The biter bit? Investigation of possible in-ovo self-envenomation in an Egyptian saw-scaled viper using region of interest X-ray microtomography

Abstract

Proven examples of self-envenomation by venomous snakes, and especially instances of death as a result of these events, are extremely rare, if not non-existent. Here we use Region of Interest X-ray microtomography to investigate a putative case of fatal in-ovo self-envenomation in the Egyptian saw-scaled viper, Echis pyramidum. Our analyses have provided unprecedented insight into the skeletal anatomy of a late-stage embryonic snake and the disposition of the fangs without disrupting or destroying a unique biological specimen.

Full text

The biter bit? Investigation of possible in-ovo self-
envenomation in an Egyptian saw-scaled viper using region of
interest X-ray microtomography
John Mulley, Richard E Johnston
Proven examples of self-envenomation by venomous snakes, and especially instances of
death as a result of these events, are extremely rare, if not non-existent. Here we use
Region of Interest X-ray microtomography to investigate a putative case of fatal in-ovo
self-envenomation in the Egyptian saw-scaled viper, Echis pyramidum. Our analyses have
provided unprecedented insight into the skeletal anatomy of a late-stage embryonic snake
and the disposition of the fangs without disrupting or destroying a unique biological
specimen.
PeerJ PrePrints | http://dx.doi.org/10.7287/peerj.preprints.624v1 | CC-BY 4.0 Open Access | rec: 19 Nov 2014, publ: 19 Nov 2014
PrePrints

1
Title page
1
2
The biter bit? Investigation of possible in-ovo self-envenomation in an Egyptian saw-scaled
3
viper using region of interest X-ray microtomography
4
5
Richard E Johnston1 and John F Mulley2*
6
7
1. College of Engineering, Swansea University, Swansea, SA2 8PP, United Kingdom
8
2. School of Biological Sciences, Bangor University, Bangor, Gwynedd LL57 2UW, United
9
Kingdom
10
11
*To whom correspondence should be addressed (j.mulley@bangor.ac.uk)
12
13
14
15
16
17
18
19
20
21
22
23
24
25
PeerJ PrePrints | http://dx.doi.org/10.7287/peerj.preprints.624v1 | CC-BY 4.0 Open Access | rec: 19 Nov 2014, publ: 19 Nov 2014
PrePrints

2
Abstract
26
Proven examples of self-envenomation by venomous snakes, and especially instances of
27
death as a result of these events, are extremely rare, if not non-existent. Here we use Region
28
of Interest X-ray microtomography to investigate a putative case of fatal in-ovo self-
29
envenomation in the Egyptian saw-scaled viper, Echis pyramidum. Our analyses have
30
provided unprecedented insight into the skeletal anatomy of a late-stage embryonic snake and
31
the disposition of the fangs without disrupting or destroying a unique biological specimen.
32
33
Keywords
34
Snake; saw-scaled viper; self-envenomation; microCT; region of interest; X-ray
35
microtomography
36
37
Background
38
Snake venom is a potent mix of proteins and peptides, honed by millions of years of natural
39
selection for rapid prey immobilisation (Casewell et al. 2013). Safely producing and storing
40
this lethal arsenal within the body prior to its use creates obvious issues, and these have to
41
some extent been overcome in snakes by the evolution of a specialised gland (the venom
42
gland (Jackson, 2003; Weinstein, Smith & Kardong, 2009)) for storing venom and by
43
production of inactive precursor proteins (zymogens) for many venom components
44
(Shimokawa et al. 1996; Portes-Junior et al. 2014). The issue of whether a venomous snake is
45
immune to its own venom is still largely unresolved, although there is some evidence of
46
possible adaptations for resistance to self-envenomation (Denson, 1976; Smith et al, 2000;
47
Takacs, Wilhelmsen & Sorota, 2001; Takacs, Wilhelmsen & Sorota, 2004; Tanaka-Azevedo
48
et al. 2004; Vieira et al. 2008). Investigations of the available literature have failed to identify
49
PeerJ PrePrints | http://dx.doi.org/10.7287/peerj.preprints.624v1 | CC-BY 4.0 Open Access | rec: 19 Nov 2014, publ: 19 Nov 2014
PrePrints

