The Epidermal Stem Cell Factor Is Over-Expressed in Lentigo Senilis (Hattori et al. 2004)

TL;DR

A human case-comparative study of 29 Japanese patients with solar lentigo (lentigo senilis, LS) established that epidermal keratinocytes — not dermal fibroblasts — are the dominant source of stem cell factor (SCF / KITLG) in aged hyperpigmented skin. Lesional epidermis showed 3.9-fold elevation of SCF mRNA (p<0.01) and 1.6-fold elevation of membrane-bound (31 kDa) SCF protein (p<0.05) relative to perilesional normal skin. The soluble isoform (18 kDa) was undetectable. Crucially, c-KIT (KIT) mRNA was unchanged (1.16-fold, NS) in lesional epidermis, suggesting receptor saturation or downstream amplification rather than receptor upregulation as the signal-transduction mechanism. The paper also documents co-upregulation of the endothelin-1 / ET-B receptor cascade, which acts synergistically with SCF/KIT to activate melanocyte melanogenesis. GROα and bFGF were not significantly elevated in this cohort. This study defined the canonical paracrine keratinocyte → melanocyte SCF/KIT axis for solar lentigo, a mechanism still cited as foundational to age-spot biology.

Background

Prior to 2004, the mechanistic basis of solar lentigo (senile lentigo; “age spots”) was poorly resolved. The lesions are characterised histologically by increased melanin deposition in the epidermis and elongated rete ridges containing increased numbers of active melanocytes ¹. Several paracrine factors had been proposed to mediate melanocyte activation by UV-damaged or aged keratinocytes, including basic fibroblast growth factor (bFGF), endothelin-1 (ET-1), hepatocyte growth factor (HGF), and stem cell factor (SCF, encoded by KITLG). SCF signals through the receptor tyrosine kinase c-KIT (encoded by KIT), which is expressed on melanocytes and mediates proliferation, survival, and melanogenic activation. The relative contribution of each paracrine factor, and whether the source was epidermal or dermal, remained contested. The Hattori 2004 study was designed to resolve the compartment question directly, using microdissected epidermal sheets from lesional and perilesional skin.

The distinction between epidermal and dermal SCF sources is biologically critical: dermal fibroblast-derived SCF (as demonstrated in coculture systems, e.g., Kovacs 2010 — see Limitations) acts via diffusion across the basement membrane, whereas epidermal keratinocyte-derived SCF is immediately proximal to resident melanocytes in the basal layer. Membrane-bound SCF (the 31 kDa isoform) acts via juxtacrine contact rather than long-range diffusion, making direct keratinocyte–melanocyte cell contact the primary signalling mode in the Hattori model.

Methods

Cohort

Total enrolled: n=29 Japanese patients with lentigo senilis (clinical diagnosis)
Age range: 35–85 years (mean ~68 years)
Sex: 16 males, 13 females
Lesional and perilesional normal skin biopsies were collected from each patient; paired design

Tissue preparation

Epidermal sheets were separated from dermal tissue by dispase digestion at 4 °C, a protocol that selectively detaches the epidermis while minimising contamination from dermal fibroblasts. This purification step is the methodological pivot of the study — it allows comparison of purely epidermal mRNA and protein levels between lesional and perilesional skin, avoiding confounding from the dermal compartment.

Assays performed

Assay	Analyte(s)	n used	Comparator
RT-PCR	SCF mRNA	7	Lesional vs perilesional epidermis
Western blot	SCF protein (31 kDa; 18 kDa)	6	Lesional vs perilesional epidermis
RT-PCR	GROα mRNA	4	Lesional vs perilesional epidermis
RT-PCR	c-KIT mRNA	6	Lesional vs perilesional epidermis
IHC	bFGF + GROα protein	10	Lesional vs perilesional skin sections

Densitometry was used to quantify Western blot band intensities for the 31 kDa (membrane-bound) and 18 kDa (soluble) SCF isoforms. RT-PCR band intensities were normalised to β-actin internal control.

Key Findings

SCF mRNA is elevated 3.9-fold in lesional epidermis

Lesional epidermal SCF mRNA was 3.9-fold higher than perilesional normal epidermal SCF mRNA in the RT-PCR cohort (n=7, p<0.01) ¹. This is the primary molecular finding and the most replicated figure from this paper.

