Altered task-specific deactivation in the default mode network depends on valence in patients with major depressive disorder

Author’s Accepted Manuscript Altered task-specific deactivation in the default mode network depends on valence in patients with major depressive disorder Bin Zhang, Shijia Li, Chuanjun Zhuo, Meng Li, Adam Safron, Axel Genz, Wen Qin, Chunshui Yu, Martin Walter www.elsevier.com/locate/jad

PII: DOI: Reference:

S0165-0327(15)31070-3 http://dx.doi.org/10.1016/j.jad.2016.08.042 JAD8519

To appear in: Journal of Affective Disorders Received date: 7 October 2015 Revised date: 16 August 2016 Accepted date: 21 August 2016 Cite this article as: Bin Zhang, Shijia Li, Chuanjun Zhuo, Meng Li, Adam Safron, Axel Genz, Wen Qin, Chunshui Yu and Martin Walter, Altered taskspecific deactivation in the default mode network depends on valence in patients with major depressive disorder, Journal of Affective Disorders, http://dx.doi.org/10.1016/j.jad.2016.08.042 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Altered task-specific deactivation in the default mode network depends on valence in patients with major depressive disorder Bin Zhang1,2,3#, Shijia Li2,4,5,6,7#, Chuanjun Zhuo8,9, Meng Li2,7, Adam Safron10 ,Axel Genz1, Wen Qin8, Chunshui Yu8, Martin Walter1,2,6,7,10,11* 1.Department

of

Psychiatry,

Otto-von-Guericke

University,

Magdeburg,

Germany 2.Clinical Affective Neuroimaging Laboratory, Otto-von-Guericke University, Magdeburg, Germany 3. Shanghai Key Laboratory of Psychotic Disorders, Shanghai Mental Health Center, Shanghai Jiaotong University School of Medicine, Shanghai, China 4.School of Psychology and Cognitive Science, East China Normal University, Shanghai, China 5.Key Laboratory of Brain Functional Genomics, Ministry of Education, Shanghai Key Laboratory of Brain Functional Genomics, Shanghai, China 6.Leibniz Institute for Neurobiology, Magdeburg, Germany 7.Department

of Neurology, Otto-von-Guericke University, Magdeburg,

Germany 8.Department of Radiology and Tianjin Key Laboratory of Functional Imaging, Tianjin Medical University General Hospital, Tianjin, China 9.Tianjin Anning Hospital, Tianjin, China 10. Department of Psychology, Northwestern University, Chicago, United States 11. Department of Psychiatry and Psychotherapy, University Tübingen, Tübingen, Germany # Equally as first author. * Corresponding author Prof. Dr. med. Martin Walter Department of Psychiatry and Psychotherapy University Tübingen Osianderstr. 24 72076 Tübingen Phone: +497071 29 86119 1

Fax: +497071 29 5901 E-mail: [email protected] Abstract Background: Major depressive disorder (MDD) is a highly prevalent psychiatric condition in which patients often have difficulties regulating their emotions. Prior studies have shown that attention bias towards negative emotion is linked to activation in regions of the default mode network (DMN) in MDD individuals. Furthermore, MDD patients showed increased resting-state functional connectivity (FC) between the medial prefrontal cortex and other DMN structures. Methods: Twenty-one MDD patients that currently experiencing depressive episodes and twenty-five healthy control participants performed the current emotional expectancy paradigm in a gradient-echo SENSE-SPIRAL fMRI. Whole brain and psycho-physiological interaction (PPI) analysis were applied to explore the task-related brain activity and FCs. Results: Relative to healthy participants, we found MDD patients had greater activity in dorsal medial prefrontal cortex as a function of positive vs. neutral expectancy conditions. PPI results revealed a significant group difference of MDD patients having relatively decreased task-dependent decoupling from DMPFC towards posterior cingulate cortex (PCC) and parieto-occipital cortex during positive vs. neutral expectancy conditions, and patients exhibited a positive correlation between PPI (DMPFC and PCC) and anhedonia as measured via SHAPS during the same conditions. Limitations: Modest sample size and lack of concurrent depressive episodes limit the generalizability of our findings. Conclusions: In MDD patients, insufficient DMN decoupling might occur in response to positive expectancy conditions. Our findings are consistent with the hypothesis that high intrinsic DMN connectivity in MDD patients interfere with the down-regulation of intrinsic focus in order to incorporate information derived from external positive events.

2

Keywords: Major Depressive Disorder; Default Psycho-Physiological Interaction; Functional Connectivity

Mode

Network;

Introduction Major depressive disorder (MDD) is a highly prevalent psychiatric condition, typically characterized by persistent sadness and ‘down states’ that dominate patients’ lives, wherein they lack interest in and happiness from pleasurable activities (anhedonia), and also experience energy exhaustion and often guilty feelings (Belmaker and Agam, 2008; Holtzheimer and Mayberg, 2011). Many of the depressive symptoms are related to the fact that MDD patients often have difficulties modulating emotion due to severe cognitive biases, in which their attention and memory retrieval are typically prone to negative emotional information (Disner et al., 2011; MacLeod et al., 2002). Previous studies that focused on attentional bias in MDD found that patients showed impaired bottom-up emotional activation and top-down attentional control interactions (Mathews and MacLeod, 2005). While healthy control (HC) participants preferentially directed attention towards positive stimuli (Gotlib et al., 2004), MDD patients had difficulties shifting their attentional focus to positive stimuli (Clark et al., 1994; Disner et al., 2011). In the review of Holtzheimer and Mayberg (Holtzheimer and Mayberg, 2011), MDD patients were prone to negative states and once they entered those states, it was very difficult for them to shift back to positive states. And, Kellough and his colleagues found that individuals in depressive episodes showed increased attention towards negative stimuli and decreased attention towards positive stimuli (Kellough et al., 2008), supporting the emotional attention bias hypothesis. Brain imaging studies have revealed that many brain regions responsible for mood regulation were functionally disrupted in MDD patients. For example, Shafritz and colleagues found that individuals with mood disorders required greater cognitive effort to divert attention away from negative stimuli, which was associated with activity in pregenual anterior cingulate cortex (pgACC) (Shafritz et al., 2006). Additionally, resting-state functional connectivity (FC) studies reported MDD patients showed abnormal signal fluctuations in ACC (Greicius et al., 2007; Zhang et al., 3

2015a), amygdala (Cullen et al., 2014; Kong et al., 2013) and medial prefrontal cortex (MPFC) (Bremner et al., 2002; Drevets, 2007; van Tol et al., 2014). Further, emotional attention bias has been linked to activation in MPFC, part of the DMN (Raichle et al., 2001) related to self-referential events and emotional processing (Gusnard et al., 2001). Grimm and colleagues showed that MDD was characterized by impaired deactivation in the anterior DMN during tasks (Grimm et al., 2009). What drives this impairment remains unclear. If MDD is characterized by biased attention, atypical attention focus may be a result of generally increased intrinsic coupling. MDD may also be characterized by atypical incorporation of explicit external information, such as increased sensitivity to actual negative stimuli. Alternatively, attentional bias may result from an inability to prepare for positive stimulation by reorienting attentional resources towards the environment. Several studies have been conducted to identify brain mechanisms contributing to emotional attention bias in MDD patients. Bermpohl and colleagues used the emotional expectancy task to investigate expectancy-induced modulation of emotional picture processing in HC group and MDD patients (Bermpohl et al., 2006). MPFC specifically regulated attentional modulation of emotion processing; they also found impairment of expectancy and perception interaction in dorsal MPFC (DMPFC) when MDD patients processed emotional pictures (Bermpohl et al., 2009). A limitation of this experimental design was that it did not distinguish positive or negative emotional expectancies, only investigating the effect of general, valence-unspecific emotional expectancy. In the present study, we investigated neural correlates of biased attention in MDD patients using a similar paradigm as in Bermpohl’s studies (with the addition of differentiated positive and negative expectancies) (Bermpohl et al., 2006; Bermpohl et al., 2009). This investigation was motivated by two hypotheses: 1) MDD patients have impaired deactivation of DMPFC (anterior DMN) during expectancy of positive events and 2) an impaired decoupling on DMPFC and PCC (anterior – posterior DMN) for positive events expectancy. In this study, the impaired deactivation of DMPFC and the impaired decoupling on DMPFC and PCC are considered as a sign for a reduced 4

positive bias (which is present in healthy subjects) and therefore in line with the negative attentional bias in MDD. We applied gradient-echo SENSE-SPIRAL fMRI a new fMRI sequence with expectancy emotion paradigm (Hypothesis 1). The SENSE-SPIRAL is a new fMRI sequence which can obtain high quality data in areas with low signal-to-noise ratios, such as the prefrontal cortex (Truong and Song, 2008). We used psychophysiological interaction (PPI) analysis to further investigate the FC alterations in MDD patients that related to our task findings (Hypothesis 2).

Methods Participants Twenty-one MDD patients that currently experiencing depressive episodes and twenty-five age- and gender-matched HC participants performed the fMRI experiment. Patients were clinically diagnosed according to the ICD-10 criteria and severity was assessed using the 24-items Hamilton rating scale for depression (HAMD) and Montgomery-Asberg depression rating scale (MADRS). We used SHAPS for psychometric evaluation of anhedonia. Exclusion criteria were major non-psychiatric medical illness, history of seizures, prior electroconvulsive therapy treatments, illicit substance use or substance use disorders, and pregnancy. The exact number of previous depressive episodes is not available for all patients. The length of the current episode was between one and twelve months. All patients were medicated according to clinical standards with a selective serotonin reuptake inhibitor (SSRI), anti-psychotic medication, or a selective noradrenalin reuptake inhibitor (SNRI), tetracyclic antidepressant (TCA), or mood stabilizers. This study was approved by the ethics committee of Tianjin Medical University General Hospital, and all participants provided written informed consent to participate.

Expectancy and emotion task The experiment was designed and administered using the Presentation software package (Neurobehavioral Systems, http://www.neurobs.com). Initially the fMRI Hardware System (Nordic NeuroLab, NNL) was used to project experimental stimuli 5

