- Email: [email protected]
Neuronal Coding of Implicit Emotion Categories in the Subcallosal Cortex in Patients with Depression
PRIORITY COMMUNICATION
Neuronal Coding of Implicit Emotion Categories in the Subcallosal Cortex in Patients with Depression Adrian W. Laxton, Joseph S. Neimat, Karen D. Davis, Thilo Womelsdorf, William D. Hutchison, Jonathan O. Dostrovsky, Clement Hamani, Helen S. Mayberg, and Andres M. Lozano Background: The subcallosal cingulate and adjacent ventromedial prefrontal cortex (collectively referred to here as the subcallosal cortex or SCC) have been identified as key brain areas in emotional processing. The SCC’s role in affective valuation as well as severe mood and motivational disturbances, such as major depression, has been largely inferred from measures of neuronal population activity using functional neuroimaging. On the basis of imaging studies, it is unclear whether the SCC predominantly processes 1) negatively valenced affective content, 2) affective arousal, or 3) category-specific affective information. Methods: To clarify these putative functional roles of the SCC, we measured single neuron activity in the SCC of 15 human subjects undergoing deep brain stimulation for depression while they viewed emotionally evocative images grouped into categories that varied in emotional valence (pleasantness) and arousal. Results: We found that the majority of responsive neurons were modulated by specific emotion categories, rather than by valence or arousal alone. Moreover, although these emotion-category-specific neurons responded to both positive and negative emotion categories, a significant majority were selective for negatively valenced emotional content. Conclusions: These findings reveal that single SCC neuron activity reflects the automatic valuational processing and implicit emotion categorization of visual stimuli. Furthermore, because of the predominance of neuronal signals in SCC conveying negative affective valuations and the increased activity in this region among depressed people, the effectiveness of depression therapies that alter SCC neuronal activity may relate to the down-regulation of a previously negative emotional processing bias.
Key Words: Cingulate cortex, deep brain stimulation, depression, emotion, humans, neuronal activity
T
he subcallosal cingulate (Brodmann areas [BA] 25 and 24) and adjacent ventromedial prefrontal cortex (BA 32, 10, 11, and 12; collectively referred to here as the subcallosal cortex [SCC]) form a core structure for emotional processing in a larger reward and salience evaluation network interconnecting the amygdala, ventral striatum, and ventral tegmental areas, among others (1). Population activity of the neuronal circuitry in the SCC has been linked to the implicit and automatic affective valuation of stimuli, to hedonic valuations measured as subjective preferences, and to motivational drive and mood states (2–9). Humans with SCC lesions show changes in their interpretation of emotional information, subjective experience of emotion, and ability to engage in emotionally appropriate social behavior (5,6). Progress in understanding the function of the human SCC is expected to follow from delineating the response characteristics of single neurons in this
From the Department of Neurosurgery (AWL), Wake Forest Baptist Medical Center, Winston-Salem, North Carolina; Department of Surgery (AWL, KDD, WDH, CH, AML), Toronto Western Hospital, Institute of Medical Science, University of Toronto, Toronto, Canada; Department of Neurological Surgery (JSN), Vanderbilt University, Nashville, Tennessee; Department of Biology (TW), York University, Toronto, Canada; Department of Physiology (WDH, JOD), University of Toronto, Toronto, Canada; and Department of Psychiatry (HSM), Emory University, Atlanta, Georgia. Authors AWL and JSN contributed equally to this work. Address correspondence to Adrian W. Laxton, M.D., F.R.C.S.C., Department of Neurosurgery, Wake Forest Baptist Medical Center, Medical Center Boulevard, Winston-Salem, NC 27101; E-mail: [email protected]. Received Nov 15, 2012; revised Feb 27, 2013; accepted Mar 19, 2013.
0006-3223/$36.00 http://dx.doi.org/10.1016/j.biopsych.2013.03.029
region, similar to recent progress in the functional characterization of adjacent orbitofrontal and dorsal cingulate cortices (2,10). One possible key function of the SCC is the specific affective valuation of stimulus content (10,11). Human imaging studies suggest that these valuational processes are triggered automatically when subjects view affectively loaded images (12). It has remained unclear, however, whether automatic valuations in the SCC proceed according to the affective valence, arousal, or complex emotion category of stimuli. Some studies suggest that the SCC primarily processes unpleasant or negatively valenced information, particularly with sad emotional content (13). Consistent with this proposal, functional imaging studies have found increased activity in the SCC evoked when subjects view sad image content and with the induction of acute sadness in healthy subjects and depressed individuals (4,14,15). Similarly, prominent biases of perceptual and cognitive processing towards negatively valenced information are evident in subjects with altered activation levels in the SCC, particularly in individuals with depression (2). These biases in affective valuations are readily evident in probabilistic choice tasks that probe individuals to choose one stimulus among alternatives dependent on whether the stimulus has been positively or negatively evaluated by feedback in previous encounters of the stimulus. In these tasks, depressed individuals overemphasize negative feedback to guide their choices and fail to effectively use positively valenced feedback to guide future decisions (16–21). Furthermore, microstimulation in a subzone of the pregenual cingulate cortex in macaques has been found to increase negative decision making (22). The surveyed evidence is consistent with the notion that the SCC conveys automatic, negative affective valuations and is implicated in negative mood states, such as anhedonia and depression, and in mood-congruent biases of attentive perceptual processing. However, several lines of evidence have raised doubts about whether the SCC’s function is restricted to negative emotional information processing. First, activation changes in BIOL PSYCHIATRY 2013;74:714–719 & 2013 Society of Biological Psychiatry
A.W. Laxton et al. the SCC that correlate with negatively valenced processes may be secondary to and thus reflect neuronal activity in the larger reward and saliency evaluation circuitry (2,6). For example, evaluation of words and faces with sad emotional content reliably activates the amygdala, and depressed patients show enhanced amygdala responses to sad content when compared with control subjects (23–27). Second, in humans undergoing deep brain stimulation of the SCC, stimulation can produce spontaneous improvements in mood state (28,29), which suggests a more general regulatory function of the SCC that goes beyond the encoding of negatively valenced valuation signals. Third, perceptual and attentional biases toward negatively valenced affective information associated with higher activity levels within the SCC do not necessarily imply an increase in negative valuations but rather could indicate the failure to appropriately appreciate positive rewards and affective information (2). Consistent with SCC involvement in mediating positive emotional valuations, about one-fifth of human functional imaging studies on affective processing report enhanced activity in SCC to positively valenced emotional processes (4). Elevated SCC activity has also been associated with reduced physiologic and subjective affective reactivity to positively and negatively valenced stimuli (30). Moreover, elevated SCC activity that has been linked to perceptual biases toward negative emotional content may in many studies be modulated by variations in arousal and attentional saliency differences between positive and negatively charged stimulus material, making it difficult to unambiguously assign a valence or emotional-category-specific function to the reported SCC activity (16,31,32). Finally, because negatively valenced stimuli tend to be more arousing than positively valenced stimuli, a preferential responsiveness to negative emotional information may actually reflect arousal responsiveness in the SCC (33). We therefore set out to better elucidate the functional role of the SCC in emotional information processing by quantifying the response characteristics of single SCC neurons in patients with treatment-resistant depression undergoing deep brain stimulation of the SCC while viewing images that varied in their categorical emotional information, valence, and arousal. We found that a large proportion of single SCC neurons conveyed reliable information about specific emotion categories that varied in positive and negative valence as well as arousal.