3
any definitive examples of self-envenomation by a venomous snake, although such tales are
50
prevalent on the internet, where they seemingly rarely cause death or long-term injury.
51
Following some breeding experiments with Egyptian saw-scaled vipers (Echis pyramidum) in
52
summer 2014, we found a single egg failed to hatch from a clutch of thirteen otherwise
53
successful eggs. Examination revealed that the developing embryo had used its eggtooth to
54
create slits in the eggshell (and was therefore within a few days of hatching) and, when
55
opened, the egg contained a dead, almost-fully-developed snake, with some un-absorbed yolk
56
(Figure 1a). A coil of the body was firmly located within the mouth (Figures 1b-1d),
57
suggesting a possible case of in-ovo self-envenomation. To definitively prove this however,
58
we needed to determine whether the fangs were penetrating the body cavity, ideally without
59
disturbing the positioning of this unique specimen.
60
High resolution X-ray microtomography (µCT, microCT) is a non-destructive method for
61
imaging internal structures in three dimensions at micron level spatial resolution based upon
62
the principle that X-ray attenuation is a function of X-ray energy and the density and atomic
63
composition of materials being scanned. The result is a 3D ‘tomogram’ (Maire & Withers,
64
2014), generated from hundreds or thousands of individual 2D X-ray projections sampled at
65
the detector while the specimen rotates between the fixed X-ray source and detector. The
66
tomogram consists of a matrix of 3D isotropic voxels, each of which is assigned a grayscale
67
value derived from a linear attenuation coefficient that relates to the density of the scanned
68
materials (Landis & Keane, 2010; Cnudde & Boone, 2013). MicroCT resolution can be of the
69
order of 100 times finer than medical CT scans (Ketcham & Carlson, 2001), enabling 3D
70
imaging and analysis of smaller internal features, although resolution is related to specimen
71
width. Successful filtered back projection reconstruction of the 3D data requires the entire
72
sample width to be encompassed within each 2D projection or ‘field of view’ at all rotations
73
(Kak & Slaney, 2001) and a typical X-ray detector panel in a laboratory microCT setup has a
74
PeerJ PrePrints | http://dx.doi.org/10.7287/peerj.preprints.624v1 | CC-BY 4.0 Open Access | rec: 19 Nov 2014, publ: 19 Nov 2014
PrePrints

4
width of around 1000-4000 pixels. For a detector with a width of 2000 pixels, the pixel size
75
(and ultimately 3D voxel size of the reconstructed tomogram) is therefore w/2000, where w is
76
the maximum width of the specimen.
77
Conventional wisdom in microCT reconstruction states that only parts of the object
78
illuminated by X-rays in all 2D projections images will be properly reconstructed i.e. the
79
whole object should lie within the field of view for all rotations during the scan. However,
80
this conventional approach produces scans of larger objects at a lower resolution. Region of
81
Interest (RoI) tomography (Kyrieleis et al. 2011) offers the potential to ‘zoom in’ to
82
particular areas of large specimens so as to provide higher resolution tomograms of key
83
regions. In this approach, parts of the specimen are within the field of view for some
84
rotations, but then rotate out of the field of view at other rotational angles. We carried out
85
Region of Interest microCT to determine the disposition of the fangs in our specimen and so
86
reveal whether the biter had indeed been bit.
87
88
Methods
89
A clutch of thirteen eggs were laid by a wild-caught Egyptian saw-scaled viper (E.
90
pyramidum) on the 4th July 2014 and, following incubation at 27°C, all but one had hatched
91
by 4th September 2014. Upon removal from its egg, the specimen was fixed in 4%
92
paraformaldehyde in phosphate buffered saline (pH7.5) and stored at 4°C. The specimen was
93
imaged using a Leica MSV269 stereoscope and an Apple iPhone 5. To minimise physical
94
disruption during shipping, the specimen was packed in paraformaldehyde-soaked cotton
95
wool in a 100ml container (Gosseline TP51-004).
96
3D geometric data was collected on a Nikon XT H 225 microfocus X-ray tomography system
97
(Nikon Metrology, Tring, UK) at the College of Engineering, Swansea University, UK.
98
Images were captured with a 1.3 Megapixel Varian Paxscan 2520 amorphous silicon flat
99
PeerJ PrePrints | http://dx.doi.org/10.7287/peerj.preprints.624v1 | CC-BY 4.0 Open Access | rec: 19 Nov 2014, publ: 19 Nov 2014
PrePrints

