Why database replication fails when database mirroring is broken?

I'm trying to understand what effect database mirroring can have on replication in SQL Server 2008.

Here is current environment: (don't ask me why it was done this way or to change it)

Server A: SQL Server 2008 (Publisher)

Server B: SQL Server 2008 R2 (Distributor)

Server C: SQL Server 2008 (Asynchronous Mirror)

Server A is setup for SQL Replication and for Database Mirroring. Mirror is going to DR site and replication is used for different purpose at local site. When database mirroring fails (Server C goes off-line or network issue or anything else) and I can see state on Server A (principal, disconnected). At that time replication failed until mirroring was turned off. My understanding was that High performance (asynchronous) database mirroring, should not prevent transaction from being committed on source. The only time transaction should not be committed to the source is when High Safety (synchronous) mode is on. Which would explain that replication not being able to pick up the changes as nothing is being committed on the source.

Is there any scenarios in which High-Performance (asynchronous) database mirroring can break database replication?