BIOL PSYCHIATRY 2013;74:714–719 715 the treatment of depression (29,35). A dual microelectrode extracellular recording technique was used [for a detailed description, see Levy et al. (36)]. This mapping is done to identify spontaneously firing single units that indicate the transition from white matter to SCC to optimize electrode placement. This mapping also allowed us to test individual SCC neuron responses to images from the International Affective Picture System (IAPS; Figure 1A) (37). Stimuli The IAPS is a set of emotionally evocative color photographs that have been categorized along several dimensions of emotional content. Images vary in valence (ranging from pleasant to unpleasant) and arousal (ranging from calm to excited) (37). During microelectrode neuronal recording, the participants sequentially viewed random series of 50 IAPS images (Figure 1C). Each image was displayed for 2 seconds followed by an image with a fixation cross for 3 to 5 seconds. The images were subdivided into five emotion categories based on valence and arousal: disturbing (low valence/high arousal), sad (low valence/low arousal), neutral (mid valence/lowest arousal), happy (high valence/low arousal), and exhilarating (high valence/high arousal). By structuring the categories in this way, it was possible to differentiate neuronal responsiveness to the five emotion categories from responsiveness to valence (high: happy ⫹ exhilarating images combined; low: sad ⫹ disturbing images combined) and arousal (high: disturbing ⫹ exhilarating images combined; low: sad ⫹ happy images combined). Neuronal Analyses The neuronal recording sites were determined from each patient’s magnetic resonance images using FrameLink on a Medtronic StealthStation (Figure 1B). To estimate the approximate site of neuronal recordings, preoperative three-dimensional spoiled gradient recalled axial images with the Leksell frame in
Methods and Materials Participants Fifteen patients (13 women and 2 men) with treatmentresistant major depressive disorder undergoing deep brain stimulation surgery of the SCC participated in the study. The participants ranged in age from 38 to 55 years (median ¼ 48 years). All participants had Hamilton Depression Rating Scale Scores (34) greater than 20, indicating current depression. All participants had normal intelligence and no structural brain abnormalities and were medically fit to undergo surgery. The participants’ demographic characteristics are summarized in Table S1 in Supplement 1. This study and all of its procedures underwent formal review and were approved by the Research Ethics Board of the University Health Network, University of Toronto. All subjects provided informed consent to participate in this study. Microelectrode Recordings The participants underwent microelectrode mapping during stereotactic implantation of deep brain stimulation electrodes for
Figure 1. (A) Intraoperative photograph of a participant viewing the image protocol on a computer screen during simultaneous microelectrode neuronal recording in that participant’s subcallosal cingulate and adjacent ventromedial prefrontal cortex (SCC, collectively). (B) Postoperative magnetic resonance imaging showing a participant with the deep brain stimulation electrodes in the SCC. (C) Schematic representation of the image protocol in which an image is shown for 2 seconds followed by 3 to 5 seconds of the fixation screen. This process was continued until 50 images had been shown to each participant.
www.sobp.org/journal
A.W. Laxton et al.
716 BIOL PSYCHIATRY 2013;74:714–719 place were transferred to a workstation. The coronal and sagittal planes were reconstructed. Images were then reformatted parallel to the intercommissural plane and orthogonal to the midline. Target coordinates and the trajectory angle of each microelectrode-recording track were plotted into the system. The location of each cell was established based on the distance in millimeters between the recording site and the target. The cell was then plotted in a grid system previously used by our group to standardize the SCC across patients [for a detailed description, see Hamani et al. (38)]. Single neuron activity was extracted from the microelectrode recordings with template matching and principal components analysis of waveforms using Spike 2 from Cambridge Electronic Design (8,39). Neurons with firing rates during image presentation 2 SD higher or lower than firing rates during fixation for that neuron were classified as responsive.
Results Overall Neuron Activity One hundred and thirty-six neurons were recorded in the SCC. The mean firing rate for all responsive neurons during fixation (4.2 ⫾ 2.7 [SD] Hz) did not differ significantly from the firing rate for responsive neurons during the 10 seconds preceding the initiation of the image presentation protocol (3.7 ⫾ 3.3 [SD] Hz; t = 1.1, p ⬎ .05). This baseline firing rate is similar to previously reported baseline firing of neurons in human dorsal anterior cingulate cortex (5.6 ⫾ 3.4 [SD] Hz) (8) and prefrontal cortex (2.3 ⫾ 1.9 [SD] Hz) (32) in nondepressed patients, suggesting that the neuronal activity in our population of depressed individuals was not nonspecifically elevated and may not be unique to depressed individuals. Responsive Neurons Fifty-six neurons (41.2%) responded to the IAPS images (for a detailed summary of responsive neuron activity, see Table S2 in Supplement 1). Of these responsive neurons, 7 neurons (12.5%) responded to a specific valence regardless of arousal level, 3 neurons (5.4%) responded to a specific arousal level regardless of valence, and 32 neurons (57.1%) were responsive to a single emotion category (reflecting a specific combination of valence and arousal). Thus, SCC neurons were more likely to respond specifically to distinct emotion categories than to valence or
arousal alone (χ2 ¼ 33.6, p ⬍ .001; Figure 2A). Fourteen of the 32 emotion-specific neurons (43.8%) were responsive to disturbing images, 7 (21.9%) to sad images, 3 (9.4%) to neutral images, 3 (9.4%) to happy images, and 5 (15.6%) to exhilarating images. SCC neurons were therefore preferentially responsive to negative emotion categories (disturbing and sad combined ¼ 21) over neutral (3) or positive emotion categories (happy and exhilarating combined ¼ 8; χ2 ¼ 16.2, p ⬍ .001; Figure 2B). Representative neurons responsive to each emotion category are displayed in Figure 3. Neurons responsive to a specific emotion category were not restricted to a specific location within the SCC (Figure 4). Fourteen neurons were responsive to multiple image types, and not specifically responsive to images of a single emotion category, valence, or arousal. Thus, the number of significant responses was greater than the number of responsive neurons because some neurons were responsive to more than one image type. Seventy-eight significant responses occurred in 56 responsive neurons. Neurons were more likely to increase rather than decrease activity (64 increases and 14 decreases; χ2 ¼ 32.1, p ⬍ .001). The mean firing rates during image presentation of responsive neurons that increased activity to disturbing, sad, neutral, happy, and exhilarating images were 6.6, 5.4, 6.5, 6.3, and 9.5 Hz, respectively, and did not differ significantly between emotion categories (F ¼ .66, p ⬎ .05). The firing rates of neurons responsive to a specific emotion category increased on average by 92.5% or decreased by 47.8% relative to fixation frequency (Figure 5).
Discussion These data demonstrate that a large subset of neurons in the SCC respond to visual images depending on the images’ specific emotional information. This selective processing of emotional content should not be confused with subjective emotional experience, which may arise only following the initial emotional valuation processes (2). Emotion-category-specific responsiveness was more common than valence- or arousalspecific responsiveness. Furthermore, SCC neurons were preferentially more responsive to negative (unpleasant) emotion categories than to positive (pleasant) or neutral emotion categories.
