Let $M$ be a matroid with an $n$-element ground set $E$. I'll assume that $M$ is connected, co-simple (so its dual has no loops or parallel elements) and has no loops. Fix a particular element $e\in E$. I would like to have some reasonable lower bound in terms of $n$ for the number of circuits of $M$ that contain $e$.
I'd be interested to find out if the exact minimum is known, but I'd also be happy with a crude lower bound that grows with $n$. Note that the conditions on $M$ are necessary in order to get a non-constant lower bound. Without the hypothesis that $M$ is connected, it could be the case that $e$ is in a small connected component so that there are not many circuits containing it. If $M$ has lots of loops or coloops other than $e$, then those loops and coloops cannot belong to circuits containing $e$. We also want the dual of $M$ to have no parallel elements to avoid a situation such as when $M$ is the graphical matroid of a cycle graph, in which each edge is contained in a unique circuit. It's also possible that I've missed some additional restrictions that would need to be placed on $M$.
Note that $M$ itself is allowed to have parallel elements. For example, if all elements of $E$ are parallel, then there are $n-1$ circuits containing $e$; they are the sets of the form $\{e,f\}$ for $f\in E\setminus\{e\}$. Is it possible that $n-1$ is a lower bound in general?