Materials and methods

Chemicals

All tissue culture media were from Life Technologies, Gibco BRL. Bovine pituary extract was purchased from Clonetics. Fatty-acid-free bovine serum albumin (BSA), 3,3,5-triiodo-L-thyronine, tocopherol phosphate disodium, the lipid TLC and high performance liquid chromatography (HPLC) standards, and all the fatty acids used were purchased from Sigma. The radioactive fatty acids were purchased from American Radiolabeled Chemicals, St. Louis, MO. All the organic solvents were HPLC grade. Chloroform was from B&J, Muskegon, MI, acetonitrile, toluene, and acetic acid from J.T. Baker, Phillipsburg, NJ, methanol from Fisher, Pittsburgh, PA, and ether from Mallinckrodt, Paris, KY. HPLC columns were from Supelco, Bellefonte, PA, and HPTLC plates from Whatman, Clifton, NJ. Scintillation counting materials were provided by DuPont and National Diagnostics, Atlanta, GA. Ultima Flo M was provided from Packard, Meriden, CT.

Maintenance and culture conditions

Samples of facial skin (otherwise to be discarded) were obtained from female patients 45–65 y of age following informed consent and IRB approval. The subjects were healthy females without a history of retinoid or laser treatment or chemotherapy. The specimens were obtained from facial lift surgery within 5 h from the end of the surgery. Upon arrival, they were washed in phosphate-buffered saline. The hair and subcutaneous fat was removed using curved surgical scissors. The remaining skin was washed again in Dulbecco's minimal essential medium (DMEM) containing penicillin (100 units per ml) and streptomycin (100 mg per ml). Two millimeter biopsy punches were taken from (these) tissues rich in sebaceous glands and cultured in serum-free medium. The medium consisted of DMEM and F₁₂ (3:1) containing, as described previously (^{Guy and Kealey, 1998}), L-glutamine, penicillin-streptomycin, sodium pyruvate (Gibco, Rockville, MD), a selenium-insulin-transferrin mixture (Gibco), 3,3,5-triiodo-L-thyronine (Sigma, St. Louis, MO), trace elements (Gibco), and bovine pituitary extract (Clonetics, San Diego, CA). Finally, prior to incubation of biopsies, the conjugated fatty acid to BSA was added to a final concentration of 50 M and 1 Ci per ml of ¹⁴C fatty acids.

Fatty acid preparation

To conjugate fatty acids to BSA, they were initially dissolved in 0.1 M of NaOH to yield a 50 mM final concentration and heated while stirring at 90°C for saturated fatty acids and 70°C for unsaturated fatty acids until dissolved (a few seconds), as previously described (^{Hovik et al, 1997}). A 0.2 ml aliquot of the resulting fatty acid solution was added while stirring into 1.2 ml of a 10% BSA solution at 37°C. After 15 min of slow stirring to allow clarification of the solution, 0.6 ml of water was added to bring the final concentration to 5 mM of fatty acid (100 solution). The solutions were filtered through acrodiscs of 0.2 m and stored at 4°C for up to 1 mo. Just prior to use, the radioactively labeled fatty acids were conjugated to the fatty acid BSA stock solution. The radioactive fatty acids, ¹⁴C-labeled fatty acid at the carboxyl carbon, were dried under argon and the fatty acid BSA (5 mM) was added to the dried radioactive fatty acid followed by sonication in a water bath sonicator until the entire radioactive material was dissolved. Ten microliters per Ci fatty acid of the fatty acid BSA 5 mM was added followed by sonication in a water bath sonicator (Branson 3510) until the radioactive material was dissolved. This process was monitored by scintillation counting.

Lipid extraction

For each experimental point, four biopsy punches were homogenized in 2 ml of CHCl₂:MetOH (2:1) using Kontes 2 ml glass homogenizers until a milky homogenate was obtained without the presence of visual debris. The homogenate was transferred to a glass tube containing 1 ml of 0.88% potassium chloride. The homogenizer was then rinsed with 2 ml of the extraction solvent twice, and these washes were added to the combined extract. After phase separation, the aqueous phase was removed and discarded and the remaining total organic extract was evaporated under a gentle stream of argon prior to HPTLC analysis.

