Layer- and age-specific functional properties of synapses onto pyramidal
cells from mouse prefrontal cortex
EWS Carbo 10416773 Supervisor: DC Rotaru; INF, VU
Improper or latent development of the prefrontal cortex (PFC) has been related to many cognitive deficits, from risk-seeking teenage behavior to schizophrenia. Information processing within the
cortex seems to be dependent on layer organization, and it has been suggested that information flows from the upper to the deeper layers. However, little research on layers of the PFC has been conducted. To understand the contribution of layer 2/3 (L2/3) and layer 5 (L5) of the PFC to the processing of information, we question: do glutamatergic and GABAergic synapses onto pyramidal cells of the PFC show different properties across layers and do they continue to develop after birth?
To address this question, we examined the properties of mini postsynaptic currents, and their events, which are the result of stochastic spontaneous vesicle release. Mature pyramidal cells were found to display more variability in EPSP/IPSP ratio than cells of young animals, suggesting a more specialized role for cells after development. A difference found in ratio across layers, suggest that L5
has more of an information-filtering role than L2/3. In contrast, the role of L2/3 might focus more on the collecting and merging of information from different sources. Significant changes in the rise time of inhibitory event of synapses onto mature L5 pyramidal cells, and decay in excitatory events
of mature L2/3 synapses suggest a possible change in respectively GABA- and AMPA receptor-composition of these specific groups. Overall this research shows that differences in receptor-composition of
synapses onto pyramidal cells are present between layers of the PFC, and changes occur during development of mice.
Introduction
The prefrontal cortex (PFC) mediates many different executive functions – e.g., impulse control, goal-directed and internally guided behavior – contributing to rational thinking and risk- and urge-control (Fuster, 2002; Kolb et al., 2012). The PFC is one of last brain regions to reach maturity, and its latent or improper development has been related to many cognitive deficits from risk-seeking teenage behavior to schizophrenia (Gogtay et al., 2004; see Selemon, 2013). As a possible candidate for these deficits, the temporal development of information processing in the cortex has recently been put forward (Yizhar et al., 2011).
Information processing within the cortex has been found to be dependent on intra- and inter-layer organization (Weiler et al., 2008). Human PFC consists out of 6 layers, running from the upper layers (L1-L2-L3) to the deeper (i.e., lower) layers (L5-L6). Many different intra-PFC pathways have been suggested, but the currently accepted hypothesis indicates that information flows from upper layers to lower layers, before being send onwards to various other brain areas.
To understand the contribution of each layer to improper information processing in these deficits, we first have to research the properties of layer development in a healthy situation. Conducted research on different cortical layers has mostly focused on the somatosensory cortex of rats and mice (e.g., Shepherd et al., 2005), and has been extensively done on the visual cortex of cats (e.g., Binzegger
et al., 2004). Studies such as these generally
found elevated electrical activity among deeper layers compared to upper layers of various cortical areas (Barth & Poulet, 2012). These findings suggest that synapses onto cells of different layers will show variability in their properties.
A.
B.
C.
1
30 EC, 4-3-2013 t/m 9-8-2013
UvA Representative/Co-assessor: Harm Krugers MSc in Brain and Cognitive Sciences
Even though research on layers has been conducted in different parts of the cortex, it is the development of the prefrontal part of the cortex, which raises some of the most
interesting clinical questions relating to cognitive deficits. Moreover, the functional maturation of synapses in different layers of the PFC has not been previously studied. Therefore, in this study we compare the properties of synapses onto pyramidal cells (PC) between layer 2/3 (L2/3) and layer 5 (L5) of the PFC. We also look at possible changes during
development in mice of different ages. To understand the development and layer-specific difference in a healthy situation, we question: do separate layers of the PFC relay information differently, and what changes occur in synapses onto PC of these layers during normal cortex development?
