The
cytokine/chemokine side of microinflammation
Why
does microinflammation take place in the
pilosebaceous unit and for what benefit
and purpose? Fig. 2
and Fig. 3
show, in a simplified sequence, that
inflammation is a multistep process which
may start from a primary event. Let us look
at the clues at the "crime scene" of AGA:
we observe a perifollicular infiltrate in
the upper follicle near the infundibulum.
2
7 This suggests that the primary causal
event for the triggering of inflammation
might occur near the infundibulum. 3,7
Supporting this point of view, improvement
of the inflammatory aspect of AGA has been
reported in a pilot study with an antimicrobial
lotion.
7
One could speculate that several inhabitants
of the scalp, such as the "triad" ( Propionibacterium
sp.; Staphylococcus sp.;
Malassezia ovalis ) or other members
of the transient flora, could be involved
in this complex inflammatory process. 7
The presence of
porphyrins (produced by Propionibacterium
sp.) in the pilosebaceous duct of
58% of AGA patients (compared with 12% of
control subjects), which are able
to induce the production of complement (C5)
chemotactic factor, is considered to be
a possible cofactor of this initial pro-inflammatory
stress. 6,7
Keratinocytes
are also known to respond within minutes
to chemical stress, pollutants, UV irradiation
or even mechanical stress. 37
Not only are radical oxygen species,
38
NO, 39
PGs, and histamine 40
produced, but also intracellularly stored
IL-1 is released 37,41
(see Fig. 2
and step 1 of Fig. 3
). By itself, this pro-inflammatory
cytokine (as well as IL-1 which binds to
the same receptor) is able to inhibit the
growth of isolated hair follicles in culture
in vitro. 9
11 This concentration-dependent inhibition
of human hair elongation and survival indicates
a high sensitivity to IL-1 of the isolated
organ in culture in vitro (IC
50 = 10 pg/mL 11
). In vivo , transgenic mice
which overexpress IL-1 in the basal
epidermis and in the outer root sheath of
their pelage hair follicles exhibit a spontaneous
cutaneous phenotype characterized by a sparseness
of hair. 42
As a response to an IL-1 signal, adjacent
keratinocytes which express receptors for
IL-1 start to engage the transcription of
IL-1 responsive genes 41
( Fig. 3
, step 2). In
vitro , following IL-1 stimulation,
this transcriptional activation cascade
is induced within 6 h in plucked human hair
follicles. 11
Alternatively, skin keratinocytes, which
may also have antigen presenting capabilities,
could theoretically induce T-cell proliferation
in response to bacterial antigens. 51 These
antigens, once they have been "tagged,"
are then selectively destroyed by infiltrating
macrophages, Langerhans cells, or natural
killer cells. 50,52 On many occasions, however,
the causal agent persists, resulting in
sustained inflammation ( Fig. 3, step 4).
This corresponds partly to the situation
which has been pictured in the progression
zone of roughly one-third of alopecia cases:
infiltrating T lymphocytes, together with
mastocytes and macrophages, located in the
upper perifollicular adventitial dermal
sheath perpetuate a local inflammatory stage.
27 This phase of inflammation
often results in tissue remodeling, where
collagenases, such as matrix metalloproteinase
(MMP)-9 (transcriptionally activated by
pro-inflammatory cytokines) or MMP-8 (directly
produced by infiltrating cells), may play
an active role. 5355 Thus, collagenases
are suspected to contribute to the tissue
changes and the so-called "perifollicular
fibrosis" by "preparing"
tissue matrix and basal membranes for macrophages
and T-cell adhesion. Accordingly,
this scenario facilitates the secretion
of membrane-anchored cytokines, such as
TNF-. 55 Other factors, such as MCP-1, have
been directly suspected to contribute to
organ fibrosis in an experimental model
of renal inflammation. 56 As MCP-1, together
with other chemokines, was found to be expressed
in human hair follicles in vitro, 11 as
well as in the eccrine ducts of sebaceous
glands in vivo, 57 it might also be actively
involved in the progression of perifollicular
fibrosis detected in AGA. 26 The development
of perifollicular fibrosis might thus appear
as the signature of a disequilibrium between
pro-and anti-inflammatory pathways.
Relations
between inflammation and steroidogenesis:
the missing link
There is no question that androgens are
major modulators of hair loss. Recently,
it was shown that testosterone inhibited
the growth of outer root sheath keratinocytes
only when they were cocultured with dermal
papilla cells derived from the bald scalp
of an adult macaque, 58 reinforcing the
hypothesis of an androgen influence on hair
growth via the dermal papilla. 59
The potent metabolite of testosterone (i.e.