onto a goggle screen worn by participants, and 8 MDD patients and 8 HC participants were tested in the goggle environment. Due to goggle dysfunction, 13 MDD patients and 17 HC then performed the task through a mirror that located in the head coil and the stimuli were projected through a beamer into the mirror. We modeled the goggle/screen effect as a non-interest regressor in order to minimize the influence of experimental environment change. The emotional expectancy paradigm is event related design used visual cues and pictures with different valences (Figure 1). A total of 60 pictures were selected from the International Affective Picture System (IAPS) (Lang et al., 2008), and pictures were divided into positive (n = 20), neutral (n = 20) and negative (n = 20) pictures according to ratings of perceived valence (mean valence ratings: negative, 2.06±0.54; neutral, 5.11±0.52; positive, 7.68±0.32). Pictures from this set were matched between valences with regard to number of people, situation complexity, and semantic content. Each picture was presented for 4 seconds. Half of trials were preceded by an expectancy task predicting (with 100% validity) upcoming stimulus types; e.g., an upwards-pointing arrow indicated that a positive picture would follow (positive expectancy); a downward-pointing arrow indicated that a negative picture would follow (negative expectancy); a horizontal arrow indicated that a neutral picture would follow (neutral expectancy). All visual cues were displayed after a fixation, and were followed by half of the pictures (expected pictures). The other half of valence counterbalanced pictures were preceded without any cue (unexpected pictures), meaning these pictures followed directly after the fixation. In order to orthogonalise expectancy and picture conditions, 15 visual cues were followed directly by fixation without pictures, and were also modeled as the conditions of expectancy effect (positive expectancy, negative expectancy and neutral expectancy). Each expectancy cue lasted 3-5 seconds. Inter-trial fixation interval was used as the baseline, which lasted 8.5-10.5 seconds.

Data acquisition and analysis 6

MRI was performed using a 3.0-Tesla MR system (Discovery MR750, General Electric, Milwaukee, WI, USA) with an 8-channel head coil. Tight but comfortable foam padding was used to minimize head motion, and earplugs were used to reduce scanner noise. A gradient-echo SENSE-SPIRAL (spiral in) sequence was performed using parameters of TR/TE = 1600/30 ms; FA = 60°, acceleration factor = 2, FOV = 220 mm × 220 mm; matrix = 64 × 64; slice thickness = 4 mm; gap = 0.5 mm; 36 interleaved transverse slices; voxel size 3.44×3.44×4 mm3; 710 image volumes in task scan, which lasts 19 minutes. Sagittal 3D T1-weighted images were acquired by a brain volume sequence with the following scan parameters: repetition time (TR) = 8.2 ms; echo time (TE) = 3.2 ms; inversion time (TI) = 450 ms; flip angle (FA) = 12°; field of view (FOV) = 256 mm × 256 mm; matrix = 256 × 256; slice thickness

=1

mm, no gap; 188 sagittal slices. Data were preprocessed using Statistical Parametric Mapping Software (SPM8, http://www.fil.ion.ucl.ac.uk/spm). The 710 volumes were corrected for time delay between different slices and realigned to the first volume. Head motion parameters were computed by estimating translation in each direction and the angular rotation on each axis for each volume. Each subject had a maximum displacement of less than 3 mm in any cardinal direction (x, y, z), and a maximum spin (x, y, z) less than 3°. Individual structural images were linearly coregistered to the mean functional image; then the transformed structural images were segmented into grey matter (GM), white matter, and cerebrospinal fluid. The GM maps were linearly coregistered to the tissue probability maps in MNI space. The motion-corrected functional volumes were spatially normalized to the individual’s structural image using the parameters estimated during linear coregistration. The functional images were resampled into 3 × 3 × 3 mm3 voxels. Finally, all datasets were smoothed with a Gaussian kernel of 8 × 8 × 8 mm3 FWHM. At the single subject level, we modeled six regressors of interest and convolved with the canonical hemodynamic response (CHR) function on the base of general linear model (GLM). The first 3 regressors indicated the effect of expectancy during cuing trails irrelevant of whether the picture was followed or not (expectancy of 7

negative pictures [Xneg], expectancy of positive pictures [Xpos] and expectancy of neutral pictures [Xneu]). The following 3 regressors indicated the effect of emotion during picture display sessions (negative pictures [Pneg], positive pictures [Ppos] and neutral pictures [Pneu]). The voxel time series were high-pass filtered at 1/128 Hz to account for non-physiological slow drifts in the measured signal and modeled for temporal autocorrelation across scans with an autoregressive model. On the group level, we conducted several t-tests to check the main effect of expectancy (Xpos vs. Xneu, Xneg vs. Xneu, Xpos vs. Xneg). For second-level analyses, single subject contrasts were entered into one-sample and two-sample t-tests across subjects. First, statistical parametric maps were estimated for the contrasts and the reverse contrasts in the healthy group (one sample t-test) with age and goggle/screen as covariates. Then, effects were compared between HC and depressed participants (two-samples t-test).

PPI analysis We used PPI methods to investigate how emotional expectancy regulated FC. Briefly, PPI analyzes possible interactions between regression slopes of different regions that can be significance tested as a measure of FC (Friston et al., 1997). We created a 6 mm radius sphere around the significant peak of the second-level result from the task. We only investigated ‘Xpos vs. Xneu’ (the only condition showing significant group difference) using PPI analysis. The extracted hemodynamic time series were deconvolved, and then the physiological variable was combined with onset times for Xpos and Xneu to derive the interaction term. To obtain data for the physiological variable, we extracted the individual mean time series within a 6 mm radius sphere that centered on the maximum peak within the DMPFC seed region under the threshold of p<0.9 (uncorrected). The physiological factor was then multiplied with the psychological factor, yielding an interaction term, and then re-analyzed in a new GLM model with 3 regressors representing PPI, BOLD and psychological condition. Subject-specific contrast images resulting from the contrast were specified [1 0 0]—where the first column represents the interaction term—and 8

then entered into a group analysis using a two samples t-test. The significance threshold used for PPI analyses was p≤0.05 with cluster level family-wise error (FWE) correction.

PPI correlated with Snaith-Hamilton pleasure scale (SHAPS) In our study, we investigated the correlation between PPI and SHAPS in MDD patients to explore the relationship between FC and anhedonia levels. SHAPS has been developed for psychometric evaluation of anhedonia and assessment of positive valence (Nakonezny et al., 2015; Snaith et al., 1995). The SHAPS contains 14 items and 4 scales: if the subjects answered ‘disagree’ or ‘totally disagree’ to an item, it was assigned a score of 1; otherwise it was 0. A total score was derived from summing the scores from each of the items. Higher SHAPS total scores indicate greater anhedonia (inability to experience pleasure), and a score of 3 or higher indicates a significant reduction in hedonic capacity (Snaith et al., 1995). Post-hoc analysis investigated the relationship between anhedonia and effective connectivity in the MDD group, restricted to regions implicated by the comparison of MDD versus healthy control participants during Xpos vs. Xneu, and controlled for group variables in order to avoid double-dipping. Multiple regression analysis was conducted to investigate PPI and SHAPS correlations, in which SHAPS total score was entered as one covariate and modeled with factorial design specifications. Small volume correction (SVC) was applied based on PPI results, constraining analyses to a 6 mm radius sphere centered at peak cluster coordinates.

Results 1.Task-fMRI findings. We studied all three expectancy contrasts for the comparison ‘MDD Vs. HC’; no significant results of the contrasts Xneg vs. Xneu or Xpos vs. Xneg were revealed. When comparing greater positive expectancy with neutral expectancy, MDD patients showed increased activity in the DMPFC (p<0.05, FWE cluster level corrected, Cluster>157, T = 5.32, x = 15, y = 48, z = 30, k = 390), compared to HCs (Figure 2a). 9

After extracting mean DMPFC values in the conditions of ‘Xpos’, ‘Xneu’ and ‘Xneg’, post hoc comparisons revealed that positive expectancy led to greater activity in DMPFC in MDD patients than HCs (p<0.05), which revealed impaired deactivation of DMPFC for patients during positive expectancy condition. We did not find any significant differences between the groups in other expectancy conditions (Figure 2b). During the positive vs. neutral expectancy condition, we correlated negative blood oxygenation level-dependent responses (NBRs) with HAMD and MADRS, but did not find any significant results (p<0.05, FWE cluster level corrected).

2. PPI results. During the positive vs. neutral expectancy condition, compared with HCs, MDD patients had increased FC from DMPFC towards posterior cingulate cortex (PCC) (Figure 3a, p < 0.05, FWE cluster level corrected, T = 4.7, x = -9, y = -36, z = 45, k = 97) and parieto-occipital cortex (Figure 3b, p < 0.05, FWE cluster level corrected, T = 5.15, x = -24, y = -60, z = 21, k = 60). Figure 3c show the mean PPI value between MPFC and PCC during positive vs. neutral expectancy, which revealed there is decoupling between MPFC and PCC for positive vs. neutral expectancy in controls, not in MDD patients. During the positive vs. neutral expectancy condition, we correlated FC with HAMD and MADRS, but did not find any significant results (p<0.05, FWE cluster level corrected).

3. PPI-SHAPS correlation. During the positive vs. neutral expectancy condition, patients exhibited a positive correlation between PPI (DMPFC and PCC) and anhedonia as measured via SHAPS (p<0.05, FWE-SVC, peak level corrected, T = 3.61). This result indicated lowest decoupling in patients with high SHAPS scores during positive compared to neutral expectancy.

Discussion 10

The present study found that the anterior part of DMN (DMPFC)—a region previously associated with abnormal attention proceeding emotional pictures in MDD patients—was also related to emotional expectancy dysfunction. More specifically, positive (but not negative or neutral) expectancy showed reduced deactivation in MDD patients compared to HCs. PPI analyses revealed significant MDD-related increased FC from DMPFC towards PCC and parieto-occipital cortex as a function of positive (vs. neutral) expectancy. This effect was driven by a significant decoupling between DMPFC and PCC during positive vs. neutral expectancy conditions in the HC group, which was absent in patients. To our knowledge, this is the first study to observe abnormal intra-DMN FC during specifically positive emotional expectancies in MDD patients. Previous studies indicated that DMPFC is involved in cognitive tasks such as the appraisal and expression of negative emotion (Etkin et al., 2011), internal cued conditions such as pleasant vs. unpleasant stimuli (Gusnard et al., 2001), expected vs. unexpected pictures (Walter et al., 2009b), action monitoring (Amodio and Frith, 2006), etc. Importantly for MDD, DMPFC is a key region during emotional attention processing, showing stronger activity when participants view positive compared with negative pictures (Bermpohl et al., 2006), and showing smaller signal decreases during cognitive tasks with high emotional load (Northoff et al., 2004). Therefore, it is unsurprising that abnormal BOLD responses in DMPFC are often related to mental disorders such as MDD. A previous clinical study found increased activity in the left DMPFC and PCC in healthy group during expected vs. unexpected emotional picture perception, but this effect was absent among MDD patients (Bermpohl et al., 2009). Insufficient deactivation in left DMPFC of MDD patients for positive picture perception was also observed in this study, suggesting that MDD patients showed DMPFC dysfunction in processing external positive information, leading to impaired preparation and thus reduced attention for positive events. However, in the above mentioned study, the researchers were not able to check the positive expectancy condition because the expectancy cues did not differentiate between positive and negative emotion. Our experiment, in contrast, examined valence during expectancy, 11

and abnormal positive expectancy activation was discovered in DMPFC among MDD patients. This finding provides evidence that MDD patients show DMPFC dysfunction in processing the expectancy of positive information. DMPFC is a core part of the DMN, and NBRs in the DMN has been related to rumination, autobiographical memory, self-referential processing, and various emotional-cognitive tasks (Fransson, 2005; Grimm et al., 2009; Grimm et al., 2006; Gusnard et al., 2001; McKiernan et al., 2003; Raichle et al., 2001; Whitfield-Gabrieli and Ford, 2012; Zhu et al., 2012). NBRs typically result from inhibiting resting-state activity during tasks due to goal-directed external-stimuli-related attentional resource allocation (Fox et al., 2005; Fransson, 2005; Greicius and Menon, 2004). Grimm and colleagues found that MDD patients showed abnormal NBRs in the anterior DMN—mostly ventromedial prefrontal cortex (VMPFC)—when performing emotion perception and emotion judgment tasks (Grimm et al., 2009). A follow-up study from (Grimm et al., 2011) found that MDD patients showed reduced NBRs in DMPFC (as well as VMPFC) during self-related judgment and passive viewing. Interestingly, they found a correlation between valence and self-relatedness ratings for positive (but not negative) emotional pictures in HCs, as opposed to MDD patients who only showed significant correlations for negative emotional pictures. Previous studies have revealed reduced mesolimbic responsiveness to positive material in depressed patients (Epstein et al., 2006; Keedwell et al., 2005), and Forbes and colleagues reported that depressed patients exhibit reduced neural response to reward anticipation (Forbes et al., 2006). In our study, we found that depressed participants exhibit greater activity in DMPFC in response to positive anticipation, because their attention is captured to a lesser extent by positive anticipation than negative or even neutral anticipations, suggesting their mind wanders to self-referential processing during the positive expectancy condition. We hypothesized that DMN network FC seeded in the DMPFC would be modulated by valence dependent expectancy. Our PPI results revealed a significant DMN decoupling between anterior DMN (DMPFC) and posterior DMN (PCC) as a function of positive vs. neutral expectancy in the HC group. However, this 12

positive-stimuli-induced decoupling was not present in MDD patients. A stronger resting state FC between posterior and anterior DMN in MDD patients—compared to HC participants—was identified in a study by (Berman et al., 2011), which also found positive correlations between rumination scores and intra-DMN FC. This study only focused on resting state FC, but a recent study reported similar FC (from subgenual anterior cingulate cortex to PCC) increases when MDD patients performed an externally focused task (Belleau et al., 2015). The same study also found task-related decreases FC in executive and salience networks, which may be a result of increased intra-DMN FC. In other words, in order to ‘normally’ perform a task requiring external focus—such as expecting a positive emotional picture when detecting an upwards-pointing cue—internally focused rumination has to be interrupted to reallocate attentional resources to the outside world; therefore rumination-related FC between anterior and posterior DMN has to be disconnected in order to adequately register external positive information. In MDD patients, however, obsessive negative rumination may prevent this intra-DMN decoupling, as indicated by our findings of strong resting state intra-DMN FC during positive expectancy (even stronger than neutral expectancy). These findings are in line with previous studies mentioned above, and further suggest that abnormal positive-valence-dependent DMPFC-PCC decoupling may contribute to rumination symptoms of MDD patients (Berman et al., 2011; Hamilton et al., 2011; Sheline et al., 2010). Severe depressive episodes are often associated with anticipatory anhedonia (Gorwood, 2008; Jensen et al., 2003; Kampe et al., 2003). We did not find any significant results for NBRs and FC correlations with HAMD or MARDS. However, we found a significant correlation between DMPFC PPI and an anhedonia scale (SHAPS) in MDD patients, consistent with the hypothesis of impaired attention toward external events with positive expectancies. Anhedonia research has flourished with the development of a basic literature describing partially dissociable neural systems for reward anticipation versus consummation (Knutson et al., 2001), as well as for learning reward cues. While anhedonia suggests reduced pleasure upon reward consumption, accruing evidence suggests that anhedonic depression may be more 13

related to blunted anticipatory pleasure (McClure et al., 2003; Yang et al., 2014; Zhang et al., 2015b). Impaired NBRs in the DMN has also been found to discriminate highly anhedonic MDD patients from patients with more mild anhedonia, with deficits in DMN activity down-regulation during perception of emotional stimuli being related to glutamatergic deficits in the MPFC (Walter et al., 2009a). In contrast to this previous study, here we found a deficit in the dorsal DMN that was apparent even before actual picture presentation. The previous association of anhedonia-related lack of DMN down-regulation during consummatory processes could thus be extended towards abnormal preparational mechanisms during anticipation of positive events. Our demonstration of correlations between severe anhedonia and reduced DMN decoupling provides yet further evidence suggesting that impairments in incorporating information about external positive events may be related to an impaired ability to down-regulate of intrinsically focused attention. We add evidence to previous clinical studies on both increased resting state FC and task-induced deactivations in MDD patients, by showing that both effects can be linked within the same paradigm. Our findings are somewhat novel compared with similar previous studies, since (Herwig et al., 2010) did not observe a DMPFC effect during pleasant vs. neutral expectancy, and (Bermpohl et al., 2009) only reported a valence-dependent effect during perception. One potential reason for these discrepancies may be due to differences in experimental paradigms. In Bermpohl et al, (2009) picture periods followed expectancies with fixed durations, which may have led to difficulties in disambiguating expectancy- and picture-related responses. Another potential reason for different expectancy-related findings may be that our sample size was slightly larger than the one reported in Bermpohl et al 2009 (n = 21 vs. n = 15 patients). Further, use of a 3 Tesla scanner and SENSE-SPIRAL sequence in our investigation may have increased our signal-to-noise ratio for detecting group differences. In contrast to these methodological differences, condition severity, gender distribution, and medication status all seem to comparable with previous participant characteristics, and thus may not be the primary sources of our novel findings. Nevertheless, it should be acknowledged that these observations are the first of their 14

kind in an Asian patient group. However, to our knowledge, potential differences in brain correlates of depression have never been systematically observed between either Chinese and German clinical samples or between Asian and Caucasian groups (Han and Northoff, 2008). As with all novel findings, the current study should be considered in light of potential limitations. Although the DMPFC PPI finding is in line with the previous work (hence providing a priori support), independent testing with larger sample sizes is needed. Further, given that our MDD patients had received different antidepressant treatments, future studies may benefit by attempting to control for these factors. In conclusion, this research is notable as the first use of an emotional expectancy task in Chinese MDD patients. Here, we identified impaired deactivation for positive valence-dependent expectancy in the DMPFC, a core region in the anterior DMN. Relative to HC participants, we observed reduced anterior to posterior DMN FC during positive expectancy. Further research should investigate the ways in which this altered DMN activity may be related to rumination and neglect of positive information in MDD patients.

References Amodio, D.M., Frith, C.D., 2006. Meeting of minds: the medial frontal cortex and social cognition. Nat Rev Neurosci 7, 268-277. Belleau, E.L., Taubitz, L.E., Larson, C.L., 2015. Imbalance of default mode and regulatory networks during externally focused processing in depression. Soc Cogn Affect Neurosci 10, 744-751. Belmaker, R.H., Agam, G., 2008. Major depressive disorder. The New England journal of medicine 358, 55-68. Berman, M.G., Peltier, S., Nee, D.E., Kross, E., Deldin, P.J., Jonides, J., 2011. Depression, rumination and the default network. Social cognitive and affective neuroscience 6, 548-555. Bermpohl, F., Pascual-Leone, A., Amedi, A., Merabet, L.B., Fregni, F., Gaab, N., Alsop, D., Schlaug, G., Northoff, G., 2006. Attentional modulation of emotional stimulus processing: an fMRI study using emotional expectancy. Human brain mapping 27, 662-677. Bermpohl, F., Walter, M., Sajonz, B., Lucke, C., Hagele, C., Sterzer, P., Adli, M., Heinz, A., Northoff, G., 2009. Attentional modulation of emotional stimulus processing in patients with major depression--alterations in prefrontal cortical regions. Neuroscience letters 463, 108-113. Bremner, J.D., Vythilingam, M., Vermetten, E., Nazeer, A., Adil, J., Khan, S., Staib, L.H., Charney, D.S., 2002. Reduced volume of orbitofrontal cortex in major depression. Biological psychiatry 51, 273-279. Clark, D.A., Steer, R.A., Beck, A.T., 1994. Common and specific dimensions of self-reported anxiety 15

and depression: implications for the cognitive and tripartite models. Journal of abnormal psychology 103, 645-654. Cullen, K.R., Westlund, M.K., Klimes-Dougan, B., Mueller, B.A., Houri, A., Eberly, L.E., Lim, K.O., 2014. Abnormal amygdala resting-state functional connectivity in adolescent depression. JAMA psychiatry 71, 1138-1147. Disner, S.G., Beevers, C.G., Haigh, E.A., Beck, A.T., 2011. Neural mechanisms of the cognitive model of depression. Nature reviews. Neuroscience 12, 467-477. Drevets, W.C., 2007. Orbitofrontal cortex function and structure in depression. Annals of the New York Academy of Sciences 1121, 499-527. Epstein, J., Pan, H., Kocsis, J.H., Yang, Y., Butler, T., Chusid, J., Hochberg, H., Murrough, J., Strohmayer, E., Stern, E., Silbersweig, D.A., 2006. Lack of ventral striatal response to positive stimuli in depressed versus normal subjects. The American journal of psychiatry 163, 1784-1790. Etkin, A., Egner, T., Kalisch, R., 2011. Emotional processing in anterior cingulate and medial prefrontal cortex. Trends Cogn Sci 15, 85-93. Forbes, E.E., Christopher May, J., Siegle, G.J., Ladouceur, C.D., Ryan, N.D., Carter, C.S., Birmaher, B., Axelson, D.A., Dahl, R.E., 2006. Reward-related decision-making in pediatric major depressive disorder: an fMRI study. Journal of child psychology and psychiatry, and allied disciplines 47, 1031-1040. Fox, M.D., Snyder, A.Z., Vincent, J.L., Corbetta, M., Van Essen, D.C., Raichle, M.E., 2005. The human brain is intrinsically organized into dynamic, anticorrelated functional networks. Proc Natl Acad Sci U S A 102, 9673-9678. Fransson, P., 2005. Spontaneous low-frequency BOLD signal fluctuations: an fMRI investigation of the resting-state default mode of brain function hypothesis. Hum Brain Mapp 26, 15-29. Friston, K.J., Buechel, C., Fink, G.R., Morris, J., Rolls, E., Dolan, R.J., 1997. Psychophysiological and modulatory interactions in neuroimaging. NeuroImage 6, 218-229. Gorwood, P., 2008. Neurobiological mechanisms of anhedonia. Dialogues in clinical neuroscience 10, 291-299. Gotlib, I.H., Krasnoperova, E., Yue, D.N., Joormann, J., 2004. Attentional biases for negative interpersonal stimuli in clinical depression. Journal of abnormal psychology 113, 121-135. Greicius, M.D., Flores, B.H., Menon, V., Glover, G.H., Solvason, H.B., Kenna, H., Reiss, A.L., Schatzberg, A.F., 2007. Resting-state functional connectivity in major depression: abnormally increased contributions from subgenual cingulate cortex and thalamus. Biological psychiatry 62, 429-437. Greicius, M.D., Menon, V., 2004. Default-mode activity during a passive sensory task: uncoupled from deactivation but impacting activation. J Cogn Neurosci 16, 1484-1492. Grimm, S., Boesiger, P., Beck, J., Schuepbach, D., Bermpohl, F., Walter, M., Ernst, J., Hell, D., Boeker, H., Northoff, G., 2009. Altered negative BOLD responses in the default-mode network during emotion processing in depressed subjects. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology 34, 932-943. Grimm, S., Ernst, J., Boesiger, P., Schuepbach, D., Boeker, H., Northoff, G., 2011. Reduced negative BOLD responses in the default-mode network and increased self-focus in depression. World J Biol Psychiatry 12, 627-637. Grimm, S., Schmidt, C.F., Bermpohl, F., Heinzel, A., Dahlem, Y., Wyss, M., Hell, D., Boesiger, P., Boeker, H., Northoff, G., 2006. Segregated neural representation of distinct emotion dimensions in the 16

prefrontal cortex-an fMRI study. Neuroimage 30, 325-340. Gusnard, D.A., Akbudak, E., Shulman, G.L., Raichle, M.E., 2001. Medial prefrontal cortex and self-referential mental activity: relation to a default mode of brain function. Proc Natl Acad Sci U S A 98, 4259-4264. Hamilton, J.P., Furman, D.J., Chang, C., Thomason, M.E., Dennis, E., Gotlib, I.H., 2011. Default-mode and task-positive network activity in major depressive disorder: implications for adaptive and maladaptive rumination. Biological psychiatry 70, 327-333. Han, S.H., Northoff, G., 2008. Culture-sensitive neural substrates of human cognition: a transcultural neuroimaging approach. Nature Reviews Neuroscience 9, 646-654. Herwig, U., Bruhl, A.B., Kaffenberger, T., Baumgartner, T., Boeker, H., Jancke, L., 2010. Neural correlates of 'pessimistic' attitude in depression. Psychol Med 40, 789-800. Holtzheimer, P.E., Mayberg, H.S., 2011. Stuck in a rut: rethinking depression and its treatment. Trends in neurosciences 34, 1-9. Jensen, J., McIntosh, A.R., Crawley, A.P., Mikulis, D.J., Remington, G., Kapur, S., 2003. Direct activation of the ventral striatum in anticipation of aversive stimuli. Neuron 40, 1251-1257. Kampe, K.K., Frith, C.D., Frith, U., 2003. "Hey John": signals conveying communicative intention toward the self activate brain regions associated with "mentalizing," regardless of modality. The Journal of neuroscience : the official journal of the Society for Neuroscience 23, 5258-5263. Keedwell, P.A., Andrew, C., Williams, S.C., Brammer, M.J., Phillips, M.L., 2005. The neural correlates of anhedonia in major depressive disorder. Biological psychiatry 58, 843-853. Kellough, J.L., Beevers, C.G., Ellis, A.J., Wells, T.T., 2008. Time course of selective attention in clinically depressed young adults: an eye tracking study. Behaviour research and therapy 46, 1238-1243. Knutson, B., Adams, C.M., Fong, G.W., Hommer, D., 2001. Anticipation of increasing monetary reward selectively recruits nucleus accumbens. The Journal of neuroscience : the official journal of the Society for Neuroscience 21, RC159. Kong, L., Chen, K., Tang, Y., Wu, F., Driesen, N., Womer, F., Fan, G., Ren, L., Jiang, W., Cao, Y., Blumberg, H.P., Xu, K., Wang, F., 2013. Functional connectivity between the amygdala and prefrontal cortex in medication-naive individuals with major depressive disorder. Journal of psychiatry & neuroscience : JPN 38, 417-422. Lang, P.J., Bradley, M.M., Cuthbert, B.N., 2008. International affective picture system (IAPS): Affective ratings of pictures and instruction manual. Technical Report A-8. MacLeod, C., Rutherford, E., Campbell, L., Ebsworthy, G., Holker, L., 2002. Selective attention and emotional vulnerability: assessing the causal basis of their association through the experimental manipulation of attentional bias. Journal of abnormal psychology 111, 107-123. Mathews, A., MacLeod, C., 2005. Cognitive vulnerability to emotional disorders. Annual review of clinical psychology 1, 167-195. McClure, S.M., Berns, G.S., Montague, P.R., 2003. Temporal prediction errors in a passive learning task activate human striatum. Neuron 38, 339-346. McKiernan, K.A., Kaufman, J.N., Kucera-Thompson, J., Binder, J.R., 2003. A parametric manipulation of factors affecting task-induced deactivation in functional neuroimaging. J Cogn Neurosci 15, 394-408. Nakonezny, P.A., Morris, D.W., Greer, T.L., Byerly, M.J., Carmody, T.J., Grannemann, B.D., Bernstein, I.H., Trivedi, M.H., 2015. Evaluation of anhedonia with the Snaith-Hamilton Pleasure Scale (SHAPS) 17