The functioning of each layer within the pathway is dependent on multiple factors, but especially the imbalance of excitation and inhibition (E/I) has been hypothesized to give rise to behavioural deficits such as autism and schizophrenia (Yizhar et al., 2011). Excitatory postsynaptic currents (EPSCs) and inhibitory PSCs (IPSCs) are caused by vesicular releases from terminals, which respectively slightly depolarizes or hyperpolarizes the cell. This change in resting membrane potential influences the cell’s proximity to the firing threshold (figure 1A). Excitatory postsynaptic
potentials (EPSPs) are a result of the
depolarizing flow of ions, such as sodium (Na+),
after binding of glutamate to an ionotropic glutamate receptor (iGluR; figure 1B). In
contrast, inhibitory postsynaptic potentials (IPSPs) are caused by a hyperpolarizing influx of chloride (Cl-), after the binding of GABA to its
receptor (figure 1B). To understand how the E/I balance changes with age in different layers of the PFC we assessed mini PSCs (mPSCs), i.e., events, which are the result of stochastic spontaneous release of neurotransmitter into the synaptic cleft.
Comparing the frequency (figure 1C) and the properties of events, leads to a functional analysis of synapse composition across layers and time and can answer the following question: do glutamatergic and GABAergic synapses on pyramidal cells of the PFC show different properties across layers and do they continue to develop after birth?
Material and Methods
Slices
Experiments were conducted in mice classified as young, meaning under 15 days old (≤P14), and mature mice, i.e., at least 2 months old (>8weeks). All adult mice used were in between 8 and 16 weeks old. Brain slices are prepared from the frontal cortex of C57BL/6
mice of either sex (Charles River), using methods described previously (Rotaru et al.,
2007). Animals were treated following procedures in accordance with Animal User Care Committee of Vrije Universiteit of Amsterdam. Coronal PFC slices of 300 mμ thickness were cut on a vibrating microtome submerged in cold modified artificial
cerebrospinal fluid (ACSF) consisting of 125 mM NaCl, 3 mM KCl, 1.25 mM NaH2PO4, 7 mM
MgSO4, 0.5 mM CaCl2, 26 mM NaHCO3 and 10
mM glucose. Slices were stored in holding chambers containing normal ACSF consisting of 125 mM NaCl, 3 mM KCl, 1.25 mM NaH2PO4, 2
mM MgSO4, 1 mM CaCl2, 26 mM NaHCO3 and 10
mM glucose, bubbled with carbogen gas (95% O2 and 5% CO2).
Electrophysiological Recordings
Normal ACSF, previously described, plus tetrodotoxin (TTX; 1 M) was used during the μ recordings. TTX was added to block Na+
voltage-gated channels, to exclude the presence of action potentials in the slice. Whole-cell recordings of EPSPs and IPSPs from pyramidal cells were made using standard
electrophysiological techniques, including both
2
Figure 1. A: Exaggerated effect for visual clearity of disturbed balance on membrane potential: balance tipped towards mEPSPs (black) or tipped towards mIPSPs (red). Firing threshold is set at -60 mV. B: Average excitatory (black) and inhibitory (red) event. C: Average excitatory (black) and inhibitory (red) frequency.
Figure 2. Double electrode patch clamp on PFC
L2/3 L5
Exc. Inh. Exc. Inh.
≤P14 9 9 6 7
single and double electrode patch clamping (figure 2). On average from each slice a L2/3 and a L5 cell were patched. However, due to health of the slices this was not always possible. Only one cell per layer per slice was patched, before moving to the next slice, to ensure that a possible difference found could not be
contributed to one deviating slice. The patch pipette (3–5 MΩ) medium contained 120 mM cesium (Cs+) gluconate, 10 mM CsCl, 8 mM NaCl,
2 mM MgATP, 10 mM phosphocreatine, 0.2 mM EGTA, 10 mM HEPES, 0.3 mM Tris-GTP and 1 mM QX-314 Cl. All recordings were made at room temperature. Only recordings with a stable series resistance of <20 MΩ were
analyzed. Cs+-based pipette solution was used to
improve voltage-clamp recordings by blocking K+-channels. Cells were held in voltage clamp at
−70 mV for recording mEPSCs and at 0 mV for recording IPSCs. Actual set-up-specific voltage clamp was calculated as -58 mV and +12 mV respectively after taking junction potential of the used solutions into account. Both the EPSPs and the IPSPs recordings took on average 15 minutes.
