I'll again repeat the usual saying that the absolute value of potential energy doesn't matter. What matters is the relative difference.
Having said that we can define the potential energy of a charge at some point relative to some other point. You can think of potential energy as the energy of a system, and not that of a single charge. When we say the charge is at infinity we mean that there are no other charges in the vicinity of the test charge and hence no force is felt by our little test charge. Hence its potential energy is zero. But you are free to define any point of your interest where potential energy is zero.
Remember, once you've defined your zero PE point, the PE's at all other points are obvious.
But what is the work done to bring it from infinity to $0$, i.e, when $r=0$?
In your case, the potential energy of both the $Q$ and the test charge will increase by the same amount. Putting $r=0$ in your formula will give infinite potential which means you'd have to spend infinite amount of energy to to reduce the distance $r$ between the charges to zero which doesn't make sense.
Can we bring it initially from infinity to $r$ and then from $r$ to $0$?
Remember potential energy is a state function. It really doesn't matter which way you do it. Only the beginning and the end points matter. The change in PE will be the same in both the cases whether you bring it directly from $\infty$ to $0$ or $\infty$ to $r$ and then from $r$ to $0$. But as I said before potential at $r=0$ doesn't make sense.