HPTLC analysis

The lipid components extracted from the explanted glands were analyzed by chromatographic separation on 10  20 cm Whatman HPTLC Silica G plates that had been previously charged with chloroform, heated in a vacuum oven at 105°C for 30 min, and cooled to room temperature in a dessicator. After spotting the lipids that were dissolved in CHCl₃:MetOH (2:1), the plates were developed three times, as previously reported (^{Doran et al, 1991}), as follows: (i) hexane to the top; (ii) toluene to the top; and (iii) hexane:ether:acetic acid (70:30:1) 10 cm from the top. Between each mobile phase, a 15 min drying time at room temperature in a standard fume hood was carried out to ensure complete evaporation of the solvent. Squalene, waxes, and sterol esters are separated during the first two phases whereas the various polar lipids are resolved during the third phase.

HPLC analysis

Performance of fatty acid analysis was modified from^{Osterroht (1987)} to simultaneously visualize the radioactively labeled and the total unlabeled fatty acids that were derivatized with a chromatophore. Because this process involves the saponification of fatty acids from lipids that were soluble only in nonpolar solvents prior to their derivatization, a technique was developed incorporating the methods of^{Doran et al (1991)}. Preliminary studies established the quantitative recovery of fatty acids. In brief, lipid extracts dissolved in 125 l of toluene were saponified with 250 l of 1 N KOH in 95% MeOH at 70°C for 1 h. The samples were then kept at room temperature for 10–20 min and 125 l of 2 N HCl plus 375 l of chloroform were added. After stirring for 15 min at room temperature and removal of the stirrers the sample were left at room temperature for 30 min to allow phase separation. If the phases were not clear the samples were centrifuged at 1000 rpm for 5 min. The lower, organic phase was transferred to a small clear reacti-vial (Pierce) and was dried under a gentle stream of argon. Then, samples were redissolved in 400 l of methanol and 20 l of 1% phenolphthalein in methanol was added prior to titrating with 0.05 N KOH in methanol to the phenolphthalein end point. After drying to completion under a gentle stream of argon, 100 l of derivatization reagent (p-bromophenacyl-8, Pierce) plus 200 l of acetonitrile were added. The samples were incubated at 80°C for 30 min with continuous stirring.

One hundred and twenty microliters of the resulting samples were injected into a 4.6  250 mm, 5.0 m C8 Supelco column connected to a Hewlett Packard model HP1090 HPLC in line with Packard model A 500 radiomatic flow scintillation counter. Chromatographic separation was obtained in which a 10 min acetonitrile:H₂O (65%:35%) isocratic elution was followed by a 30 min linear acetonitrile:H₂O gradient running from 65% to 80% acetonitrile. This was in turn followed by a 40 min 80% acetonitrile elution. The temperature was maintained at 30°C and the flow rate at 1.0 ml per min. The derivatized fatty acids were detected by absorbance at 250 nm. Radioactive fatty acids were detected and quantitated using Packard FLO-ONE software. Identification of the individual fatty acids was by coelution with pure labeled and unlabeled standards.

Phosphoimager analysis

A Bio-Rad Phosphoimager (Molecular Imager FX) with the Quantity One 4.1.0 software was used. The phosphoimager analysis was performed to quantify the amount of radioactivity in each lipid fraction of the HPTLC analysis.

Results

Incorporation of fatty acids into sebaceous lipids

For the experimental work described we chose to use biopsy punches in preference to dissected glands, because sebaceous glands cannot be dissected from the surrounding tissue without the presence of contaminating cells of dermal or epidermal origin. In addition, in preliminary experiments with dissected glands (data not shown), identical profiles of labeled sebaceous-specific lipids were observed. Consequently experiments were performed using biopsy punches, which also preserve intact sebaceous glands surrounded by their natural microenvironment.

Biopsy punches from human facial skin rich in sebaceous glands were placed in culture as described in Materials and Methods and were labeled for 24 h with radiolabeled palmitic, palmitoleic, stearic, oleic, and linoleic acids.

Kinetic analysis of label incorporation into the organ cultures demonstrated that the uptake of label was linear with time for the first 24 h. Wax ester synthesis, a marker of the differentiation state and viability of the sebaceous gland, also increased in a linear fashion for the first 24 h and declined significantly with additional time of incubation. All of the lipids that contain a fatty acid in their moiety were labeled, such as triacyl- or diacyl-glycerols, phospholipids, sterol and wax esters, as well as the free fatty acids or alcohols. In contrast to acetate labeling, lipids such as cholesterol and squalene that do not contain a fatty acid were not labeled.

HPLC analysis of fatty acids harvested from the cultured biopsies demonstrated that fatty acids conjugated to BSA were delivered to the cell and its metabolism (Figure 1). The palmitic and palmitoleic acids were metabolized by elongases, to products that were extended by two carbons. The elongation process did not proceed to longer chain fatty acids, however. This observation is in agreement with the chromatographs obtained from stearic and oleic acids, which were not further elongated to longer chain fatty acids.

Figure 1.

Biopsy punches from human facial skin take up the radiolabeled fatty acids and elongate the 16C chain but not the 18C fatty acids. Reversed phase HPLC analysis of extracted and saponified lipids from the cultured skin explants shows (a) ¹⁴C-palmitic acid (16:0) and (b) ³H-palmitoleic acid (18:1, 9) to be elongated to the 18:0 and 18:1, 11 products, respectively. Arrows point to the elongated products. (c) ¹⁴C-stearic acid and (d) ¹⁴C-oleic acid were not elongated or changed.

Full figure and legend (24K)

HPTLC analysis demonstrated that the supplied fatty acids were incorporated into the metabolism of the cell and they were utilized to synthesize the major cellular lipids (Figure 2). As shown in Figure 2, the fatty acids were esterified into the major classes of lipids including triacyl- or diacyl-glycerols, phospholipids, sterol and wax esters, and were found as free fatty acids or alcohols.

Figure 2.

Selective incorporation of fatty acids into human sebaceous lipids. Human sebaceous explants were labeled for 24 h with ³H-palmitoleic aicd (16:1), ¹⁴C-palmitic acid (16:0), stearic acid (18:0), oleic acid (18:1), and linoleic acid (18:2). HPTLC analysis of the extracted lipids revealed incorporation into all the major classes of lipids that contain fatty acids in their moiety: sterol esters (SE), wax esters (WE), triglycerides (TGs), fatty acids and alcohols, diacylglycerols (DGs) cholesterol (Cho) and polar lipids (PLs).

Full figure and legend (171K)

Selective utilization of palmitic acid

The HPTLC analysis of the extracted lipids revealed the incorporation of the fatty acids into wax esters, a product that is solely formed by the sebaceous gland biosynthetic activity. The radioactivity incorporated into the wax ester fractions was quantified by phosphoimaging analysis. The incorporation into wax esters was the highest for palmitic acid and undetectable for palmitoleic acid (Figure 3). Incorporation of palmitic acid into wax esters was at least 30% higher than that of stearic acid. In contrast the 9 desaturated 16-carbon fatty acid, palmitoleic, was not incorporated into wax esters. A comparable pattern was also observed with the 18-carbon fatty acids. Stearic acid, the saturated fatty acid, was preferred over oleic (the 9 mono-unsaturated) for wax ester synthesis. Oleic was incorporated into waxes at a significantly lower rate than stearic. Overall, palmitic was the preferred fatty acid substrate for wax ester biosynthesis and there was greater preference for saturated over mono-unsaturated fatty acids.

Figure 3.

Selective incorporation of saturated fatty acids into wax esters. Biopsy punches from human facial skin were labeled for 24 h with ³H-palmitoleic acid (16:1), ¹⁴C-palmitic acid (16:0), stearic acid (18:0), and oleic acid (18:1) and the wax ester fraction from the HPTLC was quantified by phosphoimager analysis. As shown there is a greater preference for saturated fatty acids (16:0, 18:0) and in particular for the 16-carbon palmitic acid.

Full figure and legend (13K)

Selective utilization of palmitoleic acid

The majority of the radioactive fatty acid substrates were accumulated into triglycerides and polar lipids, as shown by quantitative phosphoimager analysis of the HPTLC plates (Figure 4). Palmitic, stearic, oleic, and linoleic acids were esterified into triglycerides and polar lipids at the same rate and their distributions between triglycerides and polar lipids were very similar. Interestingly, palmitoleic acid was the only fatty acid that was preferentially fractionated into the polar lipids and was not esterified into triglycerides to the same degree as the other fatty acids used in this study.

