Using your idea:
Let $f: \mathbb N \rightarrow A$ be an appropriate bijection. Let $f(k) = x$
Let $g(n) = f(n)$ if $n < k$. Let $g(n) = f(n+1)$ if $n \ge k$.
It may be easier to see this as index rather thann functions (although they are exactly the same):
Let $A = \{a_i\}$ be an indexing for A then $x = a_k$ for some $k$. Let $b_i = a_i$ for all $i < k$ and let $b_i = a_{i+1}$ for all $i \ge k$. Then $\{b_i\} = A\setminus \{x\}$ and $\{b_i\}$ is an indexing that maps $\mathbb N \rightarrow A\setminus \{x\}$
===== thourough answer ====
$A$ is denumerable. That means we can write the terms of $A$ into a numbered list like so:
$1 \rightarrow f(1) = a_1 = w$
$2 \rightarrow f(2) = a_2 = \sigma$
$3 \rightarrow f(3) = a_3 = z$
.....
$k-1 \rightarrow f(k-1) = a_{k-1} = \psi$
$k \rightarrow f(k) = a_k = x$
$k + 1 \rightarrow f(k+1) = a_{k+1} = r$
$k + 2 \rightarrow f(k+2) = a_{k+2} = v$
.....
(I just made up those values $w, \sigma, z,$ etc. The idea being that those are the elements of $A$ and are .... whatever they are. You list them in order because... we can.)
Now we can simply remove the $k-th$ element $x$ and list the values as:
$1 \rightarrow f(1) = a_1 = w$
$2 \rightarrow f(2) = a_2 = \sigma$
$3 \rightarrow f(3) = a_3 = z$
.....
$k-1 \rightarrow f(k-1) = a_{k-1} = \psi$
empty line-- it's gone
$k + 1 \rightarrow f(k+1) = a_{k+1} = r$
$k + 2 \rightarrow f(k+2) = a_{k+2} = v$
.....
But now we are skipping over the matching of any $k-th$ term.
But that's okay. We'll just lower all the numbers by $1$ to make this list:
$1 \rightarrow f(1) = a_1 = w$
$2 \rightarrow f(2) = a_2 = \sigma$
$3 \rightarrow f(3) = a_3 = z$
.....
$k-1 \rightarrow f(k-1) = a_{k-1} = \psi$
empty line we skipped
$k \rightarrow f(k+1) = a_{k+1} = r$
$k + 1 \rightarrow f(k+2) = a_{k+2} = v$
.....
Basically we are creating a new indexing. $f(n)$ and $a_n$ is the old indexing and $g(n)$ and $b_n$ will be the new indexing.
$1 \rightarrow f(1) = a_1 = g(1) = b_1 = w$
$2 \rightarrow f(2) = a_2 = g(2) = b_2 =\sigma$
$3 \rightarrow f(3) = a_3 = g(2) = b_2 =z$
.....
$k-1 \rightarrow f(k-1) = a_{k-1} =g(k-1) = b_{k-1} = \psi$
empty line we skipped
$k \rightarrow f(k+1) = a_{k+1} = g(k) = b_k r$
$k + 1 \rightarrow f(k+2) = a_{k+2} = g(k+1) = b_{k+1} = v$
.....
To formalize we say: $A$ is denumerable so there is a bijection $f:\mathbb N \rightarrow A$. We can define $g:\mathbb N \rightarrow A$ via $g(n) = f(n)$ if $n < f^{-1}(x)$; $g(n) = f(n+1)$ if $n \ge f^{-1}(x)$. (For convenience, let's label $f^{-1}(x) = k$ i.e. $f(k) = x$.
We know this is a bijection as:
1) it is injective. If $g(j) = g(i)$ then: $g(j) = f(j')$ where $j' = j$ or $j' = j+1$ and $g(i) = f(i')$ where $i' = i$ or $i' = i +1$. But $f$ is injective so $j' = i'$. If $j < k$ then $j' = j = i' < k$. But if $i' < k$ then $i \le i' < k$and we have $i' = i$ so $j = i$.
If on the other hand $j \ge k$ and $j' = j+ 1 = i' \ge k + 1$. $i' \ge k+1$ so $i \ge i' -1 \ge k+1$ so $i' = i+1$ so $j' = i'1 = j + 1 = i + 1$. So $j = i$.
So $g$ is injective.
2) $g$ is surjective. If $w \in A \setminus \{x\}$ then $w \in A$ and there is some $j$ so that $f(j) = w$ and $j\ne k$ aos $w \ne x$.
If $j < k$ then $g(j)=f(j) = w$. If $j>k$ than $g(j) = f(j+1) = w$.
So for all $w$ there is a $j'$ where $g(j') = w$. So $g$ is surjective.