|
|
||||||||


* Department of Veterinary Basic Sciences, The Royal Veterinary College, London NW1 0TU, England; and
GlaxoSmithKline, Medicines Research Centre, Stevenage, Herts SG1 2NY, England
1To whom requests for reprints should be addressed at Department of Veterinary Basic Sciences, The Royal Veterinary College, Royal College St., London NW1 0TU, England. E-mail: kbotham{at}rvc.ac.uk
| Abstract |
|---|
|
|
|---|
Key Words: chylomicron remnants low-density lipoprotein cholesterol triacylglycerol oxidized lipoproteins human macrophages
| Introduction |
|---|
|
|
|---|
Dietary lipid is transported via chylomicrons, which are synthesized in the intestine and enter the blood via the thoracic duct (6). These particles are rapidly catabolized to smaller triacylglycerol (TG) and cholesterol-rich remnants, which are then taken up by the liver. There is now a large amount of evidence to indicate that these remnant particles are atherogenic. They have been shown to be taken up by the aorta (79) as efficiently as LDL (10), and remnant-like particles containing apolipoprotein E (apoE) have been isolated from human aortic intima and atherosclerotic plaque (11, 12). In addition, delayed clearance of remnants from the circulation is correlated with the development of atherosclerotic lesions (13, 14).
We have shown previously that chylomicron remnants induce extensive lipid accumulation in the murine cell line, J774 (4), and a similar effect has been reported in experiments with human monocytederived macrophages (HMDMs; Ref. 5). In both these studies, nonoxidized lipoprotein particles were used, highlighting an important difference between foam cell induction by chylomicron remnants and oxLDL. Another difference may be in the type of lipid accumulated, since there is evidence to suggest that other TG-rich lipoproteins such as very low density lipoprotein (VLDL) or ß-VLDL cause the accumulation of TG rather than cholesterol (15, 16). Thus, at least some of the mechanisms involved in macrophage lipid accumulation by chylomicron remnants and LDL appear to be different. Most previous studies, however, have focused on the role of oxLDL, and although the induction of macrophage lipid accumulation by chylomicron remnants has been clearly demonstrated (4, 5), little further information about the process is available. Moreover, the effects of oxLDL and chylomicron remnants have not been directly compared. Since oxidation of LDL is known to take place within the artery wall (3), it is likely that chylomicron remnants that penetrate the subendothelial space will also become oxidized. Despite this, almost nothing is known about how the oxidation of chylomicron remnants influences their induction of macrophage foam cell formation.
The aims of the present study are to characterize the effects of chylomicron remnants on lipid accumulation in macrophages of human origin in terms of the amount and type of lipid accumulated and changes in lipid synthesis in the cells, to investigate how these parameters are affected by prior oxidation of the remnant particles, and to compare the effects of chylomicron remnants and oxidized chylomicron remnants directly with those of LDL and oxLDL. The LDL was obtained from human blood, but it is difficult to obtain from this source chylomicron remnants uncontaminated with lipoproteins of similar density, such as chylomicrons and VLDL. For this reason, artificial chylomicron remnant-like particles (CMR-LPs) containing human apoE with a size, density, and lipid composition similar to that of physiological remnants were used (17). Accumulation of lipid in HMDMs and macrophages derived from the human monocyte cell line THP-1 was assessed by oil red O staining and by quantitative assay of cholesterol, cholesteryl ester (CE), and TG. In addition, the effects of the four types of lipoprotein on cholesterol esterification and TG synthesis were assessed in THP-1 macrophages by measuring the incorporation of [3H]-oleate into CE and TG.
| Materials and Methods |
|---|
|
|
|---|
Preparation and Oxidation of Lipoproteins.
The LDL was prepared from fresh human serum obtained from blood donated by healthy volunteers or from the National Blood Service, North London Centre, London, England. Serum was layered under 0.9% NaCl (d = 1.006 g/ml) and centrifuged for 5 hrs at 65,000 g (4oC), and the top fraction was discarded. The density of the bottom layer was raised to 1.063 g/ml with potassium bromide (KBr), layered under KBr (d = 1.063 g/ml), and centrifuged for 24 hrs at 23,000 g (4oC), and LDL was collected from the top fraction by tube slicing. The LDL was dialyzed against phosphate-buffered saline (PBS) (72 hrs, 15 l, pH 7.4) at 4°C before use.
