Re: How large are the proteins that can pass membranes of intestine epitelium?

The vast majority of proteins will be denatured in the stomach by the 
acid, assuming that the pH is normal.  Given the number of people taking 
drugs like Prilosec/omeprazole and that 10-20% of elderly patients have 
achlorhydria (no acid production), there are a substantial number of 
people where little denaturation occurs.  The lack of acid also diminishes 
or eliminate pepsin activation in the stomach, so there is less 
proteolysis in the stomach.  (This has no apparent physiological 
consequences however.)  

The pH in the duodenum is still slightly acid, aroudn 5.5-6.5 normally.  
We are still apparently able to digest protein adequately even if the 
stomach acid and pepsin steps are absent. Virtually all protein is 
degraded.

The epithelial lining of the intestine, assuming it is intact, does not 
normally allow uptake of large molecules.  Some smaller peptides/proteins 
probably leak through the tight junctions, but this is minor.  Amino acids 
and very small peptides can be transported across the epithelial cell 
layer and deposited in the blood.  The major exception to this is the 
immune system.  The immune surveillance system of the gut is able to take 
up immunoglobuliins and also able to sample proteins from the lumen of the 
intestine.  The immunoglobulins are taken up by specific transport 
systems, but the proteins are largly taken up by pinocytosis or some other 
bulk method.  The total amount of protein taken up is very, very small 
compared to the total protein/amino acid in the intestine.  It's purely 
for immune surveillance not for "food".

As specifically for lectins, they can bind to specific molecules on the 
cell surface and trigger endocytosis/pinocytosis.  This certainly occurs 
in the immune surveillance cells, for example, Peyer's patches.  It likely 
can occur in the epithelial cells also.  This is not however a transport 
mechanism for the purpose of obtaining protein/food.  The lectins 
recognize specific protein structures.  These in turn are simple activated 
to internalize.  This is still a kind of surveillance.  Lectins have 
evolved to recognize protein structures in cells.  This enables the cell 
they are on to attach to other cells.