Material properties of cobweb silk from the black widow spider

Page created by Wade Wagner
 
CONTINUE READING
International Journal of Biological Macromolecules
                                                            24 (1999) 277 – 282

       Material properties of cobweb silk from the black widow spider
                             Latrodectus hesperus

                                              Anne M.F. Moore a,*, Kimly Tran b
                         a
                             Department of Biological Sciences, The Uni6ersity of the Pacific, Stockton, CA 95211, USA
                         b
                             W.M. Keck Science Center, 925 N Mills A6e., Scripps College, Claremont, CA 91711, USA

Abstract

   We present the material analysis of scaffolding silk from the cobweb of the black widow spider Latrodectus hesperus. 30 strands
were tested from the webs of nine spiders. Strands were stretched at 0.211 mm/s as force and extension were recorded.
Cross-sectional area was measured under 1000 × oil-immersion light microscopy. The stress – strain curve shows that cobweb silk
is a distinct material from other known spider silks. The average breaking point for this cobweb silk is 1.1 9 0.5 GPa at
0.22 90.05 strain. All samples increased stiffness as they were stretched, but to different extents. Variation in stiffness might be
due to differential crystallization or alignment of the silk proteins during stretching. © 1999 Elsevier Science B.V. All rights
reserved.

Keywords: Latrodectus hesperus; Black widow; Material analysis

1. Introduction                                                             in the loose ends of the broken thread. Thus, cobweb
                                                                            silk functions by breaking.
   While there has been considerable interest in the                           Despite the diversity in capture mechanisms, all web-
material properties of silks from orb web spiders, silks                    spinning spiders are entirely dependent on their webs
from non-orb web spiders remain largely unstudied. All                      for food, even though silk is metabolically expensive [1].
spider webs capture flying and crawling insects, but                        Thus, spiders have, at least, some selective pressure to
they do so by distinctly different mechanical means. An                     optimize for maximal capture success using minimal
orb web captures prey when an insect sticks to the                          amounts of material. The amount of material can be
spiral capture thread. The thread dissipates the insect’s                   minimized if its properties are optimized to the specific
kinetic energy by stretching and slowly retracting.                         application. In short, the role that silk plays in the
Therefore, an orb web’s capture silk functions by                           process of prey capture might be a potential predictor
stretching, but not breaking. In contrast, a cobweb                         of novel material properties in unstudied silks. It is,
captures an insect when the insect flies into a thread at                   therefore, reasonable to hypothesize that silk from non-
the outer region of the three-dimensional meshwork of                       orb webs will display a wider range of material proper-
the cobweb. As the outer thread breaks, the insect                          ties than is currently known for spider silks.
tumbles into the center of the web while being tangled                         To investigate this possibility, we have studied the
                                                                            material properties of silk from the scaffolding region
                                                                            of cobwebs made by the black widow spider Latrodec-
  * Corresponding author. Tel.: +1-209-946-2182; fax: + 1-209-946-
                                                                            tus hesperus. The silk is mechanically distinct from any
3022.                                                                       other known silk. Its properties seem to be well suited
  E-mail address: amoore@uop.edu (A.M.F. Moore)                             to the function of prey capture in a cobweb.

0141-8130/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved.
PII: S 0 1 4 1 - 8 1 3 0 ( 9 8 ) 0 0 0 9 0 - 7
278                    A.M.F. Moore, K. Tran / International Journal of Biological Macromolecules 24 (1999) 277–282

