Rose hip and its constituent galactolipids confer cartilage protection by modulating cytokine, and chemokine expression

RHP, prepared from dried Rosa canina fruits of a selected cultivar, was obtained from Hyben Vital, Langeland, Denmark. GLGPG (galactolipid (2S)-1, 2-di-O-[(9Z, 12Z, 15Z)-octadeca-9, 12, 15-trienoyl]-3-O-β-D-galactopyranosyl glycerol, abbreviated as GOPO in [18]) was isolated from a lipophilic extract prepared from leaves of Valeriana locusta or from RHP (performed by AnalytiCon Discovery, Potsdam, Germany). Compounds were dissolved in DMSO and added to the culture medium concomitantly with the stimulating agent. Final DMSO concentration in culture medium in all treatments .was 0.5%. E. coli LPS (serotype 055:B5) and fetal bovine serum (FBS) were from Sigma (Saint-Louis, MO). RPMI 1640, DMEM, 2-mercaptoethanol and MEM non-essential amino acids (NEAA) were from Invitrogen (Carlsbad, CA). Human IL-1β and recombinant interferon-γ (IFN-γ) were from PeproTech EC (London, UK).

Murine RAW264.7 macrophage cells were from ATCC (Manassas, VA) and cultured in DMEM supplemented with 50 U/mL penicillin, 50 μg/mL streptomycin, 0.1 mM NEAA (DMEM-C) and 10% FBS. Cells were seeded into 12-well or 96-well plates at 1 and 0.05 × 10⁶ cells per well, respectively, and used after 2 days of pre-culture. Cells were starved for 18 h in DMEM-C containing 0.25% FBS before the start of treatment and stimulated with LPS (1 μg/mL) for 4-24 h in phenol red-free DMEM-C containing 0.25% FBS.

Peripheral blood leukocytes (PBL) were obtained from healthy male and female donors (Blood Donor Service, University Hospital, Basel, Switzerland). Human primary cell protocols were approved by the Swiss Federal Office of Public Health (No. A050573/2 to J. Schwager). Erythrocytes were removed by the Dextran sedimentation procedure [24]. Cell viability was determined by the Trypan Blue exclusion test and exceeded 95%. PBL (at 3-8 × 10⁶ cells/mL) were cultured in phenol-red free RPMI 1640, supplemented with 0.25% FBS, 0.1 mM NEAA, 50 U/mL penicillin, 50 μg/mL streptomycin and 5 × 10^-5 M 2-mercaptoethanol. Cells were stimulated with LPS (100 ng/mL) and IFN-γ (20 U/mL) for 2-24 h.

SW1353 chondrosarcoma cells were from ATCC and cultured in DMEM-C containing 10% FBS. Cells were seeded into 6-well plates at 0.5 × 10⁶ cells per well. Sub-confluent cell monolayers were washed and incubated overnight in DMEM-C containing 0.25% FBS and 0.2% lactalbumin hydrolysate (Bacto™ LC, Becton Dickinson, Franklin Lakes, NJ). Cells were activated with 10 ng/mL IL-1β in phenol-red free DMEM-C supplemented with 0.25% FBS and 0.2% lactalbumin hydrolysate in the presence of increasing concentrations of test compounds for 4-24 h. Batches of normal human articular chondrocytes from knee (NHAC-kn) obtained from different individuals were from Lonza and cultured in chondrocyte growth medium (Lonza, Wakersville, MD). For experiments NHAC-kn were used at passage 3 to 6. Cells were seeded into 6-well plates at 0.5 × 10⁶ cells per well and activated with IL-1β (10 ng/mL) for 4-24 h.

Cells were lysed in RLT buffer (Qiagen) after 2-4 h of culture and total RNA was extracted. Culture supernatants were harvested after 24 h of culture and stored at -80°C until use for analysis.

Total RNA was isolated using the RNeasy Mini Kits (Qiagen, Hilden, Germany) as previously described [25]. RNA quality and quantity was assessed by Nanodrop^® ND-1000 and evaluated by the ND-1000 3.2.1 software (Witec AG, Littau, Switzerland).

Total RNA was transcribed into first strand cDNA using the Superscript™ First-Strand Synthesis System for RT-PCR from Invitrogen (Carlsbad, CA) [25]. Real-time PCR analysis was performed using the ABI PRISM^® 7700 Sequence Detection System or the ABI 7900HT Fast Real-Time PCR System (Applied Biosystems [ABI], Foster City, CA). Primers and probes were designed with the Primer Express™ software purchased from ABI. PCR was performed using the Taqman^® universal PCR Master Mix (ABI). 18S rRNA primers and probes were used as internal standards. Relative gene expression quantification was performed by subtracting threshold cycles (C_T) for ribosomal RNA from the C_T of the targeted gene (ΔC_T). Relative mRNA levels were then calculated as 2^-ΔΔCT, where ΔΔC_T refers to the ΔC_T of unstimulated minus treated cells. The values were obtained from at least three independent series of experiments, in which each treatment was performed in duplicate with each being analyzed twice in RT-PCR.