Figure 2. (A) Responsive neurons of the subcallosal cingulate and adjacent ventromedial prefrontal cortex show specificity for complex emotion categories rather than valence or arousal alone (*p ⬍ .001). (B) Preferential responsiveness for negative emotion categories over positive or neutral categories (*p ⬍ .001).
www.sobp.org/journal
A.W. Laxton et al.
BIOL PSYCHIATRY 2013;74:714–719 717
Figure 3. Raster plots and peristimulus histograms of individual example neurons showing significantly enhanced or suppressed responses during processing of images from a specific emotion category. Solid vertical lines indicate the onset and offset of image presentation. The horizontal dashed lines indicate 2 SD above (first five neurons) or below (sixth neuron) the mean firing rate during fixation.
For neurons responsive to a specific emotion category, the magnitude of their response was large, with a greater than 90% increase in firing rate. Neuronal activity during emotional picture viewing was much more likely to increase than decrease. Responsive neurons were not clustered in specific regions within the SCC. The term “SCC” has been chosen as a simple descriptor of the general region from which we recorded neurons. It is important to recognize that the pictorial representation of the recording electrode locations in Figure 4 is only meant to provide a general schematic overview. The specific gyral patterns, underlying cytoarchitecture, and anatomic relationships in each of the 15 participants cannot be inferred from this single image. The observation of disturbing and sad-responsive neurons both in the dorsal and more ventral positions in the trajectories in Figure 4 indicates that such neurons can be found in these distinct subcallosal and ventral medial prefrontal cortical areas, notwithstanding their different inputs and projections. Single neuron activity in response to emotional stimuli has been reported in the ventral prefrontal cortex of epilepsy patients (9,32). This, however, is the first report of human single neuron
activity in the SCC of depressed patients and provides new insights into the neuroanatomic substrates of emotional information processing and depression. The findings help to reconcile some inconsistent results from functional imaging studies
Figure 4. Microelectrode recording sites for each emotion category responsive neuron. Midline sagittal right and left hemispheres depicted. Red ¼ image type 1 (disturbing), blue ¼ image type 2 (sad), white ¼ image type 3 (neutral), yellow ¼ image type 4 (happy), and green ¼ image type 5 (exhilarating).
www.sobp.org/journal
718 BIOL PSYCHIATRY 2013;74:714–719
A.W. Laxton et al.
Figure 5. Mean change in firing rates (as a percentage relative to the prestimulus firing rate) for each emotion category responsive neuron during its significant (A) increase or (B) decrease to a specific image type compared with its activity during all other image types.
regarding the specificity of this region for negative emotion processing. In particular, this identified pattern of categoryselective single neuron modulations in SCC is in line with the conclusion from a recent meta-review of functional imaging studies of emotional processing and regulation. This review suggests that emotion-specific information processing occurs in the SCC but is heterogeneous with its activation dependent on the interpretational demands of the stimuli or tasks (7). Importantly, this heterogeneous tuning of single SCC neurons is biased toward valuations and interpretation of negative stimulus content. In the present study, SCC neurons responded to positive and negative emotion categories, but more negative emotion processing neurons were present than positive emotion processing neurons. Because the participants in this study were patients with major depression, these results could reflect an alteration in the way that the SCC in depressed patients processes emotional information. Similarly we cannot fully ascertain if some of these responses are related or influenced by the medications that the patients were taking. This means that the generalizability of our findings to nondepressed, nonmedicated individuals is uncertain. Functional imaging studies, however, support the notion that the SCC region preferentially responds to negative emotional content in both nondepressed and depressed people, but that in depressed individuals, the SCC is overactive (13,15). Recent work has confirmed that patients with chronic depression show resting hyperactivity in the SCC region that diminishes with successful treatments, including medications (14,40), cognitive behavioral therapy (41), electroconvulsive therapy (42), or deep brain stimulation (29,35,43). Overall, these findings may also help to explain how depression therapies that alter activity in this region work. By modulating neuronal activity in the SCC and its downstream targets, these therapies may downregulate an overactive and preferentially negative emotional processing bias (22,27,44). This research was supported by the Surgeon-Scientist Program, Department of Surgery, University of Toronto, and Canadian Institutes of Health Research. Karen Davis was a Canada Research Chair in Brain and Behavior. Andres Lozano is a Canada Research Chair in Neuroscience. Drs. Lozano and Mayberg hold intellectual property in the field of deep brain stimulation for depression. Drs. Lozano, Hamani, and www.sobp.org/journal
Mayberg are consultants for St. Jude Neuromodulation. Drs. Laxton, Neimat, Davis, Womelsdorf, Hutchison, and Dostrovsky report no biomedical financial interests or potential conflicts of interest. Supplementary material cited in this article is available online at http://dx.doi.org/10.1016/j.biopsych.2013.03.029. 1. Haber SN, Knutson B (2009): The reward circuit: Linking primate anatomy and human imaging. Neuropsychopharmacology 35:4–26. 2. Murray EA, Wise SP, Drevets WC (2011): Localization of dysfunction in major depressive disorder: Prefrontal cortex and amygdala. Biol Psychiatry 69:e43–e54. 3. Der-Avakian A, Markou A (2012): The neurobiology of anhedonia and other reward-related deficits. Trends Neurosci 35:68–77. 4. Phan KL, Wager T, Taylor SF, Liberzon I (2002): Functional neuroanatomy of emotion: A meta-analysis of emotion activation studies in PET and fMRI. Neuroimage 16:331–348. 5. Hornak J, Bramham J, Rolls ET, Morris RG, O’Doherty J, Bullock PR, et al. (2003): Changes in emotion after circumscribed surgical lesions of the orbitofrontal and cingulate cortices. Brain 126:1691–1712. 6. Price JL, Drevets WC (2012): Neural circuits underlying the pathophysiology of mood disorders. Trends Cogn Sci 16:61–71. 7. Lindquist KA, Wager TD, Kober H, Bliss-Moreau E, Barrett LF (2012): The brain basis of emotion: A meta-analytic review. Behav Brain Sci 35: 121–143. 8. Davis KD, Taylor KS, Hutchison WD, Dostrovsky JO, McAndrews MP, Richter EO, et al. (2005): Human anterior cingulate cortex neurons encode cognitive and emotional demands. J Neurosci 25:8402–8406. 9. Kawasaki H, Adolphs R, Kaufman O, Damasio H, Damasio AR, Granner M, et al. (2001): Single-neuron responses to emotional visual stimuli recorded in human ventral prefrontal cortex. Nat Neurosci 4:15–16. 10. Wallis JD (2012): Cross-species studies of orbitofrontal cortex and value-based decision-making. Nat Neurosci 15:13–19. 11. Hunt LT, Kolling N, Soltani A, Woolrich MW, Rushworth MFS, Behrens TEJ (2012): Mechanisms underlying cortical activity during value-guided choice. Nat Neurosci 15:470–476. 12. Lebreton M, Jorge S, Michel V, Thirion B, Pessiglione M (2009): An automatic valuation system in the human brain: evidence from functional neuroimaging. Neuron 64:431–439. 13. Vogt BA (2005): Pain and emotion interactions in subregions of the cingulate gyrus. Nature Rev Neurosci 6:533–544. 14. Mayberg HS, Liotti M, Brannan SK, McGinnis S, Mahurin RK, Jerabek PA, et al. (1999): Reciprocal limbic-cortical function and negative mood: Converging PET findings in depression and normal sadness. Am J Psychiatry 156:675–682. 15. George MSM, Ketter TAT, Parekh PIP, Horwitz BB, Herscovitch PP, Post RMR (1995): Brain activity during transient sadness and happiness in healthy women. Am J Psychiatry 152:341–351.
