. T-ChOS™ is a ChOS blend with a chain length of 4–20 and an average degree of acetylation of 55 % [46]. A human embryonic kidney cell line (HEK293) was from ECACC, UK. A human fetal lung fibroblast cell line (HFL-l) was from ATCC, USA. The human YKL-40 cDNA, TransPass D2 Transfection Reagent, plasmid vector pCMV6-AC-His (PS100002) containing a C-terminal 6x Histidine tag and a Neomycin resistance gene, and Rapid Shuttling kit, were from OriGene, USA.

2.2. Cloning and production of YKL-40

The human YKL-40 (NM_001276.2) cDNA sequence was cloned into a plasmid vector pCMV6-AC-His (C-terminal 6x histidine tag and neomycin resistance gene) using Rapid Shuttling kit (Sgf I and Mlu I restriction enzymes). Restriction enzyme digestion and agarose gel electrophoresis confirmed successful cloning of the cDNA into the plasmid vector. HEK293 cells were transfected with the YKL-40 plasmid using TransPass D2 transfection agent, following manufacturer's instructions. Following incubation of HEK293 with YKL-40 plasmid for 72 h, transfected cells were selected using 400 μg/ml G418 selection antibiotic supplemented into the complete growth medium (GM): Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10 % fetal bovine serum (FBS), 100 IU/ml penicillin and 50 IU/ml streptomycin (1 % P/S). G418 selection antibiotic was maintained until a stable cell line expressing His-tagged YKL-40 was established. YKL40-HEK293 cells were cultured and secreted His-tagged YKL-40 protein as adherent culture at 37 °C in a 5 % v/v CO₂ humidified incubator in GM.

2.3. YKL-40 purification

Recombinant human His-tagged YKL-40 protein was purified from the culture medium in a two-step purification process. First step was purification on a pre-packed 5 ml Protino Ni-NTA FPLC column (Macherey-Nagel) according to the manufacturer's instructions with the following changes: NPI-0 buffer was used as equilibration buffer, NPI-10 was used as washing buffer, and NPI-250 was used as elution buffer (single step elution from NPI-10 to NPI-250). All buffers and sample solution were set to pH 7.4 instead of recommended pH 8.0 due to the estimated pI 7.6 for YKL-40 [51]. A conditioned cell medium was thawed, filtered, and loaded on a Ni-NTA column using a fast protein liquid chromatography system (FPLC, Pharmacia). The eluted protein fractions were pooled, centrifuged, and dialyzed against heparin binding buffer (20 mM phosphate, 50 mM NaCl, pH 7.4) using Amicon-4 10 kDa filter unit. As a second step purification, the nickel-purified YKL-40 was purified on 5 ml heparin HiTrap column. Heparin binding buffer was used as equilibrating-, loading-, and washing buffer. Heparin elution buffer (20 mM phosphate, 1 M NaCl, pH 7.4) was used to elute the bound YKL-40 protein (single step elution from heparin binding buffer to heparin elution buffer). The heparin-eluted protein fractions were pooled, centrifuged, and dialyzed against storage buffer (50 mM phosphate, 300 mM NaCl, pH 7.4) using Amicon-4 10 kDa filter unit. The resulting YKL-40 concentrate was collected and the final yield was calculated using spectrometric determination of protein concentration at 280 nm (1 Abs = 1 mg/ml) and the purity determined by SDS-PAGE. To confirm the functional activity of the isolated recombinant YKL-40 a proliferation assay was performed using the human fetal lung fibroblast cell line HFL-l. Native YKL-40 (purchased from Quidel, USA) served as a control. The assay was carried out by seeding approximately 5000 cells per well in a 96-well microplate and allowing them to attach for 24h. The cells were subsequently serum-starved for 24 h and then treated for 72 h with varying concentrations of YKL-40. Cell viability was measured at the end of the treatment period using PrestoBlue (Invitrogen) and Crystal violet (Sigma Aldrich). Similar cell viability effects were achieved by using purified His-tagged YKL-40 and native YKL-40 purchased from Quidel.

