Mechanisms for the light-cell interface in optical
neurostimulation
Master’s thesis
for the acquisition of the academic degree
Master of science
from the University of Veterinary Medicine Hannover
Created at Laser Zentrum Hannover e.V.
Biomedical optics department
Funded by the Cluster of Excellence “Hearing4all”
Submitted by
Sonja Johannsmeier
Hannover, 2016
Supervisor:
Dr. Dag Heinemann
Biomedical optics department
Laser Zentrum Hannover e.V.
First examiner:
PD Dr. Hugo Murua Escobar
Division of Medicine
University of Rostock
Second examiner: Dr. Dag Heinemann
Biomedical optics department
Laser Zentrum Hannover
Declaration of Authorship
I certify that the presented master thesis with the title “Mechanisms for the lightcell interface in optical neurostimulation” is my own unaided work and no other
sources and tools than stated were used. This thesis was not previously presented to another examination board and has not been published in any form.
Place, date
Sonja Johannsmeier
Abstract
I
Abstract
Advances in the field of biomedical engineering have pushed optical methods of
neurostimulation further to the edge of therapeutic applications. Different approaches have revealed a broad spectrum of possibilities, and the underlying
mechanisms have largely been unraveled. Considerable effort is made to refine
the methods and introduce additional flexibility. However, the overall cellular reaction is not always sufficiently understood. The safety and reliability of a new
therapeutic device can only be assessed if its effects on the targeted cells, tissue
or organ is known. Direct or nanoparticle-mediated laser stimulation of neurons
presents an uncomplicated way of performing laser-based cell manipulation. An
increase of membrane capacity has been detected as a way to evoke action potentials in response to a laser stimulus. The method might therefore be suited for
applications in neuroprosthesis. But before it can be introduced into a patient, the
processes at the light-cell interface need to be fully understood. In this thesis,
calcium signaling, formation of reactive oxygen species and membrane perforation in response to gold nanoparticle-mediated laser stimulation have been studied. Also, the influence of different inhibitors and surrounding media was tested.
Single cell manipulation of a large number of neuro-2a cells and primary mouse
cortical neurons has revealed a multicomponent cellular reaction involving intracellular calcium release, calcium influx through transient receptor potential channels and calcium-induced calcium release. These findings imply a serious interference of the external stimulus with cellular homeostasis, potentially leading to
long-term damage. While the effect depends on the cell type and the physiological environment, any long-term in vivo application of nanoparticle-mediated laser
stimulation must be evaluated with great care.
II
Table of contents
III
Table of contents
Abstract ......................................................................................................................... I
List of abbreviations ...................................................................................................... V
List of tables ................................................................................................................ VI
List of figures .............................................................................................................. VII
1
Introduction and theoretical background .............................................................. 1
1.1
Electrical stimulation .............................................................................. 1
1.2
Optical neurostimulation ......................................................................... 2
1.2.1
Infrared neural stimulation ........................................................ 3
1.2.2
Nanoparticle-mediated stimulation............................................ 3
1.2.3
Transmitter uncaging ................................................................ 4
1.2.4
Optogenetics ............................................................................ 5
1.3
Calcium signaling ................................................................................... 5
1.4
Aim of the thesis .................................................................................. 10
2
Methods ............................................................................................................. 11
2.1
Cell culture ........................................................................................... 11
2.1.1
Neuro-2a cells ........................................................................ 11
2.1.2
Primary mouse cortical neurons ............................................. 12
2.2
Immunofluorescence ............................................................................ 13
2.3
Laser stimulation .................................................................................. 14
2.3.1
Experimental setup ................................................................. 14
2.3.2
Calcium imaging ..................................................................... 14
2.3.3
Stimulation experiments ......................................................... 16
2.3.4
Image analysis ....................................................................... 17
2.4
Lipid peroxidation ................................................................................. 18
2.5
Statistical analysis and data visualization ............................................. 20
3
Results .............................................................................................................. 21
3.1
Immunofluorescence ............................................................................ 21
3.2
Laser stimulation .................................................................................. 23
3.2.1
Laser-induced calcium transients and energy dependency..... 23
3.2.2
Inhibitor studies ...................................................................... 28
3.2.3
Atypical behavior and membrane perforation.......................... 34
3.3
Lipid peroxidation ................................................................................. 41
4
Discussion ......................................................................................................... 44
4.1
The N2A cell line presents a suitable model .......................................... 44
4.2
Laser stimulation does not cause electric activation ............................. 45
4.3
The calcium response has different extra- and intracellular sources .... 47
4.4
Membrane perforation can cause deviations in the calcium response, but
does not generally affect it ................................................................... 52
4.5
Laser stimulation causes lipid peroxidation on a similar timescale ....... 54
IV
Table of contents
5
Conclusion and outlook...................................................................................... 57
6
References ........................................................................................................ 59
7
Supplementary materials ................................................................................... 67
7.1
Methods ............................................................................................... 67
7.1.1
Cell culture ............................................................................. 67
7.1.2
Preparation for laser stimulation ............................................. 69
7.1.3
Laser stimulation .................................................................... 71
7.1.4
Immunofluorescence .............................................................. 71
7.2
Materials .............................................................................................. 73
7.3
Supplementary results ......................................................................... 77
8
Acknowledgements............................................................................................ 87
List of abbreviations
V
List of abbreviations
AuNP
CICR
DAG
ER
FCS
FDR
Fluo 4-AM
INS
IP3
MCN
MEM
MPT
nACh receptor
NCLX
NF
NMDA-R
PBS
PBSC
PI
PIP2
PLC
PMCA
ROS
RyR
SOC
SYP
TRP
UV
VDCC
Gold nanoparticles
Calcium-induced calcium release
Diacylglycrol
Endoplasmatic reticulum
Fetal calve serum
False discovery rate
Fluo 4-acetoxymethyl ester
Infrared neural stimulation
Inositol trisphosphate
Mouse cortical neuron
Minimum essential medium
Mitochondrial permeability transition
Nicotinic acetylcholine receptor
Na+/Ca2+-exchanger
Neurofilament
N-methyl-D-aspartate receptor
Phophate buffered saline
PBS with Ca2+ and Mg2+
Propidium iodide
Phosphatidylinositol-4,5-bisphosphate
Phospholipase C
Plasma membrane Ca2+ ATPase
Reactive oxygen species
Ryanodine receptor
Store operated channel
Synaptophysin
Transient receptor potential
Ultraviolet
Voltage dependent calcium channel
VI
List of tables
List of tables
Table 1: Laser stimulation of different cell types at various radiant exposures ............ 25
Table 2: N2A cells displaying a PI signal and atypical behaviour ................................ 38
Table S 1: List of equipment and consumables ........................................................... 73
Table S 2: List of software .......................................................................................... 74
Table S 3: List of chemicals, solutions and dyes ......................................................... 74
Table S 4: List of inhibitors .......................................................................................... 76
Table S 5: List of antibodies. ....................................................................................... 76
Table S 6: Adjusted p-values of pairwise comparisons for ΔF/F (N2A cells). .............. 83
Table S 7: Adjusted p-values of pairwise comparisons for time to peak (N2A cells) .... 84
Table S 8: Adjusted p-values of pairwise comparisons of ΔF/F (MCNs). ..................... 85
Table S 9: Adjusted p-values of pairwise comparisons of time to peak (MCNs) .......... 85
Table S 10: Results of lipid peroxidation experiments ................................................. 86
List of figures
VII
List of figures
Figure 1: Illustration of different mechanisms of calcium uptake and storage ................ 8
Figure 2: N2A cells in medium + 10 % FCS and after three days of serum deprivation
.......................................................................................................................... 12
Figure 3: Primary mouse cortical neurons 7 days and 14 days after thawing .............. 13
Figure 4: Single cell manipulation setup ...................................................................... 14
Figure 5: Transition of Fluo 4 from non-fluorescent to the calcium-bound, fluorescent
form ................................................................................................................... 15
Figure 6: Depiction of the evaluated variables ΔF/F and time to peak ......................... 18
Figure 7: Oxidation of BODIPY 581/591...................................................................... 19
Figure 8: Immunofluorescence, control cells ............................................................... 21
Figure 9: NF and SYP in differentiated N2A cells. ....................................................... 22
Figure 10: SYP and NMDA-R in differentiated N2A cells. ........................................... 23
Figure 11: Typical fluorescence signal of an N2A cell in response to laser stimulation 24
Figure 12: Boxplots of ΔF/F and time to peak at different radiant exposures for N2A cells
and MCNs.......................................................................................................... 27
Figure 13: ΔF/F and time to peak for N2A cells in different conditions ........................ 29
Figure 14: ΔF/F and time to peak for MCNs in different conditions.............................. 31
Figure 15: Peak ΔF/F of spontaneous transients in non-irradiated N2A cells .............. 32
Figure 16: A) Proportions of non-irradiated cells displaying a spontaneous calcium
transient. B) Proportions of irradiated cells exhibiting an additional calcium
transient after a laser-induced calcium response ............................................... 33
Figure 17: A, B) Examplary signatures of fluorescence signals that reached their peak
value later than the average or not at all during image acquisition C) Occurrence
of outliers in ΔF/F. D) Occurrence of outliers in time to peak ............................. 35
Figure 18: Signature of PI influx over time .................................................................. 36
Figure 19: Percentage of N2A cells in culture medium undergoing membrane damage
after irradiation with different radiant exposures ................................................. 39
Figure 20: Proportion of PI-positive cells after laser irradiation under the different
conditions .......................................................................................................... 40
Figure 21: Lipid peroxidation in N2A cells and MCNs under different conditions ......... 42
Figure S 1: ΔF/F and time to peak vs. baseline fluorescence values of N2A cells and
MCNs ................................................................................................................ 78
Figure S 2: ΔF/F and time to peak for N2A cells in different conditions. ...................... 79
Figure S 3: Laser-induced calcium response of N2A cells under different conditions
grouped by presence or absence of a detectable PI-signal. ............................... 80
Figure S 4: Distribution of peak ΔF/F values in N2A cells without or with a slower
(spontaneous) transient following the laser-induced calcium response. ............. 81
Figure S 5: ΔF/F values of BODIPY 581/591 oxidation ............................................... 82
VIII
1
1
Introduction and theoretical background
1
Introduction and theoretical background
Neurostimulation has come a long way since for the first time, Luigi Galvani famously made a dead frog dance [1]. Technical advances have made external
stimulation of excitable cells safely available for a living target group since the
1950s [2]. While the first devices were cardiac pacemakers, stimulation of the
peripheral [3] and central [4] nervous system were soon to follow. Advances in
battery life and downscaling of the technology have allowed for the manufacturing
of reliable, convenient devices for long-term cell stimulation. However, the mechanism of electric cell stimulation per se has its limitations which sometimes make
a therapy impossible. The search for improvement has led to the emergence of
the new field of optical cell stimulation. First observations were less targetoriented and sometimes contradictory [5, 6]. But with time, methods were developed that show potential to improve existing technology or to pursue whole new
possibilities. These methods are still being improved and some of the underlying
mechanisms still need to be unraveled.
The present work will first give an overview of the current state of neurostimulation with its chances and pitfalls. Effects on the cell’s physiology will be described
as well. Secondly, an overview of the experimental methods is given. The results
of optical stimulation experiments are then presented and discussed, and a final
conclusion will connect the outcomes to possible applications of optical neurostimulation.
1.1
Electrical stimulation
The electrical stimulation of excitable cells found its way into medical therapy in
the form of pacemakers in the 1950s [2]. Today, there is a variety of medical
indications for stimulation devices, such as cardiac arrhythmia, epilepsy [7], Parkinson’s disease [8], chronic pain [9] or hearing loss [10]. Depending on the pathology, stimulation can target muscles, the peripheral nervous system, the spinal
cord, nerves of the CNS or the deep brain [9–12]. From rather mild modulations
of the vagus nerve and its targets to profound interferences in deep brain stimulation and live-saving pacemakers, the mechanisms of electrical neurostimulation
2
1
Introduction and theoretical background
are basically the same. A current is applied to the organ or nerves in question.
This is often achieved with implantable electrodes, but broad electric fields can
also be administered transcutaneously [13]. The current changes the cells’ membrane potential which will result in a reaction depending on the cell type – muscle
cells will contract if a certain threshold is exceeded, neurons will fire action potentials. In the case of neurostimulation, the stimuli can be excitatory or inhibitory,
thus positively or negatively modulating activity. The impact on the organism is
determined by the way the signal progresses to downstream targets.
In the case of sensory devices like the cochlear implant, the aim of neurostimulation is to mimic sensory activity. Since the cochlear encodes acoustic wave frequency by precise spatial distribution of bioelectric signals, the external device
needs to place the right stimulus at a very precise location to be interpreted as
the right tone pitch [14]. However, this is where electric cell stimulation reaches
its limits: Due to the imprecision of electric field propagation and the resulting
electronic crosstalk, a number of neurons will always be excited and a natural
hearing impression is far from being achieved [15, 16]. The same phenomenon
causes side effects in other therapeutic attempts, ranging from mild tingling or
twitching to pain or sleep disturbances that outweigh the benefits of the therapy
[7, 11]. A search has been going on for a way to maintain the merits but avoid the
adverse effects of these measures.
1.2
Optical neurostimulation
Biophotonics play a major role in modern biomedical research and engineering.
While imaging technologies are probably the most wide-spread and best evaluated applications, advances in molecule delivery [17–20] and cancer therapy [21–
25] have opened up new possibilities for cell manipulation.
Unlike an electric field, precisely applied light pulses would not interact with neighboring cells. Although it is not an adequate stimulus for most cells, electric activity
was found to be modifiable by UV, red and infrared light [26]. When efforts to
achieve targeted neurostimulation became more purposeful, the first method of
choice was using infrared light to evoke currents in nerve cells.
1
Introduction and theoretical background
3
1.2.1 Infrared neural stimulation
Water is a good absorber of infrared light. Thus, when using infrared neural stimulation (INS), there will always be an absorber in the laser focus, since water is
ubiquitous in biological systems. Using a pulsed infrared laser, the light will therefore rapidly heat up water molecules in the cell, leading to a short but pronounced
temperature transient [27]. Other than previously hypothesized, the resulting neural stimulation does not necessarily rely on membrane channel activation [28].
Shapiro et al. demonstrated that the temperature transient increases the membrane capacity even in artificial bilayers, leading to depolarizing currents and possibly to neuronal activation. However, temperature-sensitive channels may also
take part in the cell’s overall reaction [29]. Since infrared light is absorbed by
water all through the biological tissue, it might lack the desired precision. Also, if
too much heat is conferred to confined areas, the stimulation will damage the
cells [30]. A more precise way of absorbing the light stimulus is presented by the
use of nanomediators.
1.2.2 Nanoparticle-mediated stimulation
When irradiated at their resonance frequency, gold nanoparticles form plasmons
and heat up very quickly [19, 31]. This heat is transferred to the cell membrane
and can induce the same effects as described for direct INS. In the case of nanorods, the resonance frequency can be tuned – within a certain range – by varying the aspect ratio to meet the application’s requirements [32]. While spherical
nanoparticles have an absorption maximum at around 530 nm (depending on the
radius [33]), nanorods can be excited with near infrared light [32]. Water does not
or only weakly absorb light of these wavelengths, unspecific bulk heating of the
aqueous medium can therefore be avoided. Other materials can be used as nanomediators as well [34].
The heat developed at the particles can trigger opening of heat sensitive calcium
channels [35], thereby depolarize the cell and evoke action potentials. More importantly, neurons can be depolarized by an increase of their membrane capacitance, which is conferred by heat and independent from membrane channels [28],
[31]. Exploiting this mechanism, the generation of action potentials has been
demonstrated with high spatio-temporal precision in cultured primary cells and
4
1
Introduction and theoretical background
brain slices [31, 36, 37] as well as the generation of compound nerve action potentials [27, 38]. The effect can be enhanced by binding the nanoparticles to the
cell membrane using antibodies. By this means, the stimulation threshold is decreased. Furthermore, particles are not washed out by liquid perfusion [31, 37].
