For exact answer you can perform WB. Western blotting is a technique that separates proteins based on size - in general, the smaller the protein the faster it migrates through the gel. However, migration is also affected by other factors and so the actual band size observed may differ from that predicted. Common factors include...
post-translational modification - e.g. phosphorylation, glycosylation etc which increases the size of the protein.
post-translation cleavage - e.g. many proteins are synthesized as pro-proteins, and then cleaved to give the active form, e.g. pro-caspases
splice variants - alternative splicing may create different sized proteins from the same gene.
relative charge - the composition of amino acids (charged vs non-charged)
multimers - e.g. dimerisation of a protein. This is usually prevented in reducing conditions, although strong interactions can result in the appearance of higher bands. So the single clear band at around 25 kDa possible.
Can you specify how you run your protein. Which method you have used for its separation. Lots of factors are to be considered to be a reason for degradation of protein if any occurring while running your sample or before the procedure. What was your protein of interest? How do you collect it? How you store it? Any pre-treatment policy before running sample you have followed? Biochemical characterization of protein of interest should also be considered.
1. There are chances that proteins run differently in SDS, reasons are plenty. Before all that. Check the sample what you've loaded. High chances that the band you are looking is not the one you are working for (As you are saying it is a clear band). 42 and 25 is a big difference.
2. Study about the protein that you are purifying. There might be some associated factors running along the protein when you got that at 42kDa. A deep purification strategies might clear off those associated chains and leave the protein at a much smaller size.