Choosing an Antigen for Membrane Protein Antibody Discovery

Membrane protein antibody discovery succeeds or fails on a decision made before any animal is immunised and any library is panned: what physical form the antigen will take. A GPCR, a claudin or an ion channel cannot simply be expressed as a soluble protein, and the format you choose — more than affinity, more than library size — determines whether your binders will recognise the target on a cell.

Why the Antigen Decides a Membrane Protein Antibody Discovery Campaign

Soluble proteins make this easy. Express the protein, purify it, coat a plate or inject a mouse, and the molecule your antibody sees is more or less the molecule that exists in biology. Multi-pass membrane proteins break that assumption completely.

An integral membrane protein is held in shape by the lipid bilayer around it. Remove the bilayer and the protein does not simply become soluble — it unfolds, aggregates, or adopts a conformation that no longer resembles anything a cell displays. This is why complex membrane protein families remain difficult targets despite being among the most attractive target classes in drug development: the bottleneck is not the antibody, it is the antigen (Dodd et al., BioDrugs 2018).

The practical consequence is a failure mode that shows up late and expensively. A campaign runs, binders come out of the screen with good ELISA signal and reasonable affinities, and then none of them stain the target on a cell. The antigen was the wrong shape from the start, and every downstream number inherited that error.

There are seven workable answers to “what form should the antigen take?”. Each one fails in a specific, predictable way. This note describes what each format is, when it is the right choice, when it will mislead you, and how to interpret the data it produces.

When to Buy an Antibody Instead of Raising One

Often, buying is the right answer. If a well-characterised antibody already exists for your target and has been validated in the application you need, buy it and the rest of this note does not apply. Raising a binder is slow and expensive, and it is worth being honest about when it is unnecessary.

The cases below are the ones where buying does not solve the problem. They are not exotic — between them they cover a large fraction of membrane protein work.

• No antibody exists. Commercial coverage is concentrated on a relatively small set of well-studied targets. For a transporter, an orphan receptor, or an under-studied family member, there may be nothing to buy.
• The antibody exists but does not work. Antibody validation failure is a well-documented and expensive problem, and it falls hardest on multi-pass targets where conformational epitopes are easily lost. Confirming that a purchased antibody binds the folded, membrane-embedded protein — rather than a denatured epitope or something else entirely — requires a native-conformation antigen. This is the most common reason to buy a membrane protein antigen without any intention of raising an antibody at all.
• Wrong species, ortholog, or isoform. Work outside the standard model organisms, comparative studies across orthologs, and isoform-selective questions routinely fall outside catalogue coverage.
• You need a function, not a detection reagent. Most catalogue antibodies are characterised for Western blot, immunohistochemistry and flow cytometry. Agonism, antagonism, ligand blocking and internalisation are rarely characterised and frequently absent.
• You need a conformational tool for structural or mechanistic work. Binders used as crystallisation chaperones, cryo-EM fiducials or state-selective probes have to be generated against a correctly folded target, and generally in-house.
• You need the sequence, or you need to own it. Engineering, humanisation and bispecific construction all require the variable region, which catalogue suppliers do not provide. Programmes intended to become a therapeutic, diagnostic or CAR construct additionally need freedom to use the binder.

Scope note: this guide addresses antigen format for antibody work — raising, screening and validating binders. The same format logic applies to ligand–receptor binding studies, serology and neutralisation assays with viral envelope antigens, and defined standards for assay development. The questions below transfer directly; the weighting shifts, with quantitative applications favouring the defined-molarity formats and serology favouring native conformation.

Five Questions That Choose Your Membrane Protein Antigen Format

Work through these in order. In most cases the first three narrow the field to one or two viable options, and the last two decide between them.

1. How much of your target is actually outside the cell?

Count the extracellular residues, not the total length. A single-pass receptor with a 300-residue ectodomain is a different problem from a 4TM claudin whose extracellular loops total perhaps fifty residues, or a 7TM GPCR whose loops are shorter still. If the target has a substantial, independently folding ectodomain you may not need a membrane-format antigen at all. If the accessible surface is a handful of short loops constrained by the membrane, no soluble construct can reproduce it, and the choice collapses to the membrane-format options.

