Yes, it has to be in frame with pET28b's own start codon, otherwise you will have no or very low expression.

Hmmm, I'm not quite sure if I understand what you mean, but I'll try to help anyway ...

Since you're insert has its own fusion tag, I'm guessing you want to remove the His-tag from the pET28 vector, so your best option is to use the Nco1 site on the pET28 vector. This means your insert should have an Nco1 site at the 5' end. Since ATG is part of the Nco1 sequence, there's no way your insert can be out of frame with the pET28 vector.

If you were planning on using the His-tag from the pET28 vector to add a His-tag to your fusion tag, then you will have to use one of the restriction sites that lie after the Nco1 site and make sure your insert is in frame with that site.

You can't use the sites after the Nco1 site if you don't want the His-tag, as either it will be added to your insert if it is in frame or your insert  will not be expressed if it is out of frame with the pET28 vector's His-tag's ATG.

I hope that wasn't too confusing.

Thanks for the reply....I did got the point

Your question is a little unclear.  It sounds to me like you are removing the start codon and all tags from the vector, keeping the promoter, ribosome binding site and terminator.  You will then put your insert between the ribosome binding site and the terminator; your insert will provide the start codon, CDS of interest, and any tags you want to have.  If this is the case, it shouldn't matter where your start codon is as long as it is the only one.  However, in order to get a different frame it looks like you'd have to use the XbaI site which would also remove the rbs (ribosome binding or Shine-Dalgarno) sequence.  In this case, to get efficient translation, you'd want to add the Shine-Dlagarno sequence generally about 8 bp before your start codon-  http://en.wikipedia.org/wiki/Shine-Dalgarno_sequence.  Does that help?

Hi Amanda,

No, that wasnt my question....Mine is for example, you have two ATGs but incase if both are out of frame....second ATG is not in frame with first ATG....but second ATG has a long transcript along with fusion tags everything in place....then will the second ATG full transcript gets translated, or it wont gets translated.....

In mammalian expression vectors, I have tried having ATG along with my cDNA sequence and never bothered to look whether ATG is in frame with the ATG before MCS. My cDNA ATG will have kozak sequence and perfectly it gets translated...but for prokaryotic vector, i am not sure how it works....having two ATGs if they are not in frame will affect translation or its just that the longest transcript without any stop codon, gets translated and is stabilized.... am i making my question clear

So, your question is:

What happens in prokaryotes if you have two different start codons with different reading frames in your vector such that the first results in a small peptide and the second results in a long protein product.

I would say in general, don't have alternative start codons.  That is just adding extra complexity to your system that you don't need.  If you want to add your own tags anyway, use the NcoI site to clone into as Antonio suggested so you will have a single start codon.  It is easy to add the restriction site on the end by PCR if you don't have the whole thing.

If you have a second start codon, it looks like it would be pretty far from the ATG in the vector (from the MCS map).  If you do not also introduce an rbs, I suspect it would be too far for efficient translation.  If they both have an appropriately spaced rbs, I'm not really sure how they would compete for translation machinery.  I would repost your question with a figure attatched showing exactly how your insert would be cloned into the MCS annotated with all rbs and start codons.  Then ask how prokaryotes will deal with the alternative translation start sites as shown in the figure.  Sorry I'm not an expert in this particular aspect!