Topographic CA1 input shapes subicular spatial coding

Precise anatomical topographic mapping in the brain supports high-fidelity information transfer for computations such as sensory alignment and motor coordination.^1,2 As a brain structure central to learning, memory, and navigation, the hippocampal formation is likewise organized with striking topographic precision.^3,4,5,6,7,8 Within this circuit, the subiculum plays a critical role in forming the cognitive map by integrating spatial and non-spatial information from both CA1 and the entorhinal cortex (EC).^{9,10,11,12,13,14,15,16,17} Inputs from these regions follow a continuous topography such that the distal subiculum receives projections from proximal CA1 (pCA1) and medial EC (MEC), whereas the proximal subiculum receives projections from distal CA1 (dCA1) and lateral EC (LEC).^3,4,5,7,8 This anatomical architecture suggests that subicular computations may depend critically on the fidelity of these circuit topographies. However, because CA1 and EC provide partially overlapping spatial information, it remains unclear to what extent CA1-to-subiculum topography, independent of EC's contribution, shapes the physiological properties of subicular neurons. Here, by selectively disrupting CA1-to-subiculum topographic projections in latrophilin-2 conditional knockout mice, we show that precise topography shapes the anatomical distribution of subicular spatial coding while preserving single-cell tuning. Disrupted CA1 input topography also selectively impairs subicular boundary vector coding and long-term network stability. Thus, CA1 input topography is critical for organizing subicular spatial coding and network dynamics.