2. Are you raising antibodies, or only screening and characterising them?

These are different jobs with different constraints. Immunisation is an in vivo procedure with dose, endotoxin and immunogenicity requirements. Screening is an in vitro binding measurement with reproducibility and density requirements. Nucleic-acid formats are excellent immunogens but cannot be coated on a plate. Detergent-solubilised protein is a well-behaved biochemical reagent but a poor immunogen. Many programmes use two different formats for the two stages — a feature, not an inconsistency, provided the specificity claim is confirmed on a native-conformation reagent.

3. Does your assay require conformational epitopes?

If the antibody must recognise the folded, membrane-embedded protein on a living cell — flow cytometry, live-cell imaging, surface staining, functional modulation, and any therapeutic or cell-engineering application — the antigen has to present those epitopes. If you only need to detect denatured protein on a blot, the requirement disappears. This is the single most common source of wasted campaigns: a linear-epitope antibody raised against a peptide will work beautifully in Western blot and fail entirely in flow cytometry, because the epitope it recognises does not exist as a contiguous exposed surface in the folded protein.

4. Do you need to discriminate between close homologues?

Claudins, chemokine receptors and CD-antigen families contain members whose extracellular surfaces differ by only a few residues. If your conclusions depend on the antibody distinguishing one family member from another, you need counter-screening antigens in the same format as your primary screening antigen. Comparing a hit against a homologue presented in a different format confounds two variables at once: a negative result may mean genuine selectivity, or may mean the homologue simply presents its loops differently in a micelle.

5. What quantitative claim will you need to make?

A qualitative “binds / does not bind” result tolerates almost any format. A reported dissociation constant does not. Particle-based formats present many copies of the target on one particle, so antibody binding is avidity-enhanced and the measured affinity is apparent rather than intrinsic. Detergent-solubilised or nanodisc-reconstituted monomeric protein supports cleaner 1:1 kinetics. If a number is going into a report, decide now which compromise you are making and state it.

Seven Membrane Protein Antigen Formats, and How Each One Fails

Ordered roughly from least to most faithful to the native protein. The first two are cheap and easy and are the right answer more often than membrane protein specialists like to admit — but only for targets that meet a specific condition.

1. Synthetic peptide from an extracellular loop — linear epitopes only

A short synthetic peptide corresponding to a predicted extracellular loop or the N-terminal segment, usually conjugated to a carrier such as KLH to make it immunogenic. Fast, inexpensive, fully defined, no expression required.

Use when you need a detection antibody for denatured protein — Western blot, or immunohistochemistry on fixed and permeabilised tissue where the epitope is already unfolded. Also reasonable for anti-tag or anti-terminus reagents.

How it fails: a free peptide has no defined three-dimensional structure. Antibodies raised against it recognise a linear sequence that, in the folded protein, is bent, buried or constrained by the membrane. The result works on a blot and fails on a cell — and because it fails silently, it is usually discovered only after significant downstream investment.

BioHippo: the peptide catalogue covers bioactive and hormone peptides, not custom extracellular-loop immunogens for the membrane targets discussed here. Peptide immunogens for these targets would need to be custom-synthesised.

2. Soluble ectodomain or extracellular fragment — well-behaved, but needs a real domain

The extracellular portion expressed on its own as a secreted, soluble construct, typically with a purification tag and often as an Fc fusion. High yield, standard purification, monomeric and defined.

Use when your target is a single-pass receptor with a genuine, independently folding extracellular domain — a growth factor receptor, a cytokine receptor, an immunoglobulin-superfamily CD antigen. For these targets this is the correct format and the membrane-format options are unnecessary complexity.

How it fails: for multi-pass targets there is nothing to express. The extracellular surface of a 4TM or 7TM protein is not a domain; it is a set of short loops whose conformation is created by the transmembrane helices anchoring them (Figure 1).

BioHippo: the recombinant protein catalogue includes ectodomain and extracellular-fragment constructs for many single-pass targets, catalogued as standard recombinant proteins rather than under the transmembrane protein lines.

3. Detergent-solubilised full-length protein — defined stoichiometry, uncertain conformation

The complete membrane protein extracted from the membrane and held in solution by a detergent micelle that substitutes for the bilayer around the transmembrane helices. Full-length sequence, defined molar concentration, monomeric, standard chromatography.

Use when you need a defined, quantifiable reagent for biochemistry — clean SPR or BLI kinetics, stoichiometric binding studies, or a characterised standard. Also the practical choice when the assay format cannot tolerate particles.

How it fails: a micelle is not a bilayer. It lacks the lateral pressure, thickness and specific lipid contacts that shape the native fold, and conformational epitopes can be lost or distorted — to a degree that varies by target and by detergent, and that is rarely characterised on the product. The protein is also only stable while the detergent stays above its critical micelle concentration, so dilution during an assay can destabilise it.

BioHippo: 17 SKUs in the Detergent-Solubilized Membrane Proteins collection, including CD20 (MS4A1), HTR3A, biotinylated and SLC23A2. Several targets are stocked in both detergent and VLP format, which is what makes a direct format comparison possible within one supply chain.

4. Nanodisc-reconstituted protein — a real bilayer with a defined molarity

The purified protein reconstituted into a nanodisc: a small patch of genuine lipid bilayer whose rim is encircled by an amphipathic scaffold (membrane scaffold protein, or a peptide scaffold as in the PeptiNanodisc format). The protein sits in lipid, not detergent, but the resulting particle is soluble, monodisperse and defined in molar terms.

Use when the trade-off between the two formats above is the thing blocking you — when you need a native lipid environment and a defined molar concentration in the same tube. That combination is what makes nanodiscs the conventional reagent for SPR and BLI on multi-pass targets, for stoichiometric binding measurements, and as a defined counter-screening antigen. Because each disc carries a small, controlled number of copies, avidity is far lower than on a particle.

How it fails: reconstitution is target-specific work — lipid composition, scaffold size and protein-to-lipid ratio all matter, and not every target reconstitutes well. Orientation within the disc is not necessarily uniform, and a disc is a fragment of membrane, not a cell surface, so it does not reproduce the crowding, glycocalyx or partner interactions of a real cell. Nanodiscs are also weak immunogens relative to particles, so this is a screening and characterisation reagent rather than a primary immunogen.

BioHippo: 8 PeptiNanodisc SKUs covering high-value antibody and CAR targets — CLDN18.2, CLDN18.1, CLDN6, CCR8, CCR8 biotinylated, CCR2 biotinylated, GPRC5D and CD19. All are full-length, HEK293-expressed, and the biotinylated versions are built for streptavidin sensor immobilisation. Note the CLDN18.1 / CLDN18.2 pair: matched-format isoform counter-screening, which is exactly the problem described in question 4 above.

5. VLP-displayed full-length protein — native bilayer, particulate immunogen

The full-length protein expressed in mammalian cells and captured, still embedded in a patch of the host cell membrane, on the surface of a budding virus-like particle. The protein is never removed from a lipid bilayer at any point. Native lipid environment, mammalian post-translational modifications, conformational epitopes retained, particulate and therefore intrinsically immunogenic, yet handled like a soluble reagent.

Use when your target is multi-pass and the antibody must recognise it on a cell. This covers most GPCR, claudin, transporter and ion-channel campaigns, and it is the format that supports both stages — immunisation and screening — with the same reagent, which removes one variable from the comparison. VLPs carrying conformationally complex membrane proteins have been used as soluble probes to rank the binding kinetics of antibody panels against GPCRs (Willis et al., Biochemistry 2008).

How it fails: two limitations, both manageable if acknowledged. Protein orientation on the particle is not uniform, so extracellular-loop accessibility varies between particles and preparations. And the particle carries the host cell's own membrane proteins alongside your target — which makes a matched control particle mandatory rather than optional (see below).

BioHippo: 160 SKUs in the VLP Membrane Proteins collection, spanning chemokine receptors (CCR1–CCR9, CXCR2/4/5/6, CX3CR1), adrenergic and adenosine receptors, the claudin family, CD antigens and viral envelope proteins. Fluorescent-conjugated variants are available for a subset, for use as direct flow cytometry and imaging reagents.

6. mRNA or DNA encoding the target — the animal or cell makes the antigen

Rather than supplying protein, you supply the coding sequence. Transfected cells — or the immunised animal's own cells — express the target and display it at the cell surface in a fully native membrane, with endogenous post-translational modifications and no purification step.

Use when you are immunising and want the most faithful possible presentation, or you need to generate target-positive cells for cell-based panning, flow cytometry counter-screening or CAR functional assays. Nucleic-acid immunisation combined with display selection has been used to isolate antibodies against the short extracellular loops of a GPCR (van der Woning et al., mAbs 2016).

