HPLC Assay of HCG

[Akthmps88]

Hello --

I am wanting to setup an HPLC assay test for HCG. The matrix is water, sometimes with and sometimes without benzyl alcohol. Concentration is not an issue as I'd like to test lyophilized powders for concentration and I can scale the HPLC assay level concentration accordingly.

I tried to replicate an application note I found online using a Phenomenex Jupiter C18 300A column but the peak turned out to be benzyl alcohol and not HCG.

I have a variety of detectors I can try: charged aerosol, ELSD, refractive index & diode array. I can alternatively using LC-MS if needed but I'd rather stick with a non-MS HPLC detector, if possible.

Has anyone had any success with HPLC separation/detection of HCG? If so, would you be willing to share your method conditions?

Any feedback would be appreciated!

Sincerely,
Aaron

[Zoraku]

Could you try using an affinity column? A quick search shows that Thermofisher has an HPLC column that seems to be designed for HCG.

https://www.thermofisher.com/order/cata ... uct/A37057

[Akthmps88]

Thank you for the suggestion. I have looked this over and sent an inquiry to Thermo. I'm not familiar with this type of column. Do you think it could be used in conjunction with refractive index detection for HCG? I didn't see any chromatograms provided. The elution buffers don't look to be compatible with MS/charged aerosol/ELSD, so I'm thinking refractive index? We are essentially just wanting to do potency verifications on various compounded samples of HCG.

[janoshik]

>Do you think it could be used in conjunction with refractive index detection for HCG?


No Aaron, it couldn't.

[janoshik]

All you really have to do is run a 0.1% TFA water acetonitrile gradient on a 300A RP column and inject enough of HCG to detect it at 214/220 nm.

I used a peptide scouting gradient like that for HCG and it worked no problem.


Purchase a standard, it's cheap on Sigma. And run it.


I'd bet that you are, despite what you might believe, simply not injecting enough of HCG onto the column.

[Akthmps88]

Thank you for the information! I did try two different Phenomenex Jupiter C18 columns, both were 300A. There is a published literature note showing separation/detection of HCG using this column both a phosphate buffer + acetonitrile mobile phase. When I ran several commercial products of lyophilized HCG, I saw a good peak. Then, when I injected the Sigma HCG standard, I did not observe a peak. It turns out the HCG "peak" I was observing across most of my samples was benzyl alcohol and not HCG. I wasn't able to replicate this application note using the Phenomenex Jupiter column.

Do you have any specific recommendations for a 300A column that worked for HCG?

Also, in terms of units/mL, any thoughts on a good HPLC concentration for HCG using UV detection? Somewhere in the range of 5-10 units/mL?

Thank you!

[janoshik]

Any modern 300A column should suffice - I'm sure your does as well. I personally use ACE-C4 3um 300A 3*150mm column with 5-95% 0.1% TFA H2O:MeCN gradient.


>Also, in terms of units/mL, any thoughts on a good HPLC concentration for HCG using UV detection? Somewhere in the range of 5-10 units/mL?

I use ~ 5000 units / ml You might also want to use injection that's rather big. 20 ul or so.

Like I said in previous post, you are not using enough HCG for the analysis.
Trust me, just try it out. You can always lower the amount injected.




Also, for HCG you shouldn't be seeing a single peak - HCG has two subunits and you should see both on RP-HPLC.

[Akthmps88]

So, I ordered the Sigma HCG standard (2500 IU per vial). I reconstituted in 1mL of water and injected 40uL onto the HPLC. I used a Waters Symmetry300 C4 column, 5um, 3.9 x 150mm. Column oven is maintained at 40C, flow rate is 1.5 mL/min and mobile phase is as follows:

Mobile Phase A: 0.1% TFA in Water
Mobile Phase B: 0.1% TFA in Acetonitrile

For the first 1 minute, I maintain the gradient at 98% A + 2% B. Then, from 1-25 minutes, it is a linear gradient increasing to 95% B.

At 220nm what I see are a cluster of 4 peaks (with a few smaller peaks trailing off). These peaks present between approximately 8-10 minutes. The biggest of these peaks is around 150 mA tall.

