Quantification of Lactose in Milk by HPLC-MS/MS?

[carol]

Dear all,

I will really appreciate any comment or suggestion. I am trying do analyze lactose in milk by HPLC-MS/MS, I am using Waters application (WA60118) but instead of using UPLC I have HPLC. The mobile phase is acetonitrile, water both with 0.2% of triethylamine. I used a gradient solvent instead of an isocratic to include more sugars.

I am diluting the milk with water and then in Mobile phase, however I am not able to reach certified concentration of lactose which is 47.5 mg/ml.

I have tried to dilute the milk in ACN only, water alone but the results did not improve. I will really appreciate any comments or suggestion.

Waiting to hearing from any of you,

Thanks in advance,
Carol

[JMB]

I think that we need more info. to answer your question.

What are your MS/MS conditions ?

What level of lactose in milk can you detect by LC-MS/MS at Signal-to-Noise of 3:1 ?

What level of a standard solution of lactose can you detect at S/N 3:1 by direct injection MS/MS ?

[carol]

The ionization is ESI, negative mode,

- The concentration of lactose in milk is 47,5 mg/g
- I have to dilute the samples at least 50000 with water,
- The lowest standard of lactose I can detected successfully is with a concentration of 0.5 ug/ml

[JMB]

Carol,

I am diluting the milk with water and then in Mobile phase, however I am not able to reach certified concentration of lactose which is 47.5 mg/ml.

- The concentration of lactose in milk is 47,5 mg/g
- I have to dilute the samples at least 50000 with water,
- The lowest standard of lactose I can detected successfully is with a concentration of 0.5 ug/ml

For arithmetic convenience, suppose concn. lactose in milk is 50 mg/mL.
Then, 1:50,000 diln. gives 1 ug/mL in milk.
Are you saying that you can detect half of this level (0.5 ug/mL) in a standard, but not in a milk sample ? Is this by HPLC-MS/MS or by direct injection-MS/MS ?

HPLC will cause peak dispersion, and thereby reduce signal height relative to direct infusion. Have you considered defatting the milk by extn. with CH2Cl2 ?; this will remove components that may contaminate ion source and mass filter components (rods/traps etc).

Why the requirement to dilute milk samples by 1:50,000 ?

What are the parent and daughter m/z values in your SRM ?

[carol]

First,
Thanks for your interest!!

I mean, Lactose should be present in the sample I am inyecting through my HPLC, however, I think something is happening in the ESI that doesn't allow proper quantification.

I don't know what is the lowest level of lactose in milk I can detect, I have to check this.

I think you are right, maybe the fact content is afecting the detection...I will try to use DCM or something similar.

According to my comatogram at 1 ug/ml of lactose (injected) the S/N is 54..so maybe I will be able to detect, as i was saying earlier around 400 ng/ml.

Parent 341 dauter 100.7 and 160.7

Thanks again,

[JMB]

Carol,

Theoretical centroid values are,

m/z 341.3 (C12H21O11) ===> 161.2 (C6H9O5) and 101.1 (C4H5O3),

so there is a large calibration error. If the calibration error is constant across the m/z range of interest, this should not affect sensitivity as long as you are sitting on the tops of each of the three peaks AND have an adequate setting for the peak widths.

What is the decimal place at m/z 341, and what are your current settings of peak widths ?

However, I would like to see you re-tune the MS to get m/z calibration to the values shown above, and initially set up the MRM with a +/- 0.5 amu window on each peak. If you then have sufficient sensitivity, the peak width settings can be reduced (to increase specificity) while still maintaining adequate sensitivity; you may finally have +/- 0.2 - 0.3 amu window on each peak.

Good Luck,

JMB

[Consumer Products Guy]

Here, I just return from vacation, and already I'm confused.

Why is HPLC-MS-MS needed for quantitation of lactose level in milk, this isn't like a trace level assay, and there are already HPLC procedures for this, even with a Google search?

[carol]

Hi again,

The only detector the company has is an HPLC-MS/MS and this is what they want to use. There are 100 % for HPLC analysis of lactose but as far as I know only two for HPLC-MS/MS I have tried the same protocol than for the HPLC analysis but something is happening on the MS that doesn't allow proper molecule ionization, I believe, and consequently, unrealistic results

[JMB]

Carol,

Any progress ??

JMB

[carol]

Sorry for the late reply,

I've been fighting with lactose in milk, after trying all the base buffer available in my lab I decided to move to acid mobile phase...and it worked perfectly, no back presurure problem or anything similar.
However, I still finding problem for the quantification in milk, I tried to precipited the protein with ACN and dilute the supernatant with water, but I am not getting any satisfactory results, my next idea is to filter.
as mentioned before, I will apreciate any coment or suggestion.

Carol

[JMB]

Carol,

Have you considered that excess MeCN to ppt. protein will also ppt. lactose ?

Can you post the Waters Application Note, that you mentioned in your first post ?

JMB

[carol]

Dear all,

I am going to start again with my fight against lactose, I hope to win one day.

JMB,
Thanks for your interest,
Have you considered that excess MeCN to ppt. protein will also ppt. lactose ? to be honest no, I haven't...what do you suggest? to swish to MeOH...the pressure in the column will be higher, no?

Waters application is:

http://www.waters.com/webassets/cms/lib ... A60118.pdf

I am going to pp protein and fat using carrez 1 and 2.

I will let you known how it goes.

[BZK]

Hi Carol,try to ultracentrifuge the milk at a least 15-18000 rpm for 30' and collect the clear (opalescent) central phase. It's be a little more than water and lactose. This operation can help a lot to remove fats and proteins.

Regards.

Robertino Barcarolo, Italy