2. Materials and methods                                                2.4. Measuring cross-sectional area

                                                                           Cross-sectional area was calculated from the diame-
2.1. Spider collection and care                                         ter of the adjacent piece of thread. Silk diameter was
                                                                        measured at 1000× magnification with a compound
   Nine black widow spiders (L. hesperus) were col-                     microscope (Nikon Labophot). The diffraction around
lected at the Bernard Biological Field Station in Clare-                these thin threads limited resolution to an uncertainty
mont, CA. They were housed separately in glass                          range of 97%, which is similar to that in studies of
terrariums containing bamboo structures that provided                   other spider silks [2,4]. The cross-sectional shape was
attachment sites for webs. Each spider was fed a                        estimated to be circular based on two lines of evidence:
cricket nymph one to three times per week. They were                    (1) The diameter did not vary over a 0.5-cm length of
kept at room temperature with an approximate                            silk. (An elliptical cross-section would show a periodic
12/12-h light/dark cycle. Silk samples were taken di-                   narrowing as it twisted in the focal plane [5].) (2)
rectly from the webs that were constructed in the                       Threads appeared to be circular in cross-section in
terrariums.                                                             scanning electron micrographs. (Two views of the same
                                                                        thread, differing by 30°, showed no difference in diame-
2.2. Silk collection                                                    ter. Data not shown.)
                                                                           Those threads that broke near the grips or showed
   Threads of silk were excised from partially con-                     irregularities in cross-section were discounted. A total
structed cobwebs. Cobweb spiders continue to add to                     of 30 silk samples were successfully analyzed.
their web for weeks at a time. In our lab, it takes about
5 days before a cobweb becomes a functional insect                      2.5. Calculating stress and strain
trap. In a mature cobweb, each strand is connected to
several others, forming an irregular, complex network                      To calculate the instantaneous stress, we chose to
of threads. Excising a single strand of suitable length is              assume constant cross-sectional area, rather than con-
difficult at this stage because of the interconnections.                stant volume. The actual behavior during deformation
Therefore, silk was usually collected between 3 and 14                  of this silk lies somewhere between the constant area
days after the spider was transferred to a new terrar-                  and constant volume cases. Given the values of break-
ium. At this time, there were regions of the web with                   ing strains we measured, however, the deformation is
long, unbranched threads. These were the threads col-                   likely to lie close to the constant area case. If constant
lected and tested.                                                      volume and integrated (true) strain are used, higher
   Threads from the scaffolding (top) region of black                   calculated breaking stress results, but there is little
widow webs were excised using a four-pronged probe                      difference in other parameters and no difference in the
to maintain the in-web tension, similar to the process                  general shape of the stress–strain curves. Thus, stress
used by Denny [2]. The silk was cut between the two                     was calculated as the instantaneous force divided by the
central prongs, leaving two separate pieces of silk in-                 original cross-sectional area, and strain was calculated
tact. One piece was transferred to a microscope slide to                as the change in length over the original length.
measure the thread’s diameter. The second piece was
transferred to a cardboard mount for positioning in a                   2.6. Characteristic stiffness
tensometer.
                                                                           Stiffness, or Young’s modulus, in a non-Hookean
2.3. Measuring force and extension                                      material is not a constant value. In this study, the
                                                                        stiffness presented represents the final, higher stiffness
   The second piece of thread was glued to the card-                    of the material. A characteristic value for that region
board mount with nail polish such that 12 mm of the                     was calculated by least squares regression for each
thread was free to stretch. The mount was attached to                   sample between 7 and 15% strain. All samples had
a Chatilon TCN 201 test stand with an Omega LCL                         displayed an increased stiffness by 7% strain and no
113G thin beam load cell in series with the thread to                   strands had broken by 15% strain.
measure force to a resolution of 90.1 mN. The car-
riage of the test stand rose at 0.211 mm/s stretching the               2.7. Toughness
thread at an extension rate comparable to those used to
test other silks [2–4]. Voltage from the load cell was                    Toughness is defined as the energy absorbed before
recorded every 12 s. Force was calculated from the                      breaking per unit volume of material. Toughness was
voltage, and extension from the time and extension                      calculated for each strand by numerically integrating
rate.                                                                   the stress over the full range of strain.
A.M.F. Moore, K. Tran / International Journal of Biological Macromolecules 24 (1999) 277–282                                  279

Fig. 1. Stress – strain curves of L. hesperus scaffolding silk. Traces from all 30 samples are shown on the graph above. Note the wide range in final
stiffness, from 2.4 to 14.2 GPa. The cluster of three traces indicated by closed circles suggests a separate material that is stiffer and stronger than
the material represented by the other traces. However, the histogram of the final stiffness (Fig. 3) shows that these traces are just extremes of the
wide variability of scaffolding silk. (To reduce clutter in this figure, the traces were filtered by three-point-averaging, for this figure only. Properties
presented in the rest of this report are computed from unfiltered data).