However, the exact way in which a cell responds depends on the cell type and
intrinsic variations which makes fine-tuning of this stimulation method very difficult. Also, like INS, stimulation can only have an excitatory effect on the neuron
as it always increases the membrane potential. Although Carvalho-de-Souza et
al. found some cells to be stable stimulated over the course of 3000 action potentials [31], the long-term stability of the system and effects on cellular health
have not been investigated.
A growing concern is cytotoxicity of the nanoparticles. Results vary markedly with
the particles’ properties such as size, aspect ratio, surface charge or coating [39]–
[43]. In general, some cytotoxicity is found in in vitro studies, especially when
applying precursor solutions (AuCl4) or nanoparticles together with stabilizing
chemicals [44]. Certain coatings appear to have protective properties [42, 43]. In
vivo studies convey an ambiguous picture [40]. While some studies reported no
toxicity or lethality after long-term exposure, others pointed out severe illness and
increased lethality in test animals. A common test system is not given at this time.
Different coatings have been developed to reduce adverse side effects, and targeted delivery by antibody binding diminishes the amount of necessary particles,
but no nanoparticles have been manufactured to specifically tackle the concern
of long-term toxicity. As routine application in medical therapies might be approaching, this is still a problem for researchers to solve.
1.2.3 Transmitter uncaging
Another approach to mimic neuronal activity relies on manipulating chemical
compounds that have a natural role in cell physiology. By adding a so called caging group to a bioactive compound, it is no longer recognized by its receptors and
therefore rendered inert. These compounds can be photolytically cleaved, meaning that the chemical is set free by a light pulse of the right wavelength and sufficient energy, and can now act upon its targets. Available caged compounds include excitatory neurotransmitters like glutamate [45], inhibitory transmitters like
1
Introduction and theoretical background
5
GABA [46], or second messengers like ATP, IP3 and even Ca2+ [47]. This selection of molecules has allowed for observations in neuronal networks and synapses at a new level of spatio-temporal resolution [46, 48, 49]. The method has not
yet been evaluated for long-term in vivo applications. A steady supply and introduction of the caging compound together with the signaling molecule pose potential drawbacks. On the other hand, the selection of diverse caged compounds
would enable researchers to develop flexible methods specifically designed for
different applications.
1.2.4 Optogenetics
This rather new field completely circumvents the need for mediators by rendering
the cells susceptible to light. Several light-activated ion channels, so called opsins, have been isolated from bacteria, archaea, fungi and algae, and transferred
into other cell types [50]. Both cation and anion channels are available for transfection. Further screening has revealed proteins that operate on different time
scales and are activated by distinct wavelengths and different energies [51, 52].
The drawback of this method is the need for transfecting the cell, tissue or organ
that is to be treated. While there are sophisticated ways for targeted transfection
in an experimental environment, it has not yet been approved for use outside of
laboratories or clinical trials. Still, optogenetic manipulation appears to be a way
for precise and safe optical cell manipulation, with applications not limited to excitable tissue.
1.3
Calcium signaling
In the course of cellular evolution, calcium has turned out to be the single most
important universal messenger [53, 54]. At rest, the concentration of free intracellular calcium ions amounts to roughly 100 nM, while extracellular concentrations often reach a few millimoles per liter [54]. On top of this 10,000-fold difference constituting a steep chemical gradient across the cell membrane, a healthy
cell’s membrane potential is usually negative at rest. This creates an additional
electric gradient for the divalent cation. When calcium is allowed to enter the cell
6
1
Introduction and theoretical background
from the external medium, large amounts of the ion can potentially flood the cytoplasm in very short time. However, because of its function as a messenger
molecule, calcium entry must be tightly regulated. To maintain the electrochemical gradient, plasma membrane Ca2+ ATPases (PMCAs) extrude Ca2+ driven by
the chemical energy from hydrolyzing ATP. Na+/Ca2+-exchangers transport calcium out of the cell in exchange for three sodium ions [54]. Under pathological
conditions, these channels can run in reverse and allow calcium to enter the cytoplasm [55]. However, the most important ways for calcium entry and signaling
across the plasma membrane in neurons are ligand-gated ion channels, voltagedependent calcium channels (VDCCs), transient receptor potential (TRP) channels and store-operated channels (SOC) [54–57]. Figure 1 illustrates the different
paths of calcium flux.
Ligand-gated cation channels – namely glutamate receptors and nicotinic acetylcholine (nACh) receptors – are especially important at the chemical synapse [58].
Calcium mediates the fusion of presynaptic vesicles with the plasma membrane
and is therefore crucial for the release of neurotransmitters. This process can be
modulated by presynaptic nACh receptors. At the postsynaptic membrane, different glutamate receptors possess different permeabilities for calcium [59]. Apart
from conveying an excitatory postsynaptic potential, calcium also carries a signal
that can lead to synaptic rearrangement and long-term potentiation [60].
Voltage-gated calcium channels can substantially raise the intracellular calcium
concentration very quickly [54]. There are different types of channels that are
activated at specific voltage levels. Some channels are especially found in certain
organs – P-type channels are named after their primary locality, Purkinje neurons,
and L-type channels are found in cortical neurons, but also in muscle and bone
[57]. VDCCs are activated by changes in the membrane potential, therefore their
activity particularly accompanies action potentials. The channels’ main function
is to ensure fast propagation and modulation of action potentials, transmitter release and regulation of gene transcription [57]. These properties have enabled
researchers to map action potentials across neuronal populations by recording
the accompanying calcium signal with fluorescent dyes and applying temporal
deconvolution methods to overlapping signals from high firing rates [61–63].
1
Introduction and theoretical background
7
The fundamental insights into the nature of TRP channels originally stem from
invertebrate studies [64]. Members of this superfamily are however common to
many mammalian cell types, both excitable and non-excitable. The different subfamilies will not be described in detail here. However, the genetic identity provides
some information about activation and function of a certain channel type. Multiple
stimuli can activate TRP channels [65] and while their expression is not restricted
to excitable cells, an important purpose appears to be to translate sensations
from the outside world to the organism and from the direct surrounding to the cell
[66]. The cation channels of the TRPC subfamily react signaling from within the
cell. They can be activated by IP3, PIP2, Ca2+ and calmodulin, DAG or phosphorylation [65] and are thereby deeply embedded into intracellular signaling
pathways. Members of other subfamilies respond to temperature changes, mechanical stimuli, changes of the membrane potential, endogenous and exogenous chemical ligand, pH or changes in the cell’s redox status [65]. Hence, these
channels encode a multitude of information and enable cells, organs or organisms to properly react to changes in their environment. Heat-activated channels
of the subfamily TRPV1 have also been suggested as the mediators of neural
laser stimulation, particularly INS [32, 67, 68].
Store-operated channels are activated by depletion of intracellular calcium stores
[56]. Upon emptying of the ER, a sensor protein (stromal interacting molecule,
STIM1) forms junctions at the cell membrane, which cause localized inflow of
calcium and replenishment of the stores [69]. The sensor does not react to high
cytoplasmic calcium levels, but to low concentrations within the ER. The SOCs
are therefore specialized in activation and function. Some members of the TRP
superfamily have also been associated with SOC functionality [70].
8
1
TRPchannels
SOC
Introduction and theoretical background
stimulus
VDCCs
Ligandgated cation
channels
PIP2
PMCA
G
IP3
DAG
PLC
IP3R
MPT
pore
RyR
MCU
downstream
targets
NCLX
Figure 1: Illustration of different mechanisms of calcium uptake and storage. PMCA = plasma
membrane Ca2+-ATPase; SOC = store operated channel; TRP = transient receptor potential;
VDCC = voltage dependent calcium channel; PIP2 = phosphatidylinositol-4,5-bisphosphate; IP3 =
inositol trisphosphate; DAG = diacylglycerol; PLC = phospholipase C; IP3R = IP3 receptor; RyR =
ryanodine receptor; MPT = mitochondrial permeability transition; MCU = mitochondrial calcium
uniporter; NCLX = sodium/calcium exchanger. Calcium can enter from the extracellular space
through several types of channels that are activated by a physical stimulus (VDCCs, some TRP)
or a biochemical compound (light green; ligand-gated, SOC). Calcium influx through SOCs occurs
upon ER-depletion and is directed immediately to the ER. Generally, an elevated c(Ca 2+)i serves
a signaling purpose and the ions act on diverse downstream targets. Cytoplasmic Ca 2+ levels can
also be increased from intracellular stores. A G-protein coupled receptor mediates signals from
various extracellular components: The G-protein is released and activates PLC which in turn
cleaves PIP2 to release IP3 and DAG. IP3 can then act on its receptors in the ER-membrane to
release Ca2+. The calcium ions further stimulate Ca2+ release from IP3R and RyR. Excess calcium
is pumped out of the cell by PMCAs or sequestered by mitochondria. Passive uptake via MCUs
is the main mechanism at high cytosolic calcium concentrations. The NCLX, when operating in
forward mode, plays an important role in calcium extrusion. The MPT pore is closed under physiological conditions.
As a universal messenger, Ca2+ is involved in a multitude of signaling pathways
that will not be described in depth here. Because of its ubiquity, it provides a good
means of monitoring cellular processes. The endoplasmic reticulum and the mitochondria are of particular importance as they can act as both a calcium source
and sink. The common way in which calcium is released from the ER is by activation of phospholipase C (PLC) via G-protein coupled receptors, and subsequent formation of IP3 and DAG. IP3 binds to its receptor in the ER membrane,
which in turn releases calcium into the cytosol [71]. The IP3 receptors as well as
1
Introduction and theoretical background
9
the ryanodine receptors (RyR) of the ER membrane first amplify the calcium signal, as they are stimulated by the ion itself; a mechanism termed calcium-induced
calcium release (CICR). After a short time, Ca2+ then deactivates the channels
and stops calcium release. Both the ER and mitochondria can store calcium concentrations in the low millimolar range [54]. Under physiological conditions, the
calcium concentration within the mitochondria resembles that of the cytosol. So
an important function of the mitochondria lies in this exact ability to sequester
large amounts of calcium and prevent an overload of the cell [71]. Both organelles
are of great importance for intracellular signaling and homeostasis. Their pathways are further interconnected by the mitochondria-associated membranes [72].
When undertaking cell manipulation, attention should not only be paid to physiological signals. Calcium is also an important messenger of the cellular stress response, and can itself pose great stress on the cell. Many downstream targets
are affected by Ca2+. In physiological signaling pathways, it activates proteins
such as calmodulin by conformational change, which translate the ionic signal to
actions. Protein activity can encode level and duration of the calcium signal and
evoke short-term responses like modulation of downstream targets, or long-term
modification like alteration of gene transcription [54]. In excess amounts however,
calcium can damage organelles and macromolecules [73]. In mitochondria, calcium interacts with various targets. At higher concentrations, this can result in
increased activity of the respiratory chain [74] and an increased membrane potential [75], and both mechanisms lead to an increase in reactive oxygen species
(ROS) production. The same is true for depolarization of the mitochondrial membrane, for example by excess calcium influx. Under normal conditions, low
amounts of ROS are a by-product of the respiratory chain and can be rendered
harmless by the cell [53]; they are even assigned to a role in physiological signaling [74]. But accumulation of high ROS and/or calcium concentration can lead to
opening of the mitochondrial permeability transition pore (MPT), an event that
has been described as “catastrophic” for the cell by Duchen [76]. Through this
non-selective megachannel, large amounts of calcium and ROS can enter the
cytoplasm, as well as cytochrome c and other apoptosis-inducing factors. Free
ROS, if not caught by the cell’s defense mechanisms, can damage proteins, nucleic acids and lipids. Damaged macromolecules and disruption of cellular homeostasis will trigger the cellular stress response which will transition into apoptosis
10
1
Introduction and theoretical background
if the stress level is too high [53]. High levels of calcium and oxidative stress can
also promote necrosis [73].
ROS can also stem from external sources. However, it is important to keep in
mind the complex mechanisms and delicacy of cellular homeostasis in order to
assess short-term or long-term effects of cell manipulation, although the ubiquity
of calcium and the countless interactions of its physiological and pathological
downstream targets do not make this an easy task.
1.4
Aim of the thesis
Triggering action potentials via gold nanoparticle-mediated neurostimulation has
already been demonstrated. However, there is a considerable lack of information
on its effect on cell physiology and health. Also, involvement of other mechanisms
might open up new possibilities of this method’s application in basic research and
technical implications. The aspiration of the present work is to thoroughly elucidate the cellular response to gold nanoparticle-mediated neurostimulation and
derive implications for further use of this method. A neuroblastoma cell line and
primary mouse cortical neurons were used as a model. By examining magnitude
and temporal characteristics of the calcium signal, lipid peroxidation and membrane perforation on a single cell level, data from different endpoints were integrated to provide a hypothesis explaining the overall cellular reaction to gold nanoparticle-mediated laser stimulation as completely as possible.
2
2
Methods
11
Methods
This section gives an overview of the experimental methods. Exact protocols can
be found in the supplementary materials.
2.1
Cell culture
Depending on the scientific question, a selection of cellular models is available.
As the aim of this thesis was to examine the physiologic and pathophysiologic
reactions of neurons undergoing laser stimulation, a neuronal cell line or primary
neurons were the model of choice. Cell lines have the advantage of being virtually
immortal and can thus be cultivated continuously under standardized conditions.
Also, because these cells can be propagated ad libitum, they ensure a steady
supply for an experimental study, thereby diminishing the need for test animals.
On the other hand, immortalized cell lines often undergo substantial physiological
changes [77]. Any comparison to or prediction for in vivo models derived from a
cell line must therefore be assessed with great caution.
To allow a better translation to complex multicellular models, primary cells should
be used. These cells are isolated directly from tissue of a donor animal. If these
are not stem cells, but fully differentiated, they will usually only survive for a few
weeks to months in culture, depending on the cell type. For further experiments,
new cells need to be harvested from another animal. Firstly, this may limit the
comparability to previous experiments due to natural variations in genetic or physiologic properties. Secondly, although no animal experiment is conducted, there
is the need for a steady supply of donor animals. For these reasons, all the experiments of this thesis were first done in a neuroblastoma cell line. Then, to
examine the applicability of the methods and results to physiologically unaltered
neurons, they were partly verified in primary mouse cortical neurons.
2.1.1 Neuro-2a cells
The murine cell line Neuro-2a (subsequently referred to as N2A) was derived
from a neuroblastoma. The cells grow adherent and display a rounded phenotype
12
2
Methods
in culture. Upon serum deprivation, cells start to form neurites (Figure 2). After
about two days in low-serum or serum-free culture medium, many of the cells
display a more neuronal-like phenotype. After about four days, viability decreases. For the experiments where differentiated cells were used, the cells were
usually kept in serum-free medium for two to three days.
A)
B)
Figure 2: N2A cells in medium + 10 % FCS (A) and after three days of serum deprivation (B).
While the soma of differentiated cells generally remains rounded, they form neurites two to five
times longer than the cell body. Scale bar 20 µm.
For the N2A cell line, functionality of voltage-dependent sodium channels [78, 79]
and ionotropic glutamate receptors [80] has previously been demonstrated. Furthermore, these cells have been used as a model for neurite outgrowth [81, 82]
and neurotoxicity [83, 84]. It can therefore be assumed that the N2A cell line
would serve as a good model for first experimental insights and deductions.
2.1.2 Primary mouse cortical neurons
To study the transferability of the results obtained from N2A cells, murine primary
cortical neurons (MCN) were tested in the same experiments. The neurons were
obtained as a frozen aliquot from Thermo Fisher Scientific. The distributor ensures high viability (at least 50%), high purity (at least 98% cells of neuronal phenotype) and functionality of neurotransmitter response [85].
2
Methods
13
After about one week in culture, the cells display a neuronal phenotype with multibranched neurites and start forming synapses (Figure 3). In the following days,
they form dense clusters that are connected via neurites. While the cells still keep
maturing during the first two to three weeks, very dense clusters are not practical
for evaluating single cell laser stimulation. The experiments were therefore carried out between day 10 and 13 after thawing.
A)
B)
Figure 3: Primary mouse cortical neurons 7 days (A) and 14 days (B) after thawing. A): The cells
quickly develop multiple long branches and form connections. B): After about two weeks, dense
clusters are formed with strong interconnections. At this point, cultures are more mature, but single cell manipulation becomes difficult. Singular bright spots in both pictures are dead cells. Scale
bar 10 µm.
2.2
Immunofluorescence
A standard immunofluorescence protocol was used to examine the neuronal phenotype and effect of differentiation on N2A cells. Three antigens were chosen as
markers: Neurofilament, the intermediate filaments in neurons [86]; synaptophysin, which is incorporated into the membrane of synaptic vesicles [87]; and the
NMDA receptor, an ionotropic glutamate receptor [60]. Together, these markers
convey a good estimate of the neuronal physiology of the cellular model. N2A
cells cultured in serum-deprived medium were compared to undifferentiated N2A
cells and to a fibroblast cell line (3T3).
14
2.3
2
Methods
Laser stimulation
2.3.1 Experimental setup
The setup used for these experiments was constructed for a recent study [88]
and allows for simultaneous laser stimulation and recording of a fluorescence
signal. It is shown in Figure 4 (adapted from [88]). The sample is placed on a
commercial inverted epifluorescence microscope. A weakly focused laser is led
onto the sample vertically from above. Illumination by the mercury vapor lamp is
realized through the microscope objective. An LED provides the light source for
bright-field images.
Figure 4: Single cell manipulation setup. The laser beam (A) is weakly focused on the specimen.
A mercury-vapor lamp (B) is used for fluorescence imaging, and an LED (not shown) provides
non-coherent brightfield illumination of the sample. The manipulation laser is attenuated using a
notch filter, and the fluorescent excitation light is blocked with an emission filter. Images are acquired with a microscope camera (C).
2.3.2 Calcium imaging
To visualise intracellular calcium, cells were stained with the calcium-sensitive
