Based on the question, you misread the problem. It's asking about a case where you're designing a mutant, CREB327-119, which is mutated at Ser119. Further, the question stem says that the variants cannot be phosphorylated at Ser119. Therefore, you want a residue that 1) cannot be phosphorylated and 2) is similar structurally and in terms of properties to serine so that it doesn't disturb the protein otherwise (you want to isolate one variable here and that's phosphorylation). So, since tyrosine and threonine can both be phosphorylated, they're out. Glutamate is also a bad choice because it is acidic and charged at physiological pH while serine is not so that introduces another variable. Alanine is the best choice because it is similar structurally to serine while having the property of not being able to be phosphorylated.