Establishment of an E-RAB involves a number of control messaging among core network nodes.
Here I am assuming, UE requested bearer establishment scenario:
When UE figures out that existing bearer can not suffice new requirement. UE's NAS layer sends "Bearer Resource Allocation Request" message to MME. Once MME receives this NAS message, its sends "Bearer Resource Command" GTP-C message to SGW which in turns forward the same message to PGW/PCEF.
PCEF receives response message from PCRF and if response is positive, then PGW responds with either "Create bearer Request" or "Update bearer Request" GTP-C message.
when MME receives this response message (i.e. Create bearer Request) from SGW, it prepares "Activate dedicated EPS bearer context request" NAS message and send to eNodeB as piggybacked withing the S1-AP E-RAB setup Request message. MME also sends "Create bearer Response" GTP-C message to SGW and SGW in turn sends this message to PGW.
Once S1 E-RAB Setup Request message is received by eNodeB, it configuration its layers accordingly.
It prepares and sends "RRC Connection Reconfiguration" message to UE and wait for RRC Connection Reconfiguration complete message from UE. Then It sends S1 E-RAB setup response message to MME.
Once MME receives both "S1 E-RAB Setup Response from eNodeB" and "Activate dedicated EPS bearer context accept" NAS message from UE, It sends Modify bearer request to SGW and receives Modify bearer response from SGW.
This is how a dedicated bearer gets established.
Hope I have answered your query.