Figure 4.

Selective incorporation of palmitoleic acid into polar lipids. Biopsy punches from human facial skin were labeled for 24 h with ³H-palmitoleic acid (16:1), ¹⁴C-palmitic acid (16:0), stearic acid (18:0), oleic acid (18:1), and linoleic acid (18:2) and the polar lipid (origin) and triglyceride fraction analyzed by HPTLC were quantified by phosphoimager analysis. The results show that palmitoleic acid is preferentially fractionated into polar lipids and is not esterified into triglycerides to the same degree as the other fatty acids.

Full figure and legend (14K)

Selective utilization of linoleic acid

Of the lipids tested, linoleic acid was the only fatty acid that gave a unique pattern of distribution into lipids. As shown in Figure 2, squalene was also labeled when radiolabeled linoleic acid was applied to the culture, suggesting that this fatty acid is metabolized into two carbon precursors. Each ¹⁴C-labeled fatty acid used for this study had the label in the carboxylic carbon of the molecule. The release and utilization of this labeled carbon atom in metabolic reactions other than the esterification of the whole fatty acid into different lipids requires the activation of the -oxidation mechanism. The result of this process is the breakdown of the fatty acid from its carboxylic end and the release of labeled acetyl coenzyme A (acetylCoA). The latter is used for the formation of other lipids such as squalene and cholesterol along with other cell metabolites that are not direct products of the intact fatty acid molecule. From the HPTLC analysis, it becomes obvious that synthesis of squalene and other lipids from the radioactive linoleic acid was not observed when the cultured biopsy punches from human facial skin were labeled with other fatty acids. The area between the polar lipids and triglycerides was labeled more intensely when ¹⁴C-linoleic was used compared to the same area when other fatty acids were used. This indicates that more lipid products were formed when ¹⁴C-linoleic was used as the substrate, lending further support to the hypothesis that linoleic acid was oxidized in the organ cultures.

Additional proof for the -oxidation of linoleic acid comes from the fatty acid analysis by HPLC, shown in Figure 5. In contrast with the HPLC analysis of other fatty acids (Figure 2), the radiomatic trace from the linoleic acid experiment revealed the presence of other labeled peaks that correspond to fatty acids with shorter chains. The presence of radioactive products of shorter chain length and with one double bond can be explained only if two carbon precursors are generated from the oxidation of linoleic acid. This may explain the labeling of the two major fatty acids of the sebaceous cells, palmitic and stearic acid. These data are in agreement with the labeling of squalene when glands are incubated with ¹⁴C-linoleic acid. The degradation of the linoleic acid due to chemical oxidation was monitored by HPLC and it was insignificant.

Figure 5.

Oxidation of linoleic acid by human sebaceous glands. Biopsy punches from human facial skin were labeled with linoleic acid for 24 h. The lipids were extracted and saponified, and the derivatized fatty acids were subjected to reverse phase (C8) HPLC. Arrows point to the products generated from the oxidation of linoleic acid.

Full figure and legend (10K)

Oxidation of linoleic acid correlated with the ability of cultured glands to synthesize sebaceous lipids

As the organ culture system is derived from intact tissue, it contains a mixed population of cells such as keratinocytes and fibroblasts besides the sebaceous cells, which could contribute to the process of -oxidation. Therefore, we investigated the cellular origin of the linoleic acid oxidation.

Biopsy punches from human facial skin were labeled with ¹⁴C-acetate, shown in Figure 6, and at 24 h the sebaceous-specific lipids, squalene and wax esters, appeared as prominent bands, indicative of active biosynthesis of sebaceous-specific lipids. On the third day in culture, however, the wax ester band was diminished and at the seventh day in culture wax ester synthesis ceased and the synthesis of squalene was significantly less than in the first 3 d. No significant changes in the biosynthesis of other nonsebaceous-specific lipids were observed with time. The data are evidence that biopsy punches from human facial skin maintained under our culture conditions gradually lose their sebogenic phenotype characterized by the synthesis of wax esters and squalene, whereas the biosynthesis of the other lipids is retained.

