That depends on what you mean by "can I". Yes, you can collection WB data without load normalization; however, you'll likely not be able to publish it. The greater concern is how you will deal with experimental variance across your blot(s) - reviewers will want to see that you've addressed this concern. There are other load-controls other than b-actin and stains you can use (e.g., Panceau S) to probe the protein load across the entire blot (not specifying a specific protein). Controlling for load can be very important as often there may be a bias across your blot from loading or from uneven probing (e.g., your tray wasn't level during your overnight antibody incubation with more liquid on one side so that the binding kinetics was uneven across the blot). However, common load control proteins can cause issues. What if your experimental variable (treatment, exposure, etc) alters b-actin or another load protein (your normalization would then bias your findings between groups)? What if your load protein isn't within your sample because of your extraction / fractionation methods? Consider these questions when deciding how to proceed.
That depends on what you mean by "can I". Yes, you can collection WB data without load normalization; however, you'll likely not be able to publish it. The greater concern is how you will deal with experimental variance across your blot(s) - reviewers will want to see that you've addressed this concern. There are other load-controls other than b-actin and stains you can use (e.g., Panceau S) to probe the protein load across the entire blot (not specifying a specific protein). Controlling for load can be very important as often there may be a bias across your blot from loading or from uneven probing (e.g., your tray wasn't level during your overnight antibody incubation with more liquid on one side so that the binding kinetics was uneven across the blot). However, common load control proteins can cause issues. What if your experimental variable (treatment, exposure, etc) alters b-actin or another load protein (your normalization would then bias your findings between groups)? What if your load protein isn't within your sample because of your extraction / fractionation methods? Consider these questions when deciding how to proceed.
Can you use alpha-Tubulin or GAPDH as load controls? Otherwise I wouldn't think it could be published, since you don't have any other reference value to show that the possible differences in your blots are not due to loading errors. Ponceau, as professor Ottens points, may be accepted sometimes but definitely you should go with another stable protein.
I have previously used total loaded protein for normalization of preliminary data, if for instance a tested loading control showed a clear treatment or genotype bias that would skew my data. I've also used Ponceau S staining and a measurement of 'total lysed tissue mass' for similar normalizations.
However, I considered these to be a stop-gap means of normalization until I could find a decent loading control.
Using anything but a proper loading control leaves a lot to chance.