Multiparametric kits were obtained from BIO-RAD Laboratories (Hercules, CA) and used in the LiquiChip Workstation IS 200 (Qiagen, Hilden, Germany). The data were evaluated with the LiquiChip Analyser software (Qiagen). In all experimental series, the proteins secreted into the culture supernatants were determined.

The concentration of NO in culture supernatants was measured using the Griess Reaction [26]. Secreted PGE₂ was determined by Enzyme Immuno Assay (EIA) (Cayman Chemicals, Ann Harbor, WI).

Macrophages represent a cellular model to identify effects of substances on inflammatory processes including those associated with arthritis. RAW264.7 cells responded to LPS-stimulation by producing inflammatory mediators (e.g. NO and PGE₂). In unstimulated cells, test substances alone did not modulate the secretion of inflammatory mediators. In LPS-treated cells, RHP reduced NO production and PGE₂ secretion at IC₅₀ values of 797 mg/L and 594 mg/L, respectively (Table 1). Similarly, GLGPG significantly inhibited the NO production (IC₅₀ values of 28.6 mg/L), whereas PGE₂ secretion was diminished by 14 ± 9% at 38.7 mg/L (i.e. highest concentration tested). The data are consistent with the described effects of rose hip constituents on the activity of COX isozymes [22, 23]. At the investigated concentrations, RHP or GLGPG did not impair cell viability as determined by LDH release (data not shown).

LPS-stimulation of RAW264.7 elicited the expression of inflammatory genes. Most (e.g. genes for TNF-α, COX-2, iNOS, IL-1α, CCL5/RANTES, CXCL10/IP-10) were drastically up-regulated (Table 2 and Additional File 1). We had determined in previous experiments that at 2, 4 and 8 h of stimulation, LPS-responding genes were significantly up-regulated (data not shown); at 4 h both early- and late-responding genes were significantly expressed. At this point in time, the iNOS mRNA levels were concentration-dependently reduced by GLGPG and RHP (Figure 1). COX-2 mRNA levels were not affected, but those of prostaglandin E synthase (PGES), which converts PGH₂ to PGE₂, were lowered (Table 2 and Additional File 1). Among murine cytokines and chemokines, CCL5/RANTES and CXCL10/IP-10 were robustly down-regulated by RHP and GLGPG. The expression of MMP-9 was also significantly reduced by RHP and less profoundly by GLGPG. Remarkably, expression of the anti-inflammatory IL-10 was increased by both substances. Macrophage genes encoding transcription factor (TF) of the NF-κB signaling pathway (i.e. NF-κB1, NF-κB49, NF-κBp65, I-κBα) were up-regulated by LPS (Table 2 and Additional file 1). RHP significantly reduced all TFs towards pre-stimulation values or below. GLGPG had less robust effects on those factors and markedly reduced only NF-κB49. Thus, the test compounds modulated gene expression at the transcriptional level via elements of the NF-κB pathway.

Orally absorbed bioactive compounds reach target tissues via the vascular system, where they might exert effects on peripheral blood leukocytes (PBL). Therefore, the effects of RHP and GLGPG on LPS/IFN-γ -activated PBL was investigated. The secretion of chemokines or interleukins was modulated by RHP and GLGPG in an idiosyncratic pattern (Table 3; Figure 2a-b and Additional file 2). RHP significantly reduced the production of CXCL10/IP-10 and CCL5/RANTES; the levels of other chemokines were unaltered or only affected at the highest tested RHP concentrations (Figure 2a). GLGPG reduced MCP-1 and MIP-1α secretion. It did, however, not match the effect of RHP on CXCL10/IP-10 and CCL5/RANTES production. RHP had robust effects on secretion levels of (pro-inflammatory) IL-12p70 and (anti-inflammatory) IL-10, while it modulated IL-1β and IL-6 less profoundly. This contrasted with the effects of GLGPG, which significantly diminished IL-1β and IL-6 (Figure 2b). TNF-α and IFN-γ were concentration-dependently diminished by both RHP and GLGPG (Figure 2b, Table 3). Furthermore, in PBL RHP increased the expression of GM-CSF, an anabolic factor for chondrocytes [27]. Next, expression levels of inflammatory mediators were determined in human PBL. In most cases, RHP and GLGPG exerted effects on mRNA levels that paralleled those described for the secreted proteins (Table 4, Figure 2c and Additional file 3). In particular, RHP robustly down-regulated mRNA levels of IL-1α, TNF- α and CXCL10/IP-10, but not those of e.g. IL-6, CCL5/RANTES, COX-2, MIP-2 or MIP-3α. Other genes that are involved in the production of eicosanoids (e.g. 5-lipoxygenase) or participate in ECM remodeling (e.g. MMP-9) were not modulated by the test compounds. The results infer that the main effects of the natural substances were at the transcriptional rather than the post-transcriptional level. A notable exception was CCL5/RANTES, where mRNA levels were barely up-regulated by LPS/IFN-γ stimulation, while the secretion of CCL5/RANTES was impaired by the natural substances.

