Predicting Western Blot Band Sizes

Accurate protein band size prediction starts with the primary sequence, but real gels reflect biology. PTMs, processing, isoforms, and experimental conditions shift apparent molecular weight. Use the estimate, then verify with the diagnostic steps below.

Introduction

Western blotting separates proteins by size via gel electrophoresis (SDS-PAGE with sodium dodecyl sulfate) and detects targets on a membrane using antibodies. Predicted band size comes from sequence length, yet observed bands often differ due to biological and workflow variables. Solid sample preparation (e.g., protein extraction in the right Lysis Buffer with protease inhibitors) and correct antibody concentration improve band appearance and interpretability. Include a positive control to validate protein expression and confirm antibody detection. Understanding why bands shift—and how to test each cause—improves interpretation and reproducibility.

How to Estimate Protein Size Before Running a Western Blot

When predicting a protein’s molecular weight for Western blotting analysis, the simplest approach uses the average molecular weight (MW) of an amino acid—approximately 110 Daltons (Da). This value is derived from the mean mass of the 20 standard amino acids, adjusted for the removal of water during peptide bond formation.

Predicted MW (Da) ≈ N × 110
Convert the result to kilodaltons (kDa) by dividing by 1,000.

Example:
A protein with 170 amino acids → 170 × 110 = 18,700 Da, or ≈18.7 kDa.

This theoretical value serves as a baseline when interpreting Western blot bands. However, it assumes the protein is in its unmodified, fully denatured form—conditions that are often altered by biological processes or sample processing

Factors Affecting the Accuracy

1. Amino Acid Composition

   The 110 Da average works well for estimation, but proteins rich in aromatic or sulfur-containing residues (tryptophan, cysteine, methionine) may weigh more, while those enriched in small residues (glycine, alanine) may be lighter. These small shifts (±5–10%) can slightly change expected migration, especially for smaller proteins.

2. Signal Peptides and Pro-Peptides

   Many secreted or membrane proteins contain signal sequences that guide them through the endoplasmic reticulum and are later cleaved.

   • The predicted MW based on the full sequence (from databases like UniProt) includes these regions.
   • The mature protein observed on the blot is therefore often smaller than the theoretical value after protein transfer to the membrane.
3. Post-Translational Processing

   After translation, proteolytic cleavage or covalent modification (e.g., phosphorylation, glycosylation) can alter apparent size. These events are discussed in detail in the next section but are important to consider early when calculating expected MWs.

4. Transmembrane or Hydrophobic Domains

   Membrane proteins can migrate anomalously in SDS-PAGE due to detergent binding; consider stain-free gels or alternative systems.

   • Hydrophobic segments do not bind SDS uniformly, causing underestimation of true size.
   • Even when denatured, these proteins may aggregate or fold partially, resulting in bands running slower or faster than predicted.
   • Running a Tris-Tricine or urea–SDS gel can help improve accuracy for such proteins.
5. Isoforms and Splice Variants

   Different isoforms can appear as closely spaced bands; verify with Molecular weight marker and appropriate protein marker ranges. The difference may be small (1–2 kDa) or large (>10 kDa), depending on the alternative exon inclusion.

In general, the smaller the protein, the faster it migrates through the gel. However, migration is also affected by a few other factors. As a result, the actual band size observed in your results may differ from what was predicted by math. These other factors include, but are not limited to:

Why Observed Band Size Can Differ from Predicted

Small biochemical changes and run conditions can shift SDS-PAGE mobility, so the “predicted” kDa is a guide—not a guarantee.

1) Post-Translational Modifications (PTMs)

PTMs add mass and can hinder uniform SDS binding, slowing migration. Glycans and phosphate groups commonly create upward shifts or band “smearing/laddering.” Ubiquitin chains add ~8.6 kDa per moiety; SUMO adds ~12 kDa, often producing discrete higher bands.

How to test

• Run ± λ-phosphatase or PNGase F/O-glycosidase to remove phosphate/sugar groups.
• Expect a collapse toward the predicted kDa after treatment; include heat-inactivated enzyme controls

If available, probe with PTM-specific antibodies (e.g., anti-phospho). Use Coomassie staining or Ponceau staining to assess protein load and transfer efficiency.

2) Proteolytic Processing / Pro-forms

Zymogens or precursors are cleaved to active fragments, yielding smaller bands or multiple species (full-length + processed). Processing may be constitutive or stimulus-dependent.

How to test

• Add a broad protease inhibitor cocktail during lysis and sample prep.
• Compare ± a stimulus known to trigger processing; look for appearance/intensification of lower bands.
• Use domain-specific antibodies (N- vs C-terminal) to map the fragment detected and ensure correct antibody incubation timing.

