Primary neurons RNA silencing guide

Post-mitotic nature and transfection incompatibility

Neurons cease cell division after differentiation, creating fundamental incompatibility with conventional transfection. Electroporation can directly deliver nucleic acids to the cytoplasm and nucleus of non-dividing neurons. Primary neurons are challenging for transfection due to their sensitivity to membrane disruption and selective uptake mechanisms, and while mitosis is not a requirement for ASO or siRNA delivery, post-mitotic neurons present unique challenges for conventional transfection methods. Consequently, lipofection efficiency in DIV 7+ neurons is often very low, and electroporation can cause immediate cell death through membrane damage in non-dividing cells, though optimized protocols can preserve high viability. This makes functional studies in mature, physiologically relevant neurons challenging with conventional methods.

High impact

Electrophysiological disruption

Transfection reagents can alter neuronal membrane properties and ion channel function. Electroporation creates transient membrane pores that typically reseal within minutes to hours, though optimized protocols can preserve normal membrane properties. However, suboptimal electroporation conditions may temporarily affect input resistance and capacitance. Cationic lipids at high concentrations have been reported to affect membrane potential in some studies. These potential artifacts require careful optimization and controls for electrophysiology experiments studying neuronal excitability, synaptic transmission, and action potential propagation.

High impact

Neurite retraction and morphological damage

High-intensity transfection methods, particularly biolistic gene transfer and suboptimal electroporation, can cause morphological damage to neurons. Published studies report progressive dendritic spine loss after biolistic transfection with shRNA in hippocampal neurons, and reduced neurite outgrowth after electroporation in some neuronal subtypes. Neurons may retract neurite arbors under stress conditions, severing synaptic connections and disrupting neuronal network architecture. These morphological changes make it challenging to study neurite outgrowth, axon guidance, synaptogenesis, or circuit-level phenomena with certain transfection methods.

High impact

Viral vector toxicity in long-term cultures

Lentiviral vectors can trigger interferon signaling and innate immune activation that alters baseline gene expression, though AAV vectors show minimal immunogenicity. However, insertional mutagenesis from lentiviral integration disrupts endogenous gene regulation, creating experimental artifacts. shRNA overexpression can saturate the Exportin-5 export pathway and cause widespread off-target effects by competing with endogenous miRNAs. Moreover, the 2-4 week timeline for viral production and validation is incompatible with primary neuronal culture windows (typically 3-4 weeks total).

Medium impact

Developmental stage sensitivity

Neuronal transfection efficiency varies with developmental stage, creating a narrow and restrictive experimental window. Immature neurons (DIV 0-3) often lack fully developed uptake machinery and exhibit poor transfection tolerance. Mature neurons (DIV 7-21) develop resistance to transfection as they become fully post-mitotic, with lipofection efficiency often very low at DIV 7-21 unless the protocol is adapted. The narrow optimal window (DIV 3-5) restricts experiments to semi-mature neurons that may not fully recapitulate adult neuronal biology, limiting translational relevance.

Medium impact

Neuronal subtype variability

Different neuronal subtypes exhibit wildly variable transfection susceptibility, requiring extensive cell-type-specific optimization. GABAergic interneurons show marginally higher transfection tolerance than glutamatergic pyramidal neurons, but efficiency stays low and toxicity severe. Motor neurons are reported to be difficult with lipofection and electroporation, and the figures behind that report are not established: the group that states it published no efficiency or viability measurement of its own for either method, so read both for the method and the cargo in hand. DRG sensory neurons have large cell bodies, 30-50 μm in diameter, and that is not what decides how much reaches them: in primary adult human dorsal root ganglion cultures from organ donors, a cationic lipid reagent carrying a CRISPR plasmid reached the large majority of the neurons, and the same class of reagent reaches a substantial share of adult mouse DRG neurons with messenger RNA. What did cost viability there was the cargo rather than the lipid: two of the three plasmids reduced viability against untreated cells but not against the lipid alone. Each neuronal subtype demands different protocols, making comparative studies across cell types impractical.

