|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||



,
* Kao Biological Science Laboratories, Haga, Tochigi, Japan;
Department of Dermatology, Tokyo Medical University, Nishishinjyuku, Shinjyuku-ku, Tokyo, Japan;
The Skin Sciences Institute, Cincinnati Children's Hospital Research Foundation, Cincinnati, Ohio, USA; and
Department of Dermatology, University of Cincinnati College of Medicine, Ohio, USA
1Correspondence: Department of Dermatology, University of Cincinnati College of Medicine, 231 Albert Sabin Way-ML 592, Cincinnati, OH, USA 45267-0592, USA. E-mail: boissyre{at}ucmail.uc.edu
| ABSTRACT |
|---|
|
|
|---|
Key Words: race melanocytes melanization complexion
| INTRODUCTION |
|---|
|
|
|---|
Human skin color is determined by the amount and types (i.e., eumelanin and pheomelanin) of melanin synthesized in melanocytes located in the basal layer of epidermis (3)
. The melanocyte contains a unique lysosome-related organelle called the melanosome, which is specialized in melanin production (4)
. Both melanosomes and lysosomes are derived from the endosomal compartment, receive various gene products predominantly sorted from the trans Golgi network, and share features such as low cisternal pH (5)
. However, only melanosomes acquired melanin synthetic enzymes, such as tyrosinase, DOPAchrome tautomerase, and structural proteins, such as pmel-17 (GP100), which characterize melanosome biogenesis (6)
. Melanin is synthesized in the melanosome via a catalytic pathway using tyrosine as a precursor. Ultimately, pigmented melanosomes are translocated down dendrites (7)
and transferred to neighboring keratinocytes (8)
imparting complexion coloration.
Although these melanogenic mechanisms are common for all racial skin groups, skin color among races is dramatically different. African descent skin is darker primarily due to relatively larger, more pigmented melanosomes widely distributed throughout the epidermis. In contrast, Caucasian skin is lighter primarily due to smaller, more lightly pigmented melanosomes absent in the upper layers of the epidermis (9)
. This differential melanogenesis in racial skin groups has been attributed primarily to tyrosinase activity. Tyrosinase, a key enzyme in melanin synthesis, is quantitatively more abundant in melanocytes of dark skin as oppose to light skin (10
11
12
13)
. However, the precise genetic mechanisms within the melanocyte that regulate this variation in tyrosinase activity are still unknown.
Keratinocytes, the predominant cell type in the epidermis, have also been implicated in regulating skin color. Keratinocytes can secrete cytokines that stimulate facultative melanin synthesis. Cytokines secreted after UVB irradiation that enhance melanogenesis in melanocyte consist of bFGF (14)
, endothelin-1 (ET-1), (15)
,
-MSH (16)
, proopiomelanocortin (POMC), (17)
, and stem cell factor (SCF; ref. 18
). However, the effect of these cytokines on constitutive racial skin color is questionable.
A significant correlation among racial skin color is the distribution pattern of melanosomes in keratinocytes. Keratinocytes from dark (African descent) skin contain melanosomes that are predominantly distributed individually throughout their cytosol whereas keratinocytes from light (Caucasian descent) skin contain melanosomes primarily clustered into aggregates (19
, 20)
. Keratinocytes of Asian skin contain a melanosome distribution pattern that is intermediate between these two conditions (21)
. This differential distribution pattern of melanosomes in the epidermis between racial groups appears to be regulated by the keratinocyte (22)
.
In addition to the distribution pattern of recipient melanosomes, the phagocytotic activity of keratinocytes has also been associated with skin color differential. The protease-activated receptor-2 (PAR-2) is a seven transmembrane G-coupled receptor and known to regulate phagocytosis in keratinocytes (23)
. Recent data have demonstrated that dark skin exhibits higher expression of PAR-2 compared to the light skin (24)
, and inhibitors of PAR-2 lighten skin effectively (25
26
27
28)
. These observations suggest that PAR-2 mediated transfer of melanosomes may regulate skin color.
