I am trying to understand how to compute the scattering amplitude for the process $e^{+}e^{-}\rightarrow W^{+}W^{-}$, as a reference one could look at Peskin chapter 21.
What I do not understand is how is possible to pass from the expression (21.105) $$i\mathcal{M}_{t}=\frac{ig^{2}}{2}\bar{v}_{L}\frac{\not{k}_{+}}{M_{W}^{2}}u_{L}(l)$$ to the next one (21.106) $$i\mathcal{M}_{t}=ie^{2}\frac{1}{2s_{w}^{2}}\frac{1}{2M_{W}^{2}}\bar{v}_{L}\gamma_{\lambda}u_{L}(l)(k_{+}-k_{-})^{\lambda}$$
It's written that it's true since $m_{e}\simeq0$, but if $l$ is the initial electron momentum then $\bar{v}_{L}\not{l}u_{L}=0$ while here it seems like it has been exploited $\bar{v}_{L}\not{k}_{-}u_{L}=0$, why?
