Decision-making often requires us to make choices between biologically relevant goods that differ across many dimensions, including some whose value may be largely contextual (i.e. I value water most when I am thirsty) or even highly subjective (i.e. I like blue). How the brain accomplishes such comparisons, maintaining the relative values of disparate goods across time, while adjusting to reflect new situations and even contingencies, is a fundamental question in cognitive neuroscience. Neuroeconomics holds that a major step in this process is the compression of the relevant features defining the different goods into a single dimension of value, so that disparate goods can be compared on a universal scale (Levy and Glimcher, 2012; Padoa-Schioppa, 2011; Padoa-Schioppa et al., 2006). Single-unit and fMRI studies have typically assigned this process to the orbitofrontal cortex (OFC) based on the demonstration that neural activity there – identified either by single-unit or BOLD response – tracks value on this universal scale during economic choice (Levy and Glimcher, 2011; McGinty et al., 2016; Padoa-Schioppa, 2009; Padoa-Schioppa and Assad, 2006; Padoa-Schioppa and Assad, 2008; Plassmann et al., 2010; Plassmann et al., 2007; Rich and Wallis, 2016; Tremblay and Schultz, 1999; Xie and Padoa-Schioppa, 2016). Yet these correlates are not ubiquitous; often single-units in orbitofrontal cortex do not seem to encode value independent of other information (Blanchard et al., 2015; Kennerley et al., 2011; Kennerley and Wallis, 2009; McDannald et al., 2014; Wikenheiser et al., 2017). Further, causal evidence to support a critical role for the OFC in economic choice is sparse, owing to the difficulty in doing such experiments in humans or monkeys, where economic choice is typically studied.

To address this, we recently duplicated in rats the economic choice task used in monkeys to isolate key neural correlates of economic value in the lateral OFC (Padoa-Schioppa and Assad, 2006). Although rats’ behavior in this task exhibited all the core features of economic choice (Levy and Glimcher, 2012; Padoa-Schioppa, 2011; Padoa-Schioppa et al., 2006), it was entirely insensitive to optogenetic inactivation of the lateral OFC, inactivation which was sufficient to disrupt devaluation-sensitive changes in Pavlovian responding (Gardner, 2017. Indeed recent work suggests that lateral OFC is not generally necessary for established choice behavior (Miller et al., 2018).

While we interpreted our result as consistent with the parallel support of economic choice by a number of neural circuits, an alternative explanation is that the lateral OFC is simply the wrong part of OFC. The OFC is increasingly recognized as consisting of several subregions (Heilbronner et al., 2016; Izquierdo, 2017; Murphy and Deutch, 2018; Rudebeck and Murray, 2011a; Walton et al., 2011), and medial OFC is implicated in signaling common currency value by human imaging studies (Levy and Glimcher, 2011; Plassmann et al., 2010; Plassmann et al., 2007) and in mediating value-guided behaviors in rats (Bradfield et al., 2015; Münster and Hauber, 2017), monkeys (Noonan et al., 2010) and humans (Camille et al., 2011; Noonan et al., 2017). Here we tested whether the medial region might be the critical site in OFC responsible for calculating economic value by optogenetically inactivating this region in rats engaged in the aforementioned task.

To test whether medial OFC is critical for economic choice behavior we inactivated it while rats performed an economic choice task, developed previously (Gardner et al., 2017) to mimic procedures used in monkeys (Padoa-Schioppa and Assad, 2006). In this task, rats learned to choose between different numbers of flavored food pellets, providing the opportunity to precisely measure each rats’ relative preferences between different goods. On each trial within the task, an offer was displayed on a pair of 3.5’ touchscreens situated on either side of a central nosepoke. Each offer consisted of two symbols that conveyed two pieces of information: 1) the pellet flavors in the offer (banana, bacon, grain, chocolate, grape or cellulose), indicated by the shape of each symbol, and 2) the number of pellets of each flavor in the offer (1,2,3,4,6 and sometimes 8), indicated by the segmentations of each symbol (Figure 1A). Rats were required to nosepoke for 1s while the offer was displayed, after which they were able to make a choice by touching the screen with the preferred option. The chosen pellets were subsequently delivered in a food magazine situated on the opposite wall of the operant chamber. Each session consisted of 11 offers in which two food pellet flavors were presented in different ratios. Rats learned six visual cue –pellet pairings, and experienced the 15 possible pairings of the six pellet-types across training and testing.

Figure 1

Download asset Open asset

We have previously demonstrated that within this task rats display rational economic choice behavior (Gardner et al., 2017). Within individual sessions, rats showed a clear change in preference for the two pellet-types across the range of offers given. The offer for which rats showed no preference for the two offer-types, referred to as the indifference point (IP), was used as an estimate of the relative valuation of the two pellet flavors. Notably, indifference points between particular pairs of food pellets for individual rats ranged widely from 1:1 to greater than 8:1, suggesting that the rats were choosing based on internal subjective values (only IPs <= 6 were included in analyses). Further these indifference points were stable across sessions and exhibited transitivity across offer pairs with shared constituents, a property often taken as an indicator of rational decision making.

Given that medial OFC has been suggested to represent economic value, we expected that these hallmark features of economic decision making might be disrupted if it were taken offline in the critical choice phase of each trial. To that end, we prepared rats so that we could transiently inhibit neural activity in medial OFC using the optogenetic inhibitory protein, halorhodopsin (NpHR3.0). Rats (n = 6) were first trained on the task before undergoing surgery. AAV-CaMKIIa-eNpHR3.0-eYFP was then infused into medial OFC bilaterally, and fiberoptic probes were implanted just above each injection site (Figure 1C). Following surgery and a 2 – 3 week postoperative recovery period, rats were retrained to pre-surgical performance levels (~5 weeks, see Materials and methods), then we began sessions in which we tested the effects of inactivating medial OFC.

This phase of the experiment, lasting ~8 weeks, was designed as follows: rats experienced three separate pellet-pairs for three consecutive days for a total of 9 test sessions (Figure 1B). Rats experienced each pair in an initial warm up session that was followed by two test sessions in which they were connected to either a dummy cable, which was blocked at the ferrule to prevent light delivery to the brain, or a patent light-transmitting cable, which was unblocked at the ferrule allowing light to enter the brain. Cable type was counterbalanced across days. Each session was also separated into trials in which the laser was either on or off. During laser-on trials, green light (16 – 18 mW, 532 nm) was delivered to the fiberoptic cables. Laser onset co-occurred with the onset of the white noise that signaled that a trial could be initiated. The laser was terminated when the rat contacted one of the touch screens, and termination included a period of 300 ms during which the power was reduced linearly, a protocol which has been shown to minimize or prevent rebound excitation (Chuong et al., 2014). Note, this design was identical to our previous experiment in which we inactivated the lateral OFC, except that we did not include a group injected with virus lacking the coding region of the halorhodopsin chloride pump. Because the behavior of the eYFP group in the prior study was identical to that of the experimental group and given the presence of both blocked-fiber control sessions and no-laser control trials, we deemed this third, between-subject control to be superfluous and hence unnecessary for the current experiment.

An initial examination of data from example sessions (Figure 2A, blocked and patent cable sessions from consecutive days on the same pellet pair) revealed very little difference in choice behavior between the laser-on and –off trials, and the averaged choice behavior (Figure 2B) supported this assessment. To look more closely, we quantified the effect of medial OFC inactivation on three separate primary behavioral measures: (1) the indifference point or relative preferences between pellet-types; (2) the steepness of the preference curves; and (3) transitivity or the consistency of preferences across multiple pellet-pairs experienced over a series of days.

Figure 2 with 1 supplement see all

Download asset Open asset

If the medial OFC is necessary for the assignment of value to particular goods or outcomes with the quantity of those goods modulating the decision elsewhere, one might expect a shift of the IP towards a 1:1 preference. To determine whether this occurred with medial OFC inhibition, we ran a two-factor repeated measures ANOVA on 50 pairs of sessions with factors Laser (laser-on/-off) and Fiber (blocked/patent). This analysis revealed no significant main effects of either Laser (F(1,147) = 1.31, p=0.25) or Fiber (F(1,147) = 1.03, p=0.31) nor any significant effect of the critical interaction term Laser*Fiber (F(1,147) = 0.75, p=0.39). In addition, at the reviewers’ request, we also compared the current groups to eYFP controls, identically trained and tested in a prior experiment, and there were no significant differences (Figure 3—figure supplement 1, see Supplementary file 1 for ANOVA results)(Gardner et al., 2017). We also ran post-hoc simple effects comparing levels across each factor of the design. None of the four comparisons reached significance (paired t-test, n = 50, p>0.14, β <1e-10 for a shift to 1:1); this included the primary comparison of interest, laser-on vs laser-off trials in the patent cable condition (p=0.90). The indifference point increased by only 0.3% during the laser-on trials (mean IP for the patent fiber and laser-off trials: 1.649, mean IP for laser-on trials: 1.65), an effect which is well within the confidence interval (CI: [1.56 1.74], which corresponds ±10.6% shift of the IP). Importantly, our test would have found about a one-tenth change in the average preference significant.

Since rats do not have strong preferences between some of the pairs of food pellets, it’s possible that mild preferences close to 1:1 might mask a significant IP shift towards a 1:1 ratio with medial OFC inactivation. To test this possibility, we removed sessions within a conservative 1.45:1 preference based on an empirical distribution of variance (see Materials and Methods, 19 session pairs were excluded) and reran the two-way ANOVA. Despite our focus on sessions with strong preferences (mean IP for all conditions: 2.07), this analysis still found neither significant main effects (Fiber: F(1,90) = 1.57, p=0.21; Laser: F(1,90) = 1.19, p=0.28) nor any interaction (F(1,90) = 0.87, p=0.35). We again tested the simple effects of the design of which none were significant (4 paired t-tests, n = 31, p>0.15, β <1e-10 for a shift to 1:1) including the primary comparison, laser-on vs laser-off trials in the patent cable condition, (p=0.90). The mean IP increased 0.40% from 2.03 (laser-off) to 2.04 (laser-on) with a 95% confidence interval of [1.83 2.25] or ±17.2% of the mean laser-off IP. The exclusion of sessions with IPs close to the 1:1 can be visualized in Figure 3D in which we progressively removed the closest IP to 1:1 from the dataset up to 90% of the sessions. The vertical line represents the threshold in which we performed the above ANOVA with 31 pairs of sessions.

Figure 3 with 1 supplement see all

Download asset Open asset

Figure 3—source data 1

Text File Containing the Source Data for Figure 3.

IP data is arranged as follows. Rows correspond to experimental replications. The first two columns (B-off and B-on) contain data for the blocked fiber with the laser-off and laser-on trials respectively. The latter two columns (P-off and P-on) contain data for the patent fiber with the laser-off and laser-on trials respectively.

https://doi.org/10.7554/eLife.38963.007

Download elife-38963-fig3-data1-v1.txt

Because the full course of inactivation spanned ~8 weeks, we wanted to be sure there was no effect of time influencing our measurements of inactivation of mOFC on the indifference point. Although beginning 7–8 weeks after the initial viral injections, which should allow ample time for expression of the light-sensitive chloride pump, the effect of inactivation might have evolved over the course of the experiment. To test this we split the data into thirds, coinciding with the design for the test of transitivity. Figure 2—figure supplement 1 shows the average behavior for each of the three blocks of sessions. It is apparent from visual inspection of the plots that there was no effect of inactivation at any stage of the inactivation. A three-way ANOVA with factors Laser, Fiber and Time revealed no significant effect of the critical three-way interaction of Laser*Fiber*Time (F(2,141) = 0.08, p=0.92), nor any of the two-way interactions (Laser*Time: F(2,141) = 1.58, p=0.21; Fiber*Time F(2,141) = 0.35, p=0.70; Laser*Fiber (F1,141)=0.73, p=0.40). There was a main effect of Time (F(1,141) = 5.97, p=5.0e-3), which is likely due to the variations in the preferences across the three pellets used for each of the three time blocks.

Effects of time could also be manifest within each session. To determine whether an effect of inactivation was apparent at the beginning of a behavioral session, we included the minimum number of trials for each offer for which we could determine a reasonable estimate of the IP (n = 4 for each of the 11 offers, for both laser-on and –off trial-types, 88 trials total) which was on average, the first 29.9% of the session. We then reran the two-way ANOVA test on this early session behavior. Neither the critical interaction effect of Laser*Fiber nor the main effects displayed a significant effect of inactivation (see Supplementary file 2 for ANOVA results).

We next tested whether inhibition of medial OFC might be increasing the variance or uncertainty of the value assessment on individual trials. This effect in isolation would affect the slope of the fit without impacting the IP. Histograms and boxplots of the inverse slope, or the variance estimate of the probit fit, are shown in Figure 4A–C.

A repeated measures two-way ANOVA on these measures revealed no significant main effect of either Laser (F(1,147) = 0.69, p=0.41) or Fiber (F(1,147) = 0.11, p=0.74) nor any significant effect of the critical interaction term Laser*Fiber (F(1,147) = 0.26, p=0.61). In addition, at the reviewers’ request, we also compared the current groups to eYFP controls, identically trained and tested in a prior experiment, and there were no significant differences (Figure 4—figure supplement 1, see Supplementary file 1 for ANOVA results)(Gardner et al., 2017). Simple effects within and across factors also showed no significant differences (4 paired t-tests, n = 50, p>=0.29). The variance estimate decreased 1.0% for the laser-on trials ($\widehat{\sigma }$=0.84) as compared to laser-off trials ($\widehat{\sigma }$=0.83) in the critical patent fiber condition. A change well within the 95% confidence interval [0.76 0.91]. We again tested whether the slope might be more affected by inactivation during sessions in which the IP is reasonably distant from equal or 1:1 preference (Figure 4D). We used the same empirical threshold, 1.45:1, that we used for testing the IPs resulting in 19 fewer session pairs entering into the analysis. We repeated the ANOVA and again found no significant effects (Laser: F(1,90) = 0.19, p=0.66; Fiber: F(1,90) = 0.74, p=0.39; Laser*Fiber: F(1,90) < 0.01 , p=0.98), and testing of the simple effects across conditions showed no effects as well (4 paired t-tests, n = 31, p>0.55). For the critical comparison in the patent fiber condition the change in the variance estimate between laser-off ($\widehat{\sigma }$ = 0.82) and laser-on ($\widehat{\sigma }$ = 0.80) was well within the CI: [0.47 0.90] and was even in the direction of a steeper slope rather than the predicted flattening of the slope.

Figure 4 with 1 supplement see all

Download asset Open asset

Figure 4—source data 1

Text File Containing the Source Data for Figure 4.

Slope data is arranged as follows. Rows correspond to experimental replications. The first two columns (B-off and B-on) contain data for the blocked fiber with the laser-off and laser-on trials respectively. The latter two columns (P-off and P-on) contain data for the patent fiber with the laser-off and laser-on trials respectively.

https://doi.org/10.7554/eLife.38963.010

Download elife-38963-fig4-data1-v1.txt

We also tested effects of time on the slope paralleling our analysis of the IP - we split the dataset into thirds, coinciding with the design for the test of transitivity (Figure 2—figure supplement 1). Again it is easily apparent from the figure that there was no emergent effect of inactivation over time. A three-way ANOVA with factors Laser, Fiber and Time revealed no significant effect of the critical three-way interaction of Laser*Fiber*Time (F(2,141) = 0.06, p=0.94), nor any of the two-way interactions (Laser*Time: F(2,141) = 0.04, p=0.96; Fiber*Time F(2,141) = 2.11, p=0.12; Laser*Fiber (F1,141)=0.27, p=0.61). Unlike the IP, there was no main effect of Time (F(1,141) = 0.44, p=0.64), further indicating this effect was due to variations in the preferences of the pellets used across the time blocks.

Finally, we looked at whether the slope was affected by inactivation just at the beginning of the session. As explained for the IP analysis, we narrowed our data to the first 4 trials of each offer (on average the first 29.9% of each session) and reran the two-way ANOVA test on this early session behavior. The critical interaction effect of Laser*Fiber was not significant. The main effect of Laser was significant. Further analysis revealed a significant difference in the simple effects between the laser-on and –off trials in the blocked fiber condition, but not in the patent fiber condition (see Supplementary file 2 for ANOVA results).

After determining that medial OFC inhibition did not result in any observable within-session differences, we next examined whether medial OFC inhibition affected the stability and thereby transitivity of the rats’ economic choice behavior across sessions. Rats show transitive behavior on our task when choosing between multiple pellet-types across days. This implies that rats have a consistent and relatively stable value space onto which they map different goods in our task. Medial OFC inactivation could disrupt this mapping, resulting in stable behavior during single sessions but disrupted transitive behavior across sessions. To assess this, we plotted transitivity from 15 sets of 3 pellet-pairs presented across 6 sessions (Figure 5B). An example of the behavior for one transitivity measure, spanning 6 sessions, is shown in Figure 5A. Each point on the scatterplots represents the consistency of pellet preferences from a group of three sessions. The greatest of the three preferences is plotted on the y-axis, ${\displaystyle I{P}_{A:C}}$, and the product of the other two pairs are plotted on the x-axis, ${\displaystyle I{P}_{A:B}\times I{P}_{B:C}}$. Stability of food preferences should be observed as points falling close to the identity line such that ${\displaystyle I{P}_{A:C}\approx I{P}_{A:B}\times I{P}_{B:C}}$ (Gardner et al., 2017); Padoa-Schioppa and Assad, 2008 ). Laser-on and laser-off trials from the same three-session pairs are connected by lines on the plot in order to visualize the effect of medial OFC inactivation on the same sessions. A comparison of these points for laser-on versus laser-off trials in blocked versus patent sessions demonstrates that the rats’ behavior was again generally transitive, with only small, seemingly random fluctuations between the laser-on and laser-off trial types. Critically the points and these small fluctuations were unaffected by whether or not the fiber was patent to allow inactivation. A table showing the IPs from all transitivity session sets is shown in Supplementary file 3. To confirm this, we performed a repeated-measures, two-way ANOVA comparing the single transitivity values; this analysis found neither main effects nor any interaction (Laser: F(1,42) = 0.54, p=0.47; Fiber: F(1,42) = 3.21, p=0.08; Laser*Fiber: F(1,42) = 0.54, p=0.47).

