Mechanisms of Hair Graying: Incomplete Melanocyte Stem Cell Maintenance in the Niche

See allHide authors and affiliations

Science  04 Feb 2005:
Vol. 307, Issue 5710, pp. 720-724
DOI: 10.1126/science.1099593


Hair graying is the most obvious sign of aging in humans, yet its mechanism is largely unknown. Here, we used melanocyte-tagged transgenic mice and aging human hair follicles to demonstrate that hair graying is caused by defective self-maintenance of melanocyte stem cells. This process is accelerated dramatically with Bcl2 deficiency, which causes selective apoptosis of melanocyte stem cells, but not of differentiated melanocytes, within the niche at their entry into the dormant state. Furthermore, physiologic aging of melanocyte stem cells was associated with ectopic pigmentation or differentiation within the niche, a process accelerated by mutation of the melanocyte master transcriptional regulator Mitf.

Qualitative and quantitative changes in stem and progenitor cells have been implicated in physiological (chronological) aging (1, 2), although the changes are poorly understood and the process of stem-cell aging has not been visually observed. Involvement of stem and progenitor cells in aging of multiple organ systems has been suggested in mice defective in DNA damage repair and telomere maintenance (3), but melanocytes may be unique in that the oxidative chemistry of melanin biosynthesis can be cytotoxic (4). This led to the suggestion that differentiated, pigmented melanocytes (rather than their unpigmented progenitors) are specifically targeted in hair graying (5, 6). The recent discovery of unpigmented melanocyte stem cells, distinctly located within the hair follicle (7), creates an opportunity to determine whether the process of hair graying arises specifically from changes in differentiated melanocytes or the stem-cell pool that provides them.

Stem cells are maintained in the niche microenvironment (8). Hair follicles contain a well-demarcated structure for the stem-cell niche (within the lower permanent portion), whereas differentiated melanocytes reside in the hair bulb (at the base of the transient portion of the hair follicle) (Fig. 1A) (7, 9). Hair follicles are constantly renewing, with alternating phases of growth (anagen), regression (catagen), and rest (telogen) (fig. S1). Taking advantage of the spatial segregation of the stem versus differentiated cell compartments (7), we used melanocyte-targeted (Dct) lacZ transgenic mice (7, 10, 11) to examine the impact of aging on these melanocyte compartments.

Fig. 1.

Differentiated melanocytes are lost in the hair bulb of Bcl-2 deficient mice. (A) Hair follicle structure. Melanocyte stem cells (blue dots) are in the lower permanent portion (light blue): the bulge (Bg) area in pelage follicles and the lower enlargement (LE) in whisker follicles. APM, arrector pili muscle (Figure 1A legend in SOM). (B) Appearance of Bcl2–/– mouse at P58. (C) Hair graying of whiskers in Bcl2–/– mouse at P39. (D and E) Distribution of lacZ+ cells (melanocytes) in P39 Bcl2–/+ and Bcl2–/– mice carrying the Dct-lacZ transgene. Pigmented melanocytes in the bulb (Bb) [arrows in (D)] and lacZ+ melanoblasts in the Bg [arrowhead in (D); the inset shows the magnified view] are completely lost in Bcl2–/– follicles (E). Double arrows indicate the level of Bb or Bg. Magnification is 100×. (F to I) Whole-mount lacZ staining of the bulb of whisker follicles from Bcl2–/+ and Bcl2–/– in black and white (albino: Tyrc-2j/c-2j) backgrounds at P40. Loss of Dct-lacZ+ melanocytes was detected in the bulb of Bcl2–/– whisker follicles regardless of albino background.

