Journal of Microscopy, Vol. 204, Pt 3, December 2001, pp. 203211.

Received 15 March 2001; accepted 3 July 2001

Microscopical investigations on the epicuticle of

mammalian keratin fibres

J. A. SWIFT* & J. R. SMITH

*Developmental Toxico-Pathology, University of Liverpool, Liverpool L69 3BX, U.K.
Scanning Probe Microscopy Laboratory, School of Pharmacy & Biomedical Sciences, University of
Portsmouth, St. Michael's Building, White Swan Road, Portsmouth PO1 2DT, U.K.

Key words. Atomic force microscopy, human hair, monotremes, sheep's wool,
transmission electron microscopy.

Summary component in the original fibre giving rise to the sacs is

known as the epicuticle. The diminution or absence of an
The existence of a thin chemically resistant layer, the
Allworden reaction is used as a diagnostic test for damage to
epicuticle, close to the surfaces of all undamaged mamma-
the fibre surface and to its epicuticle (Allworden, 1916). It is
lian keratin fibres has been known since 1916. The
generally accepted that the halogen solutions result in
identification of such a specific structure within the fibre
oxidative scission of cystine disulphide crosslinks in the fibre
cuticle has remained elusive. Now, through transmission
cuticle's subsurface proteins, these being rendered water
electron microscope investigations of stained transverse
soluble through newly formed cysteic acid residues (Brad-
sections of hairs from various animal species, the epicuticle
bury & Leeder, 1972). Because the epicuticle, which is
has been tentatively identified as a sharply defined and
located close to the intracellular surface of each cuticle cell,
continuous layer approximately 13 nm thick covering the
is both resistant to the chlorine water treatment and is
entire outwardly facing intracellular surface of every cuticle
semi-permeable, the solubilized proteins cause the char-
cell. The staining behaviour of the epicuticle leads one to
acteristic Allworden sacs to be raised by osmosis. Allworden
suppose that it is rich in cystine and that thioester-bound
membranes occur in the form of thin sheets varying in
lipids might be present within its bulk. With the atomic
thickness between 5 and 14 nm (Lindberg et al., 1948;
force microscope it was established that the undamaged
Lagermalm, 1954; King & Bradbury, 1968) and comprising
outer surface of all mammalian keratin fibres, even
mainly protein (King & Bradbury, 1968; Allen et al., 1985).
including those from the monotremes, were longitudinally
However, the histological identity of the epicuticle amongst
striated. The lateral spacing of the striations was always in
the components normally seen in sections of keratin fibres
the range 0.290.39 mm. Striations only occurred on the
examined under the transmission electron microscope
freely exposed outer surfaces of the original undamaged
(TEM) remains enigmatic. A separate histological identity
fibres; evidently arising by some, as yet undefined, interac-
has even been denied (Phan et al., 1995), the authors
tion in the follicle with the cuticle of the inner root sheath.
claiming rather that Allworden membranes are a by-
By stripping off fatty acids known to be covalently attached
product from the chemical degeneration of a much thicker
to the fibre's outer surface, the striations were shown to
subsurface component of the fibre cuticle (the A-layer; c.
reflect a corresponding irregularity of the epicuticle's
100 nm thick). In the present paper we show that this is
surface.
unlikely to be the case; having identified in the TEM a new
histological component of the cuticle that is in the correct
Introduction location and of the right thickness to be considered as the
epicuticle.
Chlorine water causes small membrane-bound sacs to be
A notable feature of all mammalian keratin fibres is fatty
raised at the surfaces of all undamaged mammalian keratin
acids covalently attached to their outer cuticular surfaces
fibres. The effect, which was first recorded by Allworden
and in the b-layers of the cell membrane complex (CMC)
(1916) and subsequently shown by Herbig (1919) to occur
that, together with an intervening d-layer, separate the
also with bromine water, is now known as the `Allworden
sheet-like cuticle cells from each other (Jones et al., 1996;
reaction' and the sacs as `Allworden membranes'. The
Jones & Rivett, 1997) (Fig. 1). With few exceptions (Peet
Correspondence: J. A. Swift. E-mail: jaswift@gayton.u-net.com. et al., 1992; Jones et al., 1996) the predominant species is