Figure 6.

HPTLC analysis of the incorporation of ¹⁴C-acetate into human sebaceous lipids. Biopsy punches from human facial skin show maximal synthesis of sebum lipids (Sq, WE) upon incubation with ¹⁴C-acetate during the first 24 h in culture. With time in culture the explants gradually lose the capacity to synthesize wax esters and squalene.

Full figure and legend (57K)

In Figure 7(a), biopsy punches from human facial skin were labeled with palmitic and linoleic acid during the first 24 h in culture. Both fatty acids were incorporated into wax esters indicating functional sebaceous cells. The other lipids were also labeled, and in the case of linoleic acid the label also was incorporated into squalene. In addition, the polar lipid region of the HPTLC plate was more intensely labeled by the linoleic precursor compared to the ¹⁴C-palmitic acid control. An additional band where cholesterol migrates is visible as well in the case of linoleic-acid-labeled cultures.

Figure 7.

Oxidation of linoleic acid correlates with loss of the ability of cultured glands to synthesize sebaceous lipids. (A) HPTLC analysis of biopsy punches from human facial skin labeled with palmitic and linoleic acid during the first 24 h in culture. Both were incorporated into wax esters (WE) indicating sebaceous lipid synthetic ability. In addition, a prominent band of squalene (Sq) is visible in linoleic-acid-labeled cells. In glands cultured for 3 d in which wax ester synthesis is reduced (Figure 6) neither incorporation of linoleic into waxes nor oxidation into two carbon precursors for squalene synthesis is seen. (B) Reversed phase HPLC analysis of biopsy punches from human facial skin cultured for 1 d demonstrates the elongation of palmitic acid to stearic (a) and the oxidation of linoleic acid (b). As shown in (c) after 3 d in culture the oxidation of linoleic acid was substantially reduced and this was correlated with the loss of sebaceous lipid synthesis.

Full figure and legend (62K)

The accumulation of radioactive squalene, when biopsy punches from human facial skin are labeled with radioactive linoleic acid, can only be attributed to sebaceous gland activity. Degradation of linoleic acid by any other cell type would not yield any accumulation of squalene, as this is a sebaceous-specific property.

After culturing the biopsy punches from human facial skin for a further 3 d, both wax ester and squalene synthesis were reduced to barely detectable counts over background, when using linoleic acid as the precursor. At the same time these cultures were able to synthesize squalene from acetate (Figure 6). So obviously no radioactive two-carbon precursors are produced by linoleic acid on the third day of culture when sebaceous glands lose their differentiation potential and the ability to synthesize wax esters.

These results are further supported by the HPLC analysis of the fatty acid metabolism (Figure 7b). For the first 24 h in culture the expected elongation of palmitate to stearate and the oxidation of linoleate were observed. After 3 d in culture the oxidation of linoleic acid was substantially reduced. The significantly reduced breakdown of linoleic acid in Figure 7(b, c) is an additional indication that nonsebaceous dermal and epidermal cells did not contribute significantly towards this reaction.

Oxidation is specific to linoleic acid

The catabolism of linoleic acid could be specific for that fatty acid. Alternatively the presence of linoleic acid could induce the catabolism of other fatty acids. These questions cannot be answered with experiments where a single exogenous labeled and unlabeled fatty acid is used. Figure 8 describes two possible scenarios. In the first (Figure 8a), linoleic acid affects a receptor that signals the onset of -oxidation of all free fatty acids available. In the second (Figure 8b), linoleic acid is specifically targeted for its own oxidation, with no effect on other fatty acids present in the culture.

Figure 8.

Schematic mechanisms of the possible ways that -oxidation of linoleic acid takes place. (A) Through the action of a receptor, where LA is linoleic acid, FA-14COOH the radiolabeled fatty acid, R the receptor and CH₃¹⁴ COCoA the ¹⁴C-labeled acetylCoA. (B) The oxidation is specific for linoleic acid.

Full figure and legend (16K)

To examine these two possible scenarios, biopsy punches from human facial skin were labeled with ¹⁴C-palmitic acid in the presence of unlabeled linoleic acid. Figure 9(a) confirms the incorporation of both labeled ¹⁴C-palmitic acid and ¹⁴C-linoleic acid into wax esters, indicating sebaceous gland activity when each was used as a single labeled and unlabeled entity. As expected, when linoleic acid was used as a single entity, the labeling of squalene was observed, consistent with -oxidation. When biopsy punches from human facial skin were incubated with ¹⁴C-palmitic acid in the presence of unlabeled linoleic acid, no labeling of squalene or other lipids that could be accounted for by the generation of two carbon precursors was observed. In particular, palmitic acid was incorporated into waxes in the presence and absence of linoleic acid. In addition, there was no perceivable incorporation of ¹⁴C into squalene when ¹⁴C-palmitic was used in combination with unlabeled linoleic acid. Also, fatty acid metabolism analysis by HPLC (Figure 9b) showed no evidence supporting -oxidation of ¹⁴C-palmitic acid tracer in the presence of linoleic acid, as no labeled products were formed other than the expected elongation product, stearate. Thus, linoleic acid oxidation is specific for its own oxidation and does not activate -oxidation of other fatty acids in our culture system.

Figure 9.

Oxidation of linoleic acid is specific for that fatty acid and its presence does not induce the oxidation of palmitic acid in cultured biopsy punches from human facial skin. (A) Biopsy punches from human facial skin were labeled with palmitic acid, linoleic acid, and palmitic in the presence of linoleic during the first 24 h in culture. As shown in panel C, addition of linoleic acid to the ¹⁴C-labeled palmitic acid did not result in squalene synthesis, proving the specificity of linoleic acid oxidation. (B) Reversed phase HPLC analysis of biopsy punches from human facial skin cultured for 1 d in the presence of ¹⁴C-palmitic acid and cold linoleic acid. Label that could be accounted for by two carbon precursors arising from palmitate was not detected. Thus linoleic acid oxidation is specific for that fatty acid.

Full figure and legend (66K)

Discussion

In this study we investigated the metabolic fate of the most important and abundant fatty acids in human sebaceous glands. Our goal was to explore wax ester biosynthesis by the sebaceous cells, as they are the only cells of the human body that synthesize and accumulate these unique lipids. The natural and direct precursors of the waxes are fatty acids and fatty alcohols. Subsequently, fatty acids are also the direct precursors of the fatty alcohols (^{Kolattukudy, 1980}). The importance of fatty acid metabolism in sebaceous glands is demonstrated by the recent discovery of the asebia mouse mutation (^{Zheng et al, 1999}). The stearoyl CoA desaturase, an enzyme responsible for the initial desaturation of fatty acids, is disrupted in the asebia mice, which demonstrate abnormal maturation of the sebaceous glands (^{Sundberg, 1994}). In addition, the fatty acids of sebum are unusual for mammalian cells because they contain very long chains, branched moieties, and, as in the case of humans, 6 mono-desaturated products. Whereas the lipid composition of sebum is generally species specific, wax esters appear to be present in all species. Moreover, wax esters are present in the secretions of other modified sebaceous glands, such as the meibomian gland, the preputial gland of rodents, and the avian uropygial gland (^{Kolattukudy, 1980}). Thus, understanding the mechanism and regulation of wax ester synthesis is an important step towards the elucidation of the regulation of sebaceous gland function and differentiation.

The mechanisms by which human sebaceous cells utilize fatty acids for wax ester synthesis is poorly characterized. This is the first study to follow the fate of exogenously supplied fatty acids in human sebaceous glands. We introduced a system in which waxes as well as other cellular or secreted lipids were labeled directly from radiolabeled fatty acids. This was done in serum-free media by conjugating fatty acids to fatty-acid-free BSA, thus eliminating possible interference by other contaminating fatty acids. As parallel experiments in serum-containing media showed that no products of differentiated cells are obtained, serum-free conditions was the method of choice for the experiments reported here. The culture system employed in these experiments has the disadvantage of contamination by other nonsebaceous tissues. Thus conclusions cannot be made about the incorporation of fatty acid labeling into lipids such as triglycerides and cholesterol esters, which are synthesized by both sebaceous and nonsebaceous cells.

