1. Journal of Cerebral Blood Flow & Metabolism (2010), 1–8
& 2010 ISCBFM All rights reserved 0271-678X/10 $32.00
www.jcbfm.com
Feature Article
Functional white-laser imaging to study brain
oxygen uncoupling/recoupling in songbirds
Stephane Mottin1, Bruno Montcel2, Hugues Guillet de Chatellus3 and Stephane Ramstein1
´ ´
1
CNRS; Universite de Lyon; Universite de St-Etienne, UMR5516, Saint-Etienne, France; 2Universite de Lyon;
´ ´ ´
CREATIS-LRMN; CNRS UMR5220; INSERM U630; Universite Lyon 1; INSA Lyon, Villeurbanne, France; 3CNRS;
´
´
Universite Joseph Fourier; Laboratoire de Spectrome´trie Physique, UMR5588, St Martin d’He
`res, France
Contrary to the intense debate about brain oxygen dynamics and its uncoupling in mammals, very
little is known in birds. In zebra finches, picosecond optical tomography with a white laser and a
streak camera can measure in vivo oxyhemoglobin (HbO2) and deoxyhemoglobin (Hb) concentration
changes following physiologic stimulation (familiar calls and songs). Picosecond optical tomography
showed sufficient submicromolar sensitivity to resolve the fast changes in the hippocampus and
auditory forebrain areas with 250 lm resolution. The time course is composed of (1) an early 2-second-
long event with a significant decrease in Hb and HbO2 levels of À0.7 and À0.9 lmol/L, respectively,
(2) a subsequent increase in blood oxygen availability with a plateau of HbO2 ( + 0.3 lmol/L), and
(3) pronounced vasodilatation events immediately after the end of the stimulus. One of the findings
of our study is the direct link between blood oxygen level-dependent signals previously published in
birds and our results. Furthermore, the early vasoconstriction event and poststimulus ringing seem
to be more pronounced in birds than in mammals. These results in birds, tachymetabolic vertebrates
with a long lifespan, can potentially yield new insights, e.g., into brain aging.
Journal of Cerebral Blood Flow & Metabolism advance online publication, 20 October 2010; doi:10.1038/jcbfm.2010.189
Keywords: brain activation; cerebral hemodynamics; near-infrared spectroscopy; neurovascular coupling; optical
imaging; songbirds
Introduction suppressed by hypercapnia (Steinmeier et al, 1996;
Biswal et al, 1997; Fox and Raichle, 2007). When
Unlike other organs, the brain of mammals and birds activation occurs, step responses are complex
is a constant energy sink, consuming energy irre- (Kasischke et al, 2004; Niven and Laughlin, 2008;
spective of whether it is at rest or active, but on the Petzold et al, 2008; Vanzetta and Grinvald, 2008)
other side of the coin is its low tolerance to a long list and induce changes in blood flow and oxygen
of ‘perturbations’, such as hypoglycemia, hypoxia, transport (Ress et al, 2009). The coupling of perfu-
hypercapnia, hyperthermia, and mitochondrial dis- sion and oxidative metabolism in the resting brain
eases (Siesjo, 1978; Mottin et al, 2003). The coupling
¨ has been shown to be disrupted in the first minute
between transport and metabolism allows the bal- after the onset of a sudden functional challenge
ance between the storage and the production of (Kasischke et al, 2004; Niven and Laughlin, 2008;
adenosine triphosphate, despite the very high rate Petzold et al, 2008; Vanzetta and Grinvald, 2008).
of combustion of glucose with B5.5 dioxygen mole- This uncoupling has also been a major problem for
cules per glucose molecule (Siesjo, 1978; Sokoloff,
¨ the interpretation of brain imaging (Vanzetta and
2001). This steady state is an oscillatory regime that Grinvald, 2008). To better understand the link
is poorly understood, and low frequencies have been between tachymetabolism and this uncoupling, we
observed in various cerebral parameters reversibly developed a method for measuring the full time
course of oxygen transport in the higher-order
auditory region of the telencephalon in the zebra
´
Correspondence: Dr B Montcel, Universite Lyon 1, CREATIS-LRMN, finch, Taeniopygia guttata. The need for an im-
4 Rue V, Grignard, Villeurbanne 69616, France. proved understanding of the mechanisms underlying
E-mail: bruno.montcel@univ-lyon1.fr brain activation, especially in songbirds, has become
These experiments were supported by the Program ‘Emergence’
´ ˆ
obvious (Voss et al, 2007; Boumans et al, 2007).
of the Region Rhone-Alpes and the Agence Nationale de la
Recherche. Following the growing evidence of mammalian-like
Received 21 June 2010; revised 13 September 2010; accepted 14 cognitive abilities in songbirds (Vignal et al, 2004), of
September 2010 the neocortex-like functions of the avian pallium

2. Brain oxygen uncoupling/recoupling
S Mottin et al
2
(Jarvis et al, 2005; Reiner, 2005), and that of some neuroscience results and to measure for the first time
peculiarities of the avian metabolism (Barja and the full time course of coupling/uncoupling in small
Herrero, 1998; Turner et al, 2005; Moe et al, 2009), songbirds.