Chemicals
Biocytin was obtained from Invitrogen; all other chemicals and reagents were obtained from Sigma Chemical.
Data analysis
Events were analysed using Mini Analysis (Synaptosoft, NJ) with event detection levels for synaptic currents set at 11 pA. An average of 100 (always 95-105) events was detected for each recording. Frequency per recording and amplitude, area, rise and decay per average recording was calculated, as were these four properties for the individual events. The statistical analysis was performed using Statistica 6.1 (Statsoft). The significance of differences between groups was determined using two-way ANOVA and Fisher’s test. Differences between group means were considered significant if P < 0.05.
Results
A total of 38 pyramidal cells were used for analysis, of which for a few cells only inhibitory events were recorded (n=5). Therefore, 33 cells have been analysed for both excitatory
glutamatergic events and inhibitory GABAergic events. For precise numbers for each subgroup, see figure 3. All cells were used in inter-group ANOVA and in analysis of individual events. However, only the double recorded (respectively -70 & 0 mV; n=33) PCs were included in the ratio test.
*
*
3
Figure 3. Number of cells measured for each subgroup (excitatory vs. inhibitory; ≤P14 vs. >8weeks; L2/3 vs. L5).
Differences in mEPSC and mIPSC frequency and ratio onto PC in L2/3 and L5 in
development
Our results show that excitatory and inhibitory synapses have different developmental patterns. We observed a layer specific increase of mEPSCs frequency over age. Specifically mature animals (n=9) have a higher frequency of excitatory events in L5 than young animals (n=6;
P=0.0088; figure 4). Additionally, the frequency
of excitatory events in mature animals is higher in L5 then in L2/3 (both subgroups n=9;
P=0.023). Compared with excitatory
frequencies, inhibition showed a less clear developmental and layer specific differentiation. Only a trend in elevated frequency of inhibitory events has been found in L2/3 in mature animals (n=10), when compared to young animals (n=9; P=0.069).
The ratio of inhibitory and excitatory events was calculated for each cell. First of all, in young mice the ratio of L2/3 versus L5 was significantly different (P=0.0492; figure 5A). Secondly, within L5: mature animals had significantly more various ratios then young
≤P14 >8weeks
≤P14 >8weeks
*
*
≤P14 >8weeks
≤P14
A.
C.
B.
D.
4
Figure 5. A: Frequencies of cells in young mice over layer (L2/3 vs. L5; P=0.0492) B: Frequencies of cells in mature mice over layer (L2/3 vs. L5) C: Frequencies of cells in L2/3 over age (≤P14 vs. >8weeks) D: Frequencies of cells in L5 over age (≤P14 vs. >8weeks; P=0.0112)
Figure 4. Frequencies over age (≤P14 vs. >8weeks) and layer (L2/3 vs. L5)
animals (P=0.0112, figure 5D). Lastly, in mature animals, a trend is found for the ratios to be different across the layers (P=0.082; figure 5B).
Differences in properties of PC in layer 2/3 and layer 5 over age
To understand whether the functional
properties of synapses change over age within different layers we assessed the amplitude and kinetics of mEPSCs and mIPSCs. Decay of excitatory events is slower in L2/3 than in L5 of mature animals (P=0.0120, figure 6A). Also decay of excitatory events is slower in L2/3 of mature animals than in the same layer of young animals (P=0.0230).
Mature animals have a slower inhibitory rise time in L5, when compared to L2/3
(average 1.27 and 0.85 ms respectively;
P=0.0257, figure 6B). All other groups had an
excitatory and inhibitory rise time of average 1.04 ms ± 0.0863 ms. Amplitude and area were not significantly different for any of these groups. Also, no difference was found for any of these properties in young mice.
