Recombinant human proteoglycan 4 lowers inflammation and atherosclerosis susceptibility in female low‐density lipoprotein receptor knockout mice

Recombinant human proteoglycan 4 (rhPRG4) is a macromolecular mucin‐like glycoprotein that is classically studied as a lubricant within eyes and joints. Given that endogenously produced PRG4 is present within atherosclerotic lesions and genetic PRG4 deficiency increases atherosclerosis susceptibility in mice, in the current study we investigated the anti‐atherogenic potential of chronic rhPRG4 treatment. Female low‐density lipoprotein receptor knockout mice were fed an atherogenic Western‐type diet for 6 weeks and injected three times per week intraperitoneally with 0.5 mg rhPRG4 or PBS as control. Treatment with rhPRG4 was associated with a small decrease in plasma‐free cholesterol levels, without a change in cholesteryl ester levels. A marked increase in the number of peritoneal foam cells was detected in response to the peritoneal rhPRG4 administration, which could be attributed to elevated peritoneal leukocyte MSR1 expression levels. However, rhPRG4‐treated mice exhibited significantly smaller aortic root lesions of 278 ± 21 × 103 μm2 compared with 339 ± 15 × 103 μm2 in the aortic root of control mice. The overall decreased atherosclerosis susceptibility coincided with a shift in the monocyte and macrophage polarization states towards the patrolling and anti‐inflammatory M2‐like phenotypes, respectively. Furthermore, rhPRG4 treatment significantly reduced macrophage gene expression levels as well as plasma protein levels of the pro‐inflammatory/pro‐atherogenic cytokine TNF‐alpha. In conclusion, we have shown that peritoneal administration and subsequent systemic exposure to rhPRG4 beneficially impacts the inflammatory state and reduces atherosclerosis susceptibility in mice. Our findings highlight that PRG4 is not only a lubricant but also acts as an anti‐inflammatory agent.


Introduction
Proteoglycan 4 (PRG4, also known as lubricin, megakaryocyte-stimulating factor and superficial zone protein) is a secreted mucin-like glycoprotein that is predominantly known for its physiological role in the lubrication of joints as it reduces friction and controls the proliferation of synovial cells in articular joints (Jay & Walker, 2014;Rhee et al., 2005).While originally discovered and studied in synovial joints, it should be acknowledged that PRG4 is ubiquitously expressed, with relatively high levels, in mice, being present in lymph nodes and metabolic tissues such as the liver, muscles and the heart (Nahon et al., 2019).Changes in PRG4 levels 0 Menno Hoekstra is an assistant professor at the Division of Systems Pharmacology and Pharmacy of the Leiden Academic Centre for Drug Research at Leiden University, Leiden, The Netherlands.He is primarily interested in uncovering the role of different lipid species in health and disease.His research mainly focuses on the complex relationship between plasma lipid levels, macrophage foam cell formation, and fatty liver and atherosclerotic lesion development.However, he also has an ongoing line on of research dedicated to understanding the contribution of lipoprotein-associated cholesterol to adrenal glucocorticoid synthesis and the associated protection against sepsis.
and functionality can therefore theoretically impact many different cell types and tissues.In support of a functional role for PRG4 in the control of metabolism, we have observed that total body PRG4 deficiency protects against glucose intolerance and fatty liver disease in high-fat diet-induced obesity in mice (Nahon et al., 2019).
Importantly, PRG4 has emerged as a putative novel therapeutic target in atherosclerotic cardiovascular disease.Relatively high levels of PRG4 have been found in macrophage-rich and calcified areas of both human and murine atherosclerotic lesions as well as in human aortic valves, where the protein appears to contribute to plaque stability by modulating the smooth muscle phenotype and enhancing the development of macro-calcified nodules (Artiach et al., 2020;Karlöf et al., 2019;Nahon, Hoekstra, Havik et al, 2018;Seime et al., 2021).Macrophage-specific PRG4 deficiency does not impact atherosclerosis susceptibility in hypercholesterolaemic low-density lipoprotein (LDL) receptor knockout mice in vivo (Nahon, Hoekstra, Havik et al, 2018).However, total body PRG4 deficiency in hypercholesterolaemic mice is consistently associated with an enhanced susceptibility for the development of atherosclerotic lesions (Nahon, Hoekstra, Van Eck, 2018).Based upon these combined findings it can be hypothesized that circulating PRG4 acts as an anti-atherogenic factor and that increasing plasma PRG4 levels may thus reduce atherosclerosis burden.
Full-length recombinant human PRG4 (rhPRG4) has previously been shown to effectively replicate the lubrication function of endogenous PRG4 in a variety of preclinical experimental settings (Abubacker et al., 2016;Flannery et al., 2009;Samsom et al., 2014) and is already applied clinically as local treatment for dry eye disease (Lambiase et al., 2017).To test the hypothesis that rhPRG4 can possibly also be used as treatment in the context of atherosclerotic cardiovascular disease, in the current study we determined the effect of chronic treatment with human rhPRG4 on disease outcome in an atherosclerosis mouse model.