In this study, we utilized an in vivo xenograft system to examine the contribution of keratinocytes to skin complexion coloration. In this model system, we produced human skin substitutes (HSS) utilizing cultured keratinocytes and melanocytes derived from dark and light skin donors in various combinations. We had recently reported that melanocytes were sorted into the basal layer of epidermis in this HSS model (29)
. In addition, melanogenesis induced after UVB irradiation was also observed in this HSS system (29)
. Here we report for the first time that the 1) expression of melanogenic cytokines, 2) maturation of melanosomes, 3) melanin synthesis, and 4) melanosome distribution in HSS are differentially influenced by the racial origin of the keratinocyte.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Cell cultures
Primary cultures of normal human melanocytes, keratinocytes, and fibroblasts were established from Caucasian and African descent neonatal foreskin after routine circumcision at Christ Hospital and University Hospital (Cincinnati, OH, USA) with modifications as described previously (30)
. In short, the dermis and epidermis were separated by overnight treatment with dispase at 4°C. Melanocyte and keratinocyte were isolated from the epidermis by treatment with Trypsin-EDTA and cultured in complete M254 with HMGS and complete Epilife with HKGS-V2, respectively, plus penicillin, streptomycin, and amphotericin (Cascade Biologics). Fibroblasts were isolated from dermis with collagenase treatment and cultured in DMEM with 10% FBS and 1x antibiotics and antimyotics (Invitrogen). All cells were passaged at least three times with ten times dilution for each passage.
Grafting cells onto SCID mice
All mice were handled according to the guidelines of the Ethical Committee for Animal Experiments at the Cincinnati Children's Hospital Research Foundation (Cincinnati, OH, USA). Four to six-week-old SCID mice (Taconic, New York, NY, USA) were maintained under pathogen-free condition (Children's Hospital Research Foundation, Cincinnati, OH, USA). HSS were prepared in the SCID mice as described previously (29)
. In short, melanocytes (1x106), keratinocytes (6x106), and fibroblasts (6x106) were harvested using 0.1% trypsin/EDTA and neutralized with equal volume of FBS. The three cell types were mixed at a ratio of 1:6:6, respectively, excess medium was removed after centrifugation, and cells were resuspended in the DMEM with 10% FBS. Silicone chambers (Renner, Darmstadt, Germany) were sutured onto the dorsal skin of mice, and 3 h after grafting, a suspension of combined cells was added into the 3.5 mm opening on the top of the chamber. One week after implantation, the top parts of the chambers were removed to increase air exposure and a wire net was attached to prevent the HSS from external damage. The remaining chamber, including the base, was spontaneously released from the back of the mice by approximately 6 wk after implantation.
Measurement of melanin content
Melanin content was quantitated by 1) image analysis of Fontana-Masson stained histological sections, and 2) absorbance of melanin at 405 nm in dissolved epidermal sheets. For the image analysis, the HSS were surgically removed and fixed in 10% formaldehyde. After paraffin sections were prepared, melanin granules were augmented using Fontana-Masson staining with a nuclear fast red counter stain. The area of melanin granules was selected, and the amount of black pixels per epidermal length was quantitated using Photoshop CS2 (Adobe Systems, San Jose, CA, USA).
In another experiment, the HSS were surgically removed and incubated in 2 M NaBr to separate the epidermis from the dermis. The epidermis was weighed and incubated in 2 M NaOH until completely dissolved. Absorbance of melanin at 405nm was measured with Model 550 Microplate Reader (Bio-Rad, Hercules, CA, USA). Melanin standard curve was prepared using synthetic melanin (Sigma, St. Louis, MO, USA). Melanin content was normalized to the wet weight of epidermis.
Immunostaining
The epidermis was separated from the HSS after 30 min incubation in 2 M NaBr in PBS at 4°C. The isolated epidermis was fixed with 50% ammonium sulfate and used for routine immunohistochemistry. Human Tyrp1 antibody, Mel-5 (Signet Laboratories, Dedham, MA, USA), was utilized for melanocyte detection with 1:40 dilution. TRITC-labeled anti-mouse IgG was used as a secondary antibody. Images were obtained by Nikon PSM-4A Microscope with a Spot II Camera (Nikon, Tokyo, Japan).
