Protein Purification
The following are Science Advisory Board members' responses to the question "If money were NO object, please detail how your lab would obtain purified proteins for its research purposes (e.g., What purification protocol(s) would you use and why? Would you invest in a high-throughput protein purification platform? Why or why not? Would you purchase already purified proteins from a supplier or send your proteins to a core facility to be purified?)?" Members are identified only by their screen name, job position, and geographic location.

1) Protocol: I would use 'tagged' sequences in Bacillus secretion system. High yield, High initial purity, no generation of inclusion bodies (= accumulated malfolded proteins). 2) Currently, we do not want to invest in an automated system since we purify most of the time the same protein labeled with radioactivity at small scale for functional assays. 3) We can purchase recombinant forms of the protein, however, they lack posttranslational modifications like glycosylation and GPI anchoring. 4) We do not have a core facility to do this.
bossers, Principal Investigator, Europe

1. Due to the sensitivity of the protein we work with, we would not change the protocol, but we would buy more (fresher) reagents for every prep. We would also use a lot more resin to get more protein. 2. We would not invest a high-throughput purification platform since protein purification on the scale that we use it on is more of an art than a set science. Human reasoning is involved in every step of the way. 3. We are the only ones who purify this particular protein so we cannot buy it. Even if it was supplied, the protein (enzyme) activity lasts only a few weeks at -80 C and our experiments demand fresh protein. If there was someone in the core facility who could purify this protein and be good at it, then we would do that. Currently, we have a technician in the lab whose sole job is to harvest this protein from bacteria and purify for those who work with it.
eharri, Post Doctoral Fellow, North America

A new FPLC/HPLC rig set up for analytical work would be essential. For preparative work a high-throughput system would certainly save time and facilitate reproducibility. LC-MS on tap would be very useful for assessing purity and nature of proteins purified.
Advisory, Post Doctoral Fellow, Europe

Aptamer based affinity purification. Yes, since to monitor the dynamics of the proteome it is necessary to analyse many samples. Ideally from a supplier.
Kashman, Principal Investigator, North America

As a convenience it would be nice to have a supplier (GLP or GMP approved) to be able to purify the proteins that we want, when we want it. Of course this only works if money is no object.
dave77, Lab Director/Supervisor/Coordinator, North America

As can be seen from the survey, I am a relatively happy user of the GST system. My work is focused enough that high-throughput systems would be waste of money.
gsvanas, Department Head, North America

As we are sometimes struggling with toxic proteins and if money wouldn't be an issue, we would go for a core facility to develop new purification techniques.
proto, Lab Director/Supervisor/Coordinator, Europe

Automation requires additional manpower in the laboratory. This time could be better utilized if protein purification could be outsourced.
Kalyan, Lab Director/Supervisor/Coordinator, North America

Based on the final application of the purified protein the purification method would be important. For target validation in drug discovery we usually purchase already purified proteins from a supplier. For high-throughput screening in vitro assays sometimes a partial purification is sufficient and large amounts of active protein are needed and these are produced in-house. For crystal structure determination based upon the proprietary nature either a core facility or outsourced lab will provide purified protein.
mhari, Post Doctoral Fellow, North America

Because of the diversity in the behaviour of proteins, no single purification method is suitable to purify all proteins. In our lab we have chosen to use Gateway cloning to insert genes into multiple vectors in order to screen for the most suitable construct/purification method. A high-throughput protein purification platform, which would allow us to screen multiple buffers to find one which gives soluble, monodisperse, active protein would greatly increase efficiency in our NMR lab. If money AND time were NO object, we would definitely send proteins to a core facility for purification.
chong, Graduate Student/Research Assistant, North America

Because our purifications are from both native and cloned sources, we would use an affinity-based system combined with traditional resins for purifying the proteins of interest. Tagging of proteins and tag-based purifications have a number of problems. Many proteins don't express as well with the tag as without. The biggest problem with tags is that you still need a separate, completely different method to purify the native form of the protein. A high-throughput protein purification platform only makes sense if someone invents an expression system that produces protein from "all" genes from "all" sources. Many researchers I've talked to and my own experience points out that probably more than half of all genes don't express in E. coli at any level. Unless expression is assured, high-throughput would be a waste of time and reagent in an attempt to purify nonexpressed proteins. Purchasing from a supplier is something we would readily do. My experience with purification by core facility is that the core facility generally knows less about the purification of your protein than you do, and the core facility usually takes longer and is more expensive.
enzymedoctor, Production/Manufacturing, North America

Certainly, if money were no object we will invest in a high-throughput protein purification platform. We will not prefer to purchase purified protein from a supplier or send our protein to a core facilities to be purified because we will always have doubt in mind about homogeneity of protein and we have to reconfirm any way.
Hakim, Professor/Teacher, North America

Evaluation of our protein-chip (under development), we need within 2 years a number of different proteins, which are especially circulated in the body fluid. Allergen or some toxic proteins impure form to use standard in the food protein safety analyses.
nokihara, Lab Director/Supervisor/Coordinator, Asia

