If a protein is heterologously expressed in E. coli under the T7 promoter, what fraction of the total protein concentration in the cell is the heterologously expressed protein? What could be its concentration at most? Does it strongly depend on the size of the protein or are there certain size thresholds for low/high expression?
Miroux and Walker (1996) Over-production of Proteins in Escherichia coli: Mutant Hosts that Allow Synthesis of some Membrane Proteins and Globular Proteins at High Levels. J Mol Biol. 260: 289–298
The authors report on problems of over-expressing membrane proteins in E. coli using a T7 (pET) system. They isolate mutant strains which show improved levels of expression compared to the parent strain, BL21(DE3). In their Table 1 they document levels of expression of 17 proteins including, as a control, GFP. GFP was found as a combination of soluble and inclusion body protein at a level of 37 mg L-1 in BL21(DE3) and 140 mg L-1 in one of their mutant strains (C41(DE3)). The highest level of expression that they observed was 300 mg L-1 for bovine OSCP.
Unfortunately the authors document neither the total protein content, nor the cell density of the cultures that they analysed.
Let's assume that the cultures were 10e9 cells mL-1 = 10e12 cells L-1.
According to this source (citing Neidhardt F.C. Escherichia coli and Salmonella: Cellular and Molecular Biology. Vol 1. pp. 14, ASM Press 1996), the dry weight of an E. coli cell is 2.8e-13 g
The generally accepted figure for protein as a fraction of dry weight in E. coli = 0.55
I calculate from these numbers a value of 1540 mg protein L-1. For the three protein expression levels described above this means:
GFP in BL21(DE3) = 2.4%
GFP in C41(DE3) = 9.6%
bovine OSCP in C41(DE3) = 19.4%
I conclude (from these numbers, and from personal experience) that overexpression levels are highly variable, and are going to be strain and protein-dependent. Values in the range of 1-20% total cell protein may be anticipated.