Re: [ccp4bb] Non-specific disulfides in a secreted protein

[Luca Jovine]

Dear Tomas,

Along Radu’s comments, here’s another paper well worth looking into:

   Halff EF, Versteeg M, Brondijk TH, Huizinga EG
   When less becomes more: optimization of protein expression in HEK293-EBNA1 
cells using plasmid titration - a case study for NLRs
   https://www.ncbi.nlm.nih.gov/pubmed/24680733

We have also observed this phenomenon several times, and simply diluting the 
expression construct did indeed help in such cases.

Good luck!

Luca


Luca Jovine, Ph.D.
Professor of Structural Biology & EMBO Member
Karolinska Institutet
Department of Biosciences and Nutrition & Center for Innovative Medicine
Medicinaren 25 Neo
Blickagången 16, SE-141 83 Huddinge, Sweden
E-mail: luca.jov...@ki.se
W3: http://jovinelab.org


On 14 Dec 2018, at 17:55, Tomas Malinauskas 
mailto:tomas.malinaus...@gmail.com>> wrote:

Dear All,

we are purifying a small secreted protein from conditioned media and
have a rather unusual problem.

It is a small ectodomain (~11 kDa, pI ~6, His-tagged) of the type 1
transmembrane receptor, crystal structures are known (of the protein
that was produced in E.coli and refolded; we are secreting the same
protein using mammalian cells) so we can design reasonable constructs.
The protein is expressed and secreted by transiently transfected
HEK293T cells that work very well for other ectodomains and
extracellular proteins in our hands (PMID 17001101). The target
protein has 10 cysteines that form 5 disulfides in the crystal
structure (of E.coli-expressed and refolded protein), there should be
no free cysteines and no non-specific disulfides. Unfortunately, once
the protein is secreted, it forms non-specific dimers and higher-order
oligomers in the media (standard DMEM/2% FBS) before purification
(confirmed by Western blotting under non-reducing conditions). Using
0.5 mM DTT during SEC gives a nice monomeric peak (however, the
protein suffers as suggested by weaker interactions with its binding
partners). We don't understand how a secreted protein (which passes
trafficking quality control in the cell) with a known disulfide
pattern forms non-specific disulfide linked oligomers in the
extracellular media. We tried expressing it at 37 C and 30 C, and have
sequenced our constructs (plasmids) multiple times.

If anyone has seen this kind of problem and successfully solved it
(purified homogeneous crystallisation quality protein), please let us
know if possible. I thank you for your help.

Best wishes,
Tomas


Dr. Tomas Malinauskas
University of Oxford
Wellcome Centre for Human Genetics
Division of Structural Biology
Roosevelt Drive
Oxford OX3 7BN
United Kingdom
to...@strubi.ox.ac.uk
tomas.malinaus...@gmail.com




När du skickar e-post till Karolinska Institutet (KI) innebär detta att KI 
kommer att behandla dina personuppgifter. Här finns information om hur KI 
behandlar personuppgifter.


Sending email to Karolinska Institutet (KI) will result in KI processing your 
personal data. You can read more about KI’s processing of personal data 
here.



To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1

[ccp4bb] AutoBuild Invalid MTZ column_types

[Robson-Tull, Jacob]

Dear all,


Has anyone encountered the error "Invalid MTZ column_types for the given 
miller_array" when trying to run AutoBuild?


The last line of the log was "Adding array ['F', 'SIGF', 'DANO', 'SIGDANO'] to 
output file with type FQDQ". The MTZ files were built with the Xia2 3dii 
pipeline.


I had a look at previous messages on the bulletin board but there were no 
answers posted.


Any advice?


Best,

Jake



To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1

Re: [ccp4bb] Non-specific disulfides in a secreted protein

[Tomas Malinauskas]

Dear Zhijie et al,
thanks a lot for your thoughtful suggestions.
a) we do get a fraction with a "proper" monomer but it would be nice
to minimise losses because of non-specific disulfides;
b) yes, it has 1 Asn-linked glycan which gets nicely trimmed upon
treatment with Endo F1;
d) yes, we do get some monomer on SEC without DTT. We thought ~0.5 mM
DTT (which degraded relatively fast anyway) might be OK to reduce some
non-specific disulfides but not sufficiently high to destroy the whole
folded domain, we used it out of desperation/curiosity;
e) we will try longer tags and adding more residues at termini.
Best wishes,
Tomas



On Fri, Dec 14, 2018 at 6:57 PM Zhijie Li  wrote:
>
> Hi Tomas,
> Some thoughts:
> a) I guess the thermodynamic drive for all part of this small ectodomain to 
> fold into a single lowest energy conformation is not very strong. The cells 
> can’t know that the little artificial domain is supposed to be a monomer when 
> the oligomers are also sufficiently hydrophilic.  We also occasionally see 
> aggregates coming out of human cell lines. Sometimes barely anything good.
>
> (One wild thought: the disulfide shuffling system in eukaryotic cells have 
> something to do with the N-glycans. Does the natural form of this protein 
> have N-glycans, while the ecotodomain is somehow deprived of that?)
>
> b) Is there absolutely no monomer on the Western? If this is the case, since 
> the bacterial preps can be refolded, can you try incubating the protein with 
> Cysteine or GSH or even add PDI or DsbC to try to increase the population of 
> properly folded species?
>
> c) If you have some monomers or can manage to generate some by disulfide 
> shuffling: for a small domain with such high disulfide content a further 
> reverse-phase C18 HPLC step may give you the properly folded species without 
> irreversibly denaturing them. Or maybe a Superdex75 gel filtration or a 
> silica SEC on HPLC, without reducing reagents, see below.
>
> d) Have you ever run the gel filtration without DTT? When putting the DTT, 
> because of its strong disulfide cleaving power, you might be sending the 
> misfoled junk to the monomer peak. They look like monomers when DTT is 
> present but they are misfolded still. If you have a small fraction of 
> monomers that had passed the cell’s QC, they are more likely to be correctly 
> folded therefore active. Then the right thing to do I think is to try to 
> separate the monomers from the oligomers, rather than forcing everything into 
> “monomers”.
>
> e) Try a larger tag? Add more natural sequence back to the domain? Try 
> periplasmic expression in E coli? The NEB pMal-p vector is my favourite for 
> disulfide forming proteins, when not using mammalian cells.
>
> Zhijie
>
> > On Dec 14, 2018, at 11:57 AM, Tomas Malinauskas 
> >  wrote:
> >
> > Dear All,
> >
> > we are purifying a small secreted protein from conditioned media and
> > have a rather unusual problem.
> >
> > It is a small ectodomain (~11 kDa, pI ~6, His-tagged) of the type 1
> > transmembrane receptor, crystal structures are known (of the protein
> > that was produced in E.coli and refolded; we are secreting the same
> > protein using mammalian cells) so we can design reasonable constructs.
> > The protein is expressed and secreted by transiently transfected
> > HEK293T cells that work very well for other ectodomains and
> > extracellular proteins in our hands (PMID 17001101). The target
> > protein has 10 cysteines that form 5 disulfides in the crystal
> > structure (of E.coli-expressed and refolded protein), there should be
> > no free cysteines and no non-specific disulfides. Unfortunately, once
> > the protein is secreted, it forms non-specific dimers and higher-order
> > oligomers in the media (standard DMEM/2% FBS) before purification
> > (confirmed by Western blotting under non-reducing conditions). Using
> > 0.5 mM DTT during SEC gives a nice monomeric peak (however, the
> > protein suffers as suggested by weaker interactions with its binding
> > partners). We don't understand how a secreted protein (which passes
> > trafficking quality control in the cell) with a known disulfide
> > pattern forms non-specific disulfide linked oligomers in the
> > extracellular media. We tried expressing it at 37 C and 30 C, and have
> > sequenced our constructs (plasmids) multiple times.
> >
> > If anyone has seen this kind of problem and successfully solved it
> > (purified homogeneous crystallisation quality protein), please let us
> > know if possible. I thank you for your help.
> >
> > Best wishes,
> > Tomas
> >
> >
> > Dr. Tomas Malinauskas
> > University of Oxford
> > Wellcome Centre for Human Genetics
> > Division of Structural Biology
> > Roosevelt Drive
> > Oxford OX3 7BN
> > United Kingdom
> > to...@strubi.ox.ac.uk
> > tomas.malinaus...@gmail.com
> >
> > 
> >
> > To unsubscribe from the CCP4BB list, click the following link:
> > https:

Re: [ccp4bb] AutoBuild Invalid MTZ column_types

[Eleanor Dodson]

Looks as though the column types have been run together.. Xia3 writers?

It should read "F Q D Q" I believe..
Eleanor

On Sat, 15 Dec 2018 at 15:56, Robson-Tull, Jacob <
jacob.robson-tul...@imperial.ac.uk> wrote:

> Dear all,
>
>
> Has anyone encountered the error "Invalid MTZ column_types for the given
> miller_array" when trying to run AutoBuild?
>
>
> The last line of the log was "Adding array ['F', 'SIGF', 'DANO',
> 'SIGDANO'] to output file with type FQDQ". The MTZ files were built with
> the Xia2 3dii pipeline.
>
>
> I had a look at previous messages on the bulletin board but there were no
> answers posted.
>
>
> Any advice?
>
>
> Best,
>
> Jake
>
> --
>
> To unsubscribe from the CCP4BB list, click the following link:
> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1
>



To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1