Figure 5

Download asset Open asset

Figure 5—source data 1

Text File Containing the Source Data for Figure 5 with the Same Layout as in Supplementary file 3.

SeeSupplementary file 3 for layout of data.

https://doi.org/10.7554/eLife.38963.012

Download elife-38963-fig5-data1-v1.txt

The previous analysis demonstrated that the transitive nature of the behavior did not change during medial OFC inhibition. We also wanted to test whether the paired distances of the transitivity points (laser-on vs laser-off trials), when plotted in the two dimensional space of the ${\displaystyle \log I{P}_{A:B}+\log I{P}_{B:C}}$ (x-axis in Figure 5B) and the ${\displaystyle \log I{P}_{A:C}}$ (y-axis in Figure 5B), increased with inhibition of medial OFC, as might happen if inhibition were increasing the variability of valuation in the goods space. These distances are plotted as the black lines connecting the paired dots in Figure 5B and the cumulative distribution of the lengths of these lines are shown in Figure 5C. To determine whether these distances were affected by inactivation, we ran a repeated measures ANOVA with the fiber-type as the single factor on the distances of the pairs of laser-on and –off trials and found no effect of Fiber: F(1,14) = 0.13, p=0.73.

Since we failed to find an effect of medial OFC inactivation on the economic choice task, we conducted two additional positive control experiments to be sure that the inactivation was working as expected. In one, we tested whether light would inhibit neurons in medial OFC in our subjects. After rats had completed behavioral testing, some rats (n = 3) were euthanized and slices were taken through medial OFC. We then performed whole cell current clamp recordings in neurons (n = 10) transfected with NpHR-YFP and assessed the effect of 593 nm light exposure (Figure 6A). Resting membrane potential was significantly hyperpolarized by 10 s light exposure (One-way ANOVA, F(1.06,9.49) = 26.42, p= 5.0e-4). The effect was maximal at the onset (peak) but maintained significant hyperpolarization in the steady state condition. In addition, repetitive firing activity elicited by a square pulse current injection of 400 pA was significantly inhibited by concurrent light exposure (Paired t-test, t(5) = 4.05, p=0.01). Following light exposure, direct current injection elicited fewer spikes per unit of current as evidenced by a rightward shift in the input-output curve (Figure 6B)(Two-way repeated measures ANOVA, F(11, 55)=4.3, p=1.0e-4).

Figure 6

Download asset Open asset

With the remaining rats (n = 4), we tested whether light would impair performance on a progressive ratio task previously shown to be medial OFC-dependent in rats and mice (Gourley et al., 2010; Münster and Hauber, 2017) Rats were trained to lever press for reward for 2 days on a fixed ratio one schedule (FR1), after which they were tested on a progressive ratio schedule (PR5) across four sessions, using either the blocked or patent fiber (counterbalanced; Figure 7A) with the laser activated on all trials. A repeated measures ANOVA (rat and run were used as blocking factors) performed on the breaking point, corresponding to the number of pellets achieved, revealed a significant main effect of Fiber (F(1,10) = 9.97, p=0.01). This effect was robust with 7 of the eight blocked/patent pairs trending in the same direction (Figure 7A left). An ANOVA with the same design yet performed on the total lever presses also revealed a significant main effect of Fiber (F(1,10) = 7.85, p=0.02) (Figure 7B right). Importantly prolonged inhibition did not result in decreased motor activity; Figure (7B top) shows histograms of the inter-press intervals (IPIs) for the two conditions in which there appears to be no discernable difference. A two sample Kolmogorov-Smirnov test on the cumulative distributions of the IPIs (Figure 7B bottom) revealed no significant effect (D(1436,1045) = 0.028, p=0.74). Lastly, we checked whether medial OFC inhibition affected the animals’ motivation to consume the food pellets. To do this, animals were given a bowl of 20 grams of food pellets in the conditioning box for 30 min to assess pellet consumption with and without inhibition. This test was repeated 4 times with blocked and patent cables (counterbalanced). A repeated measures ANOVA (with rat and run as blocking factors) showed no effect of inactivation on food consumption (F(1,10) = 0.20, p=0.66).

Figure 7

Download asset Open asset

In the current study, we employed a rodent version of the economic choice task used to demonstrate neural correlates of economic value in primates (Padoa-Schioppa and Assad, 2006) to test whether medial OFC makes a critical contribution to economic choice behavior. We found that transient optogenetic inactivation of medial OFC on randomly selected trials had no effect. Inactivation in the same rats altered performance on a progressive ratio task known to be sensitive to medial OFC manipulations, and post-mortem testing confirmed that medial OFC neurons in these rats exhibited viral expression and light-sensitive reduction in spiking activity. Together these data show that the medial OFC is not necessary for economic choice in well-trained rats.

The current study builds on our prior report that optogenetic inactivation of lateral OFC also had no discernable effect on economic choice in this setting (Gardner et al., 2017). Inactivation of lateral OFC was motivated by influential reports that neural activity in lateral OFC correlates with value during economic choice (McGinty et al., 2016; Padoa-Schioppa, 2009; Padoa-Schioppa and Assad, 2006; Padoa-Schioppa and Assad, 2008; Rich and Wallis, 2016; Tremblay and Schultz, 1999; Xie and Padoa-Schioppa, 2016), and the associated proposal that lateral OFC was central to normal decision-making in this context (Padoa-Schioppa, 2011). Implicit in this proposal is that economic value is equivalent to the value revealed by devaluation underlying so-called model-based behavior. Such behavior typically depends on this part of OFC in both rats and primates (Gallagher et al., 1999; Izquierdo et al., 2004; Machado and Bachevalier, 2007; Pickens et al., 2005; Pickens et al., 2003; Rudebeck et al., 2013; West et al., 2011). Yet we found that we could inactivate lateral OFC and abolish changes in behavior due to devaluation without any effect on economic choice behavior.

While this result did not support the lateral OFC as a critical, single bottleneck in which economic value is calculated, it left open whether other parts of the OFC might be performing this function. The OFC is now recognized to be a large area, potentially divisible into discrete subregions mediating discrete functions (Heilbronner et al., 2016; Izquierdo, 2017; Murphy and Deutch, 2018; Rudebeck and Murray, 2011b; Walton et al., 2011). Within this framework, the medial OFC is increasingly associated with representing value. While neural correlates of economic value have famously been reported in the spiking activity of single units in lateral OFC (McGinty et al., 2016; Padoa-Schioppa, 2009; Padoa-Schioppa and Assad, 2006; Padoa-Schioppa and Assad, 2008; Rich and Wallis, 2016; Tremblay and Schultz, 1999; Xie and Padoa-Schioppa, 2016), human imaging studies tend to find common currency correlates in the BOLD signal in medial OFC (Levy and Glimcher, 2011; Plassmann et al., 2010; Plassmann et al., 2007). Further, lesions to medial OFC in monkeys disrupt value-guided decisions whereas lateral OFC lesions on the same behavior cause deficits in other more subtle functions, such as credit-assignment (Noonan et al., 2010). Other studies have shown that medial OFC is necessary for appropriate valuation during acquisition of cue-outcome pairings (Baylis and Gaffan, 1991), and also following selective satiation of one outcome during a choice task (Rudebeck and Murray, 2011a). Damage to medial OFC in humans is associated with similar deficits, as well as changes in economic decision-making (Camille et al., 2011; Fellows and Farah, 2007; Noonan et al., 2017). Indeed, the comparable part of OFC in rats, located along the base of the medial wall (Heilbronner et al., 2016; Stalnaker et al., 2015), has recently been shown to be important for using information about outcomes to guide behavior when those outcomes are hidden (Bradfield et al., 2015), risky (Stopper and Floresco, 2014), or temporally delayed, although there are conflicting results for this latter finding (Mar et al., 2011).

Yet despite the promising background data, we again failed to find any effect of inactivating the OFC on economic choice. Rats exhibited similar preferences among groups of flavored food pellets with or without medial OFC online, at least as indexed by indifference points, the steepness of the curves relating the two pellets, and the transitivity of the choices. This was true despite a reduction in the progressive ratio breakpoint in the same rats with inactivation, an effect that would traditionally be interpreted as an effect on valuation. These results show that medial OFC is also not the critical bottleneck for calculating a single common currency value to guide economic choice.

Instead, these and the prior data point strongly to the possibility that economic value and the resultant decision-making depend on parallel processing in multiple regions (Hunt and Hayden, 2017). Such parallel processing would be predicted if the value judgement underlying economic choice, at least as assayed in this task, reflected a variety of associative information (e.g. Pavlovian and instrumental, model-based and model-free, habitual and goal-directed). This is likely because the economic choice task used here, and those employed in other settings, generally do not try to limit or control the associative information at play. Tasks that employ such controls have clearly demonstrated that these different types of associative information do not generally converge into a single common brain region, but instead are mediated by somewhat non-overlapping brain circuits, each capable of mediating behavior somewhat independently, particularly in simple or poorly-controlled settings. Our results are more consistent with this sort of distributed representation of value.

Of course, this does not mean that the processing demonstrated in this and similar tasks in lateral and medial OFC is not important for economic choice behavior. Rather it simply means that a better controlled form of the task is necessary in order to reveal the critical contributions these areas (and others) each make to the very general function of economic choice. For example, it has been suggested that orbital areas are critical when choices are novel, rapidly changing, or involve elements that have never been experienced – such as the evocative tea-jelly and snail-porridge prepared by Behrens and colleagues (Baylis and Gaffan, 1991; Rudebeck and Murray, 2011b; Fellows and Farah, 2007; Noonan et al., 2010); Chau et al., 2014; Barron et al., 2013). Viewed in this light, the dissociation here (and also in our prior study) between economic choice and OFC-dependent behaviors, such as progressive ratio and Pavlovian devaluation for lOFC, occurs because these behaviors require a form of inference or estimation that is not necessary in a well-trained subject in the standard economic choice task like that used here.

Importantly, while this model of economic value takes us away from one in which all information is compressed into a single common currency value in a given brain area, this idea may return to be relevant in a more limited way once the precise type of information determining the role of these areas in the task can be isolated. Economic choices will depend on a particular region to the extent the choice emphasizes the processing dependent on that area. In this regard, the role of different subregions within OFC in economic choice behavior would be identical to their roles in other value-based behaviors.

Data is available in Github (https://github.com/mphgardner/Econ-Choice-Task) as well as in the source data file for Figure 5.

1. 1. Barron HC
   2. Dolan RJ
   3. Behrens TE

   (2013) Online evaluation of novel choices by simultaneous representation of multiple memories

   Nature Neuroscience 16:1492–1498.

   https://doi.org/10.1038/nn.3515

   • PubMed
   • Google Scholar
2. 1. Baylis LL
   2. Gaffan D

   (1991) Amygdalectomy and ventromedial prefrontal ablation produce similar deficits in food choice and in simple object discrimination learning for an unseen reward

   Experimental Brain Research 86:617–622.

   https://doi.org/10.1007/BF00230535

   • PubMed
   • Google Scholar
3. 1. Blanchard TC
   2. Hayden BY
   3. Bromberg-Martin ES

   (2015) Orbitofrontal cortex uses distinct codes for different choice attributes in decisions motivated by curiosity

   Neuron 85:602–614.

   https://doi.org/10.1016/j.neuron.2014.12.050

   • PubMed
   • Google Scholar
4. 1. Bradfield LA
   2. Dezfouli A
   3. van Holstein M
   4. Chieng B
   5. Balleine BW

   (2015) Medial orbitofrontal cortex mediates outcome retrieval in partially observable task situations

   Neuron 88:1268–1280.

   https://doi.org/10.1016/j.neuron.2015.10.044

   • PubMed
   • Google Scholar
5. 1. Camille N
   2. Griffiths CA
   3. Vo K
   4. Fellows LK
   5. Kable JW

   (2011) Ventromedial frontal lobe damage disrupts value maximization in humans

   Journal of Neuroscience 31:7527–7532.

   https://doi.org/10.1523/JNEUROSCI.6527-10.2011

   • PubMed
   • Google Scholar
6. 1. Chau BK
   2. Kolling N
   3. Hunt LT
   4. Walton ME
   5. Rushworth MF

   (2014) A neural mechanism underlying failure of optimal choice with multiple alternatives

   Nature Neuroscience 17:463–470.

   https://doi.org/10.1038/nn.3649

   • PubMed
   • Google Scholar
7. 1. Chuong AS
   2. Miri ML
   3. Busskamp V
   4. Matthews GA
   5. Acker LC
   6. Sørensen AT
   7. Young A
   8. Klapoetke NC
   9. Henninger MA
   10. Kodandaramaiah SB
   11. Ogawa M
   12. Ramanlal SB
   13. Bandler RC
   14. Allen BD
   15. Forest CR
   16. Chow BY
   17. Han X
   18. Lin Y
   19. Tye KM
   20. Roska B
   21. Cardin JA
   22. Boyden ES

   (2014) Noninvasive optical inhibition with a red-shifted microbial rhodopsin

   Nature Neuroscience 17:1123–1129.

   https://doi.org/10.1038/nn.3752

   • PubMed
   • Google Scholar
8. 1. Fellows LK
   2. Farah MJ

   (2007) The role of ventromedial prefrontal cortex in decision making: judgment under uncertainty or judgment per se?

   Cerebral Cortex 17:2669–2674.

   https://doi.org/10.1093/cercor/bhl176

   • PubMed
   • Google Scholar
9. 1. Gallagher M
   2. McMahan RW
   3. Schoenbaum G

   (1999) Orbitofrontal cortex and representation of incentive value in associative learning

   The Journal of Neuroscience 19:6610–6614.

   https://doi.org/10.1523/JNEUROSCI.19-15-06610.1999

   • PubMed
   • Google Scholar
10. 1. Gardner MPH
    2. Conroy JS
    3. Shaham MH
    4. Styer CV
    5. Schoenbaum G

    (2017) Lateral orbitofrontal inactivation dissociates Devaluation-Sensitive behavior and economic choice

    Neuron 96:1192–1203.

    https://doi.org/10.1016/j.neuron.2017.10.026

    • PubMed
    • Google Scholar
11. Software

    1. Gardner MPH

    (2017) RatEcon Choice Task

    Github.

    https://github.com/mphgardner/RatEconChoiceTask
12. Software

    1. Gardner MPH

    (2018) Econ Choice Task

    Github.

    https://github.com/mphgardner/Econ-Choice-Task
13. 1. Gourley SL
    2. Lee AS
    3. Howell JL
    4. Pittenger C
    5. Taylor JR

    (2010) Dissociable regulation of instrumental action within mouse prefrontal cortex

    European Journal of Neuroscience 32:1726–1734.

    https://doi.org/10.1111/j.1460-9568.2010.07438.x

    • PubMed
    • Google Scholar
14. 1. Heilbronner SR
    2. Rodriguez-Romaguera J
    3. Quirk GJ
    4. Groenewegen HJ
    5. Haber SN

    (2016) Circuit-Based corticostriatal homologies between rat and primate

    Biological Psychiatry 80:509–521.

    https://doi.org/10.1016/j.biopsych.2016.05.012

    • PubMed
    • Google Scholar
15. 1. Hunt LT
    2. Hayden BY

    (2017) A distributed, hierarchical and recurrent framework for reward-based choice

    Nature Reviews Neuroscience 18:172–182.

    https://doi.org/10.1038/nrn.2017.7

    • PubMed
    • Google Scholar
16. 1. Izquierdo A
    2. Suda RK
    3. Murray EA

    (2004) Bilateral orbital prefrontal cortex lesions in rhesus monkeys disrupt choices guided by both reward value and reward contingency

    Journal of Neuroscience 24:7540–7548.

    https://doi.org/10.1523/JNEUROSCI.1921-04.2004

    • PubMed
    • Google Scholar
17. 1. Izquierdo A

    (2017) Functional heterogeneity within rat orbitofrontal cortex in reward learning and decision making

    The Journal of Neuroscience : The Official Journal of the Society for Neuroscience 37:10529–10540.

    https://doi.org/10.1523/JNEUROSCI.1678-17.2017

    • PubMed
    • Google Scholar
18. 1. Kennerley SW
    2. Wallis JD

    (2009) Evaluating choices by single neurons in the frontal lobe: outcome value encoded across multiple decision variables

    European Journal of Neuroscience 29:2061–2073.

    https://doi.org/10.1111/j.1460-9568.2009.06743.x

    • PubMed
    • Google Scholar
19. 1. Kennerley SW
    2. Behrens TE
    3. Wallis JD

    (2011) Double dissociation of value computations in orbitofrontal and anterior cingulate neurons

    Nature Neuroscience 14:1581–1589.

    https://doi.org/10.1038/nn.2961

    • PubMed
    • Google Scholar
20. 1. Levy DJ
    2. Glimcher PW

    (2011) Comparing apples and oranges: using reward-specific and reward-general subjective value representation in the brain

    Journal of Neuroscience 31:14693–14707.

    https://doi.org/10.1523/JNEUROSCI.2218-11.2011

    • PubMed
    • Google Scholar
21. 1. Levy DJ
    2. Glimcher PW

    (2012) The root of all value: a neural common currency for choice

    Current Opinion in Neurobiology 22:1027–1038.

    https://doi.org/10.1016/j.conb.2012.06.001

    • PubMed
    • Google Scholar
22. 1. Machado CJ
    2. Bachevalier J

    (2007) The effects of selective amygdala, orbital frontal cortex or hippocampal formation lesions on reward assessment in nonhuman primates

    European Journal of Neuroscience 25:2885–2904.

    https://doi.org/10.1111/j.1460-9568.2007.05525.x

    • PubMed
    • Google Scholar
23. 1. Mar AC
    2. Walker AL
    3. Theobald DE
    4. Eagle DM
    5. Robbins TW

    (2011) Dissociable effects of lesions to orbitofrontal cortex subregions on impulsive choice in the rat

    Journal of Neuroscience 31:6398–6404.

    https://doi.org/10.1523/JNEUROSCI.6620-10.2011

    • PubMed
    • Google Scholar
24. 1. McDannald MA
    2. Esber GR
    3. Wegener MA
    4. Wied HM
    5. Liu TL
    6. Stalnaker TA
    7. Jones JL
    8. Trageser J
    9. Schoenbaum G

    (2014) Orbitofrontal neurons acquire responses to 'valueless' Pavlovian cues during unblocking

    eLife 3:e02653.

    https://doi.org/10.7554/eLife.02653

    • PubMed
    • Google Scholar
25. 1. McGinty VB
    2. Rangel A
    3. Newsome WT

    (2016) Orbitofrontal cortex value signals depend on fixation location during free viewing

    Neuron 90:1299–1311.

    https://doi.org/10.1016/j.neuron.2016.04.045

    • PubMed
    • Google Scholar
26. Preprint

    1. Miller KJ
    2. Botvinick MM
    3. Brody CD

    (2018) Value representations in orbitofrontal cortex drive learning but not choice.

     bioRxiv.

    https://doi.org/10.1101/245720

    • Google Scholar
27. 1. Münster A
    2. Hauber W

    (2017) Medial orbitofrontal cortex mediates Effort-related responding in rats

    Cerebral Cortex pp. 1–11.

    https://doi.org/10.1093/cercor/bhx293

    • PubMed
    • Google Scholar
28. 1. Murphy MJM
    2. Deutch AY

    (2018) Organization of afferents to the orbitofrontal cortex in the rat

    Journal of Comparative Neurology 526:1498–1526.

    https://doi.org/10.1002/cne.24424

    • PubMed
    • Google Scholar
29. 1. Noonan MP
    2. Walton ME
    3. Behrens TE
    4. Sallet J
    5. Buckley MJ
    6. Rushworth MF

    (2010) Separate value comparison and learning mechanisms in macaque medial and lateral orbitofrontal cortex

    PNAS 107:20547–20552.

    https://doi.org/10.1073/pnas.1012246107

    • PubMed
    • Google Scholar
30. 1. Noonan MP
    2. Chau BKH
    3. Rushworth MFS
    4. Fellows LK

    (2017) Contrasting effects of medial and lateral orbitofrontal cortex lesions on credit assignment and Decision-Making in humans

    The Journal of Neuroscience 37:7023–7035.

    https://doi.org/10.1523/JNEUROSCI.0692-17.2017

    • PubMed
    • Google Scholar
31. 1. Padoa-Schioppa C
    2. Assad JA

    (2006) Neurons in the orbitofrontal cortex encode economic value

    Nature 441:223–226.

    https://doi.org/10.1038/nature04676

    • PubMed
    • Google Scholar
32. 1. Padoa-Schioppa C
    2. Jandolo L
    3. Visalberghi E

    (2006) Multi-stage mental process for economic choice in capuchins

    Cognition 99:B1–B13.

    https://doi.org/10.1016/j.cognition.2005.04.008

    • PubMed
    • Google Scholar
33. 1. Padoa-Schioppa C
    2. Assad JA

    (2008) The representation of economic value in the orbitofrontal cortex is invariant for changes of menu

    Nature Neuroscience 11:95–102.

    https://doi.org/10.1038/nn2020

    • PubMed
    • Google Scholar
34. 1. Padoa-Schioppa C

    (2009) Range-adapting representation of economic value in the orbitofrontal cortex

    Journal of Neuroscience 29:14004–14014.

    https://doi.org/10.1523/JNEUROSCI.3751-09.2009

    • PubMed
    • Google Scholar
35. 1. Padoa-Schioppa C

    (2011) Neurobiology of economic choice: a good-based model

    Annual Review of Neuroscience 34:333–359.

    https://doi.org/10.1146/annurev-neuro-061010-113648

    • PubMed
    • Google Scholar
36. 1. Pickens CL
    2. Saddoris MP
    3. Setlow B
    4. Gallagher M
    5. Holland PC
    6. Schoenbaum G

    (2003) Different roles for orbitofrontal cortex and basolateral amygdala in a reinforcer devaluation task

    The Journal of Neuroscience 23:11078–11084.

    https://doi.org/10.1523/JNEUROSCI.23-35-11078.2003

    • PubMed
    • Google Scholar
37. 1. Pickens CL
    2. Saddoris MP
    3. Gallagher M
    4. Holland PC

    (2005) Orbitofrontal lesions impair use of cue-outcome associations in a devaluation task

    Behavioral Neuroscience 119:317–322.

    https://doi.org/10.1037/0735-7044.119.1.317

    • PubMed
    • Google Scholar
38. 1. Plassmann H
    2. O'Doherty J
    3. Rangel A

    (2007) Orbitofrontal cortex encodes willingness to pay in everyday economic transactions

    Journal of Neuroscience 27:9984–9988.

    https://doi.org/10.1523/JNEUROSCI.2131-07.2007

    • PubMed
    • Google Scholar
39. 1. Plassmann H
    2. O'Doherty JP
    3. Rangel A