Ethics approval
The experimental protocol was approved by the Animal Welfare Body of Leiden University under project licence AVD1060020185964 issued by the Central Authority for Scientific Procedures on Animals (CCD).The studies were executed according to the principles of laboratory animal care and regulations of Dutch law on animal welfare, Directive 2010/63/EU of the European Union, and the ARRIVE guidelines.All investigators understand the ethical principles under which The Journal of Physiology operates and our current study complies with the animal ethics checklist.

Mice
The two age-and weight-matched groups of 10-12-week-old female LDL receptor knockout mice used in this study were derived from an ongoing breeding programme at the Gorlaeus Laboratories of Leiden University that was initiated with breeder mice on a C57BL/6 background obtained from the Jackson Laboratories (Bar Harbor, MA, USA).From ∼12 weeks of age, mice were fed an atherogenic Western-type diet containing 15% cocoa butter, 1% corn oil and 0.25% cholesterol (Diet W, Special Diet Services, UK) and intraperitoneally injected three times per week with 200 μl of either a solution containing 0.5 mg full-length rhPRG4 (Lubris Pharma, Naples, FL;Greenwood-Van Meerveld et al., 2018;Krawetz et al., 2022;Samsom et al., 2014) or PBS as solvent control (N = 13 for both groups of mice).The mice were housed with 3-5 mice per cage in climate-controlled rooms of 21°C with a constant 12 h/12 h light/dark regimen at the Gorlaeus Laboratories and had free access to food and water at all times.Mice were weighed once a week throughout the experiment and monitored daily for their overall well-being.One mouse in the control group developed a severe skin condition and had to be withdrawn from the study.The remaining 12 control mice and 13 rhPRG4-treated mice were killed after 6 weeks of diet feeding and compound treatment.Mice were killed through a subcutaneous injection with a terminal mixture of ketamine (100 mg/kg), xylazine (12.5 mg/kg) and atropine (125 μg/kg) and exsanguination by orbital bleeding into EDTA-coated tubes.Peritoneal leukocytes were collected by lavage of the peritoneal cavity with 10 ml PBS.Following a whole-body perfusion with PBS, the heart with aortic root, spleen and liver of each mouse were collected and stored in Formal-Fixx 10% Neutral Buffered Formalin solution or snap-frozen and stored at −20°C.

Plasma analyses
Plasma was obtained from the orbital blood samples after 10 min centrifugation at 4000 g.The concentration of free cholesterol and cholesteryl esters in the plasma specimens was determined with enzymatic colorimetric assays as previously described by Out et al. (2006).The concentration of cytokines was analysed using ELISA MAX Deluxe Mouse TNF-alpha and interleukin-6 (IL-6) kits from Biolegend (San Diego, CA, USA).

