Content uploaded by Andrew Schulz
Author content
All content in this area was uploaded by Andrew Schulz on Aug 15, 2023
Content may be subject to copyright.
Second Harmonic Generation Imaging Reveals1
Entanglement of Collagen Fibers in the Elephant2
Trunk Skin Dermis3
Andrew K. Schulz1, Magdalena Plotczyk2,+, Sophia Sordilla3,+, Madeline Boyle1,
Krishma Singal4, Joy S. Reidenberg5, David L. Hu1,3,∗, Claire A. Higgins2,∗
Schools of Mechanical Engineering1, Biological Sciences3, and Physical Sciences4
Georgia Institute of Technology, Atlanta, GA 30332, USA
Department of Bioengineering2
Imperial College London, South Kensington, London, UK
Center for Anatomy and Functional Morphology4
Icahn School of Medicine at Mount Sinai, New York, NY 10029-6574, USA
4
August 11, 20235
Corresponding authors:6
c.higgins@imperial.ac.uk7
hu@me.gatech.edu8
9
Keywords:10
cross-linking, skin layers, microscopy, morphology, composition, comparative materials, elephant11
skin12
Abstract13
Form-function relationships often have tradeoffs: if a material is tough, it is often inflexible, and14
vice versa. This is particularly relevant for the elephant trunk, where the skin should be protective15
yet elastic. To investigate how this is achieved, we used classical histochemical staining and second16
harmonic generation microscopy to describe the morphology and composition of elephant trunk17
skin. We report structure at the macro and micro scales, from the thickness of the dermis to the18
interaction of 10 µm thick collagen fibers. We analyzed several sites along the length of the trunk, to19
compare and contrast the dorsal-ventral and proximal-distal skin morphologies and compositions.20
We find the dorsal skin of the elephant trunk can have keratin armor layers over 2mm thick, which21
is nearly 100 times the thickness of the equivalent layer in human skin. We also found that the22
structural support layer (the dermis) of elephant trunk contains a distribution of collagen-I (COL1)23
fibers in both perpendicular and parallel arrangement. The bimodal distribution of collagen is seen24
across all portions of the trunk, and is dissimilar from that of human skin where one orientation25
dominates within a body site. We hypothesize that this distribution of COL1 in the elephant26
trunk allows both flexibility and load-bearing capabilities. Additionally, when viewing individual27
fiber interaction of 10 µm thick collagen, we find the fiber crossings per unit volume are five times28
more common than in human skin, suggesting that the fibers are entangled. We surmise that29
these intriguing structures permit both flexibility and strength in the elephant trunk. The complex30
nature of the elephant skin may inspire the design of materials that can combine strength and31
flexibility.32
1
.CC-BY-NC 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 15, 2023. ; https://doi.org/10.1101/2023.08.11.553031doi: bioRxiv preprint
Introduction33
Elephant trunks, octopus arms, and mammalian tongues are the three canonical examples of mus-34
cular hydrostats (Kier and Smith, 1985). The elephant trunk, the subject of this work, is extremely35
flexible and can extend by up to 25% in a telescopic manner allowing the elephant to reach distant36
objects (Schulz, Boyle, Boyle, Sordilla, Rincon, Hooper, Aubuchon, Reidenberg, Higgins, and Hu,37
2022). The ventral side of the trunk contains oblique muscles that allows that part of the trunk38
to wrap around and grasp objects (Kier and Smith, 1985). It follows that the ventral surface of39
the trunk is often the primary point of contact between the trunk and the substrate during ob-40
ject manipulation (Dagenais, Hensman, Haechler, and Milinkovitch, 2021). The dorsal side of the41
trunk is not often utilized for grasping, and this surface of the trunk is more exposed to external42
mechanical forces and predators, potentially necessitating a more protective armor-like structure.43
To fulfill the different roles required of it, the skin on the elephant trunk is required to be flexible44
and tough at the same time.45
Relatively little work has been conducted to observe and document the anatomy of elephant skin.46
In 1970, Spearman published a study discussing elephant skin’s basic anatomy, including insights47
about the different vibrissal hairs on the trunk (Spearman, 1970). More recently, biomechanical48
studies have made connections between the skin properties and an elephant’s ability to grasp and49
wrap its trunk around various objects, including barbells (Dagenais et al., 2021; Schulz, Reidenberg,50
Wu, Tang, Seleb, Mancebo, Elgart, and Hu, 2023). While the skin on the elephant body is cracked51
for thermoregulation (Martins, Bennett, Clavel, Groenewald, Hensman, Hoby, Joris, Manger, and52
