Dear Colleagues,
Thank you very much for the responses regarding my query about the best probe for metabolomics, some excellent information was provided and I would like to thank the following responders: Christopher Rithner, Martha Morton, Stephen Huhn, Jerry Dallas, Ryan McKay, Mark Swanson, and Matthew Devany. Some responses were highly confidential, and I am careful to respect this. However, if the information was freely available without prompting from Bruker, then I am able to provide it.
Considerations are obviously both cost and cost-effectiveness, running costs and maintenance demands, absolute sensitivity, spectral quality and performance (baseline roll and other artifacts, excitation profile, etc.), application to other projects (including available temperature ranges), and versatility (changing probes). Obviously inverse probes for greater 1H sensitivity are favoured (assuming most are after 1H detection and that 1H sensitivity is an issue, which is probably the majority of cases) though there might be instances when a cryoprobe of normal configuration (i.e. the X-nuclei coil as the inner coil) might be a consideration if it provides sufficient 1H sensitivity for the work at hand but also excellent X-nuclei sensitivity for, say, 13C-based metabolomics or for other project work. Even a room-temperature Bruker SmartProbe is applicable under such circumstances. Exact comparisons of probe sensitivities is very difficult due to different configurations, probe design, and probe sizes (especially without clear indications of sample volumes) and a sensitivity for a Bruker nitrogen-cooled inverse probe was not available. Generally though, I took sensitivity gains from cryoprobes to be approximately 2-2.5 for a nitrogen-cooled probe and approximately 3.5-4 for a helium-cooled probe.
In terms of sensitivity, generally the best option is a helium-cooled inverse probe, and also evidently on a cost-effectiveness basis. However, probe changes involving a helium-cooled probe are quite problematic and costly (warming up the probe reduces the time between expensive scheduled maintenance overhauls). Thus the problems and added cost of changing probes means that a helium-cooled probe may not be the best option even if money is not an issue if versatility is demanded. Time gained can easily be lost with changeovers, though available NMR time/access may not be such a big issue (as opposed to getting spectra quickly due to user impatience or sample degradation). As well, temperature ranges are limited. There is also the question of whether the sensitivity advantage of a helium-cooled probe can be maintained with high salt concentrations since sensitivity drops off quicker with high Q-factor coils and if the stability of a helium-cooled probe is sufficient for metabolomics work. But these are issues to look out for and to obtain measurements on real samples before proceeding with a purchase.
Bruker are not prepared to recommend their nitrogen-cooled inverse probe for metabolomics work (one user cautioned that it is better to follow a manufacturer’s advice unless you are really confident), citing problems such as uniform excitation and flat baseline issues which are important for metabolomics work, though it is unclear why these issues are specific to their nitrogen-cooled probe and not also present on their other probes. This was insinuated by one responder with respect to the baseline issue. I suppose it is important how the data are handled - with manual or automatic baseline correction, how the particular approach handles the particular baseline aberration, etc. - whether this is a big problem or not for users and how much additional processing they are prepared to put up with. Most users that I am generally aware of are highly satisfied with their nitrogen-cooled probes and they seem to have a very good general reputation. One responder was quite surprised that a nitrogen-cooled probe was not recommended for metabolomics work as they were extremely happy with the performance of their probe, and though they were not actually doing metabolomics work, they had no problems with, for example, solvent suppression. Probe changes involving a nitrogen-cooled probe are not nearly as problematic as helium-cooled ones though it is still preferable not to warm up the probe if avoidable. Like helium-cooled probes, temperature ranges are also limited. There is also the question of whether the sensitivity advantage of a nitrogen-cooled probe can be maintained with high salt concentrations and if the stability of a nitrogen-cooled probe is sufficient for metabolomics work. But these are issues, in addition to the aforementioned ones, to look out for and to obtain measurements on real samples before proceeding with a purchase.
In terms of cost, a nitrogen-cooled probe can be 2/3 the cost of a helium-cooled probe (including platforms), or 2-3 times the cost of a comparable room-temperature probe, so the cost-effectiveness is not high for nitrogen-cooled probes. Running costs and maintenance demands are much less (ca. 1/4) for a nitrogen-cooled probe consisting of essentially ca. 80 L of liquid nitrogen per week and a maintenance overhaul of the system every two years (under 10k in cost). It has been stated by some users that the maintenance demands of a nitrogen-cooled probe are intensive, though this opinion is in the minority and may not be a concern at all for users depending on service contracts taken up.
Though gains can be made from smaller sized probes, the impediments can become counter-productive. Most metabolomics work is conducted on 5-mm or 3-mm probes, though some is also conducted using micro-probes (e.g. 1.7-mm, or even 1.0-mm probes). The gain in sensitivity for reducing to a 3-mm probe from a 5-mm probe is only about 30% on a mass consistent basis, though to a 1.7-mm probe this can amount to a 4-fold increase in sensitivity on a mass consistent basis. But there are obstacles and limitations with smaller sized probes:
1) If the sample volume is initially limited, comparable to the tube volume, a smaller probe may be a good choice. For large available volumes, sample concentration is required to take advantage of the sensitivity gain. This incurs additional sample preparation work, as well the possibility of sample alteration, loss of volatile components, and an increase in the salt concentration of the sample.
2) For 1.7-mm and 1.0-mm tubes, disposable tubes incur an additional cost and tapered tubes are expensive (the stem - the portion of the tube where the sample resides - is micro-sized while the top part is 5 mm so that a normal 5-mm spinner can be used). Sample handling and preparation and manual filling of the tubes for small volumes is more burdensome and may lead to demands for a sophisticated robotic filling system/sample changer. Similarly, the cleaning of tubes also presents additional work considerations if disposable tubes are not used.
3) Use of an added internal standard for the volumes of 1.7-mm and 1.0-mm tubes may present problems in terms of quantification errors which need to be assessed. The best solution in cases of very small available volumes may simply be to dilute the sample so that 5-mm or 3-mm probes are used in these situations.
4) Both 5-mm and 3-mm probes readily lend themselves to general use. Micro-probes much less so, especially for the handling of volatile organic solvents where again a sophisticated robotic filling system/sample changer might be required.
For metabolomics only concerned work therefore, the best probe size is essentially governed by the sample volumes that are available from the studies, assuming one wants to avoid additional sample preparation work together with the inherent risks of altering the sample for available large volumes by concentrating the sample, not to mention the loss of volatiles from the samples in addition to increasing the salt concentration. For example, for large volumes (e.g. human urine), a 5-mm probe is best, for medium volumes (e.g. cell culture media), a 3- or even a 1.7-mm probe is likely to be preferred while for miniscule volumes, a micro-probe is best. But while large volumes are not so amenable to scaling down to get the sensitivity gain with the involved concentration work and the cost of at least the loss of the volatile components, small volumes do scale up as they can be diluted, Shigemi tubes etc. can be used, or small tubes can simply be used in a larger probe.
Clearly, a helium-cooled probe, if affordable and there is not a demand for versatility, is preferable over a nitrogen-cooled probe (if willing to take the risk or put up with extra demands on processing), which is preferable to a room-temperature probe. However, there are plenty of users who recommend room-temperature probes as the best option in particular circumstances (cost, versatility, lossy samples, etc.). The final recommendation is, apart from the nitrogen-cooled inverse probe pretty much in line with Bruker’s recommendation: a 5/3-mm helium-cooled inverse probe or 5/3-mm room-temperature BBI or TXI probes (inverse broad-band and dual 13C/15N probes, respectively). However, though there is some concern about getting a nitrogen-cooled probe for metabolomics work, it may be a viable option for some facilities. It is probable that we will go with the nitrogen-cooled inverse probe (TCI Prodigy, available only in a 5-mm size) although it is not recommended by Bruker for metabolomics work. The cost is closer to our budget (initial cost + running costs) and versatility is needed since we need to change probes intermittently. For our methodology, we are not intimidated by problems with baseline issues as we deal with them quite well in our processing, at least presently. If we do not go for the TCI Prodigy probe, then the 3-mm TXI probe (slightly better 1H sensitivity than the BBI and we only rarely need other than 13C and 15N which we can do elsewhere) without 19F (again, slightly better 1H sensitivity than with and which we can also easily do elsewhere) will be our other option.
Again, many thanks to the responders for their valuable comments.
Regards,
Karel Klika
Received on Mon Sep 03 2018 - 05:15:26 MST