The pilosebaceous duct is lined by a stratified, squamous epithelium, and consists of keratinocytes that undergo a similar, but distinct, terminal differentiation process compared to the keratinocytes that comprise the interfollicular epidermis. The upper fifth of the duct is identical to the interfollicular portion; however, the remaining portion, the infrainfundibulum, is histologically different. In the infrain- fundibulum, the granular layer is frequently absent and the keratinocytes contain a tremendous amount of glycogen. In addition, the stratum corneum is only two to three layers thick and will desquamate into the central part of the pilosebaceous duct where mixing with sebum and bacteria occurs (1).
Ductal hypercornification results initially in the formation of microcome- dones, eventually leading to the production of a comedonal lesion resulting from the accumulation of corneocytes in the pilosebaceous duct. This is believed to be the result of enhanced keratinocyte proliferation, enhanced desquamation, inadequate squame separation, or a combination of factors thereof. Individuals with acne have more cornified material in their ducts when compared with control subjects (60).
Organ Culture of the Infrainfundibulum
Keratinocyte cell culture of the interfollicular epidermis is well-established, and procedures are available to expand cell populations in serum containing and serum-free medium in the presence and absence of fibroblast-feeder support (61 –
63). In addition, three-dimensional models are also well-characterized to simulate a living skin equivalent. Using these models, it is possible to monitor the cultures
to examine the effects on keratinocyte proliferation and differentiation using a variety of end-point markers.
In contrast, the culture of the infrainfundibulum is less established and only one laboratory has reported on the use of the infundibula as a model for acne (64 – 66). Using a procedure similar to that used for isolating sebaceous glands, pilosebaceous ducts can be isolated by micro-dissection of keratomed layers of skin. Interestingly, it was found that only subjects with high rates of sebum secretion contained pilosebaceous ducts large enough for positive identification and isolation.
Isolated infundibula have been successfully cultured in serum-free medium containing essential growth factors for up to seven days (66). During this culture period, morphology of the duct is retained, although desquamated corneocytes are not eliminated via the follicular canal (Fig. 8). Rather, the keratinocytes/corneo- cytes accumulate in the lumen, whereas the keratinocytes express markers consist- ent with those found in the basal and suprabasal layers.
In the presence of 1 ng/mL of IL-1a, the infundibula were found to hypercor- nify resulting in a darkly stained appearance (Fig. 8). The resulting scaling was his- tologically similar to the changes seen in a comedonal lesion in vivo and addition of an excessive amount of IL-1a receptor blocker prevented this hypercornification process from occurring. This is of note, as IL-1a is found in high concentrations in open comedones (67 – 69). The addition of 5 ng/mL EGF or transforming
FIGURE 8 Cross-sections of infundibula isolated and maintained in vitro. Infundibula from human pilosebaceous follicles can be isolated and cultured in vitro with retention of viability and architecture as a model for acne. (A) Freshly isolated infundibulum depicting stratified squamous epithelium with dispersed fibroblasts in dermal portion. (B) Infundibulum maintained in vitro for seven days depicting visible stratum granulosum and stratum corneum in the central regions. (C) Infundibulum maintained in vitro for seven days in the presence of 1 ng/mL Il-1a depicting scaling phenotype in central lumen. (D) Infundibulum maintained in vitro for seven days in the presence of 5 ng/mL epidermal growth factor depicting loss of architecture. Source: From Ref. 66.
growth factor-a resulted in a disorganized appearance of infundibular morphology, although individual cellular integrity was retained. Looking at the retinoids, 13-cis- retinoic acid (1 mM) supplemented into the medium resulted in a significant fall in the rate of DNA synthesis after seven days maintenance which can be interpreted as an inhibition of new ductal cell growth.
In a subsequent report, the effects of inflammatory cytokines on the infundi- bulum were evaluated. The addition of 100 U of interferon-g, 10 ng/mL of tumor necrosis factor-a or 10 ng/mL IL-6 resulted in the expression of intercellular adhesion molecule-1 which was not seen with IL-1a treatment.