The CMR-LPs were prepared by sonication of a lipid mixture followed by density gradient centrifugation and binding to human apoE as described (18). Briefly, a mixture containing 70% trilinolein, 2% cholesterol, 5% CE, and 25% phospholipids in 0.9% NaCl in Tricine buffer (20 mM, pH 7.4) was sonicated at 2224 µm for 20 mins at 56oC. The resulting emulsion was brought to a density of 1.21 g/ ml with KBr, layered under a stepwise density gradient (19), and centrifuged at 17,000 g for 20 mins at 20oC. The upper layer of grossly emulsified lipids was then removed and replaced with an equal volume of 0.9% NaCl (d = 1.006 g/ ml) and centrifuged at 70,000 g for 1 hr (20oC). The CMR-LPs were harvested from the top layer. These particles were incubated with the dialyzed (18 hrs, 4oC) d > 1.006 g/ml fraction of human plasma in a ratio of 1:2.5 and incubated at 37oC with shaking for 18 hrs. The CMR-LPs were then reisolated by ultracentrifugation (45,000 g, 18 hrs, 12oC), harvested from the top layer, and stored at 4oC under nitrogen until required. For the preparation of DiI-labeled CMR-LPs, DiI (1 mg/25 mg of total lipid) was added to the lipid mixture before sonication.
The LDL and CMR-LPs were oxidized by incubation with CuSO4 (10 µM) with shaking for 1823 hrs at 37°C. The CuSO4 was subsequently removed by dialysis against 0.9% NaCl for 16 hrs at 4°C or, alternatively, by the addition of the chelating resin, Chelex 100. The extent of oxidation was determined by measuring the level of thiobarbituric acidreactive substances (TBARSs) in the preparations before dialysis. All lipoproteins were filtered (0.22 µm) before addition to the cell culture and used within 48 hrs of preparation.
Culture of HMDMs and THP-1 Cells and Experimental Protocols.
Mononuclear cells were isolated from fresh human blood collected from healthy volunteers in the presence of EDTA (4 mM). Histopaque-1077 (15 ml) was added to the upper chamber of Accuspin tubes (Sigma Chemical) and centrifuged into the lower chamber (800 g, 5 mins). Blood samples (30 ml) were then pipetted into the upper chamber, the tubes were centrifuged (2000 g, 20 mins), and the plasma layer was aspirated. The mononuclear cells in the opaque interface were washed with PBS (3 x 30 ml) and finally resuspended in RPMI-1640 culture medium supplemented with 5% human serum and 1% glutamine. The cells were then incubated in 24-well plates in culture medium at 37°C in 5% CO2 and 95% air at a density of 0.5 x 106 cells/well for up to 8 days to enable differentiation to the macrophage phenotype. The purity of the monocytic fraction was estimated by differential counts of DiffQuik (Porvair Sciences Ltd., Shepperton UK) stained cell preparations and was 90% or higher. The THP-1 monocytes were maintained in suspension in RPMI-1640 culture medium containing 10% FBS, 2 mM glutamine, 100 U/ml of penicillin, 100 mg/ml of streptomycin, and 50 mM ß-mercaptoethanol at a density of 39 x 105 cells/ml at 37oC in 5% air and 95% CO2. The cells were induced to differentiate into macrophages by incubation in 24-well plates at a density of 0.51 x 106cells/well for 72 hrs in the presence of PMA (200 ng/ml). After this time, the cells adhered to the culture plates, and the medium containing the PMA and any remaining monocytes were removed. Viability of HMDMs and THP-1 macrophages as assessed by Trypan blue exclusion was more than 95%.