How it fails: you cannot coat a plate with it, immobilise it on a sensor chip, or use it in any assay that requires a defined quantity of purified antigen. Expression level varies between transfections, so antigen density is a variable you control only indirectly. This format almost always needs a protein-format partner for the screening stage.

BioHippo: 47 SKUs of membrane-protein-encoding mRNA in the mRNA Reagents collection, tagged for antibody-target and CAR-target use — including CCR4, PD-L1 (CD274), BCMA (TNFRSF17), EGFR, CD3E and Claudin 18.2.

7. Whole cells overexpressing the target — most native, least defined

Intact cells engineered to overexpress the target, used directly as immunogen or as the panning and screening surface. Fully native context, supports functional readouts, no antigen preparation.

Use when the target requires an authentic cellular environment — a partner subunit, a specific lipid, or a signalling context — or when the desired antibody function is agonism or antagonism rather than binding, and must be read out on live cells. Cell-based panning strategies have been developed specifically for membrane protein targets (Kelil et al., Commun Biol 2021).

How it fails: your target is a small fraction of the total surface protein presented. The overwhelming majority of the immune or selection response goes to irrelevant surface antigens, which makes the campaign inefficient and makes rigorous subtractive controls — parental cells lacking the target — non-negotiable rather than good practice.

BioHippo: not currently stocked as pre-made target-overexpressing lines for these targets. The mRNA line above, transfected into a host line from the cell lines catalogue, is the practical route to generating target-positive cells in-house.

The Matched Control That Makes Membrane Protein Binding Data Interpretable

Every membrane-format antigen carries material that is not your target. This is the most under-appreciated issue in membrane protein binding work, and it has a straightforward remedy.

A VLP is assembled by budding from a producer cell. It carries the host cell's own surface proteins, glycans and lipids in the same membrane patch that carries your target. A detergent preparation carries co-purifying proteins and the detergent itself. A transfected cell carries its entire native surface proteome.

So a positive binding signal against a target-displaying particle has at least three possible explanations: the antibody binds your target; the antibody binds a host-cell component the particle happens to carry; or the antibody binds the particle scaffold. Nothing in the target-particle result alone distinguishes them.

An isotype control particle is produced by the same process from the same host cell, but without your target displayed. Run in parallel at matched particle concentration, it separates target-specific binding from everything else the particle carries. This applies to every assay in the workflow: plate-coated ELISA screening, SPR and BLI reference channels, flow cytometry gating, and — often overlooked — the immunisation itself, where a control-particle bleed tells you how much of the polyclonal response went to particle background rather than to your target.

BioHippo stocks 6 matched VLP isotype controls produced from the same platform as the target-displaying particles, in untagged, His-tagged, EGFP-fluorescent and mCherry-fluorescent versions. Match the control's conjugate to your target particle's conjugate — a His-tagged target particle should be controlled with a His-tagged blank, and a fluorescent staining reagent with the matching fluorescent blank, so that detection chemistry is not itself a variable.

Worked Examples: CD20, CCR4 and Validating an Antibody You Already Have

CD20 — the same target in two formats

CD20 (MS4A1) is a four-pass membrane protein and a validated therapeutic target. Its extracellular exposure is essentially one large loop, and the clinically important epitopes are conformational — which is why CD20 is a textbook case for antigen format mattering.

BioHippo carries CD20 as both a VLP-displayed and a detergent-solubilised preparation. A sensible division of labour: use the VLP for immunisation and primary screening, because that is where conformational fidelity determines whether the campaign produces cell-binding antibodies at all; use the detergent preparation where you need a defined molar concentration and clean monomeric kinetics; and treat any affinity discrepancy between the two as informative about the epitope rather than as an error in one of them.

CCR4 — protein for screening, mRNA for cells

CCR4 is a seven-pass chemokine receptor with minimal extracellular surface. BioHippo carries it as a VLP and as mRNA. These are complementary rather than competing: the VLP is the immunogen and the plate-coating and biosensor reagent; the mRNA generates CCR4-positive cells for the cell-binding confirmation that decides whether a hit is real.

Because CCR4 belongs to a large family with shared ligands, subtype selectivity is a substantive requirement rather than a formality. The chemokine receptor VLP set — CCR1 through CCR9 plus the CXCR members — allows a matched-format counter-screening panel, so that a selectivity result reflects the antibody rather than a difference in how each receptor was prepared. Where the selectivity claim has to be quantitative, the biotinylated CCR8 and CCR2 nanodiscs give the same family in a defined-molarity, sensor-ready format.

