Skip to content
BUDGET SAVER — Save $50 on every ELISA kit. Limited-time deal.
Lab Essentials Sale — 50% Off Lab Consumables + Free Shipping.
BIG DEAL — 20% Off Transmembrane Proteins.

Lentiviral Vector Promoter Choice and Transgene Silencing

Why expression that looks perfect on day 3 can be gone by week 4 — and how to choose between CMV, EF1α, PGK and synapsin

AH

Amanda Hu

| July 30, 2026 · 12 Lentiviral vector promoter Transgene silencing EF1α vs CMV Vector copy number
Lentiviral Vector Promoter Choice and Transgene Silencing

Your lentiviral vector promoter is not a catalog attribute to skim past — in a third-generation self-inactivating (SIN) vector it is the single design decision that determines whether your transgene is still expressed a month from now. This application note explains why the internal promoter carries all of the expression in a SIN vector, how to tell transcriptional silencing apart from vector loss with one qPCR assay, and how to choose between the CMV, EF1α, PGK and synapsin promoters available across the BioHippo lentiviral vectors collection.

4Promoter options in catalog
271Lentiviral vector products
2Failure modes to distinguish
Day 3–28Recommended timecourse

Why the lentiviral vector promoter carries all expression in a SIN vector

First-generation retroviral and lentiviral vectors drove the transgene from the enhancer/promoter in the U3 region of the 5′ long terminal repeat (LTR). Third-generation self-inactivating (SIN) vectors deliberately delete that U3 enhancer/promoter from the 3′ LTR. Because reverse transcription copies the 3′ U3 into both LTRs of the integrated provirus, the result is a provirus whose LTRs are transcriptionally inactive — the design introduced by Zufferey and colleagues to remove the safety liability of an active LTR enhancer transactivating neighbouring host genes at the integration site (PMID 9811723).

That change had a direct practical consequence: with the LTR silenced by design, the internal promoter is the only element driving your cassette. There is no residual LTR activity left to mask a poor promoter choice.

Lentiviral vector promoter architecture — first-generation LTR-driven cassette versus third-generation SIN vector where the internal promoter is the sole driver
Figure 1. Schematic (illustrative, not experimental data). Deleting the U3 enhancer from the 3′ LTR inactivates both LTRs in the integrated provirus, moving the entire transcriptional burden onto the internal lentiviral vector promoter.
What this means at the bench

If expression is weak, uneven across the population, or fades with passage, interrogate the internal promoter first — before you re-titre the prep, raise the MOI, or re-clone the insert.

Transgene silencing versus vector loss: one qPCR assay separates them

“Expression is dropping” describes two mechanistically different problems that call for opposite responses, and they look identical on a flow plot. Distinguishing them takes a single quantitative PCR assay and is worth doing before changing anything else.

A
Vector loss. The transduced cells, or the vector copies themselves, are disappearing from the culture — unintegrated vector DNA diluted out in the first days after transduction, counter-selection against a transgene that is toxic or slows proliferation, or outgrowth of untransduced cells in an unselected population.
Signature: vector copy number per genome falls across timepoints. The remaining positive cells are often still brightly positive — you are losing cells, not expression.
B
Transcriptional silencing. The provirus is still present at the same copy number, but the promoter has been epigenetically shut down. The population shifts bright → dim → negative, often heterogeneously, with individual integration sites behaving differently.
Signature: vector copy number per genome stays flat while expression falls. This is the pattern Zhang and colleagues used to establish silencing rather than cell loss in SIN lentiviral vectors (DOI · PMID 20588258).
Transgene silencing versus vector loss in lentiviral vectors — vector copy number stays flat during silencing and falls during vector loss
Figure 2. Decision schematic (illustrative). Vector copy number per genome, measured by qPCR on genomic DNA against a single-copy reference gene, is the variable that separates the two failure modes.

The assay. Quantitative PCR on genomic DNA against a single-copy reference gene, run on matched aliquots at several timepoints, gives vector copy number (VCN) per cell. Sample the same culture you are analysing by flow, at the same timepoints. Flat VCN with declining reporter signal is transcriptional silencing — and the fix is promoter and cassette design, not more virus.

What transgene silencing actually is

Integrated proviral sequences are recognised and packaged into repressive chromatin by host machinery that exists to control endogenous retroviral elements. Two features of the promoter you choose determine how exposed your cassette is.

