Forty-four 8–12 week old adult male C57Bl/6 mice (Charles River Laboratories, Quebec, QC, USA) were used. Prior to surgery, mice were group housed in a temperature-controlled room on a 12 h light/dark cycle with ad libitum access to food and water. All procedures were performed in accordance with the guidelines of the Canadian Council on Animal Care (CCAC) and the University of Toronto Animal Care Committee. Surgical procedures were performed on mice under general anesthesia with isoflurane and mounted onto a stereotaxic frame. The atlas of the mouse brain by Paxinos and Franklin (2007) was used to determine coordinates for guiding the stereotaxic infusion of tracers.

For anterograde tracing experiments, AAV2/9-hSyn-hChR2(H134R)-eYFP (~10¹² GC/ml) purchased from the Vector Core at the University of Pennsylvania and a total of 18 mice were used. Two groups of six mice received unilateral infusions of 0.3 μL of channelrhodopsin-2 (ChR2) into the dorsal HPC (no angle; two sets: AP: −2.18 ML: ±2.10 DV: −1.75, n = 3 or AP: −2.70 ML: ±2.20 DV: −2.00, n = 3) or ventral HPC (10° angle away from midline; AP: −2.92 ML: ±2.15 DV: −4.90, n = 6). One group of six mice received a simultaneous co-injection of AAV2/5-hSyn-hChR2-mCherry (~10¹² GC/ml) and AAV2/5-hSyn-hChR2-eYFP into the dorsal (AP: −2.70 ML: ±2.20 DV: −2.00) and ventral HPC using a counterbalanced combination.

For the retrograde labeling experiments, a total of twenty four mice were used. Three groups of eight mice received a unilateral injection of 0.2 μL of red fluorescent retrobeads (LumaFluor Inc., Naples, FL, USA) into the mAON (10° angle toward midline; AP: +2.90 ML: ±1.10 DV: −3.42), dAON (no angle; two sets: AP: +2.80 ML: ±0.75 DV: −3.10, n = 4 or AP: +2.80 ML: ±1.5 DV: −3.10, n = 4), or lAON (no angle; AP: +3.20 ML: ±1.10 DV: −3.90), respectively. A lower injection volume was used compared to anterograde tracing experiments to limit spread of the infusion outside the boundaries of the AON and neighboring AON subregions. All infusions were made by means of pressure ejection at a rate of 0.1 μL/min through a cannula connected by Tygon tubing to a 10 μL Hamilton syringe (Hamilton, Reno, NV, USA) mounted onto an infusion pump. Following each infusion, the cannula was left in place for an additional 20 min before retraction to limit the spread of the tracer.

Approximately 3 weeks following surgery mice were transcardially perfused with Phosphate-Buffered Saline (PBS, pH 7.4), followed by 4% paraformaldehyde in phosphate buffer. Brains were extracted and postfixed overnight at 4°C and subsequently cryoprotected with PBS containing 30% sucrose. Coronal 40 μm thick sections were obtained using a cryostat (Leica, Germany). The sections were slide-mounted, counterstained with 4′,6-diamidino-2-phenylindole (DAPI) for 10 min, and coverslipped with Aquamount (Polysciences Inc., Warrington, PA, USA). The wide-field fluorescent images were captured with a 4× objective on a fluorescent microscope (Olympus, Japan). eYFP and red fluorescent retrobeads signals were captured using an U-MWIBA3 filter cube (Ex460- 495, Em510-550, DM505) and an U-MWIG3 filter cube (Ex530-550, Em575IF, DM570), respectively. Confocal images were captured through a Quorum spinning disk confocal microscope (Zeiss) using a 20× objective lens and were subsequently analyzed with Volocity Software (Perkin Elmer). eYFP and red fluorescent retrobeads signals were excited with 491 and 561 nm laser, respectively. Adobe Photoshop CS6 (Adobe Systems Incorporated, San Jose, CA, USA) was used to adjust the brightness and contrast of representative sections.

Retrobead-positive cell counting was performed using the cellSens software (Olympus, Japan). The HPC was divided into three portions such that the dorsal one-third consisted of the anterior portion of the HPC prior to its transition into its longitudinal form in coronal sections (approximately 2.46 mm posterior to Bregma). The remaining ventral two-third of the HPC was further divided into intermediate and ventral HPC separated by the rhinal fissure (rf). Calibration parameters were established using randomly chosen tissue sections and were selected such that the threshold for cell diameter was high enough to exclude the majority of falsely identified cells (noise). Calibration parameters were maintained for consistent counting of all tissue sections. Tissue sections from three animals in each of the medial, dorsal, and lateral-AON injected groups were analyzed. The density of retrobead-positive cells found across nine coronal sections (number of labeled cells in region of interest/area of region) for each subject, and mean densities for each group were calculated.

