Dear Arabidopsis news group.
The email questions below were forwarded onto me as we have been involved in
this area for some time now. As well as replying to Jon individually, I
thought the answers might be of interest to the news group:
We have been working in the area of dsRNA/ hairpin RNA / RNAi for some time
now. You might be interested to look at one of our first papers (
Waterhouse, P.M., Graham, M.W., Wang, M.-B. (1998). Virus resistance and
gene silencing in plants is induced by double-stranded RNA. Proc. Nat. Acad.
Sci USA. 95 13959-13964.) which describes how we worked out that dsRNA was
the inducer of PTGS/RNAi in plants and provides the design rules (such as
hairpins) for silencing genes - such as adopted in the Chuang and
Meyerowitz, 2000, PNAS 97:4985 paper.
Our latest paper (Smith, N.A., Singh, S.P.,Wang, M-B., Stoutjesdijk, P.,
Green, A. and Waterhouse, P.M. 2000. Total silencing by intron-spliced
hairpin RNAs. Nature 407 319-320.) shows how we have further improved the
efficiency of obtaining silenced plants using intron-spliced hairpin
constructs. Although there was not enough room in the paper to describe all
the results, we have now silenced over 10 different genes in a wide range of
species - from tobacco and Arabidopsis to barley and rice. We have made a
generic vector that we call pHANNIBAL that is basically a gene expression
vector containing an intron flanked by a small polylinker at each end. This
allows you to PCR a piece of your gene (we usually use 300-500nt) with the
appropriate restriction sites such that the amplified piece can be
directionally cloned in a sense orientation at one polylinker site and in an
antisense orientation at the other polylinker site. This vector has worked
really well for us giving approx 100% of transformants with silencing and
the silencing is generally much more profound than that obtained using
antisense or co-suppression. If you are interested in using this vector -and
associated binary please email me.
>>Dear News group:
>>We're in the process of making RNAi plasmids for our genes of
>>interest and have some questions:
>>>>1. does the orientation of the sense/antisense fragments matter?
>> <---xxxx---> vs. --->xxxx<---
NO IT DOESN'T SEEM MATTER - WE HAVE USED BOTH WAYS AND THEY BOTH WORK WELL.
ALTHOUGH WE GENERALLY USE SENSE FIRST AND ANTISENSE LAST.
>>>>2. does the size of the linker (xxxx) matter?
YES THE SIZE OF THE LINKER REQUIRED DEPENDS ON THE SIZE OF THE ARMS. IN
ORDER TO HAVE STABILITY IN BACTERIA THE RULE OF THUMB SEEMS TO BE USE A
LINKER AT LEAST AS LARGE AS THE SIZE OF AN ARM. WE ROUTINELY USE AN SPACER
OF 700-800 BASES AND ARMS OF 300-500.
>>>>3. What should the linker be? GUS has been used (Chuang and
>>Meyerowitz, 2000, PNAS 97:4985 PubMed #10781109) but could it be from
>>the gene of interest in either orientation?
THE LINKER CAN BE DERIVED FROM THE GENE ITSELF OR GUS OR AN INTRON. IN FACT,
WE FIND THAT AN INTRON WORKS VERY WELL INDEED. THIS INCREASES THE EFFICIENCY
SUCH THAT ALMOST EVERY INDEPENDENT TRANSFORMANT SHOWS SILENCING.
>>>>Thanks for your help.
Hope this is of help
Dr Peter Waterhouse
Senior Principal Research Scientist
CSIRO Plant Industry
peterw at pi.csiro.au
Ph (61) 26 2465365
Fax (61) 26 2465000