Usually different lengths could be due to a few reasons.

Long strands can break, but short ones usually don't.

How specific was the primer, so how long was it? Did it contain mostly AT's or mostly GC's? (GC's are stronger, but less selective for a primer)

Was there a contamination in the E.coli. it may have had a SnP in it's structure, or a part of the culture had this.

It could have had a separate stop site, or a place where the TAQ was kicked off.

The ones in the band are longer, not shorter, so unlikely it broke.

Hi Walton,

I'm not particularly expert in prokaryotic biology, but of course bacteria are a good tool in the lab, so some basic knowledge has come to my mind while reading your post. I'll submit a few observations for you to think about and decide consequently:

- prokaryotic mRNA is very shortlived, average lifetime is about 2 min (makes sense for a cell that replicates in 20 min), so you wouldn't expect to extract 100% intact mRNA, however accurate you are, and I would expect heterogeneity of 5' RACE fragments. To make sure this is the case (and not that your primers are unspecifically amplifying different mRNAs) I would sequence and align all your PCR fragments: if they belong to the same mRNA they should align well but terminate earlier or later. You could take the longest fragment as in principle the most faithful of the 5' UTR;

- prokaryotic mRNA is not post-transcriptionally modified, meaning there is no CAP and no polyA, besides lacking splicing, so your RACE approach is fine;

- prokaryotic translation requires an internal sequence on the mRNA (since there is no CAP and often there are more independent translation start sites on the same mRNA, polycistronic mRNAs), the Shine-Dalgarno (SD) sequence, located only 8 nt upstream of each ATG, so 5' UTRs of bacterial mRNAs are not expected to be long, probably not much longer than the 30 nt you managed to amplify (compare how bacterial expression is constructed in plasmids to confer antibiotic resistance or LacZ expression for the white/blue selection: you'll see that the -35 and -10 promoter elements are pretty close to the ATG, maybe some 40-50 nt upstream of it, meaning that the 5' is quite short.

Thanks, guys, really appreciate your info Thom Claessen Pietro Pilo Boyl

I aligned all the sequences obtained, and did find a TSS, which has conserved -10 and -35 regions. And it is correct that usually the TSS is not far from ATG, possibly 20-50bp upstream most of time.