Medium impact

Method comparison

Cell types and applications

• Primary rodent neurons (cortical, hippocampal, motor, DRG)
• Human iPSC-derived neurons (cortical, dopaminergic, motor)
• Post-mitotic mature neurons (DIV 7-21)
• Neurodegenerative disease modeling (AD, PD, ALS, HD)
• Synaptic plasticity and electrophysiology studies
• Pain research in DRG sensory neurons
• Neuronal cell lines (SH-SY5Y, PC12, N2a)
• Long-term neuronal cultures requiring preserved function

• AUMsilence™ sdASO™

Alternative products

• AUMantagomir™ sdASO™
• AUMlnc™ sdASO™

Cell-type-optimized protocols for cortical, hippocampal, motor, and sensory neurons. No transfection reagents required: preserves neuronal morphology and electrophysiology.

Quick start protocol (all primary neurons)

1. 01Culture neurons in neuronal culture medium + supplement on poly-D-lysine/laminin-coated plates
2. 02Add AUMsilence™ sdASO™ directly to culture medium at 10 μM, with no transfection reagent
3. 03Incubate 48-72 hours at 37°C, 5% CO₂ (no media change required)
4. 04Validate knockdown by qRT-PCR (mRNA), immunocytochemistry or Western blot (protein), and patch clamp electrophysiology (functional validation)

Cell-type-specific protocols

Primary cortical neurons (rat/mouse E18)

1. 01

   Step 1: Cortex dissection

   Harvest E18 rat or mouse embryos. Dissect cortices under sterile conditions in ice-cold HBSS without calcium/magnesium. Remove meninges carefully to avoid glial contamination. Keep tissue on ice.

   Materials: Pregnant E18 rat/mouse, dissection tools, ice-cold HBSS (Ca²⁺/Mg²⁺-free)

   Note: E18 is optimal for cortical neurons; E16-E17 yields more immature neurons

   Timing: Day 0 (30-45 min dissection time)

2. 02

   Step 2: Enzymatic dissociation

   Incubate cortical tissue in papain solution (20 U/mL in HBSS) for 15 minutes at 37°C. Add DNase I (100 μg/mL) to prevent cell clumping. Gently swirl every 5 minutes.

   Materials: Papain (20 U/mL), DNase I (100 μg/mL), 37°C water bath

   Note: Over-digestion (>20 min) reduces viability; under-digestion causes cell clumps

   Timing: Day 0

3. 03

   Step 3: Trituration and cell counting

   Wash tissue 3× with neuronal culture medium to remove papain. Triturate gently with fire-polished Pasteur pipette (10-15 passes) until single-cell suspension. Count viable cells with trypan blue.

   Materials: Fire-polished pipettes, neuronal culture medium, hemocytometer

   Note: Avoid bubbles during trituration (reduces viability). Expect 2-4 million cells per cortex.

   Timing: Day 0

4. 04

   Step 4: Plating on coated surfaces

   Plate neurons at 50,000-100,000 cells/cm² on poly-D-lysine (100 μg/mL, overnight) + laminin (20 μg/mL, 2h) coated plates. Use neuronal culture medium + neuronal supplement (2%) + L-glutamine (2 mM) + 0.5 mM glutamate (first 24h only).

   Materials: PDL/laminin-coated plates, neuronal culture medium, neuronal supplement, L-glutamine, L-glutamate

   Note: Higher densities (>150,000/cm²) cause excitotoxicity; lower densities (<30,000/cm²) reduce synapse formation

   Timing: Day 0

5. 05

   Step 5: Culture maturation

   Culture neurons at 37°C, 5% CO₂. Perform half-media changes (replace 50% of medium) every 3-4 days without glutamate. Neurons mature by DIV 7 (synapse formation) and are fully mature by DIV 14.

   Materials: Humidified CO₂ incubator, neuronal culture medium + neuronal supplement (no glutamate)

   Note: Media changes should be gentle (avoid pipetting directly onto cells). Minimal glia (<10%) in these cultures.

   Timing: Days 0-7

6. 06

   Step 6: AUMsilence sdASO treatment

   At DIV 7-10 (optimal maturation), add AUMsilence sdASO directly to culture medium at 10 μM final concentration. Calculate volume: For 500 μL culture, add 5 μL of 1 mM AUMsilence sdASO stock. Mix gently. No transfection reagent needed.

   Materials: AUMsilence sdASO (1 mM stock in nuclease-free water)

   Note: Do not change media after ASO addition. Gymnotic uptake begins within 2-4 hours of addition. No washing required.

