Why when we have GHZ, then we cannot send a qubit using teleportation technique? 
Suppose the following example

Alice has a qubit $|A\rangle$ and Bob has a qubit $|B\rangle$, then by teleportation technique we can send an unknown qubit state (i.e. $|\psi\rangle = \alpha |0\rangle + \beta |1\rangle )$ Alice to Bob. 

Suppose another example

Suppose that there are three parity: Alice, Bob, and Dina. All these guys have an entangled qubits that are entangled. That is, Alice has a qubit $|A\rangle$, Bob has a qubit $|B\rangle$ and Dina has a qubit $|C\rangle$. They are entangled as $\frac{1}{\sqrt{2}}(|000\rangle + |111\rangle)$. (It is understood that the general case of entangled qubits is known as GHZ-state). Now to the question. 

The Question: Can we send an unknown qubit from Alice to Bob and Dina where they (Bob and Dina) can retrieve the unknown qubit using the teleportation technique? 

I said: Yes.
Professor says: No,
I said: Why?
Professor: Because no-cloning theorem!
I said: Professor, the three qubits are entangled (I want to emphasize that the entangled qubit are different from those that are non-entangled qubits)
Professor: We cannot do this in quantum mechanics, it is impossible.
(The discussion is over) 

My opinion 

I don't really understand, so I decided to put this question here to see if there is any further information about this problem. Therefore, the question for this side is

Could someone tell me why GZA-state doesn't work as EPR-state? 

That is, if we have an entangled two-qubits in different places (EPR-state), then we are able to send an unknown qubit from differen places. But if we have three entangled qubits or more that are entangled (GHZ-state), then we cannot send an unknown qubit; because of no-cloning theorem. 

my argument

I just don't see any replication of qubit (since no-cloning theorem prevents us to construct a copy of a qubit) instead the only thing that is replicated is the 2-classical bits that is sent from Alice to Bob and Dian, so where is exactly the no-cloning theorem here?  
Thank you.
See the figure for three qubits that are entangled in different places. 

 A: A three-party GHZ state can't be used for teleportation because extra copies of a qubit's value act like de-facto measurements. Charlie's copy of the entanglement's 'secret value' prevents some crucial interference from happening when Alice and Bob interact.
You can still transfer information from Alice to Bob using a GHZ state and classical communication... it's just that you need Charlie to help out. For example, he could use LOCC erasure to destroy his share of the entanglement.
Alternatively, Alice can tweak the teleportation procedure to use a GHZ state and classical communication to send a qubit to both Bob and Charlie. But they'll end up with entangled copies of the qubit, instead of independent copies. (If they ended up with independent copies, we'd have violated the no-cloning theorem.) Here's the circuit that does that:

I suggest working out, on paper, what actually happens when you try to teleport with a GHZ state as if it was an EPR pair (i.e. if you drop the last CNOT onto the bottom qubit in the above diagram). It's analogous to what happens when you put a detector on one of the arms of a Mach-Zehnder interferometer: there's extra state space, allowing the various paths to end up in different final states instead of the same state, meaning no interference occurs, and thus no "magic".
A: No-cloning theorem says that you cannot "clone" the state. As in there is no way  to make perfect copies of a quantum state. In the simple teleportation case with two parties the Alice's state is transferred to Bob by using the EPR pair. There is no copying. But if Alice were to try and send her state to two people(Bob and Dina) then that would mean that two copies of the state would exist after the teleportation is done. This is a violation of no-cloning theorem. 
