An AAV tracing comparison is only useful if it starts from the question you are actually asking, because several tracing systems can label the same connection and only some of them can answer your particular question. This guide sets the approaches side by side — anterograde AAV, retrograde AAV, anterograde transsynaptic AAV, monosynaptic rabies, polysynaptic herpesvirus tracers and classical dyes — and states plainly what each one buys you and what it costs.
Start with one question: how many synaptic steps does your question cross?
Answer this before comparing anything else. It removes most of the options immediately, and what remains is an easy choice.
Zero steps. You want to see where one population’s axons go, or where its inputs come from. A single AAV does this, and nothing more complex is justified.
One step. You want the neurons directly connected to a starting population. Here the direction of that step decides everything, because different systems travel in different directions.
Two or more steps. You want a chain of connected regions. AAV cannot do this at any dose, and the alternatives carry real costs in toxicity and biosafety.
Six criteria that separate the AAV tracing approaches
Once the step count is settled, six criteria decide the rest. Most disagreements about which method is “better” turn out to be disagreements about which of these matters most.
- Synaptic steps. Local labelling, one step, or a chain. Does your hypothesis involve a neuron you have never labelled directly?
- Direction of that step. Forward to targets, or backward to sources. Are you asking where signals go, or where they come from?
- Whether the neurons must stay healthy. Some systems kill the cells they label, which rules out chronic experiments. Do you need to record, image, or behaviourally test these animals weeks later?
- Genetic access. Anatomy alone, or the ability to manipulate the labelled cells. Will this lead to an optogenetic or chemogenetic follow-up?
- Time available. Survival windows differ by an order of magnitude between systems. How many weeks can this experiment occupy?
- Biosafety and species. Facility approval and species suitability decide real projects. What is approved where you work, and are you working outside the mouse?
AAV tracing versus classical anterograde and retrograde tracers
Anterograde AAV versus an anterograde dye
Both fill axons and reveal terminal fields. A dye is injected, transported, then detected; an AAV delivers a gene and the neuron manufactures the label itself. That difference has two practical consequences: the AAV signal is continuously renewed, so it stays usable for months, and the same injection can later carry a sensor or an effector instead of a marker.
A dye wins on speed and simplicity. There is no expression period to wait through, no promoter to validate, and no packaging limit to respect.
rAAV2-retro versus a classical retrograde tracer
Classical retrograde tracers are taken up efficiently by axon terminals and need no expression time. Natural AAV serotypes are poor at this, which is why rAAV2-retro was produced by directed evolution: it gives retrograde access to projection neurons at an efficiency its authors compare with synthetic retrograde tracers.1 Browse the range in AAV Vector Research.
The reason to accept the extra weeks is genetic access. A retrograde dye tells you which cells project to your target. A retrograde AAV lets you record from, silence, or activate exactly those cells.
AAV1 transsynaptic tagging versus monosynaptic rabies tracing
This is the comparison that matters most, and the one most often got wrong. Both approaches cross exactly one synapse. They cross it in opposite directions, so they answer opposite questions and are not substitutes for each other.
AAV1 or AAV9 carrying a recombinase spreads forward to neurons that receive input from the injected region; a recombinase-dependent reporter at the downstream site reveals them.2 Spread depends on synaptic release and is pathway-dependent — efficient through glutamatergic and GABAergic projections, and minimal through the neuromodulatory ones tested.3 Suitable vectors sit in AAV1 Vectors and AAV9 Vectors.