dye Fluo 4-AM. The lipohilic Fluo 4-acetoxymethyl ester enters the cell where it
is cleaved by cellular esterases, thereby rendered hydrophilic and trapped inside
2
Methods
15
the cell [89]. Upon calcium binding, the fluorescence of Fluo 4 increases over
100-fold [90] (Figure 5, adapted from [91]).
Its dissociation constant (Kd) of 345 [90] makes this an intermediate-affinity dye.
The Kd describes the calcium concentration at which half of the molecules have
bound calcium. The higher the Kd, the lower is the affinity. Indicators with low
affinity to calcium are useful for detecting fast changes in concentration of the ion,
as the dye-calcium-complexes are not very long-lasting. The calcium binding
sites are freed quickly and the fluorescence signal is not artificially prolonged by
the indicator’s properties. On the other hand, a low affinity to calcium reduces the
probability that a calcium ion is bound at all. This is an important consideration
especially when measuring small changes in calcium concentration.
F
F
F
F
Figure 5: Transition of Fluo 4 from non-fluorescent (left) to the calcium-bound, fluorescent form
(right). The molecule undergoes conformational changes upon forming a complex with a calcium ion, leading to emission of green fluorescence.
If the dye’s affinity is too low, it may not pick up on small fluctuations. To make
the right decision regarding this trade-off, the expected experimental outcome
needs to be considered. In the present case, the possibility to observe action
potentials as well as unspecific calcium responses was desired. The calcium signal corresponding action potentials have been observed using dyes with both
lower [62] and higher [37, 61] calcium affinities as well as Fluo 4-AM itself [92].
Also, unspecific calcium signals have previously been observed with Fluo 4 [88].
Hence, this calcium indicator presented a sensible choice for the experiments in
this thesis. Other favorable qualities are its large dynamic range and signal to
noise ratio [90].
16
2
Methods
2.3.3 Stimulation experiments
Although promising results for nanoparticle mediated laser stimulation have been
obtained, this was achieved using ultrashort pulsed lasers and/or very tightly focused laser spots [31, 37]. Such an approach would be very difficult to realize in
an implant. In this setup, a picosecond pulsed laser (pulse duration 850 ps, repetition rate 20.25 kHz, λ = 532 nm) is loosely focused on the cell of interest. The
spot diameter is about 60 µm on average. Since the energy density is not uniform
in the whole area of the laser spot, not all cells illuminated by it are of interest for
the experiment. Usually only one, but sometimes two cells were placed in the
center of the spot and their fluorescence signal was evaluated. At a rate of five to
seven frames per second, the baseline signal was recorded for a couple of seconds before the laser pulse. Measurements were taken during the pulse and 30
seconds to two minutes afterwards, depending on the experiment. Fluorescence
data was evaluated for cells in the center of focus as well as cells not illuminated
by the laser. Cells that were located on the edge of the laser spot were not considered due to the unknown energy density in those areas.
Cellular response
For a first assessment, five consecutive laser pulses were applied to both differentiated and undifferentiated N2A cells in culture medium with or without AuNP,
and images were acquired for one minute after each pulse. The radiant exposure
was modulated between 17 and 51 mJ/cm2. Different radiant exposures were
also used with MCNs, but only one laser pulse was applied. MCNs were kept in
PBS with Ca2+ and Mg2+ for stimulation experiments. The sample size per iteration was generally ten cells. One experiment per radiant exposure was conducted.
Inhibitor studies
As described in the theoretical background, the main sources of a physiological
increase of cytoplasmic calcium concentration are the ER, the mitochondria and
channels in the cell membrane. To investigate whether the calcium transients
observed in the stimulation experiments originate from one of these sources, different inhibitors were employed. Table S 4 gives an overview of the compounds
and their respective mode of action. 2-Aminoethoxydiphenyl borate (2-APB) is
2
Methods
17
known to block the inositol triphosphate receptor and thereby inhibit calcium release from the ER [93]. It has also been suggested that it prevents calcium entry
through store-operated channels and/or TRP channels [94]. To inhibit calcium
release from mitochondria, CGP37157 (hereafter referred to as CGP) was used.
This compound inhibits the mitochondrial Na+/Ca2+ exchanger which transports
calcium ions from the mitochondrial matrix to the cytoplasm [95]. In order to exclude sodium currents as the underlying cause for calcium entry, lidocaine was
used to block voltage-gated sodium channels. It has been previously shown that
this anesthetic is able to completely block sodium currents in N2A cells [96]. To
eliminate all paths of extracellular calcium entry, stimulation experiments were
conducted with cells kept in PBS without Ca2+ or Mg2+. Any calcium transients
observed under these conditions were sure to originate from intracellular stores.
Cells kept in PBS with Ca2+ and Mg2+ (PBSC) served as a control. Experiments
applying each of the conditions individually (cells in PBS, PBSC or culture medium, individual inhibitors in culture medium) were conducted three times each at
a constant radiant exposure. The sample size per iteration was ten cells. Inhibitors were applied in several combinations to study possible interactions in a single
run with a sample size of 15-20 cells.
Membrane perforation
Viability and cell perforation were assessed with propidium iodide (PI) as shown
before [88, 97]. After recording the Fluo 4 signal, the PI signal and a bright-field
image were taken. Thereby, immediate damage to the cell like membrane perforation or swelling could be detected. However, it cannot be deduced from the PI
signal whether membrane perforation is transient or permanent. For an estimate
of the kinetics of inflow and a comparison with previous work [98], a timelapse of
PI fluorescence was taken in the same way as described above for Fluo 4. In
those cases, the Fluo 4 signal was taken before and afterwards.
2.3.4 Image analysis
All images were analyzed with Fiji (ImageJ version 2.35) [99, 100]. The mean
gray value of a cell was calculated for each time point, normalized to the respective baseline value (mean value before laser stimulus) and presented as the relative rise of fluorescence values (% ΔF/F). Changes in fluorescence that were
18
2
Methods
not induced by a laser stimulus were calculated based on the preceding minimum. The time it took a cell to rise to the peak fluorescence value was calculated
from the laser pulse for laser-induced transients, and from the start of image acquisition for other transients (see Figure 6 for clarification).
ΔF/F
peak fluorescence (% ΔF/F)
time to
peak
time (s)
Figure 6: Depiction of the evaluated variables. Change of fluorescence is plotted against time.
The peak value for ΔF/F in % (vertical red arrow) is calculated for each cell as well as the time
that passes between the stimulus and reaching that value (time to peak, horizontal red arrow).
2.4
Lipid peroxidation
ROS formation can indirectly be detected by tracing their effects on cellular components. Polyunsaturated (membrane-) lipids are prone to a chemical event
called lipid peroxidation. During this process, free radicals react with carbon-hydrogen bonds, leaving behind lipid radicals. In uncontrolled chain reactions, these
lipid radicals propagate and rearrange the double bonds in unsaturated lipids.
This can lead to the formation of different forms of oxidized lipids, but also to a
breakdown of C-C bonds and destruction of membrane lipids [101].
By introducing a fluorophore sensitive to oxidation, it is possible to visualize this
process. BODIPY 581/591 C11 (hereafter referred to as BODIPY) is a lipid pe-
2
Methods
19
roxidation sensor that is made up by a hydrocarbon chain and an aromatic headgroup. The long hydrocarbon chain acts like an anchor that fixes the molecule in
the cell membrane. In its intact form, the excitation and emission wavelengths are
581 nm and 591 nm, respectively. Upon oxidation, the most favorable location for
radicals to attack the molecule is the short unsaturated hydrocarbon chain between the aromatic systems. Oxidation will then lead to elimination of the phenyl
residue (see Figure 7). As a consequence, the system of delocalized pi electrons
will be smaller and fluorescence will undergo a blue shift. It is now excited by
488 nm and emits light at 510 nm. This shift allows for ratiometric measurements
– upon oxidation, green fluorescence will increase, and red fluorescence will decrease.
O
HO
O
F
HO
N
F
B
N
oxidation
O
B
N
OH
F
F
N
Figure 7: Oxidation of BODIPY 581/591. The unsaturated C4-chain between the aromatic rings
introduces a predetermined point for radicals to attack the molecule. Radical reactions are difficult
to control, but the molecule’s architecture directs reactions with reactive oxygen species to the
result depicted above.
Ratiometric measurements were not possible in the experimental setup available
for this thesis and photobleaching by the laser would make assessment of a decrease in fluorescence of the red emitting form very difficult. Therefore, only the
increase of the green emitting form was studied on a short time scale with the
20
2
Methods
single cell manipulation setup. Cells were treated as described for stimulation
experiments but stained with BODIPY instead of Fluo 4-AM, and the green fluorescence signal was taken for two minutes after the laser pulse. This experiment
was conducted with N2A cells in culture medium, PBS and PBSC, and MCNs in
PBSC.
2.5
Statistical analysis and data visualization
All significance testing was performed using packages and built-in methods of R
(v. 3.1.0) [102]–[104]. For experiments that were done three times total, population means were estimated from sample means and compared with one-way
ANOVA followed by Tukey’s test as post-hoc method. Sample means from single
experiments (e.g. certain combinations of inhibitors) were compared to the respective populations means (i.e. population means for respective experiments
using the individual inhibitors) using either a two-tailed Student’s t tests (for homogeneous variances) or a two-tailed Welch’s t test (for unequal variances), unless stated otherwise. To adjust p-values for multiple comparisons, the FDR (false
discovery rate) method was used. The significance level’s value was set at 5 %.
Variables that are presented by a small sample size or that do not satisfy the
assumptions (normality for t tests, equality of variances for rank-sum tests) are
presented descriptively. Significance levels are presented as follows: * p < 0.05,
** p < 0.01, *** p < 0.001.
Analysis of proportions was not subject to further significant testing due to small
sample sizes and thereby failure to fulfill the test’s assumptions. Proportions are
calculated from the pooled data for each condition as described in the respective
sections. Therefore, no error bars are available for bar plots.
Data visualization was realized with Tableau Desktop (v. 9.3). Correlation analysis was also performed using Tableau. In box plots, the bottom part of the box
(dark grey) represents the second quartile, the top part (light grey) represents the
third. The dividing line between the two parts is the median. Whiskers are drawn
to include all data points within 1.5 interquartile ranges from the bottom and top
of the box. In some depictions, horizontal lines are added to indicate the arithmetic mean.
3
Results
21
3
Results
3.1
Immunofluorescence
The first step to assessing the neural response to laser stimulation was to characterize the physiology of the model cell line. Three neuronal antigens were chosen – the NMDA receptor (NMDA-R), neurofilament (NF) and synaptophysin
(SYP). The distribution of these proteins allows for conclusions regarding the
cells’ physiology. The fibroblast control cell line was negative for all three antigens
(Figure 8).
NF + SYP
B)
C)
D)
N2A
(undiff.)
3T3
A)
SYP + NMDAR
Figure 8: Control cells. Nuclei are stained blue with DAPI. A): NF (red) and SYP (green) in 3T3
cells. Both antigens exhibit a blurry signal. B): SYP (green) and NMDA-R (red) in 3T3 cells. The
SYP antibody binds non-specifically to targets spread across the cell. No signal is obtained from
the NDMA-R antibody. C): NF (green) and SYP (red) in undifferentiated N2A cells. The neurofilament antigen is displayed in a filamentous form in the soma of some cells. The SYP antibody
gives a blurry signal with dot-shaped highlights distributed evenly in the cells. D): SYP (green)
and NMDA (red). The SYP signal is comparable to C), but brighter regions in the vicinity of the
nuclei become apparent. Only a very weak NMDA-R signal is observable (arrow). Scale bar
10 µm.
22
3
Results
Undifferentiated N2A cells express NF in the soma. Additionally, SYP displays a
widespread diffuse signal as well as bright spots near the nucleus. Distinct expression of NMDA receptors was not detectable.
After differentiation in serum deprived medium, a distinct NF signal is observable
throughout the newly formed branches (Figure 9). Expression in the cell soma is
unaffected. Similarly, SYP is concentrated near the nucleus in the same way as
in undifferentiated cells.
A)
C)
B)
D)
Figure 9: NF (green) and SYP (red) in differentiated N2A cells. A): SYP staining yields a signal
from the cell body comparable to undifferentiated cells, but is also observable in bulged areas of
the neurites. B): NF staining is observable throughout the soma and along the neurite. C, D): SYP
signal is obtained from the distal ending of neurites, where it encloses the distal NF elements
(white arrows). Filamentous expression of NF in the cell body is also observed (black arrows).
Scale bar 10 µm.
However, a distinct SYP signal can also be observed in the neurites, particularly
in bulges at the distal ending. Co-staining with NF reveals that in this zone, SYP
3
Results
23
encloses the ending of neuronal cytoskeleton. This is consistent with the physiological occurrence of synapses in neurons [86].
Like SYP, the NMDA receptor appears to be specifically located at the bulged
areas of neurites. Co-staining of the two antigens shows their concurrent occurrence in these segments (Figure 10). Distinct circular signals from NMDA-R staining are obtained from the somatic membrane of some cells.
B)
A)
Figure 10: SYP (green) and NMDA-R (red) in differentiated N2A cells. The SYP antibody exhibits
distinct signals from bulged areas of the neurites. From these areas, dot-shaped signals are also
obtained from the NMDA-R antibody (white arrows). The NMDA receptor is detectable in the cell
membrane of the soma as well (black arrows). Scale bar 10 µm.
3.2
Laser stimulation
Laser stimulation evokes various cellular reactions under different conditions.
The calcium response of the cell models was characterized and is described here
for different radiant exposures and addition of inhibitors.
3.2.1 Laser-induced calcium transients and energy dependency
For a first assessment of the cellular reactions to the laser stimulus, differentiated
and undifferentiated N2A cells as well as mouse cortical neurons (MCNs) were
exposed to a 40 ms laser pulse at different radiant exposures. Control cells were
treated in the same way but were not incubated with AuNP before the experiment.
Control cells never showed a calcium transient that was induced by the laser
24
3
Results
pulse. However, N2A cells often displayed spontaneous transients that had no
obvious connection to the laser stimulus. These were observed in irradiated cells,
in cells that were not located within the laser spot as well as in experiments were
only the fluorescence signal was taken and laser stimulation did not occur at all.
Table 1 gives an overview of the events observed after laser stimulation with different radiant exposures for all cell types. A typical course of a laser-induced calcium transient is shown in Figure 11.
Figure 11: Typical fluorescence signal of an N2A cell in response to laser stimulation. Arrows
indicate application of the laser pulses. Usually, a relatively steep rise of fluorescence values was
observed which peaked within the first ten seconds, followed by a slow return to baseline values.
The cell in this example did not respond to the second laser pulse.
After a certain threshold is reached, the probability of a cell being activated does
not seem to depend on the radiant exposure. A high proportion of each cell type
is activated by the laser stimulus. The proportion of cells responding to more than
one laser pulse is considerably smaller and independent of radiant exposure.
Spontaneous transients were only observed in N2A cells. Undifferentiated N2A
cells appear to exhibit the same calcium response as differentiated cells.
3
Results
25
Table 1: Outcome of laser stimulation of different cell types at various radiant exposures; l-i = laser-induced, n.d. = not determined, diff = differentiated, undiff = undifferentiated. One experiment
was performed per radiant exposure with n = 10 cells. The percentage of cells exhibiting at least
one calcium response evoked by the laser stimuli is given. At 25 mJ/cm2 and higher, this proportion does not depend on the radiant exposure. Undifferentiated N2A cells appear to respond to
multiple stimuli more often than differentiated cells (more than one l-i response). Spontaneous
transients could be observed in N2A cells that were irradiated but failed to show activation by the
laser (spontaneous only) or cells that exhibited a laser-induced response first (spontaneous after
l-i response). However, such transients were never observed in MCNs.
Cell
type
N2A
(diff.)
N2A
(undiff.)
MCN
Radiant exposure
(mJ/cm2)
l-i response
(%)
More than
one l-i response (%)
Spontaneous
only (%)
Spontaneous
after l-i response (%)
17
44
11
11
0
25
100
20
0
20
34
90
20
0
0
42
90
10
10
10
51
100
20
0
30
17
90
0
10
20
25
100
60
10
0
34
89
20
11
0
42
80
20
0
0
51
100
80
0
0
17
0
n.d.
0
0
25
87.5
0
0
0
34
60
n.d.
0
0
42
82
n.d.
0
0
51
61.5
n.d.
0
0
The cells usually displayed a rise to the peak fluorescence value within the first
ten seconds with mean values of approximately five seconds. Fluorescence increases by about 150 to 400 % in N2A cells and by 20 to 60 % in MCNs. Obvious
outliers in both variables (ΔF/F and time to peak) were excluded by hand from all
inferential analysis, but not from analysis of proportions. For proportions, the
26
3
Results
shape of the transient was less important than the information whether or not an
event occurred at all. Outliers are dealt with in depth in section 3.2.3.
It is reasonable to assume that the observed change in fluorescence as well as
the time that passes from a transient’s onset to its peak value depends on the
amount of calcium indicator present in the cell. If a cell appears very dark after
staining, it might either have a low resting concentration of intracellular calcium
(c(Ca2+)i), or it might not have taken up enough calcium indicator. At too low intracellular concentrations, the dye would saturate too early to be comparable to
other cells. If a cell’s baseline fluorescence value represents the staining quality
rather than the intracellular calcium level, a positive correlation between that
value and normalized peak fluorescence is expected as well as a negative correlation with time to peak. The respective trends were calculated for both cell types
based on the initial experiments described in this section and also for the pooled
data set of all stimulation experiments (485 N2A cells, 170 MCNs) (see Figure S
1). The N2A cells show no statistically significant correlation in any of those
cases. The coefficient of determination (R2) is consistently nearly zero. For the
MCNs, there is a statistically, but not practically significant negative correlation
between ΔF/F and baseline fluorescence in the initial experiments (p = 0.01, R2
= 18 %), which disappears in the pooled data set (p = 0.63, R2 = 0.3 %). The
opposite is true for time to peak values of MCNs: In the pooled data set, a statistically significant (p= 0.004) negative correlation becomes apparent; however, the
R2 is as low as 8 %. This correlation is therefore considered not to be practically
important. Consequently, it can be assumed that the dye is present at a saturating
concentration in the performed experiments. The influence of the baseline fluorescence value on the variables fluorescence increase and time to peak will be
neglected hereafter. It is noticeable however, that MCNs generally displayed
equal or higher baseline fluorescence values than N2A cells at only 1/5th the dye
concentration, indicating a higher level of c(Ca2+)i.
3
Results
27
A) N2A
B) MCN
radiant exposure (mJ/cm2)
radiant exposure (mJ/cm2)
Figure 12: Boxplots of ΔF/F and time to peak at different radiant exposures for N2A cells (A) and
MCNs (B). Each radiant exposure was applied in one experiment to n ≈ 10 cells. None of the
MCNs tested responded at 17 mJ/cm2. Neither of the variables show a dependence on radiant