5-dihydrotestosterone, 5-DHT) is considered
as a "culprit". 60 5-DHT is generated
from testosterone through the activity of
5-reductase (5-R). Two active isoforms of
5-R, which differ both in tissue site distribution
as well as in optimal pH for enzymatic activity,
have been identified and cloned. 61,62,
63 While the type II isoform is considered
to be the major isozyme in genital tissues,
61 the type I isoform is considered to be
the major isoform expressed in skin and
in the pilosebaceous unit. 64,65 Isoform
II, however, has recently been detected
in the inner root sheath of the pilosebaceous
unit by immunohistochemistry, 66,67 Northern
blotting, 67 and the pH dependence of optimal
enzymatic activity. 67 Thus, the contribution
of both isoforms in the regression of the
pilosebaceous unit is still a matter of
debate. Recently, a clinical study using
finasteride, a strong inhibitor of 5-RII
(and weak inhibitor of 5-RI), showed that
intervention in androgen metabolism could,
to some extent, modulate the progression
of AGA, when the drug was given by the oral
route, 68 but not topically. 69 After oral
ingestion, an improvement of hair growth
was observed, which was associated with
a drastic reduction of serum levels of 5-DHT,
corresponding to those observed in castrates.
68 Despite such a
reduction of circulating 5-DHT levels, however,
a number of individuals (60-70%) still remained
unresponsive to this treatment, indicating
again that simple dysregulation of 5-DHT
synthesis levels or a genetic polymorphism
of 5-R genes cannot account for all cases
of AGA, and a polygenic etiology should
be considered. 1
Thus,
to date, the only evident link that can
be established between androgen metabolism
and the complex inflammatory process is
sebum production which is controlled by
androgens.
70 As sebum harbors a large amount of microorganisms
which use lipids as nutrients, 8 it cannot
be excluded that, at least for some individuals,
androgen metabolism might facilitate the
colonization of the sebaceous infundibulum
and sebaceous ducts by such microorganisms
which may be involved in the first steps
of pilosebaceous unit inflammation.
We propose here working
hypotheses which do not invalidate the contribution
of a hereditary genetic androgen imbalance
in AGA, 60 but rather attempt to integrate
the neglected microinflammatory aspects
of alopecia into the complex etiology of
AGA. On the one hand, excessive local
and/or endocrine, genetically exacerbated
5-DHT synthesis results in sebaceous
gland enlargement; 2,60 as a consequence,
some scalps might
offer more comfortable niches to harbor
the previously mentioned pro-inflammatory
microorganisms. 6,7 On the other
hand, androgen imbalance and metabolism
may be locally exacerbated by pro-inflammatory
cytokines. For example, gingival fibroblasts
have been reported to modify their androgen
metabolism through the action of several
growth factors, such as epidermal growth
factor (EGF), transforming growth factor
beta (TGF-), and the pro-inflammatory cytokines
IL-1 and TNF-. 71 Therefore, one could speculate
that, once the inflammatory
process has been triggered, the androgenetic
mechanism of alopecia could subsequently
be locally amplified. This upregulation
of androgen metabolism by pro-inflammatory
cytokines remains, however, to be established
at the pilosebaceous unit level.
| Our
visit to the "crime scene" of AGA
yielded many clues ( Fig. 4
). We know now that, at least
in about one-third
of cases, the tool which causes the
lethal damage is a microinflammatory
process. Several factors are
present, however, which are suspected
to have handled the tool: androgens,
microbial flora, endogenous or exogenous
stress, genetic imbalance, and possibly
others. Although other suspects or
tools are likely to be discovered
in the future, it cannot be excluded
that, for each individual, the causal
agent, as well as the sequence of
events or combined factors, may be
different. The large number of molecules
claimed to be active and patented
in this field, 89
and their limited efficacy in
offering a definite and extensive
cure of AGA, confirm that the mechanism
of AGA is highly complex. Accordingly,
it appears that, due to the complexity
and multiple interactivities and cooperations
involved throughout the distinct inflammatory
pathways (partly described in Fig. 2
), an anti-inflammatory strategy
should be targeted to the appropriate
effector(s) at the right moment. For
this purpose, we have developed a
simple assay to evaluate individuals
with potentially affected hair follicles.
11
We observed that plucked hair
specimens of 33% of the 116 volunteers
evaluated could be classified as highly
inflammatory in terms of spontaneous
IL-1 production. 11
Consequently, the identification
of the "inflammatory alopecic individuals"
may help to adapt the right answer
to
the right cause. Such a selective
approach might be valuable for other
parameters, such as an imbalance in
11 HSD activity, 5-DHT synthesis,
or microorganism colonization. Encompassing
individual diversity is thus a prerequisite
for appropriately addressing the biological
conditions contributing to AGA. Our
findings and a review of the literature
suggest that inflammation in its diversity
is a potentially active player to
consider in this approach. |
|