Transmission electron microscopy
For transmission electron microscopy (TEM) analysis, HSS samples were fixed with 1/2 strength Karnovsky's fixation buffer, washed three times with 0.2 M sodium cacodylate buffer, and postfixed with 1% osmium tetroxide containing 1.5% potassium ferrocyanide. After dehydrated, tissues were embedded in Spurr's resin. Sections were obtained using a RMC-MT6000XL ultramicrotome and stained with uranyl acetate and lead citrate. Sections were viewed, and selected images were digitally photographed using a JEOL JEM-1230 transmission electron microscope.
Real-time quantitative reverse transcriptase-polymerase chain reaction
The HSS were pretreated with RNAlater (Qiagen, Valencia, CA, USA) to stabilize total RNA, and then the epidermal sheet was removed by incubation in 2 M NaBr. The total RNA from each epidermal sheet was isolated using the RNeasy microkit (Qiagen). cDNA was synthesized by reverse transcription of total RNA using the high capacity cDNA archive kit (Applied Biosystems, Foster City, CA, USA). On-demand probes for human ET-1, SCF, POMC, microphothalmia-associated transcription factor, tyrosinase, tyrosinase-related protein 1, dopachrome tautomerase, silver protein, and MART1 in Taqman Gene Expression Assays (Applied Biosystems) were used. Real-time quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) with Taqman Gene Expression Assay was performed in an ABI PRISM 7300 sequence detection system (Applied Biosystems).
Affymetrix gene chip analysis
Total RNA was isolated from Caucasian and African descent derived keratinocytes using TRIZOL (Invitrogen) and purified with RNeasy mini kit (Qiagen) for gene chip analysis. The purity of total RNA was verified using Agilent bioanalyzer and amplified using Affymetrix amplification kit. Biotin-labeled target cRNA was generated from each RNA sample, and was hybridized to Human Genome U133 Plus 2.0 chip. Data from microarray were analyzed using GeneSpring Version 7.0 (Silicon Genetics, Redwood City, CA).
Western blotting
Epidermis isolated from HSS was homogenized in T-PER tissue protein extraction reagent (Pierce, Rockford, IL, USA). Tissue lysates were separated by using 12% Ready Gel Tris-HCL gel (Bio-Rad, Hercules, CA, USA) followed by routine extraction. Monoclonal PAR-2 antibody from Invitrogen was used at 2 µg/ml concentration. Anti-mouse IgG horseradish peroxidase (HRP)-conjugated (GE Healthcare Life Sciences, Little Chalfont, Buckinghamshire, UK) was used as a secondary antibody.
Statistics
Student's t test was used for the evaluation of the melanosome number and melanosome maturation stage analysis. Mann-Whitney's U test was performed for the gene expression assays. A P value < 0.05 was considered statistically significant.
| RESULTS |
|---|
|
|
|---|
|
Total melanin content was higher in the HSS composed of dark compared to light skin-derived keratinocyte
HSS composed of dark keratinocytes and melanocytes (KDMD) exhibited significantly higher melanin content compared to the HSS composed of light keratinocyte and dark melanocytes (KLMD) assessed by both image analysis using Fontana-Masson stained histological sections (Fig. 2
A) and by spectrophotometric analysis (Fig. 2C
). Similarly, HSS composed of dark keratinocytes and light melanocytes (KDML) exhibited significantly higher melanin content compared to the HSS composed of light keratinocytes and melanocytes (KLML; Fig. 2B
). There were no differences in melanocyte density between HSS groups composed of the same origin melanocytes (i.e., KDMD vs. KLMD; Fig. 3
) or (i.e., KDML vs. KLML; not shown). Because HSS composed of melanocytes derived from dark skin combined with keratinocytes derived from either dark (KDMD) or light (KLMD) skin exhibited quantifiable higher amounts of melanin than HSS composed of melanocytes derived from light skin combined with keratinocytes derived from either dark (KDML) or light (KLML) skin, we utilized the former set (i.e., KDMD and KLMD) for further comparative analysis of the role keratinocytes in modulating skin color.