2.4. Binding affinity measurements of YKL-40 to chitin oligosaccharides

The binding affinity of various chitin ligands to YKL-40 was determined by intrinsic fluorescence change upon ligand-binding using Microscale Thermophoresis (MST) NT LabelFree instrument. Due to ligand specific fluorescence change upon protein-ligand binding, only initial fluorescence (IF) results were used to determine the binding affinity rather than the MST traces. All measurements were carried out under temperature-controlled conditions at 25 °C using Monolith NT. LabelFree capillaries, 20 % excitation power, and 600 nM YKL-40 concentration. YKL-40 chitin ligands analyzed were chitin oligosaccharides A₃, A₄, A₅, and A₆. All YKL-40 and chitin ligand stock solutions were prepared in assay buffer: 50 mM phosphate buffer, 100 mM NaCl, pH 7.4, and spun for 20 min at 14,500 rpm before measurement. Sample preparation was prepared by following the MST software instructions provided by the pre-determined “Binding-Check” and “Binding-Affinity” assay setups. All samples were incubated for at least 20 min before measurement. The K_d values were estimated using the K_d fit analysis equation in the MO-affinity analysis software (NanoTemper). The chitin oligosaccharide concentration range used in each measurement was based on approximately 50-fold estimated K_d with 2-fold serial dilutions, a total of 16 samples in each experiment. Each sample was measured once in each experiment, with 3 repetitions. If aggregation and/or adsorption were detected by the instrument, those data points were excluded.

2.5. Heparin-binding assay

Heparin Sepharose-6 Fast Flow column resin 100 μl (HS, from GE Healthcare Life Sciences) was loaded into 0.8 ml Pierce™ Micro Spin Columns (Life Technologies) for the binding assay. To determine whether chitin oligosaccharides and ChOS affect the binding of YKL-40 to heparin, YKL-40 was incubated with various concentrations of chitin oligosaccharides at different chain lengths (degree of polymerization, DP) or ChOS before loading it on heparin-sepharose columns. YKL-40 without ligand served as a blank control. The loading- and washing buffer was 20 mM phosphate buffer, 154 mM NaCl, at pH 7.4, and the elution buffer was 20 mM phosphate, 1 M NaCl, at pH 7.4. All ligand dilutions and the blank control were prepared in the loading buffer and incubated with 50 μg YKL-40 in a total volume of 600 μl (2.0 μM) for 30 min at room temperature before loading on the spin column. YKL-40 was quantified in each of the eluates using the Bradford method at 620 nm using a microplate spectrophotometer (Labsystems Multiscan® MCC/340). Serially diluted purified YKL-40 served as a standard. Six micro spin columns were run simultaneously in each experiment, with one repetition of the experiment. In the repetition of the experiment, the concentration of ligand was slightly changed, resulting that each datapoint is a single value (not an average of 2). The heparin-bound YKL-40 was calculated as a percent of the total amount measured and used to plot and calculate the effect of each ligand. The amount of YKL-40 bound to heparin in the absence of the ligand (ligand-free protein) was set as 100 %. The chitin oligosaccharide concentration is represented in micromolarity (μM), however, since the ChOS blend is a mixture of chitooligosaccharides of variable lengths and deacetylation, data were represented as μg/ml solution (w/v).

2.6. In silico analysis

Crystal structure analysis was performed with UCSF Chimera molecular modeling system (version 1.16), developed by the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco (supported by NIH P41 RR-01081)[[52], [53], [54]]. To investigate YKL-40′s structure-function modulating effect of chitin oligosaccharides on the heparin-binding of YKL-40, we examined observed conformational changes in chitin- and heparin-binding sites of YKL-40 upon chitin-binding. This was done using the RCSB Protein Data Bank (PDB) structures 1NWR, 1NWS, 1NWU, 1NWT, 1HJX, and 1HJW, and UCSF Chimera software [2,53,55]. We analyzed superimposed crystal structures of native YKL-40 vs. YKL-40 complexed with chitin oligosaccharide ligands using the structure comparison tools “MatchMaker” and “RR distance maps”. For clarity, native YKL-40 is specified as YKL-40, and the chitin-bound YKL-40 complex as YKL-40/A_n, where n is the DP of the chitin oligosaccharide. In this comparison, we focused on conformational changes in Trp99 and within both proposed heparin-binding sites using RR distance maps of Cα-Cα [54].

2.7. Statistics, mathematics, and graphing

Sigmaplot 13 was used for statistical and graphical analysis. A 3rd parameter exponential decay fit f = y0+a∗exp (-b∗x) was applied for ligand-induced decrease in the heparin-binding of YKL-40. Statistical significance was estimated using the P-value for the ligand-induced exponential decay for each chitin oligosaccharide tested. IC₅₀ was calculated using the exponential decay fit.

3. Results

3.1. Chitin oligosaccharides bind to YKL-40 with micromolar affinity

Analysis of the initial fluorescent MST (IF-MST) data showed that chitin oligosaccharides A₄, A₅, and A₆ bound to YKL-40 with μM affinity, and that the binding affinity increased with increasing chitin oligosaccharide length (Table 1). However, K_d results for A₃ showed lower affinity than the detection limit of the MST instrument (>1 mM).

Table 1.

Chitin oligosaccharides cause decreased heparin-binding of YKL-40 in a size-dependent manner. YKL-40 affinity to chitin oligosaccharides increases with increasing chitin oligosaccharide length (K_d, μM). The results show the maximal inhibition (%) of heparin-binding observed, and the half maximal inhibitory concentration (IC₅₀) for each chitin oligosaccharides tested. The IC₅₀ values were calculated using the exponential decay fit from Fig. 1.

Chitin oligomer (DP)	Maximal inhibition (%)	Half maximal inhibitory concentration (IC50, μM)	Affinity to YKL-40, Kd ± SD (μM, N = 3)
A4	16.4	25.2	88.7 ± 16.1
A5	22.9	11.0	10.4 ± 0.4
A6	37.4	7.3	4.0 ± 0.8

3.2. Chitin oligosaccharides modulate the heparin-binding affinity of YKL-40

Chitin oligosaccharides A₄, A₅, and A₆ all significantly decreased the heparin affinity of YKL-40 in a dose-dependent manner (Fig. 1). All the chitin oligosaccharides show partial maximal inhibition of heparin binding of the respective chitin oligosaccharide up to 40 % decrease in binding affinity for A₆. The half-maximal concentration (IC₅₀) decreased with increased chitin chain length (Table 1). Statistical significance of the exponential decay fit increased with increased chain length of the chitin oligosaccharides tested (regression for A₄ p = 0.0421, A₅ p = 0.0012, and A₆ p < 0.001). These results are in line with an increasing affinity of longer chitin oligosaccharides to YKL-40.

Fig. 1.

Chitin oligosaccharides modulate the heparin-binding affinity of YKL-40. A) Heparin-binding assay was performed using 50 μg (2.0 μM) YKL-40 incubated with increasing concentrations of chitin oligosaccharides. The results show the dose-response curve for each chitin oligosaccharide tested (A₄ p = 0.0421, A₅, p = 0.0012, A₆, 6 < 0.001). Each experiment was run twice with variable concentrations. The following equation was used to find the best fit between data points: exponential decay; f = y0+a∗exp (-b∗x).

3.3. ChOS modulates the heparin-binding affinity of YKL-40

ChOS decreased the heparin-affinity of YKL-40 in a dose-dependent manner (Fig. 2A) in a comparable manner to chitin oligosaccharides with chain length of 4 or greater (DP ≥ 4). To investigate this effect further, YKL-40 was pre-bound to the heparin-sepharose column and ChOS tested as an eluent at increasing concentrations. The results show that pre-heparin-bound YKL-40 was partially eluted from the heparin-sepharose column by ChOS in a dose-dependent manner (Fig. 2B).

Fig. 2.

ChOS modulates the heparin affinity of YKL-40. Heparin-binding assay was performed using 50 μg (2.0 μM) YKL-40 with varying concentrations of ChOS, (A) Dose-response of pre-incubated YKL-40 with a variable amount of ChOS in the YKL-40 heparin-binding assay. (B) Dose-response results on how ChOS affects pre-heparin-bound YKL-40 in the heparin-binding assay. The following exponential decay equation was used to find the best fit to data points; f = y0+a∗exp (-b∗x).

