When it comes to researching life’s emergence, there are surely more hypotheses and questions than solutions (i.e. proof). It is a very experimental field that allows many ideas to be explored and tested. The research of our group focuses on the possible role of coenzymes in the transition of geochemistry to biochemistry. In this paper, we focus on the most prevalent redox coenzyme nicotinamide adenine dinucleotide (NAD). Many central coenzymes are nucelotide-derived or directly share structural moieties such as adenosine monophosphate (AMP). However, the biological function of this adenosine “tail” remains unclear, as it is not directly involved in the coenzyme’s biochemistry and rather contributes to binding to the protein.
Coenzymes hypotheses often discuss prebiotic precursors of coenzymes, that would have similar functions, but be more easily obtainable from simple prebiotic chemistry (smallest chemically active denominators of coenzymes, so to speak). Having previously published about the specific reduction of NAD to 1,4-NADH with H2 and native metals, we sought out to test: if the same is reproducible with natural occurring Ni-Fe minerals; and if smaller fragments of this molecule could be as efficient and specific under the same experimental conditions.
To test the abilities of NADs precursors, we started with nicotinamide mononucleotide (NMN), which is one of the two nucleotides that form NAD, and planned to continue testing smaller and smaller fragments, down to the nicotinamide moiety, where NAD+ is reduced to 1,4-NADH. The expectations were that some fragments would be more or less efficient than NAD at being reduced, proving insight to possible prebiotic cofactors.
Our experimental setup is simple (Fig. 1): we weigh mineral and metal powders in glass vials with pierced septum lids. These reaction vessels are then placed in high pressure steel reactors that help us to simulate hydrogen partial pressures (and thus concentrations) of so-called water-rock interaction systems (“serpentinizing systems”) found in the Earth’s crust since billions of years.
To our surprise, the samples from the first experiment, testing the abiotic reduction of NMN with H2 and metals, revealed not only a much more reactive system, but also a much more complex mixture than that of NAD’s reduction. This was intriguing because NAD’s functional moiety is identical to NMN’s, as well as the neighboring sugar and phosphate group. In NAD, these connect to a second nucleotide: adenosine monophosphate (AMP). AMP is a non-reactive handle of many universal cofactors, and it is hypothesized to have been necessary only after the emergence of complex enzymes, as a recognition handle, thus smaller fragments would have sufficed to react with prebiotic catalysts. However, for NAD, the absence of the AMP-tail, significantly affected the pattern of this non-enzymatic reaction. To understand how and why it was affected however, became a much bigger task than anticipated in the beginning, and eventually became the sole focus of this project.
The NMN experiments were the first performed in our newly started laboratory in April 2023 – so we used minimal setups at the beginning, while our lab grew not only in members but also by number of steel reactors and gloveboxes.
To untangle the complexity of our reaction mixture, we turned to Dr. Nicole Paczia, head of our institute’s mass spectrometry (MS) department. MS proved useful to assign some of the products, but detailed characterization of all side products and options for quantification remained very limited. A collaboration with Dr. Xiulan Xie, the head of the NMR department at Marburg University, made it possible to identify peaks within the complex NMR spectra. Thus, also quantification and a more complete characterization of the product spectrum became possible. Results revealed that, like NAD, NMN is firstly specifically reduced to 1,4-NMNH, but much more efficiently. However, while NADH is the endpoint of NAD’s reduction, NMNH is not, continuing to to be further reduced and also hydrolyzed. Our interpretation of these results is that in a reducing mineral environment, the dinucleotide would be a more stable reductant over the mononucleotide. And indeed, when we performed reduction experiments with both, we would always see more 1,4-NADH then 1,4-NMNH. In a prebiotic setting this would give the dinucleotide an advantage as reductant (Fig. 2), hinting a role of the AMP-tail before the rise of enzymes.
So, why is NADH so much more stable than NMNH? What role does AMP have? The idea that the nucleotides could fold and stack over each other, thus protecting the nicotinamide ring from the surrounding environment was appealing yet difficult to test. It has been reported in many different ways that NAD folds in solution, but how it behaves at the interface between liquid and solid phases, where the reaction happens, is a more complex issue as also some of our reviewers pointed out.
To help solve this problem, we approached colleagues who are experts on molecular dynamics modelling of organics-surface interactions: Dr. Valentina Erastova and her PhD student Sarah Stewart. Their calculations confirmed how NAD and NMN act differently on a metal surface, underlining how AMP can influence the outcome of non-enzymatic, mineral-based NAD+ reduction. Eventually, although we sought out to find simpler precursors of NAD (without AMP), we discovered how crucial the more complex structure of the molecule could have been for geochemical selection.
Our paper is thus a major collaborative effort, also when it comes to mineral synthesis (Prof. Dr. Harun Tüysüz and Dr. Tuğçe Beyazay) an after-reaction characterization of the same (Prof. Dr. Kerstin Volz and Dr. Jürgen Belz).