Eventually, if pharmaceutical applications arise for the protein we are currently investigating outsourcing production will be essential. However, at present for crystalographic purposes the delivery buffers have to change frequently, with careful monitoring of the enzymes integrity, specific activity, and stability. The current 6X His-tagged protocol in combination with stringent washes on Clontech's Talon resin is providing ~3-5 mg purified protein per litre of IPTG induced culture. Our primary stumbling block for crystalographic purposes has been the cost of the Enterokinase digestions. Fortunately, for our protein the tag is easily clipped (at about 1/20th the normal amount of enzyme, 3 days, 4¡C). At present without genetic engineering, trypsin is not a viable alternative. My personal viewpoint is that the hardest step is optimizing high-level expression of soluble, functional proteins. The choice of a tag and subsequent removal then depends on two factors: (1) the composition of the protein, and (2) the ease and/or cost of removing the tag. Now as to your other questions: 1) I have already explained our potential need for larger quantities of protein through outsourcing production. 2) 6X His tagged protein, stringent washes on the Talon resin, in combination with Sigma's His-tag protease inhibitor cocktail, has provided a one step procedure for high purity protein. The simplicity of the procedure, and need for flexibility in the purification buffers, cleavage steps, and delivery buffers has not warranted an automated system. 3) I would purchase commercial proteins if it was something we wanted to try for a particular experiment, and we have used the HPLC core facility here at Brandeis to purify cytoskeletal proteins like centrin that have a high affinity for other proteins and fail to come of tagged affinity columns cleanly. It is just at present we have been very fortunate. If I can clarify any of these responses or expand on anything please let me know.
remillard, Post Doctoral Fellow, North America

First of all I would always purify my protein myself (its membrane proteins I am working with, and I wouldn't trust any company to do it right). Limiting to the purification in my case(s) is a lack of large-scale purification facilities like centrifuges for large volumes and large (but expensive) chromatograpgy columns. As I am not routinely preparing the same proteins every week but test new purification methods all the time, automation is of no real interest to me.
dirk30, Post Doctoral Fellow, Europe

First, I would not invest in high-throughput. The milligram quantities of protein are sufficient for downstream analyses. Second, I would use affinity prior to prep denaturing gel permeation HPLC. Third, I have not had good luck with purchasing purified proteins except for antibodies. They are not usually pure.
ansci, Principal Investigator, North America

For part of the bio-pharmaceutical/biotech companies, it is obvious that we require to purify proteins in a way that will induce very little or no stress to the molecule, with a process that can easily be scaled-up and validated for pharmaceutical production without major changes.
nony, Lab Director/Supervisor/Coordinator, Europe

High-throughput purification in large scale would be highly convenient.
Stark, Professor/Teacher, Europe

His-tagged system including conditional endogenous lysis-system for efficient lysis of cells to allow faster purification of intracellulary expressed recombinant proteins. High-throughput purification allowing simultaneous purification of at least 48 samples with purity more than 80% and yield above 5 mg/sample. In-house technology is a must.
SteD, Lab Director/Supervisor/Coordinator, Europe

I believe that we would send out our proteins to be purified.
imargiti, Laboratory Technician, North America

I believe we would prefer to use a core facility and have the proteins of interest prepared for us by them. Alternatively, we could establish an in-house core facility which would be very well equipped with automated systems for protein purification.
K.Kostelidou, Post Doctoral Fellow, Europe

I don't have a need for a high-throughput protein purification system. I would like a system that is easily scaleable to larger size preps, particularly if the protein has some commercial value.
Rocketdog, Quality Assurance/Quality Control, North America

I recommend Akta system from Amersham Biosciences as the best method for getting very good yield of proteins I use.
m__shyamal, Principal Investigator, Asia

I still remember the 'good old' days of driving to the slaughterhouse to obtain the raw material for protein purification and spending days in the cold room. The modern expression systems and purification tags have really improved our lives.
samoht, Professor/Teacher, North America

I use the same protocol (with appropriate modifications) I developed at NIH [published in 1991] because even with all the technological advances it works well, uses a few components, and it is still fast [the process] with high purity and activity. I would not invest in a high-throughput protein purification platform because I do not need to perform protein purifications from many patients' samples. Also, I am not a commercial entity that purifies proteins for organizations (that might be of use for those companies). I would purchase already purified proteins from a supplier if I wanted to develop polyclonal antibodies; otherwise, can do in-house.
scientist, Lab Director/Supervisor/Coordinator, North America

I would advocate use of transgenic plants. Post-translational modifications in plants are more faithful to those in mammalian cells than in any other system. Transgenic plants do not require tissue culture and can achieve an extremely high biomass of expressed protein. Some headway has even been made in robotic handling of transgenic plants; the only drawback for a high-throughput system would be that plants take a relatively long time to mature (several weeks or months), giving a large backlog of samples. However, they require little maintenance during this time, only illumination!
Astrobiologist54, Post Doctoral Fellow, Europe

I would choose to send out to a reputable company for purification. This would time, effort and money on my end. The most important factor would be my trust in the company to follow the correct procedures.
sheilap, Lab Director/Supervisor/Coordinator, North America

I would outsource to a company. I would have only to send a cDNA and they would sub-clone to the appropriate vector, express and purify my protein of interest.
mburow, Professor/Teacher, North America

I would prefer not to spend my time on learning all the technical tricks to have my pure proteins, but I would prefer to send the clone to the core lab and consult with them my needs. I assume that overall I would save some time and money that way.
krzysztof, Administrator, Europe

I would prefer to have to different small tags attached to my protein of interest, which should allow for greater purity and can be cleaved off together. If protein expression would be more efficient in eukaryotic cell lines like COS7 cells it would be helpful to have a cell line which can grow up to high density and the protein should be purified with two-tag strategy. I would purchase already purified proteins from a supplier only if I would need a small amount for a few experiments.
holzmax, Staff Scientist, Europe

I would probably not invest in a high-throughput system since we are studying a small number of proteins (receptors). If money were no object I would probably use expression in a mammalian cell (native source of protein, for proper folding, etc.) using a highly controllable promoter system with no tag. I would then purify the protein of interest using custom polyclonal antibodies (which I produce myself) to various domains. I could then verify that I have the functional receptor protein of interest by performing ligand blotting assays with native ligands.
mlbrown, Principal Investigator, North America