Membrane-bound SCF protein is elevated 1.6-fold; soluble isoform absent

Western blot analysis (n=6) of lesional vs perilesional epidermal lysates revealed:

31 kDa (membrane-bound) SCF: 1.6-fold elevated in lesional epidermis (p<0.05) ¹
18 kDa (soluble) SCF: NOT detected in lesional or perilesional LS epidermis ¹

The absence of the soluble isoform in epidermal preparations is consistent with the juxtacrine model: keratinocytes produce SCF predominantly in its membrane-tethered form, which signals to adjacent melanocytes via direct cell contact rather than via diffusion through extracellular matrix.

c-KIT mRNA is unchanged in lesional epidermis

RT-PCR showed c-KIT mRNA at 1.16-fold in lesional vs perilesional epidermis (NS) ¹. This null result — included in the paper and frequently omitted from secondary citations — rules out the alternative hypothesis that melanocyte KIT receptor upregulation is the primary mechanism driving increased signalling. The elevated upstream ligand (SCF) without receptor upregulation implies the signal gain is entirely ligand-driven.

GROα and bFGF are NOT significantly elevated

RT-PCR: GROα mRNA not significantly elevated (1.13-fold, n=4, NS) ¹
IHC: bFGF and GROα protein not significantly elevated (n=10, NS) ¹

These negative results distinguish LS from other hyperpigmentation conditions (e.g., UV-induced tanning) where GROα and bFGF may play more prominent roles. They also narrow the dominant paracrine axis in LS specifically to SCF and ET-1.

ET-1 / ET-B receptor cascade is co-upregulated

The paper documents upregulation of the endothelin-1 (ET-1) and its receptor ET-B (EDNRB) in lesional skin, consistent with prior work from the Imokawa group. ET-1 and SCF act synergistically — ET-1 activates melanocytes via a G-protein coupled pathway (Gq → PKC → CREB / MITF), while SCF/KIT activates via receptor tyrosine kinase (RAS/MAPK → MITF). The two pathways converge on MITF-mediated transcriptional upregulation of melanogenic enzymes (tyrosinase, DCT, TYRP1), amplifying melanogenesis beyond what either signal alone produces. needs-replication (ET-1/SCF synergy quantification in LS specifically is from mechanistic in-vitro work; in vivo synergy in LS is inferred, not directly measured in this cohort)

Proposed Mechanism

The central paradigm from this paper, which became canonical for solar lentigo:

Chronic UV exposure and/or keratinocyte aging in the basal epidermis upregulates expression of KITLG (SCF) mRNA and protein in keratinocytes
Membrane-bound SCF (31 kDa) on keratinocytes signals to adjacent melanocytes via direct juxtacrine contact, activating c-KIT (→ RAS → MAPK → MITF → melanogenesis)
ET-1 from keratinocytes simultaneously activates ET-B on melanocytes (Gq → PKC → MITF), synergising with the SCF/KIT signal
Convergent MITF upregulation drives increased tyrosinase expression and melanin synthesis
The net result is focal hyperpigmentation characteristic of solar lentigo

The ligand-side (SCF) rather than receptor-side (KIT) upregulation is the mechanistic asymmetry the paper uniquely establishes: it is not that melanocytes become more sensitive to SCF; it is that keratinocytes produce more SCF.

See kitlg and kit for detailed gene/protein pages (R40 forward-refs; pages not yet seeded). needs-canonical-id

Relationship to Kovacs 2010 (Dermal Fibroblast Model)

The Hattori 2004 model (epidermal keratinocyte as SCF source) should be distinguished from coculture studies where dermal fibroblasts co-cultured with melanocytes produce SCF that stimulates melanocyte proliferation. Notably, Kovacs et al. (2010) demonstrated fibroblast-derived bFGF and SCF in a coculture paradigm as a mechanism for fibroblast-driven melanocyte activation. The experimental compartment differs: Kovacs 2010 uses fibroblast conditioned media acting on melanocytes via soluble/secreted factors in a culture-dish geometry; Hattori 2004 uses microdissected epidermal sheets (no dermis) from intact biopsies. These are not contradictory — both dermal fibroblasts and epidermal keratinocytes may contribute SCF to the melanocyte microenvironment in vivo — but the Hattori 2004 data specifically isolates the epidermal compartment and establishes that epidermal keratinocytes alone contain the SCF elevation, without requiring a dermal fibroblast contribution. contradictory-evidence (compartment contributions in intact LS skin not directly quantified simultaneously)

Strengths

Paired within-patient design (lesional vs perilesional from same individual) eliminates inter-individual confounding in gene/protein expression comparisons
Epidermal sheet separation isolates the epidermis from dermis, directly addressing the compartment question that had been unresolved
Multi-modal confirmation: mRNA (RT-PCR) + protein (Western blot, two isoforms) + immunohistochemistry across three independent molecular read-outs
Negative controls reported: GROα, bFGF, and c-KIT all tested; two of three null results reported honestly, strengthening confidence in the SCF positive finding
Human tissue from clinically-diagnosed patients; not a model-organism study