Peritoneal leukocyte analysis
Peritoneal exudates were centrifuged at 1500 rpm for 10 min.Cell pellets were resuspended in 1 ml PBS and subjected to haematological analysis (Sysmex XT-2000iV Veterinary Haematology analyser; Sysmex Corporation) to measure the total leukocyte count and identify SSC high /FSC high lipid-laden foam cells, as previously described (Out et al., 2008).A part of the remaining cell suspension (250 μl) was cytospun onto tissue slides at 1500 rpm for 5 min.After fixation in formalin for 15 min, the slides containing the peritoneal leukocytes were stained with Oil red O (Sigma-Aldrich Corp., St. Louis, MO, USA) to visualize cellular neutral lipid deposits, and counterstained with haematoxylin.The leftover peritoneal cells were pelleted by centrifugation at 1500 rpm for 10 min and stored at −20 in a guanidinium thiocyanate solution for subsequent gene expression analysis.

Gene expression analysis
RNA was isolated using the standard guanidinium thiocyanate/chloroform/phenol extraction method (Chomczynski & Sacchi, 1987).RNA concentrations were determined with a Nanodrop Spectrometer (Nanodrop Technologies, Wilmington, DE, USA).Notably, only six randomly selected mice per group were included for gene expression analysis on their isolated spleens.In contrast, all liver specimens as well as all peritoneal leukocyte fractions that contained sufficient RNA material, i.e. from 11 control mice and 13 rhPRG4-treated mice, were included in the gene expression analysis.cDNA synthesis was performed from 1 μg of total RNA using RevertAid reverse transcriptase.Relative mRNA expression levels were determined with quantitative PCR using real-time SYBR Green technology (Eurogentec, Seraing, Belgium) essentially as described (Hoekstra et al., 2003).PPIA, 36B4 and RPL27 were used as housekeeping, reference genes for normalization.Genes with a cycle threshold value >35 were considered not reliably detectable/undetectable.

Histological analysis
Hearts fixed for 24 h in formalin were embedded in Tissue-Tek OCT compound to generate 10 μm thick cryosections of the aortic root using a Leica CM3050-S cryostat.Unfortunately, one heart of the rhPRG4 treatment group was lost during cutting.As such, sections from only 12 mice per group could be used for lesion size and composition analysis.Oil red O neutral lipid staining (Sigma-Aldrich Corp., St. Louis, MO, USA) was applied on the sections to identify atherosclerotic plaques.Sections were stained immunohistochemically for the presence of the macrophage marker CD68 with the antibody FA-11 (AB-32 421, Bio-Rad, Hercules, CA, USA) and vector BA-4001 from a MOMA-2 kit (Abcam, Cambridge, UK).Masson's Trichrome staining (Sigma-Aldrich Corp., St. Louis, MO, USA) was utilized to visualize collagen deposits with atherosclerotic lesions.Images used for lesion size and composition analysis were generated with a Leica DMRE microscope attached to a video camera and computer with Leica Qwin Imaging software (Leica Ltd., Cambridge, UK).Quantification of the different lesion parameters was performed with ImageJ software in a blinded manner (imagej.net).

Flow cytometry
A single-cell suspension of splenocytes in PBS was obtained from each mouse by straining about half of the freshly collected spleens through a 70 μm nylon mesh (Greiner Bio-One, Kremsmünster, Austria).Splenic leukocytes were stained with different cocktails of fluorescently labelled antibodies (Table 1) and subsequently analysed for the presence of specific markers by flow cytometry on a FACSCalibur II (Becton Dickinson, Mountain View, CA).Viable single cell populations were verified by plotting the signals for FSC-A versus those detected in the FSC-H channel and staining of the cells with a 7-AAD viability marker.Monocytes and macrophages were identified as being CD45 + /CD11c − /CD11b + /Ly6G − cells and discriminated into pro-inflammatory/infiltrating (Ly6C int/high ) and non-inflammatory patrolling/resident (Ly6C low ) subtypes.After excluding CD19 + B cells from the viable single cell populations, helper T cells (CD4 + ) and cytotoxic T cells (CD8 + ) were identified by co-staining for the adhesion molecules CD62L/L-selectin and CD44/H-CAM to detect naïve (CD62L + /CD44 − ), central memory (CD62L + /CD44 + ) and effector (CD62L − /CD44 + ) subclasses.A representative visualization of the gating strategy outcome is provided in Fig. 1.

