Human serum-derived hydroxy long-chain fatty acids exhibit anti-inflammatory and anti-proliferative activity

Abstract

Background: Circulating levels of novel long-chain hydroxy fatty acids (called GTAs) were recently discovered in the serum of healthy subjects which were shown to be reduced in subjects with colorectal cancer (CRC), independent of tumor burden or disease stage. The levels of GTAs were subsequently observed to exhibit an inverse association with age in the general population. The current work investigates the biological activity of these fatty acids by evaluating the effects of enriched human serum extracts on cell growth and inflammation.

Methods: GTAs were extracted from commercially available bulk human serum and then chromatographically separated into enriched (GTA-positive) and depleted (GTA-negative) fractions. SW620, MCF7 and LPS stimulated RAW264.7 cells were treated with various concentrations of the GTA-positive and GTA-negative extracts, and the effects on cell growth and inflammation determined.

Results: Enriched fractions resulted in poly-ADP ribose polymerase (PARP) cleavage, suppression of NFκB, induction of IκBα, and reduction in NOS2 mRNA transcript levels. In RAW264.7 mouse macrophage cells, incubation with enriched fractions prior to treatment with LPS blocked the induction of several pro-inflammatory markers including nitric oxide, TNFα, IL-1β, NOS2 and COX2.

Conclusions: Our results show that human serum extracts enriched with endogenous long-chain hydroxy fatty acids possess anti-inflammatory and anti-proliferative activity. These findings support a hypothesis that the reduction of these metabolites with age may result in a compromised ability to defend against uncontrolled cell growth and inflammation, and could therefore represent a significant risk for the development of CRC.

Figure 1

Total ion chromatogram of crude serum organic extract. (A) Total ion current of bulk serum following liquid/liquid extraction and HPLC-coupled mass spectrometry as explained in the methods. (B) Extracted mass spectra of all masses from (A). (C) Extracted ion chromatograms of GTAs 446, 448 and 450 from the total ion current shown in A. (D) Cell proliferation, as assayed by MTT, for SW620 cells treated with up to 80 ug/ml of the crude serum extract.

Figure 2

Mass spectrometry characterization of semi-purified GTA-ve and GTA+ve extracts. (A) Crude serum extract (as shown in Figure 1) was subject to flash column chromatography as described in the methods resulting in two adjacent eluates, one positive and one negative for the presence of GTAs. The total ion chromatogram (top), extracted mass spectra (middle), and extracted ion chromatograms for three GTAs (GTA446, 448 and 450; bottom) of the GTA-ve fraction. (B) Same as (A) for the GTA+ve fraction. (C) For comparison, the extracted ion chromatograms of GTA446, 448 and 450 from the extracts of serum pooled from 20 CRC patients and 20 controls is shown.

Figure 3

Proliferation of SW620 cells treated with GTA+ve and GTA-ve extracts. (A) SW620 cells were incubated with increasing concentrations of GTA+ve and GTA-ve extracts for 24 hours and proliferation assayed by MTT. (B) The 80 ug/ml concentration of GTA+ve and GTA-ve extracts was then used to treat cells for up to 48 hours and the effect on cell proliferation assayed by MTT. Data are expressed as percent of vehicle or 0 hrs ± 1S.D. (C) Representative Western blot analysis of caspase-mediated PARP cleavage fragments resulting from treatment with GTA+ve and -ve extracts. Experiments were repeated at least three times.

Figure 4

Treatment of MCF7 cells with GTA+ve and GTA-ve extracts. MCF7 cells were incubated with vehicle (A), 80 ug/ml GTA+ve extract (B), and 80 ug/ml GTA-ve extract (C) and cells photographed using phase-contrast light microscopy (200×). (D) Western analysis of PARP cleavage products; ns, non-specific.

Figure 5

Western analysis of NFκB, IκBα and NOS2 in SW620 cells treated with three concentrations of GTA+ve and GTA-ve extracts and doxorubicin (DOX). Representative Western blots showing protein levels of NFκB, IκBα and NOS2 in SW620 cells treated with GTA+ve and GTA-ve extracts (see methods).

Figure 6

Determination of nitric oxide status in RAW264.7 cells treated with GTA+ve and GTA-ve extracts. RAW264.7 cells were pre-treated for 4 hours with GTA+ve or GTA-ve extracts followed by the addition of LPS (1 ug/ml) for 20 hours. (A) Nitric oxide levels in cells were determined using Griess reagent, (B) NOS2 mRNA transcript levels were determined by real-time rtPCR, and (C) NOS protein (treatment with 80 ug/ml) assessed by Western blot (NS, non-specific). Asterisks indicate p < 0.05 relative to LPS treatment alone, and FA mix in (C) represents a 100 uM equal mixture of 18:1, 18:2 and 18:3 fatty acids. Data are expressed as the average of three duplicate experiments ± 1S.D.

Figure 7

TNFα response in RAW264.7 cells treated with GTA+ve and GTA-ve extracts. RAW264.7 cells were pre-treated for 4 hours with GTA+ve or GTA-ve extracts followed by the addition of LPS (1 ug/ml) for 20 hours. (A) TNFα mRNA transcripts as determined by real-time rtPCR, (B) TNFα relative protein levels in cell lysates following 80 ug/ml treatment, and (C) TNFα protein levels in conditioned media as determined by ELISA. Asterisks indicate p < 0.05 relative to LPS treatment alone. Data are expressed as the average of three duplicate experiments ± 1S.D.

Figure 8

COX2 and IL-1β response in RAW264.7 cells treated with GTA+ve and GTA-ve extracts. RAW264.7 cells were pre-treated for 4 hours with GTA+ve or GTA-ve extracts followed by the addition of LPS (1 ug/ml) for 20 hours. (A) COX2 and (B) IL-1β mRNA levels were determined by real-time rtPCR. (C) IL-1β levels following 80 ug/ml treatment in cell lysates as determined by ELISA. Asterisks indicate p < 0.05 relative to LPS treatment alone. Data are expressed as the average of three duplicate experiments ± 1S.D.