exposure in the N2A cell line or the primary neurons.
Comparing normalized increase of fluorescence and time to peak across the different radiant exposures does not reveal a dependence of either of the variables
on radiant exposure (Figure 12). Correlation analysis shows weak correlations
between ΔF/F and radiant exposure (positive) for MCNs, and between time to
peak and radiant exposure (negative) for N2A cells. However, in both cases the
28
3
Results
radiant exposure only accounts for about 15 % of the observed variability. Due to
the rather small sample size per experiment, a real causation cannot be assumed
based on those numbers. Multiple laser stimulation of the same cell did not reliably evoke more than one laser-induced calcium response (see Table 1), and
never a second calcium trace of a similar dimension as the first one. Hence, for
the following stimulation experiments, a single pulse with a radiant exposure of
25 mJ/cm2 was chosen. This stimulus evoked calcium transients with a similar
probability as higher values did, but only causes membrane perforation in a small
proportion of cells (see Figure 19).
3.2.2 Inhibitor studies
After the typical calcium response to a laser stimulus was characterized, different
approaches were taken to determine the underlying mechanisms. N2A Cells
were first tested under the following conditions: Culture medium (med),
PBS + Ca2+ + Mg2+ (PBSC), PBS, medium + 2-APB (med + A), medium + CGP
(med + C), medium + lidocaine (med + L). Radiant exposure was kept constant
at 25 mJ/cm2. Figure 13 and Table S 6 and 4 show the results of the influence of
the different conditions on probability of activation, peak fluorescence increase
and time to peak in N2A cells. While lack of extracellular Ca 2+ has the greatest
impact on the maximum rise in fluorescence, the inhibitors of intracellular calcium
release also appear to decrease ΔF/F. The variability of time to peak values does
not allow for a reliable statement, but CGP possibly slows down the laser-induced
calcium response.
3
Results
29
p=
0.099
**
p = 0.052
***
**
**
Figure 13: ΔF/F and time to peak for N2A cells in different conditions. Each data point marks the
average for one experiment with a sample size of n ≈ 10 cells each. Additional horizontal lines in
boxplots mark the arithmetic mean of the data points (see section 2.5 for clarification). Obvious
influences of the condition are only observed for ΔF/F: Absence of extracellular calcium (PBS) or
inhibition of intracellular calcium release (med + A, med + C) markedly decrease the values for
this variable. Points of reference for the displayed outcomes of significance testing are PBS and
PBSC, respectively.
One-way ANOVA was performed to test for statistically significant differences in
the variables ΔF/F and time to peak for N2A cells with the condition as influential
variable. The extent of the laser-induced calcium response of cells in PBS differs
significantly from that of cells kept in PBSC and medium + lidocaine at a significance level of 5 %, and strongly (p = 0.052) from cells in culture medium. The pvalues of all pairwise comparisons are shown in Table S 6 and 7. In time to peak,
no significant differences between the conditions depicted in Figure 13 can be
observed.
To study possible interactions between the conditions, they were applied in different combinations: medium + AC, medium + ACL, PBS + A, PBS + C, PBS +
L, PBS + AC, PBS + ACL, PBSC + ACL. Table S 6 and 7 and Figure S 2 show
the results of laser stimulation and p-values of pairwise comparisons after adjusting for multiple tests with the FDR method.
Conditions that include PBS without Ca2+ as surrounding medium plus an inhibitor
show either significant differences in ΔF/F or considerably small p-values when
tested against the respective inhibitors in medium. The difference to PBS alone
30
3
Results
is never significant. Contrarily, adding inhibitors to cells in PBSC significantly decreases ΔF/F compared to PBSC alone. This is consistent with the observation
that extracellular Ca2+ has the greatest impact on ΔF/F. Additive or annihilating
effects when adding more than one inhibitor are not observable. Time to peak
generally increases when inhibitors are added to PBS compared to PBS alone.
In PBSC on the other hand, the inhibitors slightly decrease this variable.
The proportion of cells exhibiting a laser-induced response does not seem to differ markedly depending on the condition, with the possible exception of medium + 2-APB + CGP (see Table S 6). However, only one experiment was done
with this combination. Notably, cells in PBS or PBSC display a consistently high
probability of activation.
Mouse cortical neurons kept in culture medium did not exhibit any calcium transients. This condition was therefore excluded from further analysis. One-way
ANOVA could only be performed when PBS as condition was excluded. This
measure ensured homogeneity of variances. Hence, PBS was compared pairwise to each condition using a t test followed by FDR to correct for multiple comparisons. The results are presented in Figure 14 and Table S 8 and 9. As is the
case for N2A cells, absence of extracellular calcium has the greatest impact on
fluorescence rise. ΔF/F in cells kept in PBS is significantly smaller than in all conditions involving an inhibitor. The p-value for PBS vs. PBSC is 0.11; it has to be
kept in mind that in this case, the sample mean from the PBS-cells is compared
to the estimated population mean from cells in PBSC, while in all other cases
sample means from a single experiment were compared to each other. Different
from N2A cells however, the inhibitors of intracellular calcium release do not decrease ΔF/F; by contrast, CGP appears to increase it. Time to peak is significantly
larger in PBSC + 2-APB than PBSC + CGP or lidocaine.
3
Results
p=
0.11
31
**
***
p = 0.077
**
*
*
**
Figure 14: ΔF/F and time to peak for MCNs in different conditions. In the case of PBSC, each
data point represents the average of an individual experiment with a sample size of n ≈ 10 cells
each. For all other conditions, one experiment with n ≈ 15-20 cells was performed. Outliers are
excluded. Notably, the addition of inhibitors does not appear to decrease ΔF/F in this cell type.
On the contrary, CGP increases this variable. Absence of extracellular Ca2+ markedly decreases
ΔF/F. Time to peak differs significantly between PBSC + 2-APB and the other two inhibitors. Point
of reference for the displayed comparisons is the respective leftmost column.
N2A cells not located in the laser focus displayed spontaneous calcium transients
under all conditions. Figure 15 displays the relationship between the condition
and ΔF/F of those transients where the sample size is at least three cells. Due to
different recording times, there is a sampling bias in the ΔF/F values – the values
can be assumed to be representative, but the number of transients is not comparable. Figure 16 (A) shows the respective proportion of cells that produced a
spontaneous calcium transient, independent of whether the peak value was recorded or not, for a more accurate comparison.
Slower calcium transients are also observable in some irradiated cells following
the laser-induced calcium response. Figure 16 (B) displays the concurrence of
both events depending on the condition.
32
3
Results
Figure 15: Peak ΔF/F of spontaneous transients in non-irradiated N2A cells. The sample size per
condition varies, and conditions with n < 3 were excluded. Variability is high within the groups,
and no differences between groups can be clearly attributed to the conditions.
There is a wide distribution of peak values, and the conditions that incorporate a
sufficient sample size do not exhibit any conspicuous differences. The inhibitors
2-APB and CGP possibly reduce the number of transients that occur in cells kept
in culture medium.
Cells in focus that display a laser-induced calcium response are less likely to also
exhibit a spontaneous transient within about 90 s after the laser pulse if they are
kept in culture medium. However, cells in PBS seem to undergo this combination
of calcium transients distinctly more often (Figure 16 B). Two events are considered independent if
𝑃(𝐴|𝐵) = 𝑃(𝐴) ↔ 𝑃(𝐵|𝐴) = 𝑃(𝐵) ↔ 𝑃(𝐴 ∩ 𝐵) = 𝑃(𝐴)𝑃(𝐵)
(1)
i.e. if the probability of A under the condition of B is the same as the probability
of A alone. If the probability of a spontaneous calcium transient (event A) occurring after a laser-induced response (event B) is the same as the probability of a
calcium transient without a preceding laser stimulus, the two observations can be
considered independent of each other. In medium, the probability for a spontaneous transient is 91.7 % on average. If the cell displays a laser-induced calcium
response first, the probability for a spontaneous transient drops to 17 %.
3
Results
33
A)
B)
Figure 16: A) Proportions of non-irradiated cells displaying a spontaneous calcium transient. As
for all proportions, all cells of a condition were pooled to perform calculations. 2-APB and CGP
decrease the probability of the cells exhibiting a spontaneous transient in medium but not in PBS.
Contrarily, lidocaine decreases the proportion of active cells in PBS. No influence of extracellular
calcium per se can be observed. B) Proportions of irradiated cells exhibiting an additional calcium
transient after a laser-induced calcium response. This proportion is consistently low in medium
and high in PBS. In the absence of extracellular calcium, the probability of a spontaneous transient after a laser-induced response is comparable to the probability of a spontaneous transient in
non-irradiated cells. All cells that showed a measurable onset of a transient were included, regardless of whether the peak value was recorded.
The two types of calcium transients do not seem to occur independently in cells
kept in culture medium. When external Ca2+ is removed (PBS), about 86 % of
34
3
Results
cells not stimulated by laser display a spontaneous transient. After a laser stimulus, a spontaneous transient occurs in 77 % of cells that showed a response to
the stimulus first. Absence of extracellular Ca2+ may therefore render the two observed events independent of each other.
Primary neurons undergoing laser stimulation never exhibit spontaneous transients except when kept in PBS without Ca2+ (33 %, 5 out of 13). Non-irradiated
cells showed spontaneous transients in PBSC (1 out of 21), PBS (4 out of 9) and
PBSC + CGP (2 out of 2). Considering this, PBS might generally not confer independence to the different kind of calcium transients, but the condition may be the
cause of the spontaneous transients.
Overall, extracellular calcium is the most important factor for the laser-induced
calcium response (Figure 13, 14) and also affects other aspects of calcium signaling, observable in slower transients (Figure 16). However, the overall signature of the fluorescence signal after laser stimulation remains the same in the
absence of extracellular calcium. The inhibitors do not interact with each other to
alter the cellular response.
3.2.3 Atypical behavior and membrane perforation
To characterize the typical calcium response, outliers from both variables (ΔF/F
and time to peak) have been excluded from analysis in the previous sections. In
the respective cells, other or additional mechanisms might underlie the calcium
response. Membrane perforation represents one possibility for non-physiological
calcium entry. It might affect the calcium response in general and cause distinctly
different calcium signatures. The occurrence of non-average calcium transients
will be analyzed and connected to PI influx in this section.
3
Results
35
B)
A)
10 %
ΔF/F
20 s
20 %
ΔF/F
20 s
C)
D)
Figure 17: A, B) Exemplary signatures of fluorescence signals that reached their peak value later
than the average (A) or not at all during image acquisition (B). Arrows indicate application of the
laser pulse. C) Occurrence of outliers in ΔF/F. Below-average values are only observable in medium if inhibitors are present. In PBS, above-average values are never observed in the presence
of inhibitors. D) Occurrence of outliers in time to peak. Variability is higher for this variable. Inhibitors of intracellular calcium release appear to prevent non-average behavior in medium, but promote it in PBS.
36
3
Results
In ΔF/F, outliers presented values that were considerably higher (type 1) or lower
(type 2) than the average for the respective experiment. In the case of time to
peak, values were either distinctly higher than the average (type 1), or the laser
stimulus induced a rise in c(Ca2+)i that did not peak during image acquisition (type
2). Possibly, a peak would not be reached at all in the latter case, but rather a
plateau value after steady increase of calcium levels. Therefore, the analysis will
distinguish between the two types of outliers. Figure 17 (A, B) displays exemplary
signatures of calcium transients that are non-average in time to peak.
The occurrences of outliers under the different conditions in N2A cells are depicted in Figure 17 (C, D). Noticeably, above-average ΔF/F values only occur in
PBS when no inhibitor is present. On the other hand, cells in medium only show
below-average fluorescence values with inhibitors present. In most conditions including 2-APB and/or PBS, a proportion of cells shows a distinctly low ΔF/F. Particularly high time to peak values seem to occur predominantly in the presence of
inhibitors.
Timelapse measurements of PI influx were taken from N2A cells irradiated with
25 or 51 mJ/cm2. Mainly two different signatures were observed: A quick rise to
a peak value and a slow but steady increase of fluorescence (Figure 18). These
signatures closely resemble the calcium traces shown in Figure 17 (A, B).
A)
B)
5%
ΔF/F
10 %
ΔF/F
20 s
20 s
Figure 18: Signature of PI influx over time. Signals were taken from two N2A cells irradiated with
51 mJ/cm2. Arrows indicate application of the laser pulse. Fluorescence either increases quickly
to reach a peak value within 20 s after the laser pulse (A) or increases steadily over the time of
image acquisition (B).
3
Results
37
If membrane perforation has a general effect on the calcium trace (excluding outliers), grouping the conditions by presence or absence of a PI signal should reveal
two distinguishable distributions in either of the variables. While under some conditions the values seem to differ between perforated and intact cells, there is no
general trend that can be attributed to membrane perforation (see Figure S 3).
The normal course of the laser-induced calcium transient is undisturbed by PIinflux through a perforated membrane.
If (transient) gaps in the cell membrane occur, calcium entry should be accompanied by PI influx. This might alter the typical cellular calcium response. Table 2
lists coincidences of a non-average calcium signature and a detectable PI signal
in N2A cells. Coincidences of outliers in ΔF/F and time to peak were observable
in only four cells: Type 1 ΔF/F with type 1 time to peak in three, and type 2 ΔF/F
with type 1 time to peak in one cell.
38
3
Results
Table 2: Proportions of N2A cells displaying a PI signal and a normal calcium trace (PI only),
outliers in either of the variables without a PI signal (1/2 only), or outliers together with a PI signal
(1/2 + PI). Cells were pooled for each condition, total number of cells is given in parenthesis.
Above-average values for ΔF/F only occur together with a PI signal, while below average values
are more often presented alone. Outliers in time to peak are often accompanied by PI influx in
PBS.
condition
1 only
2 only
1 + PI
2 + PI
1 only
2 only
1 + PI
2 + PI
Outlier time to peak
% PI
only
Outlier ΔF/F
med (33)
9.1
0
0
9.1
0
9.1
6.1
0
0
med + A
(30)
20
0
10
3.3
0
0
0
3.3
0
med + C
(29)
58.6
0
0
10.3
0
3.4
0
0
0
med + L
(33)
15.1
0
6.1
9
3
0
0
3
0
med + AC
(16)
12.5
0
12.5
0
0
6.3
6.3
0
0
med + ACL
(16)
18.8
0
6.3
0
0
12.5
12.5
0
6.3
PBS (31)
58.1
0
0
9.7
6.5
0
0
9.7
0
PBS + A (15)
46.7
0
6.7
0
0
0
6.7
26.7
13.3
PBS + C (20)
70
0
0
0
0
0
0
15
15
PBS + L (16)
25
0
12.5
12.5
0
0
0
6.3
0
PBS + AC
(16)
68.8
0
6.3
0
0
0
0
18.8
6.3
PBS + ACL
(16)
56.3
0
0
0
0
0
12.5
6.3
0
PBSC (32)
68.8
0
0
0
0
0
0
0
0
PBSC + ACL
(21)
23.8
0
0
0
0
0
42.9
0
9.5
It is noticeable that cells undergoing a stronger rise in fluorescence always display a PI signal as well. On the other hand, only a small proportion of cells with a
3
Results
39
below-average ΔF/F is positive for PI. In the variable time to peak, the events
outlier (of any type) and PI influx may occur both independently and concurrently.
There is no obvious connection between a slow rise of calcium levels and membrane perforation.
Outliers occur considerably less often in MCNs. Above-average ΔF/F is observed
in one cell in PBSC + 2-APB and one cell in PBS (15 cells total for each), both of
which do not display a detectable PI signal, and a third cell in PBSC (41 cells
total), where PI influx was not tested. Also, no coincidence of a non-average time
to peak and PI influx was observed. In three cells, a below-average ΔF/F is displayed (once each in PBSC, PBSC + 2-APB and PBSC + CGP), only one of which
(PBSC + CGP) is also positive for PI.
The proportion of N2A cells experiencing membrane perforation depends on the
radiant exposure and condition (Figure 19 and 20). The proportion of cells exhibiting a detectable PI signal after laser exposure increases with radiant exposure
for values greater than 25 mJ/cm2. Only one out of 59 control cells (without AuNP)
was positive for PI staining after irradiation with 51 mJ/cm2.
Figure 19: Percentage of N2A cells in culture medium undergoing membrane damage after irradiation with different radiant exposures. The proportion of damaged cells seems to increase with
radiant exposure for values greater than 25 mJ/cm2. All cells treated with the respective radiant
exposure were pooled to calculate the percentages.
40
3
Results
Cells kept in PBS are more likely to undergo membrane damage than cells in
medium (Figure 20). Conspicuous exceptions are the conditions medium + CGP
(high proportion) and PBS + lidocaine (low proportion).
Due to the smaller sample size, a similar assessment of membrane perforation
in MCNs might be less informative. However, while two out of ten tested cells in
PBSC were positive for PI (radiant exposure 25 mJ/cm2), there was never any PI
influx into cells in PBS. Furthermore, one cell out of 15 in PBSC + 2-APB and 3
out of 14 in PBSC + CGP displayed a PI signal after laser exposure. MCNs exposed to more than 25 mJ/cm2 did not show signs of perforation.
Figure 20: Proportion of PI-positive cells after laser irradiation under the different conditions. With
two exceptions (med + C, PBS + L), cells in PBS are perforated with a higher probability than
cells in medium. In general, most cells that were positive for PI staining displayed a normal
calcium transient (PI only, dark segments) rather than a non-average signature (PI + outlier, light
segments).
In conclusion, membrane perforation does not generally affect the laser-induced
calcium response (Figure S 3). However, exceptionally strong fluorescence signals are always accompanied by PI influx. No such trend is obvious for a slow
rise of calcium levels. Membrane damage is promoted by high radiant exposures
and absence of extracellular calcium in N2A. MCNs are more resistant against
perforation.
3
3.3
Results
41
Lipid peroxidation
Calcium signaling and ROS formation are closely connected [74]. Since energy
from light or heat can initiate oxidation processes and the energy is transferred
directly to the cell membrane, lipid peroxidation is likely to occur. Lipids may also
be oxidized by ROS formed within the cell. The lipid peroxidation sensor BODIPY
581/591 was used to monitor oxidation processes that may later be connected to
calcium signaling.
Lipid oxidation on short time scales was examined under the following conditions
in N2A cells: medium at 25 mJ/cm2, medium at 51 mJ/cm2, PBS, PBSC (both
51 mJ/cm2). MCNs were tested at 51 mJ/cm2 in PBSC. To adjust for photo-oxidation of the dye, the experiments were also done with control cells (no AuNP).
Figure 21 shows ΔF/F (%) and time to peak for the green emitting form of the
peroxide sensor. The values were compared to their respective controls by onetailed t tests (see Table S 10 for adjusted p-values). Both ΔF/F and time to peak
were significantly larger in N2A cells with AuNP under all conditions, except for
time to peak in medium at 51 mJ/cm2. The presence of AuNP therefore causes
the dye to oxidize both stronger and slower than the laser by itself. Values for
MCNs are not significant, but show the same trend as those observed in N2A
cells. Generally, the values from cells with AuNP are distributed more widely, but
there are no distinct differences between groups that can be attributed to the respective condition.
42
3
Results
A)
**
**
***
p = 0.13
**
B)
**
**
p = 0.2
***
p = 0.47
Figure 21: Lipid peroxidation in N2A cells and MCNs under different conditions. Significance levels and p-values of pairwise comparisons between control (without AuNP) and test cells (with
AuNP) are displayed. A) ΔF/F values for control and test cells. For N2A cells, peak fluorescence
is significantly higher in test cells under all conditions. For MCNs, the difference is observable,