q 2001 The Royal Microscopical Society 203

204 J. A . S W I F T A N D J. R . S M I T H
Fig. 1. Schematic diagram illustrating the laminated subcompo-
nents of the first cuticle cell in a transverse section through a
mammalian keratin fibre close to its surface. The diagram
summarizes existing knowledge of the structure derived from
TEM investigations. It includes the epicuticle that has previously
thought to be in this location but has been identified as an outcome
of the present work. The cell membrane complex separates the
sheet-like layers of cuticle.
the anteiso-fatty acid 18-methyl eicosanoic acid (18-MEA),
which is biosynthesized from the precursor amino acid
isoleucine (Prottey & Ferguson, 1972; Oku et al., 1994).
Leeder & Rippon (1985) found that keratin fibres, the
surfaces from which these fatty acids had been chemically
stripped, still yielded a positive Allworden reaction, their
result thereby establishing that lipids are not responsible for
the semipermeability of the epicuticle. Negri et al. (1996)
examined in the TEM sections of wool where the fibres had
been exhaustively extracted beforehand with 70 : 30 v/v
chloroform/methanol to remove non-covalently bound
lipids. The surfaces of the fibres were contaminated by a
thin layer of debris arising from the preparative procedures
but beneath this, and superficial to the cuticular cells
proper, was an unstained layer of approximately 8 nm
thickness. This was assumed to be a layer of the covalently
bound lipids. Significantly, this layer was eliminated in fibres
treated with 0.1 m methanolic KOH for 30 min at 20 8C, a
reagent particularly efficient for the cleavage of thioester-
linked fatty acids. Leeder (1986), in studies of isolated
cuticle cells, found that chlorine water raised sacs over the
entire surfaces of the cells. His result provided the important
evidence that the epicuticle is not just restricted to that part
exposed at the fibre's outer surface but is present as an
envelope bounding the entire inner surface of each cuticle
cell. Leeder & Rippon (1985) also demonstrated that the
layer of lipid covalently attached at the fibre surface, which
they refer to as the `F-layer', is not essential for a positive
Allworden reaction to be given, i.e. this layer is not
considered an integral part of the epicuticle.
The epicuticle, which is predominantly proteinaceous in
nature, is highly insoluble; this is thought to be due to
isodipeptide crosslinks, i.e. of 1-amino-(g-glutamyl)-lysine
formed in the fibre follicle by the action of transglutaminase
in joining the lysyl and glutaminyl residues of the proteins
(Folk & Finlayson, 1977). Indeed, this is consistent with
significant amounts of the isodipeptide having been found
amongst the more insoluble residues from protease diges-
tions of cuticle (Schwan, 1981). Allen (1990) noted
similarities between the epicuticle of animal hairs and the
cell envelopes (CE) of terminally differentiated dead stratified
squamous epithelia. The CE of human stratum corneum has
an average thickness of 15.3 nm (Jarnik et al., 1998), is
crosslinked by isodipeptides, contains loricrin as its principal
component but also contains many other proteins including
cystatin, elafin, involucrin, filaggrin and several small
proline-rich proteins (Steinert & Marekov, 1995). Recently
Zahn et al. (2000), using methods of multiple linear
regression analysis, obtained an excellent match to the
amino acid composition of Allworden membranes on the
basis of the membranes being composed of 42% loricrin,
51% ultrahigh sulphur cuticle protein and 7% involucrin.
Longitudinal striations on the surfaces of the hairs of
several different animal species have been seen with the
TEM using thin-film surface replication methods (Rama-
nathan et al., 1955) and occasionally seen, albeit frequently
with some difficulty, by the direct examination of fibres with
the scanning electron microscope (Wolfram, 1972).
Recently, we have carried out a comprehensive examination
of human head hair with the atomic force microscope
(AFM) and found longitudinal striations covering the entire
outer cuticular surfaces of undamaged fibres, except at the
margins of the overlapping scales (Swift & Smith, 2000a).
The striations had a lateral spacing of c. 350 nm and were
of convex profile, rising to a height of approx. 9 nm. Our
further interest in the present paper was to extend these
initial AFM observations to see whether surface striations
might be a common feature of all mammalian keratin fibres
and to elicit information on the relationship between them
and the epicuticle.
Experimental
Materials
Fibre samples were obtained from the following mammals:
domestic cat, guinea-pig, sheep (Lincoln and Corriedale),
alpaca, Asiatic lion, echidna (Tachyglossus aculeatus) and
platypus (Ornothorhynchus anatinus). Samples of head hair
were also secured from contemporary European and Asian
human subjects not given to using such damaging chemical
treatments as permanent waving, bleaching or oxidative
dyeing. In all cases, only the undamaged root end portions
of the fibres were selected for AFM and TEM work; for
q 2001 The Royal Microscopical Society, Journal of Microscopy, 204, 203211
 EPICUTICLE OF MAMMALIAN HAIRS 205

Fig. 2. AFM images of hairs from: (a) human head, (b) domestic cat, (c) guinea-pig, (d) platypus. All show the presence of surface striations.
(a), (c) and (d) are views normal to the surfaces; (b) is a tilted isometric projection. The root end of each hair is at top left of each
image. B broken scale edge, E undamaged scale edge, G scale edge `ghost', N freshly revealed intercuticular surface, S striated
surface. Micrograph edge dimensions for (a)(d) are 10.0, 9.1, 20.0 and 20.0 mm, respectively.