To examine the hippocampal innervation pattern of the AON, we first performed antero- grade tracing of hippocampal projections using virally delivered-ChR2 as an anterograde tracer (Gradinaru et al., 2009; Jennings and Stuber, 2015). AAV-hSyn-ChR2-YFP (ChR2 fused to eYFP driven by the human synapsin promoter) was infused into either the ventral (Figure 1Ai) or dorsal HPC (Figure 1Bi). Viral infusions targeting the ventral HPC led to expression of ChR2-YFP in the ventral half of the HPC below the rf, while dorsal HPC infusions led to ChR2-YFP expression in the dorsal half of the HPC above the rf (Figures 1Ai,Bi). In both cases, ChR2-YFP expression was restricted to the HPC with minimal spread into adjacent cortical structures such as the primary somatosensory cortex which does not project to the AON (Zakiewicz et al., 2014).

ChR2-YFP-positive axon fibers originating from the ventral half of the HPC were localized largely at the mAON with little signal observed in the OB, piriform cortex, or other AON subregions (Figure 1Aii). In contrast, projections from the dorsal half of the HPC were found predominantly in the dorsal, lateral, and ventral, but not medial, aspects of the AON, indicating a topographic organization of HPC inputs within the AON (Figure 1Bii). We confirmed this differential innervation pattern using two-color anterograde tracing by simultaneously expressing ChR2-YFP and ChR2-mCherry in the dorsal and ventral halves of the HPC, respectively (Figure 1Ci). Hippocampal projections established a gradient of innervation within the AON whereby YFP-positive and mCherry-positive HPC axon fibers partially overlapped at the junction between the dorsal and medial AON. The density of projections originating from the ventral half of the HPC gradually increased from the lateral to the medial AON whereas projections from the dorsal half of the HPC increased gradually from the medial to lateral AON (Figure 1Cii).

The ChR2-mediated anterograde tracing method cannot unambiguously distinguish hippocampal axon terminals from hippocampal fibers of passage at the AON. Furthermore, it remains unclear how the AON-projecting HPC cells are distributed along the dorso-ventral axis of the HPC. To overcome this limitation and further validate the topographic organization of hippocampal inputs, we performed retrograde tracing experiments by employing fluorescent retrogradely-transported microspheres (retrobeads) that are uptaken at presynaptic axon terminals. Retrobeads were microinjected unilaterally into the medial, dorsal, or lateral aspects of the AON in separate cohorts of mice (n = 4–8 for each group). The spread of injected retrobeads was restricted to the targeted AON subregions (Figures 2Ai–Di), which are denominated on the basis of their cardinal positions as previously described (Brunjes et al., 2005). In all cases, the majority of labeled cells were observed in the hemisphere ipsilateral to the injection site, with very few cells labeled in the contralateral HPC (data not shown).

Injection of retrobeads into the mAON yielded labeled cells found in their greatest density near the ventral-most part of the CA1/Subiculum subregion (Figures 2Aii, 3). These mAON-projecting cells in the ventral CA1 progressively decreased in number towards the dorsal end of the longitudinal axis, forming a gradient terminating at the level of the rf. In contrast, retrobead injections into the dAON produced the greatest density of labeled cells in the intermediate CA1 at or near the level of the rf, tapering gradually in both dorsal and ventral directions (Figures 2Bii,Cii, 3). Very few, if any, dAON-projecting cells were found at the ventral- and dorsal-most parts of the CA1. Finally, lAON injections produced retrobead-labeled cells which were found in their greatest density at the dorsal aspect of the intermediate CA1 and were progressively fewer towards the ventral end with the gradient terminating at the level of the rf (Figures 2Dii, 3). Notably, lAON injections yielded extremely sparse, if any, labeling in the dorsal one-third of the CA1 (Figure 2E); instead, labeled cells first appeared at the dorsal most extent of the ventral two-thirds of the CA1 (Bregma: ~ −2.06 to −2.54; Figure 2E), indicating that the dorsal CA1 is unlikely to represent a meaningful source of inputs to the AON. Taken together, our anterograde and retrograde tracing data revealed that hippocampal inputs originating from the ventral two-third of the HPC are topographically organized within the AON.