I would purchase protein from a supplier. It's a common protein, and the protein is just one of the tools I need to do my project.
walterse, Professor/Teacher, North America

I would purchase the Aktadesign system from Amersham Biosciences. This platform allows you to go from basic research to process development in one compact unit.
SteveL, Staff Scientist, Europe

I would use a tagged mammalian expression system for solubility, appropriate modification(s), and activity reasons. I have no need currently for a high-throughput system since the number of proteins I produce is limited. I would purchase proteins from a supplier or farm them out to a core facility if money were no object.
wtaylor, Lab Director/Supervisor/Coordinator, North America

I would use GST and antibody affinity systems and most likely would outsource the complete production of recombinant proteins.
Burkhard, Principal Investigator, Europe

I wouldn't invest in a high-throughput platform because our needs for purified proteins have varied quite widely over the years. Sometimes the whole lab will be purifying proteins every day. A year later, the projects have moved on and nobody cracks bugs and runs a column for several months. The large, expensive machine would end up gathering dust and occupying a bench that could be otherwise occupied. If money were no object, my ideal solution would be to have a dedicated technician to produce the proteins. A decent tech will always have the flexibility to contribute to the lab even if nobody wants protein expression for a while.
Digard, Principal Investigator, Europe

Ideally, we would probably have a protein purification core facility that would be able to handle all the group's protein needs.
jlk, Associate Scientist, North America

If I have sufficient funding, I will buy the protein of interest for a supplier. Protein is rather complex regarding its physical and chemical properties; one needs to have comprehensive knowledge and sound expertise to be able to purify a protein of unknown property. It is laborious and time-consuming even with all the tag-based recombinant protein purification protocols available. On the other hand, I would like to have a commercial supplier to clone all the human proteins into a high expression vector so that investigators can buy any protein they come across. There are certainly a huge market in the commercial protein. I can discuss the detailed issues with someone who are interested in commercialization of proteins in human body.
Mingang, Post Doctoral Fellow, North America

If it were possible, I would like to have an Akta system that we could use for exploring and optimizing purification protocols. Most of our work is done on novel proteins of unknown sequence and we will typically follow purification by some sort of activity assay. Making this process quicker would help our work enormously. Because we typically are interested in novel proteins we have less interest in purchasing from a supplier or sending our material to a core facility.
craigm, Principal Investigator, Australasia/Pacific

If money was no object - I would prefer to purchase the recombinant protein from a vendor. I would require quality control to be in place and a minimum specified specific activity per prep. Even though we have a large campus with many proteins being purified - I don't think automation system would make sense, due to the intricate fine tuning that is usually required for each individual protein.
virushunter, Principal Investigator, North America

If money was no object I would keep the current purification system. Expression is usually the tricky part. Not purification. I would invest in a high-throughput cloning/expression project: With a various tags, vectors and hosts combinations. Find the best one for each particular protein. Then purification is easy.
emmanuelle, Lab Director/Supervisor/Coordinator, Australasia/Pacific

If money was NO object I would send my proteins to a trustworthy core facility to be purified. The reason being that they would have a lot of experience purifying and their troubleshooting time would be a lot less than mine. The objective of purifying proteins in my research is to use them as a tool; for this reason I would not like to waste my precious time towards purification.
dr_ashishg, Principal Investigator, North America

If money were no object and it would be possible to get the proteins of interest from commercial sources we would go and buy the proteins there. Our lab would do the purification in-house, because we have some experience with the purification. Although it would save much effort and disappointments to get the purified proteins from somewhere else.
peptide, Graduate Student/Research Assistant, Europe

If money were no object and results were guaranteed, we would gladly invest in a high-throughput protein purification platform if it would provide us with large amounts of biologically active proteins. Our studies are complicated because of the fact that we generally deal with membrane bound ion channels and accessory channel proteins from Drosophila. We are interested in site directed mutagenesis of these proteins in order to define which parts of the proteins interact. Integral membrane proteins are notoriously difficult to work with because they are prone to aggregation. In addition, the ion channels that we work with (sodium and calcium channels) are large and prone to degradation during purification. I would have to be convinced that any purification system would work with our preparations.
linny34, Principal Investigator, North America

If money were no object I would obtain purified proteins from a commercial purification service. I would send them an expression vector containing my sequence of interest. For prokaryotic expression this vector would probably contain a His- or GST-tag. For eukaryotic expression likely a Strep-tag or a His-tag. The company would transfect the DNA in the E. coli strain/cell line of my choice, culture and lyse the cells, and would purify the protein using the tag. The tag would then be cleaved of and remaining impurities removed e.g. by gel filtration. I would then receive the protein in the buffer of my choice. ... I should stop dreaming and return to reality. Such service would save many worries but would probably be to expensive although the lab time expenses for the lab staff to establish and perform these purifications don't have to be neglected.
thorstEE, Post Doctoral Fellow, Europe

If money were no object I would put my gene of interest into a Gateway vector and clone it in parallel to every available expression vector and do a high-throughput screen on which system worked best and scale-up. I would also like to find an algorithm that would predict which amino acids in a protein should be changed to give better solubility. I would still purify the protein in-house to keep better control over the process.
Proteome, Staff Scientist, North America

If money were no object I would send proteins out to be purified, or would purchase directly from a supplier. Failing that, I would invest in an FPLC-type system that would I could do purifications on, as well as other analyses. I would purify from whatever worked best, first choice would be His-tagged proteins, if solubility were a problem, I would go with MBP fusions, or would use the IMPACT system if I absolutely needed native protein.
szat, Professor/Teacher, North America