Statistical analysis
Data were analysed with GraphPad Prism Software (GraphPad Software, La Jolla, California, USA).Outliers were detected with a Grubb's test.A two-tailed Student's t test with Welch's correction or two-way ANOVA with Bonferroni post hoc test was applied, where appropriate.P values <0.050 were considered significant.

Results
Genetically hyperlipidaemic female LDL receptor knockout mice were fed a Western-type diet for 6 weeks to stimulate the development of atherosclerotic lesions.To study the potential effect on atherosclerosis outcome, the atherogenic diet-fed mice were injected three times per week intraperitoneally with either a solution containing 0.5 mg rhPRG4 or the same volume of the solvent control PBS during the full course of the study.As can be appreciated from Fig. 2A, rhPRG4 treatment did not significantly affect body weight development.Control mice gained on average 1.2 g of weight during the 6 weeks of Western-type diet feeding, while rhPRG4-treated mice gained on average a total of 0.7 g (P = 0.478) (VanderLaan et al., 2009).Plasma lipid analysis at death revealed that rhPRG4 treatment was associated with a small improvement in the atherogenic index as a minor, but significant decrease in free cholesterol levels (−15%; P = 0.0152) was detected (Fig. 2A).Plasma cholesteryl ester levels did, however, not differ significantly between rhPRG4-treated and control mice (Fig. 2B).Peritoneal leukocytes were isolated during killing and subsequently subjected to routine haematological analysis to identify (macrophage) foam cells, i.e. SSC high /FSC high lipid-laden cells.Total peritoneal leukocyte counts were not significantly different between rhPRG4-treated mice (6.9 ± 0.4 × 10 6 ) and control mice (6.4 ± 0.7 × 10 6 ).In contrast, a 3.7-fold increase (P < 0.001) in the number of peritoneal foam cells was detected in response to chronic rhPRG4 exposure (Fig. 3A and 3B).As can be appreciated from Fig. 3C, Oil red O neutral lipid staining of cytospins confirmed the higher presence of lipid-laden leukocytes in peritoneal exudates from rhPRG4-treated mice as compared with those of PBS-injected controls.Quantitative real-time PCR analysis of the peritoneal leukocyte fractions revealed that rhPRG4 treatment was associated with significant 3.1-fold (P < 0.001) and 1.6-fold (P = 0.0419) increases in relative mRNA expression levels of the macrophage scavenger receptors M. Hoekstra and others J Physiol 602.9 MSR1 and CD36.Gene expression levels of the major cholesterol efflux mediator ATP-binding cassette transporter (ABCA1) were not significantly different between the two experimental groups of mice.It is thus anticipated that the increased foam cell extent can be primarily be attributed to an increased cellular lipid influx.In further support of this notion, a highly significant correlation was detected between the MSR1 gene expression levels and the foam cell extent of the individual peritoneal cell pools (R 2 = 0.556; P < 0.001).Notably, no significant correlation was found between CD36 gene expression levels and the foam cell extent (data not shown).
