Functional Imaging of Cerebral Oxygenation with Intrinsic Optical Contrast and Phosphorescent Probes [chapter]

Anna Devor, Sava Sakadžić, Mohammad A. Yaseen, Emmanuel Roussakis, Peifang Tian, Hamutal Slovin, Ivo Vanzetta, Ivan Teng, Payam A. Saisan, Louise E. Sinks, Anders M. Dale, Sergei A. Vinogradov (+1 others)
2013 Neuromethods  
Microscopic in vivo measurements of cerebral oxygenation are of key importance for understanding normal cerebral energy metabolism and its dysregulation in a wide range of clinical conditions. Relevant cerebral pathologies include compromised blood perfusion following stroke and a decrease in efficiency of single-cell respiratory processes that occurs in neurodegenerative diseases such as Alzheimer's and Parkinson's disease. In this chapter we review a number of quantitative optical approaches
more » ... optical approaches to measuring oxygenation of blood and cerebral tissue. These methods can be applied to map the hemodynamic response and study neurovascular and neurometabolic coupling, and can provide microscopic imaging of biomarkers in animal models of human disease, which would be useful for screening potential therapeutic approaches. Key words O 2 sensing, Phosphorescence quenching, Intrinsic optical signals, Energy metabolism, In vivo imaging, Hemoglobin, Two-photon microscopy, CCD loading of extrinsic O 2 -sensitive probes. Following the original demonstration [1], the intrinsic imaging method was widely used for investigation of neuro-hemodynamic coupling [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] and mapping of cortical neuronal responses [12] [13] [14] [15] (Box 1). The hemoglobin molecule is composed of four monomers, each containing an O 2 -binding heme group. Hemoglobin tetramer Box 1 Optical Imaging of Intrinsic Signals-A Historical Perspective One hundred and twenty years ago, Roy and Sherrington [16] argued that neuronal activation causes the local vasculature to respond. While unequivocal confirmation of this claim had to wait nearly a century, until radioactive methods became available [17] [18] [19] [20], large reflectance changes of brain tissue during localized seizure activity could be visualized already in the late 1930s [21]. A few decades later advances made it possible to detect and analyze the much smaller optical signals during activity of the normal cortex. These were accounted for by activity-associated changes in cerebral blood flow (CBF) and volume (CBV) [18, 22]; in addition, Chance [23] and Jöbsis [22] observed that neuronal activity is often accompanied by oximetric signals that can be detected optically by monitoring the absorption (and/or fluorescence) of hemoglobin and other intrinsic chromophores. In the late 1980s, Grinvald et al. [1] showed that the small light absorption changes induced by these activity-evoked hemodynamic responses can be used to explore cortical functional architecture in vivo, by using a CCD camera to image the cortex upon illumination at specific wavelengths during the presentation of sensory stimuli. The resulting cortical images can then be used to produce functional maps at the spatial resolution of a few tens of microns, more than enough to image the columnar structure of the mammalian neocortex. Since then, the interpretation of intrinsic signals in terms of neuronal activity-and thus their utility for functional brain mapping-has been tightly linked to our understanding of the mechanisms underlying neurovascular coupling. In fact, such knowledge is necessary to distinguish between optical signals resulting from changes in the activity of local neuronal populations and those resulting from the vascular effects of remote neuronal events (e.g., venous drain). To address this issue, Frostig et al. [4] imaged the sensory-evoked optical responses in the cortex at several wavelengths and tried to decompose them spectroscopically into CBV and O 2 saturation. A few years later, Malonek and Grinvald [6] used a continuous wavelength spectroscopic approach, optical imaging spectroscopy, to investigate the hemodynamic response in further detail. Both studies concluded that some hemodynamic events colocalize better with neuronal activity than others. In particular, an early increase in deoxy-hemoglobin concentration (Hb)-the so-called initial dip-was detected, and interpreted as resulting from local O 2 consumption induced by neuronal activity while the vasculature is still at rest. Thus, the initial dip is expected to colocalize more accurately with neuronal activity than the subsequent hemodynamic events, i.e., increased CBV and CBF, which are mediated by the complex spatiotemporal transfer function of the active vascular response [11, 24]. Which hemodynamic response component is best for functional mapping and how to optimally choose the imaging parameters (wavelength, timing of data acquisition, etc.) are therefore still the object of some debate [8, [25][26][27][28][29][30][31]. Whatever its conclusion, de facto, imaging the activity of visual stimuli in early visual cortex using oximetric wavelengths (600-630 nm, emphasizing changes in [Hb] over those in [HbO]) has allowed investigators to obtain high-quality single-condition functional maps, i.e., maps obtained by comparing stimulated conditions to rest. This has not been possible using isosbestic wavelengths, where Hb and HbO absorb with equal strength and which thus emphasize changes in CBV. At those wavelengths, differential approaches (the comparison of several stimulated conditions one to another) are needed to visualize functional architecture at the columnar level [4, 24], at least upon stimulation with large visual stimuli. Interestingly, in other cortices such as auditory [32][33][34] and rat somatosensory cortex [26], CBV-based optical signals have turned out to warrant at least as good mapping as oximetric ones. Moreover, in primate visual cortex [30] it has been shown that, upon stimulation with small (continued)
doi:10.1007/978-1-62703-785-3_14 fatcat:auqaseofhndrte2usnqphxndhq