Loading control that does not register red blood cell proteins

[synthesize]

Hello,

I am running Westerns on the cervical spinal cord. The blood on some of my samples was not removed before the protein was extracted. Therefore, there is quite a bit of red blood cell protein in the extract. Because erythrocytes contain a lot of b-actin but none of my protein of interest, some of the samples will contain a much higher proportion of b-actin to my protein of interest than would be caused by a change in experimental conditions alone. I am therefore looking for a loading control that is expressed consistently across all cell types EXCEPT red blood cells. In this way, I hope that my loading control will allow me to exclude the contribution of RBCs to the loading control's concentration.

I am considering a nuclear protein. TBP is one candidate. Lamin B1 is another.

I doubt there is much TBP in RBC's, but wouldn't expression of TBP change if there is a change in gene expression in the cell? I think I can test whethere Lamin B1 would be more than marginally present in blood by running a lane with protein extract from blood. Currently just thinking through my options, however.

Any input would be much appreciated.

[synthesize]

To add:

I also considered mitochondrial proteins, specifically COXIV and VDAC1. I decided not to pursue COXIV, because it is a marker of neuronal activity, and we are using neural tissue. And I decided not to use VDAC1, because VDAC1 is present on cellular membranes, and other VDACs are known to be expressed on RBC membranes.