In accordance with previous finding for PC of PFC (Brunel & Wang, 2013) for all age groups and layers amplitude and area were larger in inhibitory events than in excitatory events (respectively 20.6 ± 1.232, 485.6 ± 40.9 vs. 12.4 ± 0.640, 136.5 ± 11.4; P<0.001). Also, decay was faster for excitatory events (P<0.001) than for inhibitory events of all cells.
Discussion
Developmental differences in frequency
In young mice (≤P14) both excitatory and inhibitory frequencies are on average below 1 Hz, showing very little spontaneous mPSPs in pyramidal cells. However, in mature mice (>8weeks) mPSP-frequency showed more diversity. This is seen in the elevated EPSP frequency in L5 over age, and in the trend of elevated IPSP frequency in L2/3 over age. The underlying aspects of these significant
differences become more obvious, when considering not just the singular frequency, but when the E and I frequencies are combined into the ratio (figure 5). The clearly visible
differences in ratios show an increased
variation of cells of older animals compared to
cells of younger animals.
Changuex and Danchin (1973; 1976) were the first to suggest activity-dependent
stabilization and selective elimination of initially overproduced synapses, and their theory has been widely tested and accepted by now (see Petanjek et al., 2011). Our findings corresponds with earlier developmental papers, which indicate a basic ‘pool’ of cells when an organism is first born, and a later specialization and pruning of cells in development. Here we show, that the number of events resulting from spontaneous vesicle release of both glutamate and GABA is limited in young animals.
Additionally, we show that in mature animals the chance of spontaneous vesicle release is more common, but diverse for each cell (see figure 5). This indicates that the ‘basic pool of cell’ in young animals, develops into a more diverse group of cells.
Subsequently, by variation of E/I balance of mPSCs, the proximity to the firing threshold of each cell is different (figure 1A). If the E/I balance of a cell is tipped more towards excitatory signals, the steady membrane potential will be closer to the firing threshold, while the opposite holds true for a balance tipped towards inhibitory signals. This change will respectively facilitate or impede the cells ability to have an action potential, resulting in a cell’s specialized function in the network.
An increase in frequency can have two main causes: an elevated number of released vesicles per synapse or an increased number of synapses present. To test which of these two is the main culprit; a staining of the synapses in the cell can be performed. If an equal number of synapses are present onto PC of each layer over age, the main cause is within the synapse itself.
Layer dependent differences in frequency
Variation in frequency across layers would indicate differences in PC functions between L2/3 and L5 pyramidal cells. With these data we show that in mature animals, the frequency of excitatory events is higher in L5 than in L2/3. Additionally we show that in animals of ≤P14 and mature animals, the ratios of PCs of separate layers are different (significant in young, only trend in mature; figure 5). In young
5
Figure 6 A: Decay over age (≤P14 vs. >8weeks) and layer (L2/3 vs. L5) B: Rise over age (≤P14 vs. >8weeks) and layer (L2/3 vs. L5)
animals the E/I balance tips towards excitatory in L2/3, while the balance tips to inhibitory events in L5. This shows an early developmental change or possibly even a before-birth
disposition of separate PC groups within the layers. To establish at which moment in time this change occurs, further research should be conducted with more variation in age for animals under 15 days old.
In mature animals, a trend was found for the difference between ratios among layers. However, this difference is less clear than in young animals and only indicates a higher variance in frequencies of L5 cells (figure 5). By showing that each layer has its own subtype of synaptic composition on PCs, one can question: why do PCs of separate layers develop different properties? Previous research indicates that L2/3 PCs receive information from multiple sources. The pathway from L2/3 to L5 and its possible signal filtering properties has been previously described in motor cortex of mice (Weiler et al., 2008). If information indeed reaches L2/3 first, this layer’s main function could be to receive and merge all information inputs from the different brain regions.
