I have a data frame with many columns, say:
df: name salary age title John 100 35 eng Bill 200 NaN adm Lena NaN 28 NaN Jane
Try this:
df.loc[:, ['salary', 'age']].fillna(-1, inplace=True)