Cryosections from the aortic root were stained with Oil red O, MOMA-2 and Masson's Trichrome to identify atherosclerotic lesions and determine their macrophage and collagen content (Fig. 4A).Strikingly, despite the apparent foam cell enhancing action of rhPRG4, quantification of the Oil red O-positive lesion areas revealed that rhPRG4 treatment was associated with a mild, but significant protection against the development of atherosclerotic lesions (−18%; P = 0.0251).More specifically, rhPRG4-treated mice exhibited aortic root lesions of 278 × 10 3 μm 2 , while the average lesion size was 339 × 10 3 μm 2 in the aortic root of control mice (Fig. 4B).In accordance with the notion that ongoing infiltration of leukocytes into the diseased vessel wall -i.e.circulating monocytes that are locally converted into macrophagesis the driving force in the development of atherosclerotic lesions (Ross, 1999;Ye et al. 2011), the lower total plaque area was paralleled by a similar decrease in the aortic root MOMA-2 + macrophage content (Fig. 4C).Since the studies by Seime et al. (2021) have shown that rhPRG4 is able to inhibit fibrosis development, we also investigated a potential effect on lesional collagen content.As can be appreciated from Fig. 4D, the total plaque area covered by collagen was indeed decreased in rhPRG4-treated mice.However, this anti-fibrotic effect was nullified when correcting for the difference in plaque size, with average collagen content of 10.1% of total intimal lesion surface area in rhPRG4-treated mice and 9.9% in controls.
Previous studies have indicated that PRG4, through interacting with the cell surface receptors CD44 and toll-like receptor 2 (TLR2) and 4 (TLR4), can execute potent anti-inflammatory effects in vitro and in vivo (Alquraini et al. 2015(Alquraini et al. & 2017;;Al-Sharif et al., 2015;Qadri et al., 2018).We therefore evaluated whether the rhPRG4 treatment-associated atheroprotection in the context of an enhanced foam cell extent could be explained by a beneficial change in the systemic inflammatory state.Given that all immune cells contributing to the development of atherosclerotic plaques, such as monocyte/macrophages and T and B cells, are also present and interact with each other within secondary lymphoid tissues, we used the spleen as a proxy for identifying potential effects on inflammation.
Flow cytometric analysis on isolated splenocytes showed that rhPRG4 treatment did not greatly impact the overall immune cell distribution profile.Although average CD19 + B cell counts were somewhat elevated within splenocyte fractions of rhPGR4-treated mice as compared with control mice (38.5% versus 33.7%; P = 0.0433), spleens of the two groups of mice contained about equal numbers of CD4 + helper T cells, CD8 + cytotoxic T cells and CD11b + /CD11c − /Ly6G − monocytes/macrophages (data not shown).No treatment-associated change was detected in the global T cell activation status as judged from the similar distribution of both CD4 + helper and CD8 + cytotoxic T cells over their respective naïve (CD62L + /CD44 − ), central memory (CD62L + /CD44 + ) and effector (CD62L − /CD44 + ) subclasses (Fig. 5A).However, a significant shift within the splenic myeloid cell population was observed from the pro-inflammatory Ly6C int/high (infiltrating/migrating) subclass towards the non-inflammatory Ly6C low (more patrolling/resident) subtypes (Fig. 5B).As a result, the Ly6C int/high /Ly6C low cell ratio was decreased from 0.43 ± 0.2 in control mice to 0.36 ± 0.01 in rhPRG4-treated mice (P = 0.00618; Fig. 5C).