human hair this was of segments within 10 mm of the scalp
surface.
AFM methods
One set of fibres was cleaned prior to examination in the
AFM by sonication (U300H, Ultrawave Ltd, Cardiff, UK) for
2 min at room temperature in a solution of sodium dodecyl
sulphate (1%) in double distilled water (50 cm3) and was
then rinsed in copious amounts of double distilled water
before drying in air. Human hair samples were also Soxhlet
extracted for 5 h with chloroform/methanol (70 : 30 v/v)
to remove loosely bound lipids (Negri et al., 1996) and, after
allowing the solvents to evaporate, these were then
exhaustively rinsed in double distilled water before drying
in air. Some of these were treated further by soaking for 1 h

q 2001 The Royal Microscopical Society, Journal of Microscopy, 204, 203211

206 J. A . S W I F T A N D J. R . S M I T H
at room temperature in 0.1 m methanolic KOH, a procedure
known to remove covalently linked fatty acids from the fibre
surfaces (Negri et al., 1993, 1996). The latter samples were
then rinsed exhaustively in double distilled water and dried
in air before Soxhlet extracting again for a further 5 h in
chloroform/methanol to remove any lipids dislodged by the
alkali treatment but still absorbed in the fibres.
Topographic AFM observations were made with a
Discoverer TMX2000 scanning probe microscope (Thermo-
Microscopes, Bicester, UK), using standard-profile, silicon
nitride tips mounted on V-shaped cantilevers of length
200 mm and nominal force constant, K, 0.032 N m21,
respectively (ThermoMicroscopes, Sunnyvale, CA, USA).
Animal hairs were attached to carbon-loaded, double-sided
adhesive tape and were scanned using a tripod scanner with
a maximum x,y,z-translation of 70  70  12 mm. Cap-
tured images were left-shaded using a hypothetical light
source to enhance topographic features. Fourier analyses of
images were performed using the one-dimensional fast
Fourier transform function in the TopoMetrix image
analysis software (Version 3.09.06, ThermoMicroscopes).
TEM methods
The root ends of fibres were embedded with Spurr's resin
(Spurr, 1969) directly in size 0 gelatin capsules. Using a
Reichert-Jung `Ultracut' ultramicrotome transverse sections
of the fibres at a nominal thickness of 80 nm were cut with
the aid of a diamond knife and collected on 400-mesh gold
electron microscope grids previously equipped with a thin
collodion support film. Grids were stained by immersion in
ammoniacal silver reagent (0.1 m AgNO3, to which
concentrated ammonia was added until the initial pre-
cipitate had just dissolved) for 20, 40 and 60 min and then
soaked in three changes of distilled water over 1 h before
drying. The grids were examined in a Zeiss EM109 TEM
operated at 50 keV.
Results and discussion
AFM observations
In all the hairs from different mammalian species cleaned by
sonication in sodium dodecyl sulphate, longitudinal surface
striations were seen to cover the entire outward-presenting
surfaces of the undamaged scales (Fig. 2). They also termi-
nated abruptly at the point where portions of an overlying
scale had been removed by mechanical action; the freshly
revealed intercuticular surface being relatively smooth
(Figs 2(a), (b) and (d)). These observations are consistent
with what has already been seen in comprehensive AFM
studies of human head hair (Swift & Smith, 2000a).
One-dimensional Fourier analyses of our micrographs in
a direction perpendicular to the striations showed either a
pronounced single peak or sometimes multiple peaks in the
power/frequency spectrum corresponding to the lateral
periodicity of the striations. For some micrographs two or
more separate spacings were evident. Striation spacings
were variable both between different micrographs for the
same species and between hairs from different species with
no particular trend being evident. A key point, however,
was that the spacings of the striations always fell within the
range 0.290.39 mm. There seems little doubt therefore
that surface striations are a common feature of the
undamaged (i.e. root-end) hairs of all mammalian species.
Furthermore, the narrowness in the range of their lateral
spacings would seem to support a common process for their
formation.
It was particularly interesting to find striations on the
hairs of echidna (the short-nosed spiny anteater, Tachyglos-
sus aculeatus) and platypus (Ornithorhynchus anaticus) (cf.
Fig. 2d). These animals, from the mammalian order
Monotremata of primitive egg-laying mammals, are the
only surviving members (two species of spiny anteater and
the platypus) in the subclass Prototheria. The Prototherians
are thought to have diverged from Eutherian mammals at
least 150 million years ago (Griffiths, 1988; Peet et al.,
1992). Whereas all Eutherian mammals seem to have 18-
MEA (a C-21 fatty acid) as the predominant lipid covalently
attached at the surfaces of their hairs, in monotreme hairs
the bound fatty acids are of mainly 26 carbon atoms
(Rismiller & Seymour, 1991; Peet et al., 1992; Jones &
Rivett, 1997). Notwithstanding this biochemical diver-
gence, hair surface striations are present in both groups.
This highlights the likelihood that, if the striations serve
some beneficial purpose, this has been sustained since at
least the early Cretaceous period.
In samples of human hair exhaustively extracted with
chloroform/methanol, and known to remove all non-
covalently bound fatty acids (Negri et al., 1993; Jones &
Rivett, 1997), striations were still in evidence, albeit with
some slight reduction in visibility (Fig. 3). This establishes
that the striations are not due to agglomerates of loose
lipids. Yet further, in samples treated with methanolic KOH,
a reagent known to remove the covalently bound fatty acids
from the fibre surface (Negri et al., 1993; Jones & Rivett,
1997), striations were still in evidence (Fig. 4). This
provided strong support for the notion that the immediate
underlying proteinaceous component of the cuticle (the
epicuticle) has a striated surface and that the bound fatty
acids are merely conforming to this geometry. That
striations were not seen on freshly exposed intercuticular
surfaces indicates that the epicuticle's surface exhibits its
corrugated form only where the cuticle was exposed at the
original fibre surface. Specific interaction of the epicuticle
with components in the hair canal and/or earlier contact
with peripheral cells in the follicle (i.e. inner root sheath
cuticle) may thus be the cause of the striations.
q 2001 The Royal Microscopical Society, Journal of Microscopy, 204, 203211