CpG density and DNA methylation

Cytosine methylation at CpG dinucleotides within and around a promoter recruits repressive complexes and blocks transcription factor access. Promoters with high CpG density in their core regulatory sequence present more substrate for this. In SIN lentiviral vectors, silencing correlated with extensive CpG methylation of the regulatory element, while a methylation-resistant element remained almost completely unmethylated and stably expressed (Zhang et al., Mol Ther 2010).

Heterochromatin deposition and cell state

Repressive histone marks — H3K9 trimethylation and associated heterochromatin — are deposited on retroviral sequences by KRAB zinc-finger protein complexes acting through the co-repressor TRIM28 (KAP1) and the methyltransferase SETDB1 (Rowe et al., Nature 2010, PMID 20075919). This machinery is most active in pluripotent and early embryonic states, which is why embryonic stem cells, iPSC and embryonal carcinoma lines are the hardest contexts in which to maintain expression — and why a promoter that behaves well in HEK293 can fail in iPSC.

Integration-site position effects

Lentivirus integrates semi-randomly with a preference for active transcription units. Local chromatin at each integration site modulates promoter output independently of the promoter's intrinsic strength. This is why clones from the same transduction differ, and why one clone's behaviour is not evidence about the construct. The right unit of testing is several independent clones followed over time, not one clone measured once.

CMV, EF1α, PGK and synapsin: the four lentiviral vector promoters compared

Products in the BioHippo lentiviral vectors collection are filterable by promoter, with four options available. Each sits at a different point on the strength-versus-durability trade-off.

Lentiviral vector promoter comparison — CMV, EF1a, PGK and synapsin ranked by expression strength and durability with recommended use context
Figure 3. Comparative summary (illustrative; strength and durability bars are qualitative summaries of published comparisons, not measured values from a single experiment).

CMV — human cytomegalovirus immediate-early enhancer/promoter

Viral · highest peak output · least durable

The strongest available driver in permissive transformed lines, and the standard choice where maximum protein per cell over a short window is the goal. Its weakness is well documented: in a systematic comparison of eight constitutive promoters delivered by lentiviral integration across a range of mammalian cell types, most promoters held a fairly consistent rank across cell types — but CMV varied considerably from one cell type to another (Qin et al., PLOS ONE 2010 · PMID 20485554). That context-dependence is the practical risk.

Viral originHigh CpG densityHighest peak output in permissive linesMethylation-prone

Use when the readout falls within roughly two weeks of transduction, in a permissive line, and peak signal matters more than durability. Avoid when the experiment runs for weeks, uses primary or stem cells, or goes in vivo.

EF1α — human elongation factor 1α (EEF1A1)

Mammalian housekeeping · strong and comparatively durable · general default

A mammalian housekeeping promoter that combines high output with substantially better cross-cell-type durability than CMV, which is why it is the usual default for stable lines and primary cells.

It is not silencing-proof, and the literature is explicit about this. In human embryonic stem cells, EF1α was the most stable of five constitutive promoters through embryoid body differentiation — but was still downregulated in approximately half of the cells (Norrman et al., PLOS ONE 2010 · PMID 20865032). In P19 embryonal carcinoma cells, EF1α-driven expression declined rapidly within 16 days while a chromatin-opening element in the same system remained stable (PMID 20588258). In pluripotent and embryonal contexts, EF1α buys time rather than immunity.

Two forms circulate and are not interchangeable. The full-length version includes the first exon and intron and is generally the stronger and more durable of the two, at a cost of roughly 1.2 kb of packaging capacity. The core or intron-less short form (often written EFS) is a few hundred base pairs, weaker, and valued where cassette space is tight or a smaller enhancer footprint is preferred. Confirm which form a product carries before assuming published EF1α performance transfers.

Mammalian housekeepingStrong across most cell typesFull-length ~1.2 kb vs. core ~0.2 kbDurable, not silencing-proof

Use when you need one lentiviral vector promoter to work reasonably across dividing lines, primary cells and long-term culture — the sensible starting point absent a cell-type-specific reason to choose otherwise.

PGK — human phosphoglycerate kinase 1 (PGK1)

Mammalian housekeeping · moderate output · uniform

A compact housekeeping promoter that produces moderate, consistent expression. It is reliably weaker than EF1α, and that is frequently the point rather than a deficiency. Modest, even expression avoids the overexpression artefacts that confound structure–function, localisation and dose–response work, and its small size and comparatively weak enhancer activity make it a common choice for selection and marker cassettes.