Limitations and Gaps

Small per-assay n values: n=7 (RT-PCR), n=6 (WB), n=4 (GROα PCR), n=10 (IHC). While within-patient paired design improves power, the absolute sample sizes limit generalisability. needs-replication
No sex-stratified analysis reported despite mixed-sex cohort (16M / 13F); estrogen modulates melanogenesis and SCF expression; sex-specific effects are unexamined
Japanese cohort only: LS occurs across all ethnicities but morphological and biochemical features may differ by skin phototype / Fitzpatrick type; no cross-ethnic validation available from this cohort
Age range 35–85 is wide; younger patients (35–50) may represent a qualitatively different aetiology than octogenarian patients; age-stratified analysis not reported
Mechanism of SCF transcriptional upregulation not determined: the paper establishes WHAT (SCF is elevated, membrane-bound isoform predominates) but not HOW (what UV-sensing or aging-associated transcription factor drives KITLG upregulation in keratinocytes). Candidate regulators include AP-1, NF-κB, and MITF itself, but this paper does not address this. no-mechanism
ET-1/SCF synergy not quantitatively tested in vivo: the synergy model is inferred from in-vitro mechanistic work; the relative contributions of ET-1 vs SCF to the LS phenotype remain unquantified in vivo
No functional rescue experiment: the epidermal compartment origin of SCF is established, but blocking membrane-bound SCF (e.g., anti-KIT antibody or soluble KIT-ECD) in LS biopsies was not performed; causality of the SCF elevation for hyperpigmentation is inferred, not directly tested in this paper

Significance and Impact

This paper (117 citations as of 2026; citation percentile 99th within its field per OpenAlex) established the keratinocyte → melanocyte membrane-bound SCF/KIT axis as the canonical mechanism for solar lentigo. Prior to this work, the paracrine factors mediating UV/aging-driven focal melanocyte activation were debated and the cellular source was uncertain. The three key contributions that have persisted in the field:

The epidermal (keratinocyte) source of SCF, isolated by epidermal sheet separation
The membrane-bound isoform as the dominant form in lesional epidermis (juxtacrine, not paracrine diffusion)
The c-KIT receptor null result — ligand-driven signal amplification, not receptor sensitisation

In the context of aging biology, the paper connects to altered-intercellular-communication (keratinocyte–melanocyte paracrine cross-talk is altered in aged/UV-damaged skin) and to cellular-senescence (keratinocyte senescent secretory phenotype may overlap with the pro-SCF, pro-ET-1 transcriptional state). It is the primary citation on the skin-aging phenotype page for the SCF/KIT focal hyperpigmentation subsection and on melanogenesis for the SCF/KIT axis.

Extrapolation

This is a human tissue study; the standard model-organism extrapolation table is not applicable. All findings are directly in human lesional skin.

Dimension	Status	Notes
Species	Human	Japanese adults 35–85 yr; no extrapolation needed
Replicated independently?	partial	SCF/KIT axis in LS cited widely; the specific per-isoform WB data and per-assay n values need independent replication (#gap/needs-replication)
Functional intervention tested?	no	Causal inference relies on ligand–receptor biology; no blocking experiment in this paper

Cross-References

melanocytes — target cell type; KIT receptor is expressed on melanocytes
keratinocytes — source cell type; SCF overexpression established here
kitlg — the gene encoding SCF (R40 forward-ref; page not yet seeded) needs-canonical-id
kit — the gene encoding c-KIT receptor tyrosine kinase (R40 forward-ref; page not yet seeded) needs-canonical-id
skin-aging — phenotype page; this study is canonical reference for SCF/KIT focal hyperpigmentation mechanism
melanogenesis — process page; SCF/KIT and ET-1/ET-B converge on MITF-driven melanogenesis
cellular-senescence — hallmark; senescent keratinocyte SASP may include SCF/ET-1 upregulation
altered-intercellular-communication — hallmark; the keratinocyte → melanocyte paracrine axis is the exemplar for this hallmark in skin

Footnotes

hattori-2004-scf-solar-lentigo · doi:10.1111/j.0022-202x.2004.22503.x · n=29 (total; per-assay n=7 RT-PCR, n=6 WB, n=4 GROα PCR, n=10 IHC) · observational, paired within-patient · model: human lentigo senilis biopsies, Japanese adults 35–85 yr · p<0.01 (SCF mRNA), p<0.05 (membrane-bound SCF protein), NS (GROα, bFGF, c-KIT) ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷

🧬 Aging Wiki

Explorer

The Epidermal Stem Cell Factor Is Over-Expressed in Lentigo Senilis: Implication for the Mechanism of Hyperpigmentation