Fig. 3. AFM image of root-end human head hair exhaustively
extracted with chloroform/methanol. Striations are not removed
by this solvent extraction. Micrograph edge dimensions 5.0 mm.
TEM observations
The cutting of transverse sections of keratin fibres to gain
satisfactory TEM observation of the boundary between the
fibre and the resin is normally fraught with difficulty and
was the source of some problems for the present work. It
was important to use an embedding resin that was not
chemically aggressive and did not swell on polymerization
(i.e. not an acrylic). Given limited penetration of most
monomers into untreated hairs, it was also essential to use a
resin of low initial viscosity. Spurr's resin was therefore our
chosen embedding medium. Limited keying between the
resin and the fibre unfortunately was the source of a major
problem as sections reached the water reservoir behind the
ultramicrotome knife or as they were treated with aqueous
stains. Keratin fibres swell diametrically by approximately
15% in water and, because the resin itself does not swell, the
induced stresses are relieved by the sections splitting along
the poorly keyed resin/fibre interface. Only for fibres of
diameter less than 20 mm (cat's fur and sheep's wool) was it
possible to obtain sections where an intact fibre/resin
interface was preserved around the entire perimeter. For
fibres of greater diameter peripheral splitting always
occurred but, nevertheless, there was usually a small
portion where the resin/fibre boundary remained suffi-
ciently intact for appropriate TEM examination.
The perimeters of all the hairs were intensely stained by
silver (Fig. 5) and there seems little doubt that this defines
the outer bound of the cuticle's proteins. A monolayer of
lipid (18-MEA) is known to be attached via a thioester
q 2001 The Royal Microscopical Society, Journal of Microscopy, 204, 203211
EPICUTICLE OF MAMMALIAN HAIRS 207
Fig. 4. AFM image of root-end human head hair extracted with
chloroform/methanol, then treated with methanolic KOH and re-
extracted with chloroform/methanol. Striations are not removed
by this treatment that removes the surface layer of covalently
bound fatty acid. Micrograph edge dimensions 5.0 mm.
linkage to cysteine residues in the proteins underlying the
hair's immediate surface (Negri et al., 1993). The outwardly
presenting hydrocarbon tails of these lipids were sufficiently
inert not to be stained by silver. This layer therefore
exhibited no contrast relative to the adjacent unstained
embedding resin for it to be rendered separately visible
under the TEM by the present methods of specimen
preparation. As the thioester groups are stained by silver
(see later), the aforementioned boundary must therefore
represent the molecular interface between the surface lipids
and the cuticular proteins.
The perimeters of all the hairs had a gently undulating
profile but lacked the periodicity expected for transverse cuts
through the surface striations, an anomaly highlighted in
more detail in a recent paper by the authors (Swift & Smith,
2000b). There was variation neither in the thickness nor in
the intensity of staining of any of the underlying layers,
which might have correspondence with the surface stria-
tions. One surmises that the expected periodicity would only
be seen in sections precisely aligned for end-on projection of
the striations, i.e. with the longitudinal axis of the striations
exactly parallel to the imaging axis of the TEM. In none of
the present work would such alignment seem to have
prevailed. Further work using a TEM stage goniometer
would be necessary to resolve this issue.
In all the hairs examined a continuous layer of very high
staining intensity and of constant thickness was seen on the
208 J. A . S W I F T A N D J. R . S M I T H
Fig. 5. Transverse section of domestic cat hair close to the fibre
surface. The section was stained with ammoniacal silver nitrate.
This particularly shows the new layer (E) tentatively identified as
being the epicuticle. A 2 A-layer, CMC 2 cell membrane complex,
En endocuticle, Ex exocuticle, I inner layer, R embedding sufficient heavy metal to render the microfibrillar pattern of
resin.
outer-facing aspect of every cuticle cell, both where a given
cell was overlapped by another cuticle cell and where it gave
onto the fibre's outer surface (Fig. 5). The layer, which
seems not to have been identified previously, often presented
