Imaging brain activity in conscious animals using functional MRI

https://doi.org/10.1016/S0165-0270(98)00037-5Get rights and content

Abstract

Functional magnetic resonance imaging (fMRI) in humans has helped improve our understanding of the neuroanatomical organization of behavior. Unfortunately, fMRI in animal studies has not kept pace with the human work. Experiments are limited because animals must be anesthetized to prevent motion artifacts, precluding most studies involving neuroimaging of brain activity during behavior. The present study tested a newly developed head and body holder for performing fMRI in fully conscious animals. Significant changes in signal intensities were observed in the somatosensory cortex of conscious rats in response to electrical shock of the hindpaw. These changes in evoked signal ranged between 4 and 19% and were accompanied by significant increases in local cerebral blood flow. The fMRI study was performed with a 2.0-Tesla spectrometer. Using this non-invasive method of imaging brain activity in conscious animals, it is now possible to perform developmental studies in animal models of neurological and psychiatric disorders.

Introduction

Functional magnetic resonance imaging (fMRI) is a non-invasive procedure for studying the localization of brain activity. Functional MRI has greater spatial and temporal resolution than positron emission tomography and single photon emission computerized tomography and is much more convenient because it does not require the production of radioactive molecules (Neil, 1993). In addition, with the use of positron emitters the number of possible studies and serial repetitions within the same subject over a finite time period are restricted by the exposure limit to ionizing radiation. However, with fMRI the signal is not generated by radioactivity; instead hydrogen nuclei give off radiofrequency (RF) signals following RF excitation of the sample. With this non-invasive technology, subjects can be repeatedly studied within a single session and over time. This has lead to a rapid increase in fMRI methodology for use in neuroscience research pertaining to functional brain mapping in humans.

However, animal work has not kept pace with the human studies. This is unfortunate because fMRI in preclinical research would allow the study of developmental changes in behavior in animal models of neurological and psychiatric diseases that could not be done in humans. The use of fMRI in animal research has been restricted because of technical problems associated with the movement of the animal in the MR spectrometer. Any head movement not only distorts the image, but may also create a change in signal intensity that can be mistaken for stimulus-associated changes in brain activity (Hajnal et al., 1994). In addition to head movement, it has been observed that any motion outside the field-of-view can also obscure or mimic the signal from neuronal activation (Birm et al., 1996). Consequently, general anesthetics are normally used to immobilize animals for fMRI.

Functional MRI measures changes in magnetic susceptibility in tissue that is associated with increases in neuronal activity. These changes are accompanied by changes in hemodynamics such as increased cerebral blood flow (Sokolloff et al., 1977, Fox et al., 1986), increased cerebral blood volume (Fox and Raichle, 1986, Belliveau et al., 1990) and increased oxygen levels (Fox et al., 1986, Fox et al., 1988). Despite an increase in neuronal activity, the blood flow to an activated brain region exceeds the tissue oxygen uptake (Fox and Raichle, 1986, Fox et al., 1988) resulting in a net increase in the content of oxygenated hemoglobin. The precessional frequencies or `wobble' of water protons aligned in the magnetic field of the spectrometer are affected by the different magnetic properties of oxygenated and deoxygenated hemoglobin in the blood flowing through the region of interest. Changes in metabolic activity that alter the ratio of oxygenated to deoxygenated hemoglobin are revealed by changes in MR-signal intensity (Ogawa et al., 1992). These alterations in signal intensity (voxels in which paramagnetic deoxyhemoglobin content is decreased are illuminated in the image) related to blood oxygenation are termed the BOLD (blood oxygenation-level-dependent) effect (Ogawa et al., 1992). Hence, one of the critical variables influencing signal intensity in fMRI is local changes in cerebral blood flow (CBF).

General anesthetics depress CNS metabolic activity and reduce basal CBF (Ueki et al., 1992) and consequently, reduce BOLD signal in fMRI. Despite these limitations, Scanley et al. (1997)observed fMRI signal changes in the somatosensory cortex of propofol-anesthetized rats in response to median nerve stimulation. Using another general anesthetic, α-chloralose, Hyder et al. (1994), reported fMRI signal changes in the somatosensory cortex of rats in response to foot shock. Alpha-chloralose is routinely used as an anesthetic in many CNS studies because it does not depress cortical activity and provides a stable CBF that increases during brain activation (Lindauer et al., 1993). Nonetheless, changes in CBF are more robust in awake animals. Bonvento et al. (1994)used laser-Doppler flow to record changes in CBF in rats under awake and anesthetized conditions after exposure to 1 min of excess carbon dioxide in the blood. They reported an increase in CBF in awake rats compared to α-chloralose anesthetized animals. While studies performed under anesthesia are informative and show stimulus-evoked changes in brain activity, the physiological responses that contribute to the fMRI signal are dampened, limiting the sensitivity and scope of behavioral investigation.

