Yes, you should either desalt your samples by solid phase extraction OR divert the void volume of your injection to waste (do not infuse phosphate into the electrospray source) This will save you a lot of trouble later.

Because of the high polarity of the target molecules (UDP and ADP),  it is difficult to separate them from the phosphate salts by ordinary HPLC.  An online desalination by SPE or anion exchange may be a good choice to ensure the sample throughout.

For those that have not used solid phase extraction, it has other advantages. (1) samples can be extracted in parallel.  (2) Many chemistries are even offered in 96 well or syringe barrel formats for robotic extraction. (3) This is for many reasons the high through-put answer.

yes, you need to desalt them....use ziptip or stage tips and elute in 80% acetonitrile, 0.1% FA folllowed by speed vac to reduce the volume to the desired levels

Thanks for all good, helpful suggestions above. I have tried to separate my target molecules from phosphate by C18 analytical HPLC column. Just like Bin Guo said, they were difficult to separate. I think C18 analytical HPLC column is more efficient than C18 SPE and ziptip in separation. If HPLC column can't work, SPE and Ziptip have little chance to work. So I decide to try another volatile buffer system to replace the phosphate I used before.

SPE, Ziptips, etc are not a substitute for analytical separation. They function best as a way to desalt or clean-up. I freely admit that SPE sample prep development requires considerable effort and time to get it right. It functions best when the chemistry is different than your analytical chemistry. Ion-exchange, reverse-phase, Size exclusion  and even affinity are all possible.