a sharply-defined and straight interface with the immedi-
ately adjacent A-layer. Its thickness, both within and
between the hairs of animal species, was approximately
13 nm. This thickness is consistent with the 14 nm
maximum thickness reported for isolated Allworden mem-
branes (Lindberg et al., 1948; Lagermalm, 1954; King &
Bradbury, 1968). Allworden membranes as little as 9 nm
thickness have been reported by these workers. One
presumes that these lesser thicknesses have arisen by
material loss under the aggressive conditions of chlorination
and by drying shrinkage. Interestingly, the thickness at
13 nm is sufficiently close to that at 15.3 nm for the
isopeptide- and cystine-rich insoluble envelopes of cornified
cells (Jarnik et al., 1998) for one to believe similarity of
purpose and of synthesis for these layers. There seems little
doubt that the new layer identified by the present work is
the epicuticle. For convenience its presence has been
included in Fig. 1.
A continuous layer of high staining intensity was also
observed in all fibres at the inner-facing perimeter of every
cuticle cell, i.e. between the endocuticle and the unstained
cell membrane complex (CMC) (cf. Fig. 5). Its thickness was
between 6 and 14 nm and most of this variation was
accommodated by irregularity at its interface with the
endocuticle, the interface with the CMC being of relatively
straight profile. This particular layer has been recognized
many times previously and is generally known as the `inner
layer'; this name was coined by Rogers (1959) following his
pioneering TEM investigations of wool. Whilst the morphol-
ogy of the inner layer is slightly different from that of the
epicuticle, their locations close to the cell surface and
similar staining behaviours lead one to believe that they are
of similar chemical composition. Furthermore, they are
likely to be different parts of an envelope covering the entire
intracellular periphery of each cuticle cell in a fashion
similar to that found for the CE of squamous epithelial cells.
This is consistent with Leeder (1986), who found that
Allworden sacs are produced over the entire surface of
isolated cuticle cells.
Fortuitously, it would seem that staining of fibre sections
with ammoniacal silver for as little as 20 min has enabled
the present identification of the epicuticle. Although the
internal structures of keratin fibres have been the subject of
many previous TEM studies, this has usually been of
sections of hairs stained before embedding. An additional
aim in most of this other work was to load the samples with
the fibre cortex readily visible. Under such conditions one
surmises that so much stain has been introduced into the
cystine-rich A-layer that it overwhelms the adjacent
epicuticle thereby making it difficult, if not impossible, to
discriminate between them. The present staining conditions
have evidently overcome this difficulty.
We believe that micrographs in a paper by Jones et al.
(1994) illustrate some of the stages of formation of the
epicuticular layer in the wool follicle. The paper makes no
specific mention of the epicuticle but of a `developing
exocuticle layer'. The location and thickness of this latter
layer match those of the new layer found in the present
work on fully formed hairs. Thus it would seem that Jones
et al. (1994) had captured images also showing develop-
ment of the epicuticle. Changes in the structure of the
cuticle during its final hardening process occur with such
rapidity as to make it difficult to capture the precise
sequence of events. Despite these reservations, the micro-
graphs of Jones et al. (1994) indicate that the epicuticle is
laid down at about the same time as the membranes
between the fibre cuticle and the inner root sheath cuticle
are starting to degenerate. The epicuticle seemingly makes
its distinctive and separate appearance at the outer
intracellular boundary of each cuticle cell at a stage where
a significant proportion of the cell's volume is occupied by
complex folds of presumptive cysteine-rich material. Imme-
diately preceding this event the region between the complex
folds and the cell boundary is relatively electron transpar-
ent, indicating that cysteine enrichment of the epicuticle
must be occurring very rapidly. Once the epicuticle has been
laid down the complex underlying folds evidently differ-
entiate very rapidly to yield adjacent homogeneously
q 2001 The Royal Microscopical Society, Journal of Microscopy, 204, 203211

