Note: Moving to the front due to recent comments
Even in KHI Habermas is working with a notion of "transcendental" which is also "natural":
The “knowledge-constitutive interests mediate the natural history of the human species with the logic of its self-formative process” (KHI: 196). Thus they are the conditions of the possibility of experience in the sense that they constitute the object domain of possible experience (and hence they are transcendental) but they are also the product of the process of natural evolution of human species which is ultimately “contingent” (KHI: 34). Hence the term “quasi transcendental” for these interests (TP: 8).
Habermas credits Nietzsche for doing the same; however he censures him for psychologising those interests. The following summarises Habermas’ verdict on Nietzsche regarding Knowledge and interest: “Nietzsche . . . saw the connection of knowledge and interest, but psychologised it, thus making it the basis of a metacritical dissolution of knowledge as such. Nietzsche carried to its end the self abolition of epistemology inaugurated by Hegel and continued by Marx, arriving at the self denial of reflection” (KHI: 290).
By psychologising "transcendental interests" Nietzsche collapses the distinction between "transcendental" and "empirical". In TJ the same charge would be levelled against Quine and Davidson (albeit in different terms and different tone).