It is the job of particular enzymes at particular environment. Different saccharides or their different sites are evolved for several reasons, such as protecting role, to mask the protein not to be recognized for reasons such as degradation, immune-surveillance. Follwoing the genetic coding from DNA to proteins, what makes us different is the sugar. The reason for it comes as the last is because it is doing a subtle tuning, including to make the same protein existing in different glycoforms. Why? The body don't want the same protein function at the same time and at multiple sites, then how to control it? The cheapest way is work at post-translational level, i.e., use enzymes to mask proteins and unmask them when it is required. But why sugar? My theory is if the same protein run out of the sites of modification (normally limited by amino acid residue used as linkage point such as Serine, Threonine, Asparagine), sugar has tremendous ways to make difference such as 5-stereo centers in one hexapyranose which can linked to another sugar in 5 different sites and 2 anomeric stereoisomers, not mention further modification of the same sugar alone (e.g. sulfate, amino replacement (eg GlcNAc vs Glc). In addition, added saccharides also involve in protein folding, sometimes so critical for their correct 3D structures. Hope this help a bit.

Your question is twofold. First why the differences in glycosylation on the same protein in different tissues, secondly, why are proteins even glycosylated? Well the second one is easiest to answer.

Protein glycosylation has different purposes. Some proteins only attain the proper folding when glycosylated. Without glycosylation they would not be able to perform their role. In other cases the glycosylation helps recognition by other proteins or receptors. There are also examples of specific glycosylation epitopes to play a role in targeting the enzyme to the desired location. And certain glycosylation epitopes protect the protein against degradation. Particularly addition of sialic acid prolongs halflife of proteins. Also, addition of glycans in the folding process helps for quality control of the protein and determines whether a produced protein gets excreted or goes into another round of folding or ultimately targeted for degradation.

The reason for the differences between tissues is not entirely clear as far as I understand it. In some cases it may not need to be a functional reason towards the protein, but more the fact that the protein is produced in a tissue where a certain balance in glycosylation-enzymes exists, that is reflected in the glycosylation of the protein. In other cases the differences in glycosylation may be part of the specific role they have to play in this specific tissue, while they play another in another tissue and hence require different glycosylation.

Moreover, you have to take into account that wherever a protein is produced, it will be present with a very heterogeneous array of glycoforms.

In last years, has been increased the study of the function of glycosylation in the biological function of proteins because there are evidence to suggest that this modification are, in many cases, directly related to the biological function of proteins. Differences in N-glycosylation between different tissues and species are associated with the availability of glycosyltransferase in tissues. Also, proteins in a cell may differ in the glycosylation pattern because factors such as the 3D structure of the protein. Moreover, the same protein may have different oligosaccharides linked to the same glycosylation site which is known as microheterogeneity. The final structure of the glycan that binds to the protein depends on the accessibility that have the tens of glycosyltransferases and glycosidases to act on them during their transit through the secretory way. But, although there are diferent glycosforms in the proteins, the major structures will have the capacity to perform the biological function.