structured A- and exocuticle-layers of the type seen in the
fully formed fibre.
The intensity of ammoniacal silver staining of the various
components in sections of keratin fibres was roughly in
direct proportion to the amount of protein-bound cystine
they contain. Thus, in the fibre cuticle the intensity of silver
staining increased in the order: cell membrane complex
(very low intensity), endocuticle, exocuticle, A-layer (very
intense). This accords reasonably well with separate
chemical analyses for all but the cell membrane complex,
where no reliable figure is available, in which their 12-cystine
contents in moles/100 mol of protein are 5.7 (Swift & Bews,
1974), 25.1 (Swift & Bews, 1974) and 37.5 (Swift, 1979),
respectively. Although the precise chemistry of the staining
reaction remains uncertain, it is likely to involve electro-
static binding of the complex silver ammine present in the
staining solution, [Ag(NH3)2]1, to thiolate (S2) and/or to
perthiolate (S-S2) anions formed during the alkaline
degradation of cystine (Maclaren & Milligan, 1981). An
anomaly for such a scheme is that the newly identified
epicuticle stains even more intensely than the adjacent A-
layer (cf. Fig. 5). As the proteins in the first 100 nm depth of
the hair surface (i.e. mainly A-layer) contain 37.5% of their
amino acid residues as 12-cystine (Swift, 1979), and in itself a
truly remarkable concentration, it is hard to imagine the
presumptive epicuticle containing cystine at an even higher
concentration. Ammoniacal silver staining of this compo-
nent might not therefore arise from cystine disulphide
crosslinks alone. A possibility to be considered is that the
epicuticle contains fatty acids (such as 18-MEA) covalently
linked to cysteine residues within its protein framework, as
well as linked as a monolayer at its outer surface (Jones &
Rivett, 1997). Under the alkaline staining conditions it
seems likely that the thioester linkages will be hydrolysed to
yield thiolate anions, then capable of electrostatic binding
with silver ammine cations:
Protein-CH2 -S-CO-R ! Protein-CH2 -S2
! Protein-CH2 S2 AgNH3 2 1
If thioesters (such as those of 18-MEA) are indeed present
in the epicuticle, the observed rapid staining of this
structure by ammoniacal silver might be taken to indicate
that cleavage of the thioesters occurs significantly faster
than the breakdown of the cystine disulphides. In part
support of these ideas it is notable that E. Hoffmann & A.
Korner (personal communcation, 2000) have found ammo-
niacal silver nitrate to be particularly efficient at cleaving
covalently bound 18-MEA from wool. They used gas
chromatography/mass spectroscopy to measure the amount
of bound 18-MEA in wool. 650 mg g21 of bound 18-MEA
was found in untreated wool, as against 207 mg g21 after
treating the wool with 0.1 m ammoniacal silver nitrate at
pH 9.8 for 2 h and 315 mg g21 after treating for 2 h with
q 2001 The Royal Microscopical Society, Journal of Microscopy, 204, 203211
EPICUTICLE OF MAMMALIAN HAIRS 209
ammonia at pH 11.2. Yet further support for the notion
that the epicuticle contains thioester-linked lipids is
provided by the work of Dauvermann-Gotsche (1999). She
treated wool with hydroxylamine (this cleaves off thioester-
linked fatty acids to leave a thiol) and then with
monomaleimido-nanogold (thiols now specifically labelled
with small gold clusters). In sections of the fibres examined
in the TEM she found the gold label to be in a layer up to
20 nm thick close to the immediate fibre surface. Her
results, taken together with the present ones, highlight the
distinct possibility of thioester-linked lipids not only being
attached to the epicuticle's outer surface but also contained
within its bulk.
Multifarious benefits to the fibre and to the animal's
welfare would seem to be conferred by the epicuticle. On
account of the isopeptide crosslinkages it contains, the
epicuticle will undoubtedly protect the fibre surface from the
chemical insult of natural weathering processes. Together
with the underlying A-layer, which is very highly cross-
linked by cystine, it will provide resistance to external
mechanical insult and in addition afford a stable platform
for the layer of lipids covalently attached at its outer surface.
The surface striations of the native undamaged fibre, which
are imprinted on the epicuticle, may offer further benefits.
By reducing the area of interaction for frictional contact
with other surfaces and in the presence of a lubricating lipid
overlayer, the striations will ensure a low coefficient of
friction for the epicuticular surface. All mammalian keratin
fibres exhibit directional dependence of friction in which
frictional resistance is greater for tip-to-root rubbing events
(`against scales') than for root-to-tip rubbing (`with scales').
Such differential friction in the fibres has the major benefits
for the animal of ensuring maximum disentangled align-
ment of the fibres and of aiding in the migration of skin
surface detritus to the surface of the pelage (Swift, 1999).
By effecting low with-scales friction, and concomitantly
contributing to a high frictional differential, the surface
striations would seem to make an important contribution
towards the aforementioned benefits. It is pertinent to
mention in these respects that surface striations, a scale-
like surface and directionally dependent friction are key
features governing the reptation locomotion of some snakes
(Hazel et al., 1999). A further benefit of the epicuticular
striations might also be in reducing the effective area of
contact with water and thereby increasing the apparent
hydrophobicity of the surface. This may even contribute to
a process of self-cleaning by water of the animal's hair by
the so-called `lotus effect' (Barthlott & Neinhuis, 1997), in
which droplets of water roll off the corrugated hydrophobic
surface carrying hydrophilic detritus with them. Were such
