Research Undertaken
Introduction
Tetrode recording is a widely used technique for in vivo
chronic recordings of extracellular currents associated with action potentials1.
It is easy to apply in various behavioural tasks, it provides a good
signal-to-noise ratio (SNR) and enables reliable recordings of many neurons at
the same time. The major drawback of this method is that it does not provide
the information about the anatomical arrangement of the recorded cells and
their connectivity. Identifying the exact anatomical location of a particular
cell would permit the correlation of its functional properties with neural type
determined by immunostaining. One solution to this problem could be the application
of tetrode recordings together with Ca2+ imaging of neuronal
population activity and matching regions of interest. Ca2+ imaging
involves detection of the cytosolic Ca+2 concentration related to
the action potentials using fluorescent dyes (e.g. Oregon Green BAPTA (OGB),
calcium green-1 acetoxymethyl (AM) ester, Fura-2 AM etc).2, 3
The aim of my project in Prof. Fritjof Helmchen’s lab in the
University of Zurich was therefore to combine Ca2+ imaging with
tetrode recordings to record in vivo neural activity in the rat barrel
cortex. To our knowledge, this has not been accomplished before.
Progress made
Combining Ca2+ imaging with tetrode recordings involves
two main difficulties. Firstly, acute insertion of blunt tetrodes into neuronal
tissue causes noticeable tissue drag and damage, potentially leading to altered
neuronal responses. Secondly, direct two-photon laser scanning on the tetrode
surface induces a pronounced photoelectric effect, interfering with the
electrophysiological recordings by impairing the SNR4.
In order to address the first difficulty I developed a
non-conventional microdrive design (Fig.1) and worked out how to achieve a
smooth insertion of a tetrode into the brain with minimal damage.
To overcome the photoelectric noise, we applied a random access
scanning system5 developed in Prof. Helmchen’s lab and showed that
we were able to record single action potentials while scanning (Fig.2). We have
evaluated how a noise level depends on the distance from the tetrode to the
closest scanned point (Fig. 3). We could show that the closer the scanned point
is to the tip of a tetrode the bigger the noise level becomes, allowing for
only a narrow ‘working window’ (Fig. 3, the yellow box) where the SNR is
sufficient to detect single action potentials of nearby neurons. Previous
studies of chronic tetrode recording suggest that tetrodes (made with 17 µm
wires) can record signals from cells located up to 100 µm away.
We have also investigated how noise level depends on the laser
intensity. We showed that it increases drastically with an increase of laser
power (Fig.3). Depending on the properties of imaged area of interest the laser
intensity used is typically between 150-300 mW. In conclusion, our findings
showed that we should be able to use laser intensity large enough to detect Ca2+
events and scan close enough to the tetrodes to be able to record from the same
cells using both Ca2+ imaging and tetrodes.
Finally, we have noticed that using too much laser power for
prolonged periods of time may be destructive to cells surrounding the tetrodes.
This may be due to the currents flowing from the tetrodes created by the
photoelectric effect induced by laser scanning. Further studies will therefore
have to be carried out to determine the scanning time and laser intensity which
could potentially lead to cell damage.
References
1. Andersen, P. et al. The Hippocampus Book. Oxford University Press (2007).
2. Svoboda, K. et al. Nature. 385, 161 (1997).
3. Stosiek, C. et al. PNAS. 100, 7319 (2003).
4. Shew, W. L. et al. J of Neurosci Methods 192, 75 (2010).
5. Grewe, B. F. et al. Nat Methods 7, 399 (2010).
6. Nimmerjahn et al. Nat Methods 1, 31 (2004).
Page last modified on 26 jan 11 04:41