   Timing: Day 7-10 (depending on experimental design)

7. 07

   Step 7: Validation

   At 48-72h post-AUMsilence sdASO treatment: (1) qRT-PCR for mRNA (expect 70-95% knockdown), (2) Western blot or immunocytochemistry for protein (72-96h), (3) Whole-cell patch clamp to confirm normal action potential firing and resting membrane potential (-60 to -70 mV).

   Materials: RNA extraction kit, qPCR reagents, antibodies, patch clamp rig

   Note: Include viability staining (Calcein AM/Ethidium homodimer).

   Timing: Days 9-12

Primary hippocampal neurons (synaptic plasticity studies)

1. 01

   Step 1: Hippocampus dissection

   Dissect hippocampi from E18 rat or mouse embryos. Remove dentate gyrus if studying CA1-CA3 pyramidal neurons specifically. Dissection similar to cortex but requires careful isolation of hippocampal structure.

   Materials: E18 embryos, sterile dissection tools, ice-cold HBSS

   Note: Hippocampal neurons are more sensitive than cortical neurons: work quickly on ice

   Timing: Day 0

2. 02

   Step 2: Dissociation and plating

   Follow same papain digestion protocol as cortical neurons (15 min, 37°C). Plate at lower density: 30,000-50,000 cells/cm² for dendritic spine imaging, or 50,000-80,000 cells/cm² for electrophysiology. Use neuronal culture medium + neuronal supplement + L-glutamine.

   Materials: Papain, PDL/laminin-coated coverslips or plates, neuronal culture medium + neuronal supplement

   Note: Lower density allows visualization of individual dendritic arbors and spines for imaging experiments

   Timing: Day 0

3. 03

   Step 3: Synaptic maturation

   Culture for 7-14 days. Synapses begin forming at DIV 7 and are functionally mature by DIV 14. Dendritic spines appear by DIV 10-12. Half-media changes every 3-4 days.

   Materials: Neuronal culture medium + neuronal supplement, humidified incubator

   Note: DIV 14+ neurons are optimal for synaptic plasticity experiments (mature silent and functional synapses)

   Timing: Days 0-14

4. 04

   Step 4: AUMsilence sdASO treatment

   At DIV 7-14, add AUMsilence sdASO at 10 μM. Used for knocking down synaptic proteins (PSD-95, synaptophysin), plasticity regulators (Arc, BDNF, CREB), or AMPA/NMDA receptor subunits.

   Materials: AUMsilence sdASO

   Note: Compatible with live-cell imaging (calcium imaging, spine dynamics) during and after treatment

   Timing: Day 7-14

5. 05

   Step 5: Validation and functional assays

   Validate knockdown at 48-72h: (1) qRT-PCR, (2) Immunocytochemistry for synaptic markers (synaptophysin, PSD-95, Homer1), (3) Dendritic spine density quantification, (4) mEPSC recordings (miniature excitatory postsynaptic currents) by patch clamp, (5) Calcium imaging for network activity.

   Materials: qPCR reagents, synaptic antibodies, patch clamp setup, calcium imaging microscope

   Note: AUMsilence sdASO preserves dendritic spine morphology unlike lipofection which causes spine loss

   Timing: Days 9-16

Human iPSC-derived neurons (translational research)

1. 01

   Step 1: Source neurons

   Use commercially available human iPSC-derived neurons, or differentiate them in-house from iPSC lines. Commercial neurons arrive as frozen vials or differentiated cultures.

   Materials: Commercial iPSC-derived neurons or in-house differentiation protocol

   Note: Patient-derived iPSC neurons provide valuable research models for studying Alzheimer's, Parkinson's, and ALS disease mechanisms and may enable future personalized medicine approaches. iPSC-derived neurons provide scalable platforms for tau knockdown screening and therapeutic discovery

   Timing: Day 0 (or weeks 0-6 for in-house differentiation)

2. 02

   Step 2: Culture and maturation

   Plate neurons according to manufacturer instructions. Typical media: neuronal medium + neuronal supplement + BDNF (20 ng/mL) + GDNF (20 ng/mL) + NT-3 (20 ng/mL). Culture on PDL/laminin-coated plates. Human neurons mature more slowly than rodent neurons.