Among hair graying models, the melanocyte lineage in Bcl2–/– (12) and Mitfvit/vit (13) mice show relatively selective hair graying compared with mouse models of syndromic premature aging, which affect numerous cell lineages (14, 15). Hair graying in the Bcl2–/– background has been suggested to arise by chemical cytotoxity of melanin synthesis (5, 12, 16, 17). Distribution and morphology of melanoblasts among Bcl2–/–, Bcl2–/+, and Bcl2+/+ mice were normal during early development (fig. S2, a to h). Bcl2–/– mice gray after the first hair molting (Fig. 1, B and C) with white hairs. Histologically, differentiated melanocytes were almost completely absent in Bcl2–/– pelage (body hair) or whisker follicles (Fig. 1, E and G) compared with Bcl2–/+ (Fig. 1, D, F, and H) or Bcl2+/+ (18) follicles at postnatal day 39 (P39). Albino background did not protect against melanocyte loss in Bcl2–/– mice (Fig. 1I), suggesting that melanin synthesis is unnecessary for this melanocyte disappearance. In addition, Bcl2–/– follicles in the second hair cycle lack both differentiated melanocytes in the hair bulb and undifferentiated Dct-lacZ+ melanoblasts in the stem-cell niche (located at the bulge area in pelage follicles) (Fig. 1, D and E, and fig. S2, i and j), suggesting that Bcl2 might be important for survival of melanocyte stem cells.

Looking earlier at P6.5, when hair follicle morphogenesis is almost complete, Bcl2–/– follicles appear normal (Fig. 2B). In contrast, Bcl2–/– follicles at P8.5 showed sudden, nearly complete loss of melanoblasts in the niche (bulge area, Fig. 2D), whereas the number of melanocytes in the hair bulb did not show significant differences between Bcl2–/+ and Bcl2–/– mice (Fig. 2E). In both pelage and whisker follicles from Bcl2–/– animals, disappearance of niche melanoblasts begins at stage 6 of hair follicle morphogenesis [standardized hair follicle stages based on (19)], and by stage 8 they are gone (fig. S2, k to n). At this stage, niche melanoblasts undergo a morphologic change from a dendritic shape into a slender, oval shape with shrinkage to maximal nuclear/cytoplasmic ratio upon entry into the dormant state (Fig. 2, F and G). This change in morphology was seen cyclically at corresponding stages of subsequent cycles (18). Apoptosis of melanocyte stem cells was observed at the same stage on the albino or black background (Tyrc-2j/c-2j) in both pelage and whisker hair follicles (Fig. 2, H to M). The same pattern of cell loss was detected by using Dct-lacZ, KIT (c-Kit) (KIT), or microphthalmia-associated transcription factor (MITF) as markers (fig. S3 and Fig. 3, A and B). On the other hand, melanocytes in the epidermis and dermis of hairless skin (e.g., tail and soles) survived throughout the hair regeneration cycle (fig. S2, o and p). These findings indicate that BCL2 selectively protects melanocyte stem cells at the time of their transition into the dormant state in the niche and could potentially be responsible for certain forms of human presenile hair graying, although no direct supporting evidence has been reported thus far.

Fig. 2.

Loss of Bcl2–/– melanocyte stem cells upon entry into the dormant state. (A to D) Distribution of Dct-lacZ+ melanoblasts (arrows) in the Bg (top double arrow) of pelage follicles at P6.5 and P8.5. Whereas Bb melanocytes appear largely unchanged (bottom double arrow), bulge melanoblasts are lost in Bcl2–/– follicles at P8.5 [compare (D) with (C)] but not at P6.5 [compare (B) with (A)]. (E) Comparison of the total number per field of Dct-lacZ+ melanoblasts in the bulb versus in the bulge plus subbulge of Bcl2–/– and Bcl2–/+ pelage follicles at P8.5 on 7-μm sections (magnified 100×). KIT expression matches Dct-lacZ+ in bulge melanoblasts [stage 6 (F) and stage 8 (G)] of Bcl2+/+ animals (magnification, 630×). Cell size is diminished from stage 6 to stage 8. Terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick end labeling (TUNEL), lacZ, and 4′,6′-diamidino-2-phenylindole (DAPI) staining of stage 7 skin from P6.5 Bcl2–/– (H) and P6.5 Bcl2+/– (I) mice. Arrowheads show apoptotic inner root sheath keratinocytes. The inset area, marked with the arrow in (H), shows an apoptotic melanoblast. The top inset on the right shows the merged view for TUNEL (green) and DAPI (blue). The bottom inset shows the merged view for TUNEL (green) and LacZ (red). Distribution of Dct-lacZ+ melanoblasts in the niche: Bg of pelage hair follicles (J and K) and LE (double arrow) of whisker hair follicles [(L and M), double arrows] from P8.0 mice with white backgrounds (Tyrc-2j/c-2j).