In contrast, L5 is a deeper layer after which the information is redirected to other brain areas. Somewhere along this pathway, a filtering and amplification of respectively unimportant and important information could be present, to ensure the forwarding of a complete message from the PFC. As previously described in ‘developmental differences in frequency’: by increasing the chance of a mEPSP or mIPSP, the resting potential of a cell is respectively a bit more depolarized or hyperpolarized - resulting in a change to the proximity to the firing threshold of each cell (figure 1A). L5 seems to have a variable elevated chance of mPSPs, therefore cells that can amplify or facilitate (when elevated mEPSP-frequency) or diminish and impede (when elevated mIPSP-frequency) a signal, while PCs of L2/3 all have equal chances of PSPs occurring and are thus of equal ability to send a signal onwards.
Change of pyramidal cell properties
The occurrence of EPSPs and IPSPs is based on the stochastic spontaneous vesicle release
containing its corresponding neurotransmitter. Vesicle release from the axon terminal or ‘bouton’ into the synapse is regulated by a voltage-gated calcium channel. The vesicle fuses with the presynaptic membrane and
neurotransmitter molecules are released after an influx of Ca2+ into the presynaptic axon.
However, occasionally a vesicle will fuse without causal Ca2+-influx, such as an action potential in
the presynaptic cell, resulting in a spontaneous PSP on the postsynaptic cell. Since the chance of spontaneous vesicle release is small, these resulting events are often singular and can be used to research the properties of the receptor, which the neurotransmitter is binding to, and the kinetics of the corresponding channel.
Excitatory events are most often caused by the amino acid neurotransmitter glutamate binding to an ionotropic glutamate receptor (iGluR). Three iGluRs are known, of which the kianate receptor plays only a minor role at synapses (Song & Huganir, 2002). The attribution of the second type of iGluR, the NMDA receptor, is neglectable as it is restricted by a voltage-dependent Mg2+-block. Therefore
the only iGluR taken into account is the -α
amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor. The AMPA receptor is made up of 4 subunits, which form into a functional tetramer. The four
subunits – GluR1, GluR2, GluR3 and GluR4 – can form various compositions, e.g., in which some subunits are present twice, and another subunit is excluded. The composition of the tetramer influences properties of the AMPA receptor, such as amplitude and area of the
corresponding event.
Spontaneous inhibitory events are a result of the neurotransmitter GABA binding to an ionotropic receptor. These GABA-receptors are pentamers build of various subunits (e.g., , , , , , , , ) and different α β γ δ ε θ π ρ compositions occur. Again, this composition influences the kinetics of the channel, as can be seen in the amplitude and the area of the corresponding event.
Amplitude of an event can also be used to give an indication of the quantal size of
neurotransmitter present per vesicle. The higher the amount of neurotransmitter per
vesicle, resulting in more channels opening, the bigger the amplitude will be. In our data amplitude was found to be larger in inhibitory events, while decay was faster in excitatory events. This is in correspondence with earlier research on the difference between
glutamatergic and GABAergic receptors and their channels (Brunel & Wang, 2013). No differences within excitatory or inhibitory measurements in amplitude were found in either the different layers or over development. This could indicate a steady vesicle quantum size over layers and development.
Two other properties of the events
researched, suggest an interesting change of the post-synaptic channel of the receptor. First of all, the rise-time of inhibitory events was slower in L5 than in L2/3 in mature animals (figure 6B). Rise-time is influenced by the opening of the channel and the speed of the exchange of ions; slower than average rise-time suggest a change in kinetics of the GABA-channel, possibly by a change in the composition of the subunits. Secondly, decay of excitatory events was slower in L3 than in L5 in older animals (figure 6A) and excitatory events of mature animals within L3 PCs had a slower decay than excitatory events of young animals. Decay of an event is dependent on the closing of the channels and therefore on the uncoupling of the neurotransmitter, and slower decay suggests slower closing. This result suggest a change in AMPA-channel subunit composition of L3 cells in mature animals
Differences in maturity
For most electrophysiology research currently conducted in mice, animals of around P14 are used. These cells are more robust to survive the preparation process, and are more likely to result in a successful measurement than cells from older mice. Especially after the age of 8 weeks, cells become more prone to cell death during the patch clamp preparation process. By showing significant differences in firing
frequency and cell properties, an interesting developmental issue impedes the way of the standard laboratory protocols. Officially, at 8 weeks old, mice are sexually mature, and are therefore considered ‘mature’ in the whole sense of the word. However, so far unpublished data from the Mansvelder lab in Amsterdam (Rotaru, yet unpublished) shows that even more frequency and cell property changes in cortical structures take place after 8 weeks, suggesting that there is a big gap between ≤ P14, what an average researcher calls ‘mature’ and what is actually ‘fully developed’. This difference in sexual maturity versus PFC maturity is also found in humans. An average human becomes sexually mature during puberty between the ages 12 and 18. However, the PFC is not structurally mature until around age 25
(Dumontheil et al., 2010). These results should indicate a preference for the use of animals >8 weeks in experiments focused on reflecting ‘mature’ cells.