Not sure what to make of these results. Seems it would be hard to do any sort of quantitation with these small, numerous peaks. I was hoping for 2 distinctive peaks, but that is not the case.

Any thoughts on my method setup and/or results?

I'm thinking about taking this method setup and running it to a Charged Aerosol Detector rather than a diode array. Perhaps these subunits do not have a good chromophore?

Thanks for any additional thoughts.

[HPLC chemist]

Try 214 nm where the peptide peak has its maximum absorbance. What is your injection volume? I had to use 250 uL for 9.6 ug/mL (about 10 U/mL) of Oxytocin.

The peaks could be active HCG, deactivated HCG, denatured HCG, deamidated HCG.... You're lucky you only had 4 peaks! I had 6!

[janoshik]

So, I ordered the Sigma HCG standard (2500 IU per vial). I reconstituted in 1mL of water and injected 40uL onto the HPLC. I used a Waters Symmetry300 C4 column, 5um, 3.9 x 150mm. Column oven is maintained at 40C, flow rate is 1.5 mL/min and mobile phase is as follows:

Mobile Phase A: 0.1% TFA in Water
Mobile Phase B: 0.1% TFA in Acetonitrile

For the first 1 minute, I maintain the gradient at 98% A + 2% B. Then, from 1-25 minutes, it is a linear gradient increasing to 95% B.

At 220nm what I see are a cluster of 4 peaks (with a few smaller peaks trailing off). These peaks present between approximately 8-10 minutes. The biggest of these peaks is around 150 mA tall.

Not sure what to make of these results. Seems it would be hard to do any sort of quantitation with these small, numerous peaks. I was hoping for 2 distinctive peaks, but that is not the case.

Any thoughts on my method setup and/or results?

I'm thinking about taking this method setup and running it to a Charged Aerosol Detector rather than a diode array. Perhaps these subunits do not have a good chromophore?

Thanks for any additional thoughts.

Your method sounds good to me.

I honestly wouldn't bother with Charged Aerosol - if it's a peptide, which it is, it's going to have a rather nice absorbance at 214-220 nm because of the chromophore that peptide bond is.

That being said, I don't have personal experience with CAD.

Check the UV spectra of the peaks that you have, it should help you distinguish which of the peaks are the alpha and beta subunit of HCG.

If all have the same spectra, typical for peptides, than what HPLC chemist said about HCG variants could be true.

Try 214 nm where the peptide peak has its maximum absorbance. What is your injection volume? I had to use 250 uL for 9.6 ug/mL (about 10 U/mL) of Oxytocin.

The peaks could be active HCG, deactivated HCG, denatured HCG, deamidated HCG.... You're lucky you only had 4 peaks! I had 6!

214 nm and 0.1% TFA tends to introduce errors into automatic integration in my experience, so I advised him to use 220 nm.
Miniscule difference in sensitivity, much less interference from TFA.

He injected 40 ul to achieve concentration I had advised him and he got 150 mA peak in there, which, going by memory sounds about right on a normal system.

[Zoraku]

HCG is a glycoprotein, so like nearly every protein out there, should have a very strong absorbance at 280nm. Have you tried doing a simple UV/Vis absorbance test with this? If your matrix is just water and HCG or water, HCG and alcohol you shouldn't necessarily need to even bother with chromatography.

You should be able to use a regular UV/Vis spectrophotometer with a 1cm path length and you can use an extinction coefficient of 1.41x10^4 or just compare it to a standard of known concentration and be able to calculate the concentration that way. I don't know why I didn't think of this from the start.

Even if you want to stick with HPLC, set the wavelength to 280nm and see how it looks.

[Akthmps88]

Unfortunately the types of samples we are ultimately wanting to test will contain preservatives and other agents such as methylcobolamin, benzyl alcohol, etc.

I essentially have 4 good size peaks (not well separated) between ~ 8-10 minutes during the gradient. Peaks between 100-150 mAu (@ 220 nm). I monitored at both 220 and 280 nm. The peak profiles look the same @ 280 nm, but admittedly the peaks are much smaller. They are on the order of 6 - 10 mAu @ 280 nm. The signal intensity is much better at 220 nm.

We tried also running this Sigma HCG standard last night on a size exclusion method (also DAD @ 220nm). We only observed one peak using this method setup, can't say it was a "good looking" peak, but compared to a molecular weight distribution standard we ran prior to the HCG standard, it does seem to line up with a peak in the range of ~36,000 daltons.

[Zoraku]

Unfortunately the types of samples we are ultimately wanting to test will contain preservatives and other agents such as methylcobolamin, benzyl alcohol, etc.

I essentially have 4 good size peaks (not well separated) between ~ 8-10 minutes during the gradient. Peaks between 100-150 mAu. I monitored at both 220 and 280 nm but I only see these peaks @ 220 nm. When I open the 280 nm results, I essentially have just a flat baseline.

We tried also running this Sigma HCG standard last night on a size exclusion method (also DAD @ 220nm). We only observed one peak using this method setup, can't say it was a "good looking" peak, but compared to a molecular weight distribution standard we ran prior to the HCG standard, it does seem to line up with a peak in the range of ~36,000 daltons.

That's very interesting that there is no peak at all at 280. Have you tried doing a complete scan between ~200 - 400nm to locate the actual maximum?

What is the estimated concentration you have in your samples and at what concentration are you preparing the standard?

I'd also be interested in what Thermofisher has mentioned about the affinity column. While it may not work with refractive index, it should have no problems working with any type of UV/Vis detection. The affinity column works by "capturing" the HCG on the column so all other molecules flow through, then you would use a second mobile phase to elute your compound off the column. While this is normally used in purification techniques where you want to purify a specific molecule, I have used it for quantification purposes, and still think it may be worth looking into.

[Akthmps88]

I made a correction to my original post stating there was no signal @ 280 nm. I had the signal scale too far zoomed out. When I looked at it again, I saw the same profile for the peaks @ 280 nm as I observed at 220 nm. The peaks @ 220 nm were around 150 mAu in height, but the same peaks @ 280 nm were only between 6-10 mAu in height. So, significantly less signal intensity @ 280 nm, compared to 220 nm.

I think the Thermo affinity HCG column does look to be of interest. It is rather expensive, however, and I'm worried about longevity after running samples. I asked Thermo about this column and they suggested using the two mobile phases, and monitor @ 280 nm when eluting the HCG. Sadly, they did not have a chromatogram available.

We did try to replicate this application note/publication about a year ago with no success using a 300A Phenomenex Jupiter C18 column. It didn't work for us as published. Sure looks like a nice single peak for HCG, but we were unable to replicate:

https://pdfs.semanticscholar.org/1647/b ... 5bc792.pdf

[janoshik]

I made a correction to my original post stating there was no signal @ 280 nm. I had the signal scale too far zoomed out. When I looked at it again, I saw the same profile for the peaks @ 280 nm as I observed at 220 nm. The peaks @ 220 nm were around 150 mAu in height, but the same peaks @ 280 nm were only between 6-10 mAu in height. So, significantly less signal intensity @ 280 nm, compared to 220 nm.

I think the Thermo affinity HCG column does look to be of interest. It is rather expensive, however, and I'm worried about longevity after running samples. I asked Thermo about this column and they suggested using the two mobile phases, and monitor @ 280 nm when eluting the HCG. Sadly, they did not have a chromatogram available.

We did try to replicate this application note/publication about a year ago with no success using a 300A Phenomenex Jupiter C18 column. It didn't work for us as published. Sure looks like a nice single peak for HCG, but we were unable to replicate:

https://pdfs.semanticscholar.org/1647/b ... 5bc792.pdf

Wouldn't trust that publication too much then, I guess.

I just ran some HCG sample and came with 4 peaks too. Two major ones and two minor one, very close to the first major one.

My HRMS is out of order, so I can't figure out what exactly are those minor peaks, but considering how close they are to one of the hcg subunits, I'd say degradation.