5
panel detector, in reflection mode with a molybdenum target. Scans were performed with 65
100
kV X-ray tube voltage, a current of 295 µA, with an exposure of 2000 ms, capturing 1 image
101
per rotation step of 0.119°, resulting in 3016 images per scan and a voxel (3D pixel) size of
102
17.6 µm. The tomograms were reconstructed from the 2D projections using Nikon CTPro
103
version 3.1.3 software (Nikon Metrology, Tring, UK). The commercial software VGStudio
104
Max 2.1.5 (Volume Graphics, Heidelberg, Germany) and the free software Drishti (Limaye,
105
2012) were used to view the reconstructed data, 2D slices and rendered 3D volumes.
106
107
Results and discussion
108
In order to minimise handling and potential disruption of our specimen, it was decided to
109
conduct scans whilst it was still packed in its 52mm diameter container of cotton wool-
110
soaked preservative (Figure 1). Since scans of the entire specimen and its container would
111
have resulted in a lower overall resolution, with a voxel size of approximately 27µm, we
112
employed RoI tomography to ‘zoom in’ to the snake, ignoring the surrounding materials,
113
resulting in a field of view of 33.75mm and a voxel size of 17.6µm. These RoI scans have
114
provided astonishing insights into the skeletal anatomy of this specimen and clearly reveal
115
the position and orientation of both fangs (Figures 2a-e). The fangs of vipers such as E.
116
pyramidum are located on a hinged maxilla, which allows them to be folded against the roof
117
of the mouth when not in use and to swing forward to an erect position during a strike.
118
Perhaps disappointingly, we find that the fangs of this specimen are in the folded position and
119
are not penetrating the body cavity (Figure 2). It is still possible however that a bite and
120
envenomation did take place, followed by subsequent withdrawal of the fangs, where the
121
cause of death could be either a result of venom or the physical trauma associated with the
122
bite itself, especially if one or both fangs punctured a major organ. Alternatively, it is
123
possible that this animal drowned within its egg, after having non-fatally bitten itself and then
124
PeerJ PrePrints | http://dx.doi.org/10.7287/peerj.preprints.624v1 | CC-BY 4.0 Open Access | rec: 19 Nov 2014, publ: 19 Nov 2014
PrePrints

6
being either unable or unwilling to release. Whilst it may be possible that disruption of the
125
specimen may reveal traces of bite marks, we feel that the chances of identifiable marks
126
being found are not high enough to risk the permanent loss of this unique specimen.
127
Although we were unable to determine the cause of death in this case, we were easily able to
128
identify the location and orientation of the fangs and other skeletal structures in this relatively
129
small specimen. Our approach demonstrates the power and utility of non-destructive X-ray
130
microtomography and Region of Interest scanning to shed light on biological problems,
131
especially those involving rare, delicate, or unique specimens. More generally, this project
132
highlights the importance of, awareness of, and collaboration across academic disciplines, in
133
this case biological sciences and materials sciences.
134
135
Conclusions
136
We have successfully used Region of Interest scanning to determine the position of the fangs
137
in an embryonic snake that seemingly died as a result of a self-inflicted bite. Whether death
138
was a direct result of a bite involving penetration of the fangs (envenomation, organ
139
puncture/failure) or an indirect result of a non-penetrative bite (e.g. drowning) is unclear and
140
so the cause of death of this enigmatic specimen remains a mystery.
141
142
Acknowledgements
143
The authors wish to thanks Rhys Morgan for technical assistance and Twitter for facilitating
144
the initial collaboration.
145
146
Funding
147
PeerJ PrePrints | http://dx.doi.org/10.7287/peerj.preprints.624v1 | CC-BY 4.0 Open Access | rec: 19 Nov 2014, publ: 19 Nov 2014
PrePrints