an effect to occur in practice, it would be seriously
impaired by the adsorption of surfactants to the fibre
surfaces (e.g. in human hair washed with contemporary
shampoos).
210 J. A . S W I F T A N D J. R . S M I T H
Acknowledgements
J.A.S. thanks Deutsches Wollforschungsinstitut (DWI),
Aachen, Germany, for providing the facilities for him to
undertake the TEM work, and Wella AG of Darmstadt,
Germany, for financial support. He is grateful to Dr Andrea
Korner of DWI for having suggested we examine fur from
the monotremes and, with Fraulein Erika Hoffmann, for
having made available to us analyses of the 18-MEA
content of wool treated with ammoniacal silver nitrate.
J.R.S. wishes to thank the Royal Society, the Royal Society of
Chemistry and the Royal Microscopical Society for financial
support. We also wish to thank London Zoo for providing
the hair sample from the Asiatic lion, Dr Warren Bryson of
the Wool Research Organization of New Zealand for the
fibres from Corriedale and Lincoln sheep, alpaca, echidna
and platypus, and finally Dr Chris Gummer of Proctor and
Gamble, Egham, Surrey, UK, for the sample of guinea-pig
hair.
References
Allen, C.F. (1990) The composition and structure of the membrane
released from the wool fibre by chlorination. PhD Thesis,
University of Leeds, UK.
Allen, C.F., Dobrowski, S.A., Speakman, P.T. & Truter, E.V. (1985)
Evidence for lipid and filamentous protein in Allworden
membrane. Proceedings 7th Internat. Wool Textile Res. Conf, Tokyo,
1, 143151.
Allworden, I.K. (1916) The nature of sheep's wool and a new
research method for the chemical identification of damaged
wool. Z. Angew. Chem. 29, 7778.
Barthlott, W. & Neinhuis, C. (1997) Purity of the sacred lotus or
escape from contamination in biological surfaces. Planta, 202,
18.
Bradbury, J.H. & Leeder, J.D. (1972) Keratin fibres. Part 4.
Mechanism of the Allworden reaction. Aust. J. Biol. Sci. 25,
133138.
Dauvermann-Gotsche, C. (1999) Characterisation of 18-methyl
eicosanoic acid containing proteolipids from wool. PhD Thesis.
Rheinisch-Westfalischen Technische Hochschule, Aachen,
Germany.
Folk, J.E. & Finlayson, J.S. (1977) The epsilon-(gamma-glutamyl)
lysine crosslink and the catalytic role of transglutaminase. Adv.
Protein Chem. 31, 1133.
Griffiths, M. (1988) The platypus. Sci. Am. 258, 6067.
Hazel, J., Stone, M., Grace, M.S. & Tsukruk, V.V. (1999) Nanoscale
design of snake skin for reptation locomotions via frictional
anisotropy. J. Biomechan. 32, 477484.
Herbig, W. (1919) The elastikum reaction of wool. Z. Angew. Chem.
32, 120.
Jarnik, M., Simon, M.N. & Steven, A.C. (1998) Cornified cell
envelope assembly: a model based on electron microscope
determinations of thickness and projected density. J. Cell Sci.
111, 10511060.
Jones, L.N., Horr, T.J. & Kaplin, I.J. (1994) Formation of surface
membranes in developing mammalian hair fibres. Micron, 25,
589595.
Jones, L.N., Peet, D.J., Danks, D.M., Negri, A.P. & Rivett, D.E. (1996)
Hairs from patients with maple syrup urine disease show a
structural defect in the fibre cuticle. J. Invest. Dermatol. 106,
461464.
Jones, L.N. & Rivett, D.E. (1997) Role of 18-methyleicosanoic acid
in the structure and formation of mammalian hair fibres.
Micron, 28, 469485.
King, N.L.R. & Bradbury, J.H. (1968) The chemical composition of
wool. Part 5. The epicuticle. Austral. J. Biol. Sci. 21, 375384.
Lagermalm, G. (1954) Structural details of the surface layers of
wool. Textile Res. J. 24, 1725.
Leeder, J.D. (1986) The cell membrane complex and its influence
on the properties of the wool fibre. Wool Sci. Rev. 63, 335.
Leeder, J.D. & Rippon, J.A. (1985) Changes induced in the
properties of wool by specific epicuticle modification. J. Soc.
Dyers Col. 101, 1116.
Lindberg, J., Philip, B. & Gralen, N. (1948) Occurrence of thin
membranes in the structure of wool. Nature, 162, 458459.
Maclaren, J.A. & Milligan, B. (1981) Wool Science: the Chemical
Reactivity of the Wool Fibre. Science Press, Marickville, N.S.W.,
Australia.
Negri, A.P., Cornell, H.J. & Rivett, D.E. (1993) A model for the
surface of keratin fibres. Textile Res. J. 63, 109115.
Negri, A.P., Rankin, D.A., Nelson, W.G. & Rivett, D.E. (1996)
Transmission electron microscope study of covalently bound
fatty acids in the cell membranes of wool fibres. Textile Res. J. 66,
491495.
Oku, H., Yagi, N., Nagata, J. & Chinen, I. (1994) Precursor role of
branched-chain amino acids in the biosynthesis of iso and
anteiso fatty acids in rat skin. Biochim. Biophys. Acta, 1214,
279287.
Peet, D.J., Wettenhall, R.E.H., Rivett, D.E. & Allen, A.K. (1992) A
comparative study of covalently-bound fatty acids in keratinised
tissues. Comp. Biochem. Physiol. 102B, 363366.
Phan, K.H., Thomas, E. & Heine, E. (1995) Structure of the cuticle
of fine wool fibres. Proceedings 9th Internat. Wool Textile Res. Conf,
Biella, Italy, 2, 1930.
Prottey, C. & Ferguson, T.F.M. (1972) Measurement of lipid
synthesis in mouse auricular skin cultured in vitro. Br. J.
Dermatol. 87, 475495.
Ramanathan, N., Sikorski, J. & Woods, H.J. (1955) Electron
microscope studies of the surface structure of wool and other
fibres. Biochim. Biophys. Acta, 18, 323340.
Rismiller, P.D. & Seymour, R.S. (1991) The echidna. Sci. Am. 264,
8087.
Rogers, G.E. (1959) Electron microscopy of wool. J. Ultrastruct. Res.
2, 309330.
Schwan, A. (1981) On the enzymatic isolation and chemical
analysis of the cell membrane complex from wool. PhD Thesis,
Rheinisch Westfalischen Technische Hochschule, Aachen,
Germany.
Spurr, A.R. (1969) A low-viscosity epoxy resin embedding medium
for electron microscopy. J. Ultrastruct. Res. 26, 3143.
Steinert, P.M. & Marekov, L.N. (1995) The proteins elafin, filaggrin,
keratin intermediate filaments, loricrin, and small proline-rich
proteins 1 & 2 are isodipeptide cross-linked components of the
q 2001 The Royal Microscopical Society, Journal of Microscopy, 204, 203211