|
|
Melanosomes in melanocytes and keratinocytes were more mature in KDMD than KLMD, whereas the number of melanosomes in basal and suprabasal keratinocytes did not differ between the two conditions
To further assess the difference in melanin content between relatively darker KDMD and lighter KLMD HSS, the maturation stages and number of the melanosomes in both melanocytes and keratinocytes were analyzed by electron microscopy. Melanosomes were present in melanocytes as well as basal and suprabasal keratinocyte in HSS composed of both light (KLMD; Fig. 4
A) and dark (KDMD) (data not shown) keratinocytes. The melanocytes contained all four maturation stages of melanosomes in both KDMD (Fig. 4B
) and KLMD (Fig. 4C
). These stages consisted of nonpigmented spherical melanosomes (stage I), unpigmented oval melanosomes with an assembled filamentous matrix (stage II), intermediately pigmented melanosomes (stage III), and completely pigmented melanosomes (stage IV). The ratio of the stage IV melanosomes was significantly increased within melanocytes of KDMD compared to KLMD (Fig. 4D
), suggesting that melanogenesis was accelerated in the presence of dark derived keratinocytes. Concomitantly, the ratio of stage IV melanosomes was significantly increased in both basal (Fig. 4E
) and suprabasal (Fig. 4F
) keratinocytes in KDMD. In contrast, the total number of melanosome was not statistically different within melanocyte (data not shown), the basal keratinocyte, and the suprabasal keratinocyte between KDMD and KLMD (Fig. 4G
). PAR2 expression, involved in melanosome phagocytosis, was also investigated by Western blotting analysis. There was no difference in expression of PAR-2 between KDMD and KLMD (Fig. 5
), consistent with the above result demonstrating similar melanosome number transferred to basal and suprabasal keratinocytes (Fig. 4G
).
|
|
Melanosomes were more individually distributed or more aggregated into clusters in keratinocytes of KDMD and KLMD, respectively
The patterns of melanosomes distribution in keratinocytes of different races are distinct. Melanosomes are predominantly distributed individually or in membrane bound clusters in keratinocytes of dark or light skin, respectively (21)
. Also in vitro coculture studies have demonstrated that the keratinocyte regulates this distinct distribution pattern (22)
. In our HSS model, melanosomes were localized in recipient keratinocyte surrounding the nucleus and scattering throughout the cytoplasm in both KDMD (Fig. 6
A) and KLMD (Fig. 6B
). Individually distributed melanosomes were significantly increased in KDMD (Single:Cluster=83.4:16.6) than KLMD (Single:Cluster=80.2:19.8, P<0.05; Fig. 6C
), confirming the role of keratinocytes in regulating melanosome distribution patterns in the human skin.
|
Genetic expression of melanogenic cytokines and melanin synthesis-related proteins were higher in the HSS composed of dark skin derived keratinocytes
To determine the mechanisms of increased melanogenesis in HSS composed of dark keratinocytes, quantitative real-time RT-PCR analysis was performed. The mRNA expression of ET-1 and proopiomelanocortin POMC was significantly increased in the HSS composed of dark keratinocytes (KDMD) compared to HSS composed of light keratinocytes (KLMD; Fig. 7
A). In contrast, isolated keratinocytes derived from light skin donors exhibited significantly higher expression of ET-1 than isolated keratinocytes derived from dark skin donors (Fig. 8
). SCF mRNA expression was significantly different in the two HSS groups but did exhibit a higher tendency in KDMD. The mRNA expression of microphothalmia-associated transcription factor, up-regulated by ET-1 and SCF (31)
, was significantly increased in KDMD HSS (Fig. 7A
), as was its downstream enzyme tyrosinase (Fig. 7A
). The premelanosome components pmel-17 (silver/GP100) and MART1 were also significantly increased in KDMD (Fig. 7A
). Interestingly, similar tendency, albeit not statistically significant, was also observed in the most of gene expressions in the HSS group composed of light melanocytes (Fig. 7B
).
|
|
| DISCUSSION |
|---|
|
|
|---|
Our results specifically demonstrated that dark skin (African descent) derived keratinocytes, in comparison to light skin (Caucasian descent) keratinocytes, promote darker skin color by increasing the melanin content in the HSS. Concomitantly, melanosome maturation, as determined by ratio of stage IV melanosome assessed by electron microscopy, and therefore melanogenesis was increased in HSS composed of dark keratinocytes as oppose to light keratinocytes. In contrast, the total number of melanosomes, both synthesized by the melanocyte and transferred to and/or residing in the keratinocyte, was not influenced by the type of keratinocyte used to produce the HSS. Alternatively, these results could suggest that light skin derived keratinocytes suppress melanogenesis relatively.