A.W. Laxton et al. 16. Gradin VB, Kumar P, Waiter G, Ahearn T, Stickle C, Milders M, et al. (2011): Expected value and prediction error abnormalities in depression and schizophrenia. Brain 134:1751–1764. 17. Taylor Tavares JV, Clark L, Furey ML, Williams GB, Sahakian BJ, Drevets WC (2008): Neural basis of abnormal response to negative feedback in unmedicated mood disorders. Neuroimage 42:1118–1126. 18. Pizzagalli DA, Iosifescu D, Hallett LA, Ratner KG, Fava M (2008): Reduced hedonic capacity in major depressive disorder: Evidence from a probabilistic reward task. J Psychiatr Res 43:76–87. 19. Henriques JB, Glowacki JM, Davidson RJ (1994): Reward fails to alter response bias in depression. J Abnorm Psychol 103:460–466. 20. Murphy FC, Sahakian BJ, Rubinsztein JS, Michael A, Rogers RD, Robbins TW, et al. (1999): Emotional bias and inhibitory control processes in mania and depression. Psychol Med 29:1307–1321. 21. Murray LA, Whitehouse WG, Alloy LB (1999): Mood congruence and depressive deficits in memory: A forced-recall analysis. Memory 7:175–196. 22. Amemori K-i, Graybiel AM (2012): Localized microstimulation of primate pregenual cingulate cortex induces negative decisionmaking. Nat Neurosci 15:776–785. 23. Fu CHY, Williams SCR, J CA, Brammer MJ, Walsh ND, Kim J, et al. (2004): Attenuation of the neural response to sad faces in major depression by antidepressant treatment: A prospective, event-related functional magnetic resonance imaging study. Arch Gen Psychiatry 61:877–889. 24. Siegle GJ, Steinhauer SR, Thase ME, Stenger VA, Carter CS (2002): Can’t shake that feeling: Event-related fMRI assessment of sustained amygdala activity in response to emotional information in depressed individuals. Biol Psychiatry 51:693–707. 25. Surguladze S, Brammer MJ, Keedwell P, Giampietro V, Young AW, Travis MJ, et al. (2005): A differential pattern of neural response toward sad versus happy facial expressions in major depressive disorder. Biol Psychiatry 57:201–209. 26. Neumeister A, Drevets WC, Belfer I, Luckenbaugh DA, Henry S, Bonne O, et al. (2006): Effects of a [alpha]2C-adrenoreceptor gene polymorphism on neural responses to facial expressions in depression. Neuropsychopharmacology 31:1750–1756. 27. Victor TA, Furey ML, Fromm SJ, Öhman A, Drevets WC (2010): Relationship between amygdala responses to masked faces and mood state and treatment in major depressive disorder. Arch Gen Psychiatry 67:1128–1138. 28. Holtzheimer PE, Kelley ME, Gross RE, Filkowski MM, Garlow SJ, Barrocas A, et al. (2012): Subcallosal cingulate deep brain stimulation for treatmentresistant unipolar and bipolar depression. Arch Gen Psychiatry 69:150–158. 29. Mayberg HS, Lozano AM, Voon V, McNeely HE, Seminowicz D, Hamani C, et al. (2005): Deep brain stimulation for treatment-resistant depression. Neuron 45:651–660. 30. Bylsma LM, Morris BH, Rottenberg J (2008): A meta-analysis of emotional reactivity in major depressive disorder. Clin Psychol Rev 28: 676–691.
BIOL PSYCHIATRY 2013;74:714–719 719 31. Kaping D, Vinck M, Hutchison RM, Everling S, Womelsdorf T (2011): Specific contributions of ventromedial, anterior cingulate, and lateral prefrontal cortex for attentional selection and stimulus valuation. PLoS Biol 9:e1001224. 32. Kawasaki H, Adolphs R, Oya H, Kovach C, Damasio H, Kaufman O, et al. (2005): Analysis of single-unit responses to emotional scenes in human ventromedial prefrontal cortex. J Cogn Neurosci 17: 1509–1518. 33. Lewis P, Critchley H, Rotshtein P, Dolan R (2006): Neural correlates of processing valence and arousal in affective words. Cereb Cortex 17: 742–748. 34. Hamilton M (1960): A rating scale for depression. J Neurol Neurosurg Psychiatry 23:56–62. 35. Lozano AM, Mayberg HS, Giacobbe P, Hamani C, Craddock RC, Kennedy SH (2008): Subcallosal cingulate gyrus deep brain stimulation for treatment-resistant depression. Biol Psychiatry 64: 461–467. 36. Levy R, Lozano AM, Hutchison WD, Dostrovsky JO (2007): Dual microelectrode technique for deep brain stereotactic surgery in humans. Neurosurgery 60:277–284. 37. Lang PJ, Bradley MM, Cuthbert BN (1997): International Affective Picture System (IAPS): Technical Manual and Affective Ratings. Gainesville: University of Florida Press. 38. Hamani C, Mayberg H, Snyder B, Giacobbe P, Kennedy S, Lozano AM (2009): Deep brain stimulation of the subcallosal cingulate gyrus for depression: Anatomical location of active contacts in clinical responders and a suggested guideline for targeting. J Neurosurg 111: 1209–1215. 39. Spike2 (2006): Spike 2. 6.01 ed. Cambridge, UK: Cambridge Electronic Design. 40. Kennedy SH, Evans KR, Krüger S, Mayberg HS, Meyer JH, McCann S, et al. (2001): Changes in regional brain glucose metabolism measured with positron emission tomography after paroxetine treatment of major depression. Am J Psychiatry 158:899–905. 41. Kennedy SH, Konarski JZ, Segal ZV, Lau MA, Bieling PJ, McIntyre RS, et al. (2007): Differences in brain glucose metabolism between responders to CBT and venlafaxine in a 16-week randomized controlled trial. Am J Psychiatry 164:778–788. 42. Nobler MS, Oquendo MA, Kegeles LS, Malone KM, Campbell CC, Sackeim HA, et al. (2001): Decreased regional brain metabolism after ect. Am J Psychiatry 158:305–308. 43. Lozano AM, Giacobbe P, Hamani C, Rizvi SJ, Kennedy SH, Kolivakis TT, et al. (2012): A multicenter pilot study of subcallosal cingulate area deep brain stimulation for treatment-resistant depression. J Neurosurg 116:315–322. 44. Ressler KJ, Mayberg HS (2007): Targeting abnormal neural circuits in mood and anxiety disorders: From the laboratory to the clinic. Nat Neurosci 10:1116–1124.