7
JFM has been generously supported by the Biosciences, Environment and Agriculture
148
Alliance between Aberystwyth and Bangor universities. RJ is supported by the College of
149
Engineering at Swansea University.
150
151
References
152
153
Casewell, N. R., Wuster, W., Vonk, F. J., Harrison, R. A., Fry, B. G. 2013 Complex
154
cocktails: the evolutionary novelty of venoms. Trends in Ecology & Evolution. 28: 219-229.
155
156
Cnudde, V., Boone, M. N. 2013 High-resolution X-ray computed tomography in geosciences:
157
A review of the current technology and applications. Earth-Science Reviews 123: 1-17.
158
159
Denson, K. W. 1976 The clotting of a snake (Crotalus viridis) plasma and its interaction with
160
various snake venoms. Thrombosis and Haemostasis 35: 314-323.
161
162
Jackson, K. 2003 The evolution of venom-delivery systems in snakes. Zoological Journal of
163
the Linnean Society 137: 337-354.
164
165
Kak, A. C., Slaney, M. 2001 Principles of Computerized Tomographic Imaging: Society for
166
Industrial and Applied Mathematics.
167
168
Ketcham, R. A., Carlson, W. D. 2001 Acquisition, optimization and interpretation of X-ray
169
computed tomographic imagery: applications to the geosciences. Computers & Geosciences
170
27: 381-400.
171
172
PeerJ PrePrints | http://dx.doi.org/10.7287/peerj.preprints.624v1 | CC-BY 4.0 Open Access | rec: 19 Nov 2014, publ: 19 Nov 2014
PrePrints

8
Kyrieleis, A., Titarenko, V., Ibison, M., Connolley, T., Withers, P. J. 2011 Region-of-interest
173
tomography using filtered backprojection: assessing the practical limits. Journal of
174
Microscopy 241: 69-82.
175
176
Landis, E. N., Keane, D. T. 2010 X-ray microtomography. Materials Characterization 61:
177
1305-1316.
178
179
Limaye, A. 2012 Drishti: a volume exploration and presentation tool. Proc. SPIE 8506,
180
Developments in X-Ray Tomography VIII, 85060X. (DOI 10.1117/12.935640).
181
182
Maire, E., Withers, P. J. 2014 Quantitative X-ray tomography. International Materials
183
Reviews. 59: 1-43.
184
185
Portes-Junior, J. A., Yamanouye, N., Carneiro, S. M., Knittel, P. S., Sant'Anna, S. S.,
186
Nogueira, F. C., Junqueira, M., Magalhaes, G. S., Domont, G. B., Moura-da-Silva, A. M.
187
2014 Unraveling the processing and activation of snake venom metalloproteinases. Journal of
188
Proteome Research 13: 3338-3348.
189
190
Shimokawa, K., Jia, L. G., Wang, X. M., Fox, J. W. 1996 Expression, activation, and
191
processing of the recombinant snake venom metalloproteinase, pro-atrolysin E. Archives of
192
Biochemistry and Biophysics. 335: 283-294.
193
194
Smith, A., Marshall, L. R., Mirtschin, P. J., Jelinek, G. A. 2000 Neutralisation of the clotting
195
activity of Australian snake venoms by snake plasma. Toxicon. 38: 1855-1858.
196
197
PeerJ PrePrints | http://dx.doi.org/10.7287/peerj.preprints.624v1 | CC-BY 4.0 Open Access | rec: 19 Nov 2014, publ: 19 Nov 2014
PrePrints