Although the organ culture of the infundibulum suffers from the same con- straints as sebaceous gland organ culture, the advantages over keratinocyte mono- layer culture lend support as an additional model for the understanding and treatment of acne.
As previously mentioned, keratinocyte cell culture procedures are well- established and it is beyond the scope of this chapter to disclose the numerous refer- ences related to epidermal biology. However, the use of keratinocyte monolayer models in further understanding acne is reported in subsequent areas of this review.
EVALUATION OF TECHNOLOGIES TARGETING ENZYMES INVOLVED IN ANDROGEN METABOLISM AND LIPID SYNTHESIS
Enzyme Assays for In Vitro Androgen Metabolism
Human skin is a collection of androgen-responsive tissues, which have been recog- nized to contain a wide variety of enzymes capable of metabolizing topically applied drugs and endogenous substrates supplied by the body.
One of the most important classes of hormones affecting the pilosebaceous appendage, and more specifically the sebaceous glands, is the sex steroids secreted by the testes and ovaries in response to pituitary gonadotrophic hormones (70). In general, androgens stimulate sebaceous gland activity, whereas estrogens have a suppressive function (70,71). Androgens are the best-known stimulators of the sebaceous glands, and are responsible for the development and enlargement of the glands that occur at puberty (72). Castration results in sebaceous gland atrophy accompanied by an overall decrease in sebum production (73). The most effective androgens possess a 17b-hydroxyl group that includes testosterone,
5a-DHT, 5a-androstene-3b, and 17b-diol.
The enzymes involved in androgen metabolism in follicles that have been identified using biochemical and histological techniques include 3b-hydroxysteroid dehydrogenase (3b-HSD), 17b-HSD, and 5a-reductase. A schematic diagram of the pathways of cutaneous androgen metabolism and the converting enzymes involved has been effectively conveyed by Chen et al. (74) in Figure 9.
The reduction of testosterone to DHT by the enzyme 5a-reductase is one of the most important reactions involved in androgen metabolism. DHT is a more potent androgen than testosterone due to its greater affinity for the androgen recep- tor and enhanced stability. DHT is generally considered responsible for increased sebum production, as increased local formation of DHT has been documented in acne (75) and by increasing the size of the sebaceous gland (73). Two subtypes of
5a-reductase have been identified.
The Type I isozyme has an optimal pH of 6 to 9 and has been found localized in skin from the scalp, chest, dermal papilla, and sebaceous gland regions (76).
FIGURE 9 Pathways of cutaneous androgen metabolism and the converting enzymes. Abbreviations: DHEAS, dehydroepiandrosterone sulfate; 17b-HSD, 17b-hydroxysteroid dehydrogenase. Source: From Ref. 74.
Activity was found to be greater in acne-bearing facial skin and in sebaceous glands than in other cells of similar location. Activity was also higher in glands in the facial region than glands in other bodily regions. In a subsequent study (77), keratinocytes cultured from the infrainfundibulum demonstrated greater 5a-reductase activity than those obtained from the interfollicular epidermis. Overall activity was highest in the sebaceous gland followed by the duct, infrainfundibulum, and lastly the epidermis. Acne-bearing skin was found to produce 2 to 20 times more DHT than normal skin, and facial skin produced more DHT than normal back skin.
The Type II isozyme has an acidic pH optimum and is present in the epididymis, seminal vesicles, prostate, and the inner root sheath of the terminal hair follicle (78).
Specific target assays for 5a-reductase are problematic in nature as the enzyme is difficult to extract and the substrate testosterone is water insoluble. Nonetheless, several assays are available to monitor androgen metabolism in cell/tissue homogenates or in living cell/tissue culture systems. The enzymatic assay for conversion of testosterone to DHT involves adding testosterone (usually radiolabeled) together with a cell or tissue homogenate in the presence of NADPH. Inhibitors can be added and after a set period of time, the steroids formed are isolated and separated either by TLC, HPLC, or GC (79).