References
Barth AL & Poulet, JFA. (2012). Experimental
evidence for sparse firing in the neocortex. Trends in neurosciences, 35: 345–55.
Binzegger T, Douglas RJ & Martin KA.
(2004). A quantitative map of the circuit of cat primary visual cortex. J Neurosci 24:8441–8453. Brunel N. & Wang X. (2013).
What Determines the Frequency of Fast Network Oscillations With Irregular Neural Discharges ? Synaptic Dynamics and Excitation-Inhibition Balance. I:415–430.
Changeux JP, Courrège P, Danchin A. (1973). A
theory of the epigenesis of neuronal networks by selective stabilization of synapses. Proc Natl Acad Sci U S A. 70:2974–2978.
Changeux JP & Danchin A. (1976).
Selective stabilisation of developing synapses as a
mechanism for the specification of neuronal networks. Nature. 23:705–712.
Dumontheil I, Hassan B, Gilbert S &
Blakemore SJ. (2010). Development of the selection and manipulation of self-generated thoughts in adolescence.
Journal of Neuroscience. 30:7664-7671. Fuster JM. (2002). Frontal lobe and cognitive
development. J Neurocytol. 31:373–385 Gogtay N, Giedd JN, Lusk L, Hayashi KM,
Greenstein D, Vaituzis C, Thompson PM. (2004). Dynamic mapping of human cortical development during childhood through early adulthood. Proc Natl Acad Sci U S A. 101:8174–8179.
Kolb B, Mychasiuk R, Muhammad A, Li Y, Frost DO & Gibb R. (2012). Experience and the
developing prefrontal cortex. Proc Natl Acad Sci U S A. 109, Suppl 2:17186–17193.
Petanjek Z, Judaš M, Šimic G, Rasin MR, Uylings
HBM, Rakic P & Kostovic I. (2011). Extraordinary neoteny of synaptic spines in the human prefrontal cortex. Proc Natl Acad Sci U S A. 108:13281–13286. Rotaru DC, Lewis DA, Gonzalez-Burgos G.
(2007). Dopamine D1 receptor activation regulates sodium channel-dependent EPSP amplification in rat prefrontal cortex pyramidal neurons. J Physiol. 581:981-1000.
Selemon LD. (2013). A role for synaptic
plasticity in the adolescent development of executive function. Translational psychiatry, 3:238. Shepherd GMG, Stepanyants A, Bureau I,
Chklovskii DB & Svoboda K. (2005). Geometric and functional organization of cortical circuits. Nature Neuroscience. 8:782–790.
Song I & Huganir RL. (2002). Regulation of
AMPA receptors during synaptic plasticity. Trends in Neurosciences. 25:578–589.
Weiler N, Wood L, Yu J, Solla SA & Gordon MG. (2008). Nat Neurosci. 11:360–366. Yizhar O, Fenno LE, Prigge M, Schneider F,
Davidson TJ, O’Shea DJ & Deisseroth K. (2011). Neocortical excitation/inhibition balance in information processing and social dysfunction. Nature. 477:171–178.