www.sobp.org/journal
Neuronal Coding of Implicit Emotion Categories in the Subcallosal Cortex in Patients with Depression Adrian W. Laxton, Joseph S. Neimat, Karen D. Davis, Thilo Womelsdorf, William D. Hutchison, Jonathan O. Dostrovsky, Clement Hamani, Helen S. Mayberg, and Andres M. Lozano Background: The subcallosal cingulate and adjacent ventromedial prefrontal cortex (collectively referred to here as the subcallosal cortex or SCC) have been identified as key brain areas in emotional processing. The SCC’s role in affective valuation as well as severe mood and motivational disturbances, such as major depression, has been largely inferred from measures of neuronal population activity using functional neuroimaging. On the basis of imaging studies, it is unclear whether the SCC predominantly processes 1) negatively valenced affective content, 2) affective arousal, or 3) category-specific affective information. Methods: To clarify these putative functional roles of the SCC, we measured single neuron activity in the SCC of 15 human subjects undergoing deep brain stimulation for depression while they viewed emotionally evocative images grouped into categories that varied in emotional valence (pleasantness) and arousal. Results: We found that the majority of responsive neurons were modulated by specific emotion categories, rather than by valence or arousal alone. Moreover, although these emotion-category-specific neurons responded to both positive and negative emotion categories, a significant majority were selective for negatively valenced emotional content. Conclusions: These findings reveal that single SCC neuron activity reflects the automatic valuational processing and implicit emotion categorization of visual stimuli. Furthermore, because of the predominance of neuronal signals in SCC conveying negative affective valuations and the increased activity in this region among depressed people, the effectiveness of depression therapies that alter SCC neuronal activity may relate to the down-regulation of a previously negative emotional processing bias.
Key Words: Cingulate cortex, deep brain stimulation, depression, emotion, humans, neuronal activity
T
he subcallosal cingulate (Brodmann areas [BA] 25 and 24) and adjacent ventromedial prefrontal cortex (BA 32, 10, 11, and 12; collectively referred to here as the subcallosal cortex [SCC]) form a core structure for emotional processing in a larger reward and salience evaluation network interconnecting the amygdala, ventral striatum, and ventral tegmental areas, among others (1). Population activity of the neuronal circuitry in the SCC has been linked to the implicit and automatic affective valuation of stimuli, to hedonic valuations measured as subjective preferences, and to motivational drive and mood states (2–9). Humans with SCC lesions show changes in their interpretation of emotional information, subjective experience of emotion, and ability to engage in emotionally appropriate social behavior (5,6). Progress in understanding the function of the human SCC is expected to follow from delineating the response characteristics of single neurons in this
From the Department of Neurosurgery (AWL), Wake Forest Baptist Medical Center, Winston-Salem, North Carolina; Department of Surgery (AWL, KDD, WDH, CH, AML), Toronto Western Hospital, Institute of Medical Science, University of Toronto, Toronto, Canada; Department of Neurological Surgery (JSN), Vanderbilt University, Nashville, Tennessee; Department of Biology (TW), York University, Toronto, Canada; Department of Physiology (WDH, JOD), University of Toronto, Toronto, Canada; and Department of Psychiatry (HSM), Emory University, Atlanta, Georgia. Authors AWL and JSN contributed equally to this work. Address correspondence to Adrian W. Laxton, M.D., F.R.C.S.C., Department of Neurosurgery, Wake Forest Baptist Medical Center, Medical Center Boulevard, Winston-Salem, NC 27101; E-mail: [email protected]. Received Nov 15, 2012; revised Feb 27, 2013; accepted Mar 19, 2013.
0006-3223/$36.00 http://dx.doi.org/10.1016/j.biopsych.2013.03.029
region, similar to recent progress in the functional characterization of adjacent orbitofrontal and dorsal cingulate cortices (2,10). One possible key function of the SCC is the specific affective valuation of stimulus content (10,11). Human imaging studies suggest that these valuational processes are triggered automatically when subjects view affectively loaded images (12). It has remained unclear, however, whether automatic valuations in the SCC proceed according to the affective valence, arousal, or complex emotion category of stimuli. Some studies suggest that the SCC primarily processes unpleasant or negatively valenced information, particularly with sad emotional content (13). Consistent with this proposal, functional imaging studies have found increased activity in the SCC evoked when subjects view sad image content and with the induction of acute sadness in healthy subjects and depressed individuals (4,14,15). Similarly, prominent biases of perceptual and cognitive processing towards negatively valenced information are evident in subjects with altered activation levels in the SCC, particularly in individuals with depression (2). These biases in affective valuations are readily evident in probabilistic choice tasks that probe individuals to choose one stimulus among alternatives dependent on whether the stimulus has been positively or negatively evaluated by feedback in previous encounters of the stimulus. In these tasks, depressed individuals overemphasize negative feedback to guide their choices and fail to effectively use positively valenced feedback to guide future decisions (16–21). Furthermore, microstimulation in a subzone of the pregenual cingulate cortex in macaques has been found to increase negative decision making (22). The surveyed evidence is consistent with the notion that the SCC conveys automatic, negative affective valuations and is implicated in negative mood states, such as anhedonia and depression, and in mood-congruent biases of attentive perceptual processing. However, several lines of evidence have raised doubts about whether the SCC’s function is restricted to negative emotional information processing. First, activation changes in BIOL PSYCHIATRY 2013;74:714–719 & 2013 Society of Biological Psychiatry
A.W. Laxton et al. the SCC that correlate with negatively valenced processes may be secondary to and thus reflect neuronal activity in the larger reward and saliency evaluation circuitry (2,6). For example, evaluation of words and faces with sad emotional content reliably activates the amygdala, and depressed patients show enhanced amygdala responses to sad content when compared with control subjects (23–27). Second, in humans undergoing deep brain stimulation of the SCC, stimulation can produce spontaneous improvements in mood state (28,29), which suggests a more general regulatory function of the SCC that goes beyond the encoding of negatively valenced valuation signals. Third, perceptual and attentional biases toward negatively valenced affective information associated with higher activity levels within the SCC do not necessarily imply an increase in negative valuations but rather could indicate the failure to appropriately appreciate positive rewards and affective information (2). Consistent with SCC involvement in mediating positive emotional valuations, about one-fifth of human functional imaging studies on affective processing report enhanced activity in SCC to positively valenced emotional processes (4). Elevated SCC activity has also been associated with reduced physiologic and subjective affective reactivity to positively and negatively valenced stimuli (30). Moreover, elevated SCC activity that has been linked to perceptual biases toward negative emotional content may in many studies be modulated by variations in arousal and attentional saliency differences between positive and negatively charged stimulus material, making it difficult to unambiguously assign a valence or emotional-category-specific function to the reported SCC activity (16,31,32). Finally, because negatively valenced stimuli tend to be more arousing than positively valenced stimuli, a preferential responsiveness to negative emotional information may actually reflect arousal responsiveness in the SCC (33). We therefore set out to better elucidate the functional role of the SCC in emotional information processing by quantifying the response characteristics of single SCC neurons in patients with treatment-resistant depression undergoing deep brain stimulation of the SCC while viewing images that varied in their categorical emotional information, valence, and arousal. We found that a large proportion of single SCC neurons conveyed reliable information about specific emotion categories that varied in positive and negative valence as well as arousal.
BIOL PSYCHIATRY 2013;74:714–719 715 the treatment of depression (29,35). A dual microelectrode extracellular recording technique was used [for a detailed description, see Levy et al. (36)]. This mapping is done to identify spontaneously firing single units that indicate the transition from white matter to SCC to optimize electrode placement. This mapping also allowed us to test individual SCC neuron responses to images from the International Affective Picture System (IAPS; Figure 1A) (37). Stimuli The IAPS is a set of emotionally evocative color photographs that have been categorized along several dimensions of emotional content. Images vary in valence (ranging from pleasant to unpleasant) and arousal (ranging from calm to excited) (37). During microelectrode neuronal recording, the participants sequentially viewed random series of 50 IAPS images (Figure 1C). Each image was displayed for 2 seconds followed by an image with a fixation cross for 3 to 5 seconds. The images were subdivided into five emotion categories based on valence and arousal: disturbing (low valence/high arousal), sad (low valence/low arousal), neutral (mid valence/lowest arousal), happy (high valence/low arousal), and exhilarating (high valence/high arousal). By structuring the categories in this way, it was possible to differentiate neuronal responsiveness to the five emotion categories from responsiveness to valence (high: happy ⫹ exhilarating images combined; low: sad ⫹ disturbing images combined) and arousal (high: disturbing ⫹ exhilarating images combined; low: sad ⫹ happy images combined). Neuronal Analyses The neuronal recording sites were determined from each patient’s magnetic resonance images using FrameLink on a Medtronic StealthStation (Figure 1B). To estimate the approximate site of neuronal recordings, preoperative three-dimensional spoiled gradient recalled axial images with the Leksell frame in
Methods and Materials Participants Fifteen patients (13 women and 2 men) with treatmentresistant major depressive disorder undergoing deep brain stimulation surgery of the SCC participated in the study. The participants ranged in age from 38 to 55 years (median ¼ 48 years). All participants had Hamilton Depression Rating Scale Scores (34) greater than 20, indicating current depression. All participants had normal intelligence and no structural brain abnormalities and were medically fit to undergo surgery. The participants’ demographic characteristics are summarized in Table S1 in Supplement 1. This study and all of its procedures underwent formal review and were approved by the Research Ethics Board of the University Health Network, University of Toronto. All subjects provided informed consent to participate in this study. Microelectrode Recordings The participants underwent microelectrode mapping during stereotactic implantation of deep brain stimulation electrodes for
Figure 1. (A) Intraoperative photograph of a participant viewing the image protocol on a computer screen during simultaneous microelectrode neuronal recording in that participant’s subcallosal cingulate and adjacent ventromedial prefrontal cortex (SCC, collectively). (B) Postoperative magnetic resonance imaging showing a participant with the deep brain stimulation electrodes in the SCC. (C) Schematic representation of the image protocol in which an image is shown for 2 seconds followed by 3 to 5 seconds of the fixation screen. This process was continued until 50 images had been shown to each participant.
www.sobp.org/journal
A.W. Laxton et al.
716 BIOL PSYCHIATRY 2013;74:714–719 place were transferred to a workstation. The coronal and sagittal planes were reconstructed. Images were then reformatted parallel to the intercommissural plane and orthogonal to the midline. Target coordinates and the trajectory angle of each microelectrode-recording track were plotted into the system. The location of each cell was established based on the distance in millimeters between the recording site and the target. The cell was then plotted in a grid system previously used by our group to standardize the SCC across patients [for a detailed description, see Hamani et al. (38)]. Single neuron activity was extracted from the microelectrode recordings with template matching and principal components analysis of waveforms using Spike 2 from Cambridge Electronic Design (8,39). Neurons with firing rates during image presentation 2 SD higher or lower than firing rates during fixation for that neuron were classified as responsive.
Results Overall Neuron Activity One hundred and thirty-six neurons were recorded in the SCC. The mean firing rate for all responsive neurons during fixation (4.2 ⫾ 2.7 [SD] Hz) did not differ significantly from the firing rate for responsive neurons during the 10 seconds preceding the initiation of the image presentation protocol (3.7 ⫾ 3.3 [SD] Hz; t = 1.1, p ⬎ .05). This baseline firing rate is similar to previously reported baseline firing of neurons in human dorsal anterior cingulate cortex (5.6 ⫾ 3.4 [SD] Hz) (8) and prefrontal cortex (2.3 ⫾ 1.9 [SD] Hz) (32) in nondepressed patients, suggesting that the neuronal activity in our population of depressed individuals was not nonspecifically elevated and may not be unique to depressed individuals. Responsive Neurons Fifty-six neurons (41.2%) responded to the IAPS images (for a detailed summary of responsive neuron activity, see Table S2 in Supplement 1). Of these responsive neurons, 7 neurons (12.5%) responded to a specific valence regardless of arousal level, 3 neurons (5.4%) responded to a specific arousal level regardless of valence, and 32 neurons (57.1%) were responsive to a single emotion category (reflecting a specific combination of valence and arousal). Thus, SCC neurons were more likely to respond specifically to distinct emotion categories than to valence or
arousal alone (χ2 ¼ 33.6, p ⬍ .001; Figure 2A). Fourteen of the 32 emotion-specific neurons (43.8%) were responsive to disturbing images, 7 (21.9%) to sad images, 3 (9.4%) to neutral images, 3 (9.4%) to happy images, and 5 (15.6%) to exhilarating images. SCC neurons were therefore preferentially responsive to negative emotion categories (disturbing and sad combined ¼ 21) over neutral (3) or positive emotion categories (happy and exhilarating combined ¼ 8; χ2 ¼ 16.2, p ⬍ .001; Figure 2B). Representative neurons responsive to each emotion category are displayed in Figure 3. Neurons responsive to a specific emotion category were not restricted to a specific location within the SCC (Figure 4). Fourteen neurons were responsive to multiple image types, and not specifically responsive to images of a single emotion category, valence, or arousal. Thus, the number of significant responses was greater than the number of responsive neurons because some neurons were responsive to more than one image type. Seventy-eight significant responses occurred in 56 responsive neurons. Neurons were more likely to increase rather than decrease activity (64 increases and 14 decreases; χ2 ¼ 32.1, p ⬍ .001). The mean firing rates during image presentation of responsive neurons that increased activity to disturbing, sad, neutral, happy, and exhilarating images were 6.6, 5.4, 6.5, 6.3, and 9.5 Hz, respectively, and did not differ significantly between emotion categories (F ¼ .66, p ⬎ .05). The firing rates of neurons responsive to a specific emotion category increased on average by 92.5% or decreased by 47.8% relative to fixation frequency (Figure 5).
Discussion These data demonstrate that a large subset of neurons in the SCC respond to visual images depending on the images’ specific emotional information. This selective processing of emotional content should not be confused with subjective emotional experience, which may arise only following the initial emotional valuation processes (2). Emotion-category-specific responsiveness was more common than valence- or arousalspecific responsiveness. Furthermore, SCC neurons were preferentially more responsive to negative (unpleasant) emotion categories than to positive (pleasant) or neutral emotion categories.