3.4. Rotation of the Trp99 sidechain out of the chitin-binding site is important for decreased heparin affinity of YKL-40

We analyzed superimposed crystal structures of native YKL-40 vs. YKL-40 complexed with chitin oligosaccharide ligands (Fig. 3). In our analysis we found a clear difference at aa residue Trp99 when native YKL-40 or YKL-40/A₂ was compared to all YKL-40/≥A₄ complexes (Fig. 4B–E). No difference was found at aa residue Trp99 when comparing YKL-40 and YKL-40/A₂, or YKL-40/A₄ and YKL-40/A₅ (Fig. 4A and F). In the heparin-binding site at aa143-149 (H1), conformational changes were visible between YKL-40 and all YKL-40/≥A₄ complexes. By contrast, a negligible difference was observed comparing YKL-40 and YKL-40/A₂, or between YKL-40/A₄ and YKL-40/A₅ in this region of the protein (H1). Within the newly proposed heparin-binding site located in the KR-rich area at the C-terminus of YKL-40 (aa334–345, H2), no conformational differences were observed in any of our comparisons (Fig. 4A–F). Additionally, a difference was observed in loop regions at aa209-212 and aa228-232 upon chitin oligosaccharide binding. However, these conformational changes were of variable intensity and found in all comparisons, including a comparison between YKL-40/A₄ and YKL-40/A₅, independent of conformational changes in Trp99 or heparin-binding sites. Therefore, we concluded that these loop variations are not likely to contribute to the observed decrease in heparin affinity, but rather correlate to the length of the bound chitin oligosaccharide.

Fig. 3.

The Trp99 side-chain translocation in YKL-40 upon binding to chitin oligosaccharides. Analysis of YKL-40 structures performed using Chimera (version 1.16). A) Ribbon representation with transparent surface of the structures of native YKL-40 (1HJX, grey on top) and YKL-40 bound to chitin octamer (1HJW/A8, blue below), crystallized by Houston et al. [55]. The chitin octamer is colored yellow and the amino acid residue Trp99 is purple. The heparin-binding site GRRDKQH is colored red (H1), and the KR-rich heparin-binding site is colored green (H2). B) Comparison of the Trp99 side-chain location in superimposed YKL-40 structures without ligand (1NWR; brown), complexed with A2 (1NWS; blue), A4 (1NWU, green), and A5 (1NWT; purple), crystalized by Fusetti et al. [2]. Trp99 sidechains in native YKL-40 (brown) and YKL-40-A2 complex (blue) are located within the chitin-binding site while YKL-40 complexes with chitin oligosaccharides A4 (green) and A5 (purple) have rotated out of the chitin-binding site upon binding.

Fig. 4.

Comparison of superimposed YKL-40 crystal structures (1NWR native, 1NWS/A₂, 1NWU/A₄, and 1NWT/A₅) using RR distance maps. YKL-40 crystal structure analysis was performed using Chimera (version 1.16) and RR distance maps depicting the standard deviation (SD) of C_α-C_α distance among equivalent residue pairs [54]. Figures A–F highlight the SD difference in Trp99 amino acid residue (Trp99), heparin-binding site GRRDKQH (H1), two loops located near aa209-212 and aa228-232 (Loops), and the KR-rich heparin-binding site at C terminus (H2). A) Comparison of YKL-40 vs. YKL-40/A2 (1NWR vs. 1NWS). B) Comparison of YKL-40 vs. YKL-40/A4 (1NWR vs. 1NWU). C) Comparison of YKL-40 vs. YKL-40/A5 (1NWR vs. 1NWT). D) Comparison of YKL-40/A2 vs. YKL-40/A4 (1NWS vs. 1NWU). E) Comparison of YKL-40/A2 vs. YKL-40/A5 (1NWS vs. 1NWT). F) Comparison of YKL-40/A4 vs. YKL-40/A5 (1NWU vs. 1NWT). The details of the PDB structures being compared are on top of each map and the SD scale is to the right for each map.

