Using the right tool – Or how incredibly bad I am at pipetting

One elements we want measure in seaweed and foods is Iodine. It seems that the way to determine Iodine content is by the use of the Sadell -Kolthoff reaction. The basis of this is the reduction of Cerium (IV)  by As(III), according to the following:

2 Ce(IV) + As(III) -> 2 Ce(III) +  As(V)      (1)

The Ce(IV) is orange and the the Ce(III) is colourless, so it can be done by UV-Vis.  But how does one use this to determine Iodine? well reaction (1) is slow,  but it can be catalysed by Iodine, the higher iodine  concentration the faster we lose the orange colour. 

Okay this method was published in 1937, and is still used it is probably as optimised as it can get, so haven been into the literature and found a method adapted for 96 well plates. There where actually several, but the chose this one as starting point. I like doing assay like these in a 96 well plates as we have very nice BMG plate reader  and one can  measure a lot of samples (well 96 I guess) in one go. And the plate reader can be run in a “slow kinetic” mode where I can measure each well once a minute, so that we get a time trace out, which should be better than just measuring after a given time, as the most assays say the did.

So time to show my first data set:

What you see here are 5 different concentration of KI, with 2 repetitions . At first glance  it doesn’t look so bad until I reveal that each colour is one concentration in two repeats and there is only one on set where they overlap. But several where different coanc oare on top of each other. 

Now I spend a lot of time to wonder about this and didn’t really understand it. Until I applied my brain. This experiment was made by pipetting the standards into the wells with a standard Finnpipete. Now, as the lowest volume was 1 microL this could be a problem (And all you analytical chemist can stop screaming  YES, now) So today I tried again, but this time I used  my SGE eVol syringe. Now this was what I found:

Except for the red curves (Where I guess there was a fibre or similar in the well).

So today I learned do not use finnpipette for high precision work (or that I suck in pipetting)

Next step her should now be to figure out how to use the plate readers build in pump to do this, and how to automate the data analysis.  

 

 

On mixing triglycerides with phsophorlipids

For the last several years (more than I actually want to think about, I think we started around 2007) I have been looking into what happens when we mix a phospholipid and a triglyceride.  As we now  have published  the fourth paper on this subject, it might be time to give a short overview.

The project started many years ago when we, a visiting Phd student and I,  where looking at lysosomes. She had isolated the lipids from them and now she wanted to make GUV (giant unilamellar vesciles) to get pretty pictures to study the lateral structure of these. But they just didn’t want to form, so we went on to do a lot of different experiments and among them was  TLC, where we found a large amount of triglycerides. Okay the might prevent the formation of liposomes, we thought, but we where so wrong. The GUV could be formed much faster and easier from a mixture of POPC and triolein (TO)  than from pure POPC. We then started to do all the thing we normally do with lipids mixture like  DSC, EPR and fluorescence spectroscopy.

During one of the EPR experiments I discovered that a mixture of 10% TO and 90% POPC separated into to different phases when they where centrifugated.  One went to the bottom of the tube, the other to the air-water interface, like this:

The lower phase we labelled HF, for heavy fraction and the upper LF for light fraction. We also found that there spectroscopic properties where different, so we ended up  writing a paper on this, which went into chemistry and physics of lipids.

In this we had TLC’s which showed that both fractions had the same amount of TO (which is why we where talking of fractions instead of phases originally) which made it hard  to explain what was going on,  but we where bold (stupid)  enough to suggest the following model:

The “driving force” here would be that TG are not that happy with having all three FA chains parallel, like in B. But we never figured out how this could explain the difference in density.

After this got out we managed to get our resident MD simulation guy to run MD simulations on a system with 2 and 5 mol% TO and he found a different picture:

In A is the results for 5 mol% and in B and C the result for 2 mol%.  The figure is from this paper. Now this starts to make a bit more sense, once we accept that the TO molecules can go into the centre of the membrane, which was a bit hard for me.  And the biological relevance also went up. This is highly relevant for the formation of lipids droplets and  or for the so-called mobile lipids in cancer cells.

