The visual system is a selective and primary target of acute alcohol effects (Esposito et al., 2010; Abrahao et al., 2017). Previous research has indicated various changes in visual perception caused by alcohol consumption, such as spatial frequency discrimination (Watten et al., 1998), contrast sensitivity (Nicholson et al., 1995; Cavalcanti-Galdino et al., 2014; Timney et al., 2016), eye accommodation and vergence (Hill and Toffolon, 1990), visual acuity (Wilson and Mitchell, 1983; Hill and Toffolon, 1990), or motion processing (MacArthur and Sekuler, 1982; Gummel et al., 2012). Primary visual cortex (V1) is the earliest cortical processing stage in the hierarchical organization of the visual system, and its circuit connectivity and neuronal response properties are well understood (Bijanzadeh et al., 2018). Neurophysiological research has reported the negative effects of acute alcohol exposure on the response properties of visual area 17 of cat (Chen et al., 2010), functional magnetic resonance imaging (fMRI) revealed a strong enhancement of spontaneous BOLD fluctuations in V1 in an acute alcoholic state (Esposito et al., 2010), and magnetoencephalography recordings (Campbell et al., 2014) showed that stimulus-induced Gamma oscillations in human V1 were also strongly modulated by alcohol consumption.

Surround suppression (SS) is a canonical cortical computation (Bijanzadeh et al., 2018), where stimuli beyond the classical receptive field (RF) tend to suppress neuronal responses of stimuli within the RF center with similar features. SS plays a crucial role in visual perception, for example in segmentation of object boundaries, visual saliency, and rapid figure-ground segmentation of moving objects (Petrov and McKee, 2006; Nurminen and Angelucci, 2014; Bijanzadeh et al., 2018; Tadin et al., 2019). Feedforward and feedback connections between V1 and higher brain regions, V1 intracortical horizontal connections, interlaminar connections, and disinhibitory circuits have been identified to account for the generation of SS (Angelucci et al., 2002, 2017; Angelucci and Bressloff, 2006; Shushruth et al., 2013; Bijanzadeh et al., 2018; Nurminen et al., 2018; Keller et al., 2020).

Interestingly, the SS change has been linked to an alteration in GABA-ergic inhibitory cortical function (Angelucci and Bressloff, 2006; Angelucci et al., 2017). Much weaker surround suppression effects were found in human subjects with decline in efficacy of cortical GABA-ergic inhibitory systems, such as patients with schizophrenia and major depression (Tadin et al., 2006; Golomb et al., 2009), while magnetic resonance spectroscopy has provided additional evidence (Yoon et al., 2010).

Tilt illusion (TI) is a type of SS effect, where the perceived orientation of the center target is biased by the simultaneously presented surround stimulus (Figure 1A). In particular, subjects strongly misperceive the physical orientation of the center target when the surround orientation has an angular difference between 0° and 50° (repulsion effect), while a systematic weaker effect is observed for differences around 75° (attraction effect) (example in Figure 1B; O’Toole and Wenderoth, 1977; Wenderoth and Smith, 1999; Takao et al., 2020). Several researchers have proposed that repulsive TI is caused by lateral inhibition from spatially arranged orientation hypercolumns of V1 neurons (Blakemor et al., 1970, 1973; Kapadia et al., 2000; Qiu et al., 2013; Takao et al., 2020) and successfully modeled with such assumptions (Gilbert and Wiesel, 1990; Bednar and Miikkulainen, 2000; Tzvetanov, 2012). On the other hand, the attractive TI is attributed to more global orientation mechanisms, related to higher-level extra-striatal orientation processing (Wenderoth and Johnstone, 1988; Smith and Wenderoth, 1999). Dynamic causal modeling of fMRI signals further suggested that perception of the repulsive TI reflects an intra-hemispheric integration mechanism in V1 (Song and Rees, 2018). Therefore, repulsive TI measures are considered, from psychophysics, physiological, and modeling perspectives to reflect the inhibitory response properties of V1.

Given that ethanol could potentiate GABA_A receptors and promotes the spontaneous release of GABA (Abrahao et al., 2017), this should increase overall inhibition. Therefore, we hypothesized that moderate alcohol consumption would alter visual perception related to V1 by increasing perceptual effects attributed to inhibition. We investigated this hypothesis by using the center-surround TI effect as a probe of inhibition changes, such that the potentiation of the inhibitory system should increase the magnitude of TI. Participants in our current research performed TI measurements under sober, alcohol intoxication [approximately 0.6 mg/ml of blood alcohol concentration (BAC)], and placebo conditions, with a within-subjects and double-blind experiment design.

In the following, we emphasize the results of the second TI measures because they were obtained around the peak intoxication level (see Wang et al., 2018), where we expected the strongest Intoxication effect (Figures 1C,D). Then, we briefly present the results of the first and third measurements, which provide qualitatively similar conclusions.

In this study, we investigated the putative effects of alcohol onto V1 inhibitory activity by using the center-surround repulsive TI effect as a probe of inhibitory contextual interactions. It was measured under acute moderate alcohol ingestion, and contrasted to sober and placebo conditions. We observed obvious increases in the amount of repulsive TI after alcohol ingestion together with the discrimination threshold of orientation, both effects strongly supporting increased inhibitory surround interactions.

For several decades lateral inhibition between orientation sensitive neurons in V1 was proposed to account for the repulsive TI (Blakemor et al., 1970, 1973; Georgeson, 1973; Gilbert and Wiesel, 1990; Takao et al., 2020). The psychophysics observations in the center-surround configuration are considered to reflect the particular organization of spatial lateral interactions in V1, where inhibition created by the presence of stimuli beyond the classical RF is very common (Shushruth et al., 2013; Angelucci et al., 2017). Because alcohol consumption increases inhibition strength in area V1 (Chen et al., 2010), we propose that it also increases lateral inhibition. This is consistent with our alcohol-related observations of stronger repulsive TI and increased overall orientation discrimination thresholds.

Previous research has investigated plausible lateral inhibition changes under alcohol intoxication by using psychophysical measurements of “Westheimer” functions (a probe of retinal center-surround interactions) or the Hermann grid illusion (Johnston and Timney, 2008, 2013). Although a full analysis with the signal detection theory of the Westheimer function under alcohol effects is still missing (Timney et al., 2016), the “Westheimer” function results (Johnston and Timney, 2008) seem to comfort our findings of increased lateral inhibition. For the Hermann grid illusion, where people see illusory black dots on the crossings of vertical and horizontal thin white stripes drawn over a black/gray background (or vice versa), its interpretation as a probe of retinal or subcortical lateral inhibition was clearly dismissed (Schiller and Carvey, 2005; Geier et al., 2008) and thus the reported results must be carefully considered. In contrast, our within-subject simple orientation discrimination task design, complete psychometric function measure, and well characterized center-surround inhibitory effect of TI provide sufficient and reliable evidence to prove that lateral inhibition is enhanced by alcohol in area V1.

One important aspect of our data is that it allows to discard specific explanations of alcohol induced repulsive TI due to higher-level effects of orientation processing or attentional changes targeting the exact conditions where the repulsive TI appears. TI patterns were systematically modulated by surround orientations consistently across all conditions, that means participants reliably represented individual perceptual sensitivities even under alcohol intoxication. The increase in perceptual bias only occurred at surround orientations of 15° and 30°, while the attractive TI effect was unchanged. The increased lapse rates across surround orientations indicated that subjects had global changes in attention to the task, and these “high cognitive” effects were unrelated to specific surround conditions. These results showed that deficits such as more global, higher-order, orientation processing is not visibly affected and deteriorated cognition, i.e., attention, represent generalized effects.

Another source of interference with our observations is that subjects might have an unstable fixation state in the alcohol condition, but that seems unlikely. Microsaccades are one of the main type of eye movements during visual fixation in humans (Martinez-Conde et al., 2004). Alcohol significantly increases saccade latency both in low (0.4 g/kg) and high (0.8 g/kg) dose in humans (Roche and King, 2010). After ethanol administration, monkeys showed diminution of the frequency of saccades and prolongation of fixation periods (Fuster et al., 1985). The acute alcohol ingestion increased the number and duration of fixations (mean and total) in humans during visual Maze test (Silva et al., 2017) and reading (Watten and Lie, 1997), and the mean eye fixation time was above 250 ms, longer than the stimulus presentation time (200 ms) in our experiment. From these reports, we concluded that eye movements are an unlikely source of specific TI increase in the alcohol condition.

One interesting feature in the TI effect is the strong correlation of the amount of TI bias and threshold of discrimination (Solomon and Morgan, 2006, 2009) at oblique surround angles. This phenomenon in the data led Solomon and Morgan (2006) to propose that both changes in perceived value and acuity (inverse of threshold) can be explained by a stochastic recalibration mechanism, which is indistinguishable from plausible increased internal noise (Solomon and Morgan, 2006). While this explanation is debatable due to the decrease of internal background noise by alcohol administration (Chen et al., 2010), we would like to emphasize that increased inhibition also allows to explain simultaneous changes in both variables, of perceived value and threshold (Tzvetanov and Womelsdorf, 2008, their Figure 5, for an application to motion direction). While the current analyses and experimental design may seem insufficient for disentangling between both possibilities, neurophysiological results (Chen et al., 2010) and literature reports (see section “Introduction”) hint to stronger inhibition.

Low-level based neuronal explanations of the TI effect showed that various population changes of center orientation tuning characteristics can contribute to the final TI effect: amplitude inhibition, tuning width change, shift of neuronal preferred orientation, etc. (Gilbert and Wiesel, 1990; Schwartz et al., 2007). In addition, orientation surround suppression measures, obtained with the probe of apparent contrast (Cannon and Fullenkamp, 1991) or more recently comparing human psychophysics and neurophysiology (Shushruth et al., 2013), hinted toward two distinct spatial mechanisms: one narrowly tuned that is spatially restricted and the other broadly tuned that is spatially widespread. From the reported alcohol effects onto neuronal tuning properties (Chen et al., 2010), we consider, for the moment, that the alcohol-enhanced repulsive TI effect comes from a stronger surround amplitude of inhibition due to alcohol, until further evidences are available.

From a larger perspective, the GABA-ergic inhibitory system plays critical roles in V1 functions and their putatively associated visual perceptions: GABA agonists improve V1 function in senescent monkeys (Leventhal et al., 2003); it increases responsiveness and controls response gain in V1 (Katzner et al., 2011); its modulation affects spatio-temporal contrast sensitivity of healthy subjects (Blin et al., 1993) and binocular rivalry in autism patients (Spiegel et al., 2019) that have reduced GABA-ergic action (Robertson et al., 2016). Among various inhibitory neuronal subclasses, activation of parvalbumin-expressing (PV) interneuron improves neuronal feature selectivity, perceptual discrimination and response reliability in V1 (Lee et al., 2012; Zhu et al., 2015), somatostain-expressing (SOM) interneuron in V1 sharpen neuronal feature selectivity and has contrast tuning (Wilson et al., 2012; Millman et al., 2020), and vasoactive intestinal peptide-expressing (VIP) interneuron enhances responses to weak but specific stimuli in V1 (Millman et al., 2020). These interneuronal subclasses also contribute to SS (Angelucci et al., 2017). For example, the disinhibitory circuit that consists of SOM and VIP neurons regulates contextual modulation in V1 (Keller et al., 2020). Recently, a research has linked the SS to the inhibitory-based neural responses in human V1, proved by enhanced Gamma oscillations which can reflect the activation of inhibitory neurons (Orekhova et al., 2020). In our study, we hypothesized to increase inhibition strength in human brain by moderate alcohol intoxication and thus to observe enhanced SS via the repulsive TI probe. Our results provide strong psychophysics evidence linking GABA-ergic inhibition in V1 to visual perception, consistent with the above research.

In summary, we think that our data altogether demonstrates that alcohol consumption induced stronger repulsive TI that reflects a specific change at very early neural stages of orientation processing, as V1. In line of the findings about enhancement of GABA_A receptors and spontaneous release of GABA by ethanol (Abrahao et al., 2017), we propose that the V1 GABA-ergic system could account for the observed visual perception changes reported in our study. These results combined with our previous report in motion domain (Wang et al., 2018) show that alcohol enhances surround suppression effects at various hierarchical stages of the visual system, which could be explained by increased inhibitory processing in motion- or orientation-sensitive areas.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Ethics Committee of the University of Science and Technology of China. The patients/participants provided their written informed consent to participate in this study.

ZW and HW designed behavioral experiments and performed experiments. TT and HW performed data analysis and wrote the manuscript. ZW, YZ, and TT provided project supervision and funds. All authors discussed and commented on the manuscript and critically reviewed content and approved the final version for publication.

TT is currently working for his own company NEUROPSYPHY Tzvetomir TZVETANOV EIRL. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.