3. Results                                                                       strands also show higher strength and slightly lower
                                                                                 breaking strain than the others do. Nevertheless, the
3.1. Shape of stress– strain cur6e                                               histogram of stiffness for the 30 samples (Fig. 3) shows
                                                                                 a single, skewed population rather than a bimodal
   The stress–strain curves for each of the 30 samples                           distribution. Taken as a single class of material the
are presented in Fig. 1. All samples were quite compli-                          mean stiffness for all 30 samples is 69 3 (S.D.) GPa.
ant as stretching began, but began to stiffen at approx-
imately 4% strain. Together the curves present a broad                           3.3. Other properties
continuum of different shapes. They vary from a grad-
ual to an abrupt stiffening as the sample is stretched                              Material properties of L. hesperus scaffolding silk are
beyond 4%. The former extreme results in a J-shaped                              shown in Table 1. With the exception of breaking
curve represented in Fig. 2a, while the latter results in a                      strain, there is greater variation in these parameters
sharp-angle bend with two linear regions, as displayed                           than is normally found within a single material. The
in Fig. 2b. The shape of the curve could not be corre-                           extra variation is largely due to the variation in the
lated to diameter, temperature, relative humidity or age                         shapes of the stress–strain curves. Breaking stress, of
of the web from which the silk was collected.                                    course, varied with the slope, leading to a wide range in
                                                                                 breaking stress. Average strength was 1.19 0.5 GPa.
3.2. Characteristic stiffness                                                    Toughness averaged 136 J/m3, but had a very large
                                                                                 standard deviation 83 J/m3. The strands with more
   The slope between 7 and 15% strain was used to                                J-shaped curves tended to have lower strength and
estimate the stiffness in the stiffer region of the curves.                      toughness. There was no correlation between breaking
This stiffness, for most strands, fell between 2 and 8                           strain and shape of the curve. As can be seen in Table
GPa, however, three strands, represented by circles in                           1, breaking strain had a mean of 0.22 with a standard
Fig. 1, had a stiffness range of 10 – 14 GPa. These three                        deviation of 0.05.
280                     A.M.F. Moore, K. Tran / International Journal of Biological Macromolecules 24 (1999) 277–282

4. Discussion

4.1. Comparison to other silks

   The scaffolding silk of L. hesperus has some similari-
ties and some differences when compared to orb web
silks. Table 2 shows the breaking point of silks from
silk worms and various orb weaving spiders. It is clear
that scaffolding silk shares the high strength and exten-
sion of other silks. Of the silks in Table 2, scaffolding
silk most resembles dragline, frame, radial or MAS silk
of orb weaving spiders, in that the breaking stress and
breaking strain of these two silks are quite similar.
Currently, it is thought that dragline, frame and radial
silk are the same material and that each contains MAS
silk. Nevertheless, it is clear that scaffolding silk and
dragline silk are mechanically distinct materials because
the shapes of their curves are nearly inversions of each
other. Whereas scaffolding silk is first compliant then
                                                                         Fig. 3. Histogram of stiffness. The stiffness of each sample was
stiffens, dragline silk is stiffest when first stretched and             computed by least squares regression between 7 and 15% strain. This
becomes less stiff at about 2% strain. The shape of the                  final stiffness varied by almost an order of magnitude within the 30
scaffolding silk’s stress – strain curve more closely re-                samples. This histogram shows that only one population of silks can
sembles that of viscid silk than dragline silk; however,                 be discerned.

                                                                         viscid silk is ten to twenty times more extensible than
                                                                         scaffolding silk. Clearly, scaffolding silk is a novel
                                                                         material that has not been found in orb web spiders.

