Ex 3 in Transfer Learning with MobileNetV2 assignment

In Ex 3 we used a variable called base_model to store model2.layers[4]
We made so changes on the base_model:
base_model.trainable = True
for layer in base_model.layers[:fine_tune_at]:
layer.trainable = False

Like I believe we used the changes on the base_model, How did these changes affect the model2? Please clarify

Please use model2.summary(show_trainable=True, expand_nested=True) to see how the base model becomes part of the overall model and see the trainable nature of each layer.