Validating an antibody you already have

The third case involves no antibody generation at all. A laboratory has a commercial or in-licensed antibody against a multi-pass target and needs to establish, before building a body of work on it, that it recognises the folded protein rather than a denatured or off-target epitope.

The experiment is short. Titrate the antibody against the target-displaying particle and against a matched isotype control particle in parallel, at equal particle concentration. A genuine conformational binder gives a dose-dependent signal on the target particle and a flat response on the control. An antibody that responds equally to both is binding something the particle carries, not your target. An antibody that binds neither, but is positive by Western blot, is a linear-epitope reagent — still useful for blotting, but it will not perform in flow cytometry or live-cell work.

This costs a fraction of a discovery campaign and resolves a question that otherwise surfaces much later, after the antibody has been used to generate data.

Three Numbers to Read Carefully From Particle-Format Antigens

Affinities from particle formats are apparent, not intrinsic

A particle presents many copies of the target, so a bivalent IgG can engage two at once. The resulting avidity makes the measured off-rate slower and the apparent affinity tighter than the true monovalent interaction. This is not a defect — avidity is also what happens on a cell — but a K_D derived from a particle surface should be reported as apparent, and should not be compared directly against a K_D measured on monomeric protein. Where an intrinsic number is required, a nanodisc or detergent preparation is the appropriate reagent.

Antigen density is usually unknown

Copy number per particle is rarely specified, so total protein mass does not convert into a molar concentration of accessible target. That makes absolute quantification unreliable and makes run-to-run consistency dependent on using a single lot for a given comparison. Where a defined molar concentration is essential, a detergent or nanodisc preparation is the better reagent.

A negative result may be an orientation artefact

Because display orientation is mixed, a given extracellular loop may be under-represented in a particular preparation. A clean negative against one antigen lot is weaker evidence than it appears. Confirm negatives on cells before concluding that an antibody does not bind the target.

Pre-Flight Checklist Before You Order a Membrane Protein Antigen

• Count the extracellular residues. Check the topology in UniProt. This determines whether a soluble construct is even conceptually possible.
• Confirm the expressed region. Product pages state the amino acid range. A partial fragment may omit the loop your epitope sits in.
• Decide immunogen vs. screening reagent. They may be different formats. Plan both before ordering either.
• Order the matched control with the antigen, not afterwards. Isotype control particle, mock detergent prep, or parental cell line.
• Check endotoxin if injecting. Relevant for in vivo use only; irrelevant for plate coating or sensor immobilisation.
• Build the counter-screen panel now. If selectivity matters, source the homologues in the same format at the same time.
• Match conjugates across the panel. Tag and fluorophore should be consistent between target and control.
• Plan the cell-based confirmation. No format substitutes for showing the antibody binds the target on a cell.

Membrane Protein Antigen Format Decision Matrix

Conformational fidelity is necessary but not sufficient — a format can present correct epitopes and still be the wrong choice if it cannot support the assay you need to run.

FAQ: Choosing an Antigen for Membrane Protein Antibody Discovery

Can I immunize with one antigen format and screen with another?

Yes, and it is common practice — for example immunizing with mRNA or VLPs and screening on plate-coated VLPs or nanodiscs. The requirement is that at least one stage uses a native-conformation antigen and that the final specificity claim is confirmed on cells. What does not work is immunising and screening both on formats that lack conformational fidelity, because nothing in the workflow would then detect the problem.

Why is the same target offered as both a VLP and a detergent preparation?

Because they answer different questions. The VLP preserves the lipid environment and the conformational epitopes, which is what you need for raising and screening cell-binding antibodies. The detergent preparation is monomeric with a defined molar concentration, which is what you need for stoichiometric binding measurements and clean kinetics. Programmes that use both typically use the VLP for discovery and the detergent format for characterization.

What is the difference between a nanodisc and a VLP antigen?

Both keep the protein in a lipid bilayer, but they differ in scale and in what they let you measure. A VLP is a cell-derived particle carrying many copies of the target plus the producer cell's own surface proteins, which makes it a strong immunogen and an avidity-enhanced binding surface. A nanodisc is a small, defined, monodisperse bilayer patch carrying very few copies, which makes it a poor immunogen but a clean reagent for 1:1 kinetics and for matched-format counter-screening at a known molar concentration.