birds have become an important focus of interest
for comparative neuroscience (Vignal et al, 2004;
Clayton, 2007). Materials and methods
The reliability of optical measurements of changes Animals and Stimulation Protocols
in the concentration of hemoglobin in tissues has
been a challenge for several years (Vanzetta and Adult male zebra finches (T. guttata) served as subjects
Grinvald, 2008; Calderon-Arnulphi et al, 2009). for the experiments. Bred in the aviary of the University
Owing to its noninvasive nature, transcranial optical of Saint-Etienne in a 12-light/12-dark photoperiod, they
cerebral oximetry (near infrared spectroscopy, dif- received normal tutoring by adult males. Four and five
fuse optical tomography, etc.) has become a source of birds were used for spectral-POT and spatial-POT, respec-
quantitative or semiquantitative information about tively. Birds were anesthetized with 2% isoflurane under
brain oxygenation, cerebral blood flow, and volume. spontaneous breathing conditions (Vignal et al, 2008).
However, continuing technical controversies about Animal preparation and the spectral-POT setup have
signal derivation, accuracy, precision, and quantita- been described previously (Vignal et al, 2008). Anesthe-
tive ability have limited the application of transcra- tized zebra finches with the head previously plucked
nial optical cerebral oximetry. Clearly, transcranial (3 days before experiments) were fixed in a stereotaxic
optical cerebral oximetry still needs developments. frame (Stoelting Co., Wood Dale, IL, USA, adaptations for
Recent ultrafast technological advancements have birds). Body temperature was kept within a narrow range
opened up a new promising avenue in neuroscience (391C to 401C) using a feedback-controlled heating pad.
(Gibson et al, 2006; Vignal et al, 2008; Montcel et al, Optical fibers were fixed into stereotaxic manipulators
2005, 2006; Pifferi et al, 2008; Liebert et al, 2004; (Stoelting Co.) and placed directly on the skin. Positions of
Selb et al, 2006). the input optical fiber F1 providing illumination and those
As part of our broader effort to develop a of the optical fiber F2 collecting transmitted light were
noninvasive neuro-method and to improve quantita- chosen to probe the auditory regions of the telencephalon
tive measurement of absorbing chromophores into (field L, NCM (caudomedial nidopallium), and CMM
scattering brain tissues, we worked on a time (caudomedial mesopallium)). The precise anatomic locali-
domain-based device. Using a white-light super- zation of the optical fiber (Figure 1 in Ramstein et al, 2005)
continuum or ‘white laser’ (Chin et al, 1999), we and probed region (Figure 1 in Vignal et al, 2008) has been
combined picosecond optical tomography (POT) with described in previous studies. The head of the bird was
near-infrared spectroscopy (spectral-POT) (Vignal turned until the beak (rostral extremity) was perpendicular
et al, 2008) and a new POT with contact-free spatial to the body plane. This position allowed us to define a
imaging (spatial-POT). In the near-infrared spectral stereotaxic origin point (0, 0, 0) defined by the intersection
window of 650 to 850 nm, the nonmonotonic of the vertical plane passing through the interaural line and
behavior of the absorption spectrum of deoxyhemo- the sagittal suture (the vena cerebralis dorsocaudalis). The
globin (Hb) provides reliable ‘molecular fingerprints’ stereotaxic axes were chosen according to this origin point.
(Vignal et al, 2008). Furthermore, optical signals are F1 was placed more rostrally on the right hemisphere than
integrated into a selected picosecond time-of-flight F2. The distance between F1 and F2 was 5 mm (Figure 1).
window specifically defined so as to probe only the The chosen coordinates in millimeters were: F1 (2.0, 5.4,
targeted deep brain structures (Vignal et al, 2008). À2.7) and F2 (2.0, 0.4, À0.3).
This system allows us to monitor in vivo and Animals were kept in a custom-made sound-attenuated
quantify an evoked brain hemodynamic response box (48 Â 53 Â 70 cm3) equipped with 2 fixed high-fidelity
with submicromolar sensitivity and submillimeter speakers (Triangle Comete 202, Triangle SAS, Villeneuve
spatial resolution. The spatial-POT is different from Saint Germain, France). After a 1-minute baseline period,
the classic strategy of several discrete detectors each bird was subjected to a 20-second stimulus, followed
(Gibson et al, 2006). This configuration is without by 1 minute for recovery of baseline. The original auditory
contact between the skin and detectors. The position signal was a random sequence of songs and calls recorded
of this imaged segment on the head of the bird can be in the zebra finch aviary, normalized to the same intensity.
