A number of modified carbohydrates for glycan labelling are available, here we chose to test the possible incorporation of mannosamine into glycosylated structures of the diatom Phaeodactylum tricornutum. This mannose derivative is a precursor compound for the synthesis of sialic acids, which occurs at the surface of cell membranes as a compound of glycolipids or glycoproteins (for review see (Ress & Linhardt 2004)).
Mannosamine was per-acetylated and the chemical reporter-group (azide) was attached. The product is then a 1,3,4,6-Tetra-O-acetyl-N-3-azidoacetylmannosamine (Ac4ManNAz) (Figure 1A). The acetyl-groups increase hydrophobicity of the molecule and make it easier to cross cellular membranes. Azides are small functional groups that are metabolically stable, essentially inert in biological systems, and selectively reactive via Staudinger ligation or azide-alkyne [3+2] cycloaddition (Beckmann & Wittmann 2010, Prescher & Bertozzi 2006).
Therefore organic azides (R-N3) are ideal chemical reporters. In our study, a copper-free [3+2] cycloaddition reaction (Agard et al. 2004, Jewett & Bertozzi 2010) of Ac4ManNAz with the alkyne 4-Dibenzocyclooctynol (DIBO, (Ning et al. 2008)) which was conjugated to the fluorophore Alexa 488 (DIBO-Alexa Fluor 488) was used to visualise the modified glycoconjugates.
Figure 14. A: Copper-less azide/alkyne click reaction: Ac4ManNAz (1,3,4,6-Tetra-O-acetyl-N-3-azidoacetylmannosamine) reacts with the alkyne DIBO (4-Dibenzocyclooctynol) conjugated to Alexa Fluor 488., resulting in labelled Ac4ManNAz. B: Working scheme for fluorophore-labelling in the diatom
To apply MGE in biological systems, cells need to be cultivated in the presence of special monosaccharides (Figure 14). We have tested several approaches with variable combinations of modified monosaccharide and fluorophore to exclude unspecific labelling (Table 6).
Table 6: List of performed experiments. (+) marks the presence, (-) the absence in the respective experimental phase.
also same signal as in negative control 1
When the algae were grown in the presence of azide-modified N-acetyl-mannosamine and incubated with DIBO-Alexa Fluor 488, we could detect clear fluorescent signals, which were mostly restricted to the periphery of the cells (Figure 15A). Resulting from two independent experiments, we could observe slightly different phenotypes in the labelling experiment (Figure 16). The signal either surrounded the cell continuously, or was restricted to discrete spots along the outline of the cell (Figure 16). These differences might be due to different growth states of the cells, which change depending on the age of the individual cell and its phase within the cell cycle. Nevertheless, the signals were always detected at the periphery of the cell, which shows a clear difference to negative control 1. In this control, we could also detect weak fluorescent signals with the addition of DIBO-Alexa Fluor 488 without prior growth of the cells in the presence of Ac4ManNAz (Figure 15B). The possibility that Alexa Fluor 488 is binding non-specifically to compounds within the cells could be excluded by experiments using free Alexa Fluor 488 (Figure 15E). In the control cells without the addition of DIBO-Alexa Fluor 488 (Figure 15C+D), we did not observe any fluorescence.
Figure 15: Cellular localisation of P. tricornutum cells incubated with Ac4ManNAz. Alexa Fluor 488 fluorescence in green, chlorophyll autofluorescence in red and Nomarski differential interference contrast (DIC) in grey scale.
Scale bars: 10 µm. See Table 6 and Figure 14 for details on the experimental design.
Figure 16: Detail images of single cells from the labelling experiment. Alexa Fluor 488 fluorescence in green, chlorophyll autofluorescence in red and Nomarski differential interference contrast (DIC) in grey scale. 3D construction of Z stack images show chlorophyll and Alexa Fluor 488 fluorescence. Scale bars: 5 µm.
Figure 17: Detail images of single cells from a labelling experiment (A) in comparison with single cells from negative control 1 (B). Alexa Fluor 488 fluorescence in green, chlorophyll autofluorescence in red. Scale bars:
10 µm.
A third independent labelling experiment with longer incubation time of the cells with DIBO-Alexa488 (Figure 17A), showed strong similarities to the negative control 1, where the cells were incubated without Ac4ManNAz but with DIBO-Alexa488 (Figure 17B). This indicates that the signal is apparently not due to the reaction of DIBO-Alexa488 with the azide-group, but rather due to the attachment of DIBO-Alexa488 to the surface of the cells. Further, DIBO-Alexa488 seems to be able to enter the cell and might bind non-specifically to other compounds, possibly to thiol-groups (Figure 18), e.g. glutathione, an important component in living organism for detoxification processes.
Figure 18: A possible side-reaction resulting from unspecific binding of DIBO-Alexa488 to thiol-groups.
In all observed cells, we could not detect intracellular signals other than the unspecific signal in the centre of the cells. This indicates that there is no detectable amount of Ac4ManNAz inside the cells. There are three possible explanations for this absence of intracellular signal:
i) the amount of Ac4ManNAz might not be sufficient for subsequent fluorescence microscopic detection, ii) Ac4ManNAz might enter the cells and reacts with DIBO-Alexa488, but is then degraded resulting in the fluorescent accumulation of DIBO-Alexa488 also present in negative control 1 or iii) Ac4ManNAz might not be taken up by the cells in the first place, which is possibly due to the cell wall of the diatoms.
There is so far only one know example of a naturally occurring azide, namely in the dinoflagellate Karenia brevis, an organism known for the production of toxic red tides (Griffin 1994). Apart from this exception, azides are thought not to occur naturally, neither extra- nor intracellularly (Prescher & Bertozzi 2005). Therefore, to explain the intracellular signal observed in negative control 1, we have to conclude that Alexa Fluor 488 DIBO alkyne accumulates in a certain region of the cell (white arrows in Figure 15 + Figure 16 + Figure 17).
This region could either function as a place for storage of compounds which form non-specific aggregates with Alexa Fluor 488 DIBO alkyne or might be involved in degradation/detoxification, possibly by reaction of the alkyne-group with other intracellular (thiol-) compound.