Detergent-Free Membrane Protein Purification: Nanodiscs and Beyond

You've screened detergents. You found one that solubilizes your GPCR without aggregation. But there's a lingering concern: Is your protein really native-like in a detergent micelle? Are you losing activity? Could the detergent be interfering with your structural studies or functional assays?

Enter detergent-free purification methods: nanodiscs, SMALPs, and amphipols. These approaches keep membrane proteins in a native-like lipid environment (or a stabilizing polymer belt) without the complications of detergent micelles. Here's how they work, when to use them, and how to prepare samples for Cryo-EM.

Key Takeaways

• Detergent-free methods maintain native-like environment (better stability, function)

• Nanodiscs: Protein in lipid bilayer disc (~10 nm), wrapped by scaffold protein (MSP)

• SMALPs: Direct extraction from membranes, no detergent needed, preserves native lipids

• Amphipols: Polymers wrap around protein, stable for months, excellent for Cryo-EM

• β2-adrenergic receptor: Nobel Prize-winning structure used T4-lysozyme fusion + thermostabilizing mutations

• Cryo-EM requires: 1-5 mg/mL, >90% monodisperse, small micelles or nanodiscs

Why Detergent-Free Matters

Problems with Detergents

1. Protein may not be in native state

• Detergent micelles ≠ lipid bilayer

• Transmembrane regions interact with detergent, not native lipids

• Can alter protein conformation (inactive state vs active state)

2. Large micelles interfere with Cryo-EM

• Detergent micelles add ~50-100 kDa mass around protein

• Obscures protein features (lower resolution)

• Particle alignment difficult (micelle creates "blob")

3. Detergents can denature proteins

• Even "mild" detergents (DDM, LMNG) cause partial unfolding for some proteins

• Activity loss over time (hours to days)

• Cannot use for long-term studies

4. Hard to remove completely

• Low-CMC detergents (LMNG, digitonin) stay bound

• Interferes with crystallization, biophysical assays

• Detergent in sample causes artifacts (DLS, AUC)

Detergent-Free Approaches Solve These Problems

Nanodiscs:

• Protein in native lipid bilayer (2D disc)

• Defined size (~10-12 nm diameter)

• Better for Cryo-EM (uniform shape, easier particle alignment)

SMALPs:

• Extract protein directly from membranes with native lipids

• No detergent step at all (preserves annular lipid shell)

• Faster than nanodiscs (no reconstitution)

Amphipols:

• Polymer belt wraps around transmembrane region

• Very stable (proteins soluble for months at 4°C)

• Small size (better for high-resolution Cryo-EM)

Method 1: Nanodiscs (Lipid Bilayer in a Disc)

Concept

Self-assembling system:

• Membrane protein + phospholipids + MSP (membrane scaffold protein)

• MSP is an amphipathic α-helical protein (22 kDa)

• MSP wraps around lipid bilayer (forms disc, ~10-12 nm diameter)

• Protein sits in native-like lipid environment, but disc is water-soluble

Components

1. MSP (Membrane Scaffold Protein)

• Amphipathic α-helix (derived from apolipoprotein A-I)

• Wraps around edge of lipid bilayer

• Sizes:

  • MSP1D1: Standard (~10 nm diameter, holds ~160 lipids)

  • MSP1E3D1: Larger (~13 nm diameter, holds ~250 lipids, for bigger proteins)

  • MSP2N2: Extra-large (~17 nm, for large complexes)

2. Phospholipids

• Synthetic lipids: POPC, POPE, POPG, POPS

• Native lipids: Brain polar lipids, E. coli polar lipids

• Ratio: Protein:MSP:Lipids = 1:5:50 to 1:10:100 (needs optimization)

Advantages

1. Native-like lipid environment

• Protein in bilayer (not detergent micelle)

• Better stability, closer to physiological state

• Can test specific lipid requirements (e.g., cholesterol for GPCRs)

2. Defined size

• Uniform particle size (good for Cryo-EM)

• Easier to classify particles (clear shape in EM images)

3. Control lipid composition

• Test different lipids systematically

• Add specific lipids (cholesterol, cardiolipin, PIP2)

4. Functional studies

• Better for activity assays (native-like environment)

• Can study lipid-protein interactions

Disadvantages

1. Requires detergent first

• Solubilize protein in detergent (DDM)

• Then reconstitute into nanodiscs

• Two-step process (more work than SMALPs)

2. Optimization needed

• Lipid type, MSP:lipid ratio, protein:MSP ratio

• Can take weeks to optimize

3. MSP adds mass

• MSP is 22 kDa (2 copies per disc = 44 kDa total)

• Can interfere with some structural studies

• Need to account for MSP mass in EM

Protocol: Nanodisc Reconstitution

Starting material: Protein purified in detergent (DDM, from Part 1)

Step 1: Prepare Components

Protein:

• Concentrate to 5-10 mg/mL (in DDM-containing buffer)

• Keep at 4°C

MSP:

• Express MSP1D1 in E. coli (plasmid available from Addgene)

• Purify by His-tag (MSP has N-terminal His-tag)

• Store at -80°C in 20 mM Tris pH 7.5, 100 mM NaCl

Lipids:

• POPC (1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine) from Avanti Polar Lipids

• Dissolve in chloroform (25 mg/mL stock)

• Store at -20°C under argon

Step 2: Mix Components

Typical reaction (100 μL):

• Protein: 1 μM (1 nmol)

• MSP1D1: 10 μM (10 nmol, 10× excess)

• POPC: 100 μM (100 nmol, 100× excess, ~100 lipids per nanodisc)

• DDM: 0.5% (in buffer)

Buffer:

• 20 mM Tris pH 7.5, 150 mM NaCl, 10% glycerol

Calculation notes:

• Ratio optimization: Test ratios of protein:MSP:lipids = 1:5:50, 1:10:100, 1:15:150

• MSP excess ensures all protein is incorporated

Step 3: Remove Detergent (Critical Step)

Method: Bio-Beads SM-2 (polystyrene resin)

• Add 0.5-1 g Bio-Beads per 1 mL reaction (wet weight)

• Incubate overnight (12-16 hours) at 4°C with gentle rotation

• Bio-Beads absorb DDM → nanodisc self-assembles as detergent is removed

Alternative: Dialysis

• Dialyze against 1000× volume of detergent-free buffer

• Change buffer 3-4 times (every 4-6 hours)

• Slower than Bio-Beads, but works

Step 4: Remove Bio-Beads

• Transfer reaction to new tube (leave Bio-Beads behind)

• Or centrifuge briefly (1,000 × g, 5 min) and collect supernatant

Step 5: Purify Nanodiscs by SEC

Why: Separate protein-containing nanodiscs from empty nanodiscs

Column: Superdex 200 Increase 10/300 GL

Buffer: 20 mM Tris pH 7.5, 150 mM NaCl, 10% glycerol (no detergent)

Expected elution:

• Protein-containing nanodiscs: Elute earlier (~12-14 mL, larger MW)

• Empty nanodiscs: Elute later (~15-16 mL, smaller MW)

Collect: Fractions from protein-containing peak

Step 6: Characterization

TEM (Transmission Electron Microscopy):

• Negative stain with uranyl acetate

• Expected: Circular discs visible (~10-12 nm diameter)

• Confirms nanodisc formation

SEC-MALS:

• Measure MW of nanodisc

• Expected: Protein MW + 2× MSP (44 kDa) + lipids (~120 kDa total for 50 kDa protein)

Activity assay:

• Test ligand binding or transport activity

• Compare to detergent-solubilized protein

Troubleshooting Nanodiscs

Problem: No protein in nanodiscs (empty nanodiscs only)

• Cause: Wrong lipid:MSP ratio, or protein aggregated during reconstitution

• Solution: Optimize ratio, add more protein, keep cold (4°C)

Problem: Heterogeneous sample (multiple peaks in SEC)

• Cause: Multiple oligomeric states, or variable nanodisc sizes

• Solution: Optimize MSP:lipid ratio, use size-defined MSP (MSP1E3D1 for larger proteins)

Problem: Protein inactive in nanodiscs

• Cause: Wrong lipid type, or lipid:protein ratio too low

• Solution: Test different lipids (add cholesterol for GPCRs, cardiolipin for respiratory complexes)

Method 2: SMALPs (Styrene-Maleic Acid Lipid Particles)

Concept

Direct extraction from membranes:

• SMA (styrene-maleic acid copolymer) inserts into lipid bilayer

• Wraps around protein + surrounding lipids (forms nanoparticle, ~10 nm)

• No detergent needed (skip solubilization step entirely)

Advantages

1. Detergent-free from the start

• No solubilization step (no DDM, no LMNG)

• Protein never exposed to detergent

2. Preserves native lipids

• Annular lipid shell around protein retained

• More physiologically relevant than nanodiscs with synthetic lipids

3. Faster than nanodiscs

• No reconstitution step (one-step extraction)