Palmitate, palmitoleate, stearate, and oleate were mainly esterified into lipids without being catabolized to any oxidation or breakdown products. In the culture system, elongases responsible for the production of very long chain fatty acids and the desaturases, which add double bonds, left the fatty acids unchanged. This led us to conclude that the above enzymes are not expressed or active when the biopsies are placed into culture. Only the 16-carbon fatty acids were found to be partially elongated to their respective 18-carbon products. Mammalian cells have at least three elongases (^{Tvrdik et al, 2000}), which are expressed in various tissues including skin. One elongase is responsible for the elongation of 16–18 carbon chains and the other two are responsible for very long chain fatty acid synthesis, which may explain why we only observe the elongation up to 18 carbons and not more. This particular elongase is apparently expressed and active in our in vitro system.

Lipid analysis demonstrated that ¹⁴C-palmitic acid resulted in optimal labeling of wax esters compared to the other fatty acids used. In a previous study (^{Nordstrom et al, 1986}), when fatty acids from human sebaceous lipid extracts at steady state were analyzed, palmitic acid was one of the predominant fatty acids in waxes. Several reports (^{Kolattukudy, 1980;Perisho et al, 1988;Strauss et al, 1991}) also state that the fatty acid ends of the waxes are usually saturated compared with the alcohol ends, which are commonly desaturated. These data further support our observations for the preference of palmitic acid incorporation into wax esters. Generally saturated acids were incorporated to waxes at much higher rates than unsaturated ones. Palmitoleic acid was not detected in the wax ester fraction at all, and oleic acid was detected to a much lesser degree than stearic acid. The difference between saturated and mono-unsaturated substrates was even more pronounced when the 16-carbon fatty acids were compared to each other. This finding is significant considering that, overall, the 16-carbon fatty acids are predominant in sebaceous cells; they account for more than 50% of the total fatty acid fraction in sebum (^{Nicolaides, 1974;Nikkari, 1974;Stewart and Downing, 1991}). Moreover, it reinforced the fact that the selective utilization of the fatty acids in sebaceous cells can be of great importance. Consequently, further analyses were performed to elucidate the lipid fraction in which palmitoleic acid accumulates.

From the labeling experiments, we found that most of the labeled fatty acids were esterified into triglycerides or polar lipids. As was demonstrated, palmitate, stearate, oleate, and linoleate were present in almost equal amounts between the above-mentioned fractions with a similar distribution pattern at steady state conditions. Palmitoleate was the only fatty acid that accumulated to a much higher degree into polar lipids, however, and was not further esterified to sebaceous lipids, i.e., waxes or even triglycerides. This finding may in part be due to the fact that under our culture conditions the rates of the wax ester and triglyceride biosynthesis are not as high as the in vivo reported rates. It is known that waxes account for about 25% of human sebum (^{Nicolaides, 1974;Nikkari, 1974;Stewart and Downing, 1991}) and this cannot be reproduced ex vivo most probably because some of the in vivo expressed pathways are not active in our culture conditions. Therefore, it is possible that palmitoleic accumulates into a specific class of lipids, the polar lipids, as its processing enzymes are minimally expressed. Whether this observation is due to culture conditions or to the fact that these studies were done with skin of postmenopausal females, known to have decreased sebum production (^{Jacobsen et al, 1985;Downing et al, 1989}), needs to be further investigated. An alternative explanation could be that palmitoleic accumulates in the polar lipid fraction, containing such lipids as phospholipids and sphingolipids, for its putative role in sebaceous cell membrane fluidity.

The most intriguing result was the metabolism of the linoleate from the sebaceous glands. From the HPTLC analysis it was obvious that linoleic acid is processed in a very different way than the other fatty acids. Only the labeled carbon from this substrate incorporated into squalene, whereas the other labeled substrates did not. In the case of linoleic acid, oxidation of the fatty acid led to the generation of two carbon unit precursors that were incorporated into different metabolic routes as well into the de novo synthesis of squalene. The HPLC analysis demonstrated as well that the de novo biosynthesis of the major sebaceous fatty acids such as palmitic and oleic can only occur if linoleic acid is oxidized and broken down to two carbon precursors.

