Let us make no bones about it: We do not really know what causes depression.
— Andrew Solomon, The Noonday Demon: An Atlas Of Depression (pp. 29). Scribner. Kindle Edition.
I suppose that depends how we parse the logical term "cause". Because we know in great detail the neurophysiology through which depression produces its affects. Neuroreceptors are overly sensitive or not sensitive enough to individual neurotransmitters. Cells produce too little of a particular neurotransmitter, or "reuptake" (consume) too many neurochemical messages before allowing them to reach the neurons they're intended to activate. These neurophysiological errors "cause" mood changes, from anxiety and depression to fight-or-flight. But what "causes" these errors?
There appear to be genetic determinants. There's much correlation to childhood trauma such as loss or abuse. Stress and associated hormones seem to be involved. Trauma whether childhood or adult appears to generate subtle distortions in brain function which can trigger, or at least are associated with, depression.
Current science suggests that mood is regulated by complex interaction between multiple "systems" in the brain. It's an oversimplification to say that depression is "caused" by chemical imbalances which misregulate just one of these systems, say, dopamine or serotonin. It seems to be the case that interaction between all or at least several of them has to go haywire to produce depressive affects.
That's a lot of useful information. Cumulatively it leads to the explosion of treatments available since the late 1980s.
Solomon may be right, though. All of the language here must be nuanced: contributors are suspected but the mechanisms of their action are not understood.
Fortunately treatment can be effective despite these lacunae. There's about an 80% success rate treating major depression, if properly diagnosed and correctly approached.