                                                                         4.2. Variability in shape of stress–strain cur6e

                                                                            Whereas our preliminary report [6,7] described two
                                                                         mechanically distinct silks from the scaffolding region
                                                                         of this cobweb, the report was based on 13 strands
                                                                         (including all three of the circle traces in Fig. 1). These
                                                                         original 13 strands showed a strongly bimodal distribu-
                                                                         tion of stiffness, shape of curve, breaking stress and
                                                                         breaking strain. Increasing the sample size, however,
                                                                         showed that these two behaviors are actually extremes
                                                                         in a single continuum of mechanical behaviors.
                                                                            This range of behaviors could occur if scaffolding
                                                                         silk constituted a group of similar materials. Casem et
                                                                         al. [8] show that different samples of L. hesperus scaf-
                                                                         folding silk differ in the number and size of polypep-

                                                                         Table 1
                                                                         Material properties of L. hesperus silka

                                                                         Property                                      Mean         S.D.

                                                                         Diameter (mm)                                   4.0         0.8
                                                                         Breaking stress (GPa)                           1.1         0.5
                                                                         Breaking strain                                 0.22        0.05
                                                                         Stiffness (GPa) @ 7–15% strain                  6.0         3.0
Fig. 2. Variation in shape of stress–strain curve. All samples dis-      Toughness (J/m3)                              136          83
played a non-linear stress–strain curve in which the thread stiffened
as it was stretched. The rate of stiffening varied causing a gradient      a
                                                                             Means and standard deviations are shown for various mechanical
from gradually stiffening J-shaped response (a) to an abrupt stiffen-    properties for 30 strand of scaffolding silk from cobwebs of L.
ing, resulting in two Hookean regions (b).                               hesperus.
A.M.F. Moore, K. Tran / International Journal of Biological Macromolecules 24 (1999) 277–282                       281

Table 2                                                                  stretched scaffolding silk [12]. Ordered and disordered
Breaking stress and breaking strain of silks from silkworms and
                                                                         regions have been observed on the same thread imaged
various orb weaving spidersa
                                                                         under AFM, and the ordered regions are more preva-
Silk type                       Breaking stress   Breaking strain        lent in the more highly stretched threads [12].
                                (GPa)
                                                                         4.3. Cobweb function
Bombyx mori cocoon silk [3]     1.1               0.24
Nephila cla6ipes MAS [3]        1.75              0.15
Nephlia maculata dragline       1.1               0.46                      As expected from prey-capture mechanisms, the ma-
Araneus serratus frame silk     0.81b             0.24                   terial properties of silk in the black widow’s cobweb are
  [2]                                                                    distinct from orb-web silks. Nevertheless, the specific
Araneus diadematus radial       1.2               0.40
                                                                         function of this material, and thus its optimization for
  silk [4]
A. serratus viscid silk [2]     1.0b              2.00                   the task must remain speculative because of the com-
A. diadematus viscid silk [4]   1.4               4.76                   plex morphology of a cobweb. Since a cobweb is a
                                                                         highly complex, disordered, three-dimensional mesh-
  a
    Comparisons must be approximate because of variety of strain         work, it is not easy to predict the impact load on any
rates used in different studies.
  b                                                                      single thread. For a cobweb to capture a flying insect,
    Values reported as true stress, otherwise stresses reported are
engineering stress. All breaking strains have been converted to engi-    the impact of the prey must be sufficient to break at
neering strains.                                                         least one thread and the breaking of that thread must
                                                                         slow the forward momentum of the insect to some
                                                                         extent. It is equally important that most of the web
                                                                         remain intact so that the spider can get to the prey,
tides in a protein gel. Studies that correlate mechanical                once it has been immobilized.
behavior with gel analysis have yet to be done.                             Over the course of several days, the spider reinforces
   Whether the scaffolding silk possesses single or multi-               original guy lines and also adds more line. Like the
ple molecular compositions, its range of mechanical                      webs of other Latrodectus species [13], fully functional
behavior must result from variable conformational                        webs of L. hesperus contain widely spaced thick lines to
changes at the molecular level. The scaffolding silk of                  prevent collapse, and large regions of interconnected
L. hesperus stiffens when stretched, a behavior common                   thin lines, that serve as meshwork for catching prey.
to randomly coiled molecules [9]. However, unlike these                  Within these mesh regions, the wide variety of breaking
rubbery materials, scaffolding silk exhibits the same                    stress and toughness of scaffolding silk could work to
high strength as other spider silks. Given that scaffold-                the advantage of the trap. The variety might assure that
ing silk has a high percentage of alanine and glycine [8]                some threads would break, while others remain intact,
and that it is presumed homologous to other spider                       without having to pinpoint the location or direction of
silks, it is reasonable to assume that scaffolding silk has              impact. The spider would be able to reach the immobi-
a primary sequence that is similar, but not identical, to                lized prey along the intact lines. This hypothesis should
other spider silks. Thus, it may be capable of forming                   be tested with behavioral studies of prey capture.
b-sheet crystals within a random coil matrix as has been
shown to occur in other spider silks [10,11]. The initial
low stiffness and later increased stiffness suggests that
                                                                         Acknowledgements
some sort of crystallization or molecular alignment
occurs as the material is stretched past 4%. Whether
                                                                           This material is based upon work supported by the
this stiffening is due to b-sheet formation or some other
                                                                         National Science Foundation under Grant No. DBI97-
alignment is unknown.
                                                                         11031.
   The model of crystal formation or molecular align-
ment is consistent with the diversity of shapes to the
stress–strain curve observed in this material. Differ-
ences in the shape of the curves and stiffness seen                      References
between different samples can be explained if crystal
                                                                          [1] Prestwich KN. The energetics of web building in spiders.
formation occurred at varying rates and to various                            Biochem Physiol 1977;57A:321 – 6.
degrees as each strand was stretched. Thus, a sample                      [2] Denny M. The physical properties of spider’s silk and their role
that exhibits a J-shaped curve, as in Fig. 2a, may                            in the design of orb-webs. J Exp Biol 1976;65:483 – 506.
crystallize gradually as the material is stretched,                       [3] Dunaway DL, Thiel BL, Viney C. Tensile mechanical property
whereas a shape like that shown in Fig. 2b might                              evaluation of natural and epoxy-treated silk fibers. J Appl Polym
                                                                              Sci 1995;58:675 – 83.
appear if there was an abrupt transition to the more                      [4] Kohler T, Vollrath F. Thread biomechanics in the two orb
ordered state. This model is also consistent with atomic                      weaving spiders Araneus diadematus and Uloboris walckenaerius.
force microscopy (AFM) images of stretched and un-                            J Exp Zool 1995;271:1 – 17.
282                   A.M.F. Moore, K. Tran / International Journal of Biological Macromolecules 24 (1999) 277–282