in adult outpatients with major depressive disorder. Journal of psychiatric research 65, 124-130. Northoff, G., Heinzel, A., Bermpohl, F., Niese, R., Pfennig, A., Pascual-Leone, A., Schlaug, G., 2004. Reciprocal modulation and attenuation in the prefrontal cortex: an fMRI study on emotional-cognitive interaction. Hum Brain Mapp 21, 202-212. Raichle, M.E., MacLeod, A.M., Snyder, A.Z., Powers, W.J., Gusnard, D.A., Shulman, G.L., 2001. A default mode of brain function. Proceedings of the National Academy of Sciences of the United States of America 98, 676-682. Shafritz, K.M., Collins, S.H., Blumberg, H.P., 2006. The interaction of emotional and cognitive neural systems in emotionally guided response inhibition. NeuroImage 31, 468-475. Sheline, Y.I., Price, J.L., Yan, Z., Mintun, M.A., 2010. Resting-state functional MRI in depression unmasks increased connectivity between networks via the dorsal nexus. Proceedings of the National Academy of Sciences of the United States of America 107, 11020-11025. Snaith, R.P., Hamilton, M., Morley, S., Humayan, A., Hargreaves, D., Trigwell, P., 1995. A scale for the assessment of hedonic tone the Snaith-Hamilton Pleasure Scale. The British journal of psychiatry : the journal of mental science 167, 99-103. Truong, T.K., Song, A.W., 2008. Single-shot dual-z-shimmed sensitivity-encoded spiral-in/out imaging for functional MRI with reduced susceptibility artifacts. Magnetic resonance in medicine 59, 221-227. van Tol, M.J., Li, M., Metzger, C.D., Hailla, N., Horn, D.I., Li, W., Heinze, H.J., Bogerts, B., Steiner, J., He, H., Walter, M., 2014. Local cortical thinning links to resting-state disconnectivity in major depressive disorder. Psychological medicine 44, 2053-2065. Walter, M., Henning, A., Grimm, S., Schulte, R.F., Beck, J., Dydak, U., Schnepf, B., Boeker, H., Boesiger, P., Northoff, G., 2009a. The Relationship Between Aberrant Neuronal Activation in the Pregenual Anterior Cingulate, Altered Glutamatergic Metabolism, and Anhedonia in Major Depression. Archives of general psychiatry 66, 478-+. Walter, M., Matthia, C., Wiebking, C., Rotte, M., Tempelmann, C., Bogerts, B., Heinze, H.J., Northoff, G., 2009b. Preceding attention and the dorsomedial prefrontal cortex: process specificity versus domain dependence. Hum Brain Mapp 30, 312-326. Whitfield-Gabrieli, S., Ford, J.M., 2012. Default mode network activity and connectivity in psychopathology. Annu Rev Clin Psychol 8, 49-76. Yang, X.H., Huang, J., Zhu, C.Y., Wang, Y.F., Cheung, E.F., Chan, R.C., Xie, G.R., 2014. Motivational deficits in effort-based decision making in individuals with subsyndromal depression, first-episode and remitted depression patients. Psychiatry research 220, 874-882. Zhang, B., Li, M., Qin, W., Demenescu, L.R., Metzger, C.D., Bogerts, B., Yu, C., Walter, M., 2015a. Altered functional connectivity density in major depressive disorder at rest. European archives of psychiatry and clinical neuroscience. Zhang, B., Lin, P., Shi, H., Ongur, D., Auerbach, R.P., Wang, X., Yao, S., Wang, X., 2015b. Mapping anhedonia-specific dysfunction in a transdiagnostic approach: an ALE meta-analysis. Brain imaging and behavior. Zhu, X., Wang, X., Xiao, J., Liao, J., Zhong, M., Wang, W., Yao, S., 2012. Evidence of a dissociation pattern in resting-state default mode network connectivity in first-episode, treatment-naive major depression patients. Biol Psychiatry 71, 611-617.

Table 1. Demographic characteristics of study participants 18

HC

MDD

Group effect

(n = 25)

(n = 22)

P value

16/9

13/9

0.522

Gender (male/females) Age (years)

39.28 ±13.02

43.77 ±10.83

0.203

HAMD

-

28.05 ±8.04

-

MADRS

-

23.86 ±8.58

-

Legends Figure 1. Paradigm for task fMRI.

Figure 2. (a) Group difference of positive vs. neutral expectancy (Xpos > Xneu) between patient and control groups (p<0.05, cluster level FWE corrected). Patients showed significantly increased activity in the DMPFC (warm colors). (b) Mean betas of DMPFC during negative picture expectancy (Xneg), neutral picture expectancy (Xneu) and positive expectancy (Xpos). Error bars depict the standard error of the mean. * p<0.05.

Figure 3. PPI group difference between patient and control groups (p<0.05, cluster level FWE corrected). Patients showed significantly increased FC from DMPFC towards PCC (a) and parieto-occipital cortex (b).(c) Mean PPI values from DMPFC towards PCC for each group, error bars depict the standard error of the mean. Figure 4. (a) the positive effect of anhedonia level on the PPI (DMPFC and PCC) with the seed of PCC in patients with MDD (warm colors) during the positive vs. neutral expectancy (Xpos > Xneu). (b) Positive correlation between SHAPS and the PPI (DMPFC and PCC) in patients (P<0.05, small volume and FWE corrected).

19

20

Highlights MPFC activity during positive expectancy is abnormal in MDD Increased MPFC activity mirrors DMN hyperconnectivity Anhedonia correlates with DMN hyperconnectivity

21