human epidermal cornified envelope. J. Biol. Chem. 270, 17702
17711.
Swift, J.A. (1979) Minimum depth electron probe X-ray micro-
analysis as a means for determining the sulphur content of the
human hair surface. Scanning, 2, 8388.
Swift, J.A. (1999) Human hair cuticle: biologically conspired to the
owner's advantage. J. Cosmet. Sci. 50, 2347.
Swift, J.A. & Bews, B. (1974) The chemistry of human hair
cuticle. Part 2. The isolation and amino acid analysis of the
cell membranes and A-layer. J. Soc. Cosmet. Chem. 25, 355
366.
q 2001 The Royal Microscopical Society, Journal of Microscopy, 204, 203211
EPICUTICLE OF MAMMALIAN HAIRS 211
Swift, J.A. & Smith, J.R. (2000a) Atomic force microscopy of
human hair. Scanning, 22, 310318.
Swift, J.A. & Smith, J.R. (2000b) Surface striations on human hair
and other mammalian keratin fibres. Proceedings of the 10th
Internat. Wool Textile Res. Conference, Aachen HH-1, 110.
Wolfram, L.J. (1972) Topography of some cuticle cells. Textile Res. J.
42, 252254.
Zahn, H., Wortmann, F.-J. & Hocker, H. (2000) Chemical compo-
sition of the hair cuticle: cornified envelope (CE) proteins in
epicuticle and A-layer. Poster 9, p. 35. Abstracts, Annual Meeting
of the European Hair Research Society (EHRS), September 1517.