No Evidence for Early Decrease in Blood Oxygenation in Rat Whisker Cortex in Response to Functional Activation Ute Lindauer, Georg Royl, Christoph Leithner, Marc Ku ¨ hl, Lorenz Gold, Jo ¨rn Gethmann, Matthias Kohl-Bareis, Arno Villringer, and Ulrich Dirnagl Department of Experimental Neurology, Charite ´ Hospital, 10098 Berlin, Germany Received August 21, 2000; published online January 19, 2001 Using optical methods through a closed cranial win- dow over the rat primary sensory cortex in chloralose/ urethane-anesthetized rats we evaluated the time course of oxygen delivery and consumption in re- sponse to a physiological stimulus (whisker deflec- tion). Independent methodological approaches (opti- cal imaging spectroscopy, single fiber spectroscopy, oxygen-dependent phosphorescence quenching) were applied to different modes of whisker deflection (sin- gle whisker, full whisker pad). Spectroscopic data were evaluated using different algorithms (constant pathlength, differential pathlength correction). We found that whisker deflection is accompanied by a significant increase of oxygenated hemoglobin (oxy- Hb), followed by an undershoot. An early increase in deoxygenated hemoglobin (deoxy-Hb) proceeded hy- peroxygenation when spectroscopic data were ana- lyzed by constant pathlength analysis. However, cor- recting for the wavelength dependence of photon pathlength in brain tissue (differential pathlength correction) completely eliminated the increase in de- oxy-Hb. Oxygen-dependent phosphorescence quench- ing did not reproducibly detect early deoxygenation. Together with recent fMRI data, our results argue against significant early deoxygenation as a universal phenomenon in functionally activated mammalian brain. Interpreted with a diffusion-limited model of oxygen delivery to brain tissue our results are compat- ible with coupling between neuronal activity and ce- rebral blood flow throughout stimulation, as postu- lated 110 years ago by C. Roy and C. Sherrington (1890, J. Physiol. 11:85–108). © 2001 Academic Press INTRODUCTION In the brain neuronal activity, metabolism, and re- gional blood flow are tightly coupled in time and space. Changes in neuronal activity induce localized changes in the metabolism of glucose and oxygen (neurometa- bolic coupling), as well as blood flow (neurovascular coupling). Mapping of such dynamic changes form the basis of modern functional brain mapping (positron emission tomography, PET; functional magnetic reso- nance imaging, fMRI (Villringer and Dirnagl, 1995)). Due to its relatively high temporal (seconds) and spa- tial (approx. 100 m) resolution as well as its wide availability, fMRI has emerged as today’s most com- monly applied technique for functional neuroimaging. Blood oxygen level-dependent (BOLD)-fMRI uses the paramagnetic endogenous contrast agent deoxy-hemo- globin (deoxy-Hb) to detect a signal, which results from the complex interaction of stimulation-induced changes in oxygen metabolism (CMRO 2 ) and oxygen delivery (via changes in cerebral blood flow, CBF). Despite intense research efforts, a number of unre- solved issues regarding functional activation induced changes in oxygen consumption and delivery exist, which at present are center stage in the literature. In particular, it remains unclear whether neuronal acti- vation leads to early (2 s after stimulation onset) hemoglobin deoxygenation, just before (2 s after stim- ulation onset) the well-characterized hyperoxygen- ation sets in and peaks with higher amplitude. Several BOLD-fMRI studies have reported an early BOLD sig- nal decrease, which was interpreted as evidence for early deoxygenation (Ernst and Hennig, 1994; Poeggel et al., 1992; Menon et al., 1995; Hu et al., 1997), while other studies did not confirm this finding (Fransson et al., 1998; Mandeville et al., 1999a). Since the fMRI signal in principal can be influenced by many other factors beside the concentration of deoxy-Hb (see, e.g., Ernst and Hennig, 1994), great hope was based on alternative, more direct approaches to detect early de- oxygenation. Optical imaging (Malonek and Grinvald, 1996; Villringer and Chance, 1997) seems particularly suited for this task, since in its spectroscopic variant it can measure the concentration of oxy-Hb as well as deoxy-Hb of superficial brain areas. The optical method has been widely applied to functional neuroanatomy (Grinvald et al., 1986) and more recently directly to the investigation of functional activity related hemoglobin oxygenation changes. NeuroImage 13, 988 –1001 (2001) doi:10.1006/nimg.2000.0709, available online at http://www.idealibrary.com on 988 1053-8119/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.