| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Address correspondence to Michael B. Morgan, M.D., 16124 Chastain Road, Odessa, FL 33556, USA; tel 813 971 0775; fax 813 971 6675; e-mail mbkmmorgan{at}aol.com.
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionReferences |
|---|
(received 1 July 2002; accepted 17 November 2002)
Keywords: alopecia, baldness, bcl-2, testosterone, TUNEL, apoptosis, hair follicle
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionReferences |
|---|
In 1997, Sawaya and Price [2] demonstrated higher levels of androgen receptors in frontal scalp follicles than in occipital follicles, providing a micro-anatomical explanation for the "classic" AGA pattern of frontal follicular miniaturization with occipital follicular preservation. These authors also noted regional differences in the expression of androgen converting enzymes 5-
reductase I and II, and cytochrome P-450 aromatase on the scalp.
The classic pattern of AGA, while easily recognized and familiar to clinicians, has a few animal analogues: some chimpanzees and red uacaris and, chiefly, the stump-tailed macaque [1]. The latter primate shows a clinically distinct androgen-induced alopecia in both sexes, histologically resembling that seen in humans. Owing to limitations of the macaque as an experimental animal, researchers have avoided macaques in favor of the more easily handled murine model [1,3].
Cycling of hair follicles in the mouse has been intensively investigated from the histological and molecular viewpoints [3,4]. It is clear that the mammalian hair follicle is an extremely complex organ that undergoes periods of growth (anagen), involution (catagen), and quiescence (telogen). Mediators suspected of dictating the cyclic action of the follicles in the murine model include dozens of biochemicals that have been studied in other cell signaling pathways, including those involved with the initiation of apoptosis or programmed cell death [3–12]. In 1997, Lindner et al [4] mapped the expression of biochemical mediators throughout the follicular cycle of mice, demonstrating the complex association of biochemical signaling with follicular phase changes.
In this study, we investigated apoptosis using 2 different staining techniques (ie, bcl-2, an inhibitor of apoptosis, and TUNEL, an indicator of apoptosis) (a) to see if the observations in the murine model hold true in human histologic sections, and (b) to determine whether there is a significant staining difference between the regions of the scalp that are affected and unaffected by androgenetic alopecia.
Scalp biopsies (4 mm punch) were obtained during 2001–2002 from 16 men undergoing hair transplantation (age 32–69 yr, mean 46 yr). The subjects manifested male pattern baldness classified as stage II–IV (4 at Stage II, 6 at Stage III, and 6 at stage IV) according to the Norwood classification scheme [5]. There were no indications of dermatological or systemic diseases that might have contributed to the development of alopecia in these subjects.
Two biopsies were obtained from each subject, one from the frontal scalp at the apex of a coronal plane through the ear tragi, and the other from the normally appearing occipital scalp. The biopsies specimens were processed in the usual manner and embedded in paraffin after being cut both longitudinally and in cross-section. At least 25 sections, cut at 5 µm intervals, were mounted on glass slides in numerical sequence. Every fifth slide was stained with hematoxylin and eosin (H&E); the remaining slides were saved for immunohistochemical studies. Unstained sections were selected to demonstrate the dermal papilla (DP) for bcl-2 immunohistochemical staining and the infundibulum for TUNEL staining (as described below). The identical procedure was performed on 3 human volunteers (men, ages 36, 47, and 52 yr) with no alopecia, who served as controls.
The cases were assigned random numbers and independently evaluated by a single pathologist (MBM) who was blinded in respect to subject identity and anatomic site. Total numbers of follicles were enumerated, including the miniaturized type and the terminal type, as well as the percentages of anagen, catagen, and telogen types. A ratio of such types within each region was calculated for each patient. The number of cells in the dermal papillae that stained positively for bcl-2 was quantified at 400 x magnification. TUNEL staining was evaluated qualitatively (present or absent) with attention to microanatomical location. Staining indices (ie, the numbers of cells that stained positively for bcl-2 and the numbers of nuclei with TUNEL, divided by the total number of cells or nuclei visible) were calculated for each visualized follicle.
The individual follicle indices were averaged to arrive at a single frontal staining index (FSI) and occipital staining index (OSI) for each subject studied. Twelve follicles were examined at high power magnification (400 x) for each sample, with 20 nuclei or cells analyzed per follicle. The F-test was used to assess the variances between the two data sets; the significance of differences between means was computed by the paired-sample t-test.
The TUNEL procedure was performed using the "ApopTag Plus In Situ Apoptosis Detection Kit" (Oncor, Inc., Gaithersburg, MD), according to the manufacturer’s instructions with the following amendments: (a) protein digestion was performed with the Oncor protein digesting enzyme at room temperature for 10 min; (b) the DAB reagent was supplied by Sigma Chemical Co. (St. Louis, MO); and (c) counterstaining was performed with Mayer’s hematoxylin rather than methyl green. Positive controls consisted of tonsillar tissue; negative controls consisted of the identical tissue with omission of the secondary antibody from the standard staining procedure.
For bcl-2 staining, microwave antigen retrieval was performed by treatment in citrate buffer (pH 6.0) for 20 min. Bcl-2 antibody (Ventana, Tucson, AZ, prediluted) was used with an automated stainer (Ventana, Inc., Tucson, AZ), with hematoxylin counterstaining. Positive and negative external controls consisted of tonsillar tissue, as noted above. Positive internal controls consisted of eccrine ductal epithelium and dermal lymphocytes, both of which typically stain positively with bcl-2 antibody.
|
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionReferences |
|---|
The quality of the bcl-2 stain did not differ from that seen in patient biopsies and controls, each showing cytoplasmic immunoreactivity. The quality of the bcl-2 staining was similar among miniaturized and terminal follicles, and showed qualitatively less intense staining in the catagen and telogen follicles. Bcl-2 staining was demonstrated in all anagen follicles examined. The average anagen staining index for the frontal scalp (FSI) was 67%, consisting mostly of miniaturized follicles. The average anagen staining index for the occipital scalp (OSI) was 42%, consisting mostly of terminal follicles (Table 1
).
|
). The OSI for the control group was 33%, 43%, and 48 (average = 43%, Table 1; Fig. 1, panels A and B). The F statistic of the patients’ FSI versus their OSI was 1.21, indicating no significant differences in the variances of the 2 groups of biopsy staining patterns. In patients with Norwood classification II–IV the F statistics within frontal and occipital scalp were 1.12 and 1.36 respectively, indicating no significant differences in the variances of the 2 groups. The paired t-test revealed a t statistic of 2.20 (p <0.05). FSI results revealed a non-skewed distribution of indices in the studied subjects. The OSI results for bcl-2 staining were negatively skewed, indicating that the majority of follicles from the posterior scalp displayed a smaller proportion of staining cells, compared to the OSI mean. Dermal eccrine duct epithelium and lymphocytes stained with the bcl-2 antibody in all slides examined.
|
The average anagen staining index (ie, the average percentage of cells stained for TUNEL per follicle) for the FSI was 5.7% and for the OSI was 10.2% (Table 1; Fig. 1, panels C and D). For TUNEL staining in controls, the FSI was 9.8% and the OSI was 10.1%. The F statistic of the FSI versus the OSI was 1.01, and within all Norwood groups 1.41 and 1.59 respectively. The paired t-test revealed a t statistic of 3.01 (p < 0.05) for TUNEL FSI vs OSI.
|
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionReferences |
|---|
One possible explanation is that the vellus follicles identified in the frontal biopsies may represent a relatively resistant population that has not yet been "earmarked" for involution. This "survival" hypothesis suggests that these few remaining follicles are either equipped to resist involution, or have not yet been irresistibly influenced by biochemical signals that would eventually result in their disappearance. Thus, they may represent a relatively pure population of follicles (at least temporarily) resistant to apoptosis.
That the frontal follicles may represent a "pure" population is further suggested by the evenly distributed FSI values. This distribution indicates that the majority of frontal dermal papillae manifest a staining index clustered around the mean, with relatively few papillae exhibiting markedly hypo- or hyper-active staining. This finding is particularly interesting since the distribution of occipital follicle staining indices is unevenly distributed. This negatively skewed OSI distribution indicates that the great majority of occipital dermal papillae manifest a decreased staining pattern for bcl-2, with a minority exhibiting relatively hyperactive staining. This skewed pattern suggests either an asynchronous cycling follicular population, or two (or more) populations of follicles, each demonstrating varying degrees of bcl-2 staining. Observations in clinical and animal models favor the former hypothesis.
Follicular cycling involves a complex and coordinated interaction between the bulge stem cell region, the dermal papilla, and the matrix cells. It has been postulated that the dermal papilla orchestrates stem cell activation "upstream" to the events of terminal differentiation and apoptosis in the bulge and matrix areas [13,14]. In this model, the papilla, as master controller of an apoptosis-mediated follicular cycle, could theoretically be protected from the flurry of nearby follicular involution by bcl-2 expression.
Our results show that the degree of expression of bcl-2 in frontal follicles relative to occipital follicles is increased in individuals with clinically evident AGA. As our study did not include a large number of biopsies from non-AGA affected individuals, it is unclear what level of bcl-2 expression would be considered "normal" in human frontal follicles. Whether the presence of bcl-2 in frontal follicles of patients with AGA is a pathological expression or overexpression is unanswered. If subsequent studies reveal a pathological overexpression in frontal AGA follicles, one could speculate that persistently elevated bcl-2 expression could lead to immortalization of the dermal papilla. This could potentially exhaust the follicular stem-cell pool, resulting in progressively smaller follicles.
Our study set is too small to reach meaningful associations between bcl-2 staining indices and follicular stage, although others have attempted this in mouse studies [4,9]. When bcl-2 overexpression was studied in outer root sheath keratinocytes in transgenic mice treated with cyclophosphamide, increased catagen induction was observed [15,16]. Specific catagen-promoting factors were not identified, however. As one of the regulators of the hair cycle [3] and a promoter of catagen, FGF-5 was suggested, though not proven [16].
The results obtained with TUNEL staining were anatomically different with a higher rate of apoptosis obtained in the occipital region. Given the antithetical relationship between bcl-2 expression and apoptosis as assessed by the TUNEL method, it would be expected that the apoptotic rate would be higher in the occipital region and lend credence to the bcl-2 findings. The intranuclear staining observed in the bulge-isthmus region of a follicle (Fig. 1, panel D) correlates well with one of the four TUNEL-associated apoptotic "hot spots" described in the murine model [4].
In summary, our findings suggest that the propensity to inherit AGA may represent increased susceptibility to mediators of bcl-2 expression or qualitatively abnormal bcl-2 responsive elements (receptors) within the DP of the alopeciaprone frontal areas of the scalp. Further studies of AGA are needed to assess the expression of IGF-1, BAD, FGF-5, TGF-ß, and possibly other genes in the bcl-2 family that have been implicated in murine follicular cycling. We plan future investigations that will focus on the dermal papilla, its important role as a coordinator of hair follicle cycling, and the alterations responsible for androgenetic alopecia.
|
References |
|---|
|
TopAbstractIntroductionResultsDiscussionReferences |
|---|
reductase type I and II, aromatase, and androgen receptor in hair follicles of women and men with androgenetic alopecia. J Invest Dermatol 1997;109:296–300.[Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
