The occurrence and strength of vasoconstriction depends on the photostimulation frequency of pyramidal cells.

(a) Representative examples of the voltage responses of a layer II-III pyramidal cell (upper traces light grey to black traces) induced by photostimulations (470 nm, 10 s train, 5 ms pulses) delivered at 1, 2, 5, 10 and 20 Hz (cyan lower traces) and mean spike success rate (middle trace, n= 4 cells from 3 mice). The SEMs envelope the mean traces. The red dashed lines represent a spike success rate of 100%. (b) Representative example showing IR-DGC pictures of a layer I penetrating arteriole (1) before a 20 Hz photostimulation, (2) at the maximal diameter decrease, and (3) after 10 minutes of recording. Pial surface is upward. Yellow calipers represent the measured diameters. White dashed lines indicate the initial position of the vessel wall. Scale bar: 25 µm. (c) Kinetics of arteriolar diameter changes induced by photostimulation (vertical cyan bars) at 1 Hz (n= 4 arterioles from 3 mice), 2 Hz (n= 10 arterioles from 8 mice), 5 Hz (n= 6 arterioles from 6 mice), 10 Hz (n= 5 arterioles from 5 mice) and 20 Hz (n= 10 arteriole from 9 mice). The SEMs envelope the mean traces. The blue trace represents the kinetics of the diameter changes of the arteriole shown in (b). (d) Effects of the different photostimulation frequencies on AUC of vascular responses during 10 min of recording. Data are presented as the individual values and mean ± SEM. * statistically different from 20 Hz stimulation with p<0.05.

Photostimulation of pyramidal cells elicits a time-locked firing and a frequency-dependent calcium increase.

(a). Voltage response (top trace) and kinetics of relative fluorescence changes (red bottom trace) induced by photostimulation at 20 Hz. Insets, IR-DGC (top), Rhod2 fluorescence (bottom) pictures of an imaged layer II/III pyramidal cell. The somatic region of interest is outlined in white. Pial surface is upward. Scale bar: 20 μm. (b) Mean relative variations of Ca2+ fluorescence in response to photostimulation at 2 Hz (grey) and 20 Hz (black). Dashed line represents the baseline. The vertical cyan bar indicates the duration of photostimulation. SEMs envelope the mean traces. Inset, Maximum increase in relative fluorescence changes induced immediately after photostimulation, indicated by the black arrow. The data are shown as the individual values and mean ± SEM. * statistically different with p< 0.05.

Optogenetically-induced vasoconstriction requires AP firing and partially glutamatergic transmission.

Effect of TTX (1 μM, brown, n= 6 arterioles from 5 mice) and cocktail antagonists of AMPA/kainate (DNQX, 10 μM), NMDA (D-AP5, 50 μM), mGluR1 (LY367385, 100 μM) and mGluR5 (MPEP, 50 μM) receptors (gray, n= 10 arterioles from 6 mice) on (a) kinetics and (b) magnitude of arteriolar vasoconstriction induced by 20 Hz photostimulation (cyan bar). The SEMs envelope the mean traces. Dashed lines represent the initial diameter. The shaded traces correspond to the kinetics of arteriolar vasoconstriction in control condition (Fig. 1c – 20 Hz). Data are presented as the individual values and mean ± SEM. * and *** statistically different from control condition (Fig. 1c – 20 Hz) with p<0.05 and p<0.001, respectively.

Layer II-III pyramidal cells express PGE2 and PGF2α synthesizing enzymes.

(a) Voltage responses of a layer II-III pyramidal cell induced by injection of current (bottom traces). In response to a just-above-threshold current pulse, the neuron fired long-lasting action potentials with little frequency adaptation (middle black trace). Near saturation, it exhibits the pronounced spike amplitude accommodation and marked frequency adaptation characteristic of regular spiking cells (upper grey trace). (b) Agarose gel analysis of the scRT-PCR products of the pyramidal cell shown in (a) revealing expression of vGluT1, COX-2, mPGES2, cPGES, PM-PGFS and CBR1. Φx174 digested by HaeIII (Φ) was used as molecular weight marker (c) Histogram summarizing the single-cell detection rate of PGE2 and PGf2α synthesizing enzymes in layer II-III pyramidal cells (n= 16 cells from 6 mice). PGES (green zone) corresponds to mPGES1, mPGES2 and/or cPGES and PGFS (blue zone) to PM-PGFS, CBR1 and/or AKR1B3. (d) Co-expression of PGE2 and PGf2α synthesizing enzymes in pyramidal cells. The box size is proportional to the detection rate. Note the absence of co-expression between COX-1 (purple) and COX-2 (red). Co-expression of a PGES (left, green) and a PGFS (right, blue) with COX-1 (up) and COX-2 (bottom).