Milinkovitch, 2018), the trunk, in contrast, has wrinkles and folds on its ventral and dorsal surfaces,53
respectively (Schulz et al., 2023). The structure also varies with position along the length of the54
trunk : the distal trunk skin (on both ventral and dorsal surfaces) is characterized by wrinkles,55
while the proximal dorsal trunk skin has folds. These differing skin characteristics enable the trunk56
to extend to reach faraway objects, with the dorsal surface stretching more than the ventral (Schulz57
et al., 2022).58
In this work, we used both classical and newly developed microscopy techniques to investigate59
the structure of elephant trunk skin. We focused our analysis on collagen, a foundational protein60
that governs the structure of many body tissues, including muscle, blood vessels, and skin, and61
provides bio-inspiration across scales (Eder, Amini, and Fratzl, 2018). Collagen I (COL1) is the62
primary collagen found within the skin; it has a fibrillar structure and can therefore be detected63
with second harmonic generation (SHG) imaging. SHG is a nonlinear optical imaging technique64
that selectively detects noncentrosymmetric molecules, including type I and II collagen with no65
labelling (Boddupalli and Bratlie, 2015; Chen, Nadiarynkh, Plotnikov, and Campagnola, 2012).66
SHG microscopy works by viewing the skin sample at a specific frequency that excites the67
fibrillar structure of COL1; the resulting image exhibits half the wavelength of the original wave-68
length used, hence the term ”second harmonic.” The fibrillar structure of COL1 fibers allows the69
microscopy technique to detect COL1 in the tissue, while the resulting image is related to the70
amount of pre-strain on the COL1 fibers (Turcotte, Mattson, Wu, Zhang, and Lin, 2016). The71
SHG technique is label-free and therefore accrues less error compared to traditional histochem-72
istry since there is not a chained sequence of staining that can vary based on the specific timing73
that segmented skin spends in various chemical baths(Haggerty, Wang, Dickinson, O’Malley, and74
Martin, 2014). The SHG technique is specific to collagen and does not pick up the other fiber75
structures, such as elastin or keratin that are present within the skin (Chen et al., 2012). In skin,76
COL1 networks are characterized by variations in fiber orientation, thickness, density, strain, and77
weaving with neighboring fibers - this last feature is a phenomenon known as entanglementDay,78
Zamani-Dahaj, Bozdag, Burnetti, Bingham, Conlin, Ratcliff, and Yunker (2023). Analysis of SHG79
2
.CC-BY-NC 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 15, 2023. ; https://doi.org/10.1101/2023.08.11.553031doi: bioRxiv preprint
images of skin allows quantification of all these variations in COL1 fibers.80
We here used SHG to analyze COL1 architecture in elephant trunk skin. We conducted morpho-81
logical and compositional analyses on skin samples along the trunk at several locations, including82
seven sites for SHG microscopy and eight for histochemical staining (Figure 1). We show key83
differences in collagen architecture along the length of the trunk, and differences between COL184
architecture in elephant and human skin.85
Experimental Methods86
Dissection of elephant trunk skin87
Icahn School of Medicine at Mount Sinai, New York, provided access to a dissected frozen trunk88
from a 38-year-old female African elephant (Loxodonta africana) that initially lived in a Virginia89
zoo. The elephant was euthanized for health issues in 2011.90
We accessed the trunk when it was on loan from the National Museum of Natural History91
(NMNH), Smithsonian Institution. The elephant’s body weight before death was approximately92
4000 kg. The trunk was cut into several parts and initially stored in a freezer at −20◦Cuntil it93
was dissected in July 2016.94
In March 2019, eight samples of the trunk skin were further dissected at the Icahn School of95
Medicine at Mount Sinai. These samples included five dorsal and three ventral samples ranging96
from the proximal to the distal end of the trunk. These samples were shipped on dry ice to Impe-97
rial College London by the Smithsonian Institute Collections Department as a scientific exchange98
between the two CITES-registered institutions. The Animal Plant and Health Agency in the UK99
(authorization number ITIMP19.0822) approved the tissue shipment. The samples were stored at100
Imperial College London at −80◦C until embedding, sectioning, and imaging were conducted from101
January to March 2020.102
Histology and Morphometrics103
The eight samples were further dissected to enable analysis in the trunk’s longitudinal direction.104
Samples were embedded in OCT (optimum cutting temperature) medium and 20 µm-thick sections105
were cut on a cryostat (Figure S1). The tissue sections were stained using hematoxylin and106