   Materials: Neuronal medium, neuronal supplement, neurotrophic factors, PDL/laminin plates

   Note: Human neurons require 14-28 days for full maturation (compared to 7-14 days for rodent neurons)

   Timing: Days 0-28

3. 03

   Step 3: AUMsilence sdASO treatment

   Treat fully mature neurons (DIV 14-28) with AUMsilence sdASO at 10 μM. Incubate 72-96h, which is longer than rodent neurons need.

   Materials: AUMsilence sdASO

   Note: Extended incubation (72-96h) accounts for slower protein turnover in human neurons

   Timing: Day 14-28

4. 04

   Step 4: Disease modeling applications

   Use patient-derived neurons for disease-specific studies: Alzheimer's (APP, PSEN1, MAPT mutations), Parkinson's (SNCA, LRRK2, GBA mutations), ALS (SOD1, C9orf72, TDP-43 mutations). Knock down disease genes or modifiers to study pathogenic mechanisms.

   Materials: Patient-specific iPSC neurons, disease-relevant ASOs

   Note: Human neurons recapitulate disease pathology better than rodent models for age-related neurodegeneration

   Timing: Days 17-32 (post-treatment)

5. 05

   Step 5: Validation

   Validate at 72-96h: qRT-PCR, Western blot, immunocytochemistry. Functional validation by patch clamp (human neurons exhibit slower action potential kinetics than rodent). Assess disease phenotypes: Aβ aggregation (Alzheimer's), α-synuclein inclusions (Parkinson's), TDP-43 mislocalization (ALS).

   Materials: Standard validation reagents, disease-specific readouts

   Note: Human neurons provide translational relevance for drug target validation

   Timing: Days 17-32

Dorsal root ganglion (DRG) neurons (pain research)

1. 01

   Step 1: DRG isolation

   Harvest DRG from adult rat or mouse spinal column (L4-L5 ganglia for hindpaw innervation, C6-C8 for forepaw). Trim connective tissue and nerve roots under dissection microscope. Place in cold HBSS.

   Materials: Adult rat/mouse, dissection microscope, sterile tools, ice-cold HBSS

   Note: Adult DRG neurons (not embryonic) are required for mature sensory neuron phenotypes

   Timing: Day 0

2. 02

   Step 2: Enzymatic digestion

   Incubate DRG in collagenase (1 mg/mL) + dispase (2.5 mg/mL) in DMEM for 45 minutes at 37°C with gentle agitation. Wash 3× with DMEM. Triturate gently to dissociate neurons.

   Materials: Collagenase, dispase, DMEM, 37°C shaker

   Note: DRG require longer digestion than embryonic neurons due to dense extracellular matrix

   Timing: Day 0

3. 03

   Step 3: Plating and culture

   Plate DRG neurons at 5,000-10,000 cells/cm² on PDL/laminin-coated glass coverslips (for patch clamp). Culture in DMEM/F12 + 10% FBS + NGF (50 ng/mL) for nociceptive neurons, or GDNF (50 ng/mL) for non-peptidergic neurons.

   Materials: PDL/laminin coverslips, DMEM/F12, FBS, NGF or GDNF

   Note: Low density allows single-cell electrophysiology. NGF maintains nociceptive (TRPV1+, substance P+) phenotype.

   Timing: Day 0

4. 04

   Step 4: AUMsilence sdASO treatment

   Treat DRG neurons at 24-48h post-plating with AUMsilence sdASO at 10 μM. Target pain-relevant genes: Nav1.7 (SCN9A), Nav1.8 (SCN10A), TRPV1, TRPA1, P2X3.

   Materials: AUMsilence sdASO targeting pain-relevant genes

   Note: DRG neurons take up AUMsilence sdASO from the medium, at the same starting concentration as the other subtypes.

   Timing: Day 1-2

5. 05

   Step 5: Functional validation by patch clamp

   At 48-72h post-treatment, perform whole-cell patch clamp to measure sodium currents (for Nav1.7/1.8 knockdown), capsaicin-evoked currents (for TRPV1 knockdown), or action potential firing. Validate that knockdown reduces target channel function without affecting other channels.

   Materials: Patch clamp setup, capsaicin, specific channel blockers (TTX, A-803467)

   Note: Patch clamp is essential for DRG neurons: validates functional impact on nociceptor excitability

   Timing: Days 3-5

Motor neurons (ALS/SMA research)

1. 01

   Step 1: Motor neuron source

   Dissect spinal cord from E12.5-E13.5 mouse embryos (motor neurons are born early). Alternatively, use human iPSC-derived motor neurons from commercial vendors.

