Improvements in Yield of 15N labelled proteins

By: Julia Smith, Ben Davies, Vernalis (R&D) Ltd, Granta Park, Cambridge, CB21 6GB

Product Nos. B21200, B21201


Nuclear Magnetic Resonance (NMR) spectroscopy is an important technique for structure–based drug design. NMR can both characterise protein molecules and be used as a screening tool for the binding of ligands to protein targets, with the key advantage of being able to detect and quantify interactions with high sensitivity. NMR can provide structural information on both the target and the ligand to aid subsequent optimisation of weak binding hits into high affinity leads. A common approach to help characterise protein-ligand or protein-protein interactions is to label the protein with stable isotopes such as 15N. Subsequent binding of a ligand alters the chemical environment around the binding site and so will perturb the chemical shift of magnetic nuclei at this site. These changes can be observed using 15N/1H correlation spectra. Isotope labelling can result in increased sensitivity and resolution and in reduced complexity of the NMR spectra.

However, this technique requires large quantities of high purity 15N labelled protein for analysis. Traditional methods for labelling protein have involved using a M9 minimal media and subsequent addition of several reagents making the set up laborious. Furthermore, the yield of labelled protein is generally low. The use of a stable isotope for not much return makes this an expensive process.

BioSilta have developed EnPresso® B Defined Nitrogen-free (EPBDNF) media which is a pre-sterilised, chemically-defined growth system designed to increase the yield of 15N-labelled proteins. Here a comparison study has been conducted between EBDNF and traditional M9 media to evaluate the yield and efficiency of 15N labelling a protein of interest.


Protocol for 15N Labelling using M9 media (Vernalis)
1L culture

  • An inoculum from a glycerol stock or fresh transformation was prepared
  • 5x M9 salts without N2 source were autoclaved
  • To each 750ml sterile dH20 in 2.5L glass baffled flask, the following constituents were added;
    • 200ml of M9 salts
    • 20ml of 20% glucose solution
    • 2ml 1M MgSO4
    • 100ul 1M CaCl2
    • 1ml 15NH4SO4 (1g/ml stock)
    • Trace vitamins and metals
    • Antibiotic
    • 5mls inoculum washed and resuspended in PBS
  • Cells were shaken at 37 °C, 200rpm until O.D. 600nm = 0.4. The temperature was reduced to 18 °C and at O.D. 600nm = 0.6, the cells were induced with 1mM IPTG [final] and incubation continued overnight. The cells were harvested the next morning.

Protocol for 15N Labelling using Biosilta EnPresso B Defined Nitrogen‐free media
100ml culture
(Instructions provided by BioSilta)

  • Prepare an inoculum from a glycerol stock or use a single colony grown overnight on an agar plate. Inoculate 1 mL of LB medium containing antibiotics.
  • Incubate with vigorous shaking at 37 °C for 6 h or until the inoculum has reached OD 2 - OD 4.
  • Add aseptically into a sterile 1000 mL shake flask
    • 90 mL sterile water
    • Four tablets (two bags) of EnPresso B Defined Nitrogen‐free
    • 6 mL 15NH4Cl (40 g/L, final concentration 2.5 g/L)
    Note: 2.5 g/L final concentration of 15NH4Cl is recommended as a starting concentration. Concentrations as low as 1.0 g/L and as high as 4.0 g/L have also been used successfully.
  • Immediately shake vigorously at 37 °C until the tablets have fully dissolved. After dissolution, small crystals of magnesium salts may occasionally be visible, but these will not affect performance.
  • Add required antibiotics. If using Ultra Yield Flasks, add antifoaming agent, such as 100 μL/L AntiFoam 204.
  • Inoculate with 1:100 of the pre-­culture inoculum (1 mL).
  • Add 50 μL Reagent A (final concentration 1.5 U/L)
  • Close the flask securely.
  • Incubate overnight at 30 °C, 250 rpm.
  • After 16 – 20 h, add
    • induction agent such as  0.2 mM IPTG
    • 170 μL Reagent A (final concentration 5 U/L)
  • Continue incubation at 30 °C, 250 rpm
  • Sample after 6 to 8 hours induction to evaluate protein production.
  • Continue incubation overnight. Harvest next morning.

Recommended conditions
Culture volume
: 10% of flask volume (20% if using Ultra Yield flasks)

Trial 1 – 1L EPBDNF/flask

The BioSilta Protocol was used but modified for the volume of culture. 1L of EPBDNF media in a 2.5L glass baffled flask with a foam bung was used, shaken at 200rpm. This is 40% of the flask volume and above the recommended culture volume conditions. A total of 5L was grown with a final concentration of 2.5g/L 15(NH4)2SO4 was used. The cell strain was BL21 (DE3) pLysS.


Lysis of the cells and subsequent 2‐step purification yielded 162mgs of protein. This is a 13 fold improvement over M9 media, see Figure 1.

Yield of protein of interest comparing EBDNF and M9 media

Figure 1. Yield of protein of interest comparing EBDNF and M9 media.

Trial 2 – Recommended Culture Volumes

Although the yield is a vast improvement it was thought a further increase in yield could be produced if the recommended culture conditions were followed.

  • The BioSilta Protocol was used but modified for the volume of culture. An experiment was performed to test three different culture volumes/flask combinations:
  • 10% culture volume in 2.5L standard glass baffled flask with foam bung
  • 10% culture volume in 2.5L standard glass baffled flask with Air-O-Top seal
  • 20% culture volume in 2.5L Ultra Yield flask with Air-O-Top seal.

Flasks were shaken at 200rpm. A final concentration of 2.5g/L 15(NH4)2SO4 was used.


Following lysis of the cells and subsequent 2-step purification it was clear yields of protein were greatly improved. The highest yield of 160mg/L was observed using 10% culture volume in a baffled flask with a foam bung, see Figure 2.

Yield of protein of interest comparing culture volumes and flasks

Figure 2. Yield of protein of interest comparing culture volumes and flasks.


NMR validation

2D 15N-­1H HSQC NMR spectra were acquired on 15N labelled protein expressed in both M9 and EBDNF media. No significant differences were observed in the NMR spectra, with a high level of 15N in both samples.


  • Use of EPBDNF dramatically increases the yield of 15N labelled protein from 2.5mg/L in M9 media to 160mg/L.
  • EPBDNF provides a high level of 15N incorporation allowing NMR analysis of the protein of interest.
  • Use of EPBDNF reduces set up time and reduces cost but improves yield of 15N labelled proteins.




  1. Pellecchia M et al. (2002). NMR in Drug Discovery Nature Reviews Drug Discovery. 1 (3), 211-­219.
  2. Maniatis T, Fritsch E F, Sambrook J. (1982). Molecular Cloning, A Laboratory Manual 1st ed.


EnPresso is a registered trademark of BioSilta Oy.

Related Links