This result is significant as it reinforces our initial hypothesis that there is a selective utilization of the fatty acids in sebaceous cells. Linoleic acid was not oxidized when sebaceous cells were cultured under conditions that prevented sebaceous cell differentiation. This finding suggests that the linoleic acid oxidation is associated with differentiating sebaceous cell activity and function. The ability to synthesize wax esters correlated with the ability to oxidize linoleic acid. Loss of both activities, however, did not result in the loss of any other labeled lipid class in culture. All other lipids were labeled as expected without a significant change to the metabolic activity of the culture.

The significance of this result is further reinforced by the fact that linoleic acid is a specific target of the -oxidation enzymes. We have demonstrated that linoleate did not trigger the general onset of -oxidation, and other fatty acids were not -oxidized in the presence of linoleate. This is quite intriguing and suggests that linoleate is targeted by enzymes that lead to its degradation due to its unique -6 double bond. We hypothesize the involvement of a lipoxygenase that targets specifically linoleic acid to form the hydroxyperoxide intermediate, a moiety that is highly susceptible to -oxidation (^{Feussner et al, 1997}). As the other unsaturated fatty acids used in this study were not oxidized the lipoxygenase probably targets the double bond between the 12 and 13 carbons. Hydroxyperoxides are either converted to hydroxylated fatty acids or rapidly oxidized for removal as they are toxic to cells.

The selective attack of linoleic acid in sebaceous cells suggests that the human gland has evolved a mechanism to protect itself from this essential fatty acid. This may be because linoleic acid is a known target for 6 desaturation to form -linolenic acid. One possibility stems from the fact that the most abundant fatty acid in human sebum is sapienic acid, which is the 16:1, 6 that is produced by the desaturation of palmitic acid (16:0) by 6 desaturase. This fatty acid is not abundant anywhere else in the human body and is a unique feature of the sebaceous gland. 6 desaturase activity, which is responsible for the synthesis of sapienic acid, is high in the liver, adrenal gland, and skin. Low levels of sapienic acid in other tissues than sebaceous glands can be accounted for by the fact that 6 desaturase is preferably converting linoleic acid to -linolenic acid (^{Svensson, 1983;Christiansen et al, 1991}). Apparently 6 desaturase has higher affinity for linoleic acid (^{Svensson, 1983}), which most probably results in underutilization of palmitic acid as a substrate for this enzyme. As a result sapienic acid is not being made in significant quantities anywhere else other than in sebaceous glands. The accumulation of sapienic acid in sebum could be due to the degradation of linoleic acid in the sebaceous cells enabling the 6 desaturase to desaturate palmitic to sapienic acid. Consequently, palmitic acid, which is the predominant fatty acid in sebaceous cells, is not competing with linoleic acid for the same enzyme (Figure 10). An alternative explanation is that the 6 desaturation of linoleic leads to the formation of -linolenic acid, which is the first step in the production of the proinflammatory cyclooxygenase products. The catabolism of linoleic acid in human sebaceous cells would minimize the initiation of this cascade by the high amounts of 6 desaturation that takes place in human sebaceous glands as witnessed by its specific production of sapienic acid.

Figure 10.

Schematic mechanisms of the possible ways that sapienic acid is produced by sebaceous cells. Sapienic acid is synthesized by palmitic through the action of 6 desaturase, which can also utilize linoleic acid to produce -linolenic. Depletion of linoleate acts synergistically to the sapienic acid production. LA-OOH is the hydroxyperoxide from linoleic acid and LOX the lipoxygenase.

Full figure and legend (23K)

The exploration of the fatty acid metabolism in sebaceous cultures is indeed intriguing and complex. Further knowledge of the metabolism and the fate of the fatty acids in sebaceous glands may provide significant knowledge towards the regulation of their function and differentiation.

References

References

Acknowledgments

We would like to express our sincere and deepest gratitude to Dr. Magdalena Eisinger for her continuous advice and the criticism that she provided throughout this work as well for her critical review and corrections of the manuscript. My thanks are extended to Wen-Hwa Li for technical help, Dr. Steve Prouty for scientific advice, and Dr. Kurt Stenn for critical review of this document.