• Less optimization required

4. Works for difficult proteins

• Proteins that denature in detergent may be stable in SMALPs

Disadvantages

1. SMA polymer is negatively charged

• pH-dependent (only works at pH >7)

• Can interfere with some proteins (charge interactions)

2. Cannot use high divalent cations

• Ca²⁺ or Mg²⁺ >1 mM precipitates SMA

• Limits buffer choices

3. Size heterogeneity

• SMALP size variable (8-12 nm)

• Less uniform than nanodiscs (harder for Cryo-EM classification)

4. SMA polymer can interfere with some assays

• Absorption at 280 nm (interferes with protein quantification)

• Can affect some biophysical measurements

Protocol: SMALP Extraction

Starting material: Crude membranes (from Part 1, Stage 2)

Step 1: Add SMA Polymer to Membranes

Membranes:

• Resuspend to 10 mg/mL total protein (in buffer)

SMA polymer:

• Add to 2.5% (w/v) final concentration

• Ratio: 1:2.5 (protein:SMA, typical starting point)

Buffer:

• 50 mM Tris pH 7.5 or 50 mM HEPES pH 7.5 (pH must be >7 for SMA to work)

• 150 mM NaCl

• Protease inhibitors

• No Ca²⁺ or Mg²⁺ >1 mM (precipitates SMA)

Incubate:

• 2 hours at room temperature (RT) or 4 hours at 4°C

• Gentle mixing (rotator or nutator)

Step 2: Clarify

Centrifugation:

• 100,000 × g, 30 min, 4°C

Supernatant: Protein in SMALPs (solubilized) Pellet: Unsolubilized membrane (discard)

Step 3: Affinity Purification

Same as detergent protocol (Part 1), but no detergent in buffers:

Wash buffer:

• 50 mM Tris pH 7.5, 150 mM NaCl

• 20 mM imidazole

• 10% glycerol

• No detergent

Elution buffer:

• 50 mM Tris pH 7.5, 150 mM NaCl

• 250 mM imidazole

• 10% glycerol

Step 4: SEC (Polishing)

Column: Superdex 200

Buffer: 20 mM Tris pH 7.5, 150 mM NaCl, 10% glycerol (no detergent)

Expected: Single peak (protein in SMALP)

Step 5: Characterization

TEM: Observe SMALP particles (circular, ~10 nm)

Activity assay: Test function (compare to detergent)

SEC-MALS: Measure MW (protein + SMA + lipids)

Troubleshooting SMALPs

Problem: Low solubilization efficiency (<20%)

• Cause: SMA:protein ratio too low, or pH too low

• Solution: Increase SMA (2.5% → 5%), check pH >7.5

Problem: Precipitate forms during incubation

• Cause: Ca²⁺ or Mg²⁺ in buffer

• Solution: Use Ca²⁺/Mg²⁺-free buffer, add EGTA (1 mM) to chelate residual Ca²⁺

Problem: Protein inactive in SMALPs

• Cause: SMA polymer interferes with functional site

• Solution: Try alternative polymers (DIBMA, SMA-QA with different charge properties)

Method 3: Amphipols (Amphipathic Polymers)

Concept

Polymer belt wraps around transmembrane region:

• Amphipols are amphipathic polymers (hydrophobic + hydrophilic segments)

• Wrap around protein like a belt (shield TM helices)

• Protein soluble in aqueous buffer, no detergent needed

Types of Amphipols

A8-35 (most common):

• Poly(acrylic acid) backbone with octyl side chains

• MW ~4-5 kDa per polymer chain

• Negatively charged (works best at pH 7-8)

PMAL-C8:

• Poly(maleic acid) with octyl chains

• Alternative to A8-35

Advantages

1. No CMC (don't form micelles)

• Stay bound to protein (don't dialyze out)

• Stable association

2. Very stable

• Proteins remain soluble for months at 4°C

• Better than detergents (no slow aggregation)

3. Small size

• Amphipol belt is thin (~3-4 nm)

• Better for high-resolution Cryo-EM (less obscuring)

4. Compatible with many biophysical techniques

• NMR (smaller than detergent micelles)

• Cryo-EM (excellent, minimal background)

Disadvantages

1. Expensive

• ~$100-200/g (more than most detergents)

2. Can be hard to remove

• If you need to later (e.g., for lipid reconstitution), amphipols are difficult to remove

3. Can block functional sites

• Polymer may sterically block binding sites or conformational changes

• Not ideal for all functional assays

Protocol: Amphipol Exchange

Starting material: Protein in detergent (DDM or LMNG)

