went below \u20132.4 volts, which switched the
device of . T is on/of behaviour allows the
structure to be used as a memory device.
T is switching ef ect also occurs i\ue004
the \ue003MV is replaced by a non-biological
polymer nanorod o\ue004 a similar size10. But
the dif erence in switching speed (\u2018on\u2019 to
\u2018of \u2019 state) indicates that both the shape and
(bio)chemistry o\ue004 the nanorod can in\ue001uence
the system. T is leaves plenty o\ue004 room \ue004or
optimizing the device. When the platinum
particles were not connected to a rod-like
structure but randomly distributed within
the polymer, the system \ue004ailed to display
this switching ef ect. Instead, the authors
saw a relatively high conduction between
nanoparticles occurring through electronic
Yang and colleagues speculate that in
the absence o\ue004 applied voltage, there is a
charge trans\ue004er between the nanoparticles
and the \ue003MV. T is blocks the tunnelling
between platinum nanoparticles and results
in an \u2018of \u2019 state (Fig. 1a). A su\ue002 ciently high
voltage reverses the charge trans\ue004er and once
again allows current to \ue001 ow by tunnelling
processes between the nanoparticles, thereby
turning the device \u2018on\u2019 (Fig. 1b). T is charge
trans\ue004er apparently can be switched back
and \ue004orth reversibly many times.
What is the outlook \ue004or practical use
o\ue004 these virus-based memory elements?
Current electronics operate at gigahertz
\ue004requencies and there is a push towards
terahertz operation. Although the
viruses can currently be switched at
megahertz \ue004requencies at best, and with
a comparatively low number o\ue004 possible
switching cycles, this disadvantage could be
a stimulus \ue004or more-detailed investigation
and improvement o\ue004 the principle. I\ue004 we
understand how to design electronic
systems that can reliably communicate
with biological systems in a controlled
On the basis o\ue004 what we will learn \ue004rom
this primitive biological/nanoparticle hybrid
memory device and \ue004rom \ue004uture experiments,
we might \ue000 nally begin to see new commercial
electronic devices made with biological
components including virus-based memory
sticks. And i\ue004 we are unlucky, these might one
day be in\ue004ected by \u2018real\u2019 computer viruses.
6. Beijerinck, M. J. Verh. Kon. Akad. Wetensch.65, 3\u201321 (1898).
7. Kausche, G. A. & Ruska, H.Kol l oid - Z .89, 21\u201326 (1939).
8. Shenton, W., Douglas, \ue003., Young, M., Stubbs, G. & Mann, S.Ad v.
are at1Centre for Health and Environment, School
of Life Sciences, Napier University, Edinburgh EH10
5DT, UK;2MRC/University of Edinburgh Centre for
In\ufb02 ammation Research, Queen\u2019s Medical Research
Institute, Edinburgh EH16 4TJ, UK.
nanoparticles that vary in size,
shape, charge, chemistry, coating and
solubility. \ue003ake carbon nanotubes, \ue004or
example, which have been intensively
studied because they have new and
unusual mechanical, electronic and other
properties. T e potential toxicity o\ue004 these
materials has attracted attention because o\ue004
their apparent similarities to asbestos and
other carcinogenic \ue000bres (Fig. 1). Carbon
nanotubes are long, thin (just nanometres
in diameter) and insoluble \u2014 all \ue004actors
that contribute to \ue000 bre toxicity in the
lungs1. A study by Andre Nel o\ue004 the
University o\ue004 Cali\ue004ornia, Los Angeles and
co-workers now suggests that the hazards
are best predicted by examining which
However, when testing the toxicity
induced by carbon nanotubes, should we
consider single- or multiwalled tubes? Long
or short nanotubes? How long or short?
Should we remove metal catalysts? Do we
use \ue004unctionalized or non-\ue004unctionalized
particles? Should we use pristine tubes
or should they be tested in the \ue004orm in
which they might actually be used? T e
list o\ue004 variations is endless and poses a real
problem \ue004or toxicologists.
However, the challenge \ue004or
toxicologists is not to test every variation
o\ue004 a new nanoparticle generated but,
instead, to identi\ue004y key \ue004actors or tests
that can be used to predict toxicity,
permit targeted screening and allow
materials scientists to generate new, sa\ue004er
nanoparticles with this structure-toxicity
in\ue004ormation in mind.
Nel and co-workers have now taken a
major step in this direction by systematically
studying how nanomaterials can induce
oxidative injury within cells2. T e basic idea
is that i\ue004 we know how this process occurs,
we can begin to compare the toxic potential
nanotubes that are longer than the distance that
the cell can stretch, which means that the rat
cannot remove such nanotubes from the body.
This optical microscopy image is superimposed
with confocal images of the protein cytoskeleton
that gives the cell structure and its ability to move.
F-actin is shown in red; tubulin in green. (Image
This action might not be possible to undo. Are you sure you want to continue?