In the next paper in we focused on microscopy and mechanics, showing confocal picture  and a video on the dynamics. We also showed that TG doesn’t  make the liposomes more prone to fusion, as we had speculated as mechanism for the fast and efficient formation of liposomes.  And we showed that the presence of TO lowered the bending rigidity significantly,  which could be relevant for cancer cells as they also have reduce bending rigidity.

And now the forth and probably last on only POPC TO is now online.  Here is the abstract:

Abstract

We have in this study investigated the composition, structure and spectroscopical properties of multilamellar vesicles composed of a phospolipid, POPC, and up to 10 mol% of triolein (TO), a triglyceride. We found in agreement with previous results that the mixtures with 10 mol% TO spontaneously separate into two distinct phases, heavy (HF) and light (LF), with different densities and found this also to be the case for 2 and 5 mol% TO. The compositions of the two phases were investigated by quantitative lipid mass spectrometric analysis, and with this method we found that TO had a solubility maximum of about 4 mol% in the HF, whereas it was markedly up-concentrated in the LF. Electron paramagnetic resonance spectroscopy indicated POPC membranes of all tested concentrations of TO in both phases to be almost unperturbed by the presence of TO and to exist as vesicular structures containing entrapped water. Bilayer structure of the membranes was supported by small angle X-ray scattering that showed the membranes to form a lamellar phase. Fluorescence spectroscopy with the polarity sensitive dye Nile red revealed, that the LF samples with more than 5 mol% TO contained pure TO domains. These observations are consistent withan earlier MD simulation study by us and  co-workers suggesting triglycerides to be located in lens shaped, blister-like domains between the two lipid bilayer leaflets (Khandelia et al. (2010)).

And the ToC picture:

Seen here are the Nile red spectra of o, 2, 5 and 10 mol%  TO samples. Nile red is used for detecting lipid droplets in cells as is changes it emission wavelength depending on the polarity of the surrounding medium. On the right are the spectra from the LF fraction shown and it shows pretty clearly that it contains TG domains, based on the increased intensity at  571 nm. This combined with the SAXS measurements, where we show the presence of lamellar structures   is al in all experimental evidence for the model suggested from the MD simulations.

And the pretty pictures I posted some time ago were made for that paper.

First phycology paper

So already back on the blog, but a bid before expected our first phycology paper is online (in Journal of Applied Phycology).

Here is the abstract:

The red seaweed dulse (Palmaria palmata) is one of the more popular seaweed species for human consumption in the Western world. With a documented historical use up to present days in Ireland, Brittany (France), Iceland, Maine (USA), and Nova Scotia (Canada), it has remained a snack, a food supplement, and an ingredient in various dishes. The trend towards more healthy and basic foodstuffs, together with an increasing interest among chefs for the seaweed cuisine, has posed the need for more quantitative knowledge about the chemical composition of dulse of relevance for human consumption. Here, we report on data for amino acid composition, fatty acid profile, vitamin K, iodine, kainic acid, inorganic arsenic, as well as for various heavy metals in samples from Denmark, Iceland, and Maine.

The data in this paper does nor come from our new equipment, but is done by various other groups or on equipment we could sneak samples in on. But is these kind of studies we would like to continue. One the “problems” in the paper is the fact we in many case have n=1 (sic), as it is nor so easy so get all the samples needed measured when it done via collaborations. But that we have our own HPLC and AAS, we going to look further into these compounds with n >> 1.

What I have been up to

In case there still should be a  single reader some where (probably in long forgotten RSS feed) I have now realised that my last post a bit more than a year old (sic).

The last year has seen  some changes in my job. I am still working in the same  lipid biophysics group , but I am slowly changing from doing primarily calorimetry and EPR on lipid model system to becoming more of an analytical chemist as seen by my two new “toys”.

First of we now have our own HPLC-MS. Choosing a HPLC turned out to more of a challenge than I was used to when buying instruments. For DSC there is basically two model to choose from (for what we do), but boy, there where many different vendors and models to choose from when looking for an HPLC. In the end I ended up with an Shimadzu LCMS-2020. I could come up with many argument (all taken from the vendor webpages) but in the end I chose this because or id Danish distributor, Holm & Halby, which always sell good stuff.  And they have  a very good service section.  One of the nice turned out to been the training. Include in the deal was all in all three days of training, performed by a very knowledgeable person from ALSAC  in Sweden.

And of course we had to make some choices, with detectors etc. I decided on a simple UV detector instead of PDA, but the also got an fluorescence detector and a ESLD detector  Further we spend the money and got  fraction collector just  in case and of course an auto-sampler, with holder for both 2 mL glass vials and 96-well plates.

I hope to make  separate post on it soonish

I have until now been analysing sterols  (with success) and I am trying to analyse amino acids. The only progress here is that my earlier post, is not reliable :( , but I hope to have good new on that in a week or two.

The second new toy is an Atomic absorption spectrometer for determining the amount of metals in different materials. Again  I chose an Shimadzu, a model AA7000, with all the extras we could by. This means both, high temp flame, graphite furnace and  metal hydride generator. Again I choose it for all the same reasons as for the HPLC.

I had my two days of training last week, so I don’t have that much to report on it yet , but I hope to get the first numbers out of it next week.

And why do we do all this? Well we want o make more paper like this: http://www.flavourjournal.com/content/1/1/4

I will try to post something soon on these things, as more papers are in the pipeline here

Okay that’s all for now

Getting Matlab to operate on a folder

The last days I have been doing some EPR for a few inorganic chemists, and they all would like to have their spectra as a text file from which the can easily make a nice figure. I used to doe the conversion from Brukers file format to ascii file with he nice function in the Kazan wiever but since I upgraded my computer and now runs a win7 64 bit I can’t get the conversion to work, so needed an other way to do it. I could of course just use the save as text function in Bruker Xenon software, but somehow  it is nicer to just do it a batch operation as the last thing.  Therefore I have written this small Matlab script:

files=dir(‘*.DTA’);

for i=1 : numel(files)
[B,I]=eprload(files(i).name);
S=[B,I];
savename=strrep(files(i).name,’DTA’,'txt’)
save(savename,’S',’-ascii’);
end

 

The “trick here is the genius dir   function, which gets all the file names etc into a structure, NICE!  And then I have  also use the brilliant EPR matlab toolbox Easyspin by Stefan Stoll.

I really like this dir function I can think of a lot things which could be analysed by batch operations on a folder.

Pretty Pictures

On of the big thing in  our research group is confocal microscopy of lipids structure. I have never really been into that line of research, always preferred some good old spectroscopy or a  thermodynamic technique, and constantly complaining  about how people just are making pretty picture with very little information in.

Now for the project I am working on now, it turns out that I have to go and make “pretty picture”. Therefore I have now spent a few day in front of our Zeiss confocale and recorded a few hundred pictures. Most of the are only of scientific interest, but a few are actually good just looking (but from a scientific point of view not really interesting) , so I though I would share a few of these here.

All pictures are fall colour images, made by lase scanning microscopy.

Cheap lipids

Being a lipid biophysicist, I need lipids to wok with, and I especially I need a lot of phospholipids. Usually I, like most other, buy them from Avanvti  in the US. Now there lipids are good, and nobody questions you work if you use lipid from them. But they are pricey and the shipping also cost a lot of money, so I am always looking for alternatives. One of the more promising alternative vendors are Corden Pharma from Switzerland. The have most of the “standard” lipids and they are significantly cheaper than Avanti, Roughly the cost around 1/3 of the Avanti price. Also the shipping is cheaper.

Other nice things are that they ship wery fast. Last order was shipped the same day as I order, where as Avanti only ships on Friday and Monday and  usually take more than  week from they get the order till they ship. Finally Corden includes an certificate with the result of their  analysis, which includes purity, fatty acid composition and water contend. Also a nice thing.

But are the lipids a s good as the Avanti? Well I will go to the DSC next week and check the DSC, in  detail, but until now I have not seen difference in what I done so far.