AN OPEN LETTER TO THE PFGE USERS OF CAMPYNET: ISSUES FOR CONSIDERATION AND DEBATE AT THE CAMPYNET MEETING 23-24 SEPTEMBER 2000
From: Stephen On, PFGE Subgroup leader
As you know, a prototype standard method for PFGE typing of. C. jejuni and C coli was made publicly available on the Website on July 7th. I know some of you are already trying the method and look forward to hearing your views. This open letter is intended to raise specific points for discussion at the meeting, in order to help in building a secure, standardised PFGE system that allows the best possible comparability between laboratories.
To build a "definitive" DNA-fragment based typing system, two key criteria must be fulfilled.
The recommendations are based on the basic remit to the PFGE Subgroup: to devise a PFGE method for typing both C. jejuni and C coli. This was not going to be easy from the outset, since some (but not all) C coli strains yield far more complex patterns than those of C. jejuni and with a size distribution that is also rather different (many low MW bands). The running conditions suggested do a good job of resolving SmaI-based PFGE profiles of both species. However, some C coli profiles remain highly complex and it can be difficult - to say the least - to identify discrete bands unequivocally. Two photographs of profiles derived from the "Campynet strain set" are shown below to illustrate my point.
Fig. 1. PFGE analysis of Campynet strains. Lane 22 is a C. coli strain. The potential problems of high MW bands in C. jejuni strains in lanes 6-9 are discussed below.
Fig. 2. PFGE analysis of Campynet strains. Lane 8 is a C. coli strain. The potential problems of high MW bands in the C. jejuni strain in lane 3 are discussed below.The standard for gel-to-gel normalisation was selected with the initial remit in mind - i.e. "definitive" PFGE typing of both C. jejuni and C coli. Bearing in mind the rather wide molecular size range to be covered, it was a challenge to identify an ideal standard. I personally examined various digests of various Campylobacter species, and also of the standards used by the CDC for their "Pulsenet" standard typing system for E. coli and Salmonella. Frankly - nothing I came across appeared to be a "perfect" standard in my experience.
The suggested standard - CNET 068 - is good but it is a compromise. It covers most of the molecular size range exhibited by C. jejuni and C coli profiles but definitive typing of some strains - i.e. accurate matching of any field strain pattern with a named reference type - will be difficult and probably impossible to attain with certainty. I estimate that about 18% of the Campynet reference strain set would probably be problematic in that respect. That would include some C. jejuni strains too, since some profiles contain a high MW band exceeding the highest marker in the present normalisation standard. See lanes 6-9 in Fig. 1 and lane 3 in Fig. 2 for examples. I do feel it likely that the position of such high MW bands in these C. jejuni profiles might not pose a problem at all, and may serve as "anchor" bands of invariant mobility. That does, however, remain to be proven.
However: if the present running conditions are maintained, a better reference standard could be chosen that would optimise comparison of all C. jejuni strains, and cover some C coli. Profiles of all strains would of course be resolved and could always be compared by eye. The difference is that I believe we could come much closer to definitive PFGE typing of all C. jejuni strains.
In summary: at present, I believe we have two options:
The Subgroup and I respond to the needs of its users. We need to know which of the above options you would prefer. I urge you to consider these issues for discussion at the meeting. If you are not attending the meeting, or are not an official participant, then you are invited to send your views to me by E-mail. Click here.
I look forward to your comments.
With best wishes, Stephen