The present studies were undertaken to test a customized head and body holder that would minimize motion artifact in fMRI studies of fully conscious animals, and to explore the applications of this device by assessing changes in brain activity associated with mild foot shock.

Section snippets

Animal care

Male Sprague-Dawley rats were obtained from Harlan Sprague-Dawley Laboratories (Indianapolis, IN). Animals were group housed in Plexiglas cages, four to a cage, and maintained in ambient temperature (22–24°C) on a 12:12 light:dark (lights on at 09:00). Food and water were provided ad libitum. All animals were acquired and cared for in accordance with the guidelines published in the `NIH Guide for the Care and Use of Laboratory Animals' (National Institutes of Health Publications N. 80-23,

Experiment I—evaluating motion artifact

The data in Experiment I support the premise that fMRI can be performed in conscious animals provided there is adequate head and body restraint. Examination of individual data sets as shown in the top row of Fig. 2, revealed no ostensible motion artifact that could have occurred over the data acquisition period. When multiple baseline images were subtracted, on a pixel-by-pixel basis, from the initial baseline as shown in the bottom row of Fig. 2, there was minimal fluctuation in the signal

Discussion

Random movement of the subject is a major concern in all forms of MR imaging. Motion artifact is a potential concern during phase encoding for any particular data set and during repetitive sampling in the same subject over time. Passive restraining devices, e.g. bite plates and head rests, have been successfully used in human studies where the subject is instructed not to move. It would be difficult to train animals to remain motionless for extended periods of times, i.e. minutes to hours,

Conclusions

In summary, the present work demonstrates the feasibility of fMRI in conscious animals. Current methods utilizing anesthetized animals, which are known to dampened neuronal activity and CBF, limit the application of fMRI in the study of behavior. Since the level of arousal (conscious versus anesthetized) is inextricably linked to behavior, the future use of awake animals will provide a clearer understanding of the neural circuitry underlying the organization of behavior. More importantly, this

Acknowledgements

The authors would like to thank Art Allard, Borislav Ristic, M.Sc. and Richard Carano, M.Sc. for their kind assistance in these studies. Support for this research was obtained from the National Institute of Health (MH52280 to CFF). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Institute of Mental Health. Part of this work was performed during the tenure of an Established Investigatorship from the American Heart

References (25)

  • G.M. Hathout et al.

    Quantitative physiological model of blood oxygenation for functional magnetic resonance imaging

    Invest Radiol

    (1995)
  • F.A. Henn et al.

    Melancholia in rodents: neurobiology and pharmacology

    Psychopharmacol Bull

    (1985)
  • Cited by (103)

    • Multi-band SWIFT enables quiet and artefact-free EEG-fMRI and awake fMRI studies in rat

      2020, NeuroImage
      Citation Excerpt :

      Movement occurring inside the FOV is still problematic, although radial sampling offers better resilience to movement as the center of the k-space is typically oversampled compared with Cartesian sampling (Glover and Pauly, 1992). There is a growing interest in awake rodent fMRI imaging (King et al., 2005; Lahti et al., 1998) to allow for behavioral study settings and thus increased translational value of the preclinical experiments. Early awake rat FC studies demonstrated anatomically well-localized signal fluctuations between several brain regions important for emotional and cognitive functions (Zhang et al., 2010).

    • Open-source hardware designs for MRI of mice, rats, and marmosets: Integrated animal holders and radiofrequency coils

      2019, Journal of Neuroscience Methods
      Citation Excerpt :

      This is most notably observed in the ubiquitous solution for rodent imaging, where local surface coils are placed superior to the ear bars, thereby avoiding the ear bars, but reducing the sensitive region of the receive coil. Although commercial vendors and researchers have developed animal holders for anesthetized and awake mice (Akselrod-Ballin et al., 2016; Boretius et al., 2009; Herrmann et al., 2014; Kokuryo et al., 2010; Madularu et al., 2017), rats (Cheng et al., 2009; Lahti et al., 1998; Xu et al., 2013), and marmosets (Belcher et al., 2013; Papoti et al., 2013), they are not typically made freely available (thereby not allowing for easy reproduction) and many are not developed in conjunction with the RF coil, as required to maximize performance. This manuscript describes fully integrated solutions for mice and rat imaging, as well as an MR-compatible stereotactic frame for marmoset imaging: all computer-aided-design (CAD) files are provided for components of the animal holders, including anesthesia delivery, immobilization, and the RF coil.

    • Functional MRI of Synaptic Plasticity

      2018, Handbook of Behavioral Neuroscience
    View all citing articles on Scopus
    View full text