but not statistically significant. B) Time to peak values. Except for 51 mJ/cm2, values are significantly higher for N2A cells with AuNP. For MCNs, no clear difference is observable.
3
Results
43
Similar to what was observed from calcium imaging, 71 % (20 out of 28) of N2A
cells that were not irradiated by the laser pulse demonstrated a spontaneous rise
in fluorescence of BODIPY (green). This was observed under all three conditions
(medium, PBS, PBSC) at a similar extent (Figure S 5).
A fast lipid peroxidation process takes place in cells bearing AuNP. In these cells,
the oxidation is stronger and slower than the initial photo-oxidation process observed in control cells (Figure 21).
44
4
4
Discussion
4.1
The N2A cell line presents a suitable model
Discussion
The model cell line was chosen to test neuronal behavior in a readily available
system. Transmembrane sodium currents have been detected before [96, 101]
as well as a fast rise of the membrane potential in response to mechanical stimuli
[106]. The functionality of ionotropic glutamate receptors has been demonstrated
as well [80]. Nevertheless, distinct evidence of N2A cells firing action potentials
cannot be found in the more recent literature. This is not surprising, considering
that the clone was isolated in 1969 [107]. It is however interesting to note that in
1971, action potentials were recorded from different clones isolated from the
same tumor line [108]. Even in those cells, that had only been kept in vitro as a
clonal cell line for some months, electrophysiology differed from that of primary
cells. It is likely that without physiological stimuli, and given their unstable genotype [109], the clone has lost some of its electrophysiological properties; mainly
(if it ever possessed) the ability to fire action potentials.
The cell line’s other properties still justify its application as a neuronal test system.
Immunofluorescence has revealed that N2A cells express the neuronal antigens
neurofilament, synaptophysin and NMDA receptors. The localization of synaptophysin and NMDA-R in particular suggest that the formation of synapses is possible [86]. Bulged areas in the neurites that were shown to contain higher concentration of both proteins might resemble axonal boutons or dendritic spines,
structures that form synaptic active zones in neurons. Further investigation of the
occurrence of synaptic proteins like PSD-95 or Bassoon [108, 109] should reveal
whether synapses can actually be formed between two cells or if the cells merely
produce synaptic vesicles and receptors for neurotransmitters on their own.
Comparable areas cannot be found in N2A cells cultured in medium containing
10 % FCS. Serum deprivation can therefore be considered a simple and effective
means of inducing both morphological and biochemical changes in the cells.
Primary mouse cortical neurons were used to compare the cell line’s reaction to
laser stimulation to that of a more physiologic system. Neither spontaneous nor
4
Discussion
45
laser-induced spiking was observed in these cells. Since the aim of their use was
to observe neurophysiologic reactions to the same conditions applied to N2A
cells, MCNs were not kept in a bath solution that is usually utilized in electrophysiology. Also, spontaneous spiking is generally observed in primary neurons after
a longer culture period [92]. It is therefore possible that the experimental conditions presented a cause for the observed lack of electric activity. However, as will
be discussed further on, this did not prove to be a disadvantage when comparing
the general reactions to laser exposure of a clonal cell line with primary cells.
4.2
Laser stimulation does not cause electric activation
Cells bearing gold nanoparticles on their membrane reproducibly displayed a
strong rise in fluorescence – and therefore c(Ca2+)i – within about five seconds
after laser stimulation (see Figure 11 and Table 1 in 3.2.1). Neither the extent of
increase in fluorescence nor the time to peak reveal a definitive dependence on
radiant exposure. Repeated stimulation does not reliably evoke repeated calcium
transients, and never a second one of a magnitude comparable to the first.
Hence, most cells (excluding outliers) appear to exhibit an all-or-nothing kind of
response to the laser stimulus in combination with nanoparticle mediators. Kalies
et al. [88] observed an energy-dependence in ΔF/F in a different cell type that
displayed an otherwise comparable calcium response in the same experimental
setup. This deviation from the present observations might be attributed to differences in the cells’ physiology or a small sample size in either of the experiments.
However, cells in the previous work did not show a conclusive trend in terms of
correlation between peak values and radiant exposure. In the present work, there
are apparent variations in ΔF/F values as well (see Figure 12), and even a weak
correlation with radiant exposure in the case of MCN cells. Nevertheless, since
peak values scatter widely and no overall trend based on modulations of radiant
exposure is obvious, the main part of variation in ΔF/F should be attributed to
natural variability of the cells due to lack of further knowledge of the biological
system.
Additionally, Kalies et al. observed calcium oscillations in cells located in the vicinity of the stimulated cell. Calcium transients have been observed in non-irradiated N2A cells in the present work. It has to be noted though that these transients
46
4
Discussion
occurred independently of whether a laser stimulus was applied. Also, there is no
apparent correlation between distance from a spontaneously active cell to the
laser focus and time of onset of the transient. Likewise, cells in the vicinity of the
laser spot exhibit spontaneous transients in an independent timely manner. Altogether, these observations provide no prove for intercellular signaling induced by
laser stimulation.
Overall, the rise of c(Ca2+)i in N2A cells is too slow and likely too strong to be
caused by an action potential, and at the same time too uniform to be caused by
a burst [61, 62, 92]. The signature strongly hints at an unspecific stress signal
[53, 54, 88, 110]. The laser stimulus heats up the gold nanoparticles, which in
turn transfer the heat to the cell membrane [19]. Previously, a fast thermal transient was connected to an increase in membrane capacity and thereby induction
of transmembrane currents [28]. However, excessive heat may also damage the
cell membrane, e.g. by denaturing membrane proteins, oxidizing lipids and/or formation of ROS, which in turn may harm cellular proteins, lipids and nucleic acids.
As described in the introduction, a rise of the intracellular calcium level can originate from multiple sources in the case of cellular stress, many of which interact
with or depend on each other. In the observed cases, no modulation in response
to the strength of the stimulus occurs. Although it cannot be ruled out that irradiation induced transmembrane currents, a depolarization strong enough to activate voltage-gated calcium channels (VDCCs) did not likely occur. In that case,
calcium would enter the cell from the surrounding medium. That event should be
observable in response to repeated stimuli and also be energy-dependent: The
more the membrane heats up, the greater the change in capacity and the resulting depolarization, and the more VDCCs should be activated. While it is possible
that the nature of N2A cells does not allow for an activation of VDCCs (because
of non-physiological function or small numbers of channels) or the overall cellular
response would mask their activity, the observed reactions of N2A cells are most
likely not to be ascribed to electric activity. Depletion of intracellular stores and/or
other (slower) modes of calcium entry are more likely at this point.
In MCNs, a visible reaction to laser irradiation is only observed in PBS and PBSC,
but never in culture medium. The Neurobasal® Medium was specifically designed
for primary neuronal cell culture. Possibly, the components protect the neurons
4
Discussion
47
against stress. The ΔF/F values reached by MCNs are generally smaller than in
N2A cells, but the shape of the calcium signature is comparable. Also, a higher
resting level of intracellular calcium might contribute to a less vigorous increase.
Similar mechanisms might hence be the cause of the increase in intracellular
calcium levels in both cell types. Repeated stimulation was only tested exemplarily on two cells, none of which showed a second calcium response. Considerations for the involvement of VDCCs are the same as for N2A. Additionally, as
discussed above, it is possible that more mature MCN cultures or neurons in a
more defined bath medium might react differently to laser stimuli. Regarding the
transferability of insights into the cellular stress response evoked by gold nanoparticle mediated laser irradiation from N2A cells, the primary neurons seem to
pose a suitable test system.
In neither of the cell models did the variables time to peak and ΔF/F markedly
correlate. A greater rise of the intracellular calcium level does therefore not depend on a longer inflow. This points at the involvement of more channels or pores
for calcium entry rather than longer opening times. Increase of intracellular calcium levels appears to be stopped and counteracted within a cell specific timely
manner.
4.3
The calcium response has different extra- and
intracellular sources
The experimental conditions were modified by introducing different channel
blockers and/or removing extracellular calcium. The most important takeaway is
twofold: 1) Absence of extracellular calcium evokes the greatest change in the
calcium trace, and 2) extracellular calcium is not a precondition for a calcium response. The influence of intracellular calcium is most apparent when comparing
ΔF/F of cells kept in PBS vs. cells in PBSC. Peak values are about 13-fold higher
on average for N2A cells in PBSC (p < 0.001) and about sevenfold higher in medium (p = 0.052). In MCNs, values for cells in PBSC increase about 13-fold as
well; however, this difference is not significant (p = 0.11). It has to be kept in mind
though, that in this case, the experiment with PBS as condition was only done in
one iteration.
48
4
Discussion
In N2A cells, inhibition of intracellular calcium release decreased the peak rise in
fluorescence, albeit not significantly. ΔF/F in the conditions medium + 2-APB and
medium + CGP differ significantly from PBSC, which in turn displays considerably
higher values than cells kept in medium. Cells in medium + lidocaine exhibit a
similar behavior as cells in medium only. Time to peak values vary stronger within
the conditions, and the only groups that might have an effect on this variable are
medium + CGP and PBSC, which also yield the lowest p-value when paired in an
ANOVA (p = 0.26).
The fact that absence of extracellular calcium significantly decreases the peak
rise in fluorescence strongly suggests that in conditions where calcium is present
in the surrounding medium, an inflow occurs. However, since the calcium response does not completely disappear in PBS and the signature is generally unchanged, the observed calcium traces must have multiple origins, at least one
extracellular and one intracellular. It is hardly possible to block every kind of intracellular calcium release due to the many pathways and their interactions [54,
111], non-specifity or side effects of inhibitors [94] or simply lack of adequate
blockers. When combining the two inhibitors 2-APB and CGP in either culture
medium or PBS, no additional effect becomes apparent compared to the respective surrounding medium alone (see Figure S 2). However, the smaller sample
size in the single experiments must be kept in mind. In medium + 2-APB + CGP,
only 6 out of 16 cells showed a calcium response, two of which were excluded as
outliers. Also, possible disturbance of cell physiology by inhibitors and subsequent secondary altering of the calcium response should be considered. Applying
lidocaine to cells in PBS alone or in combination with 2-APB and CGP in PBS or
culture medium did not reveal any further alteration of the calcium response. Notably, when all three inhibitors were applied to cells in PBSC, ΔF/F values
dropped to the level of cells in PBS + inhibitors, and distinctly below values of
medium + inhibitors (see Figure S 2).
Overall, the observations suggest that more than one pathway is involved in the
reaction of N2A cells to a laser stimulus. The heat conferred to the cell membrane
by irradiated AuNP presents a disruption of cellular homeostasis. This stress
stimulus might lead to calcium release from intracellular stores [73, 112]. The time
course of the observed calcium trace fits that of intracellular calcium release as
4
Discussion
49
reported elsewhere [113, 114]. A sudden depletion of these calcium resources
should activate store-operated calcium channels [70]. However, this mechanism
does not present a likely explanation for the observed calcium responses, as they
are activated much more slowly and do not increase c(Ca 2+)i to a comparable
magnitude [56, 69]. Voltage-gated calcium channels, when present in a sufficient
number, can largely increase intracellular calcium levels on a very short timescale
[54, 57]. However, the lack of a measurable effect of lidocaine on the calcium
trace does not hint at a physiological activation of such channels. Moreover, the
N2A cells’ calcium trace appears to rise too slowly to be evoked by channels
involved in fast electric signal transmission. Aside from these considerations,
Leung et al. [79] were unable to detect any calcium current accompanying a
transmembrane sodium current. This leaves transient receptor potential (TRP)
channels as the most likely source of calcium inflow. These channels can be activated on a millisecond timescale [117] by a number of stimuli, such as PLC or
DAG [116, 117], heat [115, 118], membrane potential [65], mechanical stimuli
[65] and even lipid peroxidation [121]. It cannot be distinguished here which stimulus is crucial for the onset of the calcium response, but all cells bearing AuNPs
should experience transient heating. It is also likely that this transient increase in
temperature induces a change in membrane potential [28] and/or leads to lipid
peroxidation. A ligand-mediated activation of TRPC channels is possible as well.
Although TRP channels in N2A cells have scarcely been investigated, Koike et
al. [122] observed morphological changes when culturing the cell line with TRP
modulators. Furthermore, 2-APB has also been described as an unspecific
blocker of TRP channels [94, 120]. Depending on the expression of such channels in N2A cells, mutual activation of different types with different time courses
presents a possibility. Regardless of the exact activation mechanism, it explains
the weak or missing dependence on radiant exposure (because once activated,
the channels always operate in the same way), the missing correlation between
rise of fluorescence and time to peak and the weak impact of inhibitors on time
to peak (because of specific time constants of the channels [123]), the great influence of extracellular calcium, the missing effect of lidocaine and the decrease
of ΔF/F by inhibitors of intracellular calcium release and – in the case of 2-APB –
of TRP channels.
50
4
Discussion
Furthermore, substantial increase of c(Ca2+)i is very likely to result in calcium induced calcium release (CICR) from the ER by positive feedback of Ca2+ on IP3
and ryanodine receptors or stimulation by ROS [54, 74] and/or from mitochondria,
which would mainly act as a calcium sink first, but release calcium when overloaded [76]. If these organelles accumulate too much calcium, the mitochondrial
permeability transition pore may open (see theoretical background). Opening of
this megachannel can occur transiently in a flickering way [76], but permanent
opening will most likely lead to cell death through energy failure or release of
apoptosis-inducing molecules [71, 76]. Since high cytoplasmic calcium levels
were not sustained in stimulated N2A cells, it is not likely that permanent MPT
pore opening occurred in a high number of mitochondria. However, there is a
reason to assume CICR from the present data: Repeated stimulation never leads
to a second rise in fluorescence close to values observed for the initial calcium
response. Also, spontaneous calcium transients only take place after a laser-induced response if the stimulated calcium transient was of a particularly small amplitude. Across the whole data set, the average ΔF/F is 39.5 ± 44 % in cells that
display another calcium transient after a laser-induced response, and 205 ±
165 % in cells that do not (see Figure S 4). Since spontaneous transients are
unaffected in non-irradiated cells in PBS, the calcium source must be intracellular. This supports the assumption that a strong laser-induced calcium response
exhausts intracellular calcium stores. If the initial response is weaker, for example
because less or no extracellular calcium has entered the cell and CICR may not
have occurred, the stores still contain enough Ca2+ to release another ionic flow
within one to two minutes after laser stimulation.
This assumption is also supported by the striking observation that cells in PBS
are much more likely to exhibit a slower calcium transient after a laser-induced
response than cells in culture medium (see Figure 16 B). Because calcium levels
cannot be elevated via TRP channels in cells in PBS, cytoplasmic c(Ca 2+) does
not rise as much and CICR may not be as pronounced. Therefore, intracellular
stores still contain enough calcium to undergo a second, slower kind of release.
In irradiated cells, the second transient might be attributable to the stress caused
by laser. Also, the intracellular stores cannot be compensated for Ca2+ loss in
PBS. This might pose additional stress on the ER and induce the second calcium
transient. Averaged over the whole data set, slower transients following a direct
4
Discussion
51
response in irradiated cells peak markedly sooner than spontaneous transients
in non-irradiated cells (66 ± 30 s vs. 147 ± 82 s, measured from start of image
acquisition). The slower transients in irradiated cells may therefore be part of the
stress response triggered by the laser, the mechanism might differ from that in
non-irradiated cells.
Inhibitors did not have an obvious effect on the spontaneous transients observed
in non-stimulated N2A cells, neither were the transients of neighboring cells synchronized or depended on the distant to the laser spot. Hence, no informed assumptions about the cause or source of these transients can be made. But since
they can be observed in N2A cells even when no stimulus is applied at all, the
laser can be excluded as a reason for this behavior.
Most of the conclusions drawn so far can be transferred to the primary neuron.
The calcium trace, though generally of lower amplitude, largely behaves in the
same way, with one exception: Addition of inhibitors does not have an obvious
effect on the rise of fluorescence. This might be due to sample size, insufficient
uptake by the cells or too low inhibitor concentrations. Possibly, already displaying a higher c(Ca2+) at rest, MCNs were less prone to calcium release from intracellular stores. If a stress signal triggers the observed calcium response, the neurons’ culture medium appears to have a protective effect on the cells, since no
calcium trace whatsoever occurred under this condition. PBSC, a less physiological medium, may allow the laser pulse to pose greater stress on the cell. Still, it
is possible that a widespread stress signal does not occur, thus fewer TRP channels would be activated and the calcium signal would be less pronounced. Possibly, the MCNs express certain kinds of TRP channels only at a very low level
or not at all. Channels activated by heat or a change in membrane potential
should have produced a calcium signal in culture medium as well.
A laser-induced calcium signal still occurs in the absence of extracellular calcium,
albeit of considerably small amplitude (see Figure 14). Also, only MCNs in PBS
ever exhibit a second calcium transient after the laser-induced response (5 out of
15 cells). This observation is consistent with the hypothesis described before: If
the laser pulse poses enough stress on the cell to cause a calcium transient, this
reaction first consists of a relatively fast release of Ca2+ ions from intracellular
stores and, if they are not depleted, a slower transient might follow. This possibly
52
4
Discussion
occurs from the ER as an additional stress response. Since these transients are
only observed in a small fraction of non-irradiated cells (22 %) and consistently
display a very low amplitude (4 ± 3 % ΔF/F on average), it is again reasonable to
assume different causes and/or pathways in stimulated and non-stimulated cells.
Noticeably, half of the non-stimulated MCNs exhibiting a calcium signal was kept
in PBS. A calcium-free medium per se might impose enough stress on the cells
to induce non-physiologic behavior. This distinction could not be made for nonstimulated N2A cells.
4.4
Membrane perforation can cause deviations in the