Fig. 3.

Effect of aging and Mitf mutation on melanocyte stem cells. (A and B) Coincident expression of Dct-lacZ, KIT, and MITF in hair follicle melanoblasts or melanocytes (magnification, 200×). (See fig. S3 for more details.) (C) Ectopically pigmented melanoblasts (lacZ+, blue) in the bulge region (arrow) of 3.5-month-old Mitfvit/vit follicles. (D) Magnified view of pigmented bulge melanoblasts. (E) Absence of pigment in lacZ+ bulge melanoblasts of age-matched Mitf+/+ follicles. (F) Quantitation of niche melanocytes (lacZ+), either unpigmented (classical stem cells, blue) or ectopically pigmented (green), in LE of whisker follicles (positions a2 and a3) (see fig. S1i for positions). (G) Number of unpigmented niche melanoblasts in whisker follicles (positions a2, a3, c5, and d5) with black, gray, and white hair in 18- to 22-month-old (18–22M) wild-type mice. Asterisk indicates statistical significance (P < 0.01). (H and I) Ectopically pigmented melanoblasts in the niche (LE of whisker follicles) of aging wild-type mice (whole-mount view).

In contrast to Bcl2–/–, the Mitfvit/vit (13) graying mouse model exhibited a gradual decrease of melanocyte stem cells rather than abrupt loss (figs. S4 and S6). This strain contains a mild hypomorphic mutation in Mitf, the melanocyte master transcriptional regulator [(20, 21) and references therein]. At early to mid-anagen of the third hair cycle, lacZ+ cells left in the niche of Mitfvit/vit pelage follicles and Mitfvit/+ whisker follicles often produced melanin pigment and exhibited a bipolar or dendritic morphology (Fig. 3, C and D, and fig. S4, j and s). These pigmented cells are unusual because the niche of wild-type controls contains only unpigmented melanocyte stem cells. We provisionally use the term ectopic pigmentation or differentiation for this reproducibly observed population because it is uncertain by which pathway these cells became pigmented, although they were absent in age-matched controls whose niche melanoblasts remain undifferentiated (Fig. 3E and fig. S4u).

Physiologic (senile) aging in mice also produces hair graying (fig S5), which could be caused by loss of melanocyte stem cells. Indeed during physiologic aging, niche melanoblasts (lacZ+) were lost in a gradual and progressive fashion (Fig. 3, F and G). Moreover, whole-mount cross sections of 8-month-old follicles revealed pigment-containing melanocytes within the stem-cell niche in addition to their scattered distribution in the outer root sheath below the niche in whisker follicles (Fig. 3, H and I, and fig. S5, k to n). The appearance of these pigmented melanocytes in the niche is reminiscent of pigmented niche melanocytes observed during the accelerated graying of Mitf-vit mutants. Quantitative analysis revealed that the presence of these cells was accompanied by simultaneous loss of the typical unpigmented Dct-lacZ+ melanoblasts in the niche and correlated closely with aging (Fig. 3, F and G). Thus, self-maintenance of melanocyte stem cells is essentially complete in young animals but becomes defective with aging.