Step 1: Mix Protein with Amphipol

Ratio: 1:5 (protein:amphipol, weight ratio)

• Example: 1 mg protein + 5 mg amphipol A8-35

Incubate: 2 hours at 4°C (gentle mixing)

Step 2: Remove Detergent

Method 1: Bio-Beads

• Add Bio-Beads (0.5 g per mL)

• Incubate overnight at 4°C

• Bio-Beads absorb detergent → amphipol remains bound to protein

Method 2: Dialysis

• Dialyze against amphipol-free buffer (1000× volume)

• Change buffer 3-4 times

Step 3: Remove Free Amphipol (Optional)

SEC: Superdex 200

• Protein-amphipol complex elutes earlier (higher MW)

• Free amphipol elutes later

Step 4: Characterization

DLS: Measure particle size (Rh should be small, ~5-8 nm for 50 kDa protein)

SEC-MALS: Confirm MW (protein + amphipol)

Activity assay: Test if protein is functional in amphipols

Functional Validation: Proving Your Protein Is Alive

Test 1: Ligand Binding (for GPCRs)

Radioligand binding assay:

• Mix protein (1-10 nM) + radiolabeled ligand (³H or ¹²⁵I, 0.1-10 nM)

• Incubate (equilibrium binding, 1-2 hours at RT or overnight at 4°C)

• Filter (separate bound from free ligand)

• Count radioactivity (scintillation counter)

Expected:

• Specific binding: Kd = 1-50 nM (high-affinity ligands)

• Bmax = protein concentration (1:1 binding)

Alternative: Fluorescence Polarization (FP)

• Use fluorescent ligand (no radioactivity)

• Measure polarization change upon binding

Test 2: Thermal Stability (DSF)

Differential Scanning Fluorimetry:

• Mix protein + SYPRO Orange dye (binds hydrophobic regions)

• Heat from 20°C → 95°C (gradual)

• Measure fluorescence (increases as protein unfolds)

Expected:

• Tm (melting temperature): Inflection point

• Good detergent/nanodisc: Tm >50-60°C

• Bad detergent: Tm <40°C (unstable)

Use: Compare detergents, identify stabilizing ligands, validate nanodiscs

Test 3: Activity Assay (for Transporters/Enzymes)

Example: GLUT1 (glucose transporter)

Reconstitute into proteoliposomes:

• Mix protein + lipids, remove detergent

• Protein inserts into lipid vesicles (liposomes)

Transport assay:

• Add radiolabeled glucose (¹⁴C-glucose) outside vesicles

• Incubate (glucose transported inside)

• Separate vesicles (filtration or centrifugation)

• Measure radioactivity inside vesicles

Expected:

• Time-dependent uptake (linear for 5-10 min)

• Specific transport (blocked by inhibitors)

Case Study: β2-Adrenergic Receptor (Nobel Prize-Winning Purification)

The Challenge

Target: β2-adrenergic receptor (GPCR, 7 TM helices)

Historical context:

• GPCRs extremely difficult to crystallize (unstable, flexible)

• No GPCR structure until 2007 (despite decades of effort)

The Solution (Brian Kobilka's Lab)

Expression: Sf9 insect cells, baculovirus (5-10 mg/L yield)

Construct Engineering (Critical for Success):

1. N-terminal truncation

• Removed disordered region (residues 1-33)

• Reduces flexibility

2. ICL3 replacement

• Replaced flexible intracellular loop 3 (ICL3) with T4 lysozyme

• T4-lysozyme acts as fiducial marker for crystallization (provides crystal contacts)

• Rigidifies receptor

3. C-terminal truncation

• Removed phosphorylation sites (residues 365-413)

• Reduces heterogeneity

4. Thermostabilizing mutations

• 6 point mutations (identified by alanine scanning)

• Increased Tm by ~20°C (from 40°C → 60°C)

• Mutations: L48A, E122W, Y132G, R149L, Y219A, H296A

Purification:

1. Membrane preparation: Ultracentrifugation

2. Solubilization:

• 1% DDM + 0.2% CHS (cholesteryl hemisuccinate, stabilizing lipid)

3. Affinity:

• Ni-NTA (His-tag at C-terminus)

4. Ligand binding:

• Add alprenolol (inverse agonist, 10 μM)

• Stabilizes inactive state (locks conformation)

5. SEC:

• Superdex 200

• Monodisperse peak (single, sharp)

6. Concentrate:

• To 40-50 mg/mL (for crystallization)

Crystallization:

• Lipidic cubic phase (LCP): Protein in monoolein lipid matrix

• Crystals grown over weeks

• X-ray diffraction: 2.4 Å resolution

Impact:

• First high-resolution GPCR structure (2007, Nature)

• Brian Kobilka awarded Nobel Prize (2012, with Robert Lefkowitz)

• Enabled structure-based drug design for GPCRs (>500 drugs target GPCRs)

Key Lessons:

1. Construct engineering is critical

• Truncations, fusions, thermostabilizing mutations = success

2. Ligand stabilization matters

• Inverse agonist (alprenolol) locks receptor in stable conformation

• Without ligand: Receptor is too flexible, cannot crystallize

3. Lipid matters

• CHS (cholesteryl hemisuccinate) stabilizes GPCR

• Mimics cholesterol (native lipid for GPCRs)

From Purification to Cryo-EM

Sample Requirements

Concentration:

• 1-5 mg/mL (optimal for grid preparation)

• Too low: Particles sparse

• Too high: Aggregation on air-water interface

Monodispersity:

• >90% monodisperse (sharp SEC peak)

• Polydisperse samples: Difficult to classify particles

Stability:

• Stable >1 week at 4°C (time needed for grid screening)

• If unstable: Add crosslinker (GraFix: glutaraldehyde gradient during SEC)

Detergent/nanodisc considerations:

• Small micelles better: DDM, LMNG (vs Triton X-100, large)

• Nanodiscs/amphipols preferred: Uniform shape, easier particle alignment

Grid Preparation (Basic Protocol)

1. Glow-discharge grid

• Makes surface hydrophilic (30 seconds, 15 mA)

2. Apply sample

• 3 μL protein (1-3 mg/mL)

• Onto grid (Quantifoil R1.2/1.3 or UltrAuFoil)

3. Blot

• Use filter paper (remove excess liquid, thin film remains)

4. Plunge into liquid ethane

• Vitrify (freeze at -180°C, forms amorphous ice)

Screen grids:

• Check particle distribution (TEM)

• Optimize concentration, blotting time

• Goal: Uniform particle distribution, thin ice, no aggregation

Practical Checklist

Before Starting

• [ ] Confirm protein is in membrane fraction (Western blot)

• [ ] Decide: Detergent vs detergent-free (nanodiscs, SMALPs)

• [ ] Prepare detergent screening panel (DDM, LMNG, digitonin)

Purification

• [ ] Screen detergents (solubilization + stability)

• [ ] Affinity purification (His-tag)

• [ ] SEC (monodispersity check)

• [ ] Validate function (ligand binding, activity assay)

For Nanodiscs

• [ ] Express MSP, prepare lipids

• [ ] Optimize protein:MSP:lipid ratio

• [ ] Remove detergent (Bio-Beads)

• [ ] SEC to separate filled/empty nanodiscs

For Cryo-EM

• [ ] Concentrate to 1-5 mg/mL

• [ ] SEC-MALS (confirm monodispersity >90%)

• [ ] Prepare grids (glow-discharge, apply, blot, vitrify)

• [ ] Screen grids (optimize concentration/blotting)

The Bottom Line

Detergent-free purification is the future of membrane protein structural biology.

The advantages:

• Native-like environment (nanodiscs, SMALPs)

• Better stability (amphipols, months at 4°C)

• Better Cryo-EM (uniform particles, less background)

The methods:

• Nanodiscs: Best for functional studies, controlled lipid environment

• SMALPs: Fastest, preserves native lipids, no detergent

• Amphipols: Most stable, excellent for Cryo-EM

The workflow:

• Express → Solubilize (detergent or SMA) → Purify → Reconstitute (nanodiscs) or Exchange (amphipols) → Validate → Cryo-EM

Modern tools accelerate success:

• AI-driven construct design (Orbion: thermostabilizing mutations, boundary optimization)

• Automated screening (96-well detergent screens)

• Cryo-EM (no crystallization needed)

The difference between a failed membrane protein project and a high-resolution structure is often construct engineering + right solubilization method.

Ready to Optimize Your Membrane Protein?

If you're struggling with membrane protein stability, expression, or construct design, Orbion can help.

Orbion provides:

• Thermostabilizing mutation predictions (increase Tm, reduce aggregation)

• Construct boundary design (truncate disordered regions, identify stable domains)

• Membrane topology prediction (TM helices, loop regions, signal peptides)

• Expression system recommendations (E. coli vs insect vs mammalian)

• PTM prediction (glycosylation sites requiring specific expression systems)