Blood sample normalisation compound


We are using whole blood samples to determine cobalt concentrations in untreated and cobalt-treated animals. The issue is that the starting volume of whole blood used for sample preparation is not accurate, which introduces bias in the results (i.e. more blood = more cobalt). 

We are wondering if there is anything we could use in order to normalise our cobalt results? It would need to be a compound that is consistently present in whole blood (preferably plasma as cobalt increases the number of circulating red blood cells) and be unaffected by cobalt (i.e. there is a possibility that Co2+ might be able to displace and replace other divalent cations such as Ca2+ etc). 

I realise this is a total long-shot, but if we can analyse any other element (maybe zinc?) in our samples then we could relate the total cobalt levels back to the starting volume of blood and normalise across treatment groups. 



  • Hi Sandy, maybe you can add a known amount of an internal standard (e.g. Ge) to each sample. Then use the ratio of counts Co to counts Ge to calculate the Co concentration in each sample.

  • Hi, thank you for your response.

    I don't think that would solve my problem though. Some of the animals were treated with cobalt, some weren't and we are measuring blood cobalt content (amongst other samples such as tissues and cells) to ensure that cobalt is indeed circulating and taken up. This is why it is important to normalise the total blood cobalt content against the amount of starting material. For example, an untreated sample could have a comparable cobalt count if I used a huge amount of starting material versus a tiny amount of treated sample starting material. 

    By adding the internal standard to each sample I would still not be able to relate back to how much overall material there was to begin with. What I am looking for is an inherent internal standard (I am running a cobalt standard curve alongside my samples), which is present consistently in blood (preferably plasma) and doesn't change in the presence of elevated cobalt. 

  • Okay. Sorry, but in that case I don't know the solution to your problem. I am in chemical analysis and don't know anything about biological compositionScream 

  • Hello ,

    Thank you for bringing this challenging application to the community.

    I would like to better understand this - when you say: "The issue is that the starting volume of whole blood used for sample preparation is not accurate, which introduces bias in the results (i.e. more blood = more cobalt).", you mean that the blood volume/mass used for sample prep (by the way - what is the sample prep you are using?) is not consistent? Are you measuring the amount of blood by mass or using micropippetes?

    If so, we can enter in all Agilent's Atomic Spec instruments the initial mass/volume and final mass/volume, so the final concentration will be based upon the reading and the dilution made during sample preparation, regardless if you used more or less sample for each sample prepared.

    If this is totally not related to your question, could you please share some more information on the sample preparation and the study so we can better understand your request?

    Thank you,


  • Hi Rodolfo, 

    Thanks for helping. We have obtained a rough amount of 500ul whole blood from each animal (treated or untreated with cobalt), which was then digested using nitric acid and hydrogen peroxide on a heat block. This digested solution was then diluted in ultrapure water and run through the ICP-MS. For the digestion and dilution all volumes were consistent across the samples, however what was not consistent is the '500ul' of whole blood. Sometimes there may have been 400ul, sometimes 550ul. The actual amount is unknown, as obtaining the blood (aspirating trunk blood with a 1ml pipette) was time sensitive and dependent on how quickly the blood coagulated (or indeed how much trunk blood pooled in the first place). 

    So in order to relate the cobalt results from the ICP-MS back to the amount of blood used, we would need a compound that is consistently present in the blood - like a control compound. Otherwise, an increase in measured cobalt content may not be due to an increase in cobalt uptake, but could be due to an increase in blood volume used for digestion and analysis. 

    I hope that makes sense. Happy to clarify further if there is still some confusion. 

    Many thanks, 


  • Hey ,

    Iron came to my mind as an element that is in relatively high concentration (maybe in the order of ppm) and can be directly related to the amount of blood.

    If you have a chance to digest one sample with an exact initial and final volume, you can define the amount of Fe in those samples (please prepare a blank with the acids to discount any contamination). Assuming that Fe doesn't vary significantly from sample to sample, you can measure Fe in each sample and correct the Co concentration with it.

    For example - Fe concentration of the "known" sample (the one you know the volume) is 20 ppm. An "unknow" sample has 16 ppm Fe.

    16/20 = 0.80

    Divide the Co concentration by this factor and then you should get your data normalized.

    Just emphasizing that we are assuming here that Fe does not change from sample to sample. I believe it does - but it should not be relevant to the final determination.

    I don't know if you ran already the tests and don't have the samples anymore. Anyhow, you would have to digest blanks and a sample (n=3) and determine Fe on them.

    Hope it helps!


  • Hi Rodolfo, 

    Iron is a great idea, however not all haemoglobins will have 4 irons bound at all times which will introduce variation (especially since every red blood cell contains a vast number of haemoglobins). Moreover, cobalt is well known to activate a cellular signalling pathway which leads to an increase in the production of red blood cells and therefore results in a much larger concentration of iron in whole blood. 

    I was thinking that globulins (proteins found in plasma) may be a good target, but don't think ICP-MS can identify whole proteins/molecules? Or looking at Zinc may be an option, but once again it will also be present in red blood cells and I am not sure that it is not affected by the presence of cobalt. 

    Perhaps I am worrying too much about something not as significant as I think it is. Maybe I need to accept that our data is imperfect and simply an indicator of changes in blood cobalt concentrations between groups, as opposed to using the data as a direct readout of cobalt uptake... :( 

  • Hi ,

    I'm afraid you are right. Unfortunately ICP-MS cannot detect globulins (directly, maybe hyphenated with LC, but not sure). 

    Sorry for not being able to help you, maybe someone else can bring a light in this case.

    Best regards,

  • Hi Sandy, 

    Perhaps I am also oversimplifying, but would an anticoagulant such as EDTA added to the blood at time of collection be beneficial and eliminate some of the issues with the preservation? In clinical settings the addition of EDTA to the collection vials is standard practice so I figured it might be worth suggesting, 


Was this helpful?