Studies on asteroid body composition would suggest their formation involves abnormal mineralization. You may want to consult atherosclerosis models of macrophages, in which abnormal calcium deposition and lipid oxidation is part of the pathology. 

http://www.ncbi.nlm.nih.gov/pubmed/11274065

http://www.nature.com/eye/journal/v22/n10/full/eye200835a.html

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3927347/

In bone, macrophages can be induced to form giant cells (osteoclasts) which are responsible for mineral resorption. 

Thank you for your kind reply.

As you mention, it has been reported that  quasimolecular ion of 1,2-dipalmitoyl-glycero-3-phosphoethanolamine (DPPE) is the main component of ABs.

However, asteroid body is specific to giant cells in granulomas exclusively of sarcoidosis, but not of tuberculosis (Langhans giant cells), Touton giant cells in the xanthoma.  

What makes the difference in the arrangement of multiple nuclei and formation of asteroid body of giant cells depending on types of granulomas???

I mean the disarrangement of multiple nuclei in giant cells as the following pictures. Phenotype of multi-nucleated giant cells varies depending on the local environment and the chemical and physical (size) nature of the agent to which the macrophage-derived giant cells (MGCs) and their monocyte/macrophage precursors are responding. There are numerous kinds of giant cells with multiple nuclei in granulomas; 

1. Macrophage derived

a. Langhans' giant cells

b. Foreign body giant cells (FBGCs)

c. Touton giant cells: Xanthelasmatic giant cells.

2. Epidermal cell derived

a. Tzanck giant cells

b. Multi-nucleated epidermal giant cells.

3. Melanocyte derived

a. Starburst giant cells

b. Giant cells in melanocytic nevus.

i. Balloon cells

ii. Giant nevus cells.

4. Other giant cells

a. Floret-like multi-nucleated giant cells (FMGCs).

At first glance, it would be intuitive to assume that nuclear positioning is regulated by cytoskeletal dynamics (quite likely by actin networks). However, as for what controls the cytoskeleton's actions, it is a separate and more difficult puzzle. 

See some references on nuclear positioning regulation:

https://www.google.com.hk/webhp?sourceid=chrome-instant&rlz=1C1CHWL_enHK630HK630&ion=1&espv=2&ie=UTF-8#q=nuclear%20positioning%20mechanisms%20and%20functions

https://www.researchgate.net/publication/51681485_Nuclear_positioning_Mechanisms_and_functions

Some references on actin-related regulators of nuclear positioning:

https://www.google.com.hk/webhp?sourceid=chrome-instant&rlz=1C1CHWL_enHK630HK630&ion=1&espv=2&ie=UTF-8#q=nuclear%20positioning%20actin

It has been suggested that microtubules could also be involved in coordinating push-and-pull actions between the cell cortex and nucleus:

https://www.google.co.uk/#q=nuclear+positioning+microtubule

In early studies,  some soluble factors (among others) can induce giant cell formation from monocytes: Perhaps such factors and the likes are the upstream signals triggering cytoskeletal dynamics that ultimately determines nuclear positioning. 

http://jem.rupress.org/content/155/1/168.full.pdf

Article Nuclear positioning: Mechanisms and functions

To investigate the mechanisms of multicleation of different unrelated cell lineages would be difficult if not impossible. You may want to look at the mechanisms first in the better studied cell types, (to me esp.) e.g. macrophages.

You may want to consult literature on the effects of RANKL and MAC-1 on macrophages in relation to their transformation into multinucleated osteoclasts (which do have physiological functions, and are perhaps among the best characterized giant cells). 

https://www.google.com.hk/webhp?sourceid=chrome-instant&rlz=1C1CHWL_enHK630HK630&ion=1&espv=2&ie=UTF-8#safe=active&q=RANK+macrophage+osteoclast

https://www.google.com.hk/webhp?sourceid=chrome-instant&rlz=1C1CHWL_enHK630HK630&ion=1&espv=2&ie=UTF-8#safe=active&q=mac-1%2C+osteoclast

References for "RANKL, cytoskeleton" (the word osteoclasts spontaneously jumps out):  

https://www.google.com.hk/webhp?sourceid=chrome-instant&rlz=1C1CHWL_enHK630HK630&ion=1&espv=2&ie=UTF-8#safe=active&q=RANKL+cytoskeleton

Thank you for the introduction of several informative references. The change in cytosklelton such as p-MLC, as you mention, contributes to the difference in the arrangement of multiple nuclei. RANKL is famous for the formation of osteoclasts in the bone remodeling and is also expected to be related to the heterogeneity of giant cells depending on variable types of granulomas. 

Thank you for sharing. Good luck on your studies. 

It is worth noting that asteroid bodies, while interesting, are not specific for sarcoidosis and are not useful as a diagnostic criterion in histopathology. This is true for virtually all of the curious macrophage inclusions (calcium oxalate, Schaumann bodies, Hamazaki-Wesenberg bodies, etc.). Biopsies of non-necrotizing well formed granulomas usually get a pathologic diagnosis that is descriptive and concludes with something like 'consistent with sarcoidosis in the appropriate clinical context', since other causes of similar granulomas are not fully excluded (e.g. berylliosis, et al.) by histopathology.