Radiolabelling proteins overexpressed in E.coli

Peter Gegenheimer peterg at
Fri Nov 25 02:10:11 EST 1994

In <3aje6o$lmv at>, szsclark at ([user unknown]) writes:
>Dear Netters,
>I am overexpressing a smallish protein (ca. 20 kD) in e. coli using the
>Novagene pEt vectors in BL21/DE3 (pLysS).  My protein expresses quite well
>and i get yields of several hundred micrograms from my preps.  My problem
>is the specific activity of the protein.  I am labelling the protein with
>3H-leucine.  For various reasons I can't use other radiolabels and 3H is
>best for my purposes, but theoretical calculations of the potential
>specific activity achievable if all the leucines in my protein are 3H'd
>are about 1000x what I actually get.  The specific activities I am
>achieving are within the range reported by other people who have expressed
>similar proteins so I don't think there is anything 'wrong' with my
>protocol, which is pretty standard.  I am wondering if there is some
>mechanism which prevents the 3H-leu being incorporated into every
>position, eg. competition with an intracellular pool of leucine that is
>synthesized by the coli?  Is 3H -leucine not taken up by the coli as
>efficiently as non-3H amino acids?  If I increase the amount of 3H-leucine
>I add to the prep, the specific activities appear to increase
>proportionally, suggesting the available 3H-leucine is non-saturating.  Or
>is there something I can do to improve my protocol and my specific
>activities?  I haven't been able to find any suggestions in the literature.  
>Your suggestions are welcome. 
>My abbreviated protocol is;
>Grow fresh cells in M9 + Bactotryptone + Carbenicillin (100 ug/ml) until 
>OD600 ca. 0.3
>Wash in M9, resuspend to 1/3 rd vol. in M9+Carb+Chloramphenicol. Store 
>o/n at 4oC. (I titre here to check plasmid stability)
>Inoc. 300 ml fresh M9+carb+chlor with 6 ml of starter culture. Grow to 
>OD600 =0.2.
>Wash cells 2x in M9. Resusp. in 1/10th vol of M9+C+C. Add IPTG to 0.4 mM. 
>Grow at 37oC for 40 min.
>Add 10 ml of culture to flask containing 1.5 mCi of dried 3H-leucine. 
>Grow for 4 to 5 hours.
>spin down cells, resuspend in 10ml of 20mM Tris 7.5, 20% sucrose, 1mM EDTA. 
>Incubate 10 min on ice.  
>Spin down cells, resuspend in 5ml of ice-cold H2O with 1mM PMSF, 20 ug/ml 
>aprotinin, 1 ug/ml leupeptin. Incubate 10 min on ice. Add equal vol 2 
>PBS. Freeze overnight.
>Lyse by probe sonication. Wash inclusion bodies 3x with 1xPBS, 5 mM EDTA, 
>25% Sucrose, 1% triton X100. wash 3x with cold H2O.
>I usually get between 200 and 300 ug of protein of between 600, 000 and 1 
>x 10e6 dpm/ug.
>thanks for reading all this!  any suggestions welcomed.
>Sonya Clark

I have some suggestions, although it's too late for me to read your protocol 
thoroughly. My experience with metabolic labeling comes from my and other's
with 3H-uridine labeling of RNA many years ago, but I think the principles are 
general. Our lab also has extensive current experience with the pET system.

You are correct in assuming that there is a pool of endogeneous leucine which dilutes 
the label you add. Since the bacteria continually synthesize Leu, the endogeneous 
pool probably does not decrease much during the labeling period.  You should also be 
aware that high INTRAcellular concentrations of an anabolite (Leu, Ura, etc.)  often 
inhibits the uptake of that metabolite, or represses synthesis of the appropriate 

One classical way to overcome this problem is to do the labeling in a
strain auxotrophic for the anabolite in question (i.e., a Leu- strain). Some 
modifications to your protocol (necessary for labeling a Leu- cell) might also
increase specific activity. 

(I assume that CARB is the selective agent for the pET vector, and
CAM is selective for the pLys[E or S] plasmid; if not, I'm not following your 
protocol. If your host has pLysE/S, then you need only freeze-thaw 2-3X to lyse; 
sonicate to shear the DNA.)

Grow the cells in M9+glucose+Leu (or glucose+casamino acids); just
before labeling, wash and starve the cells -- at the same concentration -- in M9 +
glucose (+ antibiotics) minus Leu. (The duration of starvation could be previously
determined by monitoring cell density, and by withdrawing aliquots of cells and 
labeling them to determine total [3H]Leu incorporation into TCA-insoluble material.) 
After starvation. label as you are doing. I can't recommend whether to induce the T7 
RNA polymerase before or after adding cell to label -- that would depend on the 
kinetics of synthesis of your protein. Remember that the host cells stop growing very
soon after induction. I would probably add the [3H]Leu before induction, to get the 
most label taken up into the cells beforehand.

If you are unable to construct a Leu- host, you should be able to use some of these 
suggestions with the "normal" host cells. You'd need to know the kinetics of your 
protein's accumulation: after what period of time does accumulation appear to stop? 
Also, after what period of time does [3H]Leu incorporation into your protein level
off? If protein synthesis stops before Leu incorporation, you may well be able to 
slow down protein synthesis by reducing the IPTG concentration to 25 - 40 micromolar.
(This was shown by Fletterick's lab to reduce the rate of synthesis of
phosphorylase, and give mostly soluble enzyme rather than inclusion bodies.)

If I have missed something, and others don't give better advice, please send me a 

|  Peter Gegenheimer            |  pgegen at             |
|  Departments of Biochemistry  |  voice: 913-864-3939                   |
|    and of Botany              |                                        |
|  University of Kansas         |  FAX  : 913-864-5321                   |
|  2045 Haworth Hall            | "The sleep of reason produces          |
|  Lawrence  KS  66045-2106     |  monsters."              Goya          |

More information about the Proteins mailing list