Figure 2. (A) Responsive neurons of the subcallosal cingulate and adjacent ventromedial prefrontal cortex show specificity for complex emotion categories rather than valence or arousal alone (*p ⬍ .001). (B) Preferential responsiveness for negative emotion categories over positive or neutral categories (*p ⬍ .001).
www.sobp.org/journal
A.W. Laxton et al.
BIOL PSYCHIATRY 2013;74:714–719 717
Figure 3. Raster plots and peristimulus histograms of individual example neurons showing significantly enhanced or suppressed responses during processing of images from a specific emotion category. Solid vertical lines indicate the onset and offset of image presentation. The horizontal dashed lines indicate 2 SD above (first five neurons) or below (sixth neuron) the mean firing rate during fixation.
For neurons responsive to a specific emotion category, the magnitude of their response was large, with a greater than 90% increase in firing rate. Neuronal activity during emotional picture viewing was much more likely to increase than decrease. Responsive neurons were not clustered in specific regions within the SCC. The term “SCC” has been chosen as a simple descriptor of the general region from which we recorded neurons. It is important to recognize that the pictorial representation of the recording electrode locations in Figure 4 is only meant to provide a general schematic overview. The specific gyral patterns, underlying cytoarchitecture, and anatomic relationships in each of the 15 participants cannot be inferred from this single image. The observation of disturbing and sad-responsive neurons both in the dorsal and more ventral positions in the trajectories in Figure 4 indicates that such neurons can be found in these distinct subcallosal and ventral medial prefrontal cortical areas, notwithstanding their different inputs and projections. Single neuron activity in response to emotional stimuli has been reported in the ventral prefrontal cortex of epilepsy patients (9,32). This, however, is the first report of human single neuron
activity in the SCC of depressed patients and provides new insights into the neuroanatomic substrates of emotional information processing and depression. The findings help to reconcile some inconsistent results from functional imaging studies
Figure 4. Microelectrode recording sites for each emotion category responsive neuron. Midline sagittal right and left hemispheres depicted. Red ¼ image type 1 (disturbing), blue ¼ image type 2 (sad), white ¼ image type 3 (neutral), yellow ¼ image type 4 (happy), and green ¼ image type 5 (exhilarating).
www.sobp.org/journal
718 BIOL PSYCHIATRY 2013;74:714–719
A.W. Laxton et al.
Figure 5. Mean change in firing rates (as a percentage relative to the prestimulus firing rate) for each emotion category responsive neuron during its significant (A) increase or (B) decrease to a specific image type compared with its activity during all other image types.
regarding the specificity of this region for negative emotion processing. In particular, this identified pattern of categoryselective single neuron modulations in SCC is in line with the conclusion from a recent meta-review of functional imaging studies of emotional processing and regulation. This review suggests that emotion-specific information processing occurs in the SCC but is heterogeneous with its activation dependent on the interpretational demands of the stimuli or tasks (7). Importantly, this heterogeneous tuning of single SCC neurons is biased toward valuations and interpretation of negative stimulus content. In the present study, SCC neurons responded to positive and negative emotion categories, but more negative emotion processing neurons were present than positive emotion processing neurons. Because the participants in this study were patients with major depression, these results could reflect an alteration in the way that the SCC in depressed patients processes emotional information. Similarly we cannot fully ascertain if some of these responses are related or influenced by the medications that the patients were taking. This means that the generalizability of our findings to nondepressed, nonmedicated individuals is uncertain. Functional imaging studies, however, support the notion that the SCC region preferentially responds to negative emotional content in both nondepressed and depressed people, but that in depressed individuals, the SCC is overactive (13,15). Recent work has confirmed that patients with chronic depression show resting hyperactivity in the SCC region that diminishes with successful treatments, including medications (14,40), cognitive behavioral therapy (41), electroconvulsive therapy (42), or deep brain stimulation (29,35,43). Overall, these findings may also help to explain how depression therapies that alter activity in this region work. By modulating neuronal activity in the SCC and its downstream targets, these therapies may downregulate an overactive and preferentially negative emotional processing bias (22,27,44). This research was supported by the Surgeon-Scientist Program, Department of Surgery, University of Toronto, and Canadian Institutes of Health Research. Karen Davis was a Canada Research Chair in Brain and Behavior. Andres Lozano is a Canada Research Chair in Neuroscience. Drs. Lozano and Mayberg hold intellectual property in the field of deep brain stimulation for depression. Drs. Lozano, Hamani, and www.sobp.org/journal
Mayberg are consultants for St. Jude Neuromodulation. Drs. Laxton, Neimat, Davis, Womelsdorf, Hutchison, and Dostrovsky report no biomedical financial interests or potential conflicts of interest. Supplementary material cited in this article is available online at http://dx.doi.org/10.1016/j.biopsych.2013.03.029. 1. Haber SN, Knutson B (2009): The reward circuit: Linking primate anatomy and human imaging. Neuropsychopharmacology 35:4–26. 2. Murray EA, Wise SP, Drevets WC (2011): Localization of dysfunction in major depressive disorder: Prefrontal cortex and amygdala. Biol Psychiatry 69:e43–e54. 3. Der-Avakian A, Markou A (2012): The neurobiology of anhedonia and other reward-related deficits. Trends Neurosci 35:68–77. 4. Phan KL, Wager T, Taylor SF, Liberzon I (2002): Functional neuroanatomy of emotion: A meta-analysis of emotion activation studies in PET and fMRI. Neuroimage 16:331–348. 5. Hornak J, Bramham J, Rolls ET, Morris RG, O’Doherty J, Bullock PR, et al. (2003): Changes in emotion after circumscribed surgical lesions of the orbitofrontal and cingulate cortices. Brain 126:1691–1712. 6. Price JL, Drevets WC (2012): Neural circuits underlying the pathophysiology of mood disorders. Trends Cogn Sci 16:61–71. 7. Lindquist KA, Wager TD, Kober H, Bliss-Moreau E, Barrett LF (2012): The brain basis of emotion: A meta-analytic review. Behav Brain Sci 35: 121–143. 8. Davis KD, Taylor KS, Hutchison WD, Dostrovsky JO, McAndrews MP, Richter EO, et al. (2005): Human anterior cingulate cortex neurons encode cognitive and emotional demands. J Neurosci 25:8402–8406. 9. Kawasaki H, Adolphs R, Kaufman O, Damasio H, Damasio AR, Granner M, et al. (2001): Single-neuron responses to emotional visual stimuli recorded in human ventral prefrontal cortex. Nat Neurosci 4:15–16. 10. Wallis JD (2012): Cross-species studies of orbitofrontal cortex and value-based decision-making. Nat Neurosci 15:13–19. 11. Hunt LT, Kolling N, Soltani A, Woolrich MW, Rushworth MFS, Behrens TEJ (2012): Mechanisms underlying cortical activity during value-guided choice. Nat Neurosci 15:470–476. 12. Lebreton M, Jorge S, Michel V, Thirion B, Pessiglione M (2009): An automatic valuation system in the human brain: evidence from functional neuroimaging. Neuron 64:431–439. 13. Vogt BA (2005): Pain and emotion interactions in subregions of the cingulate gyrus. Nature Rev Neurosci 6:533–544. 14. Mayberg HS, Liotti M, Brannan SK, McGinnis S, Mahurin RK, Jerabek PA, et al. (1999): Reciprocal limbic-cortical function and negative mood: Converging PET findings in depression and normal sadness. Am J Psychiatry 156:675–682. 15. George MSM, Ketter TAT, Parekh PIP, Horwitz BB, Herscovitch PP, Post RMR (1995): Brain activity during transient sadness and happiness in healthy women. Am J Psychiatry 152:341–351.