For uptake studies, DiI-labeled CMR-LPs (30 µg/ml of cholesterol) were incubated with fully differentiated macrophages for 4 or 18 hrs, and the cells were then viewed with Zeiss LMS 510 laser scanning confocal microscope. For studies on lipid accumulation, human LDL, oxLDL, CMR-LPs, or oxCMR-LPs (30 µg/ml of cholesterol) were added to the macrophages, and the incubation was continued for a further 48 hrs. For determination of the incorporation of [3H]oleate into CE and TG, THP-1 macrophages were incubated in 24-well plates at a density of 1 x 106 cells/well and exposed to lipoproteins for the times indicated in the text. The medium was then replaced with culture medium (1 ml) containing [9,10-3H] oleic acid (37 KBq/ml, 55,000 dpm/µM) and 2% fat-free bovine serum albumin in RPMI-1640, and the incubation was continued for 1 hr. The cells were then washed with PBS (2 x 1 ml), and lipids were extracted with 5 ml of chloroform-methanol (2:1 v/v) in the presence of [1-14C] cholesteryl oleate (2500 dpm) as a recovery marker and separated by thin-layer chromatography (hexanediethyl etherformic acid, 80:20:2 v/v/v). Radioactivity in the bands corresponding to CE and TG was then determined by liquid scintillation counting. Preliminary experiments showed that [3H]oleate incorporation into CE and TG was linear for a period of 2 hrs in all the conditions used.
Analytical Methods.
The diameter of the CMR-LPs was measured after electron microscopy (20). Images were captured using a Gatan MSC 791 TEM CCD camera, and the mean diameter of CMR-LPs and oxCMR-LPs was determined using a randomly chosen sample of 60 particles taken from four different fields. The total cholesterol, TG, and phospholipid content of the lipoproteins was determined using commercially available kits (Boehringer Mannheim, Mannheim, Germany), and cholesterol, CE, and TG in macrophages were assayed by high-performance thin-layer chromatography (21) using hexaneheptanediethyletheracetic acid (63:18.5:18.5:1, v/v/v/v) as the mobile phase. Lipoprotein oxidation was determined by measuring the TBARS content of the preparations as described by Steinbrecher et al. (22). Tetraethoxypropane, which yields malondialdehyde, was used as a standard. The apolipoprotein composition of CMR-LPs was evaluated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (23).
The toxicity of LDL, oxLDL, CMR-LPs, and oxCMR-LPs in HMDMs and THP-1 macrophages was assessed using a (4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT)-based toxicology assay kit (Sigma Chemical) according to the manufacturers instructions. Lipid accumulation in THP-1 monocytederived macrophages and HMDMs was assessed by oil red O staining. The cells were washed twice with PBS and once with 60% propan-2-ol, and oil red O (0.5%, w/v) in 40% propan-2-ol/H2O (v/v) was added. After 15 mins, the stain was removed and cells were washed twice with PBS. Images of the cells were captured using a Canon microscopically mounted camera and the extent of staining analyzed using Media Cybernetics LP Optimas software, version 6.5. Proteins were determined using the method described by Lowry et al. (24).
Significance limits were calculated using analysis of variance. Where necessary, the P value taken to indicate statistical significance was adjusted using Dunnets post hoc test to simultaneously test the control with each treatment group and to take account of the multiple comparisons made.
| Results |
|---|
|
|
|---|
|
|
|
|
|
|
CI976 abolished the increase in total lipid content of HMDMs caused by exposure to oxLDL, CMR-LPs, or oxCMR-LPs, and this was reflected in the levels of cholesterol, CE, and TG, which were not significantly different from those in cells incubated without lipoproteins (Tables 4
and 5
). The increases in total lipid in THP-1 macrophages caused by LDL and oxLDL were also prevented by the inhibitor, with the levels of cholesterol, CE, and TG all reduced. Compared with cells incubated in the absence of lipoproteins, however, the amount of TG remained 70%85% higher, whereas the content of CE was reduced by 43%. Inclusion of CI976 in the medium of THP-1 cells incubated with CMR-LPs led to a decrease in total lipid accumulation (Table 4
), but the levels remained higher than those found in cells not exposed to lipoproteins. Further analysis showed that this was due to raised TG levels, since cholesterol and CE values were similar to those observed in control cells, but TG concentrations remained 4-to 5-fold higher (Table 5
).
The Effect of LDL and CMR-LPs on the Incorporation of [3H]oleate Into CE and TG in THP-1 Cells.