It has been reported that the PAR-2 is a keratinocyte receptor that regulates keratinocyte phagocytosis (23)
. Expression of PAR-2, as well as its enzymatic activation, is enhanced in darker skins (24)
and the amount of PAR-2 can be induced by UVB exposure (32)
. These results suggest that PAR-2 can regulate skin coloration in part. However, we did not detect any difference in PAR-2 protein expression between our various HSS constructs nor did we demonstrate differences in number of melanosome transferred to keratinocytes in HSS composed of dark vs. light skin donor-derived keratinocytes. This indicates that PAR-2 may not be a factor in regulating the differential constitutive pigmentation of our short-term HSS model.
We demonstrate in this study that the genetic expression for tyrosinase was elevated in HSS composed of KD as oppose to KL (Fig. 7)
. In contrast, it has previously been demonstrated that the abundance of tyrosinase does not differ between cultures of melanocytes developed from Black or Caucasian individuals (33)
. These contrasting results suggest that racially derived keratinocytes from dark or light skin can either increase or decrease, respectively, the relative expression of melanogenic genes resulting in part in the respective differential skin color.
To further understand the mechanisms underlying enhanced melanin synthesis in the HSS composed of dark keratinocytes, we demonstrated quantitative differences in the expression of specific melanogenic cytokines between our various HSS constructs. In the HSS composed of dark keratinocytes, the expressions of ET-1 and POMC were significantly higher as opposed to light keratinocytes. In contrast, ET-1 mRNA expression levels in isolated, cultured keratinocytes was decreased in dark derived as oppose to light derived neonatal foreskin. Our preliminary data suggested that ET-1 mRNA expression was higher in the epidermis of dark foreskin compared to light foreskin (data not shown). These findings suggested a possible interaction between keratinocytes and melanocytes to stimulate the expression of melanogenic cytokine in keratinocytes, resulting in increased melanogenesis in HSS composed of dark skin-derived keratinocytes. Both ET-1 and
-MSH, a derivative of POMC, are cytokines secreted by keratinocytes after UVB irradiation and involved in the subsequent facultative melanogenesis. It has been reported that ET-1 is secreted from keratinocytes after UVB irradiation and induces melanogenesis in melanocytes via a PKC pathway (15
, 34)
. The production of
-MSH has been also known to be induced by UVB irradiation in both melanocytes and keratinocytes (17
, 35)
and to increase melanogenesis via cAMP, followed by the up-regulation of tyrosinase and Tyrp1 (36
, 37)
. Differences in the quantitative expression of ET-1 and/or
-MSH between racial groups have not been reported. However, differences in the blood plasma expression levels for ET-1 have been reported to be higher in African descent people who have a higher prevalence of hypertension and other cardiovascular disease (38
, 39)
. In addition, the levels of
-MSH in plasma have been reported to be greater in more pigmented individuals than less pigmented individuals (40)
. These findings indicate that differential regulation of these cytokines exists among the human race in general and possibly in the skin specifically to putatively regulate constitutive complexion coloration.
It has also been demonstrated that dermal fibroblasts may play a role in the constitutive skin color (41
, 42)
. Skin substitutes composed of normal melanocytes and keratinocytes originating from the same donors (of phototype II–VI) seeded on dead de-epidermized dermis (DDD) demonstrated that DDD influenced macroscopic pigmentation, epidermal morphology, melanin distribution, and melanocyte basal location but without any correlation to their original donor phototype, suggesting a role of fibroblast secreted soluble factors in influencing skin color variation (42)
. In addition, in lesional skin of systemic sclerosis, SCF expression in the dermal fibroblasts was stimulated, resulting in the increase in mast cell number and epidermal pigmentation (43
, 44)
. The enhanced expression of SCF and hepatocyte growth factor, another melanocyte stimulating cytokine, in dermal fibroblasts was demonstrated in pigmented neurofibromatosis type 1 lesions characterized by epidermal hyperpigmentation (45)
. These studies demonstrate that dermal fibroblast may regulate skin pigmentation in part and warrant further analysis.
The pattern of melanosome distribution within the recipient keratinocytes also contributes to the differential skin coloration between dark and light skin. It has been demonstrated that the ratio of aggregated vs. individual melanosomes in keratinocytes differs among races (21)
and that the keratinocyte itself governs this differential melanosome distribution pattern (22)
. In agreement with these previous studies, our data also demonstrated that the ratio of individual melanosomes was significantly higher, albeit 4%, in HSS composed of dark keratinocytes as oppose to light keratinocytes, confirming the role of keratinocyte in its melanosome distribution pattern. Complexion coloration is regulated by 1) type and amount of melanin synthesized by the melanocytes, 2) distribution pattern of the melanosomes within the recipient keratinocyte, and 3) degradation of melanosomes by the keratinocytes. Therefore, the distribution pattern may only represent a minor component of this total, which would be consistent with the small difference we identified in the distribution pattern between HSS utilizing dark vs. light keratinocytes as presented in Fig. 6
. Currently unknown are the cellular/molecular mechanisms utilized by the keratinocyte to govern whether the recipient melanosomes are packed into clusters or retained as individual organelles.
It would be very interesting to utilize this HSS model for the functional analysis of cutaneous cells derived from lesions of various pigmentary disorders. Cooper et al. (46)
utilized a conventional HSS model by overlaying keratinocytes on a collagen-glycosaminoglycan matrix containing fibroblast for the analysis of human giant congenital nevomelanocytic nevi (GCNN). A GCNN lesion was recapitulated in the HSS composed of the cells derived from nevi, indicating that the lesional cells maintained, after the isolation, their pathological characteristics. In contrast, Bessou et al. (47)
constructed in vitro skin equivalence using cells derived from the lesional site of vitiligo and demonstrated no difference in pigmentation between HSS composed of vitiligo-derived cells and normal/perilesional skin derived cells, supporting the hypothesis that vitiligo requires an extracellular triggering factor to elicite melanocyte death (48
49
50)
.
Our in vivo HSS model demonstrates for the first time that the number of melanocyte sorted into the basal layer was consistent regardless of the different racial skin derived keratinocytes utilized. This observation suggests that the ratio of keratinocytes to melanocytes in human skin was intrinsically determined by the type of melanocytes. This distinct ratio of mammalian epidermal cells within the skin uniforms to a geometrical array corresponding to the phi proportionality (51)
. In hyperproliferative skin diseases, this ratio is altered as the epidermal turnover rate is increased. It would be interesting to compose HSS using cutaneous cells from hyperproliferative diseases to investigate the intracellular and extracellular communication between various epidermal cells. Our in vivo HSS technique also enables us to maintain the skin structure for 6 months, allowing sufficient time to assess mechanisms underlying long-term skin diseases, such as melanoma. In addition, this HSS model can be developed to utilize epidermal cells containing transfected genes for the analysis of their functions in cutaneous homeostasis.
In conclusion, our results demonstrate that the HSS model is an attractive and promising melanogenic system for elucidating the functions in each cutaneous cell type. More specifically, we demonstrated that keratinocytes, in addition to melanocytes, play a significant role in skin color determination by producing cytokines involved in the melanogenesis as well as regulating the distribution pattern of melanosomes in the epidermis. These findings provide a new insight into the complex cellular and molecular processes in the skin underlying racial skin color differences.
Received for publication August 2, 2006. Accepted for publication March 8, 2007.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
A. Hachiya, P. Sriwiriyanont, T. Fujimura, A. Ohuchi, T. Kitahara, Y. Takema, W. J. Kitzmiller, M. O. Visscher, R. Tsuboi, and R. E. Boissy Mechanistic Effects of Long-Term Ultraviolet B Irradiation Induce Epidermal and Dermal Changes in Human Skin Xenografts Am. J. Pathol., February 1, 2009; 174(2): 401 - 413. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Ni-Komatsu, C. Tong, G. Chen, N. Brindzei, and S. J. Orlow Identification of Quinolines that Inhibit Melanogenesis by Altering Tyrosinase Family Trafficking Mol. Pharmacol., December 1, 2008; 74(6): 1576 - 1586. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |