Abstract
To navigate real-world listening conditions, the auditory system relies on the integration of multiple sources of information. However, to avoid inappropriate cross-talk between inputs, highly connected neural systems need to strike a balance between integration and segregation. Here, we develop a novel approach to examine how repeated neurochemical modules in the mouse inferior colliculus lateral cortex (LC) allow controlled integration of its multimodal inputs. The LC had been impossible to study via imaging because it is buried in a sulcus. Therefore we coupled two-photon microscopy with the use of a microprism to reveal the first-ever sagittal views of the LC to examine neuronal responses with respect to its neurochemical motifs under anesthetized and awake conditions. This approach revealed marked differences in the acoustic response properties of LC and neighboring nonlemniscal portions of the inferior colliculus. In addition, we observed that the module and matrix cellular motifs of the LC displayed distinct somatosensory and auditory responses. Specifically, neurons in modules demonstrated primarily offset responses to acoustic stimuli with enhancement in responses to bimodal stimuli, whereas matrix neurons showed onset response to acoustic stimuli and suppressed responses to bimodal stimulation. Thus, this new approach revealed that the repeated structural motifs of the LC permit functional integration of multimodal inputs while retaining distinct response properties.
Introduction
The brain must constantly integrate information from multiple senses—such as hearing and touch—without allowing these signals to become "tangled." This balance of integration and segregation is a fundamental challenge for neural systems. The inferior colliculus (IC) is a major hub in the auditory midbrain where this integration occurs. Specifically, the Lateral Cortex (LC) of the IC is known for its unique "patchy" structure, consisting of neurochemical modules embedded within a surrounding matrix.
However, the LC is physically buried deep within a brain fold (a sulcus), making it inaccessible to standard optical imaging. Until now, researchers have had to rely on electrophysiology—which makes it difficult to pinpoint exactly which structural motif a neuron belongs to—or static histology. This study introduces a novel method to overcome these anatomical constraints.
Methods
To "see around the corner" into the brain's folds, the researchers developed a surgical approach involving the implantation of a 45° microprism into the sulcus between the IC and the cerebellum. This redirected the laser of a two-photon microscope to capture sagittal views of the LC surface.
The study utilized hybrid mice expressing GAD67-GFP to identify GABAergic modules and jRGECO1a (a red calcium indicator) to monitor real-time neuronal activity. Neuronal responses were recorded while mice were presented with pure tones, amplitude-modulated (AM) noise, and somatosensory input (whisker deflections). Experiments were conducted in both anesthetized and awake mice to account for state-dependent processing differences.
Key Results
The microprism technique provided stable, high-resolution sagittal images of the LC, which were validated through histological markers and laser lesions. The findings revealed that the LC is not a uniform "processing soup" but is instead functionally segregated.
- Functional Segregation: Neurons within the GABAergic modules act as multimodal hubs. They are significantly more responsive to touch, often show "offset" responses to sound (firing when a sound stops), and exhibit enhancement when auditory and touch stimuli are presented together.
- Matrix Specialization: Matrix neurons are more sensitive to complex sounds (like AM noise) and typically show "onset" responses. Interestingly, they often show suppression during bimodal (sound + touch) stimulation.
- Lack of Tonotopy: Unlike the neighboring Dorsal Cortex, which has a neat "piano-key" arrangement of frequencies, the LC shows a fragmented and inconsistent frequency map.
- State-Dependent Shifts: The transition from anesthetized to awake states significantly altered response properties. For example, somatosensory responses in modules increased from ~14% to over 63% in awake preparations.
Discussion and Future Work
This study demonstrates that the repeated structural motifs of the LC allow for the functional integration of multimodal inputs while maintaining distinct response properties. However, some limitations remain. The authors note that the Thy1 promoter used for the calcium indicator might not express evenly across all cell types, potentially biasing recordings. Additionally, somatosensory input was limited to whisker deflections; future work could explore other inputs like proprioception.
The dramatic differences between awake and anesthetized data suggest that many previous findings in midbrain research may need re-evaluation in awake animals. Understanding the specific local circuitry connecting the modules and matrix remains a primary target for future investigation.
Summary updated