Glycoprotein-deleted (ΔG) rabies virus spreads backward to neurons that send input to a defined starter population.6 Restriction to a single step comes from deleting the viral glycoprotein and supplying it only in the starter cells; EnvA pseudotyping plus the TVA receptor restricts which cells can start.7
| AAV1 transsynaptic | Monosynaptic rabies | |
|---|---|---|
| Direction of the step | Forward, to postsynaptic targets | Backward, to presynaptic sources |
| Neuron survival | Neurons remain viable; long survival times are routine | First-generation ΔG vectors are cytotoxic, which confined the method largely to anatomical endpoints9 |
| Usable window | Weeks to months | Days for first-generation vectors; tissue is commonly collected about one week after the rabies injection |
| Chronic follow-up | Practical | Not practical with first-generation vectors |
| Completeness | Pathway-dependent3 | Incomplete and strain-dependent — CVS-N2c(ΔG) markedly enhances transsynaptic transfer over SAD-B19(ΔG)8 |
| Main false-positive risk | A weaker reverse-direction component3 | Non-specific labelling that scales with helper-AAV concentration11 |
| Biosafety | Lower containment | Higher containment; requires facility approval |
Two developments soften the toxicity limit. The CVS-N2c(ΔG) strain reduces neuronal toxicity and markedly enhances retrograde transsynaptic transfer relative to the SAD-B19(ΔG) strain.8 Second-generation double-deletion vectors go further: over 90% of both source and presynaptic cells survived a full 12-week two-photon imaging period, and visual response properties appeared normal for 10 weeks.9 If chronic access to input-defined neurons is essential, read these before concluding that rabies is unsuitable.
Rabies versus herpesvirus tracers when you need more than one step
For a chain of connected regions the choice moves to herpesvirus-derived tracers, and direction decides again. The H129 strain of herpes simplex virus type 1 is the standard anterograde transneuronal tracer and in its unmodified form spreads across multiple synapses;12 attenuated pseudorabies virus strains such as PRV Bartha are the corresponding retrograde polysynaptic tracers.14 Vesicular stomatitis virus vectors can be built to travel either way depending on the glycoprotein used — the lymphocytic choriomeningitis virus glycoprotein gives anterograde spread, the rabies glycoprotein retrograde.10
The costs are substantial and shared. Survival windows are measured in days rather than weeks. Spread continues while the animal lives, so the number of steps is set by when you collect tissue rather than by the vector itself. Containment requirements are higher. And H129 tracers carry known limitations of their own — high toxicity, low sensitivity, and non-specific retrograde labelling that contaminates an intended forward result with backward labelling.13
Local stereotaxic injection versus a systemic BBB-crossing capsid
These are not competing tracing methods, and treating them as such is a common error. A local injection defines the labelled population by where the needle went. A systemic injection labels neurons across the whole central nervous system, so the injection site defines nothing.
Systemic delivery suits brain-wide expression of a sensor or reporter. It cannot map a projection, because there is no starting population to project from.
Full AAV tracing comparison matrix
| Approach | Steps | Direction | Cells survive? | Genetic access | Typical window |
|---|---|---|---|---|---|
| AAV, standard serotype | 0 | Forward (outputs) | Yes | Full | 3–4 weeks |
| rAAV2-retro | 0 | Backward (inputs) | Yes | Full | 3–4 weeks |
| AAV1 or AAV9 + recombinase | 1 | Forward | Yes | Full | ~4 weeks |
| Rabies ΔG, first generation | 1 | Backward | No | Limited by the window | Days |
| Rabies, N2c or second generation | 1 | Backward | Substantially improved89 | Yes, over longer periods | Weeks |
| HSV-1 H129 | Multiple | Forward12 | No | Limited | Days |
| PRV Bartha | Multiple | Backward14 | No | Limited | Days |
| Classical dyes and tracers | 0 | Either, by tracer | Yes | None | Days |
Windows are typical rodent planning figures for orientation. Confirm against published work in your own pathway.
AAV and monosynaptic rabies tracing are used together, not chosen between
Monosynaptic rabies tracing is often described as an alternative to AAV. In practice its most widely used implementation depends on AAV. Before the rabies vector is injected, the starter neurons must already express two things: the TVA receptor, so that the EnvA-pseudotyped rabies virus can enter them, and the deleted glycoprotein, so that the virus can make exactly one step.7 Both genes are supplied by recombinase-dependent AAV helper vectors.
