Hello, everybody.
I have been having a serious hard time expressing a recombinant protein and I would like to check any source I can in order to achieve some results in my project. Herein I will explain some of the details of my issue and, before anything, I would really like to thank you for any information you can provide me.
First, the protein I am trying to produce is one of the subunits of a cytochrome protein. The DNA coding for it has been cloned in a pBAD/gIII (A) vector (L-arabinose inducible) and the presence, orientation and completeness of the insert has been confirmed by PCR and digestion experiments in several clones (sequencing is being performed in this moment).
I have tried the expression of the protein according to the instructions manual of the vector. In it, there is a suggested "primary" expression in which 10 mL cultures are used in order to test the system. Several arabinose levels used for induction are suggested to be tested and the incubation of the cells with the arabinose is carried for a total of 4 hours. The cells being used for transformation and expression are TOP10 cells. All growth media is supplemented with 50 ug/mL ampicillin.
In the experiments I've performed I allowed the 10 mL cultures to grow to an OD600 of 0.4-0.5 at high speed shaking (200 rpm) and 37ºC. After this initial growth, I've induced the cultures with L-Arabinose to the final concentrations of 0.00002%, 0.002% and 0.2%. The cultures were incubated in the previously mentioned conditions for 4 additional hours and afterwards they were harvested by centrifugation. The pellets were stored overnight at -20ºC and lysis was performed in denaturing urea buffer using silica-zirconia beads in a FastPrep machine. The whole process was repeated with 6 colonies.
Another experiment was performed in which induction was performed with L-arabinose (in the same conditions previously mentioned) and δ-Aminolevulinic acid (ALA) to a final 1 mM concentration. ALA was added to the culture at the moment of induction (OD600 0.4-0.5).
Detection of the recombinant protein was performed by Western Blot with Anti-His 6x (coupled to HRP) antibody. The detection mechanism was autoradiography using the LumiGLO chemiluminescence kit.
No protein expression has been detected in any the previously mentioned experiments. Positive control (calmodulin) seems to work properly.
Do you have any information about why there is absolutely no detectable expression of the protein in the selected system?
I really, really appreciate your help in this situation. Please feel free to request me any additional information that you consider relevant.
Best regards,
Andrés.
It seems an interesting question for me. This is due to hydrophobicity of membrane proteins. Traditional detection methods, such as SDS-PAGE, western blotting, and immunological method, are not applicable to the hydrophobic membrane proteins at all (please see the file; IEF for hydrophobic proteins).
Calmodulin is soluble protein with hydrophobicity 0.456 similar to avidin of 0.451, and the protein detection is easy to perform as expected. However, please use the quantitative methods as described below in order to precisely determine the expression levels.
On the other hand, most of the subunits of cytochrome proteins are hydrophobic membrane proteins.
Although hepatoma HepG2 has not, fetal normal liver Hc cells has membrane protein of Cytochrome c oxidase subunit 2 (hydrophobicity of 0.635; similar to hydrophobicity of olfactory receptor (OR) of 0.637) at 18.6 μg/mg of cell protein.
Liver specimens usually have cytochrome proteins.
LC tissue (with leprosy) has Cytochrome P450 XIA1/Cholesterol desmolase/Cholesterol side-chain cleavage enzyme, mitochondrial at 4.1 μg/mg of tissue protein, whose hydrophobicity is 0.558 (similar to insulin receptor (IR) of 0.533).
HCC tissue (with PBC) has Cytochrome P450 IID6/cytochrome P450-DB1 at 0.16 μg/mg of tissue protein, whose hydrophobicity is 0.692 (similar to insulin of 0.676), Cytochrome P450 11B2, mitochondrial/Cytochrome P450Aldo at 3.6 μg/mg of tissue protein, whose hydrophobicity is 0.580 (similar to E. coli KdpD protein of 0.578), and Cytochrome P450-C17/Steroid 17-alpha-hydroxylase/17,20 lyase at 2.7 (total 6.5) μg/mg of tissue protein, whose hydrophobicity is 0.542 (similar to Guinea pig liver lipoamidase of 0.540), respectively.
LC tissue (No.6) has Cholesterol 7-alpha-monooxygenase/Cytochrome P450 7A1/ Cytochrome P450 VII/CYPVII at 1.8 μg/mg of tissue protein, whose hydrophobicity is 0.543 (similar to human carbonic anhydrase of 0.545).
Serum of 33ｙ healthy male has Cytochrome P450 3A3/Cytochrome P450 3A4 at 9.6 μg/mg of serum protein, whose hydrophobicity is 0.568 (similar to human urinary/kidney biotinidase of 0.573).
Serum of 1y girl (at 44w of biotin therapy) has Cytochrome c oxidase subunit 4 isoform 1, mitochondrial at 3.4 μg/mg of serum protein, whose hydrophobicity is 0.498 (similar to BSA of 0.496). This protein is membrane protein, but has hydrophilic nature.
Serum of 4mo boy (biotin therapy at 13w) has Cytochrome P450 4F12 at 3.2 μg/mg of serum protein, whose hydrophobicity is 0.562 (similar to human beta-lactoglobulin of 0.562).
Therefore, most cytochrome proteins are difficult to detect except Cytochrome c oxidase subunit 4 isoform 1, mitochondrial (this gene is described in the host cell nucleus, and not present in mitochondrial DNA).
Then, quantitative methods such as Short-Column-RP-HPLC, HPLC-Soap-SEC, and protein-direct-microsequencing-deciphering (PDMD) method are expected to be useful for the determination of hydrophobic proteins (please see my publications uploaded in this ResearchGate).
I'm assuming that detection of calmodulin is with anti-hist as well? if not then you need a positive control for your anti hist antibodies.
I know very little about cytochrome proteins but there are general considerations to look at when expressing proteins.
1. Confirm sequence: analysing lots of clones makes more work and may be counterproductive if it does not work and you need to troubleshoot problems
2. positive/negative controls for expression.
3. Make sure your anti Hist antibodies work, use a positive control on gels
4. I would carry out a time course of expression and take samples at 1, 2 4, 8hrs etc to check for expression. Do you see any detectable protein expression by SDS-PAGE/Coomassie or do you have to WB?
5. Observe the culture: do cells grow ok. It is possible that some proteins are detrimental to the cells when induced (Toxic). Induce at higher OD's and for shorter periods (1-2hrs)
other variable we may use are:
1. Temperature of induction
2. Media: LB, 2xYT, TB etc
3. Different strains: JM109, Top10, XL1blue
4. Heat shock on ice
You may also consider the possibility of having a cleavage site in your protein. Did you try to locate a His-Tag in the other site of the protein?
The answer of Dr Hayakawa is really interesting tough, something I will honestly never think about... And I agree with Dr Wilkinson that a positive control is essential to draw conclusions from your experiment.
Anyway, I will wait for the sequencing result. It can be a problem of frame shift that you won't be able to detect by PCR/Digestion.
Amended on 21 December 2016
Furthermore, I would like to add one more serious case.
Dr. Sandesh Sharma Pandit (Northern Illinois University, DeKalb, IL, USA) has asked similar question as follows; i.e.,
"Rosetta gami b transformation, no transformants?"
It is surely very sorry for both employee and employer since such commercial kits are extremely high cost. Surely, I also have an unforgettable bitter experience during an Arbeit at elsewhere.
"I am trying to transform a plasmid into Rosetta GamiB strain , I used both competent cells from the manufacturer as well as I made it myself, but I am failing to get any transformants, even in positive control. What could be the reason ?".
In this serious case, I have answered as follows, since Dr. Sandesh Sharma Pandit could not get transformant even in the positive control.
“I would like to express my opinion. I have recently found that c.a. 12% of cellular proteins in Lactobacillus casei (Shirota) are of phages' (120 ug/mg of cell protein) by using PDMD method (Protein-Direct-Microsequencing-Deciphering Method) (please see file; HepG2 Fucoidan). Therefore, transformation using small amount of DNA may be difficult due to interferences by other bacteriophages (bacterial viruses).
I have once succeeded in the co-transduction experiment (please see file; P1kc RodA E.coli), however it has been all in vain to apply the gene transfection technology at all in the next study using Pseudomonas K62 (mercury-binding protein resides in the outer-membrane of the bacterial cell; please see file; Resistance to Hg).
I have now estimating that transfer of genetic information (gene transduction) may be strongly related to the region of host genetical map; i.e., my first paper has handled with the co-gene expression of inner-membrane proteins of rodA (protein for determining the rod shape of E. coli) and of lip (lipoic acid synthetase) at E. coli chromosome map of 15 min.
Further, my former late teacher Dr. Kazutomo Imahori （Prof. Emeritus, Department of Agricultural Chemistry, The university of Tokyo; 1 June, 1920 – 8 May, 2016) has once questioned me that "why phage lambda is integrated into the position of the lambda attachment site attB in E. coli chromosome at 17 min between gal and bio", which I have not possible to answer yet.
In order to answer the benefactor's important question, I have tried to answer as far as I can. I have calculated the median values of hydrophobicity (lipophilicity; LipPhil; about definition please see file, JMBT Alopecia) of genes of Escherichia coli. The lambda attachment site attB is present at 17 min. In the article of JMBT Alopecia, we have defined membrane proteins as having the LipPhil values larger than 0.55.
purE 13min; median LipPhil 0.598
gal 17min; median LipPhil 0.565
lac 10min; median LipPhil 0.553
his 38min ; median LipPhil 0.549
thyA 55min; median LipPhil 0.541
The lambda attachment site attB is present among the hydrophobic protein region (above 0.550 of LipPhil). The hydrophobic region is relatively rich in Thymine (T) residues. Therefore, DNA of lambda phage may be easier to attach and to integrate at 17 min.
Thus, mutated rodA- gene insertion simultaneously induces the UV resistance (reducing T-dimer formation) (please see file; P1kc RodA E. coli).
Therefore, DNA integration into E. coli chromosome is easier among the hydrophobic (T rich) region. Then, recombinant protein is easier to express and obtain in these hydrophobic-protein region. Thus, genes of hydrophobic membrane protens are easier to integrate and to express. Therefore, it is important to purify and to get the gene products of infected-hydrophobic- membrane protens should be done more hydrophobic and direct conditions with high concentrations of non-ionic detergent, since His-tag (hydrophilic tag) method becomes invalid for purification of membrane proteins using buffers containing non-ionic detergent.
In my experience, when you put an eucaryotic protein into bacteria and vice versa, expression levels usually are between very low and not detectable, due to codon usage issues. Of course, you need to get the contruct right (i.e. Shine-Delgarno vs. Kozak consensus sequence etc.).
For codon optimization, there are nice online tools available (e.g. from GeneArt/Invitrogen or who they currently are) which allow taking account of restriction sites etc.
It also simply might be that the protein of interest is toxic to the host, contains cryptic protease sites, is degraded for no known reason (you may extend the list of potential nightmares at ad libitum (and ad infinitum) yourself).
Find a friend or colleague who is experienced in recombinant expression to troubleshoot your project or simply outsource it (then you have somebody to blame :) . .
- Badges
- Science method