Interesting spot, especially with a guy 3 betting who is playing his first hand.  Ill try to break down some things to think about in this hand.

AQo is obviously a great hand, its usefulness in a spot liek this is tricky though.  Although before any action you can probably assume AQo is the best hand, its become less obvious once the 3 bet comes in, especially over an UTG+1 open. 

First thing we have to do is analyze what information we do have about the 3 bet, and thats its position and size.  We have a 3 bet over an opener who has 17 BBs.  Thiios liekly means he plans on getting it in with the UTG+1 opener.  We also have a very smalish 3 bet of 2bb->4bb.  This is usually done for one of three reasons, one is to decrease amount they lose when they fold, one is to induce shoves, the other is not not get a fold when the 3 better has KK/AA(can sort of fall into inducing subgroup).  They dont want to flat aa and kk with so many people behind who can flat, but they dont want to scare off the short stack.  The last peice of information, already partly covered, is that the 3 bet is coming from MP over an EP raise, which we cna tend to belevie means strength.

All of that info should allow us to sort of understand a range for MP1.  80-90% of the time we expect a hand that he is willing/wants to get in with the 17 BB stack.  This range can look something like AQ+,77+.  With the other times he has a hand he may be 3b/folding with as a pure bluff.

Now lets disect the options we have that are not folding:

If you call here(cold call the 3 bet), you can expect UTG+1 to not fold very often, so you are now playing AQ 3 way in a bloated pot, but you do have position working for you.  You could feasibly have the best hand but its far from assured.  UTG+1 may shove if you flat, which we then see what MP1 does, and have to fold if MP 1 calls/shoves, but can easily be fine with calling if he does fold, since there is dead money in the pot.  More often than not I expect UTG+1 to call, and occasionally he will fold.  Against Mustrum, our value in playing post doesnt come from hitting hands, it rather comes frmo getting him to fold on boards that dont hit him(getting AK to fold when no pair, and getting pairs to fold on scary baords).  There is some value to hitting an A agasint KK/QQ or hitting a Q high flop agaisnt JJ/TT and the occasional AQx board agasint AK.  But more often then not we wont get much value from this hand post.  We get hands to fold when we are ahead, and we will get valuetwoned by AK on A high boards.  So the value comes from him checking and us taking down a pot when both of us miss(YAY position!).

In 4 betting we actualyl have a decent spot here.  we are obviously trying to rep a hand stronger then we have, specifically we are basically trying to rep QQ+/AKs,  we have blockers ourself to AA/QQ/ and even a small blocker to AK.  If we assume villains 3 bet range is the 77+/AQ+ rnage we assigned earleir.  We dont expect to get him to fold QQ+/AK ever.  At these stakes its also unliekly we get JJ to fold.  And we probably get TT to fold 50% of the time.  Lets assume we can get AQ,and 77-99 and half of TT to fold.  Now lets do some combo counting:

Not folding:

Ak = 12 combos

AA = 3 combos

KK = 6 combos

QQ =  3 combos

JJ = 6 combos

.5 of TT = 3 combos

Total = 33 combos

Folding:

.5 TT = 3

99 = 6

88 = 6

77 = 6

AQ = 9

Total = 30 combos

So hes folding approx 30/63 = 47.6% of the time.

So 47.6% of the time we profit

4800 (open)+ 9600 (3 bet) + 2700(antes) +2400(bb) + 1200(sb) = 20.7k

52.4% fo the time we lose 20k(the amount you 3 bet).

(.476*20.7)-(.524*20) = -626 chips.

So if we fold out 4 bet we have a slight negative expectation.

hwoever lets see if we are +EV at this point to call a 5 bet shove.

Theres 20.7k in the pot plus our 20k raise so 40.7k total.  we have to call approx 44.3k more.  So 44.3/(44.3+40.7+44.3) = .342.  So we need to have 34.2% equity to call versus our assumed range.

Running some stove:

Agianst JJ+,TcTd,TcTh,ThTs,AKs,AKo (note i included 3 differnt TT combos) we have 29.5% equity, so we cant call.]

However I skipped 1 crucial part, and that is that hehas hands in his range that are not in the 77+AQ+, so assuming he does have some random hands, a 4 bet fold becomes slightly profitable, we need him to have about 3 combos of hands not in that range to become profitable.

Wow great point, toally looked overthe % where OR has QQ+ and isnt folding.  Makes a 4bet/fold even more marginal, maybe a fold pre is better.  Will have to relook over hand in the morning.