This has a practical consequence worth knowing before you plan. Both the efficiency of transsynaptic spread and the degree of non-specific labelling in wild-type control animals depend strongly on how much helper AAV is delivered — so helper concentration is a variable to control deliberately, not an incidental detail.11
- TVA and glycoprotein from one Cre-dependent vector: AAV-nEF1α-DIO-NLS-mCherry-F2A-TVA-T2A-RVG, or the neuron-restricted AAV-hSyn-DIO-EGFP-F2A-TVA-T2A-RVG
- Optimised glycoprotein, supplied separately: AAV-EF1α-DIO-oG
- Reduced-affinity TVA, to limit background starter labelling: AAV-CAG-FLEX-TVA(E66T)-H2B-EGFP
- Flp-driven rather than Cre-driven designs: AAV-nEF1α-FDIO-TVA-P2A-NLS-dTomato
These helper vectors are AAVs. The rabies vector itself is a separate reagent and is not supplied here.
If your question is this, start with that
“Where does this region send its axons?”
A single AAV of a standard serotype carrying a soluble reporter. Nothing more complex is warranted.
“Which regions project into this one?”
rAAV2-retro if you may want to manipulate those neurons later.1 A classical retrograde tracer if pure anatomy is enough and time is short.
“Which neurons here receive input from there?”
AAV1 or AAV9 carrying a recombinase, with a recombinase-dependent reporter at the downstream site.2
“Which neurons send input directly to this cell type?”
Monosynaptic rabies tracing, with AAV helper vectors in the starter population. Choose an N2c or second-generation vector if the animals must survive.89
“What is the whole chain downstream of this nucleus?”
An anterograde herpesvirus tracer, accepting a short survival window and higher containment.12 Consider first whether two one-step experiments would serve better.
“I want a sensor expressed across the whole brain.”
A systemic blood–brain-barrier-crossing capsid, validated in your own mouse line. This is expression, not tracing.4
What none of these circuit tracing approaches can tell you
- Synaptic strength. Every method here reports presence, not weight. Density of labelling is not a measure of influence, and no tracing result substitutes for a physiological recording.
- A complete census. Anterograde transsynaptic spread is pathway-dependent,3 and monosynaptic rabies transfer is incomplete and strain-dependent.8 Comparisons between conditions are safer than absolute counts.
- Function. A labelled connection may be silent, modulatory, or developmental. Anatomy constrains hypotheses; it does not confirm them.
- Transfer across species. Capsid behaviour and receptor dependence differ between strains and species.45
AAV tracing comparison FAQ
Is AAV or rabies better for circuit tracing?
Neither is better in general, because they cross a synapse in opposite directions. AAV1 transsynaptic tagging labels neurons that receive input from your injection site;2 monosynaptic rabies labels neurons that send input to a starter population.7 Choose by the direction your question needs, then weigh toxicity and survival window.
What is the difference between anterograde and retrograde tracing?
Anterograde tracing follows axons forward from the injected cell bodies to their terminal fields, answering where a population sends its output. Retrograde tracing moves backward from axon terminals to the cell bodies of origin, answering which regions project into a target. Some reagents do one, some the other, and a few can be configured either way.10
Why can AAV not trace more than one synapse?
The systems that cross many synapses replicate and spread continuously, which is also what makes them toxic and what raises their containment requirements. AAV vectors are replication-incompetent. The forward spread reported for AAV1 and AAV9 reaches directly connected neurons rather than continuing along a chain.3
Do I need AAV if I am doing monosynaptic rabies tracing?
In the most common implementation, yes. The starter neurons must express the TVA receptor and the deleted glycoprotein before the rabies vector arrives, and recombinase-dependent AAV helper vectors supply both.7 How much helper you deliver strongly affects both spread efficiency and non-specific background.11
Can I still use rabies tracing if the animals must survive?
Look at the newer versions rather than ruling the approach out. The CVS-N2c(ΔG) strain reduces neuronal toxicity while enhancing retrograde transfer,8 and second-generation double-deletion vectors supported 12 weeks of two-photon imaging with over 90% of labelled cells surviving.9
Are classical dye tracers obsolete?
No. For a purely anatomical question with a short deadline, a dye is faster, cheaper, and involves fewer variables. The reason to prefer a viral vector is genetic access — the option to return to the same neurons with a sensor or an effector.
How do I choose between AAV serotypes once I have picked an approach?