The rate of cholesterol esterification and TG synthesis in THP-1 macrophages exposed to LDL, oxLDL, CMR-LPs, or oxCMR-LPs was measured by the incorporation of [3H]oleate into CE and TG, respectively. After 48-hr incubation, the amount of label incorporated into CE was significantly increased in cells incubated with oxLDL but not LDL, and an increase was also observed after incubation with CMR-LPs and oxCMR-LPs (Table 6
). Incorporation of the label into TG was increased to a similar extent by LDL compared with oxLDL and by CMR-LPs compared with oxCMR-LPs, but the rate found with the CMR-LPs after 48-hr incubation was approximately 2-fold greater than that observed with LDL (Table 6
). Treatment of the cells with the ACAT inhibitor CI976 (10 µM) led to a decrease of approximately 85% in incorporation of [3H]oleate into CE but had no significant effect on the amount of radioactivity found in TG (Table 6
).
|
|
| Discussion |
|---|
|
|
|---|
Lipid accumulation in response to LDL or CMR-LPs was evaluated in HMDMs and in the human monocyte cell line THP-1, which can be induced to differentiate into macrophages by treatment with phorbol ester (35). In agreement with the previously reported effects of LDL and oxLDL in macrophages (3638), native LDL did not cause any change in lipid accumulation in HMDMs or THP-1 cells as assessed by quantitative evaluation of histological staining with oil red O (Table 3
), whereas oxLDL caused a marked increase. In contrast, when the macrophages were incubated with either CMR-LPs or oxCMR-LPs, extensive lipid accumulation was observed (Table 3
). Since it has been calculated that chylomicron remnants carry approximately 40 times more cholesterol per particle than LDL (39) and the cholesterol content of the LDL used in our experiments was approximately 2-fold that of the CMR-LPs (Table 1
), it is clear that considerably fewer CMR-LP than oxLDL particles are required to produce an equivalent level of lipid accumulation. Yu and Mamo (5) have demonstrated previously that chylomicron remnants induce lipid accumulation in HMDMs, and we have found a similar effect in J774 macrophages (4) and after short-term incubation (8 hrs) of this murine cell line with both native and oxidized chylomicron remnants (40). This is the first report, however, of the effects of oxidized remnants on lipid accumulation in human macrophages.
The type of lipid accumulated in human macrophages exposed to chylomicron remnants has not been studied in earlier work, although we have found that levels of both TG and total cholesterol are raised in J774 cells after short-term (8-hr) incubation with this type of particle (40). In addition, substantially increased TG levels have been reported after exposure of HMDMs to VLDL (16), and ß-VLDL has been shown to cause increases in both the CE and TG components in rat peritoneal macrophages (41). Liposomes containing phospholipid and cholesterol have also been found to lead to the massive accumulation of lipid droplets containing CE and TG in mouse macrophages (42). On the other hand, oxLDL has been shown to load macrophages with CE (38), but previous studies have paid little attention to any possible effects of oxLDL on TG accumulation. Recent evidence from experiments on the effects of TG depletion on the physical state of CE and measurements of TG mass in THP-1 macrophages and HMDMs exposed to acetylated LDL, however, indicates that accumulation of TG and CE occurs in these cases (43, 44). Indeed, in the study of Kritharides et al. (43), the levels of TG found in HMDMs and THP-1 cells exceeded those of total cholesterol both before and after their exposure to acetylated LDL.
Measurement of the total lipid mass (total cholesterol + TG) in HMDMs and THP-1 macrophages in the present work (Table 4
) gave results broadly similar to those obtained by determining the percentage of oil red O staining per cell; thus, a considerable increase was observed with oxLDL, CMR-LPs, and oxCMR-LPs. As expected, oxLDL caused a large rise (approximately 3-fold) in the CE content in both types of macrophages, and in agreement with the findings of Aviram with THP-1 cells (38), this was accompanied by a marked rise (2- to 2.5-fold) in intra-cellular cholesterol levels (Table 5
). In addition, however, we also observed a 2- to 3-fold increase in TG in both HMDMs and THP-1 cells, and this had the effect of maintaining the ratio of total cholesterol to TG at a level (1.52) similar to that found in macrophages incubated without lipoproteins. In THP-1 cells, although not in HMDMs, unmodified LDL also caused a significant increase in TG mass (Table 5
), and this led to a decrease in the ratio of total cholesterol to TG to 1.1. We conclude, therefore, that oxLDL and, to a much more limited extent, LDL have a similar effect to acetylated LDL (43, 44) in causing mixed lipid droplets containing CE and TG to accumulate in macrophages during foam cell formation. Our findings also show that, in contrast to oxLDL, CMR-LPs and oxCMR-LPs induced greater increases in the TG content of macrophages than in the cholesterol or CE levels, particularly in THP-1 cells (Table 5
), resulting in a decrease in the ratio of total cholesterol to TG from 1.5 to 0.9 (CMR-LPs) or 1.1 (oxCMR-LPs) in HMDMs and from 2.1 to 0.8 (CMR-LPs) or 1.0 (oxCMR-LPs) in THP-1 cells. These findings are in agreement with our previous, more limited study with chylomicron remnants and J774 macrophages (35). Thus, these results clearly show that macrophage foam cells induced by chylomicron remnants differ from those induced by oxLDL in that TG makes up a greater proportion of the accumulated lipid.