If money were no object I would try and find a custom service for purifying our proteins as this would free up my postdocs/students to do more interesting experiments with the resulting proteins.
sueRT, Principal Investigator, Europe

If money were no object, a high-throughput automated platform would be used. The increase in throughput would allow for greater productivity. In our business, greater output means greater profits. Also, if money was not an issue, purchasing purified proteins from a supplier would be advantageous over the time investment of purifying in-house.
gamab, Staff Scientist, North America

If money were no object, I would both buy and purify my own proteins and compare activity. I would get the vectors with multiple cloning sites with the most flexibility and with the greatest number of tags because I could buy any antibody I wanted. I would consider high-throughput depending on how quickly the lab went through their supplies of proteins.
cancun1991, Principal Investigator, North America

If money were no object, I would defer protein production to an outside supplier. This would enable me to focus on the research at hand and not worry about the production of a "reagent" protein. Putting this task into the hands of a company or individual whose primary focus is the production of protein would also yield a better final product.
Dick P, Lab Director/Supervisor/Coordinator, North America

If money were no object, I would have a tech dedicated to the purification of proteins using manual methods. I have not been impressed by the yield from the automated systems to date, and I use large quantities of purified protein in various biochemical and biophysical experiments as a protein biochemist. I would invest in a French press and a full range of gravity column matrices and an FPLC or HPLC. Additionally, I would have a bench top Mass Spec to determine purity and possibly sequencing.
saxl, Post Doctoral Fellow, North America

If money were no object, I would most definitely purchase purified proteins from a supplier or send them out to a core facility to be purified. Though automation saves times, its nothing like getting the job done by someone else without cluttering my lab with complicated automation. I can use the time saved for more important work involving the purified protein.
vashisht, Post Doctoral Fellow, North America

If money were no object, I would outsource the production/purification of specific recombinant proteins. My time would be better-spent developing assays for using the protein(s) of interest.
rjubinlin, Staff Scientist, North America

If money were no object, I would prefer producing my proteins in eukaryotic systems, perhaps using the adenoviral expression system, so that the posttranslational modifications of the protein would be closer to the physiological ones. A high-throughput purification platform would be very interesting for the purpose of producing high quality protein preparations for immunization studies. If money were no object, we would buy purified proteins and would send material to a core facility for purification, no doubt.
Acid, Professor/Teacher, Europe

If money were no object, I would purchase the purified proteins from a supplier and avoid the headaches involved in doing it myself.
jethrod1, Principal Investigator, North America

If money were no object, I would send my protein to a core facility to be purified, because they are the people who have the expertise for this type of techniques. It would be useful to concentrate these techniques and automate them to a certain extent in order for the ordinary research groups to be able to concentrate on the real biological questions and not just in technical details.
eroine, Post Doctoral Fellow, Europe

If money were no object, I would send my proteins to a company to be purified. For me, a purified protein is a tool to establish an assay in order to address biological questions. Because I'm not working with unknown proteins, or trying to discover new proteins, to purify proteins is a waste of my time. I have to do it because nobody else would do it for me, or because I have no money to send it to somebody else to do it for me. It is a time-consuming activity, it's never straight forward, the protocols available not always work, there are so many variables that can ruin the purification of a protein, and I think one need to be a bit of a specialist in the field. Ideally, if I could have a protein chemist working for me I would be in seventh heaven.
chivite, Graduate Student/Research Assistant, Europe

If money were no object, I would simply purchase them from a core facility.
academian, Professor/Teacher, North America

If money were no object, I'd send my proteins to a core facility to be purified. They'd be able to do it more quickly with better consistency than my lab would. This is why I send my sequencing to a core facility.
pamiam, Professor/Teacher, North America

If money were no object, it would be great to be able to pick proteins of interest off of a shelf just like we currently do with restriction enzymes.
donny, Post Doctoral Fellow, North America

If money were no object, it would not make sense to tie up lab personnel and resources to purify proteins. We would definitely use a core facility for purification if the proteins were not available from a supplier.
Dr.Diagnostic, Lab Director/Supervisor/Coordinator, North America

If money were no object, of course everyone would prefer to purchase purified, active proteins from a vendor. In the real world, however, the cost to purchase enough protein for HTS is prohibitive for even the biggest companies. In addition, a lot of proteins are proprietary and must be purified in-house for confidentiality reasons.
klundeen, Staff Scientist, North America

If money were no object, we would adapt our protein purification protocol to automation using large-scale column chromatography and automated fractionation. I would also adapt our current fractionation system to incorporate automated regeneration of the columns so that repeated runs could be obtained without manual intervention. Furthermore, I would automate the protein de-salting process to incorporate a system that would desalt *and* lyophilyze the resultant protein products. As very few laboratories are skilled at purifying collagens from tissues, it is unlikely that we would be able to send our proteins out for commercial purification.
dbrand, Principal Investigator, North America

If money were no object, we would obviously purchase already purified proteins if available, or send a construct out for custom expression/purification if not available commercially. As it is, a post-doc's time is cheap in comparison.
Umber, Laboratory Technician, North America

If money were no object, we would outsource all of our protein purification needs. Why bother building capacity, investing in expensive capital equipment costs, and hiring highly qualified personnel if a third-party provider can express, purify and ship proteins directly? As for core facilities, with no budget restrictions we would definitely use third-party private sector providers. Academic core facilities tend to be slow, buried in collaborative projects and under-equipped in the automation required for high-throughput purification.
rwintle, Staff Scientist, North America

If money were no object, we would send plasmids to a facility for expression and purification of recombinant proteins.
bblumberg, Principal Investigator, North America