   Materials: E12.5 mouse embryos or commercial iPSC motor neurons

   Note: E12.5 is critical: later stages have fewer motor neuron progenitors

   Timing: Day 0

2. 02

   Step 2: Dissociation and purification

   For mouse spinal cord: papain digestion (20 U/mL, 15 min). Optionally purify motor neurons by p75 immunopanning (removes non-motor neurons) or density gradient centrifugation. Expect 5-10% motor neurons in crude spinal cord prep.

   Materials: Papain, p75 antibody (optional), density gradient medium (optional)

   Note: Purification increases motor neuron purity to 60-80% but reduces yield. For most applications, crude prep is sufficient.

   Timing: Day 0

3. 03

   Step 3: Culture conditions

   Plate motor neurons on PDL/laminin at 20,000-40,000 cells/cm². Culture in neuronal culture medium + neuronal supplement + BDNF (10 ng/mL) + GDNF (10 ng/mL) + CNTF (10 ng/mL). These neurotrophic factors are essential for motor neuron survival.

   Materials: Neuronal culture medium, neuronal supplement, BDNF, GDNF, CNTF

   Note: Motor neurons are extremely fragile and require all three neurotrophic factors for >7 day survival

   Timing: Day 0

4. 04

   Step 4: AUMsilence sdASO treatment

   Treat motor neurons at DIV 3-7 with AUMsilence sdASO at 10 μM. Motor neurons are highly sensitive to osmotic stress. Target ALS-relevant genes: SOD1, TDP-43 (TARDBP), FUS, C9orf72.

   Materials: AUMsilence sdASO (10 μM working concentration)

   Note: Dial the concentration down if the cells prove sensitive. Motor neuron fragility requires conservative dosing.

   Timing: Day 3-7

5. 05

   Step 5: Validation and morphology

   Validate knockdown at 48-72h by qRT-PCR and Western blot. Assess motor neuron morphology by non-phosphorylated neurofilament H antibody or ChAT (choline acetyltransferase) staining. Confirm preserved axon length and branching complexity.

   Materials: qPCR reagents, non-phosphorylated neurofilament H or ChAT antibodies, ImageJ for morphometry

   Note: AUMsilence sdASO preserves motor neuron morphology; lipofection causes severe axon retraction

   Timing: Days 5-10

Neuronal cell lines (Neuro2A, SH-SY5Y, PC12)

1. 01

   Step 1: Cell culture

   Culture Neuro2A (mouse neuroblastoma) in DMEM + 10% FBS. Culture SH-SY5Y (human neuroblastoma) in DMEM/F12 + 10% FBS. Culture PC12 (rat pheochromocytoma) in RPMI + 10% horse serum + 5% FBS.

   Materials: Cell line-specific media formulations

   Note: Cell lines proliferate rapidly unlike primary neurons. Useful for screening, and they do not fully recapitulate primary neuron biology.

   Timing: Ongoing maintenance

2. 02

   Step 2: Differentiation (optional)

   For neuronal phenotype: differentiate SH-SY5Y with retinoic acid (10 μM, 5 days) + BDNF. Differentiate PC12 with NGF (50 ng/mL, 5-7 days). Neuro2A can be used undifferentiated or with retinoic acid.

   Materials: Retinoic acid, BDNF, NGF

   Note: Differentiation increases neuronal characteristics (neurite outgrowth, synaptic markers) but reduces proliferation

   Timing: 5-7 days for differentiation

3. 03

   Step 3: AUMsilence sdASO treatment

   Add AUMsilence sdASO at 10 μM. Treat for 24-48h. Cell lines have faster protein turnover: shorter incubation sufficient.

   Materials: AUMsilence sdASO (10 μM)

   Note: Cell lines useful for initial target validation, dose-response curves, and screening before expensive primary neuron experiments

   Timing: 24-48h treatment

4. 04

   Step 4: Screening applications

   Use cell lines for: (1) ASO sequence optimization (test 3-5 ASOs per target), (2) Concentration optimization, (3) Time-course studies, (4) Preliminary target validation. Then confirm findings in primary neurons.