The chondrosarcoma SW1353 cell line was used as a surrogate for primary chondrocytes. In response to exogenous IL-1β, these cells altered the expression of a similar set of genes as primary chondrocytes [28] (Tables 5 and 6). IL-1β activated SW1353 cells displayed increases in gene expression levels of MMP-1, -3, -13 and ADAMTS-4. In contrast, MMP-2 and ADAMTS-5 were barely affected. Treatment of IL-1β-activated SW1353 cells with RHP or GLGPG led to a significant inhibition of gene expression of MMP-1, -3, -13 but not of ADAMTS-4 or MMP-2 (Table 5). IL-1β treatment only slightly influenced expression levels of anabolic genes (i.e. aggrecan, collagen), whereas other mediators such as COX-2, TNF-α, iNOS, IL-6 and leukemia inhibitory factor (LIF) [14, 29] were markedly altered by IL-1β stimulation (see fold changes in Table 5). Basal chemokine gene expression was low in SW1353 cells; IL-1β treatment induced dramatic increases of chemokine mRNA levels, with CCL5/RANTES and MIP-3α being the most responsive. RHP significantly decreased the expression levels of five chemokine genes (MIP-2, MIP-3α, IL-8, CCL5/RANTES and CXCL10/IP-10) (Table 5) and cytokine genes including IL-1α, IL-1β and IL-6. Whereas GLGPG significantly reduced CXCL10/IP-10 expression, it augmented that of MIP-2 or IL-8. This feature was also observed with regard to anabolic genes, where GLGPG, in contrast to RHP, increased aggrecan and collagen expression. It should be stressed that the chosen in vitro conditions preferably trigger catabolic activity in chondrocytes, while anabolic events would have been favored at different culture conditions.

Normal human articular chondrocytes from knee (NHAC-kn) were activated by IL-1β in the presence of increasing concentrations of RHP or GLGPG. Basal expression levels of the monitored genes were comparable in SW1353 and NHAC-kn except for ADAMTS-5, collagen 2A1, four chemokine genes and IL-6, which were more expressed in NHAC-kn (Tables 5 and 6). The IL-1β induced changes in NHAC-kn gene expression levels were as marked as those observed in SW1353 cells. IL-1β significantly up-regulated catabolic genes (ADAMTS-4, MMP-1 and MMP-3), interleukins (IL-1α, IL-1β, IL-6), chemokines (IL-8, CCL5/RANTES, MIP-2 and MIP-3α) and COX-2 (Table 6). LIF experienced strong up-regulation by IL-1β, a feature that was not shared by the receptor for IL-1 (IL-1RI) and IL-1 receptor antagonist (IL-1Ra). The patho-physiological stimulus (i.e. IL-1β) weakly affected anabolic genes like collagen 1, collagen 2 and aggrecan (Table 6).

RHP significantly diminished the expression level of MMP-1, MMP-3, MMP-9 and MMP-13, but had only moderate effects on ADAMTS-4, TIMP-1 or ADAMTS-5 (Table 6). Similarly, MIP-2, MIP-3α, CCL5/RANTES and IL-8 were down-regulated by up to 75%. The effects on IL-1β, IL-1RI, and IL-6 were also statistically significant. In contrast, GLGPG weakly influenced mRNA levels of most of the tested genes (p value ~0.1). The effects of the substances were concentration-dependent (Figure 3). It should be noted that RHP had more potent effects on gene expression in NHAC-kn than in PBL (Figure 2). Anabolic genes were barely altered at 4 h of IL-1β stimulation.

Murine macrophages, human PBLs and chondrocytes revealed idiosyncratic patterns of reactivity to RHP and GLGPG as exemplified for IL-1α and CCL5/RANTES (Figure 4). Treatment of cells with LPS and IL-1β, respectively, elicited tissue-specific responses; IL-1α was induced in all three cell populations. Concomitantly, the effect of test compounds was robust with RHP being more potent than GLGPG. Conversely, CCL5/RANTES induction was strong in RAW264.7 cells and NHAC-kn but virtually absent in PBL. In general, RHP and GLGPG had effects on genes that were strongly activated by IL-1β or LPS; conversely, non-induced genes were virtually unaffected.