I want to show that that the following two notions of separation in a metric space are equivalent
$\textbf{1st Definition:}$ Let $(X,d)$ be a metric space then a subset $E$ of $X$ is said to be separated if there exists two non-empty and disjoint open sets $U_1$ and $U_2$ in $X$ such that $E \subset U_1 \cup U_2$ and $E\cap U_1 \neq \emptyset \neq E \cap U_2$.
$\textbf{2nd Definition:}$ Let $(X,d)$ be a metric space then a set subset $E$ of $X$ is said to be separated if there exists non-empty and disjoint open sets $V_1$ and $V_2$ in metric space $(E,d\restriction _{E\times E})$ such that $E = V_1 \cup V_2$.
$\textbf{My attempt:}$ It's pretty easy to show that $(1st) \implies (2nd)$ but I am stuck at the reverse implication. If a subset $E$ of $X$ is separated according to 2nd defnition then $E = V_1 \cup V_2$ for some non-empty and disjoint open sets $V_1$ and $V_2$ in $E$. Therefore there exists open sets $U_1$ and $U_2$ in the ambient space $X$ such that $V_1 = E \cap U_1$ and $V_2 = E \cap U_2$. $U_1$ and $U_2$ satisfies all the conditions of 1st definition except disjointness. If I take $U_3 = U_1 \cap (\overline{U_2})^{c}$ to replace $U_1$ in order to impose disjointess I find it difficult to prove that $V_1 = E \cap U_3$.