I don't think it's common to use DEPC treated water for DNA (and some now don't recommend it for RNA). As mentioned, DNAse contamination is rare.
90ng/ul in a Nanodrop is not a low reading. If it's not DNA, and the ratio is odd, then it sounds like there's contamination of something else. Are you sure you used a normal elution solution, such as water or low concentration Tris , maybe with EDTA? The negative yield, and then positive yield with a bad ratio suggests some kind of contamination. Consider the other sources of contamination, such as all the solutions you used, and whether they were properly washed away, and removed from the column.
Since it's a Nanodrop, you can look at the curve and see where there's an unusual peak, and sometimes that can help diagnose it. Phenol absorbs about 270, I think (not the case now). THe Qiagen gel dissolving reagent absorbs way to the left, but you shouldn't have that for a PCR clean up. That one often is often not completely removed and can raise the 260/280. Oh, I just checked and this is a gel purification: as I said it's not uncommon to have a low absorbance contaminant peak leftover, and for low DNA yields it' becomes relatively big. I see a higher 260/280 ratio for gel pure samples, but not that high. But, if you use a different kit or had a technical issue removing the solutions the result may be different. It's also important that the agarose be completely dissolved. If not, I can imagine undissolved agar harboring these contaminants and affecting the OD. The final spin of the column after removing the flowthrough (qiagen protocol) is probably important to remove the dirty wash solution.
If you get a single clean band on the gel, you can use a PCR clean up column instead of the gel purification, and this is a more reliable method.