An endonuclease of Escherichia coli active on a DNA treated with methylmethane sulfonate has been separated from an endonuclease active on depurinated sites. The former enzyme is designated here as endonuclease II, while the latter enzyme is designated as apurinic acid endonuclease. Endonuclease II is also active on DNA treated with-methylnitrosourea, 7-bromomethyl-12-methylbenz[a anthracene, and 'y-irradiation. A third fraction which contains activities for both depurinated and alkylated sites needs further study. Endonuclease II, molecular weight 33,000, has been purified 12,500-fold and does not have exonuclease III activity. Apurinic acid endonuclease, molecular weight 31,500, has been purified 11,000-fold and does not have exonuclease III activity. Exonuclease III, molecular weight 26,000, has been purified 2300-fold and does not have endonucleolytic activity at depurinated reduced sites or at alkylated sites in DNA. Therefore, these are three separate proteins. Exonuclease III can produce, presumably by its exonucleolytic activity, double-strand breaks in heavily alkylated DNA under conditions where it does not make single-strand endonucleolytic breaks at either depurinatedreduced or alkylated sites. The first purpose of this paper is to define endonuclease II of Escherichia coli as an activity different from the apurinic acid endonuclease of E. coli. Strauss and Robbins first described an endonucleolytic activity in extracts of Bacillus subtilis that recognized alkylated DNA (1). In this laboratory, an enzyme in extracts of E. coli, active on heavily alkylated DNA, was partially purified, characterized, and designated endonuclease II of E. coli (2, 3). The substrate used for these experiments was DNA that was entrapped in a polyacrylamide gel and then alkylated with methylmethane sulfonate [MeSO2OMe (MMS)] at an MeSO2OMe-to-nucleotide ratio of 6000 to 1. A partially purified preparation of endonuclease II was also found to have an endonucleolytic activity on depurinated reduced DNA (4), and this activity was thought to be due to the same enzyme that was active on MeSO20Me-treated DNA. However, Verly et al. (5, 6), using the purification procedure originally described in this laboratory, obtained an enzyme that was active on depurinated DNA but not on alkylated DNA. Subsequently, we succeeded in separating the activity on depurinated sites in DNA from the activity on MeSO2OMe-treated DNA (7,8). The former we designate as the apurinic acid endonuclease of E. coli, while the latter we designate as endonuclease II of E. coli. Endonuclease II of E. coli is also active on DNA treated with methylnitrosourea, 7-bromomethyl-12-methvlbenz-[alanthracene, and y-irradiation (7-11).The second purpose of this paper is to demonstrate that endonuclease II, the apurinic acid endonuclease, and exonuclease III are separate proteins. Originally, Yajko and Weiss (12) demonstrated that a number of E. coli mutants deficient in exonuclease III were also deficient in "endonuclease II" and vice versa. The "endonucleas...