That is a separate decision, and it depends on region, species, promoter, and how far the vector should spread from the needle. See the companion guides on how anterograde and retrograde AAV labelling work and how to design an AAV tracing experiment. If an experiment is already running and the tissue looks wrong, start from AAV tracing troubleshooting.
Match the approach to the question, then pick the reagent
Once the number of steps and the direction are settled, the reagent list is short. Browse AAV vectors for circuit research, or talk to a BioHippo technical specialist about which helper and reporter combination suits a monosynaptic design in your pathway.
References
- Tervo DGR, Hwang BY, Viswanathan S, et al. A designer AAV variant permits efficient retrograde access to projection neurons. Neuron. 2016;92(2):372–382. PubMed · DOI
- Zingg B, Chou XL, Zhang ZG, et al. AAV-mediated anterograde transsynaptic tagging: mapping corticocollicular input-defined neural pathways for defense behaviors. Neuron. 2017;93(1):33–47. PubMed · DOI
- Zingg B, Peng B, Huang J, Tao HW, Zhang LI. Synaptic specificity and application of anterograde trans-synaptic AAV for probing neural circuitry. J Neurosci. 2020;40(16):3250–3267. PubMed · DOI
- Hordeaux J, Yuan Y, Clark PM, et al. The GPI-linked protein LY6A drives AAV-PHP.B transport across the blood-brain barrier. Mol Ther. 2019;27(5):912–921. PubMed · DOI
- Hordeaux J, Wang Q, Katz N, Buza EL, Bell P, Wilson JM. The neurotropic properties of AAV-PHP.B are limited to C57BL/6J mice. Mol Ther. 2018;26(3):664–668. PubMed · DOI
- Wickersham IR, Finke S, Conzelmann KK, Callaway EM. Retrograde neuronal tracing with a deletion-mutant rabies virus. Nat Methods. 2007;4(1):47–49. PubMed · DOI
- Callaway EM, Luo L. Monosynaptic circuit tracing with glycoprotein-deleted rabies viruses. J Neurosci. 2015;35(24):8979–8985. PubMed · DOI
- Reardon TR, Murray AJ, Turi GF, et al. Rabies virus CVS-N2c(ΔG) strain enhances retrograde synaptic transfer and neuronal viability. Neuron. 2016;89(4):711–724. PubMed · DOI
- Jin L, Sullivan HA, Zhu M, et al. Long-term labeling and imaging of synaptically connected neuronal networks in vivo using double-deletion-mutant rabies viruses. Nat Neurosci. 2024;27(2):373–383. PubMed · DOI
- Beier KT, Saunders A, Oldenburg IA, et al. Anterograde or retrograde transsynaptic labeling of CNS neurons with vesicular stomatitis virus vectors. PNAS. 2011;108(37):15414–15419. PubMed · DOI
- Lavin TK, Jin L, Lea NE, Wickersham IR. Monosynaptic tracing success depends critically on helper virus concentrations. Front Synaptic Neurosci. 2020;12:6. PubMed · DOI
- Zeng WB, Jiang HF, Gang YD, et al. Anterograde monosynaptic transneuronal tracers derived from herpes simplex virus 1 strain H129. Mol Neurodegener. 2017;12(1):38. PubMed · DOI
- Su P, Ying M, Han Z, et al. High-brightness anterograde transneuronal HSV1 H129 tracer modified using a Trojan horse-like strategy. Mol Brain. 2020;13(1):5. PubMed · DOI
- Card JP, Enquist LW. Transneuronal circuit analysis with pseudorabies viruses. Curr Protoc Neurosci. 2001;Chapter 1:Unit 1.5. PubMed · DOI
- Fischer KB, Collins HK, Pang Y, et al. Monosynaptic restriction of the anterograde herpes simplex virus strain H129 for neural circuit tracing. J Comp Neurol. 2023;531(5):584–595. PubMed · DOI
This guide summarises published methodology to support method selection. It does not replace the original protocols and it is not a biosafety assessment — containment requirements for rabies and herpesvirus tracers must be established with your own institutional biosafety committee. Capsid behaviour, survival windows and labelling efficiency must be validated in your own model, species and pathway. All products are for research use only.