4. Discussion

In this study, we aimed to investigate if ChOS affects the heparin-binding affinity of YKL-40. Our research uncovers a novel mechanism of YKL-40 where chitin oligosaccharides and ChOS act as allosteric effectors for its heparin-binding affinity. Specifically, these oligosaccharides modulate YKL-40′s heparin-binding affinity based on their size and concentration. Our research extends prior findings by demonstrating that chitin oligosaccharides larger than A₃ bind to YKL-40 with micromolar affinity. Furthermore, ChOS affects YKL-40′s heparin-binding affinity in a manner similar to chitin oligosaccharides of size A₄ or larger. Likewise, we show that pre-heparin-bound YKL-40 can be partially eluted by ChOS in a dose-dependent manner, indicating that ChOS can suppress the heparin activity of YKL-40 after it has been activated. These results show that chitin oligosaccharides and ChOS can modulate the binding affinity of YKL-40 to heparin, potentially influencing its downstream signaling in various biological processes, such as tissue remodeling, tumor angiogenesis, and immune modulation.

YKL-40 binds chitin/ChOS [2,46,55] and heparin/heparan sulfate [2,4,5] at distinct ligand-binding sites within the protein. To investigate the allosteric effects observed in our heparin-binding assay, we analyzed YKL-40′s conformational changes upon chitin oligosaccharide binding using available crystal structures. The exact location of the functional heparin-binding site has been debated, with suggestions pointing to a KR-rich domain at the C-terminus [4,38], while a heparin-binding motif (BXXBXB) at aa residues 143–149, GRRDKQH, has also been proposed [2,5]. Molecular dynamics simulations by Kognole et al. suggested that heparin binds to the heparin-binding site at aa143-149 in silico, while non-sulfated heparan sulfate does not bind to YKL-40, indicating that heparin binding is dependent on the degree of sulfation [5]. Fusetti et al. describe how a chitin tetramer (A₄) and longer oligosaccharides bind to the central part of YKL-40′s chitin-binding site (subsites −3 to +3), while shorter oligosaccharides preferably bind to the more distant subsites (−5 and −6) [2]. The extent of conformational changes induced by ligand binding to the chitin-binding site depends on the length of the bound oligomer, where Trp99 rotation is found to depend on binding to the central part of the chitin-binding site [2,55]. Based on the analogy between chitin binding to different regions of the chitin-binding site, and the varying heparin-binding affinity of YKL-40 depending on the chitin oligosaccharide length, we hypothesized that the rotation of Trp99 is the crucial factor connecting these two structure-function interactions. Our systematic analysis of superimposed crystal structures of native YKL-40 vs. YKL-40 complexed with chitin oligosaccharide ligands revealed a clear connection between Trp99 conformational changes and the heparin-binding site at aa143-149, but not with the newly proposed KR-rich heparin-binding site at the C-terminus. Our observed decrease in heparin affinity by chitin oligosaccharides is independent of the location of the heparin-binding site. Inferred from our structural analysis results and the molecular dynamic results by Kognole et al., we consider it more likely that the heparin-binding site is located at aa143-149. Notably, since we only observe partial inhibition of the heparin binding affinity, it is also plausible that there is more than one heparin-binding site, and only one of the binding sites is allosterically controlled by chitin oligosaccharides. Further research, including crystallization of heparin-bound YKL-40, is needed to firmly establish the identity of the heparin-binding site(s). Fusetti et al. attempted to crystallize a heparin-bound YKL-40 but were unsuccessful [2]. Alternative approach to identify the heparin-binding site(s) of YKL-40 could be molecular dockings, molecular dynamics simulation, and/or machine learning-based predictions techniques [[56], [57], [58], [59], [60]]. Computational simulation (in silico) can bring advantages such as lower cost and faster results compared to laboratory (in vitro) and clinical (in vivo) experiments [[61], [62], [63], [64]].

Identifying biomarkers and therapeutic targets is crucial for diagnosing diseases early and in the development of targeted treatments [65]. YKL-40 is overexpressed in variable pathological conditions, limiting its use as a standalone biomarker to distinctly identify a disease or condition. Instead, its value increases when combined with other markers or used to assess the treatment response over time, supporting the evaluation of the regimen's effectiveness [18,31]. Targeting of YKL-40 using knockout constructs of YKL-40, neutralizing antibodies, and small molecule inhibitors have shown promising results reducing inflammation, tumor angiogenesis, tumor progression, and enhanced cancer cell sensitivity to cancer treatments [9,22,25,[31], [32], [33], [34], [35], [36],66].

YKL-40 binds at least three heparin-carrying receptors; syndecan-1 [40], syndecan-4 [7], and CD44v3 [8]. YKL-40 binds to the heparan-sulfate chains of syndecan-1 in endothelial cells causing coordination of the membrane proteins syndecan-1 and integrin α_vβ₃, activating FAK-MAPK signaling pathway that, in turn, leads to increased tumor angiogenesis [40]. Blocking the heparin binding interaction between YKL-40 and the heparan-sulfate chains of syndecan-1 eliminated the observed elevation in angiogenesis [40]. Additionally, YKL-40 binding to heparan-sulfate chains of syndecan-4 stimulated the migration and tube formation of HUVECs, where downregulation of syndecan-4 abolished YKL-40-induced angiogenesis [7]. Geng et al. found that YKL-40 binds to the membrane protein CD44v3 and promotes metastasis and epithelial-to-mesenchymal transition (EMT) through β-catenin/Erk/Akt signaling in gastric cancer [8]. Their result shows that YKL-40 binds exclusively to CD44v3, which is the only heparan sulfate (HS)-modified isoform of the CD44 transmembrane glycoprotein, but not to CD44v6 or the standard isoform CD44. Interestingly, the expression of CD44v3, over CD44, is greatly increased in pancreatic tumor tissues [41]. This suggests that the heparin-binding feature of YKL-40 plays a significant role in cancer development, causing poor outcomes for cancer patients with elevated YKL-40 by interacting with several heparin-binding cell surface receptors. Therefore, targeting the heparin-binding affinity of YKL-40 could reduce cancer progression through various mechanisms, leading to broad therapeutic benefits.

In conclusion, our results show that the heparin affinity of YKL-40 can be suppressed by chitin oligosaccharides. For further research, our results provide valuable insights into a potential mode of action for YKL-40 within the surrounding tumor microenvironment. Since ChOS has many beneficial properties, such as being non-toxic and biodegradable, these results suggest that chitin and/or ChOS can be used to modulate YKL-40 activity, establishing YKL-40 as a promising therapeutic target.

CRediT authorship contribution statement

Unnur Magnusdottir: Writing – review & editing, Writing – original draft, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Finnbogi R. Thormodsson: Writing – review & editing, Supervision, Project administration, Methodology, Conceptualization. Lilja Kjalarsdottir: Writing – review & editing, Writing – original draft, Supervision, Methodology, Formal analysis, Conceptualization. Hordur Filippusson: Writing – review & editing, Supervision, Conceptualization. Johannes Gislason: Writing – review & editing, Supervision, Methodology, Conceptualization. Kristinn Ragnar Oskarsson: Writing – review & editing, Investigation. Jens G. Hjorleifsson: Writing – review & editing, Supervision. Jon M. Einarsson: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization.

Competing financial interest

This project was funded by Genis hf.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Unnur Magnusdottir reports financial support was provided by Genis hf. Jon M. Einarsson reports financial support was provided by Genis hf. Unnur Magnusdottir reports a relationship with Genis hf that includes: employment and equity or stocks. Jon M. Einarsson reports a relationship with Genis hf. That includes: equity or stocks. I, Unnur Magnusdottir, am a doctoral student at the University of Akureyri and this work is part of my Ph. D project, which Genis hf funds. Genis hf is a biotech company producing short-chain chitooligosaccharides, patented as T-ChOS™. Jon M. Einarsson is a co-supervisor of Unnur's Ph.D. project. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.