Do I really need an isotype control particle if I already have an isotype control antibody?

Yes — they control for different things. An isotype control antibody tests whether your detection antibody binds non-specifically. An isotype control particle tests whether your antibody binds something on the particle that is not your target. A hit can pass the first control and fail the second, and that failure is precisely the one that shows up later as an antibody that binds the antigen but not the cell.

My antibody binds the VLP but not cells. What went wrong?

The most likely explanations, in rough order: the antibody binds a host-cell component carried by the particle rather than your target (run the isotype control particle); the epitope is on an intracellular loop exposed by inverted particles but inaccessible on an intact cell; expression on your test cells is below the detection threshold; or the epitope is masked on the cell by a binding partner or glycan not present on the particle. The first two are format artefacts, and are why the control particle and the topology check belong at the start of the workflow.

How much membrane protein antigen should I order?

The two applications have very different requirements. A rodent immunisation campaign consumes substantially more material than a screening campaign, because dosing is repeated across several animals; plate-based screening consumes very little per plate, since coating typically uses microgram-per-millilitre concentrations. Scope the immunisation requirement first, since it dominates, and order screening quantities separately — ideally as a single lot for any comparison you intend to make across plates.

Does endotoxin specification matter for my application?

Only if the material is going into an animal. Endotoxin drives innate immune activation and can distort an immunisation response, so a low endotoxin specification is meaningful for in vivo use. For plate coating, biosensor immobilisation, or any purely in vitro binding assay it has no bearing on the result.

Can I use these antigens for structural work?

Detergent-solubilised and nanodisc preparations are the conventional starting points for structural studies, though structural work generally requires target-specific optimisation of construct boundaries, detergent or lipid, and stabilising mutations well beyond what a catalogue preparation provides. VLP preparations are heterogeneous in composition and are not suitable for high-resolution structure determination, though negative-stain electron microscopy is routinely used to confirm particle integrity.

Membrane Protein Antigens Available From BioHippo, by Format

Catalogue coverage for each format discussed above. Counts are of active listings at the time of writing; confirm current specifications, expressed region, tag and pack size on each product page before ordering.

Synthetic extracellular-loop peptides and pre-made target-overexpressing cell lines are not currently stocked for these targets. Several targets — including CD20, CCR8 and members of the claudin family — are available in more than one format, which allows a direct format comparison within a single supply chain.

If your target is multi-pass and the antibody has to work on a cell, start with a native-bilayer format and order the matched control at the same time. If you need a defined molar quantity for kinetics, add a detergent or nanodisc preparation. If you are immunising, consider whether a nucleic-acid format gives you a better presentation than any purified protein can. Tell a BioHippo technical specialist your target, its topology, and which stage you are at, and we will scope the format and the control panel with you. Request a quote or technical consultation →

References

1. Dodd RB, Wilkinson T, Schofield DJ. Therapeutic monoclonal antibodies to complex membrane protein targets: antigen generation and antibody discovery strategies. BioDrugs. 2018;32(4):339–355. doi:10.1007/s40259-018-0289-y (PMID 29934752)
2. Willis S, Davidoff C, Schilling J, Wanless A, Doranz BJ, Rucker J. Virus-like particles as quantitative probes of membrane protein interactions. Biochemistry. 2008;47(27):6988–6990. doi:10.1021/bi800540b (PMID 18553929)
3. van der Woning B, De Boeck G, Blanchetot C, et al. DNA immunization combined with scFv phage display identifies antagonistic GCGR specific antibodies and reveals new epitopes on the small extracellular loops. mAbs. 2016;8(6):1126–1135. doi:10.1080/19420862.2016.1189050 (PMID 27211075)
4. Kelil A, Gallo E, Banerjee S, Adams JJ, Sidhu SS. CellectSeq: in silico discovery of antibodies targeting integral membrane proteins combining in situ selections and next-generation sequencing. Commun Biol. 2021;4:561. doi:10.1038/s42003-021-02066-5 (PMID 33980972)

Catalogue counts, formats and product availability are accurate at the time of writing and are subject to change. Figures in this note are illustrative schematics, not experimental data. Confirm current specifications, expressed region and intended-use statements on each product page before purchase.

All products referenced are supplied for Research Use Only (RUO). Not for use in diagnostic or therapeutic procedures.