Complementary to using skin homogenates, activity of enzymes involved in androgen modulation can be assessed using cell monolayer cultures and living skin equivalents. Although keratinocyte methods for cultivation of each are well- established, cell culture may be of limited use for evaluating water-insoluble com- pounds or formulations. For the latter, the use of a reconstituted three-dimensional living skin model has been reported by Bernard et al. (80) and Slivka (81).
Harris et al. (82) reported on the expression of 5a-reductase in COS cells, whereas Sugimoto et al. (83) constructed 5a-reductase expression vectors that were transfected into a human adrenal carcinoma cell line. Type I 5a-reductase was strongly inhibited by the chloride salts of cadmium, copper, and zinc (inhi- bition constant or Ki less than 5 mM), moderately by nickel and iron (Ki less than
250 mM), and no inhibition by manganese, magnesium, calcium, or lithium. Only the Type II isozyme was inhibited by copper at 11 mM. Additionally, minimal effects were seen when the chloride counter ion of zinc was replaced with sulfate, chloride, or acetate, with acetate being the least effective.
In general, 5a-reductase inhibitors are classified into two categories: steroidal and nonsteroidal, and there is very little overlap in efficacy between the two different isozymes. In fact, many of the Type II inhibitors have little to no effect on sebum pro- duction in vivo. This is possibly the result because the two enzymes have only 50% homology in amino acid sequences. The gene for Type I isozyme is located on chromo- some 5, whereas the gene for Type II isozyme is located on chromosome 2 (84).
Using foreskin keratinocytes (85), testosterone was utilized as a substrate in the production of 5a-reduced metabolites including DHT (5a-reductase), androste- nedione (17b-HSD Type II), androsterone, and androstanediol (3a-HSD activity). A similar pattern was reported with fibroblasts, although scalp skin fibroblasts metab- olized testosterone only to a limited extent. Using cultured keratinocytes obtained from the infrainfundibulum and epidermis (86), Thiboutot et al. found that mean enzyme activities were slightly higher in the acne group, but not statistically differ- ent from the control group. Overall activity was greatest in the infrainfundibulum.
Altenburger and Kissel (87) cultured the human keratinocyte cell line HaCaT as a model to express the enzyme systems involved in testosterone metabolism, which were subsequently identified by LC/MS. HaCaT cells were found to predo- minantly express the Subtype I isoform of 5a-reductase. Similar studies have been reported utilizing keratinocyte cultures derived from breast skin (88), in cultured fibroblasts (89), and living skin equivalents (90).
Assays for the Rate-Limiting Enzymes Involved in Cholesterol and Fatty Acid Synthesis
In the formation of squalene and cholesterol, 3-hydroxy-3-methylglutaryl- coenzyme A (HMG-CoA) is the rate-limiting enzyme involved in the cholesterol synthesis pathway. As reported by Shapiro et al. (91) and Smythe et al. (92), radio- labeled HMG-CoA is added to skin homogenates containing NADP (as a necessary cofactor) along with glucose-6-phosphate and glucose-6-phosphate dehydrogenase to form radiolabeled mevalonate. The reaction then involves a spontaneous conver- sion of mevalonate to mevalonolactone, which can then be separated by TLC and quantified.
Using this assay with isolated sebaceous gland homogenates, Smythe et al. (92) reported that HMG-CoA activity decreased with subject age, and enzyme activity was inactivated via phosphorylation.
In the study of fatty acid synthesis, acetyl-coenzyme A carboxylase (acetyl- CoA) has been identified as a rate-limiting step. This assay involves adding radio- labeled bicarbonate to skin homogenates in the presence of ATP and acetyl-CoA. Total radioactivity incorporated is then quantified via direct scintillation counting.
All of the enzyme assays described are amenable for high throughput screen- ing to assess compounds that may modulate lipid synthesis.