eosin (H&E) and then imaged on a Zeiss inverted microscope at 3x magnification. Images were107
automatically segmented using the wand tool in FIJI (ImageJ) based on the stained color differences108
from H&E.109
To quantify the thicknesses of each layer (the stratum corneum (SC), the viable epidermis (VE),110
and dermis (D) shown on Figure 2), a MATLAB script was used to divide each H&E image (1000111
pixels wide) into vertical strips of one-pixel width. The pixels corresponding to each layer were112
counted and recorded. To compare samples, we reported the thickness for each layer, defined as the113
thickness of the layer divided by the sum of all layers (Table 1, Figure 3A, Figure S2, Figure114
S3).115
Second Harmonic Generation116
Samples embedded in OCT were sectioned at 100µm thickness for second-harmonic generation117
(SHG) imaging. Images were taken from an upright confocal microscope (Leica SP5) coupled to a118
Ti: Sapphire laser (Newport Spectra-Physics). Raw images were received as a stacked TIF file with119
10 µm between each image of the TIF file at a maximum of 255 nm with green luminescence. Stacks120
3
.CC-BY-NC 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 15, 2023. ; https://doi.org/10.1101/2023.08.11.553031doi: bioRxiv preprint
were then processed using a workflow in Fiji (ImageJ), including setting the minimum-maximum121
range to (0,4000), applying the blur filter (σ= 0.5), and subtracting background (rolling ball radius,122
40 pixels). A machine-learning and segmenting open-source software, ilastik, was used to analyze123
the difference between fibers and background. Completed images are shown on Figure 4A-B.124
Collagen Fiber Orientation and crossings125
We used the open-source software CurveAlign to quantify the collagen fiber orientation in SHG126
images (Bredfeldt, Liu, Pehlke, Conklin, Szulczewski, Inman, Keely, Nowak, Mackie, and Eliceiri,127
2014). Images were broken into regions of interest of size 600 µm×450 µm with at least a 150128
pixels distance from the boundary (Figure 5A). For this study, we only examined individual fibers129
instead of the entire fiber network.130
We considered two broad categories of fiber orientation shown in Figure 5B. Fibers perpen-131
dicular to the skin, shown in blue in the schematic, have angles of 0 ±5◦and 180 ±5◦, where 0◦is132
defined as outward normal from the skin as shown in inset of Figure 5A. Parallel fibers (orange)133
have angles of 90±5◦. To report the number of fibers of these orientations, we report the percentage134
of fibers oriented in each direction. A histogram of fiber arrangement is constructed and analyzed135
for the perpendicular and parallel orientation ratios (Figure 5A).136
We measured the number of collagen fiber overlaps from a dorsal section 133 cm from the tip.137
The region was a 200 ×200 pixel square and an extruded depth of 100 µm. These crossings were138
counted using ImageJ. In reporting individual collagen fibers, we compared the SHG images of139
human skin given by Boyle et al. with that of the elephant skin samples in our study (Boyle,140
Plotczyk, Villalta, Patel, Hettiaratchy, Masouros, Masen, and Higgins, 2019). We measured the141
average number of overlaps per unit volume and compared this between humans’ plantar and142
non-plantar tissue and that of elephants.143
Statistical Methods144
All calculations, including statistical analysis, were performed with MATLAB 2022A. In the ta-145
bles and on the figures, values are reported as mean plus or minus standard deviation. We used146
the MATLAB function ttest for t-test to find statistically significant differences between dorsal147
versus ventral values, difference of values at different positions along the trunk, and differences in148
perpendicular versus parallel values.149
Results150
Macrostructure of the elephant trunk skin151
The outermost layer of the skin, the stratum corneum (SC), is composed of denucleated, keratinized152
epithelial cells with lipids in between. Underneath the SC is the viable epidermis (VE) which is153
a sheet of epithelial cells with tight junctions in between them, which gives the skin its barrier154
function. Beneath this is the structural support layer for the overlying epithelium, known as the155
dermis (D)(Boyle et al., 2019). To quantify differences in elephant skin morphology across trunk156
locations, we used H&E image analysis to segment the skin into the SC, the VE, and D (Figure157
2,Figures S1). Below we will make comparisons of dorsal and ventral skin at the same distance158
from the tip of the trunk.159
Starting with the stratum corneum, we found that on the dorsal trunk, the SC was thickest in160
the proximal base, with a mean thickness of 2 mm (Table 1, Figure 3A), which is significantly161
4
.CC-BY-NC 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 15, 2023. ; https://doi.org/10.1101/2023.08.11.553031doi: bioRxiv preprint
different from the ventral SC, with a thickness of 0.34 mm (p<0.001). The remainder of the SC162
on the dorsal trunk varied from 0.25 mm to 1 mm on average (Figure 3A). In contrast, SC of the163
ventral trunk had a relatively constant thickness of 0.40 mm.164
The viable epidermis thickness remained broadly consistent throughout the length of the trunk165
and between ventral and dorsal sites (Table 1, Figure S2). The overall thickness of the VE166
remained nearly constant at around 0.3-0.4 mm for both the dorsal and ventral elephant surfaces.167
An exception was the very distal tip of the dorsal skin, 3 cm from the tip (finger at the tip of the168
trunk), which had a tiny layer of VE at only 0.05 mm thick (Figure S2). This thickness displayed169
a statistically significant difference from the rest of the skin analyzed (p<0.001).170
Together, the SC and VE are considered to be the armor for the skin as they serve as the first171
layers of protection against environmental insults. Compared to other species’ armor layers, such172
as scales or shells, the elephant skin on the dorsal trunk reaches 2.2 mm thick - this is double the173
thickness of a pangolin scale and four times that of a human fingernail (Figure 3B). Additionally,174
the epidermal thickness of the elephant trunk is nearly 100 times thicker than the epidermis on an175
adult human’s torso.176
The next skin layer beneath the VE is the dermis. We observe two regions of increased thickness177
of the dermis, the tip and the proximal base. At the tip, the ventral dermis is 1.5 thicker than178
the dorsal dermis (2.3 mm versus 1.46 mm thickness, respectively) ((Table 1, Figure S3). This179
thickening makes sense: at the tip, the thicker ventral dermis is where the trunk grasps and180
manipulates objects. The dermis appears to thicken where the trunk increases in diameter as well.181
At the proximal base, the dermis along the dorsal trunk is 700% thicker than the dermis at the182
distal tip (5.44 mm versus 0.8 mm, respectively).183
Micro-structure of elephant skin184
To characterize compositional differences in COL1 between the skin samples from the elephant185
trunk, we used Second harmonic generation imaging (SHG). SHG can identify the macro and186
micro-level structures of the skin, such as COL1 fiber density, intensity, and orientation (Figure187
S4A). The color intensity in SHG images can be used as a proxy for fiber strain, indicating the188
mechanical state of the tissue (Turcotte et al., 2016). (Figure 4A-B) showed the ventral trunk189
has an overall higher intensity than the dorsal trunk, indicating ventral fibers have more pre-strain190
than dorsal (Figure 4C). At the tip of the trunk, the ventral skin has an SHG intensity twice that191
(p < 0.001) seen in the dorsal (Figure 4D). This trend was accentuated at the trunk base, where192
the ventral skin SHG intensity was six times (p < 0.001) the intensity of the dorsal skin (Figure193
4D). The differences in SHG intensity observed here indicate that dorsal skin has less pre-strain194
imposed on the collagen allowing more stretch-ability than ventral skin.195
We next used the SHG images to assess the collagen fiber angle (Figure S4,Figure 5A). Two196
fiber angle orientations, perpendicular and parallel relative to the skin surface, are of particular197
relevance to the physical properties of the skin (Figure 5B). As discussed in the methods, we198
define zero degrees as the outward normal of the skin surface (Figure 5A). Perpendicular fibers199
resist axial trunk loading from forces perpendicular to the skin (Figure 5B). The parallel fibers200
are oriented 90 degrees to the outward normal. Parallel fibers primarily assist with extension and201
shear loading tolerance (Figure 5B).202
Upon analysis of the collagen orientation from the SHG images, we found that dorsal skin203
samples are composed of bi-modal orientation peaks, with COL1 fibers oriented in both the per-204
pendicular and parallel directions (Figure 5C). All samples of dorsal skin analyzed have over 20%205
of perpendicular and 20% of parallel fibers in the skin, indicating a bi-model peak of fiber dis-206
tribution. Additionally, we see a significant difference when we compare the fiber orientation at207
5
.CC-BY-NC 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 15, 2023. ; https://doi.org/10.1101/2023.08.11.553031doi: bioRxiv preprint
specific sites along the trunk. Along the dorsal surface of the trunk at 3, 27, and 81 cm from the208
tip of the trunk, we see significant differences between the percentage of perpendicular and parallel209
fibers. The proximal base (100 and 133 cm from the tip) on the dorsal surface, however, shows no210