Mammalian housekeepingCompact (~0.5 kb)Moderate, uniform outputWeak enhancer activity

Use when physiological-range or uniform expression matters more than magnitude — dose-sensitive constructs, rescue experiments, selection and marker cassettes, or any transgene toxic at high copy. Avoid when you need maximum protein yield or a bright reporter above background.

Synapsin — human synapsin 1 (SYN1) promoter

Neuron-specific · moderate output

Unlike the three above, synapsin is not constitutive — it is a neuronal promoter that restricts expression largely to neurons, sparing astrocytes, microglia and other non-neuronal cells in a mixed culture or an injected brain region. In mixed primary cultures and in vivo CNS work this specificity is usually worth more than raw strength, because it removes the confound of glial expression without a transgenic driver line.

Two caveats worth stating plainly. Output is moderate rather than high, so pair it with a sensitive reporter or an amplification step where signal is marginal. And promoter specificity in a viral vector is relative, not absolute — validate cell-type restriction in your own preparation and target region by co-staining rather than assuming it.

Neuron-restrictedCompactModerate outputSpecificity requires local validation

Use when the experiment needs neuronal restriction — mixed primary cultures, in vivo CNS injection, circuit-level work. Avoid when the target is non-neuronal, or the transgene must be expressed uniformly across all transduced cells.

Matching lentiviral vector promoter to cell type and timeline

The table below reflects general behaviour reported in the promoter-comparison literature. Treat it as a starting hypothesis to test in your own cells, not a substitute for a stability timecourse.

Context First choice Rationale Watch for
HEK293, HeLa, other permissive lines — short-term, high yield CMV Highest peak output in exactly the contexts where CMV performs well. Output decline if the line is carried for weeks.
Stable cell line generation EF1α Best available balance of strength and durability for long-term culture. CMV is a common cause of “the line stopped expressing.”
Primary cells — fibroblasts, MSC, lymphocytes EF1α CMV was the most cell-type-variable promoter in direct comparison (PMID 20485554). Transduction efficiency is a separate variable — do not read low signal as silencing without a VCN check.
iPSC / ESC and differentiated progeny EF1α + silencing-resistant element Most stable of five promoters through differentiation, though still lost in roughly half of cells (PMID 20865032). The hardest context there is. Budget for a stability timecourse.
Dose-sensitive or toxic transgene PGK Moderate, uniform output avoids overexpression artefacts and counter-selection. May fall below detection with insensitive readouts.
Neurons — mixed culture or in vivo CNS Synapsin Restricts expression to neurons without a driver line. Moderate strength; verify restriction by co-staining in your own preparation.
Selection or marker cassette PGK Compact, uniform, sufficient for resistance; leaves capacity for the transgene. Separate promoters can be silenced independently — see below.

For shRNA vectors, the listed Pol II promoter may not drive your effector

This trips people up regularly. Short hairpin RNA and microRNA cassettes are conventionally transcribed by RNA polymerase III from small nuclear RNA promoters such as U6 or H1, because Pol III produces the short, defined, non-polyadenylated transcripts that hairpin processing requires. The RNA polymerase II promoter listed as a product attribute — CMV, EF1α, PGK — typically drives the accompanying reporter or selection marker, not the hairpin itself.

1
Reporter loss does not necessarily mean knockdown loss. If a Pol II reporter cassette silences while the Pol III hairpin cassette remains active, you lose your visual or selectable readout while the knockdown persists. Confirm knockdown at transcript or protein level rather than inferring it from reporter signal.
2
The converse is also possible. Cells under selection can retain a functioning resistance marker while the effector cassette is silenced, producing a population that survives selection but has lost the phenotype. Independent cassettes have independent fates.
Design response: linking marker and transgene in a single transcript — via an IRES or a 2A peptide sequence under one promoter — couples their fates, so selection genuinely reports transgene expression. An IRES gives markedly lower downstream expression; 2A gives near-stoichiometric output but leaves a short peptide remnant on the upstream protein.

Before ordering. Check the individual product datasheet for the full cassette architecture — which promoter drives the effector, which drives the marker, and whether they are linked. The promoter filter narrows the catalog; the datasheet is the authoritative record for any specific product.