To further investigate the influence of oxLDL and chylomicron remnants on cholesterol esterification and TG synthesis during foam cell formation, experiments were performed to measure lipid accumulation in the presence of the ACAT inhibitor CI976. CI976 is a potent inhibitor of hepatic and intestinal ACAT activity (45), and in our experiments a concentration of 10 µM reduced cholesterol esterification by almost 90% but had no effect on TG synthesis (Table 6
). Histological staining with oil red O clearly demonstrated that lipid accumulation in response to oxLDL, CMR-LPs, and oxCMR-LPs was considerably reduced in the presence of the inhibitor (Table 3
). The finding that control levels were reached in the case of oxLDL but not CMR-LPs or oxCMR-LPs, in both HMDMs and THP-1 macrophages (Table 3
), however, provides further evidence to suggest that TG accumulation is quantitatively more important than CE accumulation in foam cell induction by chylomicron remnants compared with oxLDL.
Inhibition of cholesterol esterification might be expected to lead to increased accumulation of cholesterol within the cells, particularly since no cholesterol acceptor was present in the medium in our experiments. In fact, although the mass of CE was reduced to or below control levels in both HMDMs and THP-1 cells, the cholesterol content was also decreased compared with that seen in macrophages exposed to oxLDL, CMR-LPs, or oxCMR-LPs in the absence of the inhibitor. These findings contrast with those reported for the ACAT inhibitors E5324, HL-004, and NTE-122, which reduced CE mass in macrophages loaded with acetylated LDL but increased the mass of cholesterol in the absence of a cholesterol acceptor (4648). Inhibition of ACAT using 58-035, however, was found to increase cholesterol efflux from HMDMs after treatment with acetylated LDL (49), and it was suggested that apoE secreted by the cells may act as an acceptor.
Although CI976 did not appear to affect TG synthesis in the absence of lipoproteins (Table 6
), TG levels were reduced when HMDMs were exposed to lipoproteins, with values falling to control levels with both oxLDL and CMR-LPs (Table 5
). This effect was less marked in THP-1 macrophages, particularly with CMR-LPs and oxCMR-LP; nevertheless, a small decrease in TG levels was recorded in the presence of the inhibitor. These results suggest that the metabolism of the extra lipid taken up by the macrophages from the lipoproteins is affected by CI976, possibly by diversion of the fatty acids derived from the TG into other substrate pools, such as those for the blocked cholesterol esterification pathway or for phospholipid synthesis.
ACAT converts excess cholesterol taken up by macrophages to CE, which is stored in the cytosol, but undergoes a continuous cycle of hydrolysis by CE hydrolase and re-esterification by ACAT (50). Measurement of the incorporation of radiolabeled fatty acid into CE, therefore, gives an indication of the overall rate of cholesterol esterification. Previous work using this technique has shown that acetylated LDL increases cholesterol esterification in THP-1 macrophages (11, 12, 41) and in HMDMs, whereas LDL has a lesser effect (51). In our experiments, incubation of THP-1 macrophages with oxLDL, but not LDL, led to a significant increase in cholesterol esterification, and this is consistent with the relative changes in cellular CE content induced by the two types of lipoprotein (Tables 5
and 6
). Similarly, CMR-LPs and oxCMR-LPs caused a significant increase in cholesterol esterification, and both types of particles caused CE to accumulate in the cells (Tables 5
and 6
). Yu and Mamo (52) have also reported that chylomicron remnants increase the rate of cholesterol esterification in rabbit alveolar macrophages. In contrast, in our previous work with J774 cells, we found no effect of remnants on cholesterol esterification (40); however, in these experiments the cells were exposed to the particles for only 4 hrs rather than 48 hrs as in the present study.