PGE2 mostly derived from COX-2 activity and its EP1 and EP3 receptors mediates vasoconstriction induced by optogenetically activated pyramidal cells.

(a, b) Ex vivo effects of the COX1/2 inhibitor indomethacin (magenta, n= 10 arterioles from 9 mice), the COX-1 inhibitor SC-560 (purple, n= 10 arterioles from 7 mice), and the COX-2 inhibitor NS-398 (red, n= 7 arterioles from 6 mice) on kinetics (a) and AUC (b) of arteriolar vasoconstriction induced by 20 Hz photostimulation (vertical cyan bar). (c) Optogenetic stimulation was achieved in vivo with an optic fiber through a chronic cranial window over the barrel cortex. (d) Left, diameter of pial arterioles labeled with fluorescein dextran (i.v) was measured with line-scan crossing the vessel (white line). Right, Representative examples of vascular response upon photostimulation (10 Hz, 10 s) under control (top) and indomethacin condition (bottom). (e) Diameter changes upon photostimulation under control (black; n = 5 arterioles, 4 mice) or indomethacin (magenta; n = 4 arterioles, 4 mice) conditions. (f) Area under the curve of the diameter change in control (black) or indomethacin (magenta) conditions calculated between 20 and 40 s (unpaired, two-tailed Mann Whitney test, * p<0.05). (g, h) Effects of the EP1, EP3 and FP antagonists, ONO-8130 (10 nM, dark green, n= 9 arterioles from 7 mice), L798,106 (1 μM, light green, n= 9 arterioles from 5 mice) and AL8810 (10 μM, dark blue, n= 9 arterioles from 7 mice), respectively, on kinetics (g) and AUC (h) of arteriolar vasoconstriction induced by 20 Hz photostimulation. The data are shown as the individual values and mean ± SEM. Dashed line represents the baseline. The SEMs envelope the mean traces. The data are shown as the individual values and mean ± SEM. The shaded traces correspond to the control condition (Fig. 1c – 20 Hz). *, ** and *** statistically different from 20 Hz control condition with p< 0.05, 0.01 and 0.001, respectively.

NPY Y1 receptors activation and 20-HETE synthesis mediates the vasoconstriction induced by pyramidal neurons.

Effects of paxilline (1 μM, orange, n= 10 arterioles from 6 mice), HET-0016 (100 nM, blue-grey, n= 10 arterioles from 7 mice) and BIBP3226 (1 μM, yellow, n= 10 arterioles from 6 mice) on (a) kinetics and (b) AUC of arteriolar vasoconstriction induced by 20 Hz photostimulation (vertical blue bar). Dashed line represents the baseline. The SEMs envelope the mean traces. The shaded traces correspond to the control condition (Fig. 1c – 20 Hz). The data are shown as the individual values and mean ± SEM. * and *** statistically different from 20 Hz control condition with p< 0.05 and 0.001.

Possible pathways of vasoconstriction induced by pyramidal neurons.

20 Hz photostimulation induces activation of pyramidal neurons expressing channelrhodopsin-2 (ChR2H134R) and increases intracellular calcium (Ca2+). Arachidonic acid (AA) is released from membrane phospholipids (MPL) by phospholipases (PL) activated by intracellular Ca2+ and is metabolized by type-1 and type-2 cyclooxygenases (COX-1 and COX-2) and prostaglandin E2 synthases (PGES) to produce prostaglandin E2 (PGE2). Three non-exclusive pathways can be proposed for arteriolar vasoconstriction in layer I: 1) PGE2 released into the extracellular space may act directly on arteriolar EP1 and EP3 receptors to induce smooth muscle cell constriction. 2) Glutamate released from pyramidal cells may activate neuropeptide Y (NPY) interneurons and NPY is released to act on Y1 receptors to constrict smooth muscle cells. Glutamate can also activate astrocytes to induce constriction through the 20-HETE and the COX-1/PGE2 pathways. 3) PGE2 may act on pre- and postsynaptic EP2 receptors to facilitate glutamate release and NPY interneuron activation, respectively.