Diagnostic workflow: working out whether you have a silencing problem

1
Establish a timecourse, not a timepoint. Measure expression by flow cytometry at approximately day 3, 7, 14 and 28 post-transduction on the same culture. A single early reading cannot distinguish a stable cassette from one about to fail — and day 3 is typically the most flattering timepoint you will ever see.
2
Run VCN alongside it. Quantitative PCR for vector copies per genome on matched aliquots. Flat VCN with declining expression indicates silencing; falling VCN indicates cell or vector loss.
3
Look at distribution, not just the mean. Silencing usually appears as a growing dim or negative subpopulation rather than a uniform downward shift. Report percentage of positive cells and median intensity of the positive gate separately — a stable median with a shrinking positive fraction is a distinctive silencing signature that a mean fluorescence value hides.
4
Test the epigenetic hypothesis directly. Brief treatment with a DNA methyltransferase inhibitor or a histone deacetylase inhibitor, with recovery of expression, supports epigenetic silencing.
Interpret with care: these compounds are broadly cytotoxic and genome-wide in effect. They are a diagnostic, not a remedy. For a direct measurement, bisulphite sequencing or methylation-specific PCR across the promoter region quantifies CpG methylation without pharmacology.
5
Rule out counter-selection. Run a reporter-only vector in parallel at matched MOI. If the reporter-only control is stable and your transgene vector is not, the transgene product may impose a fitness cost and the population is selecting against it. That is a biology result, not a vector fault.

Reducing lentiviral transgene silencing risk in the construct

Choose a mammalian housekeeping promoter for anything long-term

The single highest-yield decision. For experiments running beyond about two weeks, in primary cells, or in vivo, a mammalian housekeeping promoter is a better starting point than a viral one.

Add a chromatin-opening element where the context is hostile

The A2UCOE, a methylation-free CpG island spanning the divergent promoters of the human HNRPA2B1CBX3 housekeeping locus, resists methylation-mediated silencing and can protect a linked heterologous promoter. In the study establishing this in SIN lentiviral vectors, a 1.2 kb A2UCOE remained completely stable in P19 embryonal carcinoma cells under conditions where both an SFFV LTR and an EF1α promoter declined within 16 days, and placing the element upstream of the SFFV LTR protected that element from both methylation and silencing (Zhang et al., Mol Ther 2010). The cost is packaging capacity — 1.2 kb is a substantial fraction of a lentiviral cassette, and insert size trades against titre.

Distinguish transcriptional from post-transcriptional fixes

The woodchuck hepatitis virus post-transcriptional regulatory element (WPRE) raises transgene expression by improving transcript processing and stability, and is present in many vector backbones (Zufferey et al., J Virol 1999, PMID 10074136). It acts on the RNA after transcription and therefore does not protect against promoter silencing. WPRE and a durable promoter address different problems.

Transduce at low MOI and screen several clones

Low multiplicity of infection favours single-copy integration, reducing position-effect artefacts and copy-number-driven overexpression. It also means clones will differ, so pick several independent clones and carry them for at least four weeks before committing. Bank early passages — a clone stable at passage 5 may not be at passage 30.

Couple your marker to your transgene

Where continued selection is part of the experiment, a single-promoter bicistronic cassette (IRES or 2A) makes marker retention a genuine proxy for transgene expression, closing the gap described in the Pol II / Pol III section above.

What to record and report

Expression figures without their measurement conditions are not interpretable and will not survive review. For any lentiviral expression result, capture:

  • Promoter identity and form — including whether EF1α is full-length or core, since the two behave differently.
  • MOI and the titre method behind it — functional titre in transducing units per mL, with the cell line the titration was performed in. Physical particle counts and functional titres are not interchangeable.
  • Cell type, passage number and culture duration at each measurement.
  • Vector copy number per genome at each expression timepoint, so silencing and vector loss can be told apart by a reader.
  • Percentage positive and median intensity of the positive gate, reported separately rather than as a single mean.
  • Number of independent clones characterised, and the spread between them.

Frequently asked questions about lentiviral vector promoter choice

My GFP was bright at day 3 and gone by week 4. Is that transgene silencing?

Possibly, but you cannot tell from expression data alone. Run qPCR for vector copy number on the same cultures across those timepoints. Flat copy number with falling signal indicates transcriptional silencing, and the response is to change promoter or cassette design. Falling copy number indicates you are losing transduced cells, and the response is selection, a lower-expressing promoter if the transgene is toxic, or a re-examination of transduction efficiency.

Can I just switch from CMV to EF1α and stop worrying about silencing?

