Re: ProtSeq: HP 241 vs Procise cLC
Heinz Nika (hnika@pacbell.net)
Sat, 31 Oct 1998 01:20:08 -0800
Erland Holmberg wrote:
>
> Thanks for the comments on this matter!
>
> (Ken, additional information is listed at the end)
>
> Our interest is solely in the high-sensitivity N-terminals sequencing.
> For the HP machine we would Never use the C-terminal chemistry, in order
> to keep the sensitivity for the N-terminal part.
>
> Unfortunately we haven't got more results from Perkin Elmer. But since
> cLC seems to be the well recognized machine for low sensitivity samples
> we havent pushed them so hard. HP has of course been very eagerly to
> provide data as they see a chance to match or at least compete for the
> cLC. And their results, are fairley good, and are provided with a machine
> equipped with a 2 mm column and a flow cell with less volume.
>
> With our HP G1005A we can sequence samples down to 2 pmol (10 cycles). We
> have the 1100 system as on-line system for LC-MS so we know what (and how
> to do it) it can do in separation at low levels. As I wondered
> previously: what if You provide the HP with a more narrow column (and/or
> a shorter one?) and smaller volume flowcell? HP claim that a shorter
> column (12.5 cm) would double the signal.
>
> Thus, the results from HP make us puzzled. Before these results, we were
> certain that we would end up buying the cLC. HP calls signals at 40 fmol
> (20 cycles) for a 750 fmol samples.
>
> What are the amounts (as injected) people have sequenced on either
> machines? I expect people to have more experience on low amoumnts for the
> cLC. But has anyone really tried the 241 yet for low amounts?
>
> I have to admit that our first impression about the levels beeing similar
> might be misleading since those samples seuqenced were at a rather "high"
> level - 2.5 to 5 pmole (see below). Joseph Fernandes has previously
> reported that HP definitely sequences further at these levels, but the
> cLC is the choice for the really low amounts. Thus, this might be an
> important matter - maybe the matter that makes the choice.
>
> The problem I have is to make a decision rather urgently.Should I go for
> an instrument with a, so far, more proven record - cLC - or invest in the
> "joker" 241 and go for improvments in columns/ flowcell etc. A furhter
> matter is that we are rather experienced in the HP instrumentation.
>
> Erland Holmberg
>
> -----------------------------------------------------------
> cLC HP
> flow rate 40 ul/min 300
> path lengt 6 mm 6
> wl 269 269
>
> Correction: run on a vWL detector in both case! HP is used to come with a
> DAD since a semi-micro cell was not available for the wWL before.
>
> We provided two proteins: human growth hormone and apolipoprtein A1
>
> Results we have obtained so far are:
> cLC
> 5 pmol PVDF blots cut in halves (2.5 pmol) from both proteins:
> Relatively high background
>
> HP
> 5 pmol PVDF blotsfrom both proteins:
> Low background
> High R.Y.
> (Double singnal as compared to cLC - as it should be 5 vs 2.5 pmole)
> 2 pmol sample on biphasic column
> I.Y. 0.9 (45%) RY (97.2)
> all aa's clearly identified
>
> 0.75 pmol sample om biphasic column:
> I.Y. 0.65 (87%) RY (95.4) ambigous results for 4 aa's where experienced
> and manual interpretation were needed
> Aa's were called at 40 fmoles.
>
> ----------------------------------
> Erland Holmberg, PhD
> Manager, Protein Identification
> Research, Pharmacia & Upjohn, Sweden
> Tel: +46 8 695 9094
> Mobile +46 8 817 0619
> Fax: +46 8 695 7640
>
> e-mail:erland.holmberg@eu.pnu.com
> ----------------------------------
Dear Erland:
The results you described regarding blots are also obtainable on the
G1005A sequencer using method PVDF 3.1 since the presence of the strong
anion exchange column (SAX) effectively prevents sample washout. When
developing this method, I found it essential to sonicate the blots for
10 min in water prior to insertion into the upper, empty sequencing
column. This simple procedure that had been incorporated in a HP
technical note provides for exceptionally clean backrounds and helped to
unambiguously interpret the first two or three signals.
The data you described for liquid samples sequenced on the bi-phasic
column are certainly impressive. With the G1005A/1090 analyser I
considered 1 picomole of protein as the lowest sequencable amount
provided that the material was applied onto the RP column in 100 ul of
SLS or less to avoid sample adsorption to the loading funnel as
described in a poster presentation (ABRF 96, San Diego). With Beta-lac
at that level, 25 residues could be unambiguously identified,
the 25th residue alanine at the 200 fmole level. What is it that enables
the G1005A/1100 combo to identify signals at the 40 fmole level? I would
really like to know the technical issues behind this improvement.
Heinz Nika
hnika@pacbell.net