If money were not an object, I would order purified protein from a commercial source or outsource antibody production from serum free cell culture for molecules that are not commercially available. I don't purify proteins frequently enough to justify having an automated system in the lab, unless money was not an object or the automated machine was extremely cost effective and didn't take up much space.
apoptosis47, Post Doctoral Fellow, North America

If money would be no issue I would like to have a resource of pre-purified proteins where I can purchase what I need.
rslany, Principal Investigator, Europe

In a perfect world we would definitely automate our protein purifications. Currently we have to run multiple (up to 4) FPLC columns per enzyme subunit and our pet enzyme is a heterodimer, which increases the amount of work to purify a functional enzyme markedly. Given that site directed mutagensis is now routine, we as a lab are spending inordinate amounts of time standing in a cold room watching (and willing) peaks to appear. I believe we would be reticent to out source the protein purification and would much prefer an in-house high-throughput system.
Melisa, Post Doctoral Fellow, Europe

In an ideal world, we wish to perform ideal science with the best possible model system. For instance, if we are studying a group of human proteins, I would invest as much time and energy as possible in using a mammalian/human cell expression/purification system to obtain these recombinant proteins in their most biologically relevant form. The inherent disadvantage to such a system is low protein yields, therefore I would scale-up my production of these proteins and using my unlimited resources, I would employ a number of technicians to mass produce milligram quantities of these recombinant protein from the gallons of cell cultures that they would be maintaining. Because my highly active lab would be pursuing a number of projects, a high-throughput protein purification system would be purchased for maximum efficiency and turnover. Furthermore, I would ensure antibodies were made to these proteins so that I could perform affinity purification of these proteins to ensure maximal purity. Therefore, with the remainder of my riches, I would ensure that at least three antibodies per protein were produced and would make sure each were outsourced to a customized antibody production company. In addition, I would use HPLC purification to separate differentially post-translationally modified proteins so that specific forms of these proteins could be studied.
gradstudent, Graduate Student/Research Assistant, North America

It all boils down to the level of expression and the expertise available in the lab, but my personal choice would be working with inclusion bodies at higher scale of expression and purification because it accounts for the major pit falls in expression and purification, i.e., toxicity of recombinant protein on cells producing it, protein degradation, level of expression, the most crucial step being its refolding, because I consider purification from inclusion bodies as a reproducible method provided a good established expertise and analytical tools as well are available for refolding of proteins, which can be followed by ion-exchange, and if needed size exclusion chromatography (for oligomeric proteins). High-throughput purification system would be definitely a good choice as this would ensure speedy, efficient and reproducible purification, as we know TIME is as precious as gold in purification moreover it saves on unnecessary labour associated with it. Purchasing a purified protein from supplier is a good idea as it would work as control in initial experimental designs to develop a purification strategy, but best would be to develop the strategy in-house, as it would mean less dependence on external sources and better control over further experiment designs. Sending sample to a core facility would not be the best idea, as anyway it would need standardization again under lab conditions and dependence on other sources as mentioned earlier.
amar, Graduate Student/Research Assistant, Asia

It is critical that we purify high yields of isotopically labeled protein for NMR spectroscopy. We cannot have tags on such proteins in most cases and it would be ideal if we could subclone large numbers of proteins, express them, and have quick small scale purification tests that would scale-up easily. Our single largest problem is not purifying tagged proteins, but cleaving the tag efficiently and cost effectively. For these reasons, we often do not even try tags on our first try, but rather we try purifying the protein directly from the overexpression cells. The cells that we use for expression have to be able to grow on defined media, so the labels are incorporated (13C, 15N, 2H).
juanny, Professor/Teacher, North America

Large scale in mammalian expression system; FLAG tagged to minimise interference with functional/pharmacological properties. High-throughput not necessary for my interests, use it for protein-protein interation studies in the main. Prefer to purify protein myself rather than send away.
apul3, Professor/Teacher, Europe

Large-scale in vitro translation in retic lysates. They work the best for most of our proteins with respect to solubility and activity, but are prohibitively expensive for large-scale purification.
lguarino, Professor/Teacher, North America

Low scale high-throughput analysis of proteins (50-96 analysis for presence ELISA and activity at a time).
Bibikov, Staff Scientist, North America

My dream for protein purification would be to receive them pure from a reliable source, ready to start my experiments.
tblumens, Post Doctoral Fellow, North America

My favorite purification tool is FPLC system. The reason is simple: flexibility. There FPLC allows you to use several different strategies, and one can be easily fed into the next. Also, there is a certain amount of creativity that goes into designing a successful protein purification scheme, and this is something I enjoy.
tritty, Principal Investigator, North America

My lab is interested in producing large amounts of highly pure proteins (mostly enzymes) for biophysical analysis, crystallography and mechanistic studies. In general, production of enzymes with high specific activity free from contaminating enzymatic activities is paramount. Despite what the manufacturers might say, no single affinity chromatography step can ever produce a truly clean protein prep. It is simply a case of whether the protein is clean enough for your downstream application! If you're going to spend weeks, months or even years investigating your protein, then it makes sense to spend a week putting it down a couple of complementary columns, e.g. cation exchange, anion exchange and/or gel filtration. To quote Kornberg: "don't waste clean thinking on dirty enzymes"! Typically we run 1 to 2 litre bacterial fermenters with custom-made heat shock promoters. Native lysis followed by ammonium sulphate cuts and dialysis. Heparin columns figure quite often, as they are good for DNA binding proteins and our nucleases at neutral to slightly basic pH. Then an anion exchanger followed by a sulfpropyl cation exchanger at slightly acid pH. Monodispersed resins are best, but expensive. Disposable columns are best as they cut down the chances of cross-contamination. Purity is most important followed by high yields in our line of work.
jrsayers, Professor/Teacher, Europe

My opinion on high-throughput protein purification for BIOLOGICAL purposes is that someone needs to develop a robust process for production of specific affinity reagents, similar to the ideas behind single chain antibodies. With these reagents, a single batch of cells or bacteria need to be grown under carefully controlled and defined conditions, and the protein pool can be excised from a single experiment. Trying to express thousands of (tagged) proteins under "similar" conditions is a short-term fix to underdeveloped technology in protein affinity purification. Even efforts to express many differently tagged proteins simultaneously in batch culture is simply unlikely to ever be proteome-scale. If money were no object, I would invest in the technology to manufacture unique affinity reagents for each protein, and protein derivative, in an organism.
Bluegenes, Staff Scientist, North America

No, we would not invest in a high-throughput protein purification platform. That would not be practical considering the types of experiments with recombinant proteins that we perform. But we might consider sending protein to a core facility for purification, raising antibodies etc.
marek, Staff Scientist, Europe

None of the purification step gives 100% purity and desired yield. But it is possible to select series of optimized steps to get maximum purity and maximum yield. The beauty of protein chemistry allows us to select and optimize the conditions to screen desired molecule from among thousands of undesired macromolecules.
kumarkurle, Graduate Student/Research Assistant, Asia

Often there are a number of problems associated with purification - particularly for increasing the yield. Therefore, for routine jobs where the yield of recombinant protein is important, and in cases where stability is not a factor, it may be helpful if there is some core facility for purification jobs. Core facilities (special expertise) may be useful where copurification problems are encountered. However, it may be necessary at the initial stages to perform a pilot study to confirm the purification - the functioning of the tags and the accessibility of the tags (e.g., to the Ni or Cobalt resin systems.
tssen, Professor/Teacher, Asia

Our lab would purchase an automated protein purifier. We would use a simple GST column for purification, using an insect cell line (Sf9) to produce the protein of interest. We would invest in a high-throughput system to help in the production of protein, since we use this protein for enzymatic studies and protein structure studies. Secondly, we would like to have a high-throughput method that would prevent inclusion bodies. If we could outsource a clone that we have already made, this would be our method of choice since we do have inclusion bodies and are currently having problems identifying the technical problem with our system.
Gruther, Staff Scientist, North America

Our laboratory is so small that some sort of automated purification system would need to be available for automation to make any dent. So my ideal would be an instrument that could automatically isolate tagged proteins, analyze the purity, cleave the tag, and provide mg quantities of 100% protein of interest -- all for under 25K.
Oberon, Principal Investigator, North America

Our proteins have to be purified from scratch by detecting their interactions with specific RNAs. If money were no object, I would send the protein: RNA complexes off to a facility for purification & protein identification.
hollings, Principal Investigator, North America

Perhaps HPLC (reverse phase or similar, that can separate larger proteins >55kDa). Our proteins have a large number of expressed isoforms in vivo, and when we use recombinant proteins, we are only dealing with one particular isoform. Therefore, if we could isolate all the expressed isoforms of each protein from tissue, this may give more representative results in activity assays. We would do this in-house, but our HPLC has been broken for years and demand for use is too low to get it fixed and running.
jreed, Post Doctoral Fellow, Europe

Probably we would still use the low-cost methods we are using now. Due to the nature of our work, gram scale isolation and purification of a small number of specific proteins, we have found that the higher cost methods (in terms of money) are much more time-consuming than the methods that we have developed ourselves. I would expect the story might be different for micro-scale purification of large numbers of different proteins. In terms of purchasing purified proteins from a supplier, or using a core facility, we have found that we are able to do a much better job of purification if we do it ourselves. For anything we buy, we buy the lowest purity, and then we purify it ourselves.
jterner, Principal Investigator, North America

Protein purification is a technological tool. If money were no object, we would use a core facility, and allow our staff to devote more time to the real scientific questions.
dvstirling, Lab Director/Supervisor/Coordinator, Europe

Protein purification is an empirical science. It seems, in our hands, that no two proteins behave the same during purification, even under similar conditions (expression and purification). That being the case, if money were NO object, it would be a difficult decision between purchasing quality controlled proteins from a supplier and hiring additional technicians to perform purification in-house. Nothing beats a pair of experienced hands when detailing a protein purification protocol. We currently use Akta FPLC units for rapid purification of all our protein samples. If we chose to drive protein purification in-house, we would also supply one Akta for each technician hired. The Akta FPLC is indispensable when it comes to rapidly repeating purification protocols. Money being no issue, I'd vote for purchasing quality controlled proteins from a vendor and devote our full attention to the business at hand--drug discovery.
socalsmithers, Staff Scientist, North America

Protein purification is NOT the challenge; expression/synthesis of high-levels of soluble protein IS. Currently we often need to resort to multiple runs of 100 Liter fermentations, followed by large-scale purification in order to generate 10 mg of pure, soluble protein.
tstout, Lab Director/Supervisor/Coordinator, North America

Protein purification is quickly leaving the specialized lab and becoming a part of every researcher expertise. The speed and ease factors are becoming the most important for most people I work with today.
kemtoi, Staff Scientist, North America

Protein tagged with Fc portion of IgG and it's a soluble protein. In our culture it comes in supernatant and then we purify it using different IgG tagged column.
pranaymeenu, Post Doctoral Fellow, North America

Purchase from a supplier!
robynap, Professor/Teacher, North America

Purchase purified proteins on a custom basis.
pharmurr, Staff Scientist, North America

Purchasing ready-made proteins would be to good to be true. Especially for labs that need truncated forms which may not have a membrane spanning region, or only carry a catalytic domain. For this reason, ideally we would be able to use a quick expression system that could produce mg quantities, purify on the benchtop by means of spin columns, and purify to at least 65% purity. We would be willing to pay more for something this quick and disposable, as long as it worked well.
lahn1, Staff Scientist, North America

Purification would stay in-house. Ideally, we would scale-up to several Millipore K40 systems, as I find these to be best suited to our purposes.
royal, Production/Manufacturing, Australasia/Pacific

Recombinant protein purification is becoming an rapid growth area. High throughput methods are needed to achieve greater purity and efficiency. In this regard development of new matrices is a very key area of research.
sraju, Principal Investigator, North America

The ideal source of purified protein is to have collaborators provide them in a pure and active form. But, back in the real world, we work with a variety of proteins and consequently do quite a few different methods to purify tagged, untagged, native, recombinant proteins in a variety of expression systems. If money were no object, then I would be more likely to hire more staff and purchase robotic, high-throughput systems to purify some proteins. Certainly, robotic devices could be used to take lysed cells of various organisms and produce purified proteins. The robot would have to be versatile enough to change the protocols easily. Core facilities are great for specific and repetitive purposes, but they then to fail when you keep changing the jobs you want them to do.
rstevens, Post Doctoral Fellow, North America

The key to successful protein purification is having people with the skills and expertise to make it work reproducibly. Having good and adaptable is considerably more important than automation.
terrance, Principal Investigator, Australasia/Pacific

The most important technique wanted in the field of protein purification is tag-resin specificity. GST, His, or even biotin can non-specifically absorb detectable amount of unrelated proteins. We need a resin which only recognize the tag protein.
Takeshi, Lab Director/Supervisor/Coordinator, Asia

The protein(s) we are working with has been previously characterized. So regardless of the cost, we would still use the small-scale kits because it is the most appropriate size for the scale of the experiments we are performing. While we are not a rich lab, money is not the primary factor is adopting a protocol especially protein purification. The most influential factor in choosing a kit or supplier for a particular assay that I am new to is from relevant publications. Each protein can have its own unique set of problems when purifying. If it has been purified before, we obviously follow their protocols. New or unknown proteins are obviously another matter all together.
etoainshrdlu, Graduate Student/Research Assistant, North America

There is no doubt that affinity-based purification techniques are of fundamental importance, particularly for the proteomics analysis of protein complexes, and also for the removal of common contaminating proteins (e.g., serum albumin). The use of ligand dye chromatography is largely overlooked.
Nice, Principal Investigator, Australasia/Pacific

Use the Roche RTS system for almost all proteins since they have a guarantee for getting soluble active protein.
stev434, Staff Scientist, North America

We are a small lab and could not afford or care to have high-throughput protein purification. Most of our applications do not require very high protein concentrations.
biofrog, Staff Scientist, North America

We are currently forced to use maltose binding protein tags to achieve high solubility and purity for the protein that is currently our major interest. If money were no object, we would pay a core facility to purify and concentrate it from a medium to large scale fermentation vessel (at least 10 liters media). I would not invest in a high-throughput system because I don't need it. I would purchase the highly-pure proteins if they were available and of high concentration and activity.
Lukac, Principal Investigator, North America

We are in the process of express and purify parasitic proteins in the new lab, in the process of set-up. We shall be interested to invest in a HTP purification using good bacterial expression system because impossible to extract the essential proteins from parasite.
indirag, Staff Scientist, Asia

We don't purify enough to need high-throughput.
RBFree, Post Doctoral Fellow, North America

We plan to purify a secreted protein from CHO cell expression, but the expression amount is very low. We are considering how to improve the yield.
z9615, Staff Scientist, North America

We use GST purification because its high affinity and easy to purify. If we have enough money, we would invest in a high-throughput protein purification platform, that would greatly improve our research process. We already cost hundreds of dollars for purchase one purified, recombined protein, just because it is too expensive, we can not buy as much as we want, there are a lot of interested targets for researching. Each protein have several different domains, if you could purify each domain separately, that will greatly improve function research. If there has a core facility, of course I will send my proteins for purifying, just like sending my DNA sample for sequencing.
y27xiang, Post Doctoral Fellow, North America

We use ion exchange, gel filtration and affinity chromatography. If you automate these into one giant machine, that would be great. Then we can tailor make the protein purification protocol to suit our needs.
Venk, Professor/Teacher, North America

We use purified proteins in studies aimed at determining the biological properties and function of the protein. Hence, they must be as pure as possible within the time constraints that we work. We are unlikely to use an automated system because there is no automated system that would purify our protein. We cannot use immunoaffinity columns because "our" protein is denatured by such columns; we cannot use any other type of affinity column because "our" protein is a protease that autolyzes quickly unless kept in an inactive state; the protein precipitates at its isoelectric point so isoelectric focusing cannot be used; and a combination of ion-exchange, hydrophobic interaction, and size exclusion and 5-6 successive column runs are required to purify the protein. It is a difficult one to purify and a large number of people have been working at it for over 25-years. Finally, it is difficult to express (requires two subunits for activity). Our expression studies use a second protein that is easy to express.
Darrel, Professor/Teacher, North America

We work with membrane protein complexes and their components; therefore the full purification scheme includes also electrophoresis (1-D or 2-D, standard or not depending on the complex). For our purposes it would be ideal to have a combination of electrophoresis and electroelution with HPLC (ion exchange, gel filtration and reverse phase columns) in micro and nano scales equipped with different detectors including mass spectrometry. We don't need high-throughput; instead, we are interested in isolation of unknown complex components that are present in low amounts per cell in quantities enough for analysis of the protein structure including post-translational modification. This task is not routine and not for the core facility.
Natbat, Post Doctoral Fellow, Europe

