In our LPFC recordings, the frequency of recording from such neurons appeared to be much higher (46% of the total number of modulated cells during the BFS condition; n = 54/118) than the respective percentage observed in cortical areas lower in the visual hierarchy, like V1 (10%, n = 10/104; Keliris et al., 2010), V4 (30%, n = 8/26; Leopold and Logothetis,

1996), and MT (26%, n = 12/46; Logothetis and Schall, 1989). However, encountering such cells is most likely the result of weak and variable stimulus preferences. In our data, when interaction effects (Stimulus × Condition) were explicitly tested using an ANOVA, only 2% of cells (n = 10/577) were found to be significantly modulated LY2157299 (p < 0.05) only during BFS. We also studied whether local cortical processing reflected in the local population spiking activity within a prefrontal cortical site could represent subjective visual perception. When nonsorted multiunit spiking activity was examined (MUA, i.e., the sum of the spikes recorded from a tetrode before spike sorting), we BKM120 found further evidence that the spiking activity of neuronal populations in the LPFC follows reliably phenomenal perception. Our results show that 20% of the total number of recorded sites (n = 42/211) were significantly modulated during physical

alternation (Wilcoxon rank-sum test, p < 0.05). In the large majority of these sites, MUA was also found to be significantly modulated during BFS (n = 31/42, or 74%). During BFS, sensory preference was retained in 94% (n = 29/31) of these sites, and in only 6% of the sites (n = 2/31), neuronal discharges were found to reverse their preference and increased their firing rate when a preferred stimulus was perceptually suppressed (ANOVA, Mannose-binding protein-associated serine protease Stimulus × Condition interaction effect, p < 0.05). We compared the magnitude of sensory and perceptual modulation for the 42 MUA sites found to be significantly modulated

during physical alternation (Figure 2C). We found that MUA modulation during BFS was significantly decreased and reached 56% of the modulation observed during physical stimulus alternation (d′sensory MUA = 1.1 ± 0.14 and d′perceptual MUA = 0.62 ± 0.13, Wilcoxon rank-sum test p = 0.019). Both distributions were significantly different from zero (t test, p < 10−9 for d′sensory MUA, p < 10−6 for d′perceptual MUA), thus indicating that the level of mean perceptual modulation was also adequate to distinguish between preferred and nonpreferred stimuli during subjective perception. Similar to SUA, we found that in cases where MUA exhibited a particularly strong sensory modulation (d′sensory MUA > 1), the percentage of perceptually modulated recording sites was even higher. In such strongly modulated cases, 92% of the sensory selective recording sites (n = 12/13) were also significantly modulated during BFS (Figure 2D).