Goal: To quantify the HIV-1 reverse transcriptase (RT) activity for a number of mutated RTs.


  1. Put RT into a vector in E.coli.

  2. Deliver GFP mRNA to E.coli.

  3. RT will process the mRNA.

  4. The resulting DNA will be integrated into E. coli.

  5. RT activity correlates with the GFP we observe.

First question: Does this logic make sense?

Second question: An alternative procedure?

  • 1
    $\begingroup$ Wouldn't it be far, far easier to do the whole experiment in vitro instead of inside E. coli? You're introducing a whole lot of problems in your plan, and I'm having a had time understanding how you would get a quantitative result from it. You might want to add more detail about your specific goal. $\endgroup$ Apr 17, 2012 at 15:07

3 Answers 3


the devil is in the details here - the logic is okay, but there is no experiment here. Too many unanswered questions.

Are you doing anything specific to see optimize RT activity?

How are you preventing your GFP mRNA from being turned into protein, producing much larger signal than you will ever see from RT activity?

Once its integrated into the chromosome, how will this gene be expressed when it has been suppressed in all these other stages?

How do you know that the plasmid isn't just being incorporated into the chromosome by some other means?

Where is the RT coming from ? E coli doesn't have the gene. this experiment is very likely to give no results.

When/how does this RT work? If its some sort of phage RT ( i don't know if there are such things) any sort of viral RT will probably kill the cell the way normally it would.

so many questions before this could work - need to do some homework here.

  • $\begingroup$ Thanks for the reply. (1) I should have mentioned that it is RT from HIV-1. (2) I would like to quantify the RT activity in a number of conformations of RT (mutant RTs).(3) I did not quite get why it may not be expressed after the integration? (4) I do not have any idea how frequent the incorporation into the plasmid by other means occurs - any estimate? (5) E.coli uptake of the RT is indeed an issue - but I suppose E.coli can handle the viral RT? $\endgroup$
    – Shinta
    Apr 17, 2012 at 16:59
  • $\begingroup$ just because the GFP gene sequence is in the chromosome, does not mean it will be copied into mRNA. You need to have a RNA polymerase binding site nearby upstream of the gene and an intact shine delgarno sequence 8 bp up from a start codon. how will you get the HIV RT into the e coli? expressed from a plasmid? HIV cannot infect e coli - they will just sit in solution until the coli eats it, which probably wont do for your experiment. Even if it did get inside the e coli, I doubt it would be active in a bacteria at all.... lots of little details to add to your question - feel free to edit. $\endgroup$
    – shigeta
    Apr 17, 2012 at 17:34

I'm with @Mad Scientist. Save yourself the trouble and do this in an in vitro system. At the moment, you don't even know if the HIV-RT will express or will be active in E .coli.

Rather, express mutated HIV-RT in T-Cells or Rabbit Lysate. Purify.

Add in a RNA template with the appropriate primer.

Hydrolyze your RNA

Detect your cDNA.


I like that you are considering a simple and standardized approach. However, there are a number of potential pitfalls to consider with this proposal. First of all, what specific RT activity are you trying to quantify? Is the the transcription speed or accuracy, or are you trying to characterize something else entirely? If you are measuring speed, how and what will you measure, reliably and will be reproducible, for speed? If you are going for accuracy, you will need a whole lot of sequencing to check the bases are correct. Both of these can quickly become expensive.

Also, is it necessary to use HIV-1 RT? Since you mention E. coli, would you want to consider using msr, a prokaryotic RT?

Though I've never heard or tried it myself, is it possible to transform mRNA into bacteria? I would imagine that it would be degraded as a natural defense mechanism. You may need to consider some alternative form of gene delivery, such as a standard expression plasmid. Genomic integration may not be necessary to achieve your goal.

As mentioned, you cannot just have naked cDNA to translate a protein. You need at least the promoter, RNA pol binding site and polyA signal. You could take advantage of viral infection systems (a replication defective HIV) to deliver your GFP and RT genes into a mammalian system (like HEK-293 cells), but this adds a lot of complexity and expense that may not be worthwhile.

The last thing off the top, you would benefit greatest from taking advantage of a directed evolution experiment to generate some potential leads. For example, you may PCR amplify the RT cDNA with an error-prone Taq to purpose create RT mutants for use in your assay.

Keep thinking about this, and maybe check the websites of the some of the RT commercial vendors. It is possible one of these companies have already done something similar.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .