AP down should have generated a critical alert on prime, follow this post, slightly older but same logic applies to newer prime as well.
https://supportforums.cisco.com/document/130381/configuring-mail-server-prime-infrastructure
**rate helpful posts**
-hope this helps-