Histological Study of Sheep Skin Transformation in the Course of Parchment Manufacturing Process

We report a simple histological study on skin biopsies from young domestic sheep following each step in transformation from skin to parchment production. Histological analyses were conducted; before and after lime treatment, hair removal, and stretching. Sections were �xed and stained using a variety of histological stains to identify the presence of different molecular classes and the �brous proteins, collagen and elastin. The results reveal surprisingly few histological changes in most steps in the production process. However, very visible changes in the supramolecular ordering of skin macromolecules (elastin, collagen) occur during the �nal stage of parchment production when stretched on the frame. Collagen �bres and hair follicles were all strongly re-oriented in the direction of strain. Surprisingly, despite the thinness of the lambskin and the exhaustive treatment in lime, not all fats were saponi�ed and even in the �nal product Oil Red O stained fat bodies were detectable on the hair side of the skin.

Recent research [15] has revealed that sheep skins have been widely employed for parchment production throughout history, typically for legal documents.In this article, we explore the histological changes that occur to fresh sheep skins during the different stages employed in parchment production.Using an old manufacturing recipe [16], and by studying the impact of changes in the ingredients and processes, we hope to better interpret the results of previous studies of old manuscripts.
There have been few previous histological studies of parchment [e.g.13,[17][18][19] and none have followed the process of parchment manufacture.By documenting how each stage is processed and by taking successive samples, it will be possible to correlate with the non-invasive procedures currently used to decipher the production of parchment (SEM, XRF, Raman, etc) from unknown historic production methods.The resulting data from experimental manufacture can then be compared with non-invasive methods applied to historical material [15,20,21], as it was in a recent article [22] in which the authors conclude that "the skin studied was rst tawed with alum and then dressed with gypsum".

Skin structure
Sheepskin, like the skin of other mammals, is divided into two layers: the (outer) epidermis and the (inner) dermis (Figure 1).The epidermis is an epithelial tissue made mainly by cohesive cells called keratinocytes.The differentiation of these cells generates four classical layers in the epidermis: stratum basale, stratum spinosum, stratum granulosum and stratum corneum.
The sublining dermis is a connective tissue rich in collagen bers.Invaginations of the epidermis form hair follicles in the dermis, with their attached sebaceous glands, containing sebocytes rich in lipids.Smooth muscle cells from the hair erector muscle attach the hair follicle to the epidermis.
The greasiness that is noted in the processing of sheep skins stems from the distinct distribution of fat speci c to sheep because of the abundance of secondary follicles in ne wooled sheep, to which are attached sebaceous glands [23].The lipids produced and secreted by these glands (notably lanolin) generate a lipid layer below the epidermis separate from the usual subcutaneous fat.The greater the number of follicles, the more lipids that will be produced.It is this additional lipid layer that contributes to the 'delaminating' effect observed in sheep parchment, where the skin can be effectively split into two thinner layers.

Parchment preparation
The skins from four lambs (Ovis aries) were taken from the Ovine Center of the University of Namur (natural deaths); the animals were cared for according to procedures conforming to the European requirements on farm animals (EC directive 86/609).The skins -two from stillborn animals and two from animals less than 2 weeks old (approximately a rectangular shape and size of 16 cm by 20 cm or lower) were manufactured into parchment (Figure 2).Recipes for parchment production exist from the 8thC [14], the rst well documented example being John Beale in "The Art of making Parchment, Vellum, Glue etc" read to the Royal Society in 1664 [24].Here, we used the later description of Diderot & D'Alembert [16], see Box 1.

Preparation
Fresh skins were washed in water then left in a light lime solution (3g/L) for ve days.

Dehairing
Two dehairing processes were used (each treatment was applied on skins from one born dead animal and from one animal less than 2 weeks old), both at ambient temperature, either: 1. the esh side of a skin was uniformly covered with a milky lime solution and left for one week folded in two, esh side together (pancake method), or 2. the full skin was put into a lime bath and left for one week (bath method).Skins were dehaired using successively stronger lime Ca(OH) 2 solutions (5 g Ca(OH) 2 /l, 15 g/l and 30 g/l), equivalent to 'plein mort', 'plein gris' and 'plein vif' for 1 week each, with a wash in clean water for one day in between each different solution.

Scraping
The hair was removed by rubbing the skin with hands and nished with a semi lunar knife.The skins were then put on a framed, scraped, chalked and pounced, to produce the nal parchment.