3) Alternative Splicing / Isoforms

Distinct transcripts encode proteins with different lengths; a single antibody can detect multiple isoforms, appearing as closely spaced bands.

How to test

• Check UniProt/NCBI for annotated isoforms and expected sizes.
• Validate with isoform-specific RT-PCR or antibodies targeting unique regions.
• If possible, express a tagged reference isoform to confirm migration.

4) Multimerization / Disulfide-Linked Species

Incomplete reduction or strong noncovalent interactions can preserve dimers/trimers that migrate at higher apparent molecular weight.

How to test

• Ensure strong reducing/denaturing conditions (fresh DTT/β-ME; adequate heating).
• Alkylate cysteines post-reduction (e.g., IAA) to prevent re-oxidation.
• Run a non-reducing lane in parallel to confirm disulfide dependence. Verify using a non-reducing lane and review background problems in the Western Blot Troubleshooting Guide.

5) Charge and Composition Effects

Extreme pI, proline-rich or highly hydrophobic regions, and low-complexity domains alter SDS binding and shape, shifting mobility independent of true mass.

How to test

• Try an alternative gel/buffer system (e.g., Tris-Tricine for <30 kDa targets; Bis-Tris gels).
• Compare against multiple molecular weight ladders covering your range.
• Assess a truncated/Tagged construct to see if mobility normalizes.

6) Experimental Variables

Gel percentage, sample load, buffer freshness/pH, transfer settings, and membrane type affect band position and sharpness; artifacts can mimic true size shifts.

How to test

• Match gel % to target size (e.g., 8–10% for 50–150 kDa; 12–15% for 10–50 kDa).
• Use fresh running/transfer buffers; verify pH and ionic strength.
• Confirm even transfer (no blow-through/under-transfer): check gel post-transfer and use Ponceau S on the membrane.
• Manage signal strength with proper Chemiluminescent substrate and washes (e.g., Tween 20 in TBST). Keep exposure within the linear range to avoid “bloomed” bands.

Additional best practices:

• Confirm membrane choice (PVDF membrane/polyvinylidene fluoride vs nitrocellulose membrane) based on target and detection chemistry.
• Remove air bubbles during stack setup to avoid patchy western blotting transfer.
• For long wet transfer under wet conditions, control temperature and transfer time.
• If using enzyme-conjugated detection, avoid Sodium azide with HRP systems.

Different types of post-translational modifications

Aquaporin-1’s expected band is 28kDa. However, after AQP1
glycosylation, another band is observed around 50kDa.

Quick Reference Table: Common Causes of Band-Size Shifts

Workflow pointers: Validate transfer efficiency with Ponceau staining; confirm protein load; use appropriate protein marker/Molecular weight marker.

Example Calculation and Interpretation

Protein: 412-aa receptor with a 20-aa signal peptide and an N-glycosylated ectodomain.
Predicted MW: 412 × 110 ≈ 45.3 kDa for the nascent chain; after signal peptide removal (−20 aa → −2.2 kDa), mature core ≈ 43.1 kDa.
Observed: A ~55 kDa major band in standard lysates; after PNGase F, a ~42–43 kDa band.

Interpretation:

• The ~12–13 kDa upward shift relative to the mature core is consistent with N-glycosylation (e.g., multiple complex N-glycans contributing several kDa each).
• PNGase F collapses the broad/higher band to the predicted deglycosylated core (~43 kDa), confirming that the mass increase is glycans rather than dimerization or another PTM.
• A slightly broad or fuzzy appearance of the ~55 kDa band is typical of heterogeneous glycoforms.

Recommended confirmations:

• Endo H vs PNGase F: Distinguish high-mannose/ER-type (Endo H sensitive) from complex Golgi glycans (Endo H resistant, PNGase F sensitive).
• Tunicamycin control: Block N-glycosylation to verify loss of the ~55 kDa band and enrichment of the ~43 kDa species.
• Reducing conditions: Ensure fresh DTT/β-ME and proper heating to rule out disulfide-linked dimers.
• Domain antibody mapping: N- vs C-terminal antibodies should detect the same deglycosylated core if no proteolysis occurred.
• Ladder and gel choice: Use a ladder bracketing 40–60 kDa and, if needed, a Bis-Tris or Tris-glycine system to confirm migration consistency across gels.

References

• Kurien BT, Scofield RH. “Western blotting: an introduction.” Methods in Molecular Biology. 2015;1312:17–30. Indexed on PubMed and available via PMC; it’s a chapter in the Western Blotting: Methods and Protocols volume. PubMed+2PMC+2
• Mahmood T, Yang P-C. “Western blot: technique, theory, and troubleshooting.” North American Journal of Medical Sciences. 2012;4(9):429–434. Indexed on PubMed and available via PMC (DOI: 10.4103/1947-2714.100998). PubMed+1