To further confirm an anti-inflammatory effect of our rhPRG4 treatment on splenic myeloid cells, we also investigated the impact on gene expression levels of macrophage-derived chemokines and cytokines (Fig. 5D).A 77% increase in splenic relative mRNA expression levels of the anti-inflammatory cytokine IL-10 was detected, which failed to reach significance due the large intra-group variance (P = 0.0707).
No change was found in expression levels of the pro-inflammatory/pro-atherogenic mediators IL-1beta, IL-18 and monocyte chemoattractant protein-1 (MCP-1), while levels of IL-6 were too low to be reliably detected.However, a significant decrease (−42%; P = 0.0352) in relative mRNA expression levels of the pro-inflammatory cytokine tumour necrosis factor-alpha (TNF-alpha) was detected in spleens from rhPRG4-treated mice (Fig. 5D).Importantly, a similar rhPRG4 treatment-associated decrease in TNF-alpha gene expression was measured within the peritoneal leukocyte fractions (−54%; P = 0.00442; Fig. 5E) as well as in livers of rhPRG4-treated mice (−55%; P < 0.001; Fig. 5F).Notably, the decrease in hepatic TNF-alpha gene expression levels could not be attributed to a major change in the macrophage content, as judged from the only minor difference in CD68 mRNA expression levels between livers of rhPRG4-treated and control mice (Fig. 5F).It thus appears that rhPRG4 treatment was able to induce a systemic decrease in the production of TNF-alpha by macrophages.In agreement, the TNF-alpha protein concentration was 36% lower in plasma samples of rhPRG4-treated mice than in plasma samples obtained from control mice (P = 0.00126; Fig. 5G).As can be appreciated from Fig. 5G, no change was detected between the two groups of mice in their plasma levels of the pro-inflammatory cytokine IL-6.
High TNF-alpha gene expression is an established feature of pro-inflammatory M1 macrophages.We therefore evaluated whether the selective decrease in TNF-alpha expression upon rhPRG4 treatment was reminiscent of a shift in macrophage phenotype.Hence, we quantified relative mRNA expression levels of additional M1 macrophage markers as well as classical markers of M2, anti-inflammatory macrophages using the cDNA generated from our peritoneal leukocyte fractions (Fig. 6A).Gene expression levels of the established M1 markers CD80 and CD86 were not significantly different between rhPRG4-treated mice and controls.However, we observed significant increases in the expression levels of M2 gene markers MRC1 (3.7-fold; P = 0.00243), FIZZ1/RETNLA (10-fold; P < 0.001), and in particular CHI3L3/YM1 (321-fold; P < 0.001).These combined findings suggest that rhPRG4 treatment indeed shifted the macrophage polarization from a M1-dominated pro-inflammatory phenotype towards a more M2-like inflammation resolving phenotype.It has previously been suggested by Wan et al. (2018) that, upon its local release by M2 macrophages, CHI3L3/YM1 promotes eosinophil infiltration into inflamed tissues to aid in the resolution of inflammation.In accordance, the major rise in peritoneal leukocyte CHI3L3/YM1 gene expression levels was paralleled by a significant 26-fold increase in peritoneal eosinophil counts (P < 0.001; Fig. 6B).The observation that peritoneal cell expression levels of all three M2 markers in our experimental setting correlated significantly with the peritoneal eosinophil content (Fig. 6C) underlines that the shift of macrophages towards the M2-like phenotype likely drives the accumulation of eosinophils within the peritoneal cavity in response to the rhPRG4 treatment.