significant difference, with around 25% perpendicular and 25% parallel fiber orientation (Figure211
5C).212
Dorsal and ventral surfaces show statistically significant differences in collagen fiber orientation.213
At the distal tip of the trunk (27 cm from the tip), the ventral skin has significantly more COL1214
fibers in the parallel direction (p < 0.01) compared to the dorsal skin at the same site (Figure 5C).215
When we look at the proximal base (133 cm from the tip), the dorsal skin has more perpendicular216
collagen (p < 0.001), and less parallel collagen (p < 0.001) relative to ventral skin at the same217
location.218
As mentioned above, we see a bi-modal distribution of fiber orientation in the elephant with219
large percentages in both the perpendicular and parallel directions. In our previous work looking220
at human skin, we found that both plantar (skin on the sole of the foot) and non-plantar (body)221
skin contained COL1 fibers with preferential fiber orientation (perpendicular or parallel) in just a222
single direction(Boyle et al., 2019), as opposed to the bi-model distribution observed in elephant223
skin. Given the differences in fiber orientation between human and elephant skin, we postulated224
that there would also be differences in the entanglement of COL1 fibers. To assess COL1 fiber225
overlap or entanglement (Figure 6A), we analyzed a 200 x 200-pixel SHG image segment from226
dorsal skin 133 cm from the tip. We found that the average number of fiber crossings per µm3in227
the elephant trunk is 5.85 (Figure 6B). This value is six times higher than that observed in both228
human plantar (p < 0.01) and non-plantar skin (p < 0.01).229
Discussion230
We set out to evaluate if elephant trunk skin has variations in its architecture along the length of231
the trunk that may explain the different functions of the trunk. We found variations in morphology232
and composition along the trunk length at both the macro and micro scale. The dorsal portion of233
the trunk, including the trunk’s dorsal finger (3 cm from tip) and dorsal root (133 cm from tip),234
had the thickest SC layers. The distal tip of the trunk, or finger, is regularly used to manipulate235
objects, and the dorsal root is more exposed to external stimuli(Dagenais et al., 2021). These236
functions may explain the thicker dorsal finger and root SC layers.237
When we combine the thickness of the SC and VE in this dorsal root and compare it to other238
species, we see the elephant may have the thickest dermal armor among extant animals; elephants239
have a dermal armor thickness twice that of a pangolin scale and four times a human thumbnail240
Figure 3B)(Wang, Yang, Sherman, and Meyers, 2016; Wollina, Berger, and Karte, 2001).241
While the elephant uses skin for protection, aquatic and arctic species use thick fat layers for242
protection and insulation (Liwanag, Berta, Costa, Budge, and Williams, 2012). In humans, the sole243
also has a fat pad that protects the skeleton from heel strike impact. Unlike the fat layers in arctic244
species, this fat pad does not protect the skin – instead, foot skin has adapted to be thicker and245
stiffer than body skin, which allows it to withstand mechanical loading. In other species, we see246
a range of morphological structures, such as shells and scales (Figure 3B), where the skin armor247
has adapted to provide additional protection against environmental pressures (Wang et al., 2016).248
Our study was limited by having material samples from only one elephant specimen and one249
elephant species. While many dry skin samples are available in museums, frozen samples, which250
allow preservation and histological analysis, are much rarer. Moreover, this specimen was an251
African bush elephant (Loxodonta africana), just one of three elephant species. There may be252
6
.CC-BY-NC 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 15, 2023. ; https://doi.org/10.1101/2023.08.11.553031doi: bioRxiv preprint
intrinsic differences between species that we could not address in our study. Asian elephants have253
only one finger at the tip, with the ventral finger composed of a cartilage bulb. This difference in254
trunk tip morphology is partly due to Asian elephants being grazers (eat low-lying vegetation). In255
contrast, African elephants are browsers (also eat high-growing vegetation) and require a prehensile256
finger to grip and pull leaves off branches for nutrients.257
Boyle et al. found that in comparing human skin samples, skin on different body sites had258
COL1 fibers oriented preferentially in either a parallel or perpendicular direction, depending on259
the functional requirements for skin at that site (Boyle et al., 2019). The dorsal surface of the260
elephant trunk expressed relatively even amounts of parallel and perpendicular collagen. The261
ventral root portion of the trunk had more parallel collagen. We envisage that these observations262
will give inspiration to future biomimetic studies. While collagen fiber entanglement is still being263
understood, the general belief is that the structure on the micro-scale leads to unique mechanical264
responses on the macro scale. There has been increased interest in understanding the macro265
physical properties that stem from micro-scale entanglements. Such work may influence the design266
of soft robotic manipulators(Becker, Teeple, Charles, Jung, Baum, Weaver, Mahadevan, and Wood,267
2022). Our studies of the impacts of woven fiber structure inside the skin are reminiscent of268
the impact of patterning in knitted fabric structures. Knitting is a centuries-old activity that269
involves manipulating a string-like material, traditionally yarn, into a complex fabric with emergent270
elasticity. These fabrics can exhibit vastly different mechanical properties based on how the stitches,271
specific slipknots formed by the yarn, are patterned and structured(Singal, Dimitriyev, Gonzalez,272
Quinn, and Matsumoto, 2023). These structural differences leading to robustness are also challenges273
in the public health sector. Collagen fibers in skin constructs are always oriented parallel to the274
skin dermis as they govern how skin contracts. Orienting perpendicular fiber alignment could make275
skin grafts more robust in their mechanical and flexibility utility.276
In summary, we compared the trunk along the distal-proximal and dorsal-ventral anatomical277
axes, finding differences in the morphology and composition across the elephant trunk and giving278
insights into the form-function relationships. Elephant trunks have some of the thickest dermal279
armor in the animal kingdom, with a 2.2 mm thick epidermis. This armor is paired with parallel and280
perpendicular collagen in the dermis, allowing strength and flexibility. Furthermore, the bi-model281
orientation of collagen in the dermis leads to individual fiber overlap and interaction, showcasing282
the entanglement of fibers inside the skin. This work shows the complex nature of elephant skin283
and provides bio-inspiration for materials that require strength and flexibility.284
Acknowledgements285
Thank you to the European Hair Research Society, who supported the travel of AS to Imperial286
College London. We thank J. Ososky and the Smithsonian Institution Museum of Natural History287
for assistance with information regarding the frozen elephant trunk. We thank the Facility for288
Imaging by Light Microscopy (FLIM) at Imperial College London, which is in part, supported by289
funding from the Wellcome Trust (104931/Z/14/Z) and BBSRC (BB/L015129/1). CH was funded290
by a project grant from the Engineering and Physical Sciences Research Council (EP/N026845/1).291
References292
K. Becker, C. Teeple, N. Charles, Y. Jung, D. Baum, J. C. Weaver, L. Mahadevan, and R. Wood.293
Active entanglement enables stochastic, topological grasping. Proceedings of the National294
Academy of Sciences, 119(42):e2209819119, Oct. 2022. 10.1073/pnas.2209819119.295
7
.CC-BY-NC 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 15, 2023. ; https://doi.org/10.1101/2023.08.11.553031doi: bioRxiv preprint
A. Boddupalli and K. Bratlie. Multimodal imaging of harmonophores and application of high296
content imaging for early cancer detection. Materials Discovery, 1:10–20, Jan. 2015. 10.1016/297
j.md.2015.11.002.298
C. Bordoloi. RRC: A note on microscopic anatomy of the skin of great Indian rhino calf (Rhinoceros299
unicornis), Aug. 2021.300
C. J. Boyle, M. Plotczyk, S. F. Villalta, S. Patel, S. Hettiaratchy, S. D. Masouros, M. A. Masen,301
and C. A. Higgins. Morphology and composition play distinct and complementary roles in the302
tolerance of plantar skin to mechanical load. Science Advances, 5(10):eaay0244, Oct. 2019.303
10.1126/sciadv.aay0244.304
J. S. Bredfeldt, Y. Liu, C. A. Pehlke, M. W. Conklin, J. M. Szulczewski, D. R. Inman, P. J. Keely,305
R. D. Nowak, T. R. Mackie, and K. W. Eliceiri. Computational segmentation of collagen fibers306
from second-harmonic generation images of breast cancer. Journal of Biomedical Optics, 19307
(1):1 – 11, 2014. 10.1117/1.JBO.19.1.016007.308
X. Chen, O. Nadiarynkh, S. Plotnikov, and P. J. Campagnola. Second harmonic generation mi-309
croscopy for quantitative analysis of collagen fibrillar structure. Nature Protocols, 7(4):654–310
669, Apr. 2012. 10.1038/nprot.2012.009.311
R. K. Chintapalli, M. Mirkhalaf, A. K. Dastjerdi, and F. Barthelat. Fabrication, testing and mod-312
eling of a new flexible armor inspired from natural fish scales and osteoderms. Bioinspiration313
