Oxygen Isotopes in Atmospheric Oxygen and the Dole Effect

Oxygen bubbles from algae

         One of the more intriguing intersections, to me, between botany and biogeochemistry was the “Dole Effect”. Malcolm Dole, a chemist at Northwestern University, had been measuring oxygen isotope fractionations in inorganic experimental systems for over 15 years starting in the 1940s. He and others, including Harold Urey, noted that the oxygen in the atmosphere had an excess of ¹⁸O relative to what it should be if atmospheric O₂ was in equilibrium with water (Dole, 1935; Morita, 1935). In 1956, Lane and Dole (1956) published experiments that measured oxygen isotope fractionation during respiration by various organisms. Earlier experiments had shown that oxygen evolved from photosynthesis had the same isotopic composition as that of the water in the plant or the medium (Dole and Jenks, 1944). In the 1956 study, respiration was associated with an oxygen isotope fractionation between 10-25‰. The authors concluded that, “the O¹⁶/O¹⁸ ratio of atmospheric oxygen has risen to a point such that the ratios for photosynthetic oxygen delivered to the atmosphere and the oxygen extracted from the atmosphere by respiration are equal.” The problem of the Dole Effect persisted. Studies by Michael Bender, then at the University of Rhode Island, showed that oxygen isotope fractionation during respiration by marine phytoplankton and microbes in oceans was considerably less than what had been measured in the oxygen of the atmosphere (23.5‰) (Bender and Grande, 1987).

         In 1985, I was looking for additional challenges to my work as an isotope biogeochemist. I had spent eight years at the Geophysical Laboratory in the company of geochemists and earth scientists, far removed from biologists and botanists at the start of my career. I met Joe Berry at Carnegie’s Plant Biology Department at a Carnegie Institution of Washington meeting, and we immediately struck up a conversation about how we could weave together experiments on modern plant biology to tackle the potential factors that cause the Dole Effect. Berry, Chris Field, and Olle Bjorkman, all staff members at Carnegie’s Department of Plant Biology, were studying photorespiration in leaves, as well as conducting in situ work with Rubisco. If conditions are right, Rubisco has the potential to fix O₂, rather than CO₂. This can occur when temperatures are high, and the plant has made an excess of oxygen that might damage the plant via free radical production.

         At that time, we could not account for the factors that caused the Dole effect, which had implications for how biological processes were balanced globally.  Clearly, the measurements of isotope fractionation from respiration of O₂ or from net oceanic or terrestrial processes could not resolve the problem Thus, I started on a major discovery endeavor to describe the oxygen isotope systematics of photosynthesis, photorespiration, and the other plant processes that could explain why the atmosphere has a δ¹⁸O of +23.5‰.

         Berry and I submitted a proposal to the Department of Energy and received the following reviews:

“The investigators underestimate the pitfalls of their experimental technique. The trapping of oxygen “on a molecular sieve column” (page 8) and the application of an oxygen electrode (pages 8,9) shift the ratio of the two oxygen isotopes...The hypothesis that photorespiration can account for part of the photosynthetic fractionation is interesting. Is there any reason (literature, preliminary data) to expect it to be so?”

Amazingly, the proposal was funded and on we went.

         To purify atmospheric oxygen from argon in air samples, Tom Hoering and I constructed our first vacuum line at the Geophysical Laboratory in Washington, DC. After oxygen was separated from argon, it was converted quantitatively to CO₂. In late February 1985, we deconstructed the line piece by piece and shipped it to the Department of Plant Biology in Palo Alto, California. Working closely with postdoctoral fellow Robert Guy, we designed and built the “biological” part of the extraction line where enzyme reactions and cells would be incubated for experiments. Experiments began in earnest in 1986, when I spent a year’s sabbatical leave at Plant Biology.

         Our first experiments were with Anacystis nidulans, an easy-to-culture cyanobacterium. Cells were grown in the lab, then transferred to a collapsible bag that was sparged with helium to remove all traces of atmospheric oxygen. We had an oxygen electrode mounted in the bottom of the apparatus, which was constantly stirred. Light was turned on, and oxygen evolved. To kill the reaction, we added phosphoric and salicylic acids, then extracted the dissolved oxygen in a vacuum system. We confirmed earlier studies that showed there was little to no oxygen isotope fractionation during photosynthesis (e.g. Stevens et al., 1975). These first photosynthetic O₂ measurements allowed us to refine our techniques and understand the various pitfalls. Typically, we would start on Monday to grow cells, get the chamber and line in good shape on Tuesday, run the first set of experiments on Wednesday, then repeat Thursday and Friday. Samples of CO₂ were sealed into Pyrex ampules. Every two weeks or so, I traveled down to NASA’s Ames Center at Mountain View to work in David DesMarais’s lab, who kindly allowed me to use his isotope ratio mass spectrometer.

         Rubisco, which stands for ribulose 1,5 bisphosphate (Rubis) carboxylase (c) oxygenase (o), is an unusual enzyme in that it catalyzes the uptake of both CO₂ and O₂. Our second set of experiments was a “revisit” to my Ph.D. dissertation and the re-measurement of both carbon and oxygen isotope fractionation by Rubisco. Times had changed. Rubisco from spinach, cyanobacteria, and R. rubrum could now be expressed in E. coli; no need for growing massive amounts of cells or purifying enzymes. For these experiments, rather than the time-consuming purification and crystallization of PGA, we simply measured the isotopic composition of the remaining CO₂. It turns out that the values I measured for my Ph.D. were several ‰ more negative than those I measured in California, because the δ¹³C of RuBP in my earlier experiments was influenced by unknown contamination. By measuring only CO₂, we avoided that problem. We measured an isotope fractionation for carbon in CO₂ of 29.4‰, similar to recent experiments by Roeske and O’Leary (1984).

         Oxygen isotope fractionations were calculated using Rayleigh equations that compared the oxygen isotope compositions of the O₂ in our reaction chamber at the start of the experiment to the oxygen isotope compositions of O₂ when we took a sample. Using our oxygen electrode, we could calculate the fraction of O₂ that was consumed (Guy et al., 1993). We determined the isotope fractionation from the oxygenation part of the Rubisco reaction to be very similar to what was required to explain the Dole Effect. Approximately 30% of all O₂ produced by photosynthesis is “fixed” by Rubisco prior to its being released. Another 20% is taken up by the Mehler reaction in which oxygen radicals are converted to peroxide, then converted via catalase to H₂O. The isotope fractionation during this reaction yielded further isotope fractionation. We followed with measurements of oxygen isotope fractionations by glycolate oxidase in vitro and cytochrome oxidase. Experiments with cytochrome oxidase, the enzyme important for aerobic respiration, were another story. To perform these experiments, we needed large quantities of the enzyme cytochrome oxidase, which we could obtain from Sigma Chemical Company. Cytochrome c, on the other hand, was expensive--about $500 per sample---and it was destroyed at the end of each experiment. We managed to only have one or two successful runs, measuring an isotope fractionation of 6.6‰, hardly enough to explain the Dole Effect. 

         Rob Guy took the lead on respiration experiments (Guy et al., 1989) in which we investigated an alternative respiration pathway in plants that was cyanide-resistant. We used cells, isolated mitochondria, sub-mitochondrial particles, whole seedlings, as well as purified enzymes to show that whole plant respiration had a oxygen isotope fractionation slightly different from the cyanide resistant respiration, the type of respiration in plants like skunk cabbage. We measured oxygen isotope fractionation by almost all of the enzymes that could possible take up molecular oxygen. The work remains important—and the only studies to date—for understanding the biogeochemical cycles of oxygen on Earth.