controlled by eye by shining the intermediate slit Among the 20 seconds of stimulus recorded, 94% repre-
using a He-Ne laser and by checking and adjusting sented songs and calls, whereas 6% represented silence.
the position of its image on the skin. In the case of For each animal, 15 stimuli were used, with 9 random
such small animals, this imaging system allows the white-noise stimuli. After experiments, all animals were
analysis of the resolution limits of POT. kept in the recording room for 24 hours for physiologic
Having developed a spectral-POT, we have pre- and behavioral verifications. All experimental proce-
viously been able to measure oxyhemoglobin (HbO2) dures were approved by the University’s animal care
and Hb changes following hypercapnia (Vignal et al, committee. Statistical methods have been described previ-
2008). We now address the task of mapping the ously (Vignal et al, 2008) (multiple comparison procedure,
acoustic field with the best possible spatial resolu- one-way ANOVAs (analyses of variance) for repeated
tion to show that POT is able to reproduce classic measures, Tukey–Kramer test, Statistics toolbox, Matlab,
Journal of Cerebral Blood Flow & Metabolism (2010), 1–8

3. Brain oxygen uncoupling/recoupling
S Mottin et al
3
Figure 1 The imaging setup based on two lenses (L4 and L1) conjugates the surface of the skull with the plane of the slit of the
streak camera. An afocal system made of two lenses (L2 and L3) is placed between the two imaging lenses. The slit (F) is placed at
the focal point of the afocal system. The position of the 5-mm-long segment on the head of the bird can be controlled by eye, by
shining a He-Ne laser through the intermediate slit and adjusting the position of its image on the skull (with mirror M). A narrow
bandwidth filter (IF) is placed just before the streak camera. The 5-mm-long segment is located 5 mm away from the white-laser
input optical fiber (F1).
The Mathworks, Natick, MA, USA). The variation of the long segment located 5 mm apart from the fiber (Figure 1).
time-resolved transmittance spectrum was also fitted to We put a narrow bandwidth filter (IF) (10 nm full-width
the spectra of HbO2 and Hb known in mammals by classic half-maximum) centered at 700 nm, at which the difference
linear least-squares procedure. The same procedure was of absorption between the two hemoglobin species was
applied to calculate variations in the concentration of maximal. The position of the imaged segment on the head
HbO2 and Hb. These concentration variations could be of the bird could be controlled by eye by shining the
expressed using an absolute scale (mmol) because our time- intermediate slit with a He-Ne laser and by checking and
resolved detection system could measure the mean optical adjusting the position of its image on the surface of the
path through the bird’s head owing to the mean arrival time head (mirror M). The spatial resolution along the slit was
of photons (Vignal et al, 2008). determined by imaging a white sheet of paper half covered
with black ink. The image consisted in the response of the
system to a Heaviside step and characterized the spatial
Experimental Setup of Spatial-Picosecond Optical resolution of the setup. Its resolution along the slit was
Tomography nearly 250 mm.
The single shot streak camera measured the propagation
We used the same setup and the same laser fiber position as time of the photons through tissues. All measurements
described previously (Vignal et al, 2008), with the omis- were carefully corrected from shading effects. Each
sion of the polychromator and with an imaging system measurement consisted in a frame integrating 33 laser
between the head of the animal and the streak camera pulses. The 5-mm-long segment (of 150 mm thickness) was
(Hamamatsu Streakscope C4334, Hamamatsu, Bridgewater, imaged onto the slit of the streak camera and at the end,
NJ, USA). The imaging setup based on two lenses (L4 and transformed to 640 pixels. The 2.1-nanosecond deflection
L1) (100 mm focal length) optically conjugated the surface time was converted to 480 pixels. The temporal resolution
of the skull with the plane of the entrance slit of the streak of the system was set by the temporal width of the trace
camera (Figure 1). An afocal system composed of two on the CCD camera of a femtosecond pulse. Instrument
lenses (L2 and L3) (100 mm focal) was placed between the response function was obtained by directly sending a
two imaging lenses, and a slit (F) was put at the focal point leaking of a femtosecond laser pulse. Owing to the jitter
of the afocal system. This slit was optically conjugated (with 33 laser pulses), the resolution was 25 picoseconds
with the entrance slit of the streak camera, resulting in a (6 pixels).
great simplification of the alignment procedure and in the An advantage of this setup comes from the possibility of
possibility of controlling the intensity of the light by contact-free measurements. The versatility of the optical
narrowing the slit as well. The magnification of the setup design we implemented has another interesting advantage
was 1. Intrinsic filtering properties of the imaging setup in terms of imaging. In the near future, by simply tilting
enabled to collect only the photons emerging from a 5 mm- the M mirror, we will sweep the imaging segment. This
Journal of Cerebral Blood Flow & Metabolism (2010), 1–8

4. Brain oxygen uncoupling/recoupling
S Mottin et al
4
possibility is way more flexible than fiber bundles and transmittance were equivalent for spatial-POT and
leads to a narrower spatial resolution, not limited by the spectral-POT. The maximum change in transmit-
diameter of the optical fibers, but only by the numerical tance induced by the auditory stimulus was 1.03. To
aperture and the properties of the optical setup. establish a calibrated functional technique, acoustic
response experiments were carried out under the
same conditions as the 7% normoxic hypercapnic
Results experiments (Vignal et al, 2008). The functional
signal under these conditions was found to be
Figure 2 shows the full time courses of picosecond equivalent to 10% of the hypercapnic changes
time-resolved transmittance measured by spectral- (Vignal et al, 2008). Our results showed that the
POT (Figure 2A) and by spatial-POT (Figure 2B). The dynamics of these physiologic changes required at
shape of the time courses and the level of variation of least a 2-second time resolution (Figures 2 and 3).
Significant Hb and HbO2 changes were obtained
by linear unmixing (Vignal et al, 2008) and were
840 analyzed with a 0.667-second time resolution. Dur-
ing the 2 seconds following the onset of acoustic
stimuli, Hb and HbO2 levels significantly decreased
to À0.7 and À0.9 mmol/L, respectively (Figure 3). The
HbO2 level then increased significantly (during
Wavelength (nm)
Transmittance 12.4 seconds, 100 concentration measurements) to
reach a plateau of 0.3 mmol/L (P = 0.015 when
compared with the 100 concentrations preceding
the stimulus). Immediately after the end of the
stimulus, Hb and HbO2 pulses reached + 0.7 mmol/L.
Changes were significantly localized (Figures 2B
and 4A) above the auditory forebrain areas (such as
the NCM, field L, CMM). A small contribution could
have derived from the hippocampus (dorsal and
672
0 silence 20 stimulus 40 silence 60
Time (s) *
1 **
**
*
** * * * *
Hb (microMole/L)
Left
-1
0
Transmittance
Length (mm)
0
***
-1
Silence Stimulus Silence
1 0 20 40 60
Time (s)
2
** *
Right
HbO2 (microMole/L)
1
3
0 silence 20 stimulus 40 silence 60 0
Time (s)
Figure 2 Time course of the picosecond time-resolved transmit- -1
*
tance. (A) The full time course of the picosecond time-resolved *
transmittance spectra was measured by spectral-POT. The
near-infrared spectral window is 668 to 844.4 nm, with Figure 3 Hb and HbO2 concentration changes. (A) Hb and (B)
20 spectral windows of 8.83 nm. (B) The time course of the HbO2 concentration changes obtained by linear unmixing of the
picosecond time-resolved transmittance for 20 spatial regions picosecond time-resolved transmittance spectra. Each point is
of 0.25 mm was imaged by spatial-POT. The 695 to 705 nm an average of five concentrations along the time axis, allowing a
spectral window was used for spatial-POT. Each point of time resolution of 0.667 seconds. Bars corresponds to P = 0.05
measurement corresponds to 33 milliseconds. For illustration for multiple comparisons (one-way ANOVAs for repeated
purposes, these results were filtered to get rid of high-frequency measures) between periods. The limits of significance of Hb
noise, using a Chebyshev window only along the time axis, and HbO2 are 0.42 and 0.75 mmol/L, respectively. The asterisk
1 second for spectral-POT and 2 seconds for spatial-POT. (*) indicates significant difference (P < 0.05) from the detection
The 0-mm position corresponds to the sagittal midline. POT, limit. ANOVA, analysis of variance; Hb, deoxyhemoglobin;
picosecond optical tomography. HbO2, oxyhemoglobin.
Journal of Cerebral Blood Flow & Metabolism (2010), 1–8

5. Brain oxygen uncoupling/recoupling
S Mottin et al
5
1.02 *
§ **
§
§ 1.01
1
ns
1.01
0.99
ns
ns
ns
ns
ns
Transmittance
ns
* *
1
1.01
# # # # 1
Transmittance
0.99
0.99 *
* * * * *
* * * *
1.01
0.98 1
3 1.5 0 -1.5
0.99
Length (mm)
* *
Figure 4 Spatial-POT transmittance changes. Spatial-POT
averaged 20-second-long transmittance changes for (A) the 1.01
stimulus (20 to 40 seconds) period and for (B) the poststimulus 1
(40 to 60 seconds) period. The transmittance of the rest period 0.99
(on 0 to 20 seconds) is normalized to 1 in all regions. Bars *
corresponds to P = 0.01 for multiple comparisons (one-way * * * * * *
ANOVAs for repeated measures) between the rest period and the Silence Stimulus Silence
two respective periods. Points without the NS (nonsignificant) 0 20 40 60
symbol indicate areas with significant changes when compared Time (s)
with the rest period (P < 0.01). The symbols for stimulus period
Figure 5 Time course of spatial-POT transmittance. Time course
and for the poststimulus period are y and # (P < 0.01) for the
of transmittance measured by spatial-POT at four positions
left hemisphere and right hemisphere, respectively, compared
(A: left hemisphere, position À1.25 mm; B: sagittal midline,
with more lateral positions (2.75 to 3 mm). The position 0 mm
position 0 mm; C: right hemisphere, position + 1.25 mm; and
corresponds to the sagittal midline and positions (0.25 to
D: right hemisphere position + 1.75 mm). Each bar (P = 0.01)
1.25 mm and À0.25 to À1.25 mm) to the auditory-hippocam-
allows multiple comparisons (one-way ANOVAs for repeated
pal areas. ANOVA, analysis of variance; POT, picosecond optical
measures) between the prestimulus period and the two others
tomography.
periods (the symbol * indicates P < 0.01). ANOVA, analysis of
variance; POT, picosecond optical tomography.
posterior areas) (Vignal et al, 2008). Furthermore, other methods (Clayton, 2007; Voss et al, 2007;
there was a significant bilateral increase in transmit- Poirier et al, 2009). In addition to bilateral responses
tance (Figure 4A) when compared with more lateral in these areas, lateralized activation has been
positions (2.75 and 3 mm). During the poststimulus suggested to take place (Voss et al, 2007; Poirier
period (Figure 4B), all areas showed significant et al, 2009). With our less specific stimulation para-
decreases in transmittance when compared with digm, we obtained a bilateral response without
the rest period. These results show that recoupling lateralization (Figures 2B and 4). Unlike electrophy-
was less localized than uncoupling. Compared with siology and immediate early gene expression, func-
more lateral positions and within the stimulation tional magnetic resonance imaging and POT measure
period, a significant bilateral increase was observed. blood dynamics in relation to the activity of large
In contrast, a significant bilateral decrease was clusters of cells. In birds, links between blood
observed within the poststimulus period. Further- oxygen level-dependent (BOLD), hemodynamics,
more, the number of Hb and HbO2 pulses was less and neuronal activation was not previously known
high for the auditory-hippocampal areas (0.25 to (Boumans et al, 2007; Voss et al, 2007; Vignal et al,
1.25 mm and À0.25 to À1.25 mm, respectively) than 2008). By choosing anesthetic conditions, sequences
for the more lateral positions, showing that recou- of stimuli and spatiotemporal parameters similar to
pling was faster in these areas (Figures 2B and 5). other published BOLD experiments (Table 1), we
sought to establish a robust correlation between the
BOLD signal and changes in Hb and HbO2 concen-
Discussion trations.
The diffuse optical method has the potential to
Our study shows for the first time the changes of differentiate hemoglobin dynamics; however, they
blood oxygen in a small songbird during stimulation have limited spatial resolution. Conversely, BOLD
in vivo. The most intense responses to similar stimuli functional magnetic resonance imaging has achieved
have been observed in the NCM and field L, using high spatial resolution but is more susceptible to
Journal of Cerebral Blood Flow & Metabolism (2010), 1–8

6. Brain oxygen uncoupling/recoupling
S Mottin et al
6
Table 1 Comparisons between BOLD spatiotemporal parameters and POT parameters
Male zebra finch Male zebra finch Male starling Male zebra
(Voss et al, 2007) (Boumans et al, 2007) (Van Meir et al, 2005) finch (POT)
Time resolution 4 seconds Between 3 and 5 seconds 0.033 seconds for DTr
6 seconds 0.666 seconds for DC
Time stimulus paradigm 32 seconds 40 seconds 30 seconds ON–30 seconds OFF 20 seconds
ON–32 seconds OFF ON–40 seconds OFF and 60 seconds ON–60 seconds OFF ON–60 seconds OFF
Thickness of slice 1 mm 0.5 mm 0.7–0.8 mm 0.25 mm
Number of slices 8 sagittal slices 1 Tilted coronal slice 2 sagittal slices 20 sagittal slices
BOLD, blood oxygen level dependent; POT, picosecond optical tomography.
BOLD spatiotemporal parameters in male zebra finches (Voss et al, 2007; Boumans et al, 2007) and male starlings (Van Meir et al, 2005) are shown. For POT
parameters (DTr, transmittance changes; DC, concentration changes of hemoglobin), the thickness of the ‘diffuse slice’ is measured at the surface of the scalp.
1 Mole/L hematological features of tachymetabolic vertebrates
appear to have converged to an equivalent oxygen
supply. In mammals, the coupling of blood trans-
HbO2 Concentration changes
-(Hb) Concentration changes
port and cerebral metabolic rates in physiologically
o o BOLD Signal Changes active brain states has been the subject of debate,
o o o and different theoretical models for it have been
o o
proposed (Banaji et al, 2008; Ress et al, 2009). In
0
humans, contrary to the ‘canonical hemodynamic
o response function’ (used by software packages,
o i.e., SPM; http://www.fil.ion.ucl.ac.uk/spm), BOLD
responses reveal several disparities (Aguirre et al,
1998), and some authors (Vanzetta and Grinvald,
2008; Ress et al, 2009) suggest that the differences
-1 Mole/L between BOLD responses are related to differences in
Stimulus Silence the properties of blood vessel networks between
mammals. Despite these differences, the BOLD over-
Figure 6 Typical time course of BOLD signal (dashed line) in
zebra finches is compared with the time course of HbO2 (curve shoot is always observed in the sensory systems of
with the symbol ‘o’) and Hb concentration changes measured by tachymetabolic vertebrates. Nevertheless, in the
POT in the caudal-medial areas. Minus Hb (ÀHb) concentration zebra finch, transient events seem to be more
is shown for better comparison with the BOLD signal. BOLD, pronounced than in comparable small rodents,
blood oxygen level dependent; HbO2, oxyhemoglobin; POT, perhaps because of differences in blood vessel
picosecond optical tomography. networks.
The biphasic changes in HbO2 (early decrease and
increase) measured by POT reveal a temporal pattern
limited ability to monitor hemoglobin dynamics. The similar to the biphasic response of tissue oxygen
BOLD time courses exhibit a sharp increase and (decrease and increase) in the auditory cortex
an overshoot at the beginning of the stimulus and (Masamoto et al, 2003) and to the biphasic response
an undershoot during the poststimulus period. The (initial constriction followed by dilation) of isolated
BOLD undershoot reveals two oscillations that have penetrating cerebral arterioles following an elevation
not been discussed previously, and these poststimu- of K + or adenosine triphosphate (Dietrich et al,
lus BOLD pulses are more pronounced in field L than 2009). For the poststimulus period, recoupling seems
in the NCM (Voss et al, 2007; Boumans et al, 2007). to be more complex than expected because (1) HbO2
As known in mammals, we showed (Figure 6) that in and Hb pulses were less localized than during
birds there was a direct link between the BOLD activation (Figure 2B), (2) the recoupling of the
signal and minus Hb (ÀHb). However, ÀHb exhibited activated auditory regions was faster than for other
a faster response to changes in stimulus (Figure 6), regions (Figure 5), and (3) the early HbO2 pulse
suggesting that the BOLD signal is a more ‘convo- arrived before the Hb pulse (Figure 3). Therefore,
luted response’ to hemodynamic than ÀHb. BOLD and POT recoupling pulses in birds seem to be
Birds and mammals have well-pronounced pial more spatiotemporally structured than nonlinear
arterial ramifications (Mc Hedlishvili and Kuridze, ‘passive elastic sloshing’, as expected (Ress et al,
1984). Birds have (1) bigger red blood cells by a factor 2009).
of 3 and a far larger capillary diameter, (2) less red We imaged the sinus sagittalis superior (position
blood cells per volume of blood by a factor of 1/3, 0 mm in Figure 2B), and no significant changes were
and (3) a hemoglobin concentration in red blood observed during the activation period. However, this
cells comparable with that of small rodents (Altman result should be considered carefully because distin-
and Dittmer, 1971). The pial angioarchitecture and guishing arterial, capillary, and venous compartments
Journal of Cerebral Blood Flow & Metabolism (2010), 1–8

7. Brain oxygen uncoupling/recoupling
S Mottin et al
7
is not straightforward in optical neuroimaging (Hillman fluctuations in the human motor cortex during rest
et al, 2007). using echo-planar MRI. J Cereb Blood Flow Metab 17:
Systems with a low ratio of energy storage to 301–8
energy consumption, end products, and heat genera- Boumans T, Theunissen FE, Poirier C, Van Der Linden A
(2007) Neural representation of spectral and temporal
tion have to respond ‘instantaneously’ to activation.
