Notice that PIE /kʷ/ in wlkʷo- cannot give /p/ in Latin, therefore the PIE inheritance needs further explanation.

According to De Vaan (2008)

The earlier derivation of lupus from PIE *uIkʷo- 'wolf' requires two ad hoc assumptions: metathesis to *lukʷo- in Italic (as is often assumed for Greek; λύκος), and borrowing by Latin of the Sabellic outcome *lupo~. Alternatively, one could derive lupus from PIE *ulp- / lup- 'marten' (whence, among others, Av. urupi-, Lat. volpes 'fox'). This would amount to a semantic shift from 'marten' to 'fox' to 'wolf, the latter one maybe by tabooistic replacement of an earlier word for 'wolf'. The disadvantage of this approach is that the stem *ulp- is already continued in Latin volpēs, so that an additional reflex in lupus would require further special pleading.