   Materials: Multiple ASO sequences per target

   Note: Cell line data should always be validated in primary neurons for publication

   Timing: Varies by application

Essential controls for neuronal RNA silencing

• Non-targeting control ASO: Distinguish sequence-specific knockdown from off-target effects
• Vehicle control (no ASO): Establish baseline neuronal function and gene expression
• Positive control (GAPDH or ACTB knockdown): Verify neuronal competence for ASO uptake and RNase H1 activity
• Independent ASO sequence verification: Rule out off-target effects from single ASO sequence

Optimization guide for challenging neuronal applications

Knockdown validation by qRT-PCR

Protein validation by Western blot and immunocytochemistry

Electrophysiology: Whole-cell patch clamp

Viability and neuronal health assessment

Morphological analysis and neurite integrity

Synaptic function and plasticity

Calcium imaging (network activity)

Microelectrode array (MEA) recording

Critical controls for validation

• Untreated neurons
  Purpose: Baseline expression, viability, and electrophysiological properties
  Culture neurons identically but without ASO addition. Use for all comparisons (qPCR, Western blot, patch clamp).
• Non-targeting control ASO
  Purpose: Control for off-target effects, immune activation, and ASO-related toxicity
  Use AUM non-targeting control ASO at the same concentration and timing as the experimental ASO. Should show no knockdown of target gene, normal viability, and normal electrophysiology.
• Positive control ASO (housekeeping gene)
  Purpose: Verify ASO uptake and RNase H1 activity in neurons
  Optional but recommended: Use GAPDH or ACTB-targeting ASO to confirm 70-95% knockdown. Validates that neurons are competent for ASO uptake.
• Viability control (dead neurons)
  Purpose: Validate viability assay sensitivity
  Induce neuronal death with glutamate (100 μM, 30 min) or hydrogen peroxide (200 μM, 1h) as positive control for cell death assays. Should show >80% death.
• Electroporation/lipofection comparison
  Purpose: Compare AUMsilence sdASO with a conventional transfection method in the same experiment
  Transfect parallel cultures with cationic lipid reagents or electroporation systems. Compare viability, morphology, and electrophysiology. Toxicity and neurite retraction can appear in the transfected cultures.
• Time course control
  Purpose: Assess optimal validation timepoint for target gene
  Harvest neurons at 24h, 48h, 72h, 96h post-treatment. Perform qPCR and Western blot to define mRNA and protein knockdown kinetics. Accounts for protein half-life variability.
• Dose-response control
  Purpose: Confirm concentration-dependent knockdown and identify optimal dose
  Test at minimum 3 concentrations (e.g., 5 μM, 10 μM and 20 μM AUMsilence sdASO). Knockdown should correlate with concentration. Plot dose-response curve to identify EC50.
• Independent ASO sequence verification
  Purpose: Confirm on-target specificity with second independent ASO
  Design and test 3-5 different ASOs targeting non-overlapping regions of same mRNA. Concordant knockdown and phenotype across independent ASOs confirms on-target specificity and rules out off-target effects. This is the standard for target validation in neuroscience.

Best practices

• Safety note: Always verify target gene function before knockdown to avoid unexpected phenotypes. Some genes are essential for neuronal survival or have unknown roles in neuronal maintenance
• Use biological triplicates (n=3 independent neuronal cultures, different dissection dates) for statistical analysis
• Validate knockdown at both mRNA (qRT-PCR, 48-72h) and protein (Western/ICC, 72-96h) levels
• Perform electrophysiology (patch clamp or MEA) to confirm neuronal function is preserved: this is the functional validation for AUMsilence sdASO in neurons
• Include morphological analysis (neurite length, spine density) to demonstrate lack of neurotoxicity
• Use non-targeting control ASO at same concentration to rule out off-target effects
• For functional assays (LTP, calcium imaging), verify knockdown in same neurons used for functional readout
• Report both knockdown efficiency and cell viability in all publications and presentations
• Include an electroporation or lipofection comparison in parallel cultures
• For ion channel targets, correlate the mRNA and protein knockdown you measure with the functional current reduction you measure in the same cells
• Use appropriate statistical tests: t-test for 2-group comparisons, one-way ANOVA with post-hoc for multiple groups, two-way ANOVA for interaction effects (e.g., genotype × treatment)
• Archive representative electrophysiology traces, calcium imaging movies, and microscopy images for supplementary data