[5] Dunaway DL, Thiel BL, Srinivasan SG, Viney C. Characterizing       [10] Thiel BL, Guess KB, Viney C. Non-periodic lattice crystals in
    the cross-sectional geometry of thin non-cylindrical twisted            the hierarchical microstructure of spider (major ampullate) silk.
    fibers. J Mater Sci 1995;30:4161–70.                                    Biopolymers 1997;41:703 – 19.
[6] Moore AMF, Tran K. Material properties of cobweb silk pro-         [11] Grubb DT, Jelinski LW. Fiber morphology of spider silk: the
    duced by the black widow spider Latrodectus mactans. Am Zool            effects of tensile deformation. Macromolecules 1997;30:2860–7.
    1996;36:54A.                                                       [12] Gould SAC, Tran KT, Spagna JC, Moore AMF, Shulman JB.
[7] Holden C. Super silk. Science 1997;275:163.                             Short and long range order of the morphology of silk from
[8] Casem ML, Turner D, Houchin K. Protein and amino acid                   Latrodectus hesperus as characterized by atomic force mi-
    composition of silks from the cob weaver, Latrodectus hesperus          croscopy. Int J Biol Macromol 1999: in press
    (black widow). Int J Biol Macromol 1999: in press.                 [13] Szlep R. The web-spinning process and web-structure of La-
[9] Vincent J. Structural Biomaterials. Princeton, NJ: Princeton            trodectus tredecimguttatus, L. pallidus, and L. re6i6ensis. Proc
    University Press, 1990.                                                 Zool Soc London 1965;145:75 – 89.

                                                           .
You can also read