How can I release the N-linked glycans from my glycoprotein?

There are two main techniques commonly used to liberate glycoprotein N-linked glycans for glycosylation analysis - hydrazinolysis and endoglycosidase treatment. Our Guide to Glycosylation Analysis covers some of the considerations for choosing between these release methods.

How can I stop the sialic acids falling off my glycans?

Sialic acids are tricky things. In our experience, the purer the glycan sample the more likely the sialic acid residues are to fall off. There are some sensible precautions though:

  1. Avoid the combination of acidic pH and elevated temperature. Wherever possible, keep the pH in the range 6-9 and the temperature below 30 C.
  2. Be careful when drying glycan samples by centrifugal evaporation or rotary evaporation. When you use these procedures try to keep the temperature below 25 C. Lyophilization (freeze-drying) can be safer with respect to sialic acid degradation but with this technique there is a greater danger of losing the entire sample when the vacuum is released.

Which are the best fluorescent tags for glycan analysis?

The following table shows the most widely used fluorescent glycan tags for glycan analysis by HPLC, CE and MS:

Fluorescent Glycan Label Main Applications Fluorescence Wavelengths and Common Analysis Methods
Oligosaccharide Profiling by HPLC-FD and MS λex = 250 or 330 nm, λem = 428 nm

HPLC (amide, C18, AEX, HPAE-FD,
Proc (procainamide) Oligosaccharide Profiling by HPLC-FD and sensitive MS λex = 310 nm, λem = 370 nm

HPLC (amide, C18, AEX, PGC), ESI-MS
(2-amino-benzoic acid)
Oligosaccharide Profiling by HPLC-FD and MS
Monosaccharide Profiling by HPLC-FD

λex = 250nm or 360nm, λem = 425 nm

Oligosaccharides: HPLC (amide, C18, AEX, PGC), ESI-MS, MALDI-MS
Monosaccharides: HPLC (C18)
Oligosaccharide Profiling by CE λex = 455 - 488 nm, λem = 510 - 520 nm

VTag IgG Glycopeptide Profiling by HPLC-FD λex = 250 nm, λem = 360 nm

HPLC (amide), MALDI-MS
Profiling of Human vs Non-
Human Sialylation by HPLC
λex = 373 nm, λem = 448 nm

HPLC (C18)

Key: AEX = anion exchange, HPAE-FD = high pH anion exchange with fluorescence detection, PGC = porous graphitic carbon

FAQs - Quantitative Sialic Acid Release and DMB Labelling Kit

In general we recommend against storing the left over solution because we cannot guarantee the chemical quality after storage. The reason that it is generally too risky to store this solution is that it contains a mixture of two reducing agents and a light sensitive dye – if this solution is not protected from the air (oxygen in the air with react with the reducing agents) and the light then the efficiency of the labelling reaction will be impacted and the quality of the resulting data will suffer. As this mixture is in an ampule – which is required for shipment and storage under inert atmosphere – and is now broken open there is no way to reseal unless you transfer the solution – we consider this too big a risk and potentially too variable of a process from lab to lab to suggest this to our clients as best practice.

Some salts/buffers commonly used with proteins may interfere with the sialic acid analysis process. This is dependent on the amount of sample taken compared to the volume of buffer (as a large amount of buffer can affect the acidity of the solution during acid hydrolysis). In our experience buffers such as PBS are not a problem where the sample concentrations are above 1mg/mL and between 50 and 200 µg of sample is taken for analysis. This will need to be tested for each sample type and we recommend a buffer blank also be taken though to the process to establish if there are any components in the buffer that interfere with the quantitation of sialic acids.

We have seen this happen before under 2 circumstances: 1) When the amount of protein in the sample is too high you can end up with a thick solution after acid treatment. We were able to solve this problem by using less starting material (you can run the assay with 50-200µg starting material). Consequently, if you find that your signal is too low when running the samples on LC, do not dilute them prior to injection. 2) Certain buffers can also lead to a thick solution. Our solution to this was to perform a buffer exchange into water prior to acid release. We use vivaspin cartridges with a molecular weight cut off of 10 kDa for buffer exchange.

The conditions that we have suggested for the DMB kit were developed using the existing scientific literature as a starting point. The sialic acids are released from glycoproteins by mild acid hydrolysis (2M acetic acid, 80°C for 2h) using conditions that preserve the N-acetyl, N-glycolyl and O-acetyl groups. We looked for the best balance between sialic acid release and minimal degradation (both of the sialic acids and of the of the acetyl groups). It was important for us to maintain the acetyl groups and limit their migration as our clients are often interested in the relative abundance of the Neu5,9Ac2 and other acetylated species.

We have performed sialic acid analysis with IgG-type samples where we used 50µg to 1 mg of starting material. For this type of sample the optimal amount of starting material was 200-600µg (linear proportion of sample amount range in validation study). We found that there is no need to increase the volume of any of the reagents. Furthermore, as indicated in the guide if you have low sialylation in your sample, you can run your samples without dilution on the LC.

The experiment was set up this way to ensure maximum solubility and efficiency in the first step (trying to dissolve something in 5µl while heating is not feasible). After this, you only require a portion of this sample for an adequate response on FLR-LC after DMB labelling hence only a 5µL portion is taken through.

We have tried to aliquot fetuin standards from one vial before and our data was also poor. It is best practice not to aliquot the standards but rather use them as provided. This is the only way to ensure that no sample is lost.

We have used this method with many formulation buffers and we have yet to encounter a problem. Regardless, we always run a buffer blank in order to establish if there are any potential interfering species.

Fetuin is a glycoprotein which has both NANA and NAGA. This standard has been studied well in our labs using the DMB assay and we have set acceptance criterion which is based on several years of historical data - this criterion is constantly being monitored and been updated as we collect more and more data about this standard.

GPEP is a glycopeptide with a biantennary disialylated glycan. This is a purified homogeneous glycopeptide that has been quantitated using qNMR. The acceptance criteria is set based on this qNMR value.

There are pros and cons to both standards. Fetuin is a glycoprotein which may match your own sample of interest better than a glycopeptide, however, the glycopeptide is quantified in a more robust way (since it is a smaller molecule that can be subject to NMR). Also, fetuin contains both sialic acids - which may be an important aspect for you, while, GPEP only has NANA. You will need to assess what is important for your assay - we always use both but this is a choice you will need to make for your specific situation.

In terms of sensitive steps to take into account – when we use this kit in-house or for client services the hitch points are often the following:

  1. Accurate transfer of 5ul of the sample from the acid release step for the labelling experiment. Considerable care should be taken because of the small volume transferred
  2. After adding the labelling reagents that the samples are mixed well
  3. Perhaps the step with the most opportunity to impact quantitation is when preparing the LC samples for analysis. These must be prepared with care and mixed very well to ensure a uniform solution is injected into the LC
  4. Also in general – protecting the labelled samples from degradation by exposure to light is important – provided that they weren’t left sitting out for an extended period of time they should be fine – this is easily identified as the samples tend to go pink in colour

We always run triplicates in order to ensure that the data we have is trustworthy. Especially with the quantitative assays it is easy to have some degree of operator error - for instance in this assay transferring 5ul for labelling is a pinch point. It allows the analyst to quickly see if there is an outlier