I import data from JSON file which contain squence names and DNA sequences, and other things. How do I make multisequence alignment, extracts positions that has SNPs (flanking each end of a SNP with some character, 'N' or '-' for easy differentiation), and visual the alignment in a graph (i.e. ggplot)?

As you correctly point out, the pipeline you have in mind consists of three steps:

• I suggest you try the 'rjson' package to read your JSON file. How to do that exactly depends on the actual format of your JSON file.
• Suppose you have read the JSON file such that you finally have a named character vector x. You can use the 'msa' package and either run msa(x, type="dna") or something like msa(DNAStringSet(x)). (use the method argument to choose between the three different alignment algorithms; the default is ClustalW)
• Once you have an alignment, you can start analyzing SNPs. Do I interpret your question correctly that you have previously unknown sequences (e.g. from a previously not sequenced species) and you want to identify novel SNP locations from sequences? If so, I am not aware of any Bioconductor package for doing that, but there might be other tools that are available freely.