This ties in with extensive prior conversations on how to support visual (or sometimes semantic) markup within titles: