There are a number of things you can do. First, although AlexaFluor 488 is not a bad dye, PE may be a better choice when working with weak signals. Make sure your cells are healthy and try and use propidium iodide or some other dye to distinguish live from dead cells. Dead cells tend to be ...

I generally ask one of the authors for a copy, nearly always works.

I have a lot of experience withe hematopoietic colony assays in soft agarose. I use Low Melting Temperature Agarose (not agar). This way you can keep it in the hood while you work in diluted form (0.36-0.5% is what I've used as the final concentration fro top and bottom layers). I make up 10X s ...

What size is the protein and what size is the membrane?

This isn't my area of expertise but... If you were cooking pasta I'd say the best way to keep the foam down is adding some olive oil to the water. Oils are the best way of keeping proteins from foaming, but I have no ida how this would affect the rest of your experiment.

I grow human DC in serum-free, not mouse. When I've grown mouse myeloid colonies (myeloid, not necessarily DC) I've found the serum-free formulation is still improved with some added serum (5-10%, rather than 20%). Can you try the same protocol with IMDM 20%FBS - that is an easier change than go ...

Cell death suggests that either the cells were overgrown or had reached the end of their natural lifespan. My guess is that they were overgrown, do you than that is a possibility? Did the medium turn yellow? FBS can also be a variable factor. I usually avoid FBS and go serum-free, but if needed ...

I clearly should stop answering dilution problem questions late at night. Dave is absolutely correct.

concentration is 2 ug/ml you need 3ul for the 120ul volume (3 times the volume, thus 3 times the Ab).

Is the size (weight) of the mice accounted for? Is there an age range or sex of the mice specified for the assay? Among these variables, I think the size of the mouse is most important. Older male mice can vary a lot in size within a cage and that could affect the variance in your data.