We also analyzed the distribution of melanoblasts in aging human hair follicles with the use of MITF immunostaining (Fig. 4). MITF+ small unpigmented melanoblasts were found in the outer root sheath preferentially around the bulge area where the arrector pili muscle attaches below the level of the sebaceous gland (Fig. 4, A to C), similar to previously described amelanotic melanocytes (22, 23) that express PMEL17 (24, 25), a transcriptional target of MITF (26). These cells have been suggested to be a reservoir population for differentiated melanocytes (23) and exhibit very similar morphology to melanocyte stem cells in mice. Whereas MITF+ immature melanoblasts were abundant in follicles from 20- to 30-year-old subjects (2 to 3% of the total basal keratinocytes in the bulge area), they were absent from most hair follicles of 70- to 90-year-old subjects (Fig. 4J). MITF+ melanocytes in the uppermost area (infundibulum) of the outer root sheath did not decrease significantly with aging, thus serving as a control population in these studies (fig. S7).

Fig. 4.

Melanoblast and melanocyte distribution in human hair follicles from different age groups. Human scalp specimens were immunostained with antibodies against MITF. (A and B) MITF+ cells (arrows) are distributed on the outer root sheath in the bulge of follicles from 20 to 30-yearold individuals. Magnification in (A), 200×; (B), 630×. (C and D) Representative views of the bulge from follicles of 40- to 60- and 70- to 90-year-old people, respectively (magnification, 630×). (E) Schematic for human hair follicle with pigmented hair. Immature MITFlow melanoblasts (blue) are located in the lower permanent portion (the bulge). MITFhigh melanocytes are located in epidermis, infundibulum (If) (brown), and hair matrix (M, green). ORS, outer root sheath. See Fig. 1A for abbreviations. (F to H) The bulb region of follicles from different age groups. Mature melanocytes in the hair matrix express MITF (arrowheads). yo, years old. (I) An MITF+ melanocyte that contains abundant melanin granules and long dendrites is detected in the bulge and subbulge of follicles specifically from middle-aged individuals. (J) The frequency of MITF+ cells per basal keratinocytes in the bulge. Asterisks indicate statistical significance (P < 0.01).

Follicles from intermediate-aged individuals (40 to 60 years old) revealed intermediate loss of bulge melanoblasts (Fig. 4, C and J). Bulge melanoblasts were found more in pigmented follicles than in gray follicles (18), as shown recently with PMEL17+ bulge melanoblasts of middle-aged individuals (27). In addition, as with aged or Mitfvit mouse follicles, ectopically pigmented MITF+ cells were occasionally observed in the bulge area or just below. These cells closely resembled the dendritic melanocytes described by Narisawa et al. in the bulge area of human follicles (28). The ectopically pigmented or differentitated melanocytes were seen exclusively in middle-aged follicles but did not accumulate in the bulge area, suggesting that they are not self-maintaining.

Our results demonstrate that Bcl2 is selectively critical for maintenance of melanocyte stem cells, specifically for entry into the dormant state. Bcl2 was previously shown to modulate hematopoietic stem-cell pool size (29). Different lineages might use distinct antiapoptotic mechanisms to resist the specific stress signals for dormancy. Although melanin biosynthesis has been elegantly shown to be cytotoxic in the context of a certain genetic mutation (6), stem-cell disappearance in Bcl2 null mice does not require melanogenesis. Bcl2 is a transcriptional target of MITF (30), but Bcl2 does not appear to fully account for melanocyte loss in the context of the weakly hypomorphic Mitfvit/vit allele.

Our data suggest a previously unknown pathophysiologic explanation for hair graying. Loss of melanocyte stem cells can be observed and temporally precedes the loss of differentiated melanocytes in the hair matrix. Thus, incomplete maintenance of melanocyte stem cells appears to cause physiologic hair graying through loss of the differentiated progeny with aging. This is associated with ectopic melanocyte pigmentation or differentiation within the niche. Possible explanations include premature differentiation or activation of a senescence program [which induces pigmentation in vitro (31)]. Acceleration of this process in Mitfvit follicles implicates MITF in the self-renewal of melanocyte stem cells. The precise roles for stem-cell apoptosis versus ectopic differentiation remain to be determined but may similarly contribute to stem-cell loss in other aging organ systems.

Supporting Online Material

Materials and Methods

Figs. S1 to S7

References and Notes

View Abstract

Stay Connected to Science

Navigate This Article