& Biomimetics, 9(3):036005, Mar. 2014. 10.1088/1748-3182/9/3/036005.314
P. Dagenais, S. Hensman, V. Haechler, and M. C. Milinkovitch. Elephants evolved strategies315
reducing the biomechanical complexity of their trunk. Current biology: CB, 31(21):4727–316
4737.e4, Nov. 2021. 10.1016/j.cub.2021.08.029.317
T. C. Day, S. A. Zamani-Dahaj, G. O. Bozdag, A. J. Burnetti, E. P. Bingham, P. L. Conlin,318
W. C. Ratcliff, and P. J. Yunker. Entanglement in living systems, June 2023. Pages:319
2023.06.13.544814 Section: New Results.320
M. Eder, S. Amini, and P. Fratzl. Biological composites—complex structures for functional diversity.321
Science, 362(6414):543–547, Nov. 2018. 10.1126/science.aat8297.322
J. M. Haggerty, X. N. Wang, A. Dickinson, C. J. O’Malley, and E. B. Martin. Segmentation of323
epidermal tissue with histopathological damage in images of haematoxylin and eosin stained324
human skin. BMC Medical Imaging, 14(1):7, Dec. 2014. 10.1186/1471-2342-14-7.325
D. Han and B. A. Young. The rhinoceros among Serpents: Comparative anatomy and experimental326
biophysics of Calabar burrowing python (Calabaria reinhardtii) skin. Journal of Morphology,327
279(1):86–96, 2018. 10.1002/jmor.20756.328
W. M. Kier and K. K. Smith. Tongues, tentacles and trunks: the biomechanics of movement in329
muscular-hydrostats. Zoological Journal of the Linnean Society, 83(4):307–324, Apr. 1985.330
10.1111/j.1096-3642.1985.tb01178.x.331
H. E. M. Liwanag, A. Berta, D. P. Costa, S. M. Budge, and T. M. Williams. Morphological332
and thermal properties of mammalian insulation: the evolutionary transition to blubber in333
pinnipeds. Biological Journal of the Linnean Society, 107(4):774–787, Dec. 2012. 10.1111/334
j.1095-8312.2012.01992.x.335
A. F. Martins, N. C. Bennett, S. Clavel, H. Groenewald, S. Hensman, S. Hoby, A. Joris, P. R.336
Manger, and M. C. Milinkovitch. Locally-curved geometry generates bending cracks in the337
African elephant skin. Nature Communications, 9(1), Dec. 2018. 10.1038/s41467-018-06257-3.338
A. K. Schulz, M. Boyle, C. Boyle, S. Sordilla, C. Rincon, S. Hooper, C. Aubuchon, J. S. Reidenberg,339
C. Higgins, and D. L. Hu. Skin wrinkles and folds enable asymmetric stretch in the elephant340
trunk. Proceedings of the National Academy of Sciences, 119(31):e2122563119, Aug. 2022.341
10.1073/pnas.2122563119.342
A. K. Schulz, J. S. Reidenberg, J. N. Wu, C. Y. Tang, B. Seleb, J. Mancebo, N. Elgart, and D. L.343
8
.CC-BY-NC 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 15, 2023. ; https://doi.org/10.1101/2023.08.11.553031doi: bioRxiv preprint
Hu. Elephant trunks use an adaptable prehensile grip. Bioinspiration & Biomimetics, 18(2):344
026008, Feb. 2023. 10.1088/1748-3190/acb477.345
K. Singal, M. S. Dimitriyev, S. E. Gonzalez, S. Quinn, and E. A. Matsumoto. Programming346
Mechanics in Knitted Materials, Stitch by Stitch, Feb. 2023. arXiv:2302.13467 [cond-mat].347
R. I. C. Spearman. The Epidermis and its Keratinisation in the African Elephant ( Loxodonta348
Africana ). Zoologica Africana, 5(2):327–338, Jan. 1970. 10.1080/00445096.1970.11447400.349
R. Turcotte, J. M. Mattson, J. W. Wu, Y. Zhang, and C. P. Lin. Molecular Order of Arterial Colla-350
gen Using Circular Polarization Second-Harmonic Generation Imaging. Biophysical Journal,351
110(3):530–533, Feb. 2016. 10.1016/j.bpj.2015.12.030.352
B. Wang, W. Yang, V. R. Sherman, and M. A. Meyers. Pangolin armor: Overlapping, structure,353
and mechanical properties of the keratinous scales. Acta Biomaterialia, 41:60–74, Sept. 2016.354
ISSN 17427061. 10.1016/j.actbio.2016.05.028.355
U. Wollina, M. Berger, and K. Karte. Calculation of nail plate and nail matrix parameters by356
20 MHz ultrasound in healthy volunteers and patients with skin disease. Skin research and357
technology, 7(1):60–64, Feb. 2001. 10.1034/j.1600-0846.2001.007001060.x.358
W. Yang, H. Quan, M. A. Meyers, and R. O. Ritchie. Arapaima Fish Scale: One of the Toughest359
Flexible Biological Materials. Matter, 1(6):1557–1566, Dec. 2019. ISSN 2590-2393, 2590-2385.360
10.1016/j.matt.2019.09.014.361
9
.CC-BY-NC 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 15, 2023. ; https://doi.org/10.1101/2023.08.11.553031doi: bioRxiv preprint
Figures362
Dorsal Thickness
Distance from Tip SC (mm) VE (mm) D (mm)
3 0.39 ±0.11 0.05 ±0.06 0.8±0.097
27 0.17 ±0.16 0.36 ±0.21 1.46 ±0.27
81 0.39 ±0.27 0.27 ±0.21 6.6±0.28
100 0.87 ±0.61 0.58 ±0.51 5.8±0.90
133 1.83 ±0.68 0.42 ±0.36 5.44 ±.64
Ventral Thickness
Distance from Tip SC (mm) VE (mm) D (mm)
27 0..29 ±0.32 0.50 ±0.31 2.4±0.44
51 0.12 ±0.18 0.32 ±0.49 4.90 ±0.54
133 0.34 ±0.22 0.22 ±0.22 4.19 ±0.23
Table 1: Table displaying the thickness of each skin layer in mm displayed in mean ±standard
deviation. Results of each layer are displayed as SC (Figure 3A), VE (Figure S2), and D (Figure
S3).
