TRACR: an anterograde transneuronal tracing system for genetic access across synapses and longitudinal circuit analysis

Understanding how neural signals travel through the brain's intricate web of connections is a cornerstone of modern neuroscience. To map these circuits, researchers have traditionally relied on transneuronal tracers—biological tools that move from one neuron to its connected neighbor to leave a visible mark. However, existing methods face a fundamental trade-off: they are often either permanent "snapshots" that cannot track changes over time, or highly toxic viral agents that kill the very cells they are meant to study.

A new study introduces TRACR (TRanssynaptic Anterograde Circuit Readout), a system that moves away from the physical transfer of materials and toward the transfer of information. Instead of injecting a virus that must physically crawl across a synapse to infect a new cell, TRACR uses a synthetic signaling mechanism. This tells the postsynaptic neuron to turn on a reporter gene. This shift allows for a reversible, non-toxic way to monitor how circuits assemble, degenerate, and repair themselves in living organisms.

The limitations of physical material transfer

Mapping the "direction" of information flow—knowing which neuron sends a signal and which receives it—requires anterograde tracing. Historically, this has been achieved using engineered viruses like Herpes Simplex Virus (HSV) or Yellow Fever Virus (YFV-17D). While powerful, these tools suffer from several systemic failures.

First, they are often cytotoxic (toxic to cells). The process of a virus replicating and moving across a synapse can damage or kill the neurons. This limits the window for long-term observation. Second, they struggle with genetic segregation. It is difficult to manipulate the "sender" neuron without accidentally affecting the "receiver" neuron. Finally, most current tracers act like permanent ink. Once a virus has moved and triggered a reporter, that mark stays forever. This makes it impossible to study dynamic processes, such as whether a synapse has recently been lost or newly formed due to injury or learning.

Converting synaptic contact into a genetic signal

The authors address these gaps by adapting the synthetic Notch (synNotch) system. This is a tool originally designed for engineering cell-to-cell communication in laboratory settings. TRACR functions through a coordinated three-component architecture:

1. The Sender: Delivered via adeno-associated virus (AAV), the sender neuron expresses a ligand. Specifically, it uses a Green Fluorescent Protein (GFP) tethered to the transmembrane and cytoplasmic domains of Neurexin-3β (a protein that helps stabilize synapses). This ensures the "signal" is concentrated precisely at the presynaptic terminals [Figure 1B].
2. The Receiver: The target neuron expresses a chimeric synNotch receptor. This receptor consists of an extracellular nanobody (a small antibody fragment) that acts as a sensor for the GFP ligand. It also includes a Notch core that acts as a mechanical switch and an intracellular transcriptional activator called tTA (a protein that triggers gene expression).
3. The Reporter: When the Sender’s ligand binds to the Receiver’s receptor across the synaptic cleft, mechanical force triggers a proteolytic cascade (a series of enzymatic cuts). This releases the tTA into the nucleus. There, it drives the expression of a reporter gene, such as Red Fluorescent Protein (RFP) [Figure 1A].

The authors used AAV2/9 or AAV2.7M8 serotypes (specific versions of the virus used for delivery) to distribute these components. Crucially, they refined this into "Receiver 2.0" by adding a hydrophobic protein sequence (RAM7) to the Notch core. This modification minimizes "leaky" activation. This refers to instances where the receptor triggers the reporter without a ligand being present [Figure 1B].

Validating connectivity across scales

The researchers tested TRACR across two vastly different scales of neural architecture: long-range projections and local microcircuits.

In the retinothalamic projection—where axons travel from the eye to the thalamus—the authors demonstrated that TRACR could specifically label postsynaptic targets in the dorsal lateral geniculate nucleus (dLGN). It did this while leaving neighboring, unconnected nuclei untouched [Figure 1I]. This confirms that the system can map long-distance highways of information.

In the denser, more crowded environment of the inner plexiform layer (IPL) of the retina, the task became harder. Here, hundreds of different cell types intermingle in tight layers. The authors used TRACR to trace the targets of starburst amacrine cells (SACs). They found that the reporter signal perfectly matched the expected anatomical locations of these connections. They measured a cosine similarity index (a mathematical measure of how closely two patterns overlap) of $0.89 \pm 0.05$ [Figure 2D, E]. This high score indicates a very strong match between the sender and the reporter. Furthermore, functional electrophysiology confirmed the identity of the labeled cells. They exhibited high direction-selectivity, a hallmark of the specific retinal circuit they were targeting [Figure 3F].

The authors also proved that TRACR does not signal simply because two cells are close. In experiments where they placed the Sender in Müller glia (support cells that wrap around, but do not synapse with, bipolar cells), the reporter failed to activate. This occurred despite the extreme physical proximity [Figure 4G, H].

Measuring the pulse of circuit remodeling

Because the tTA-driven reporter relies on continuous signaling, TRACR is inherently reversible. This opens the door to longitudinal studies—watching the same circuit over days or weeks.

The authors demonstrated this in models of retinal degeneration. In the rd1 mouse model, where photoreceptors die off, the TRACR signal progressively vanished as the synapses disappeared [Figure 5B, C]. Similarly, in models where specific adhesion molecules (like ELFN1) were genetically removed to prevent synapse formation, the TRACR signal dropped significantly [Figure 6B, C].

The ultimate test of the system's utility was a "rescue" experiment. After disrupting a connection, the researchers re-introduced the missing protein (ELFN1). TRACR successfully detected the subsequent reassembly of the synapses. It showed an increase in reporter activation that mirrored the recovery of visual function [Figure 6G, H].

Assessing the toolkit

TRACR is a significant step forward for circuit mapping, but it is not a universal panacea. The authors note that the system's sensitivity varies. While it worked well in the photoreceptor-bipolar synapse (achieving ~60% activation), it was less efficient in the inner retina [Figure 4C]. Additionally, the GFP ligand is enriched at synapses but not strictly confined to them. This means there is always a risk of low-level background noise if viral doses are not carefully titrated (adjusted to the correct amount).

However, for researchers looking to move beyond static anatomical maps, TRACR is a powerful option. It provides the necessary genetic segregation to manipulate one side of a synapse without affecting the other. Its reversibility transforms tracing from a one-time measurement into a dynamic monitor of health and repair. The next frontier will involve optimizing these receptors to ensure consistent sensitivity across the diverse range of synapses found in the mammalian brain.

Figures from the paper