    (2010) Appetitive and aversive goal values are encoded in the medial orbitofrontal cortex at the time of decision making

    Journal of Neuroscience 30:10799–10808.

    https://doi.org/10.1523/JNEUROSCI.0788-10.2010

    • PubMed
    • Google Scholar
40. 1. Rich EL
    2. Wallis JD

    (2016) Decoding subjective decisions from orbitofrontal cortex

    Nature Neuroscience 19:973–980.

    https://doi.org/10.1038/nn.4320

    • PubMed
    • Google Scholar
41. 1. Rudebeck PH
    2. Murray EA

    (2011a) Balkanizing the primate orbitofrontal cortex: distinct subregions for comparing and contrasting values

    Annals of the New York Academy of Sciences 1239:1–13.

    https://doi.org/10.1111/j.1749-6632.2011.06267.x

    • PubMed
    • Google Scholar
42. 1. Rudebeck PH
    2. Murray EA

    (2011b) Dissociable effects of subtotal lesions within the macaque orbital prefrontal cortex on reward-guided behavior

    Journal of Neuroscience 31:10569–10578.

    https://doi.org/10.1523/JNEUROSCI.0091-11.2011

    • PubMed
    • Google Scholar
43. 1. Rudebeck PH
    2. Saunders RC
    3. Prescott AT
    4. Chau LS
    5. Murray EA

    (2013) Prefrontal mechanisms of behavioral flexibility, emotion regulation and value updating

    Nature Neuroscience 16:1140–1145.

    https://doi.org/10.1038/nn.3440

    • PubMed
    • Google Scholar
44. 1. Schweimer J
    2. Hauber W

    (2005) Involvement of the rat anterior cingulate cortex in control of instrumental responses guided by reward expectancy

    Learning & Memory 12:334–342.

    https://doi.org/10.1101/lm.90605

    • PubMed
    • Google Scholar
45. 1. Stalnaker TA
    2. Cooch NK
    3. Schoenbaum G

    (2015) What the orbitofrontal cortex does not do

    Nature Neuroscience 18:620–627.

    https://doi.org/10.1038/nn.3982

    • PubMed
    • Google Scholar
46. 1. Stopper CM
    2. Floresco SB

    (2014) What's better for me? Fundamental role for lateral habenula in promoting subjective decision biases

    Nature Neuroscience 17:33–35.

    https://doi.org/10.1038/nn.3587

    • PubMed
    • Google Scholar
47. 1. Takahashi YK
    2. Chang CY
    3. Lucantonio F
    4. Haney RZ
    5. Berg BA
    6. Yau HJ
    7. Bonci A
    8. Schoenbaum G

    (2013) Neural estimates of imagined outcomes in the orbitofrontal cortex drive behavior and learning

    Neuron 80:507–518.

    https://doi.org/10.1016/j.neuron.2013.08.008

    • PubMed
    • Google Scholar
48. 1. Tremblay L
    2. Schultz W

    (1999) Relative reward preference in primate orbitofrontal cortex

    Nature 398:704–708.

    https://doi.org/10.1038/19525

    • PubMed
    • Google Scholar
49. 1. Walton ME
    2. Behrens TE
    3. Noonan MP
    4. Rushworth MF

    (2011) Giving credit where credit is due: orbitofrontal cortex and valuation in an uncertain world

    Annals of the New York Academy of Sciences 1239:14–24.

    https://doi.org/10.1111/j.1749-6632.2011.06257.x

    • PubMed
    • Google Scholar
50. 1. West EA
    2. DesJardin JT
    3. Gale K
    4. Malkova L

    (2011) Transient inactivation of orbitofrontal cortex blocks reinforcer devaluation in macaques

    Journal of Neuroscience 31:15128–15135.

    https://doi.org/10.1523/JNEUROSCI.3295-11.2011

    • PubMed
    • Google Scholar
51. 1. Wikenheiser AM
    2. Marrero-Garcia Y
    3. Schoenbaum G

    (2017) Suppression of ventral hippocampal output impairs integrated orbitofrontal encoding of task structure

    Neuron 95:1197–1207.

    https://doi.org/10.1016/j.neuron.2017.08.003

    • PubMed
    • Google Scholar
52. 1. Xie J
    2. Padoa-Schioppa C

    (2016) Neuronal remapping and circuit persistence in economic decisions

    Nature Neuroscience 19:855–861.

    https://doi.org/10.1038/nn.4300

    • PubMed
    • Google Scholar

Author details

1. Matthew PH Gardner

   NIDA Intramural Research Program, Baltimore, United States

   Contribution

   Conceptualization, Resources, Formal analysis, Supervision, Funding acquisition, Methodology, Writing—original draft, Project administration, Writing—review and editing

   For correspondence

   matthew.gardner2@nih.gov

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-9146-5043

2. Jessica C Conroy

   NIDA Intramural Research Program, Baltimore, United States

   Contribution

   Conceptualization, Data curation, Software, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing

   Competing interests

   No competing interests declared

3. Clay V Styer

   NIDA Intramural Research Program, Baltimore, United States

   Contribution

   Data curation, Software Development, Investigation, Methodology

   Competing interests

   No competing interests declared

4. Timothy Huynh

   NIDA Intramural Research Program, Baltimore, United States

   Contribution

   Software, Investigation

   Competing interests

   No competing interests declared

5. Leslie R Whitaker

   NIDA Intramural Research Program, Baltimore, United States

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

6. Geoffrey Schoenbaum

   1. NIDA Intramural Research Program, Baltimore, United States
   2. Department of Anatomy & Neurobiology, University of Maryland School of Medicine, Baltimore, United States
   3. Solomon H. Snyder Department of Neuroscience, The Johns Hopkins University, Baltimore, United States
   4. Department of Psychiatry, University of Maryland School of Medicine, Baltimore, United States

   Contribution

   Resources, Data curation, Formal analysis, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing

   For correspondence

   geoffrey.schoenbaum@nih.gov

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-8180-0701

• 1,809

  views

• 254

  downloads

• 26

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.