[3H]oleate may be incorporated into TG and CE, and this can be used to give an indication of the rate of TG synthesis from fatty acids. Our experiments demonstrate that LDL, oxLDL, CMR-LPs, and oxCMR-LPs all increased TG synthesis in THP-1 macrophages and, furthermore, that the level of [3H]oleate incorporated into TG compared with CE was many fold higher. These results indicate that the TG pool is more metabolically active than the CE pool in these cells and are in agreement with the findings of Kritharides et al. (43) in similar studies using THP-1 macrophages loaded with acetylated LDL. The TG synthesis was increased to a similar extent by LDL and oxLDL, accounting for the similar levels of TG accumulation induced by the two lipoproteins in THP-1 cells (Table 5
). In addition, the greater increment in TG synthesis observed with CMR-LPs and oxCMR-LPs compared with LDL and oxLDL is also in accord with the greater increases in TG mass found (Table 5
).
Although the course of the stimulation of cholesterol esterification in the macrophages was similar when oxLDL, CMR-LPs, or oxCMR-LPs were used, that for TG synthesis differed in that oxLDL caused an immediate sharp rise, whereas with CMR-LPs or oxCMR-LP there was a lag phase of approximately 6 hrs (Fig. 2
). These differences may reflect different pathways of uptake of the two types of particles. The oxLDL is taken up by macrophages via a number of different receptors, including class A and B scavenger receptors and the leptin-like oxLDL receptor-1 (53). The receptors involved in macrophage chylomicron remnant uptake, however, are not yet clearly defined but may be different from those used for oxLDL. The LDL receptor may be involved (54, 55), and Elsegood et al. (55) have found that chylomicron remnants bind to unidentified 43-kDa proteins in THP-1 macrophages. A receptor expressed on macrophages reported to recognize apoB48 has also been described (56).
In this study, the response of HMDMs and THP-1 macrophages to CMR-LPs was found to be similar in most respects, indicating that the THP-1 cell line provides a suitable model for study of the effects of chylomicron remnants on macrophage lipid accumulation. The major novel findings from the results obtained with the two types of human macrophages are that CMR-LPs, without prior oxidation, cause lipid accumulation in human macrophages to an extent that is comparable to that caused by oxLDL; a greater proportion of the lipid accumulated in response to the CMR-LPs is in the form of TG rather than CE, which accumulates in response to oxLDL; TG synthesis is increased to a greater extent by CMR-LPs compared with oxLDL; and cellular TG levels after exposure to oxLDL or CMR-LPs are decreased when cholesterol esterification is inhibited, although TG synthesis was unaffected, suggesting that fatty acids are diverted into other substrate pools in these circumstances. In addition, the experiments demonstrate that oxidation of CMR-LPs does not change their effects on the amount or type of lipid accumulated or on CE or TG synthesis in the cells. Thus, in contrast to LDL, oxidation of chylomicron remnants does not appear to enhance their induction of foam cell formation. We conclude, therefore, that the characteristics of lipid accumulation in human macrophages exposed to chylomicron remnants differ in many important respects from those found in cells exposed to oxLDL. Overall, these findings provide good evidence to support the idea that chylomicron remnants have a role in foam cell formation that is distinct from that of oxLDL and highlight the need for further studies of their effects.
| Footnotes |
|---|
Received for publication November 13, 2003. Accepted for publication March 7, 2004.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
P. M. Hutchins, R. M. Barkley, and R. C. Murphy Separation of cellular nonpolar neutral lipids by normal-phase chromatography and analysis by electrospray ionization mass spectrometry J. Lipid Res., April 1, 2008; 49(4): 804 - 813. [Abstract] [Full Text] [PDF] |
||||
![]() |
V. Saraswathi and A. H. Hasty The role of lipolysis in mediating the proinflammatory effects of very low density lipoproteins in mouse peritoneal macrophages J. Lipid Res., July 1, 2006; 47(7): 1406 - 1415. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |