We have used Nanodrop for several years and found the readings are always accurate. The highest concentration we have measured is around 30 mg/ml. The differences between diluted and concentrated samples are within dilution error. Nanodrop spectra at low concentration are noisier. We actually prefer to have protein concentrations over 0.5 mg/ml.
The key to get good Nanodrop reading is to have a nice liquid column. Take a look whiling waiting for the results. High concentration samples with detergent or glycerol may produce bubbles (giving an error message sometimes) or improperly formed liquid column (giving incorrect numbers). Use more samples (up to 5 ul) and gently push down the head will help the liquid column formation. 1 ul is not enough for most protein samples. If the reading is not good, wipe clean the surface and load a new sample. Repeated reading of same sample gave huge variation. The liquid column probably has difficulty to form properly after the first reading. We provide protein purification contract services and send proteins to many users around the globe. The proteins have been used for crystallography, ITC, DSC, Biacore, and activity assays. So far we havent heard any complain about our concentration measurement using Nanodrop. We also use Bradford to measure concentration if required by our customers or when concentration cant be measured by A280. The error range for Bradford is much wider and easily off by a factor of 2 (by repeating or comparing to Nanodrop). There are different brands of Bradford reagent. Each gives a different number. We once had a big discrepancy with a customer. It turns out the home-made BSA standard lot was off. After adjusting the BSA concentration with Nanodrop and using the same Bradford reagent and same curve fitting method, the concentrations agreed. The lesson leaned is to have good fresh standard solution. The color of Bradford changes over time, its better to use a plate reader than single sample spectrometer. Estimation by eye is just as good as a spectrometer:-) --Chun _____ From: CCP4 bulletin board [mailto:[EMAIL PROTECTED] On Behalf Of Juergen Bosch Sent: Saturday, December 06, 2008 12:29 AM To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] AW: [ccp4bb] suggestions for UV spectrometer Hi, really strange, I always dilute my protein when taking an absorption spectra. I try to adjust my expected concentration to a readout of ~ 0.2-0.3 OD280. And the Bradford is just as well as 'picking house numbers', depending on your protein, you can underestimate your protein concentration by a factor of 2. For my current approach I use a 100 µl volume cuvette and 2 µl of my concentrated protein (most of the times). The Structural Genomics lab (MSGPP) has a Nanodrop and I was also always happy with those results (but again I dilute my protein sample before taking a measurement) Jürgen On 5 Dec 2008, at 16:00, wangsa tirta ismaya wrote: Dear all, Thanks a lot for raising the issue with the not reproducibility of protein measurement with Nanodrop. We use the instrument as a workhorse in the lab. Indeed, recently I observed that the protein concentration suggested by Nanodrop is sometimes differ to the usual colorimetric measurement (Bradford method, measured with our Pharmacia's Ultrospec 2000 spectrometer). Since the cell in Nanodrop is very small, could it be due to the homogeneity of the sample in the cell? Also what I have observed, we have to be sure that the cell is cleaned properly before use. Another thing is, at high protein concentration I obtained noisy absorption curve at the peak (like a seismograf ....) thus the protein concentration measured is doubtful, I have to dilute the sample to have good curve (thank God it requires only 2 mikroliter for a measurement). Well, I think nanodrop is a good, fast, and powerful instrument, however it would be better if we established a reference of our daily practice to a normal spectrometer measurement. cheers, Wangsa 2008/12/5 Martin Hallberg <[EMAIL PROTECTED]> Which brings us back to the Hellma "TrayCell" solution where you can, from the same spectrometer, have both the cuvette option and the quickness of the NanoDrop/NanoVue system. Anyone that can comment on the performance of the TrayCell from Hellma? Cheers, Martin On Dec 5, 2008, at 9:06 AM, Gregor Witte wrote: Agree! I think for crystallographic use the nanodrop is perfectly okay to see if the protein is 5mg/ml or 30mg/ml. But in fact I also do not trust our instrument if it comes to more important issues like preparing solutions for titrations or assays. And due to the small pathlength I do not trust absorptions of small concentrated samples at all. I always prefer a "real" 2-beam spectrophotometer (monochromators) with a quarz-cuvette and a nice pathlength. Of course, you cannot reliably measure solutions exceeding Abs 1 or maybe 1.5 OD in a spectrophotometer with 1cm pathlength. There's also one quite strange thing about the nanodrop they sell the "calibration check solution" (which is some kind of yellow chromate-solution with known absorption), then you check your nanodrop with it and maybe find out that it's off to some certain extend: But then you're stuck(!), because you cannot calibrate it on your own. I guess it would be quite easy to integrate a calibration-option into the software, but at the moment the instrument tells you "calibration failure" and you have to call the service guys who then carry it home and calibrate it by turning of the two small screws at the top of it and then glue them with locktite. Anyway, at least for our mid to high concentrated samples the nanodrop is not showing large fluctuations so we are happy with it. But everyone using a nanodrop should check it from time to time as I found out that ours was off more than 20% at one day - which raised some trouble of course Cheers, Gregor Von: CCP4 bulletin board [mailto:[EMAIL PROTECTED] Im Auftrag von Filip Van Petegem Gesendet: Donnerstag, 4. Dezember 2008 22:20 An: CCP4BB@JISCMAIL.AC.UK Betreff: Re: [ccp4bb] suggestions for UV spectrometer I want to add I absotely hate the nanodrop. We've had a demo for it, and found the readouts to be very unreliable. Fluctuations of 20% and more. Just leaving the same drop in and measuring the sample multiple times gives different values (going in both directions, so not only due to evaportations). Sure, it's easy and fast, and maybe good to have a rough idea about your protein concentration, but I would never want to use it for exact measurements such as needed for e.g. a CD or an ITC instrument. I've heard other labs in our department have similar issues. We've also had a demo for the Nanovue from GE Healthcare: same issues - very large fluctuations from one sample to another. I suppose this is simply an inherent problem with small volumes... Cheers Filip Van Petegm On Thu, Dec 4, 2008 at 12:48 PM, Patrick Loll <[EMAIL PROTECTED]> wrote: At the risk of dragging this discussion even further afield from crystallography: How can you get realistic numbers for concentrated solutions using the Nanodrop? I understand that the instrument reduces absorbance by using a very short path length. However, I thought that in order for the Beer-Lambert formalism to be applicable, the solution needs to be sufficiently dilute so that the chance of molecules "shadowing" one another is negligible. Isn't this condition violated for concentrated solutions (even with short path lengths)? Pat On 4 Dec 2008, at 1:27 PM, Michael Giffin wrote: We also like the Nanodrop... ---------------------------------------------------------------------------- ----------- Patrick J. Loll, Ph. D. Professor of Biochemistry & Molecular Biology Director, Biochemistry Graduate Program Drexel University College of Medicine Room 10-102 New College Building 245 N. 15th St., Mailstop 497 Philadelphia, PA 19102-1192 USA (215) 762-7706 [EMAIL PROTECTED] -- Filip Van Petegem, PhD Assistant Professor The University of British Columbia Dept. of Biochemistry and Molecular Biology 2350 Health Sciences Mall - Rm 2.356 Vancouver, V6T 1Z3 phone: +1 604 827 4267 email: [EMAIL PROTECTED] http://crg.ubc.ca/VanPetegem/ . B. Martin Hallberg, PhD Assistant professor Department of Cell and Molecular Biology Medical Nobel Institute Karolinska Institutet Von Eulersv. 1 SE-171 77 Stockholm Sweden Fax: +46-8-339380 -- Wangsa Tirta Ismaya ===================================== Josef-Israelstraat 66, 9718 GN Groningen The Netherlands - Jürgen Bosch University of Washington Dept. of Biochemistry, K-426 1705 NE Pacific Street Seattle, WA 98195 Box 357742 Phone: +1-206-616-4510 FAX: +1-206-685-7002 Web: http://faculty.washington.edu/jbosch
