Generally speaking the answer would seem to be:
“No, a single polypeptide chain is encoded by a single mRNA from a
single transcript of a single gene.”
Obviously if one uses the epithet, protein, for a hetero-multimer like haemoglobin, one could say
So far, nothing new since the question was posed two years ago. But, at the risk of being shot down in flames, let me suggest two possibilities.
The vast diversity of antibodies arises not from the fact that there are a vast number of immunoglobulin genes in the genome, but that there is a repertoire corresponding to different sections of each antibodody chain, the recombination of which in individual B-cells gives rise to different genes. See, for example, this section of Alberts et al. As far as the precursor cell is concerned, the mature immunoglobulin chains are the products of different genes. As far as the mature individual B cells are concerned, each chain of their gene products arise from a single recombined gene.
You pays your money and you takes your choice.
Let us consider the metalloprotein like urease. This enzyme, which breaks down urea to ammonia and carbon dioxide, requires Nickel to function. One can adopt the standpoint that the protein is not complete without the metal cofactor. The genetics of urease has been well studied in certain pathogentic bacteria like Helicobacter pylori, where it is part of a gene cluster. Other components of this gene cluster encode proteins required to deliver nickel ions to the inactive enzyme.
One could argue that without Nickel the protein is incomplete, so that functional urease is encoded by multiple genes of that cluster.
Depends what rules we are playing.