A.W. Laxton et al. 16. Gradin VB, Kumar P, Waiter G, Ahearn T, Stickle C, Milders M, et al. (2011): Expected value and prediction error abnormalities in depression and schizophrenia. Brain 134:1751–1764. 17. Taylor Tavares JV, Clark L, Furey ML, Williams GB, Sahakian BJ, Drevets WC (2008): Neural basis of abnormal response to negative feedback in unmedicated mood disorders. Neuroimage 42:1118–1126. 18. Pizzagalli DA, Iosifescu D, Hallett LA, Ratner KG, Fava M (2008): Reduced hedonic capacity in major depressive disorder: Evidence from a probabilistic reward task. J Psychiatr Res 43:76–87. 19. Henriques JB, Glowacki JM, Davidson RJ (1994): Reward fails to alter response bias in depression. J Abnorm Psychol 103:460–466. 20. Murphy FC, Sahakian BJ, Rubinsztein JS, Michael A, Rogers RD, Robbins TW, et al. (1999): Emotional bias and inhibitory control processes in mania and depression. Psychol Med 29:1307–1321. 21. Murray LA, Whitehouse WG, Alloy LB (1999): Mood congruence and depressive deficits in memory: A forced-recall analysis. Memory 7:175–196. 22. Amemori K-i, Graybiel AM (2012): Localized microstimulation of primate pregenual cingulate cortex induces negative decisionmaking. Nat Neurosci 15:776–785. 23. Fu CHY, Williams SCR, J CA, Brammer MJ, Walsh ND, Kim J, et al. (2004): Attenuation of the neural response to sad faces in major depression by antidepressant treatment: A prospective, event-related functional magnetic resonance imaging study. Arch Gen Psychiatry 61:877–889. 24. Siegle GJ, Steinhauer SR, Thase ME, Stenger VA, Carter CS (2002): Can’t shake that feeling: Event-related fMRI assessment of sustained amygdala activity in response to emotional information in depressed individuals. Biol Psychiatry 51:693–707. 25. Surguladze S, Brammer MJ, Keedwell P, Giampietro V, Young AW, Travis MJ, et al. (2005): A differential pattern of neural response toward sad versus happy facial expressions in major depressive disorder. Biol Psychiatry 57:201–209. 26. Neumeister A, Drevets WC, Belfer I, Luckenbaugh DA, Henry S, Bonne O, et al. (2006): Effects of a [alpha]2C-adrenoreceptor gene polymorphism on neural responses to facial expressions in depression. Neuropsychopharmacology 31:1750–1756. 27. Victor TA, Furey ML, Fromm SJ, Öhman A, Drevets WC (2010): Relationship between amygdala responses to masked faces and mood state and treatment in major depressive disorder. Arch Gen Psychiatry 67:1128–1138. 28. Holtzheimer PE, Kelley ME, Gross RE, Filkowski MM, Garlow SJ, Barrocas A, et al. (2012): Subcallosal cingulate deep brain stimulation for treatmentresistant unipolar and bipolar depression. Arch Gen Psychiatry 69:150–158. 29. Mayberg HS, Lozano AM, Voon V, McNeely HE, Seminowicz D, Hamani C, et al. (2005): Deep brain stimulation for treatment-resistant depression. Neuron 45:651–660. 30. Bylsma LM, Morris BH, Rottenberg J (2008): A meta-analysis of emotional reactivity in major depressive disorder. Clin Psychol Rev 28: 676–691.
BIOL PSYCHIATRY 2013;74:714–719 719 31. Kaping D, Vinck M, Hutchison RM, Everling S, Womelsdorf T (2011): Specific contributions of ventromedial, anterior cingulate, and lateral prefrontal cortex for attentional selection and stimulus valuation. PLoS Biol 9:e1001224. 32. Kawasaki H, Adolphs R, Oya H, Kovach C, Damasio H, Kaufman O, et al. (2005): Analysis of single-unit responses to emotional scenes in human ventromedial prefrontal cortex. J Cogn Neurosci 17: 1509–1518. 33. Lewis P, Critchley H, Rotshtein P, Dolan R (2006): Neural correlates of processing valence and arousal in affective words. Cereb Cortex 17: 742–748. 34. Hamilton M (1960): A rating scale for depression. J Neurol Neurosurg Psychiatry 23:56–62. 35. Lozano AM, Mayberg HS, Giacobbe P, Hamani C, Craddock RC, Kennedy SH (2008): Subcallosal cingulate gyrus deep brain stimulation for treatment-resistant depression. Biol Psychiatry 64: 461–467. 36. Levy R, Lozano AM, Hutchison WD, Dostrovsky JO (2007): Dual microelectrode technique for deep brain stereotactic surgery in humans. Neurosurgery 60:277–284. 37. Lang PJ, Bradley MM, Cuthbert BN (1997): International Affective Picture System (IAPS): Technical Manual and Affective Ratings. Gainesville: University of Florida Press. 38. Hamani C, Mayberg H, Snyder B, Giacobbe P, Kennedy S, Lozano AM (2009): Deep brain stimulation of the subcallosal cingulate gyrus for depression: Anatomical location of active contacts in clinical responders and a suggested guideline for targeting. J Neurosurg 111: 1209–1215. 39. Spike2 (2006): Spike 2. 6.01 ed. Cambridge, UK: Cambridge Electronic Design. 40. Kennedy SH, Evans KR, Krüger S, Mayberg HS, Meyer JH, McCann S, et al. (2001): Changes in regional brain glucose metabolism measured with positron emission tomography after paroxetine treatment of major depression. Am J Psychiatry 158:899–905. 41. Kennedy SH, Konarski JZ, Segal ZV, Lau MA, Bieling PJ, McIntyre RS, et al. (2007): Differences in brain glucose metabolism between responders to CBT and venlafaxine in a 16-week randomized controlled trial. Am J Psychiatry 164:778–788. 42. Nobler MS, Oquendo MA, Kegeles LS, Malone KM, Campbell CC, Sackeim HA, et al. (2001): Decreased regional brain metabolism after ect. Am J Psychiatry 158:305–308. 43. Lozano AM, Giacobbe P, Hamani C, Rizvi SJ, Kennedy SH, Kolivakis TT, et al. (2012): A multicenter pilot study of subcallosal cingulate area deep brain stimulation for treatment-resistant depression. J Neurosurg 116:315–322. 44. Ressler KJ, Mayberg HS (2007): Targeting abnormal neural circuits in mood and anxiety disorders: From the laboratory to the clinic. Nat Neurosci 10:1116–1124.
www.sobp.org/journal