calcium response, but does not generally affect it
Analyzing atypical calcium transients and their occurrence under specific conditions can provide further insights into the underlying mechanisms. Above-average fluorescence values were only observed in cells that also demonstrated PIinflux. This is not surprising: A perforated membrane presents additional sites for
exchange of ions and molecules with the surrounding medium. The steep electrochemical gradient should facilitate fast Ca2+ inflow through newly formed unspecific pores. It has been shown before that the laser system used for these
experiments can create holes in a cell membrane large enough for efficient influx
of 10 kDa dextran [19], which has a hydrodynamic radius of about 2 nm [124].
This is considerably larger than the radius of hydrated calcium ions in solution
(412 pm [125]) and should also allow for PI influx [97]. If a laser pulse does cause
ruptures in the membrane, more calcium should enter the cell quickly and also
stimulate the CICR pathways, leading to a particularly strong calcium signal. Fluorescence usually does not appear to be constantly elevated after an above-average rise, indicating that holes in the membrane are only transient. PI and calcium signals could not be observed simultaneously, but it is reasonable to assume that the described process is accompanied by a fast PI inflow as depicted
in Figure 18 (A).
If 2-APB blocks calcium release from the ER and/or calcium influx through TRP
channels, it should cause particularly low ΔF/F values. Indeed, most conditions
including 2-APB exhibit a number of type 2 outliers in fluorescence rise. The same
is true for cells in PBS with or without inhibitors.
4
Discussion
53
Unusually long time courses of the calcium response might intuitively be explained easily by damaged membrane. However, the two events do not seem to
depend on each other (see Table 2). Characterization of PI influx demonstrated
that some cells do not experience a sudden rise in fluorescence, but rather a slow
but steady increase. Laser irradiation may therefore not only cause large membrane pores through which molecules can enter before they are closed, but also
sites of leakage. These sights may be too small or unstable for substantial inflow
of molecules, but the steady increase in fluorescence from PI shows that this kind
of damage is not repaired as quickly as the holes discussed before. Calcium
might also enter slowly but steadily through such leakage sites. Such a channelindependent means of entry would particularly fit the type 2 outliers in time to
peak (see Figure 17). These are however not always accompanied by PI influx.
Since Ca2+ is smaller than PI [97], small pores that allow for impaired entry of
Ca2+ should form more often than pores large enough for PI. Another explanation
for the unusually long time to peak values might be variability in the cells’ membrane composition. Different kinds of TRP channels have different kinetics [65].
If different stress signals are generated by laser irradiation, various types of channels might be addressed. If the cells vary in their expression patterns, different
time courses for the calcium response are a logical consequence.
The proportion of perforated cells is markedly higher in PBS than in culture medium. This is not surprising, since the divalent cations Mg2+ and Ca2+, that are
absent in PBS, are known to stabilize membrane proteins [124, 125]. PBSC as
surrounding medium also causes a relatively high proportion of cells to be damaged (see Figure 20). It is likely that this medium lacks other stabilizing factors
for membrane integrity. Two conspicuous exceptions from the overall trend are
the conditions medium + CGP and PBS + lidocaine, which display a markedly
high or low rate of membrane perforation, respectively. These effects did not appear in combination with other inhibitors. It is interesting to note, however, that
lidocaine was previously associated with membrane damage [128]. In the observed case, lidocaine might rather confer membrane protection from laser irradiation.
Primary neurons were not as prone to membrane damage. PI influx was analyzed
in 96 out of 173 cells, only six of which were exhibited a detectable signal (6.3 %).
54
4
Discussion
None of the 15 cells tested in PBS were positive for PI staining. Membrane stability therefore seems to be highly dependent on the cell type. Unlike N2A cells,
two MCNs display an above-average value for ΔF/F without a corresponding PI
signal. This might again be attributed to natural variability of membrane channels
or signaling proteins.
The most important conclusion to draw from the measurements is that membrane
perforation does not appear to affect the laser-induced calcium response in general. While sudden membrane ruptures may increase the strength of the response, holes in the membrane large enough for PI influx are neither the cause
of the calcium transient nor does their presence or absence generally influence
the signature of the fluorescence signal (see Figure S 3). Damaging the cellular
membrane by laser stimulation needs to be a concern especially in long-term
applications. However, the mechanisms underlying the cellular calcium response,
be it physiological or pathophysiological, are intrinsic and only underlie natural
variations in the cells’ biology.
4.5
Laser stimulation causes lipid peroxidation on a
similar timescale
The lipid peroxidation sensor BODIPY 581/591 was used with cells with and without AuNP to correct for photo-oxidation by the laser pulse. Although photobleaching effects have been described as low [129], the red emitting form was
strongly bleached by laser irradiation and could therefore not be used as intended. The green emitting form on the other hand, that increases in fluorescence
when lipid peroxidation occurs, proved to be a useful tool to study this chemical
process in single cells on a short time scale. In all experiments, the dye’s signal
differed significantly in N2A cells with AuNP from cells without particles (see Figure 21). This difference was observed for ΔF/F as well as time to peak (with the
exception of N2A cells in medium at 51 mJ/cm2). It can hence be concluded that
in the presence of AuNP, an oxidation process takes place that is both stronger
and slower than mere photo-oxidation – an observation consistent with lipid peroxidation induced by ROS formation. Reactive oxygen species may be created
more or less directly by the laser pulse, or they could be formed as part of the
cellular stress response, caused by heat shock or calcium overload [53, 128].
4
Discussion
55
Secondary ROS formation however usually happens on another time scale [129,
130]. After the initial peak, reached within about six to twenty seconds after the
laser pulse, no further oxidation of the dye can be observed within the following
90 seconds. It is possible, that at this time point the photobleaching effect by the
HBO lamp exceeded fluorescence increase by oxidation.
Oxidation processes in the cell membrane have thereby been demonstrated on
a time scale of a few seconds in N2A cells, and, although not significantly different
from the control, the same trend is visible in MCNs. A few things are worth noticing. First, the extent of lipid peroxidation does not vary depending on the condition. If absence of extracellular Ca2+ (PBS) does pose additional stress on the
cells, it does not manifest in a detectable ROS increase. Second, the time course
of lipid peroxidation is just a little slower than that of the calcium response. It is
therefore not likely that ROS released by mitochondria due to calcium overload
[34, 67] is the underlying cause. Still, this process is likely to take place when
mitochondria sequester large amounts of Ca2+ from the cytoplasm, leading to a
further increase of ROS levels on a different timescale. The most striking observation is that in non-irradiated N2A cells as well as some irradiated cells in PBS,
a spontaneous, slow and rather weak increase of BODIPY fluorescence is observed that matches the spontaneous calcium transients (see Figure S 5). In nonirradiated cells, the only reasonable external cause for ROS formation is the fluorescence lamp [133]. Calcium and ROS signaling however are closely interrelated [74]. Cause and effect cannot be distinguished based on the available data
on the slow transients. As mentioned before, failure to compensate the ER for
calcium depletion might cause ER stress and ROS release in irradiated cells kept
in PBS [114]. This supports the hypothesis that different causes and/or mechanisms cause the slower calcium transients in irradiated and non-irradiated cells.
Different types of TRP channels are activated by lipid peroxidation [121]. So far,
the expression of TRP channels in N2A cells is only hinted at by a few observations regarding mechanosensitivity [134] and the influence of certain agonists
[122]. However, the observations in the present data set can be thoroughly explained by a combination of intracellular calcium release, calcium entry through
TRP channels and subsequent calcium induced calcium release. Failure to restore physiologic conditions in the ER might be the source of additional stress
56
4
Discussion
signaling. The whole cascade is triggered by laser stimulation, which causes lipid
peroxidation and most likely ROS formation, possibly on different time scales.
Long-term effects of (repeated) laser stimulation on calcium buffering, mitochondrial health, accumulation of ROS and finally cell death by apoptosis or necrosis
have not been studied. But the extent of the cellular reaction at least in the case
of N2A cells strongly suggests that long-term damage will occur if homeostasis
is not restored between the stimuli.
5
5
Conclusion and outlook
57
Conclusion and outlook
The aim of this thesis was to characterize the cellular response to gold nanoparticle mediated laser stimulation in a neuroblastoma cell line and primary cortical
neurons. This was achieved by collecting data on calcium signaling, membrane
perforation and/or ROS formation from a total of about 770 individual cells. Single
cell observations provide detailed information on the magnitude and time course
of the events, and the large data set allows for deductions based on probabilities
and correction for natural variations. By integrating all the information provided
by the experiments, the following hypothesis was constructed:
The laser stimulus induces calcium release from the ER and/or mitochondria and
opening of TRP channels, most likely through the formation of ROS. The elevated
intracellular calcium level further is further increased by calcium-induced calcium
release. Either complete store depletion, channel inactivation or a less favorable
electrochemical gradient prevent another steep increase of c(Ca2+)i in response
to a second stimulus within 90 s after the first one. Failure to replenish ER calcium
resources poses additional stress on the cell. Membrane perforation may occur
depending on radiant exposure and surrounding medium, but does not have a
general effect on the cell’s response. Although the laser pulse presents a serious
stressor, occasional swelling was the only sign of irreversible damage. Long-term
exposure has not been tested but supposedly leads to cell death if the observed
severe disruption of cellular homeostasis cannot be counteracted.
The expression pattern of TRP channels in N2A cells has not yet been described.
The strong calcium response observed in this work suggests a high level of expression, potentially making them a good model cell line to study TRP channel
activation and functionality. A far lower expression level of some or all channel
types that might mediate the cellular response is suggested for primary mouse
cortical neurons, at least within the first two weeks in culture.
Since direct or nanoparticle-mediated stimulation arose as a potential new means
of precisely triggering neural cells, successful efforts have been made to unravel
the underlying mechanism of activation [28, 31, 36, 68, 133]. However, the studies have mostly focused on evoking transmembrane currents with the ambition
58
5
Conclusion and outlook
to transfer the method to in vivo applications. But neglecting how it influences
cellular health makes proper adjustments of the method impossible. The present
work strongly suggests that if the stimulus were to be applied at high frequencies,
cells are likely to be damaged permanently. This would be fatal for a new implant
technology, since functionality at the biological interface is an essential requirement for good performance of any implant interacting with the organism.
Transmitter uncaging or optogenetics might be considered as alternative approaches for optical neuroimplants. For a chemically caged transmitter however
there would be the need of constant supply at the site of action. Also, there might
be adverse effects from high transmitter concentrations – for example, glutamate
can cause excitotoxicity in neurons [136]. Although this method has proved to be
very useful in neuroscience research, a lot of fine-tuning would be necessary before it can be considered for in vivo applications.
Optogenetically altered organs are already being examined in vivo [51, 135, 136].
Using a sophisticated transfection method, stimulation would be possible with
single cell precision but without the need for a tightly focused light spot. Relatively
low excitation thresholds of the channels and lack of any further foreign substances arguably make this method the safest of those mentioned here – given
that transfection does not cause adverse effects in the cell or organism. While the
introduced ion channels themselves do not seem to cause any problems, gene
therapy is only applied to humans in some clinical trials so far, as transfection in
general has not yet proven to be safe and reliable in the long term. Still, optogenetic approaches probably present the most promising idea in this field, with current obstacles likely to be overcome in the future.
This thesis has shed light on the so far neglected overall cellular response to gold
nanoparticle-mediated laser stimulation. Although it implicates that successful in
vivo application of this method is very questionable, at least if the stimuli need to
be applied at high frequencies, the provided hypothesis also opens up new
chances. N2A cells are potentially a good model to study TRP channel functionality. Since these channels are widespread, they present an interesting target for
laser-based methods – moving optical stimulation beyond excitable tissues.
6
6
References
59
References
[1] M. Piccolino, “Luigi Galvani and animal electricity: two centuries after the foundation
of electrophysiology,” Trends Neurosci., vol. 20, no. 10, pp. 443–448, 1997.
[2] K. Mullett, “State of the art in neurostimulation,” Pacing Clin. Electrophysiol., vol. 10,
no. 1, pp. 162–175, 1987.
[3] J. C. Cauthen and E. J. Renner, “Transcutaneous and peripheral nerve stimulation
for chronic pain states.,” Surg. Neurol., vol. 4, no. 1, pp. 102–104, 1975 1975.
[4] Y. Hosobuchi, J. E. Adams, and B. Rutkin, “Chronic thalamic stimulation for the control of facial anesthesia dolorosa,” Arch. Neurol., vol. 29, no. 3, pp. 158–161, 1973.
[5] A. Arvanitaki and N. Chalazonitis, Excitatory and inhibitory processes initiated by
light and infra-red radiations in single identifiable nerve cells (giant ganglion cells of
Aplysia). Pergamon Press, 1961.
[6] J. Booth, A. von Muralt, and R. Stampfli, “The photochemical action of ultra-violet
light on isolated single nerve fibres.,” Helv. Physiol. Pharmacol. Acta, vol. 8, no. 2,
pp. 110–127, 1950.
[7] E. Ben-Menachem, “Neurostimulation-past, present, and beyond,” Epilepsy Curr.,
vol. 12, no. 5, pp. 188–191, 2012.
[8] G. Deuschl, C. Schade-Brittinger, P. Krack, J. Volkmann, H. Schäfer, K. Bötzel, C.
Daniels, A. Deutschländer, U. Dillmann, W. Eisner, and others, “A randomized trial
of deep-brain stimulation for Parkinson’s disease,” N. Engl. J. Med., vol. 355, no. 9,
pp. 896–908, 2006.
[9] R. Rokyta and J. Fricová, “Neurostimulation methods in the treatment of chronic
pain,” Physiol. Res., vol. 61, p. S23, 2012.
[10] S. Berrettini, A. Baggiani, L. Bruschini, E. Cassandro, D. Cuda, R. Filipo, I. Palla, N.
Quaranta, and F. Forli, “Systematic review of the literature on the clinical effectiveness of the cochlear implant procedure in adult patients,” Acta Otorhinolaryngol. Ital.,
vol. 31, no. 1, pp. 299–310, 2011.
[11] C. Wu and A. D. Sharan, “Neurostimulation for the Treatment of Epilepsy: A Review
of Current Surgical Interventions: A Review of Neurostimulation for Epilepsy,” Neuromodulation Technol. Neural Interface, vol. 16, no. 1, pp. 10–24, Jan. 2013.
[12] R. Levy, T. R. Deer, and J. Henderson, “Intracranial neurostimulation for pain control:
a review,” Pain Physician, vol. 13, pp. 157–165, 2010.
[13] J. M. DeSantana, D. M. Walsh, C. Vance, B. A. Rakel, and K. A. Sluka, “Effectiveness of transcutaneous electrical nerve stimulation for treatment of hyperalgesia and
pain,” Curr. Rheumatol. Rep., vol. 10, no. 6, pp. 492–499, 2008.
[14] G. Ehret, “Stiffness gradient along the basilar membrane as a basis for spatial frequency analysis within the cochlea,” J. Acoust. Soc. Am., vol. 64, no. 6, pp. 1723–
1728, 1978.
[15] H. J. McDermott, “Music perception with cochlear implants: a review,” Trends Amplif., vol. 8, no. 2, pp. 49–82, 2004.
[16] N. R. Peterson, D. B. Pisoni, and R. T. Miyamoto, “Cochlear implants and spoken
language processing abilities: Review and assessment of the literature,” Restor.
Neurol. Neurosci., vol. 28, no. 2, pp. 237–250, 2010.
[17] U. K. Tirlapur and K. König, “Cell biology: Targeted transfection by femtosecond laser,” Nature, vol. 418, no. 6895, pp. 290–291, Jul. 2002.
[18] M. Schomaker, J. Baumgart, D. Motekaitis, D. Heinemann, J. Krawinkel, M. Pangalos, W. Bintig, E. Boulais, R. Lachaine, B. St.-Louis Lalonde, A. Ngezahayo, M.
Meunier, and A. Heisterkamp, “Mechanisms of gold nanoparticle mediated ultrashort
laser cell membrane perforation,” 2011, vol. 7925, p. 79250F–79250F–6.
60
6
References
[19] D. Heinemann, M. Schomaker, S. Kalies, M. Schieck, R. Carlson, H. M. Escobar, T.
Ripken, H. Meyer, and A. Heisterkamp, “Gold Nanoparticle Mediated Laser Transfection for Efficient siRNA Mediated Gene Knock Down,” PLoS ONE, vol. 8, no. 3, p.
e58604, Mar. 2013.
[20] P. Chakravarty, W. Qian, M. A. El-Sayed, and M. R. Prausnitz, “Delivery of molecules
into cells using carbon nanoparticles activated by femtosecond laser pulses,” Nat.
Nanotechnol., vol. 5, no. 8, pp. 607–611, Aug. 2010.
[21] X. Huang, P. K. Jain, I. H. El-Sayed, and M. A. El-Sayed, “Gold nanoparticles: interesting optical properties and recent applications in cancer diagnostics and therapy,”
Nanomed., vol. 2, no. 5, pp. 681–693, Oct. 2007.
[22] W. Cai, T. Gao, H. Hong, and J. Sun, “Applications of gold nanoparticles in cancer
nanotechnology,” Nanotechnol. Sci. Appl., vol. 2008, no. 1, Sep. 2008.
[23] J. Lee, D. K. Chatterjee, M. H. Lee, and S. Krishnan, “Gold nanoparticles in breast
cancer treatment: Promise and potential pitfalls,” Cancer Lett., vol. 347, no. 1, pp.
46–53, May 2014.
[24] R. Popovtzer, A. Agrawal, N. A. Kotov, A. Popovtzer, J. Balter, T. E. Carey, and R.
Kopelman, “Targeted Gold Nanoparticles Enable Molecular CT Imaging of Cancer,”
Nano Lett., vol. 8, no. 12, pp. 4593–4596, Dec. 2008.
[25] B. Kang, M. A. Mackey, and M. A. El-Sayed, “Nuclear Targeting of Gold Nanoparticles in Cancer Cells Induces DNA Damage, Causing Cytokinesis Arrest and Apoptosis,” J. Am. Chem. Soc., vol. 132, no. 5, pp. 1517–1519, Feb. 2010.
[26] C.-P. Richter, A. I. Matic, J. D. Wells, E. D. Jansen, and J. T. Walsh, “Neural stimulation with optical radiation,” Laser Photonics Rev., vol. 5, no. 1, pp. 68–80, Jan.
2011.
[27] J. Wells, C. Kao, P. Konrad, T. Milner, J. Kim, A. Mahadevan-Jansen, and E. D.
Jansen, “Biophysical Mechanisms of Transient Optical Stimulation of Peripheral
Nerve,” Biophys. J., vol. 93, no. 7, pp. 2567–2580, Oct. 2007.
[28] M. G. Shapiro, K. Homma, S. Villarreal, C.-P. Richter, and F. Bezanilla, “Infrared light
excites cells by changing their electrical capacitance,” Nat. Commun., vol. 3, p. 736,
Mar. 2012.
[29] E. J. Katz, I. K. Ilev, V. Krauthamer, D. H. Kim, and D. Weinreich, “Excitation of primary afferent neurons by near-infrared light in vitro:,” NeuroReport, vol. 21, no. 9,
pp. 662–666, Jun. 2010.
[30] J. E. Olson, W. Schimmerling, and C. A. Tobias, “Laser action spectrum of reduced
excitability in nerve cells,” Brain Res., vol. 204, pp. 436–440, 1981.
[31] J. L. Carvalho-de-Souza, J. S. Treger, B. Dang, S. B. H. Kent, D. R. Pepperberg,
and F. Bezanilla, “Photosensitivity of Neurons Enabled by Cell-Targeted Gold Nanoparticles,” Neuron, vol. 86, no. 1, pp. 207–217, Apr. 2015.
[32] C. Paviolo, A. C. Thompson, J. Yong, W. G. A. Brown, and P. R. Stoddart, “Nanoparticle-enhanced infrared neural stimulation,” J. Neural Eng., vol. 11, no. 6, p.
65002, Dec. 2014.
[33] S. Link and M. A. El-Sayed, “Size and Temperature Dependence of the Plasmon
Absorption of Colloidal Gold Nanoparticles,” J. Phys. Chem. B, vol. 103, no. 21, pp.
4212–4217, May 1999.
[34] N. Farah, A. Zoubi, S. Matar, L. Golan, A. Marom, C. R. Butson, I. Brosh, and S.