Diderot & D'Alembert (1765)
Once the skins are peeled, they are washed in the river, then allowed to drain for some time; next they are put into 'dead-lime' that is to say, lime which has lost almost all its strength.Skins are left in baths of dead-lime for about twenty-four hours; They are removed and allowed to fully drain again.
Two days after the skins are removed from the dead-lime, they are immersed in another bath with less aged lime and left there for about two or three days with agitation, after which they are removed put to drain… If necessary, they are then plunged back in the lime bath again, this operation is repeated for between six weeks to two months during the heat of summer, but in winter in the open it is necessary for the skins to soak for at-least three months.

Loose translation, Matthew Collins
Two subsamples of skins were subsampled at each step of the manufacturing processes described by Diderot & D'Alembert [16] and placed immediately in formaldehyde for subsequent histological analyses: one for the para n embedding and the other one for the frozen section to enable alternative histochemical staining to be used.Samples of approximately one square centimeter were taken in periphery of the skin, after speci c steps of the manufacturing process (Figure 2): fresh skin (sample #1), skin after hair removal (sample #2), skin after the 3 rd lime bath (sample #3), parchment (sample #4).

Method: Histological procedures
The sheepskin and parchment samples were xed in aceti ed formalin (formol 4%, acetic acid 1%).Six-micron thick sections of the para n and of the frozen blocks were placed on SuperFrost + glass slides.
Sections from the para n blocks were dewaxed, rehydrated in graded alcohols and stained using the ve following methods (the two rst being generalized topographic stains, while the three others are speci c for speci c macromolecules): 1) Hemalun, Erythrosin and Saffron stain (HES) stains nucleic acid blue/purple, cytoplasmic proteins red and collagen yellow-orange; 2) Green trichrome stains nucleic acid in blue/purple, cytoplasmic proteins red and collagen green; 3) Orcein and hematoxylin stains elastin bers red brown and nuclei blue/purple; 4) Picro-Sirius Red and hematoxylin stains collagen I bers and nuclei blue/purple; 5) Wilder and green light stains reticulin bers, mainly collagen III ber, black with a green background.
Sections from the frozen blocks were stained with Oil Red O (ORO) and hematoxylin which stains lipids red and nuclei blue.

Results And Discussion
From a histological point of view, samples taken from the four skins shared similar changes during parchment production; there were no signi cant differences between stillborn and two week old animals.Nor did the two different dehairing processes (i.e.pancake method versus bath method) result in any histological differences in the appearance of tissues.

Pre-treated skins
Histological sections of the pre-treated skin (samples #1; Figure 3) exhibited typical histological features, expected from fresh skin.The density of the hairs is highest in the grain layer, where they occupy at least half the total thickness of the skin.We did not observe any histological structure related to the subcutaneous tissue.This may be due to our sampling at the periphery of the skin (given the unevenly distributed layer thickness over the body), the (young) age of the animal and/or the way the skin was removed.
Orcein staining shows that the elastin bers are preferentially located in the deep dermis, close to the (missing) subcutaneous layer and are more abundant around hair follicles.Picrosirius stain highlights the ubiquitous presence of collagen everywhere in the dermis in abundant bundles.The Wilder stain shows presence of reticulin bers underlining the basement membranes, and perhaps some sparse bers in the dermis.The speci c staining of ORO reveals the presence of fats in the sebocytes of the sebaceous glands con rming previous observations by Haines [12].Fats are present discreetly in the stratum granulosum and rarely in the stratum corneum.

Dehaired skins
The second samples were taken from the skin after light liming and hair removal (Figure 4).It is known that after soaking in lime (Ca(OH) 2 ), the dermal ber network in the skin expands [25].We observed the loss of hairs, whose roots only persist here and there in the stratum basale of the epidermis (Figure 4).All the stained components observed in the pre-treated skins were all still present after dehairing.

Prolonged liming
After the liming processes (three baths of increasing lime concentration; Figure 5), collagen bers seem now to have lost the characteristic brillar structure and appear to form a homogenous mass.Maxwell et al. [26] observed that "after salting there is an increase in the distance between the collagen molecules in the hydrated state, and even more expansion after liming, which is carried through the hierarchical levels up to bril packing; deliming did not appear to reverse these increases, indicating that the effects of salting and liming are permanent".Elastin and reticulin bers seem to be present only in traces.We observed that, despite remaining for three weeks in the lime solution, fat cells were still present, albeit most likely in a lower concentration (Figure 5).The ability to remove fats from the skin is known to be related to duration in the lime bath (alkali causing saponi cation) and to the freshness (alkalinity) of the solution [27]; this aspect is worthy of further investigation.

