Good news here - there is no hard and fast rule, although I'd say watch it on the time - you definitely shouldn't be asking 2 minutes of questions before you come up with your structure. Here is what you definitely need to do before the structure.
1) Replay your understanding of the situation
2) Clarify that you understand the exact objectives, ensure there aren't any others
3) (This is what I think you are asking about it) - ask any quick questions before your structure. I'd start by thinking "What would be general parts of the analysis that I would do?" And leave that to be covered in your structure and later. This would include any trend performance, number data, market/competitor/company information/performance, etc. It's hard to define everything that you would cover in your structure, so I'll move to some of the common things it would be okay to ask before the structure.
Definitions - if you simply don't understand a term or description that was given, it's okay to ask. In fact, it's best to ask early (either before or shortly after the structure). Asking later can look pretty bad, whereas asking early shows you aren't afraid to get clarification on things early and avoids the embarressment later. This is especially acceptable if it's a very technical/nuanced industry, or for example you are from a different country/industry etc and lack some basic context (for example, if you are an Indian national and don't know what exactly an American interviewer means by a "strip mall", better to clarify than assume it has something to do with stripping - you may laugh, but I've seen that go wrong in the past)
Scope of the problem/options - related to objectives, but you can clarify what exactly the focus/options for consideration are. For example, for an large MNC, you might want to clarify if adjacent markets/products/ or divesting is something the company can consider or if there are any "sacred cows" which can't be touched. Don't force this though, and only bring it up at this stage if you have some interesting potential insight or it would shape your framework. You can also address this within the structure itself and discover it through your analysis.
Major contextual scenarios/assumptions - If I were to give you a case on an oil producer, you may want to clarify the time period before assuming the structure. If we are in the post 2015 world with low oil prices, that's very different than 2013 with $100/barrel prices. You can see how the different contexts would influence your framework and where you focus your analysis. Again, don't force this, but if some elements of the context would greatly change how you approach the case, good to quickly ask. It isn't essential, but it can help shape your framework and make it more contextualized, less generic which is always a good thing.
The rule of thumb I'd give is to always stop short of getting data or analysis related to the case - if you find yourself asking for detailed information, that's too much before a framework. Think quick questions, especially with a focus on definitions you don't understand and any major scope/context questions if it would help you better frame the case.