Discussion
We have shown that systemic treatment with rhPRG4 is associated with a beneficial shift in the monocyte/macrophage activation state, reduced plasma TNF-alpha levels, and mildly reduced atherosclerosis susceptibility in the context of an enhanced foam cell extent in Western-type diet-fed LDL receptor knockout mice (Fig. 7).These findings provide further support for our working hypothesis that endogenous PRG4 is an anti-atherogenic factor and that increasing plasma PRG4 levels may constitute a novel therapeutic approach to treat patients at risk of developing atherosclerotic cardiovascular disease.
The increased susceptibility for the development of foam cells found after rhPRG4 treatment can be regarded as surprising when considering that we have previously observed that a genetic lack of PRG4 in macrophages is associated with enhanced susceptibility for the development of foam cells upon exposure to cholesterol-rich beta-VLDL in vitro (Nahon, Hoekstra, Havik et al., 2018).However, the marked increase in macrophage MSR1 gene expression levels and shift towards a more M2-like inflammation resolving (TNF-alpha low CHI3L3/MRC1/FIZZ1 high ) macrophage phenotype noted in our current in vivo setting do concur with the assumption that PRG4 executes its biological actions primarily by blocking TLR4 function (Alquraini et al., 2015).Onyishi et al. (2023) have found that genetic TLR4 deficiency in bone marrow-derived macrophages similarly increases MSR1 expression levels.Furthermore, Perera et al. (2001) and Mann et al. (2004) have shown that genetic lack of TLR4 is also associated with a reduced transcription and secretion of TNF-alpha in response to B. bronchiseptica, lipopolysaccharide or taxol exposure.Disruption of TLR4/MyD88 signalling in macrophages similarly reduces TNF-alpha gene expression levels and increases relative mRNA expression levels of CHI3L3/YM1 and the M2 marker CD206 (Mittal et al., 2016).Sánchez-Tarjuelo et al. (2020) have shown that a proper TLR4 functioning in mice is required to accommodate the S. pneumoniae infection-associated increase in the lung monocyte Ly6C high /Ly6C low ratio.In addition, treatment with the TLR4 inhibitor TAK-242 reduces the Ly6C high monocyte fraction in cardiac tissue from mice subjected to ischaemia-reperfusion (Fujiwara et al., 2019).As such, the anticipated rhPRG4-mediated blockade of TLR4's pro-inflammatory action probably also underlies the decrease in the myeloid cell Ly6C int/high /Ly6C low cell ratio observed in our experimental setting.Impairment of TLR4 functioning can also explain the rhPRG4 treatment-associated  decrease in the hypercholesterolaemia extent, since Ding et al. have previously shown that plasma cholesterol levels are also significantly reduced in LDL receptor knockout mice that genetically lack TLR4 (Ding et al., 2012).
Previous studies have suggested that the infiltration of the Ly6C high subclass of monocytes that express the chemokine receptor CCR2 is a key step in the early-stage atherosclerotic lesion development (Boring et al., 1997;Bot et al., 2017;Dawson et al., 1999;Guo et al., 2003Guo et al., , 2005;;Tacke et al. 2007;Ye et al., 2011) and thus probably more relevant in lesion initiation than the susceptibility of the monocyte-derived macrophages to subsequently become foam cells.Given that we have shown a clear beneficial impact of rhPRG4 on the monocyte Ly6C phenotype that is reminiscent of a less infiltrating subtype and see no reason to assume that the relative importance of monocyte infiltration to atherogenesis is different within the rhPRG4 treatment group of mice (as they also contain early-stage atherosclerotic lesions), it is therefore anticipated that rhPRG4 treatment was still able to reduce the atherosclerosis burden despite the marked increase in the peritoneal macrophage foam cell content as a result of a reduction in monocyte infiltration.However, it should be noted that the (systemic) reduction in TNF-alpha levels may also have contributed to the protection against atherosclerosis associated with rhPRG4 treatment, since Niemann-Jönsson et al. have shown that medial TNF-alpha expression precedes lesion formation in hyperlipidaemic apolipoprotein E × LDL receptor double knockout mice (Niemann-Jönsson et al., 2007).In addition, Branen et al. have found that total-body and bone marrow-specific TNF-alpha deficiency reduces atherosclerosis in mice (Brånén et al., 2004).
In conclusion, we have shown that chronic systemic exposure of Western-type diet-fed LDL receptor knockout mice to rhPRG4 is associated with a reduced susceptibility for the development of atherosclerotic lesions.Our findings highlight that intraperitoneal rhPRG4 treatment can execute effects both locally (i.e. on the peritoneal macrophage polarization status) and systemically (i.e. on plasma cytokine levels).Furthermore, the observation that rhPRG4 treatment reduces gene expression and plasma protein levels of the pro-inflammatory cytokine TNF-alpha in vivo substantiates our working hypothesis, derived from in vitro studies (Alquraini et al., 2017;Menon et al., 2021Menon et al., , 2022)), that PRG4 is not only a lubricant but also acts as an anti-inflammatory agent.Patients suffering from cardiovascular disease that enter the clinic generally contain highly inflamed (complex) atherosclerotic lesions.It will therefore be interesting to study in a preclinical setting whether rhPRG4 treatment is also able to inhibit the progression and/or induce regression of previously established atherosclerotic lesions.