9
Takacs, Z., Wilhelmsen, K. C., Sorota, S. 2001 Snake a-Neurotoxin Binding Site on the
198
Egyptian Cobra (Naja haje) Nicotinic Acetylcholine Receptor Is Conserved. Molecular
199
Biology and Evolution. 18: 1800-1809.
200
201
Takacs, Z., Wilhelmsen, K., Sorota, S. 2004 Cobra (Naja spp.) Nicotinic Acetylcholine
202
Receptor Exhibits Resistance to Erabu Sea Snake (Laticauda semifasciata) Short-Chain A-
203
Neurotoxin. Journal of Molecular Evolution 58: 516-526.
204
205
Tanaka-Azevedo, A. M., Torquato, R. J. S., Tanaka, A. S., Sano-Martins, I. S. 2004
206
Characterization of Bothrops jararaca coagulation inhibitor (BjI) and presence of similar
207
protein in plasma of other animals. Toxicon. 44: 289-294.
208
209
Vieira, C. O., Tanaka, A. S., Sano-Martins, I. S., Morais, K. B., Santoro, M. L., Tanaka-
210
Azevedo, A. M. 2008 Bothrops jararaca fibrinogen and its resistance to hydrolysis evoked
211
by snake venoms. Comparative Biochemistry and Physiology Part B: Biochemistry and
212
Molecular Biology 151: 428-432.
213
214
Weinstein, S. A., Smith, T. L., Kardong, K. V. 2009 Reptile Venom Glands: Form, Function,
215
and Future. In Handbook of Venoms and Toxins of Reptiles (ed S. P. Mackessy), pp. 65-91:
216
CRC Press.
217
218
219
220
221
222
PeerJ PrePrints | http://dx.doi.org/10.7287/peerj.preprints.624v1 | CC-BY 4.0 Open Access | rec: 19 Nov 2014, publ: 19 Nov 2014
PrePrints

10
Figure captions
223
224
Figure 1. Photographs of an Egyptian saw-scaled viper (Echis pyramidum) that failed to
225
hatch, most likely as a result of complications from a self-inflicted bite. Panel A was taken
226
immediately after removal from the egg (panel E, showing slits from “pipping”) and contains
227
some substrate (vermiculite). The yolk evident in this panel suggests that death occurred prior
228
to the absorption of the yolk mass. The specimen was preserved in 4% paraformaldehyde in a
229
52mm diameter Gosseline 100ml container (F) and packed in cotton wool for shipping and
230
scanning (panel G). LJ = lower jaw.
231
232
Figure 2. Microtomography (µCT) scans show that the fangs (shaded red) are in the folded
233
position and do not penetrate the body. A. whole specimen; B. frontal view; C. magnified
234
view of the head/fang region from A; D. right view; E. left view, with digital dissection to
235
‘remove’ sections of the body for clarity.
236
PeerJ PrePrints | http://dx.doi.org/10.7287/peerj.preprints.624v1 | CC-BY 4.0 Open Access | rec: 19 Nov 2014, publ: 19 Nov 2014
PrePrints

1
Figure 1
Photographs of an Egyptian saw-scaled viper (Echis pyramidum) that failed to hatch, most
likely as a result of complications from a self-inflicted bite. Panel A was taken immediately
after removal from the egg (panel E, showing slits from “pipping”) and contains some
substrate (vermiculite). The yolk evident in this panel suggests that death occurred prior to
the absorption of the yolk mass. The specimen was preserved in 4% paraformaldehyde in a
52mm diameter Gosseline 100ml container (F) and packed in cotton wool for shipping and
scanning (panel G). LJ = lower jaw.
PeerJ PrePrints | http://dx.doi.org/10.7287/peerj.preprints.624v1 | CC-BY 4.0 Open Access | rec: 19 Nov 2014, publ: 19 Nov 2014
PrePrints

PeerJ PrePrints | http://dx.doi.org/10.7287/peerj.preprints.624v1 | CC-BY 4.0 Open Access | rec: 19 Nov 2014, publ: 19 Nov 2014
PrePrints

2
Figure 2
Microtomography (µCT) scans show that the fangs (shaded red) are in the folded position and
do not penetrate the body. A. whole specimen; B. frontal view; C. magnified view of the
head/fang region from A; D. right view; E. left view, with digital dissection to ‘remove’
sections of the body for clarity.
PeerJ PrePrints | http://dx.doi.org/10.7287/peerj.preprints.624v1 | CC-BY 4.0 Open Access | rec: 19 Nov 2014, publ: 19 Nov 2014
PrePrints

PeerJ PrePrints | http://dx.doi.org/10.7287/peerj.preprints.624v1 | CC-BY 4.0 Open Access | rec: 19 Nov 2014, publ: 19 Nov 2014
PrePrints