10
.CC-BY-NC 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 15, 2023. ; https://doi.org/10.1101/2023.08.11.553031doi: bioRxiv preprint
Figure 1: Schematic of the elephant trunk with experimental outputs from H&E Staining and
SHG microscopy shown as insets.
Figure 2: Macroscopic image of a cross section of elephant skin showing subcutaneous tissue and
muscle. The skin layers are shown in a schematic of the Stratum Corneum (SC), Viable Epidermis
(VE), and Dermis (D).
11
.CC-BY-NC 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 15, 2023. ; https://doi.org/10.1101/2023.08.11.553031doi: bioRxiv preprint
Figure 3: A) Relationship between Stratum corneum (SC) thickness and position on the trunk.
The position is the distance from the trunk tip in cm. Stars indicate the statistical significance of
the difference between dorsal and ventral sites: (*** p < 0.001) B) Thickness of different dermal
armors across species. Non-elephant data taken from (Bordoloi, 2021; Chintapalli et al., 2014; Han
and Young, 2018; Wang et al., 2016; Wollina et al., 2001; Yang et al., 2019). Silhouettes and animal
images taken from Adobe CC Images.
12
.CC-BY-NC 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 15, 2023. ; https://doi.org/10.1101/2023.08.11.553031doi: bioRxiv preprint
Figure 4: A-B) SHG stacked image of dorsal and ventral sections of the proximal trunk. C)
Schematic displaying the relationship between the intensity of SHG in fibers and the indicative
strain of a fiber. D) Relationship between SHG intensity and position on the trunk. Stars indicate
the statistical significance of the difference between dorsal and ventral sites: (*** p < 0.001). Scale
bars A,B: 100 µm.
13
.CC-BY-NC 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 15, 2023. ; https://doi.org/10.1101/2023.08.11.553031doi: bioRxiv preprint
Figure 5: A) Stacked SHG image of the distal ventral elephant trunk with inset of CurveAlign
output showing collagen fiber orientation. Inset histogram showing collagen fiber alignment. B)
Schematic of parallel and perpendicular collagen fibers in the dermis. C) Relationship between the
percentage of collagen fibers and position on the trunk. Parallel fibers are shown in orange and
perpendicular fibers in blue. Blue and Orange stars indicate statistical significance between dorsal
and ventral sites, with stars placed over the larger value. Black stars indicate statistical significance
between perpendicular and parallel comparisons within a single site. Stars indicate the following
significance: (* p < 0.05, ** p < 0.01, *** p < 0.001). Scale bar A: 200 µm.
14
.CC-BY-NC 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 15, 2023. ; https://doi.org/10.1101/2023.08.11.553031doi: bioRxiv preprint
Figure 6: A) Schematic of a cross-linked and non-cross-linked collagen fiber. B) Collagen crossings
per cubic micron for elephant skin (dorsal region 133 cm from the tip) and human plantar and
non-plantar skin. Published SHG images of human skin reanalyzed from (Boyle et al., 2019). Stars
indicate the statistical significance of the difference between elephant and human skin: (** p < 0.01)
Silhouettes of African elephant (Loxodonta africana)from phylopic artist Agnello Picorelli.
15
.CC-BY-NC 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted August 15, 2023. ; https://doi.org/10.1101/2023.08.11.553031doi: bioRxiv preprint
