In Proof-of-Work, a miner tries to solve PoW puzzle based on the hash of previous block, such that in case of changing block B_i, all blocks after that (blocks B_i+1, B_i+2, ...) are no longer valid since puzzle of every block has been solved based on the hash of previous block.

Now, in a scheme without mining consensus such as Proof-of-Stake, does a validator have to use/insert the hash of previous block, B_i, in the new block, B_i+1 ?

If yes, How ?

And if No, how are blocks chained such that in case of changing a block, B_i, all blocks after this modified block will become invalid? We would need to chain blocks to each other to ensure immutability of the data stored in the blockchain.