Use the dot product identity,

where
is the measure of the angle between the two vectors
. Note that

Applying the identity to
and using the distributive and commutative properties of the dot product, we get

which is equivalent to

Now, we're given
,
, and
, so that
