Graduate research project: N-Nitroso compound mutagenicity
mechanisms and comparison with carcinogenicity
Since the project with mutagen-treated transforming DNA wasn’t going to
work, I ended up testing some nitrosamines. [Willie was always
recruiting people to work on nitrosamines. He gave Bill Russell ENU and
that became the animal super mutagen]. Willie had said Don’t
go to medical school, science is much more interesting”; and
later: “Just go into the lab and do experiments; things will
happen!”.
It was fortuitous that Evelyn Witkin’s graduate student, Donna George,
provided some of Evelyn’s DNA repair deficient E. coli strains
when she joined Jane’s lab as a new post-doc. Thus, Willie’s
nitrosamines and Evelyn’s E. coli strains, along with Jane’sHaemophilus influenzae, morphed naturally into my thesis project[a sequence of random events] .
I had a set of methylating agents, MNU, MNNG and Nitrosocarbaryl
(nitrosated insecticide Sevin) and some ethyl homologs. Nitrosocarbaryl
was mutagenic and carcinogenic. It was likely the first nitrosamine
considered a possible human environmental risk (Elespuru et al. 1974).
Why were the mutagenicities of diverse methylating agents so different?
They broke down to the same intermediate, in principle, but mutagenicity
and carcinogenicity varied a lot. Also, how did mutagenicity and
carcinogenicity compare?
Differences greater than 3 orders of magnitude in the mutagenic potency
of methylating agents (MNU, MNNG and N-nitrosocarbaryl) could be
accounted for by differences in their uptake into Haemophilus
influenzae (using 3H-labeled nitroso compounds)
(Elespuru, 1979). The results correlated roughly with the octanol/water
coefficient, indicating the critical effect of lipophilicity on uptake.
This result was surprising and likely important, but uninteresting from
a mechanistic standpoint, as it was not based on a function of active
intermediate generation. There goes our hypothesis . Moreover, the
differences we found did not correlate with carcinogenicity (Lijinsky
and Elespuru 1976; Lijinsky and Schmael 1978). Nitrosocarbaryl was the
most mutagenic but not the most carcinogenic methylating agent. Was
lipophilicity and distribution an issue with different effects in
mammalian systems? We have no idea.
As a contribution to mechanistic studies, I performed mutagenicity
experiments with deuterium replacing the alpha carbon hydrogens of NDMA,
for comparison with the carcinogenicity studies. Diminished mutagenicity
matched diminished carcinogenicity for NDMA, and thus alpha carbon
metabolism was key for both mutagenic and carcinogenic effects (Keefer
et al. 1973; Elespuru RK 1978; Lijinsky 1986).
Post-doc Donna George brought E. coli WU3610 from Evelyn Witkin’s
lab. This was a UV repair-deficient strain with two mutational markers
at 2 amino acid auxotrophic loci, amber and ochre, both with AT
base-pair targets. Using this strain with methylating and ethylating
nitrosamines, I noticed something really interesting: mutations induced
by ethylating agents appeared in one day, while those induced by
methylating agents took two days. Thus, the E. coli strain was
able to differentiate what was happening at the molecular level. DNA
targets of ethylating and methylating agents were different (most likely
direct alteration of AT base pairs vs. alteration of slower growing
GC-based suppressors). Thus, I could differentiate between methylating
and ethylating intermediate-caused mutations. With this tool we could
examine the effects of nitrosomethylethylamine (NMEA) as compared with
nitroso dimethyl- and diethyl-amine. Would the mutagenicity results
resemble that of NDMA or NDEA, neither or both? The mutagenicity results
in E. coli at AT base pairs indicated a preponderance of
ethylation, a result which it seems I didn’t publish. However,
alkylation of DNA in vitro or in vivo indicated an overwhelming bias
toward methylation vs. ethylation of N7-alkylguanine by NMEA alkylation
(von Hofe et al., 1986, 1991). [Although O6 alkylation was not always
studied, differences between methylation and ethylation would not alter
the conclusions]. Deuterium isotope effects for carcinogenicity of
NMEA indicated complex results, including targeting of esophagus, an
organ targeted by neither NDMA or NDEA (lijinsky and Reuber, 1980,
Streeter et al. 1990). A later paper expanding on the idea of studying
differential chemical interactions using mutational endpoints was
probably the first use of the term “mutational fingerprints”, a paper
I thought novel and worthy of PNAS, but not accepted by Dick Albertini
(Elespuru et al. 1991).