Some interesting facts 
- A tetramer of 4 identical subunits
- Each subunit is 120kD.
- Active only as a tetramer.
- Mutations in some of the codons of the N-terminal 60aa or C-terminal 100aa results in an inactive, dimeric β-galactosidase. 
- If the above sequences are deleted, the missing protein fragment can be replaced by the corresponding peptide. This is called intracistronic alpha or omega complementation respectively.
- Its N-terminal 23 residues can be replaced by any amino acid residues without affecting the enzymatic activity.
- A mutant with an internal deletion of codons 21-41 of the lacZ gene does not produce any active β-galactosidase.
- A mutant with a deletion of everything past residue 60 (i.e. it expresses only the first 60 N-terminal amino acids) does not produce any active β-galactosidase.
- You can measure lacZ activity using flow cytometry. See A flow cytometric study of stationary phase gene expression in E.coli using lacZ reporter gene fusion
- Alpha-complementation of β-galactosidase does not seem to yield activities equal to wildtype β-galactosidase. Depending on the fragment, the activity can be up to 24% of wildtype . (If anyone has a better reference comparing results from a Miller assay of alpha-complementated β-galactosidase with wildtype, please include it here.)
- Plovins A, Alvarez AM, Ibañez M, Molina M, and Nombela C. . pmid:7811104.
- Zamenhof PJ and Villarejo M. . pmid:4552986.
- Ullmann A. . pmid:1345751.
- [by] Jeffrey H. Miller. Experiments in molecular genetics. [Cold Spring Harbor, N.Y.] Cold Spring Harbor Laboratory, 1972. isbn:0879691069.
original β-galactosidase assay by Miller