features of song in the auditory forebrain. Eur J Neurosci
The understanding of the time course of uncoupling/ 26:2613–26
recoupling and spontaneous oscillations (Steinmeier Calderon-Arnulphi M, Alaraj A, Slavin KV (2009) Near
et al, 1996; Biswal et al, 1997; Fox and Raichle, 2007; infrared technology in neuroscience: past, present and
Ress et al, 2009) emphasize the role of nonisochoric future. Neurol Res 31:605–14
processes in this response. Despite the constraints Chin SL, Petit S, Borne F, Miyazaki K (1999) The white
of an instantaneous response, the sensory systems light supercontinuum is indeed an ultrafast white light
of all tachymetabolic animals seem to be robust laser. Jpn J Appl Phys Part 2 Letters 38:L126–8
when faced with the ordinary perturbations they Clayton CF (2007) Molecular neurobiology of bird song. In:
were designed to handle, but fragile when faced Handbook of Neurochemistry and Molecular Neurobiol-
ogy (Lajtha A, ed), US: Springer, 373–417
with unexpected or strong perturbations (Niven and Dietrich HH, Horiuchi T, Chuanxi X, Hongo K, Falck JR,
Laughlin, 2008). Our approach shows the occurrence Dacey RG (2009) Mechanism of ATP-induced local and
of strong reactivity in the cerebral vessels of the bird, conducted vasomotor responses in isolated rat cerebral
an animal with a long lifespan (Barja and Herrero, penetrating arterioles. J Vasc Res 46:253–64
1998; Moe et al, 2009). Several studies (Mitschelen Fox MD, Raichle ME (2007) Spontaneous fluctuations in
et al, 2009) indicate that age-related changes in brain activity observed with functional magnetic reso-
vascular reactivity are important contributing factors nance imaging. Nat Rev Neurosci 8:700–11
to mild cognitive impairment in aging mammals. Gibson AP, Austin T, Everdell NL, Schweiger M, Arridge
Contrary to the accepted dogma, the role of oxidative SR, Meek JH, Wyatt JS, Delpy DT, Hebden JC (2006)
stress as a determinant of longevity is still open to Three-dimensional whole-head optical passive motor
evoked responses in the tomography of neonate. Neuro-
question (Mitschelen et al, 2009; Moe et al, 2009). image 30:521–8
Our results could thus shed light on this crucial Hillman EMC, Devor A, Bouchard MB, Dunn AK, Krauss
question, i.e., the link between brain aging and GW, Skoch J, Bacskai BJ, Dale AM, Boas DA (2007)
vascular reactivity. Depth-resolved optical imaging and microscopy of
vascular compartment dynamics during somatosensory
stimulation. Neuroimage 35:89–104
Jarvis ED, Gunturkun O, Bruce L, Csillag A, Karten H,
¨ ¨ ¨
Acknowledgements Kuenzel W, Medina L, Paxinos G, Perkel DJ, Shimizu T,
The authors thank Clementine Vignal and Nicolas Striedter G, Wild JM, Ball GF, Dugas-Ford J, Durand SE,
Mathevon for their technical participation and Hough SE, Husband S, Kubikova L, Lee DW, Mello CV,
Powers A, Siang C, Smulders TV, Wada K, White SA,
bioacoustic data. The authors also thank Colette Yamamoto K, Yu J, Reiner A, Butler AB (2005) Avian
Bouchut, Sabine Palle, and Pierre Laporte for brains and a new understanding of vertebrate brain
their help. evolution. Nat Rev Neurosci 6:151–9
Kasischke KA, Vishwasrao HD, Fisher PJ, Zipfel WR, Webb
WW (2004) Neural activity triggers neuronal oxidative
Disclosure/conflict of interest metabolism followed by astrocytic glycolysis. Science
305:99–103
The authors declare no conflict of interest. Liebert A, Wabnitz H, Steinbrink J, Obrig H, Moller M,
¨
Macdonald R, Villringer A, Rinneberg H (2004) Time-
resolved multidistance near-infrared spectroscopy of
References the adult head: intracerebral and extracerebral absorp-
tion changes from moments of distribution of times of
Aguirre GK, Zarahn E, D’Esposito M (1998) The variability flight of photons. Appl Opt 43:3037–47
of human, BOLD hemodynamic responses. Neuroimage Masamoto K, Omura T, Takizawa N, Kobayashi H, Katura T,
8:360–9 Maki A, Kawaguchi H, Tanishita K (2003) Biphasic
Altman PL, Dittmer DS (1971) Respiration and Circulation. changes in tissue partial pressure of oxygen closely
Bethesda, MD: Federation of American Societies of related to localized neural activity in guinea pig
Experimental Biology, 149–52 auditory cortex. J Cereb Blood Flow Metab 23:1075–84
Banaji M, Mallet A, Elwell CE, Nicholls P, Cooper CE Mc Hedlishvili G, Kuridze N (1984) The modular organiza-
(2008) A model of brain circulation and metabolism: tion of the pial arterial system in phylogeny. J Cereb
NIRS signal changes during physiological challenges. Blood Flow Metab 4:391–6
PLoS Comput Biol 4:e1000212 Mitschelen M, Garteiser P, Carnes BA, Farley JA, Doblas S,
Barja G, Herrero A (1998) Localization at complex I and Demoe JH, Warrington JP, Yan H, Nicolle MM, Towner
mechanism of the higher free radical production of R, Sonntag WE (2009) Basal and hypercapnia-altered
brain nonsynaptic mitochondria in the short-lived rat cerebrovascular perfusion predict mild cognitive im-
than in the longevous pigeon. J Bioenerg Biomembr pairment in aging rodents: an MRI study using fair and
30:235–43 bold imaging. Neuroscience 164:918–28
Biswal B, Hudetz AG, Yetkin FZ, Haughton VM, Hyde JS Moe B, Ronning B, Verhulst S, Bech C (2009) Metabolic
(1997) Hypercapnia reversibly suppresses low-frequency ageing in individual zebra finches. Biol Lett 5:86–9
Journal of Cerebral Blood Flow & Metabolism (2010), 1–8

8. Brain oxygen uncoupling/recoupling
S Mottin et al
8
Montcel B, Chabrier R, Poulet P (2005) Detection of cortical Siesjo BK (1978) Brain Energy Metabolism. Chichester, NY:
¨
activation using time-resolved diffuse optical methods. J. Wiley&Sons
Appl Opt 44:1942–7 Sokoloff L (2001) Historical review and developments
Montcel B, Chabrier R, Poulet P (2006) Time-resolved in the field of cerebral blood volume and metabolism.
absorption and haemoglobin concentration difference In: Ischemic Blood Flow in the Brain (Fukuuchi Y,
maps: a method to retrieve depth-related information on Tomita M, Koto A, eds), Tokyo, Japan: Springer, 3
cerebral hemodynamics. Opt Express 14:12271–87 Steinmeier R, Bauhuf C, Hubner U, Bauer RD, Fahlbusch R,
¨
Mottin S, Laporte P, Cespuglio R (2003) Inhibition of Laumer R, Bondar I (1996) Slow rhythmic oscillations of
NADH oxidation by chloramphenicol in the freely blood pressure, intracranial pressure, microcirculation,
moving rat measured by picosecond time-resolved and cerebral oxygenation—dynamic interrelation and
emission spectroscopy. J Neurochem 84:633–42 time course in humans. Stroke 27:2236–43
Niven JE, Laughlin SB (2008) Energy limitation as a Turner N, Else PL, Hulbert AJ (2005) An allometric
selective pressure on the evolution of sensory systems. comparison of microsomal membrane lipid composition
J Exp Biol 211:1792–804 and sodium pump molecular activity in the brain of
Petzold GC, Albeanu DF, Sato TF, Murthy VN (2008) mammals and birds. J Exp Biol 208:371–81
Coupling of neural activity to blood flow in olfactory Van Meir V, Boumans T, De Groof G, Van Audekerke J,
glomeruli is mediated by astrocytic pathways. Neuron Smolders A, Scheunders P, Sijbers J, Verhoye M,
58:897–910 Balthazart J, Van der Linden A (2005) Spatiotemporal
Pifferi A, Torricelli A, Spinelli L, Contini D, Cubeddu R, properties of the BOLD response in the songbirds’
Martelli F, Zaccanti G, Tosi A, Dalla Mora A, Zappa F, auditory circuit during a variety of listening tasks.
Cova S (2008) Time-resolved diffuse reflectance using Neuroimage 25:1242–55
small source-detector separation and fast single-photon Vanzetta I, Grinvald A (2008) Coupling between neuronal
gating. Phys Rew Lett 100:138101 activity and microcirculation: implications for func-
Poirier C, Boumans T, Verhoye M, Balthazart J, tional brain imaging. HFSP J 2:79–98
Van der Linden A (2009) Own-song recognition in the Vignal C, Boumans T, Montcel B, Ramstein S, Verhoye M,
songbird auditory pathway: selectivity and lateraliza- Van Audekerke J, Mathevon N, Van der Linden A,
tion. J Neurosci 29:2252–8 Mottin S (2008) Measuring brain hemodynamic changes
Ramstein S, Vignal C, Mathevon N, Mottin S (2005) In vivo in a songbird: responses to hypercapnia measured with
and noninvasive measurement of a songbird head’s functional MRI and near-infrared spectroscopy. Phys
optical properties. Appl Opt 44:6197–204 Med Biol 53:2457–70
Reiner A (2005) A new avian brain nomenclature: why, Vignal C, Mathevon N, Mottin S (2004) Audience drives
how and what. Brain Res Bull 66:317–31 male songbird response to partner’s voice. Nature
Ress D, Thompson JK, Rokers B, Khan RK, Huk AC (2009) 430:448–51
A model for transient oxygen delivery in cerebral cortex. Voss HU, Tabelow K, Polzehl J, Tchernichovski O, Maul KK,
Front Neuroenergetics 1:3 Salgado-Commissariat D, Ballon D, Helekar SA (2007)
Selb J, Joseph DK, Boas DA (2006) Time-gated optical Functional MRI of the zebra finch brain during song
system for depth-resolved functional brain imaging. stimulation suggests a lateralized response topography.
J Biomed Opt 11:044008 Proc Natl Acad Sci USA 104:10667–72
Journal of Cerebral Blood Flow & Metabolism (2010), 1–8