We would certainly consider Roche's machine if it weren't so expensive.
mtuvin, Lab Director/Supervisor/Coordinator, North America

We would do very large scale preps (>10 liters), and use larger columns for purification. It is also possible we would use affinity chromatography, though we are concerned about how the elution conditions would affect the activity of the proteins. We would not be interested in a high-throughput system, since we are doing large scale preps. Since we are looking at unique hybrid proteins, we could not purchase them from an outside supplier.
dfritzin, Principal Investigator, North America

We would further invest in a core facility to purify proteins for in-house needs. We appreciate the responsiveness of our in-house group.
Jim110, Principal Investigator, North America

We would invest in an automated system for the following reasons: 1) In-house means fast, reliable, customized 2) We can run it any time (day, night, weekend) 3) Would allow us to obtain many more samples under a variety of conditions in one experiment 4) Expected to reduce the sample to sample variability Purification protocols to be used would include: 1) Ni-NTA metal chelate affinity chromatography 2) GST- and other TAG purifications 3) TAP-tagging purification: Protein A-calmodulin binding domain.
cruiser, Professor/Teacher, North America

We would love to have an affordable system for automated protein purification based on affinity tags. In our case we need a system that can purify gram quantities as efficiently as possible due to our using them to form biomaterials. We could afford the cost of automated equipment if the cost of supplies (particularly the resins) were lower. Ease of use is also a factor as we can utilize skilled undergraduates for many processes, especially if they were automated.
spidersilk, Professor/Teacher, North America

We would outsource the protein production, put it into skilled hands for scale-up and purification. Quantity of recombinant protein is always the limiting factor in experimental design; any way of removing that obstacle would increase productivity.
Edward98, Principal Investigator, North America

We would prefer a tandem purification that is potentially automated, and we definitively invest in a high-throughput protein purification platform to be able to quickly optimize expression conditions for multiple constructs simultaneously and to have sufficient material for initial activity and structural screens before scaling up for production. As starting point affinity chromatography combined with gel filtration would be optimal. We would not buy commercial protein since most of our proteins are novel targets anyway. Of course purification in large scale should be done in a core facility to ensure quality and throughput.
steschmi, Lab Director/Supervisor/Coordinator, Europe

We would probably continue to use the GST purification system for its ease of use, generally good yields and reliable success rate. This purification doesn't require a huge time commitment, and yields enough protein for many experiments, so we would probably continue to do it ourselves.
baumina, Post Doctoral Fellow, North America

We would probably invest in a high-throughput system to rapidly screen kinetic properties of rationally engineered variant proteins. For subsequent biophysical analysis of interesting variants, however, a high-throughput system would not be able to produce the necessary yields. For this, we would use standard scalable expression and purification techniques.
cchang2, Post Doctoral Fellow, North America

We would probably invest in high-throughput protein purification platform. The main problem that we experience is insolubility of isolated proteins; there is some evidence that using some high-throughput platforms leads to higher protein solubility. Ideally we would send the proteins to the core facility on the campus.
kegenzab, Graduate Student/Research Assistant, North America

What purification protocol(s) would you use and why? For crystallisation purpose we have noticed that affinity chromatography (e.g., His-tagged proteins) followed by size exclusion chromatography is sufficient for 90% of purified proteins. A three-step purification protocol can be pursued if crystallisation does not work. The protocol then would be affinity, ion exchange (anion or cation exchange chromatography) and size exclusion chromatography. Thus one can purify the target protein from other co-eluting proteins or impurities. Would you invest in a high-throughput protein purification platform? Why or why not? I would definitely invest money if we have that demand of purifying many proteins at a go. Would you purchase already purified proteins from a supplier or send your proteins to a core facility to be purified?. We buy purified proteins from the suppliers where purity is not that crucial. For the target proteins if a constant/ regular supply is needed, I may utilise the core facility. Otherwise I would rather have the control myself on the protein and purify in my lab. These days, there are so many ready made kits and solutions that one does not need to be an expert in purification, specially for the tagged proteins.
Ramagauri, Production/Manufacturing, Europe

When cleave GST fusion protein by Factor Xa, sometimes N-terminal sequence are random.
Eisaku, Staff Scientist, Asia

When starting with a new project/protein, it could be useful to have a set of host/vector systems (as a starter kit) to check and screen which one will be the more efficient or at least suitable.
jojema, Professor/Teacher, Europe

With the increasing number of ORFs being predicted from different organisms and the deduction of interactions between the major players in the cell, the research in future has to be definitely high-throughput as well as automated. Already there are systems available that would give researchers proteins directly after addition of PCR products into the reaction mixtures. Advancements like these will definitely propel the sciences very fast towards newer discoveries as these simplify the mundane steps that are well standardized now. There will be a greater flexibility in getting the proteins purified from a core facility as then one can get the proteins for e.g., with or without tags which might be useful at the stage of analysis like during immobilization on chips etc. Thus in future purification should be automated and custom required which will make the whole process faster and also according to the need of each system under study.
sci2k1, Graduate Student/Research Assistant, Asia

Yes, I would rather prefer to invest in high-throughput protein purification platform rather that purchasing already purified proteins from a supplier. This will give me the ease of obtaining proteins of different purity, which are interacting with other co-purifying proteins, which I can use in my assay system to understand and decipher the role of interaction on the activity of my protein.
thyrox, Post Doctoral Fellow, Asia

To go back to the summary, please click here.

To request information on purchasing the full report, please contact Mike Kibler at The Science Advisory Board.

[ Back to Past Studies ]