It is the right first move and will resolve many cases, but no — EF1α is more durable, not immune. In human embryonic stem cells it was the most stable of five constitutive promoters through differentiation while still being downregulated in about half the cells, and in P19 embryonal carcinoma cells EF1α-driven expression declined within 16 days. In pluripotent and embryonal contexts, plan a stability timecourse and consider adding a chromatin-opening element.

Why would I ever choose the weaker PGK promoter?

Because maximum expression is often the wrong target. Overexpression produces mislocalisation, aggregation, non-physiological interactions and dominant-negative effects that are artefacts of level rather than features of the protein. If your question is about function near a physiological level, or the transgene is toxic at high copy, PGK's moderate uniform output is the more informative choice. It is also compact, which matters when packaging capacity is tight.

Does the synapsin promoter guarantee my transgene is only in neurons?

It strongly biases expression toward neurons and will spare most glia, but promoter specificity in a viral vector is relative rather than absolute, and it can vary with target region, vector dose and preparation. If cell-type restriction is load-bearing for your conclusion, validate it in your own material by co-staining with neuronal and glial markers and quantifying the overlap. Report the quantification rather than citing the promoter's reputation.

My cells survive puromycin selection but I have lost the phenotype. How?

If the resistance marker and the effector cassette run from separate promoters, they can be silenced independently. Cells that silence the effector while retaining the marker will survive selection and enrich in the population precisely because they no longer bear the cost of the transgene. Confirm effector expression directly at transcript or protein level, and for future builds use a single-promoter bicistronic design.

Can I add WPRE to fix a silencing problem?

No. WPRE increases expression through post-transcriptional effects on the transcript, so it raises output from a promoter that is working but does nothing for one that has been transcriptionally shut down. Silencing is upstream of the mechanism WPRE acts on. Address it with promoter choice and, where warranted, a chromatin-opening element.

Which promoter do the products I need actually use?

Filter the lentiviral vectors collection by promoter alongside your other requirements — vector type, application, species, reporter and selection marker. Then confirm the full cassette architecture on the individual product page. If the combination you need is not listed, a quote request can cover custom builds.

Match the promoter to your cell type before you order

BioHippo's lentiviral vector catalog spans 271 products filterable by promoter, vector type, application, species, reporter and selection marker. For related delivery formats see the AAV vector collection and the cell lines catalog. If you are unsure whether CMV, EF1α, PGK or synapsin fits your cell type and experimental timeline, send a quote request with your target cell type, experiment duration and readout.

References

  1. Zhang F, Frost AR, Blundell MP, Bales O, Antoniou MN, Thrasher AJ. A ubiquitous chromatin opening element (UCOE) confers resistance to DNA methylation-mediated silencing of lentiviral vectors. Mol Ther 2010;18(9):1640–9. DOI · PMID 20588258
  2. Rowe HM, Jakobsson J, Mesnard D, et al. KAP1 controls endogenous retroviruses in embryonic stem cells. Nature 2010;463(7278):237–40. DOI · PMID 20075919
  3. Qin JY, Zhang L, Clift KL, Hulur I, Xiang AP, Ren BZ, Lahn BT. Systematic comparison of constitutive promoters and the doxycycline-inducible promoter. PLOS ONE 2010;5(5):e10611. DOI · PMID 20485554
  4. Norrman K, Fischer Y, Bonnamy B, Wolfhagen Sand F, Ravassard P, Semb H. Quantitative comparison of constitutive promoters in human ES cells. PLOS ONE 2010;5(8):e12413. DOI · PMID 20865032
  5. Zufferey R, Donello JE, Trono D, Hope TJ. Woodchuck hepatitis virus posttranscriptional regulatory element enhances expression of transgenes delivered by retroviral vectors. J Virol 1999;73(4):2886–92. PMID 10074136
  6. Zufferey R, Dull T, Mandel RJ, et al. Self-inactivating lentivirus vector for safe and efficient in vivo gene delivery. J Virol 1998;72(12):9873–80. PMID 9811723

Research Use Only. Not for use in diagnostic or therapeutic procedures. Promoter behaviour described here summarises published comparative studies conducted in defined cell systems and does not predict performance in every cell type. Verify expression stability, and where relevant cell-type specificity, in your own model. Confirm cassette architecture, promoter form, titre, titration method and species reactivity on each product page before purchase. Catalog product count and promoter options reflect the BioHippo lentiviral vectors collection at the time of writing.



Shop Related Collections



Ask a Scientist →