Alerting: Interval is not taken from data source but 1s by default

### What happened? We noticed huge memory consumption in our system, and the reason was some alert definitions. In alert definitions, the default interval is 1s. In our system, the scrape interval is 1m, so 1s is unnecessary. I set the 1m interval in the Prometheus datasource, but the alert definitions do not respect it and use the default 1s. ### What did you expect to happen? Take the interval value from the selected datasource, and do not let the set lower value than what is set in the data source. ### Did this work before? No. ### How do we reproduce it? 1. Set interval in Promethues datasource settings 2. Open alert creation, or edit any alert 3. Select Prometheus data source 4. Click on the option next to the selected data source 5. The interval will be empty ### Is the bug inside a dashboard panel? _No response_ ### Environment (with versions)? Grafana: 12.0.2 OS: Linux Browser: Chrome, Firefox ### Grafana platform? None ### Datasource(s)? _No response_