Figure 1 .
Figure 1.Representative visualization of the flow cytometry gatingThe upper panel depicts the gating strategy for identifying the different CD4 + helper T cell and CD8 + cytotoxic T cell subtypes, while the gating sequence for identification of the pro-inflammatory (Ly6C int/high ) and non-inflammatory (Ly6C low ) monocyte subtypes is presented in the lower panel.[Colour figure can be viewed at wileyonlinelibrary.com]

Figure 2 .
Figure 2. Effect of rhPRG4 treatment on body weight and plasma lipid levels Body weight development (A) and plasma free cholesterol and cholesteryl ester levels (B) of Western-type diet-fed female LDL receptor knockout mice that were intraperitoneally injected 3×/week with 0.5 mg rhPRG4 (black dots) or PBS control (white dots) for 6 weeks.Values in graphs represent individual mice.Respective group averages are displayed as horizontal lines.

Figure 3 .
Figure 3.Effect of rhPRG4 treatment on peritioneal leukocyte cholesterol homeostasis B, quantification of the peritoneal foam cell counts, D peritoneal leukocyte relative mRNA expression levels of genes involved cholesterol uptake and efflux and E, the correlation between the peritoneal MSR1 gene expression levels and foam cell extent in Western-type diet-fed female LDL receptor knockout mice that were intraperitoneally injected 3×/week with 0.5 mg rhPRG4 (black dots) or PBS control (white dots) for 6 weeks.Values in the graphs represent individual mice.Respective group averages are displayed as horizontal lines.A, representative images of Sysmex plots showing the presence of SFL high /SSC high lipid-laden foam cells (FC).C, representative images of Oil red O-stained cytospins from peritoneal exudates showing exacerbated neutral lipid deposition in leukocytes from rhPRG4-treated mice.Scale bar = 500 μm.[Colour figure can be viewed at wileyonlinelibrary.com]

Figure 4 .
Figure 4. Effect of rhPRG4 treatment on atherosclerosis outcome Representative images of aortic root sections (A) and quantification of total atherosclerotic lesion area (B) and lesional macrophage (C) and collagen (D) content of Western-type diet-fed female LDL receptor knockout mice that were intraperitoneally injected 3×/week with 0.5 mg rhPRG4 (black dots) or PBS control (white dots) for 6 weeks.Values in graphs represent individual mice.Respective group averages are displayed as horizontal lines.Scale bar = 200 μm.[Colour figure can be viewed at wileyonlinelibrary.com]

Figure 5 .
Figure 5.Effect of rhPRG4 treatment on the systemic immune profileThe distribution of splenic helper CD4 + and cytotoxic CD8 + T cells (A) and monocytes (B) over their different subpopulations, splenic monocyte subpopulation ratios (C), splenic (D), peritoneal (E) and hepatic (F) relative mRNA expression levels of cytokines and inflammatory cell markers, and plasma cytokine protein levels (G) in Western-type diet-fed female LDL receptor knockout mice that were intraperitoneally injected 3×/week with 0.5 mg rhPRG4 (black dots) or PBS control (white dots) for 6 weeks.Values in graphs represent individual mice.Respective group averages are displayed as horizontal lines.

Figure 6 .
Figure 6.Effect of rhPRG4 treatment on macrophage polarizationA, relative mRNA expression levels established M1 and M2 macrophage gene markers in peritoneal leukocyte fractions and B, peritoneal eosinophil counts of Western-type diet-fed female LDL receptor knockout mice that were intraperitoneally injected 3×/week with 0.5 mg rhPRG4 (black dots) or PBS control (white dots) for 6 weeks.Values in graphs represent individual mice.Respective group averages are displayed as horizontal lines.C, the linear relationships between peritoneal leukocyte fraction relative mRNA expression levels of M2 macrophage markers and the peritoneal eosinophils counts.

Figure 7 .
Figure 7. Schematic overview of the key findings of our current studyChronic treatment with rhPRG4 is associated with a reduction in plasma cholesterol levels and the frequency of Ly6C int/hi pro-inflammatory lesion-infiltrating monocytes.In addition, MSR-1-mediated uptake of lipids by macrophages and the subsequent formation of foam cells is stimulated.However, rPRG4 treatment also beneficially impacts the macrophage immunophenotype, resulting in an overall decreased atherosclerosis susceptibility.The figure was generated using a premium BioRender account of the journal.[Colour figure can be viewed at wileyonlinelibrary.com]