Final parchment
The fourth samples were taken from the fully processed skin, i.e. the nal parchment.During this nal manufacturing step, the skin was put on a frame, scraped, bleached, cleaned and polished under tension.
In the parchment samples (Figure 6), the bulk of the material took the form of a two dimensionally structured layer of uniformly oriented bers.The rest of the glands and hair follicles tend to reorient horizontally in the direction of strain.Somewhat surprisingly -despite the fact that skins were taken from young or born dead animals and the nal parchment was very thin < 0.1 mm -the production process did not remove all the fat cells (i.e.saponify all of the lipid).Oil red O staining revealed residual lipid deposits on the hair side of the skin.
The major histological changes observed at the different stages of skin transformation are revealed by the result of the six methods applied on samples and summarized in Table 1.

Persistence of lipids
From an optimal tissue preservation standpoint, it is assumed that high levels of lipid dripping from heated parchment initiates combustion of the membranes.Ghioni et al [28] report variable, but in some cases, substantial levels of lipid in historical and modern parchments.They noted a correlation between higher levels of lipid and greater degradation of the collagen, and argue that lipid peroxidation which has been documented in living tissues, may generate free radicals causing the observed reduction in level of collagen integrity as measured by X-ray diffraction.However, a more likely explanation is the production process itself, by necessitating longer exposure to lime (to saponify the lipids) and therefore compromising the structural integrity of the collagen.
The problem of residual grease bedevilled parchment makers in the past.Saxl [14] writes that "Sheep parchments have often a high grease content in the grain layer"."Magister Bernardus" Bodleian Library Oxford.Canonici Misc.128 recipe for parchment production deals with the subject of the removal of grease with warm dusting powders of chalk or other calcium salts or of wood ashes which melted the grease and acted like a "dubbin"; penetration was brought about by compressing the parchment.
Saxl's analysis of parchment recipes revealed that "degreasing was extensively practised in Italy in the fourteenth and fteenth centuries and in France", and the practice was still being used by Messrs. Russell of Hitchin using a mixture of wood ash and soda added to the stretched (sheep) parchment on the frame which was then heated in a warm stove.
We conducted species identi cation [15] on six of the samples analysed by Ghioni et al [28], and in support of Saxl con rmed that the two membranes with high levels of lipid were from sheep (Figure 7).The production of ne wool has resulted in the selection of breeds with larger numbers of secondary follicles, resulting in a greater density of sebaceous glands, and thus sebocytes in sheep when contrasted with goat-and calf-skin.This results in both the higher levels of residual lipid and also, due to the disruption in the dermal collagen network, a tendency for sheepskin to delaminate (that is for the parchment to split).
Parchment maker Jessie Mayer (Pergamena, pers comm.2015) concurs, and notes the di culty of preparing parchment from wool sheep skins, due the excessive levels of lipid, which are not removed by the liming process.

Conclusions
Our histological study shows the modi cation of the biological material and more speci cally collagen bers in the course of a parchment manufacturing experiment starting from skins of immature sheep.In particular, the modi cation and, ultimately, the disappearance of collagen type III bers was observed (using the Wilder staining method) in addition to a progressive loss of elastin bers.Another interesting observation is the persistence of fats, despite the decrease in quantity of the material stained with oil red O (so most likely lipids) as well as the genetic material observed with hemalaun staining.
Whilst we were unable to demonstrate any histological differences caused by the alternative dehairing methods, the manufacture of parchment itself was shown to induce dramatic histological changes from the fresh skin.The transformation involved a progressive evolution from a classical full and complex three dimensional framework of biological material to a lamellar structure essentially composed of co-aligned collagen bers.The most dramatic changes happen during the processes of drying and stretching on a frame.
During this phase, the thickness of the skin reduces and the collagen brils realign in the direction of the applied force.

Declarations Figures
Comparative       Species identi cation of samples analysed for lipid content [28].Sheepskin parchment has the highest levels of lipid, consistent with the histological investigation reported here (see Fig. 5).

Supplementary Files
This is a supplementary les associated with this preprint.Click to download.

Figure 3 Section
Figure 3

Figure 4 Skin
Figure 4

Table 1
The major histological changes observed at the different stages of skin transformation