Shoham, “Holographically patterned activation using photo-absorber induced neural–thermal stimulation,” J. Neural Eng., vol. 10, no. 5, p. 56004, Oct. 2013.
[35] S. A. Stanley, J. E. Gagner, S. Damanpour, M. Yoshida, J. S. Dordick, and J. M.
Friedman, “Radio-Wave Heating of Iron Oxide Nanoparticles Can Regulate Plasma
Glucose in Mice,” Science, vol. 336, no. 6081, pp. 604–608, May 2012.
[36] J. Yong, K. Needham, W. G. A. Brown, B. A. Nayagam, S. L. McArthur, A. Yu, and
P. R. Stoddart, “Gold-Nanorod-Assisted Near-Infrared Stimulation of Primary Auditory Neurons,” Adv. Healthc. Mater., vol. 3, no. 11, pp. 1862–1868, Nov. 2014.
[37] F. Lavoie-Cardinal, C. Salesse, É. Bergeron, M. Meunier, and P. De Koninck, “Gold
nanoparticle-assisted all optical localized stimulation and monitoring of Ca2+ signaling in neurons,” Sci. Rep., vol. 6, p. 20619, Feb. 2016.
6
References
61
[38] K. Eom, J. Kim, J. M. Choi, T. Kang, J. W. Chang, K. M. Byun, S. B. Jun, and S. J.
Kim, “Enhanced Infrared Neural Stimulation using Localized Surface Plasmon Resonance of Gold Nanorods,” Small, vol. 10, no. 19, pp. 3853–3857, Oct. 2014.
[39] A. M. Alkilany, P. K. Nagaria, C. R. Hexel, T. J. Shaw, C. J. Murphy, and M. D. Wyatt,
“Cellular Uptake and Cytotoxicity of Gold Nanorods: Molecular Origin of Cytotoxicity
and Surface Effects,” Small, vol. 5, no. 6, pp. 701–708, Mar. 2009.
[40] A. M. Alkilany and C. J. Murphy, “Toxicity and cellular uptake of gold nanoparticles:
what we have learned so far?,” J. Nanoparticle Res., vol. 12, no. 7, pp. 2313–2333,
Sep. 2010.
[41] Y. Pan, S. Neuss, A. Leifert, M. Fischler, F. Wen, U. Simon, G. Schmid, W. Brandau,
and W. Jahnen-Dechent, “Size-Dependent Cytotoxicity of Gold Nanoparticles,”
Small, vol. 3, no. 11, pp. 1941–1949, Nov. 2007.
[42] N. Lewinski, V. Colvin, and R. Drezek, “Cytotoxicity of Nanoparticles,” Small, vol. 4,
no. 1, pp. 26–49, Jan. 2008.
[43] C. Paviolo, J. W. Haycock, P. R. Stoddart, and S. L. McArthur, “Effects of laserexposed gold nanorods on biochemical pathways of neuronal cells,” 2013, p.
89231A.
[44] E. E. Connor, J. Mwamuka, A. Gole, C. J. Murphy, and M. D. Wyatt, “Gold Nanoparticles Are Taken Up by Human Cells but Do Not Cause Acute Cytotoxicity,” Small,
vol. 1, no. 3, pp. 325–327, Mar. 2005.
[45] E. Fino, “RuBi-Glutamate: Two-photon and visible-light photoactivation of neurons
and dendritic spines,” Front. Neural Circuits, vol. 3, 2009.
[46] E. Fino and R. Yuste, “Dense Inhibitory Connectivity in Neocortex,” Neuron, vol. 69,
no. 6, pp. 1188–1203, Mar. 2011.
[47] D. Warther, S. Gug, A. Specht, F. Bolze, J.-F. Nicoud, A. Mourot, and M. Goeldner,
“Two-photon uncaging: New prospects in neuroscience and cellular biology,” Bioorg.
Med. Chem., vol. 18, no. 22, pp. 7753–7758, Nov. 2010.
[48] R. Araya, J. Jiang, K. B. Eisenthal, and R. Yuste, “The spine neck filters membrane
potentials,” Proc. Natl. Acad. Sci., vol. 103, no. 47, pp. 17961–17966, 2006.
[49] V. Parpura and P. G. Haydon, “Physiological astrocytic calcium levels stimulate glutamate release to modulate adjacent neurons,” Proc. Natl. Acad. Sci., vol. 97, no.
15, pp. 8629–8634, 2000.
[50] E. Boyden, “A history of optogenetics: the development of tools for controlling brain
circuits with light,” F1000 Biol. Rep., vol. 3, May 2011.
[51] A. E. Hight, E. D. Kozin, K. Darrow, A. Lehmann, E. Boyden, M. C. Brown, and D. J.
Lee, “Superior temporal resolution of Chronos versus channelrhodopsin-2 in an optogenetic model of the auditory brainstem implant,” Hear. Res., vol. 322, pp. 235–241,
Apr. 2015.
[52] J. Y. Lin, P. M. Knutsen, A. Muller, D. Kleinfeld, and R. Y. Tsien, “ReaChR: a redshifted variant of channelrhodopsin enables deep transcranial optogenetic excitation,” Nat. Neurosci., vol. 16, no. 10, pp. 1499–1508, Oct. 2013.
[53] D. Kültz, “Molecular and evolutionary basis of the cellular stress response,” Annu.
Rev. Physiol., vol. 67, no. 1, pp. 225–257, Mar. 2005.
[54] D. E. Clapham, “Calcium Signaling,” Cell, vol. 131, no. 6, pp. 1047–1058, Dec. 2007.
[55] J. L. Cross, B. P. Meloni, A. J. Bakker, S. Lee, and N. W. Knuckey, “Modes of Neuronal Calcium Entry and Homeostasis following Cerebral Ischemia,” Stroke Res.
Treat., vol. 2010, pp. 1–9, 2010.
[56] A. B. Parekh, “On the activation mechanism of store-operated calcium channels,”
Pflüg. Arch. - Eur. J. Physiol., vol. 453, no. 3, pp. 303–311, Nov. 2006.
[57] B. A. Simms and G. W. Zamponi, “Neuronal Voltage-Gated Calcium Channels:
Structure, Function, and Dysfunction,” Neuron, vol. 82, no. 1, pp. 24–45, Apr. 2014.
[58] H. S. Engelman and A. B. MacDermott, “Presynaptic ionotropic receptors and control
of transmitter release,” Nat. Rev. Neurosci., vol. 5, no. 2, pp. 135–145, Feb. 2004.
[59] S. Nakanishi, “Molecular diversity of glutamate receptors and implications for brain
function,” Science, vol. 258, no. 5082, pp. 597–603, Oct. 1992.
62
6
References
[60] S. Cull-Candy, S. Brickley, and M. Farrant, “NMDA receptor subunits: diversity, development and disease,” Curr. Opin. Neurobiol., vol. 11, no. 3, pp. 327–335, 2001.
[61] D. Smetters, A. Majewska, and R. Yuste, “Detecting action potentials in neuronal
populations with calcium imaging,” Methods, vol. 18, pp. 215–221, 1999.
[62] E. Yaksi and R. W. Friedrich, “Reconstruction of firing rate changes across neuronal
populations by temporally deconvolved Ca2+ imaging,” Nat. Methods, vol. 3, no. 5,
pp. 377–383, 2006.
[63] B.-Q. Mao, F. Hamzei-Sichani, D. Aronov, R. C. Froemke, and R. Yuste, “Dynamics
of spontaneous activity in neocortical slices,” Neuron, vol. 32, no. 5, pp. 883–898,
2001.
[64] C. Montell, “The venerable inveterate invertebrate TRP channels,” Cell Calcium, vol.
33, no. 5–6, pp. 409–417, May 2003.
[65] T. Numata, S. Kiyonaka, K. Kato, N. Takahashi, and Y. Mori, “Activation of TRP
Channels in Mammalian Systems,” in TRP Channels, M. X. Zhu, Ed. Boca Raton
(FL): CRC Press/Taylor & Francis, 2011.
[66] K. Venkatachalam and C. Montell, “TRP Channels,” Annu. Rev. Biochem., vol. 76,
no. 1, pp. 387–417, Jun. 2007.
[67] A. Y. Rhee, G. Li, J. Wells, and J. P. Y. Kao, “Photostimulation of sensory neurons
of the rat vagus nerve,” 2008, p. 68540E.
[68] E. S. Albert, J. M. Bec, G. Desmadryl, K. Chekroud, C. Travo, S. Gaboyard, F. Bardin, I. Marc, M. Dumas, G. Lenaers, C. Hamel, A. Muller, and C. Chabbert, “TRPV4
channels mediate the infrared laser-evoked response in sensory neurons,” J. Neurophysiol., vol. 107, no. 12, pp. 3227–3234, Jun. 2012.
[69] R. M. Luik, M. M. Wu, J. Buchanan, and R. S. Lewis, “The elementary unit of storeoperated Ca 2+ entry: local activation of CRAC channels by STIM1 at ER–plasma
membrane junctions,” J. Cell Biol., vol. 174, no. 6, pp. 815–825, Sep. 2006.
[70] A. B. Parekh, “Store-Operated Calcium Channels,” Physiol. Rev., vol. 85, no. 2, pp.
757–810, Apr. 2005.
[71] G. E. Kass and S. Orrenius, “Calcium signaling and cytotoxicity.,” Environ. Health
Perspect., vol. 107, no. Suppl 1, p. 25, 1999.
[72] J. E. Vance, “MAM (mitochondria-associated membranes) in mammalian cells: Lipids and beyond,” Biochim. Biophys. Acta BBA - Mol. Cell Biol. Lipids, vol. 1841, no.
4, pp. 595–609, Apr. 2014.
[73] S. Fulda, A. M. Gorman, O. Hori, and A. Samali, “Cellular Stress Responses: Cell
Survival and Cell Death,” Int. J. Cell Biol., vol. 2010, pp. 1–23, 2010.
[74] A. Görlach, K. Bertram, S. Hudecova, and O. Krizanova, “Calcium and ROS: A mutual interplay,” Redox Biol., vol. 6, pp. 260–271, Dec. 2015.
[75] I. Lee, E. Bender, S. Arnold, and B. Kadenbach, “New Control of Mitochondrial Membrane Potential and ROS Formation – A Hypothesis,” Biol. Chem., vol. 382, pp.
1629–1636, 2001.
[76] M. R. Duchen, “Contributions of mitochondria to animal physiology: from homeostatic
sensor to calcium signalling and cell death,” J. Physiol., vol. 516, no. 1, pp. 1–17,
1999.
[77] J. Lee, S. Kotliarova, Y. Kotliarov, A. Li, Q. Su, N. M. Donin, S. Pastorino, B. W.
Purow, N. Christopher, W. Zhang, J. K. Park, and H. A. Fine, “Tumor stem cells
derived from glioblastomas cultured in bFGF and EGF more closely mirror the phenotype and genotype of primary tumors than do serum-cultured cell lines,” Cancer
Cell, vol. 9, no. 5, pp. 391–403, May 2006.
[78] D. J. Combs, H.-G. Shin, Y. Xu, Y. Ramu, and Z. Lu, “Tuning voltage-gated channel
activity and cellular excitability with a sphingomyelinase,” J. Gen. Physiol., vol. 142,
no. 4, pp. 367–380, Oct. 2013.
[79] Y.-M. Leung, Y.-H. Kuo, C.-C. Chao, Y.-H. Tsou, C.-H. Chou, C.-H. Lin, and K.-L.
Wong, “Osthol is a Use-Dependent Blocker of Voltage-Gated Na + Channels in
Mouse Neuroblastoma N2A Cells,” Planta Med., vol. 76, no. 1, pp. 34–40, Jan. 2010.
[80] J. B. F. Van der Valk and H. P. M. Vijverberg, “Glutamate-induced inward current in
a clonal neuroblastoma cell line,” Eur. J. Pharmacol., vol. 185, pp. 99–102, 1990.
6
References
63
[81] R. Salto, J. D. Vílchez, M. D. Girón, E. Cabrera, N. Campos, M. Manzano, R. Rueda,
and J. M. López-Pedrosa, “β-Hydroxy-β-Methylbutyrate (HMB) Promotes Neurite
Outgrowth in Neuro2a Cells,” PLOS ONE, vol. 10, no. 8, p. e0135614, Aug. 2015.
[82] J. B. Olmsted, K. Carlson, R. Klebe, F. Ruddle, and J. Rosenbaum, “Isolation of
microtubule protein from cultured mouse neuroblastoma cells,” Proc. Natl. Acad.
Sci., vol. 65, no. 1, pp. 129–136, 1970.
[83] K. T. LePage, R. W. Dickey, W. H. Gerwick, E. L. Jester, and T. F. Murray, “On the
use of neuro-2a neuroblastoma cells versus intact neurons in primary culture for
neurotoxicity studies,” Crit. Rev. Neurobiol., vol. 17, no. 1, pp. 27–50, 2005.
[84] E. Cañete and J. Diogène, “Comparative study of the use of neuroblastoma cells
(Neuro-2a) and neuroblastoma×glioma hybrid cells (NG108-15) for the toxic effect
quantification of marine toxins,” Toxicon, vol. 52, no. 4, pp. 541–550, Sep. 2008.
[85] “Primary Mouse Cortical Neurons - Thermo Fisher Scientific.” [Online]. Available:
https://www.thermofisher.com/order/catalog/product/A15585. [Accessed: 10-Aug2016].
[86] R. Perrot, R. Berges, A. Bocquet, and J. Eyer, “Review of the Multiple Aspects of
Neurofilament Functions, and their Possible Contribution to Neurodegeneration,”
Mol. Neurobiol., vol. 38, no. 1, pp. 27–65, Aug. 2008.
[87] G. Thiel, “Synapsin I, synapsin II, and synaptophysin: marker proteins of synaptic
vesicles,” Brain Pathol., vol. 3, no. 1, pp. 87–95, 1993.
[88] S. Kalies, G. C. Antonopoulos, M. S. Rakoski, D. Heinemann, M. Schomaker, T.
Ripken, and H. Meyer, “Investigation of Biophysical Mechanisms in Gold Nanoparticle Mediated Laser Manipulation of Cells Using a Multimodal Holographic and Fluorescence Imaging Setup,” PLOS ONE, vol. 10, no. 4, p. e0124052, Apr. 2015.
[89] C. Grienberger and A. Konnerth, “Imaging Calcium in Neurons,” Neuron, vol. 73, no.
5, pp. 862–885, Mar. 2012.
[90] K. R. Gee, K. A. Brown, W.-N. U. Chen, J. Bishop-Stewart, D. Gray, and I. Johnson,
“Chemical and physiological characterization of fluo-4 Ca2+-indicator dyes,” Cell
Calcium, vol. 27, no. 2, pp. 97–106, Feb. 2000.
[91] “Fluo 3 | Dojindo.” [Online]. Available: http://www.dojindo.com/store/p/539-Fluo3.html. [Accessed: 10-Aug-2016].
[92] L. Bozzo, J. Puyal, and J.-Y. Chatton, “Lactate Modulates the Activity of Primary
Cortical Neurons through a Receptor-Mediated Pathway,” PLoS ONE, vol. 8, no. 8,
p. e71721, Aug. 2013.
[93] J. M. Diver, S. O. Sage, and J. A. Rosado, “The inositol trisphosphate receptor antagonist 2-aminoethoxydiphenylborate (2-APB) blocks Ca2+ entry channels in human platelets: cautions for its use in studying Ca2+ influx,” Cell Calcium, vol. 30, no.
5, pp. 323–329, Nov. 2001.
[94] M. D. Bootman, T. J. Collins, L. MacKenzie, H. L. Roderick, M. J. Berridge, and C.
M. Peppiatt, “2-aminoethoxydiphenyl borate (2-APB) is a reliable blocker of storeoperated Ca2+ entry but an inconsistent inhibitor of InsP3-induced Ca2+ release,”
FASEB J., vol. 16, no. 10, pp. 1145–1150, 2002.
[95] A. Ruiz, E. Alberdi, and C. Matute, “CGP37157, an inhibitor of the mitochondrial
Na+/Ca2+ exchanger, protects neurons from excitotoxicity by blocking voltage-gated
Ca2+ channels,” Cell Death Dis., vol. 5, no. 4, p. e1156, Apr. 2014.
[96] Y.-M. Leung, B.-T. Wu, Y.-C. Chen, C.-H. Hung, and Y.-W. Chen, “Diphenidol inhibited sodium currents and produced spinal anesthesia,” Neuropharmacology, vol. 58,
no. 7, pp. 1147–1152, Jun. 2010.
[97] A. M. Bowman, O. M. Nesin, O. N. Pakhomova, and A. G. Pakhomov, “Analysis of
Plasma Membrane Integrity by Fluorescent Detection of Tl+ Uptake,” J. Membr. Biol.,
vol. 236, no. 1, pp. 15–26, Jul. 2010.
[98] S. Kalies, S. Keil, S. Sender, S. C. Hammer, G. C. Antonopoulos, M. Schomaker, T.
Ripken, H. M. Escobar, H. Meyer, and D. Heinemann, “Characterization of the cellular response triggered by gold nanoparticle–mediated laser manipulation,” J. Biomed. Opt., vol. 20, no. 11, pp. 115005–115005, 2015.
64
6
References
[99] J. Schindelin, C. T. Rueden, M. C. Hiner, and K. W. Eliceiri, “The ImageJ ecosystem:
An open platform for biomedical image analysis,” Mol. Reprod. Dev., vol. 82, no. 7–
8, pp. 518–529, Jul. 2015.
[100] J. Schindelin, I. Arganda-Carreras, E. Frise, V. Kaynig, M. Longair, T. Pietzsch,
S. Preibisch, C. Rueden, S. Saalfeld, B. Schmid, J.-Y. Tinevez, D. J. White, V.
Hartenstein, K. Eliceiri, P. Tomancak, and A. Cardona, “Fiji: an open-source platform
for biological-image analysis,” Nat. Methods, vol. 9, no. 7, pp. 676–682, Jul. 2012.
[101] John A. Buege and Steven D. Aust, “Microsomal lipid peroxidation,” in Methods
in Enzymology: Biomembranes - Part C: Biological Oxidations, vol. 52, Elsevier,
1978, pp. 302–310.
[102] R Development Core Team, R: A language and environment for statistical computing. Vienna, Austria, 2008.
[103] T. Hothorn, F. Bretz, and P. Westfall, “Simultaneous inference in general parametric models,” Biom. J. Biom. Z., vol. 50, no. 3, pp. 346–363, Jun. 2008.
[104] J. Fox and S. Weisberg, An R Companion to Applied Regression. SAGE, 2010.
[105] C.-S. Kuo, C.-Y. Kwan, C.-L. Gong, M.-F. Tsai, S. Nishibe, J. Tatsuzaki, and Y.M. Leung, “Apocynum venetum leaf aqueous extract inhibits voltage-gated sodium
channels of mouse neuroblastoma N2A cells,” J. Ethnopharmacol., vol. 136, no. 1,
pp. 149–155, Jun. 2011.
[106] S. G. Brohawn, Z. Su, and R. MacKinnon, “Mechanosensitivity is mediated directly by the lipid membrane in TRAAK and TREK1 K + channels,” Proc. Natl. Acad.
Sci., vol. 111, no. 9, pp. 3614–3619, Mar. 2014.
[107] R. J. Klebe and F. H. Ruddle, “Neuroblastoma-Cell culture analysis of a differentiating stem cell system.,” J. Cell Biol., vol. 43, no. 2, 1969.
[108] P. G. Nelson, J. H. Peacock, T. Amano, and J. Minna, “Electrogenesis in mouse
neuroblastoma cells in vitro,” J. Cell. Physiol., vol. 77, no. 1, pp. 337–352, 1971.
[109] “Neuro-2a ATCC ® CCL-131TM Mus musculus brain neuroblastoma.” [Online].
Available: https://www.lgcstandards-atcc.org/products/all/CCL-131.aspx?geo_country=de#characteristics. [Accessed: 19-Aug-2016].
[110] K.-O. Cho, C. A. Hunt, and M. B. Kennedy, “The rat brain postsynaptic density
fraction contains a homolog of the Drosophila discs-large tumor suppressor protein,”
Neuron, vol. 9, no. 1, pp. 929–942, 1992.
[111] S. Dieck, L. Sanmartí-Vila, K. Langnaese, K. Richter, S. Kindler, A. Soyke, H.
Wex, K.-H. Smalla, U. Kämpf, J.-T. Fränzer, and others, “Bassoon, a novel zincfinger CAG/glutamine-repeat protein selectively localized at the active zone of presynaptic nerve terminals,” J. Cell Biol., vol. 142, no. 2, pp. 499–509, 1998.
[112] M. Boyce and J. Yuan, “Cellular response to endoplasmic reticulum stress: a
matter of life or death,” Cell Death Differ., vol. 13, no. 3, pp. 363–373, 2006.
[113] D. Sanders, J. Pelloux, C. Brownlee, and J. F. Harper, “Calcium at the crossroads
of signaling,” Plant Cell, vol. 14, no. suppl 1, pp. S401–S417, 2002.
[114] R. J. Kaufman, “Stress signaling from the lumen of the endoplasmic reticulum:
coordination of gene transcriptional and translational controls,” Genes Dev., vol. 13,
no. 10, pp. 1211–1233, 1999.
[115] Y. Huang and J. W. Putney, “Relationship between Intracellular Calcium Store
Depletion and Calcium Release-activated Calcium Current in a Mast Cell Line (RBL1),” J. Biol. Chem., vol. 273, no. 31, pp. 19554–19559, 1998.
[116] A. Martin, T.-H. Tang, J. Erickson, A. Martin, Y.-W. Huang, H.-J. Wang, R. Reichard, C.-T. Chang, R. Aronstam, and E. Shannon, “Oxidative stress disruption of receptor-mediated calcium signaling mechanisms,” 2013.
[117] F. Qin, “Time-Resolved Activation of Thermal TRP Channels by Fast Temperature Jumps,” in TRP Channels, M. X. Zhu, Ed. Boca Raton (FL): CRC Press/Taylor
& Francis, 2011.
[118] T. Hofmann, A. G. Obukhov, M. Schaefer, C. Harteneck, T. Gudermann, and G.
Schultz, “Direct activation of human TRPC6 and TRPC3 channels by diacylglycerol,”
Nature, vol. 379, no. 1, pp. 259–263, 1999.
6
References
65
[119] G. Vazquez, B. J. Wedel, O. Aziz, M. Trebak, and J. W. Putney, “The mammalian
TRPC cation channels,” Biochim. Biophys. Acta BBA - Mol. Cell Res., vol. 1742, no.
1–3, pp. 21–36, Dec. 2004.
[120] C.-P. Richter and X. Tan, “Photons and neurons,” Hear. Res., vol. 311, pp. 72–
88, May 2014.
[121] S.-I. Choi, S. Yoo, J. Lim, and S. Hwang, “Are Sensory TRP Channels Biological
Alarms for Lipid Peroxidation?,” Int. J. Mol. Sci., vol. 15, no. 9, pp. 16430–16457,
Sep. 2014.
[122] S. Koike, N. Shibuya, H. Kimura, K. Ishii, and Y. Ogasawara, “Polysulfide promotes neuroblastoma cell differentiation by accelerating calcium influx,” Biochem.
Biophys. Res. Commun., vol. 459, no. 3, pp. 488–492, Apr. 2015.
[123] J. Yao, B. Liu, and F. Qin, “Kinetic and Energetic Analysis of Thermally Activated
TRPV1 Channels,” Biophys. J., vol. 99, no. 6, pp. 1743–1753, Sep. 2010.
[124] J. K. Armstrong, R. B. Wenby, H. J. Meiselman, and T. C. Fisher, “The Hydrodynamic Radii of Macromolecules and Their Effect on Red Blood Cell Aggregation,”
Biophys. J., vol. 87, no. 6, pp. 4259–4270, Dec. 2004.
[125] A. K. Campbell, Intracellular Calcium. John Wiley & Sons, 2014.
[126] J. Turnay, N. Olmo, M. Gasset, I. Iloro, J. L. R. Arrondo, and M. A. Lizarbe, “Calcium-dependent conformational rearrangements and protein stability in chicken annexin A5,” Biophys. J., vol. 83, no. 4, pp. 2280–2291, 2002.
[127] S. I. Belli, A. Sali, and J. W. Goding, “Divalent cations stabilize the conformation
of plasma cell membrane glycoprotein PC-1 (alkaline phosphodiesterase I),” Biochem. J., vol. 304, no. 1, pp. 75–80, 1994.
[128] S. Onizuka, R. Tamura, T. Yonaha, N. Oda, Y. Kawasaki, T. Shirasaka, S. Shiraishi, and I. Tsuneyoshi, “Clinical dose of lidocaine destroys the cell membrane and
induces both necrosis and apoptosis in an identified Lymnaea neuron,” J. Anesth.,
vol. 26, no. 1, pp. 54–61, Feb. 2012.
[129] G. P. Drummen, L. C. van Liebergen, J. A. O. den Kamp, and J. A. Post, “C11BODIPY 581/591, an oxidation-sensitive fluorescent lipid peroxidation probe:(micro)
spectroscopic characterization and validation of methodology,” Free Radic. Biol.
Med., vol. 33, no. 4, pp. 473–490, 2002.
[130] J. Yoon, S. Ryu, S. Lee, and C. Choi, “Cytosolic Irradiation of Femtosecond Laser
Induces Mitochondria-dependent Apoptosis-like Cell Death via Intrinsic Reactive Oxygen Cascades,” Sci. Rep., vol. 5, p. 8231, Feb. 2015.
[131] D. Yang, S. G. Elner, Z.-M. Bian, G. O. Till, H. R. Petty, and V. M. Elner, “Proinflammatory cytokines increase reactive oxygen species through mitochondria and
NADPH oxidase in cultured RPE cells,” Exp. Eye Res., vol. 85, no. 4, pp. 462–472,
Oct. 2007.
[132] S. Tan, Y. Sagara, Y. Liu, P. Maher, and D. Schubert, “The regulation of reactive
oxygen species production during programmed cell death,” J. Cell Biol., vol. 141, no.
6, pp. 1423–1432, 1998.
[133] V. L. Kolossov, J. N. Beaudoin, W. P. Hanafin, S. J. DiLiberto, P. J. A. Kenis, and
H. Rex Gaskins, “Transient light-induced intracellular oxidation revealed by redox
biosensor,” Biochem. Biophys. Res. Commun., vol. 439, no. 4, pp. 517–521, Oct.
2013.
[134] B. Coste, J. Mathur, M. Schmidt, T. J. Earley, S. Ranade, M. J. Petrus, A. E.
Dubin, and A. Patapoutian, “Piezo1 and Piezo2 Are Essential Components of Distinct Mechanically Activated Cation Channels,” Science, vol. 330, no. 6000, pp. 55–
60, Oct. 2010.
[135] C.-P. Richter, S. Rajguru, and M. Bendett, “Infrared neural stimulation in the cochlea,” 2013, p. 85651Y.
[136] K. Szydlowska and M. Tymianski, “Calcium, ischemia and excitotoxicity,” Cell
Calcium, vol. 47, no. 2, pp. 122–129, Feb. 2010.
66
6
References
[137] N. G. Laxpati, B. Mahmoudi, C.-A. Gutekunst, J. P. Newman, R. Zeller-Townson,
and R. E. Gross, “Real-time in vivo optogenetic neuromodulation and multielectrode
electrophysiologic recording with NeuroRighter,” Front. Neuroengineering, vol. 7,
Oct. 2014.
[138] U. Nussinovitch and L. Gepstein, “Optogenetics for in vivo cardiac pacing and
resynchronization therapies,” Nat. Biotechnol., vol. 33, no. 7, pp. 750–754, Jun.
2015.
[139] A. D. Edelstein, M. A. Tsuchida, N. Amodaj, H. Pinkard, R. D. Vale, and N. Stuurman, “Advanced methods of microscope control using μManager software,” J. Biol.
Methods, vol. 1, no. 2, p. e10, Nov. 2014.
7
Supplementary materials
7
67
Supplementary materials
This section lists the protocols for the experimental methods, the materials as
well as additional figures and tables.
7.1
Methods
7.1.1 Cell culture
N2A cells
N2A cells were cultured in 100 mm TC-dishes in MEM Eagle with EBSS. The
culture medium was supplemented with 1 % penicillin/streptomycin and 10 %
FCS. Cells were passaged 1-2 times per week according to the following standard protocol:

Discard old culture medium

Rinse cells with warm PBS w/o Ca2+/Mg2+

Add 1-2 mL warm trypsin/EDTA, incubate for 5 min at room temperature

Add the same volume of culture medium, rinse TC-dish, transfer all of the
liquid to a 15 mL centrifuge tube

Centrifuge at 500 rcf, room temperature for 5 min

Discard liquid

Resuspend pellet in 1 mL fresh medium
o Optional: Further dilute cells, use a Neubauer or Fuchs-Rosenthal
chamber to count cells

Transfer about 1/10th to 1/5th of the cells to a new TC-dish containing
10 mL pre-warmed culture medium with supplements

Incubate cells at 37°C, 5 %CO2

If cells are only passaged once a week: Exchange culture medium 3-4
days after passaging
Cells were frozen for easy transportation:
68
7
Supplementary materials

Follow the protocol above until cells are counted

Transfer 1-2*106 cells to a new centrifuge tube, centrifuge at 500 rcf, room
temperature for 5 min

Re-suspend pellet with chilled freezing medium (culture medium + supplements, containing 10 % sterile DMSO)

Quickly transfer cells to labeled cryo tube and immediately freeze tube at
-80°C
Thawing:

Quickly thaw cells in water bath (37°C)

Transfer liquid from cryo tube to 10 mL pre-warmed culture medium in a
15 mL centrifuge tube

Centrifuge at 500 rcf, room temperature for 5 min

Discard liquid

Re-suspend pellet in fresh culture medium, optionally count cells

Transfer cells at the desired density to new TC-dishes
MCN
Primary mouse cortical neurons were thawed and cultured according to the supplier’s protocol, adapted for the intended use:

Coat 15 glass bottom dishes (35 mm, ibidi) with sterile filtered poly-d-lysine (0.1 mg/mL, about 300 µL per dish), incubate for approx. 30 min, rinse
with sterile H2O and let dry

Culture medium: Neurobasal® Medium containing 0.5 mM GlutaMAXTM-I
and 2 % (v/v) B-27®

Rinse a 50 mL conical tube with pre-warmed culture medium

Quickly thaw frozen vial in water bath (37°C)

Rinse a 100 µL pipette tip with culture medium and gently transfer cells to
the pre-rinsed 50 mL tube

Rinse the vial with 1 mL of pre-warmed culture medium and very slowly
add the medium to the 50 mL tube, mix gently

Slowly add 2 mL of culture medium to the cells, mix gently

Distribute cells evenly to the coated ibidi-dishes
7
Supplementary materials
69

The next day, aspirate half of the culture medium and replace with new
medium; feed cells this way every three days

Use for laser stimulation after 10-13 days in culture
Note: upon the first thawing, the cells were split to four aliquots à 10 6 cells each,
only one of which was treated as described while the other three were immediately refrozen. One of those aliquots was later used for laser stimulation as well.
The viability was decreased and cells were cultured for 10-20 days. Cell physiology did not appear to be altered otherwise.
3T3
3T3 fibroblasts were treated like N2A cells, but cultured in DMEM/F12 containing
10 % FCS and 1 % Pen-Strep.
7.1.2 Preparation for laser stimulation
MCNs were stimulated in the ibidi-dishes they were cultured in. N2A cells were
seeded in 35 mm glass bottom dishes according to the protocol above at a density of 65,000 to 90,000 cells per dish. Cells were allowed to settle for 1-2 days,
then the culture medium was replaced by MEM Eagle containing 1 % Pen-Strep
but no FCS. The initial seeding density was adjusted to the incubation time in
medium + FCS, as the cells continue dividing during that time. For stimulation
experiments, a confluency of about 60-70 % was desired. Cells were kept in serum-free medium for 2-3 days.
On the day of the stimulation experiment, cells (N2A or MCN) were prepared in
the way required for the respective experiment:
Gold nanoparticles
For all stimulation experiments, cells were treated with AuNP. The nanoparticles
were added to the cells at a density of 0.5 µg/cm2 (85 µL per dish with the concentrations used) in the respective culture medium. The cells were then incubated
for approx. 3 hours to allow the AuNP to settle. This step was not performed with
control cells for all experiments.
70
7
Supplementary materials
Fluo 4-AM staining
To perform calcium imaging, cells were stained according to the following protocol:

Prepare a 2 mM stock solution of Fluo 4-AM in DMSO

Final dye concentration for N2A cells: 5 µM; for MCN: 1 µM

Mix Fluo 4-AM solution 1:1 with Pluronic F-127 (10 % w/v in DMSO)

Add 1 mL of medium (serum-free for N2A, OptiMEM for MCN) to Fluo 4 +
Pluronic F-127

Exchange medium on cells for staining solution

Incubate for 30 min (MCN) to 1 h (N2A) at 37°C

Rinse with PBS

Add new medium according to subsequent experiment
Cells were allowed to complete processing of the dye for about 15 minutes before
imaging. Dishes were kept at 37°C or room temperature.
Propidium iodide staining
Right before the stimulation experiment, propidium iodide (1 mg/mL in water) was
added to the dish at a final concentration of 2 µg/mL. Incubation time before the
start of the experiment was approx. 10 min at room temperature.
BODIPY 581/591 staining
To assess lipidperoxidation, cells were stained with BODIPY 581/591 according
to the following protocol:

Prepare a stock solution of the dye (5 mM in DMSO)

Add dye to the cells at a final concentration of 10 µM

Incubate for 30 min at 37°C, 5 % CO2

Rinse with PBS three times

Add respective test medium to cells
Inhibitor studies
To test the influence of the different blockers on the calcium response, they were
added after Fluo 4-staining. In the case of PBS and PBSC, the culture medium
7
Supplementary materials
71
was replaced with the respective buffer after staining. The inhibitors were added
to the cells in medium or buffer. For solvents and concentrations see Table S 4.
Cells were incubated with the inhibitors for approx. 15 minutes at room temperature before laser stimulation.
7.1.3 Laser stimulation
Single cell stimulation
The dish was placed in the setup depicted in figure 3. Laser stimulation and image
acquisition was performed with the following parameters:

Pulse duration: 40 ms

Exposure time: 200 ms; 150 ms for calcium imaging in MCN
o Image acquisition at live resolution

Recording time
o Before laser stimulus: 5-10 s
o After stimulus: 30 s for MCN, 60-90 s for N2A, 2 min for initial experiments with BODIPY 581/591 staining (both cell types)
The images were then analyzed with Fiji.
7.1.4 Immunofluorescence
Immunolabeling of the antigens neurofilament (NF), synaptophysin (SYP) and
NMDA-receptor (NMDA-R) was performed according to the protocol below. For
concentrations and properties of the antibodies see Table S 5.

Seed about 30,000 cells per well in a black 96-well plate (protocol for passaging see section 7.1.1)
o Culture N2A cells in serum-free medium to induce differentiation,
see section 7.1.2
o keep non-differentiated N2A cells and 3T3 cells with medium + FCS

Rinse cells with PBS + 1 mM Ca2+ + 1 mM Mg2+
72
7
Supplementary materials

Fix in PFA (4 % in PBS) for 15 min at room temperature

Wash three times with PBS + 100 mM glycine (5 min each)

Permeabilize cells with PBS + 0.5 % Tween 20 for 30 min
o Not for NMDA-R: epitope is extracellular

Wash three times with PBS (5 min each)

Block with 10 % BSA in PBS for at least 30 min

Add primary antibody at respective concentration in blocking solution, incubate for 1 h at room temperature or overnight at 4°C

Wash three times with PBS (5 min each)

Add secondary antibody diluted in blocking solution, incubate 1 h in the
dark

Wash three times with PBS and once with dH2O
7
7.2
Supplementary materials
73
Materials
Table S 1: List of equipment and consumables used.
Device
Model
Nd:YAG laser
Manufacturer
Horus Scientific, France
Microscope camera
ProgRes MF cool
Optical components
Thorlabs, Germany
Eco safe basic plus
Clean Bench
Maxisafe 2020 Class II
Incubator
Centrifuge
Water bath
Jenoptik, Germany
ENVAIR, Germany
Thermo Fisher Scientfic,
USA
Midi 40
Thermo Fisher Scientific,
Heracell VIOS 160i
USA
Z 206 A
Hermle, Germany
witeg Labortechnik, Germany
Scale
Sartorius, Germany
Pipettes
Eppendorf, Germany
Pipette tips 10 / 200 /
Sarstedt, Germany
1000 µL
TC-dishes 60 / 100 mm
Glass
bottom
dishes
35 mm
Centrifuge tubes 15 /
50 mL
Microcentrifuge
tubes
1.5 / 2 mL
Cryo tubes
Disposable
Sarstedt, Germany
Ibidi, Germany
Sarstedt, Germany
Sarstedt, Germany
Sarstedt, Germany
serological
pipettes, 2 / 5 / 10 /
Sarstedt, Germany
25 mL
Nitril gloves
VWR, USA
96-well plates, black
Corning GmbH, USA
74
7
Supplementary materials
Table S 2: List of software used to evaluate and/or present data.
Software
Version
Distributor
MS Office
2010, 365
Microsoft, USA
Fiji [99, 137]
2.35
NIH, USA
Tableau Desktop
9.3
Tableau, USA
R [102]–[104]
3.1.0
R Foundation for Statistical Computing,
Austria
Inkscape
0.91
FSF, USA
Zotero Standalone
4.0.29.10
Center for History and New Media,
George Mason University, USA
Shutter Controler SC10 1.0
Thorlabs, Germany
Micro-Manager [139]
NIH, USA
1.4.5
Table S 3: List of chemicals, solutions and dyes.
Chemical
Supplier
MgCl2 * 6 H2O
Carl Roth, Germany
CaCl2 * 2 H2O
Carl Roth, Germany
Paraformaldehyde
Carl Roth, Germany
Bovine serum albumin
Sigma-Aldrich, USA
Tween 20
Carl Roth, Germany
Glycine
Carl Roth, Germany
DMSO
Sigma-Aldrich, USA
PBS (with or without Ca2+, Mg2+)
PAN-Biotech, Germany
MEM Eagle
PAN-Biotech, Germany
OptiMEM
Thermo Fisher Scientific, USA
7
Supplementary materials
75
Chemical
Supplier
DMEM/F12
PAN-Biotech, Germany
Neurobasal® Medium
Thermo Fisher Scientific, USA
GlutaMAXTM-I
Thermo Fisher Scientific, USA
B-27®
Thermo Fisher Scientific, USA
Fetal calf serum
Merck, Germany
Penicillin-Streptomycin
Merck, Germany
Typsin/EDTA
PAN-Biotech, Germany
Gold nanoparticles, 200 nm
Kisker Biotech, Germany
Propidium iodide
Thermo Fisher Scientific, USA
Fluo 4-AM
Thermo Fisher Scientific, USA
BODIPY 581/591 C11
Thermo Fisher Scientific, USA
76
7
Supplementary materials
Table S 4: List of inhibitors with the respective solvents, final concentrations in an experiment and
mode of action on the cellular level. All inhibitors were obtained from Cayman Chemical Company,
USA.
Inhibitor
Solvent
Final concentration
Mode of action
2-APB
DMSO
75 µmol/L
Blocks IP3 receptors in the
ER-membrane; also acts
upon TRP channels unspecifically
CGP37157 DMSO
20 µmol/L
Selectively inhibits the mitochondrial Na+/Ca2+ exchanger and thereby Ca2+
extrusion from mitochondria
Lidocaine
DMSO
100 (N2A) /
Blocks voltage-gated so-
300 (MCN) µmol/L
dium channels, therefore
suppresses
transmem-
brane sodium currents
Table S 5: List of antibodies, their final concentration as used in the experiments and the suppliers.
Antigen
Dilution/concentration
Supplier
Glu-N2B (NMDA-recep- 15 µg/mL
Acris Antibodies, Ger-
tor)
many
Synaptophysin
1:100
Acris Antibodies, Germany
Neurofilament
1:10000
Abcam, United Kingdom
7
7.3
Supplementary materials
Supplementary results
A) N2A
77
78
7
Supplementary materials
B) MCN
Figure S 1: ΔF/F and time to peak vs. baseline fluorescence values of N2A cells (A) and MCNs
(B). For each cell displaying a laser-induced calcium response, either in the initial experiments
(section 3.2.1, left columns) or the whole data set (right columns), the variables of the calcium
response are plotted against the cell’s baseline fluorescence value. A) In N2A cells, no trends
become apparent that suggest a dependence of either of the variables on baseline fluorescence.
B) When pooling all data of MCNs, a very weak negative correlation between baseline fluorescence and time to peak is revealed. A negative correlation between the cells’ baseline fluorescence and the peak value of the laser-induced response is observable after the initial experiments, but not in the pooled data set.
7
Supplementary materials
79
A)
***
*
***
*
*
*
B)
*
**
*
p=
0.06
*
p = 0.06
p = 0.07
Figure S 2: ΔF/F (A) and time to peak (B) for N2A cells in different conditions, med = medium,
A = 2-APB, C = CGP, L = lidocaine. For all conditions not included in Figure 13, one experiment
was performed with n ≈ 15-20 cells. Outliers are excluded. For reasons of clarity, significant levels
are only indicated for comparisons that are not shown in Figure 13. Adding more than one inhibitor
to culture medium does not further change values of ΔF/F. Also, none of the inhibitors further
decreases fluorescence change in cells in PBS. Time to peak on the other hand increases when
adding inhibitors to PBS. Reference points for the indicated comparisons are always the rightmost
columns.
80
7
Supplementary materials
Figure S 3: Laser-induced calcium response of N2A cells under different conditions grouped by
presence or absence of a detectable PI-signal. Outliers are excluded. Under some conditions,
ΔF/F or time to peak appears to vary between perforated and intact cells, although the sample
size is often small. In general, membrane perforation accompanied by PI-influx does not affect
either of the properties of the laser-induced calcium response.
7
Supplementary materials
81
205.3 ± 165
39 ± 44
Figure S 4: Distribution of peak ΔF/F values in N2A cells without (left) or with (right) a slower
(spontaneous) transient following the laser-induced calcium response. Each data point represents
ΔF/F of an individual cell. Cells are separated according to condition and experimental iteration
for a better overview of the data points. Arithmetic means and standard deviation of all respective
cells are given. Cells that display a second calcium transient exhibit a markedly weaker calcium
response to the laser stimulus.
82
7
Supplementary materials
Figure S 5: ΔF/F values of BODIPY 581/591 oxidation. Data points represent spontaneous increase in fluorescence in non-irradiated N2A cells. Spontaneous oxidation processes do not depend on the condition or presence of AuNP.
7
Supplementary materials
83
Table S 6: Adjusted p-values of pairwise comparisons for ΔF/F (N2A cells). Values < 0.05 are
displayed in bold. The proportion of cells responding to the laser stimuli is listed as well. Total
number of cells per condition is given in parentheses.
Adjusted p-values for ΔF/F
response
(%)
med
med +
med +
med +
A
C
L
PBS
PBSC
med (33)
85
-
0.628
0.544
0.866
0.052
0.099
med + A (30)
73
0.628
-
1
0.149
0.523
0.007
med + C (29)
90
0.544
1
-
0.118
0.606
0.006
med + L (33)
85
0.866
0.149
0.118
-
0.008
0.483
44
0.4
0.324
0.324
-
-
-
88
0.78
0.249
0.249
0.824
-
-
PBS (31)
97
0.052
0.523
0.606
0.008
-
< 0.001
PBS + A (15)
93
-
0.228
-
-
0.228
-
PBS + C (20)
100
-
-
< 0.001
-
0.606
-
PBS + L (16)
75
-
-
-
0.019
0.352
-
100
-
0.023
0.023
-
0.342
-
100
-
0.101
0.101
< 0.001
0.285
-
100
0.099
0.007
0.006
0.483
< 0.001
-
100
-
0.234
0.234
< 0.001
-
< 0.001
med + AC
(16)
med + ACL
(16)
PBS + AC
(16)
PBS + ACL
(16)
PBSC (32)
PBSC + ACL
(6)
84
7
Supplementary materials
Table S 7: Adjusted p-values of pairwise comparisons for time to peak (N2A cells). Values < 0.05
are displayed in bold. Total number of cells per condition is given in parentheses.
Adjusted p-values for time to peak
med
med + A
med + C
med + L
PBS
PBSC
med (33)
-
0.992
0.657
0.992
1
0.965
med + A (30)
0.992
-
0.922
1
0.999
0.756
med + C (29)
0.657
0.922
-
0.918
0.789
0.262
med + L (33)
0.992
1
0.918
-
1
0.761
med + AC (16)
0.602
0.602
0.602
-
-
-
med + ACL (16)
0.074
0.42
0.773
0.42
-
-
PBS (31)
1
0.999
0.789
1
-
0.9
PBS + A (15)
-
0.123
-
-
0.127
-
PBS + C (20)
-
-
0.019
-
0.001
-
PBS + L (16)
-
-
-
0.369
0.128
-
PBS + AC (16)
-
0.209
0.395
-
0.014
-
PBS + ACL (16)
-
0.083
0.18
0.083
0.083
-
PBSC (32)
0.965
0.756
0.262
0.761
0.9
-
PBSC + ACL (6)
-
0.058
0.011
0.058
-
0.259
7
Supplementary materials
85
Table S 8: Adjusted p-values of pairwise comparisons of ΔF/F (MCNs). Values < 0.05 are displayed in bold. The proportion of cells (total number given in parenthesis) is also given.
Response
(%)
Adjusted p-values of pairwise comparisons of ΔF/F
PBSC
PBS
PBSC + A
PBSC + C
PBSC + L
PBSC (41)
73.2
-
0.11
0.38
0.077
0.93
PBS (15)
86.7
0.11
-
0.006
8*10-7
0.002
PBSC + A (15)
66.7
0.38
0.006
-
0.65
0.49
PBSC + C (14)
92.9
0.077
8*10-7
0.65
-
0.045
PBSC + L (10)
90
0.93
0.002
0.49
0.045
-
Table S 9: Adjusted p-values of pairwise comparisons of time to peak (MCNs). Values < 0.05 are
displayed in bold. Total number of cells per condition is given in parenthesis.
Adjusted p-values of pairwise comparisons of time to peak
PBSC
PBS
PBSC + A
PBSC + C
PBSC + L
PBSC (41)
-
0.49
0.39
0.93
0.69
PBS (15)
0.49
-
0.7
0.1
0.06
PBSC + A (15)
0.39
0.7
-
0.01
0.003
PBSC + C (14)
0.93
0.1
0.01
-
0.86
PBSC + L (10)
0.69
0.06
0.003
0.86
-
86
7
Supplementary materials
Table S 10: Results of lipid peroxidation experiments. Right-tailed t tests were performed to compare fluorescence increase and time to peak in cells with and without AuNP. Even without cellular
involvement, some oxidation of the dye is expected to occur due to photooxidation. In N2A cells
treated with AuNP, die oxidation process is significantly stronger and slower than in the control
under almost all conditions. In MCNs, the same trend is observable. This suggests an oxidation
process beyond mere photooxidation.
Cell
Condition
type
Radiant expo-
Adjusted p-
Adjusted p-value
sure (mJ/cm2)
value ΔF/F
time to peak
N2A
medium
25
0.004
0.01
N2A
medium
51
0.002
0.25
N2A
PBS
51
0.004
0.01
N2A
PBSC
51
0.002
0.01
MCN
PBSC
51
0.13
0.47
8
8
Acknowledgements
87
Acknowledgements
First of all, I want to thank Dr. Dag Heinemann for welcoming me into his group,
including me in ongoing research, enabling me to work independently while preventing me from getting lost, and of course being the second examiner for this
thesis. I am thankful to PD Dr. Hugo Murua-Escobar for being the first examiner
of my thesis. I also want to thank all members of the biomedical optics department
for creating a pleasant working climate, assisting me at various opportunities, including me in after-work activities and broadening my horizon during lunch
breaks. Lastly, I want to thank friends and family for their support and understanding especially during the writing of this thesis.