Artturi Virtanen

According to our experiments with peas, nitrogen is still fixed in a nutrient solution containing 50 to 100 mg ammonium nitrogen per litre, nevertheless, the plants simultaneously also utilize the ammonium nitrogen available. The nodules are much more sensitive to nitrate. Just 25 mg of nitrate nitrogen in a litre of nutrient solution almost completely inhibits nitrogen fixation. Depending on the concentration in the soil of soluble nitrogen compounds and particularly nitrate, nitrogen fixation in legume cultures varies. In nitrogen-poor soils, legumes are nourished practically only by atmospheric nitrogen. In soils rich in soluble nitrogen compounds, the plants draw partly on them. We have recently developed a method to determine nitrogen fixation in pea cultures in various soils and have found that in a humus-rich clay soil containing 0.3 percent nitrogen, a very nitrogen-rich soil, peas receive somewhat over 70 percent of their total nitrogen from the air. Because the roots along with the lower parts of the stem remain in the soil, a significant portion of the nitrogen is incorporated in the soil. Of the other factors that affect nitrogen fixation, I would like to mention the acidity of the nutrient medium. Both free and symbiotically living nitrogen-fixing bacteria thrive, with a few exceptions, in a neutral or weakly acidic milieu. Acidobacteria occur altogether only in chalk-rich, approximately neutral soils. It has been observed that some legumes themselves are not as sensitive to acidity as the bacteria. Here we see the growth of hybrid clover with bacteria, without nitrogen compounds. We can see that in practical terms growth is diminished between pH 5 and 4.5, but with ammonium nitrate we still have growth below that level, even down to pH 4. But it is not worth cultivating legumes at this acidity, because the soil is weak, even with ammonium or nitrate fertilisation. Nitrogen fixation is so closely linked to the life of the cells that so far it has proved impossible to simulate the process outside living cells. This situation has of course made it difficult to elucidate the chemical mechanism of nitrogen fixation and the enzyme system at work in the process. As early as the end of the last century, when researchers began discussing the mechanism of biological nitrogen fixation, Winogradsky expressed the view that molecular nitrogen is reduced directly to ammonia in this process, which therefore constitutes the primary product of nitrogen fixation. No direct evidence to support his theory was available. However, also in the light of more recent findings, ammonia appears naturally to be a product of nitrogen fixation because it has been shown to play a key role in the synthesis of amino acids. I am thinking specifically about aspartase enzyme and glutamic dehydrogenase, which Huber and Adler investigated in great detail. As we will see from the following, the production of ammonia as a final product of nitrogen fixation has recently actually been proved. The production of ammonia, however, sheds little light on nitrogen fixation itself, namely about the first phases of the process. Because ammonia does not necessarily have to be a direct reduction product of nitrogen, as shown by the reduction of nitrate, an already highly oxidised nitrogen compound, to ammonia. In my opinion, the key question to ask today in relation to the mechanism of nitrogen fixation is as follows: Is the first phase of nitrogen fixation oxidative or reductive? The latter would appear natural, if we did not have to take into account the formation of bound hydroxylamine observed during aerobic nitrogen fixation. Bloom was the first to identify molecular nitrogen as a nitrogen source for hydroxylamine in Acetobacter cultures. And Endres later confirmed that oxine nitrogen is formed in Acetobacter cultures both with molecular and with nitrate nitrogen, but not with ammonium nitrogen. Because the cultures exhibited the nitrite reaction after sulphuric acid hydrolysis and oxidation with iodine. We have observed that such compounds are also produced in the root nodules of legumes. The occurrence of free hydroxylamine in living cells and in bacterial cultures is unlikely, because the possible hydroxylamine structure reacts very rapidly with key keto acids in the cell, such as pyruvic oxaloacetic acid and ketoglutaric acid. The free hydroxylamine observed by Bloom and others evidently originated during the treatment of the material from oximes or from hydroxamic acids. The enzymatic production of hydroxamic acids from hydroxylamine and certain organic acids or their amides was demonstrated by Speck, Elion and by us as well. Here we see photos of glutamates and hydroxylamine from which glutamohydroxamic acid is formed We also found that comminuted liver produces benzohydroxamic acid … yes, what is that in German … both from benzoic acid and from the corresponding amide. The observations about the production of bound hydroxylamine support the theory that hydroxylamine is produced both during aerobic nitrogen fixation and during the reduction of nitrate. However, Acetobacter also produce bound hydroxylamine using ammonium nitrogen as the nitrogen source. According to our measurements however much more slowly than with molecular or nitrate nitrogen. Endres was unable to observe the production of oximes with ammonium salts. Nevertheless, oximes are formed, but more slowly. Thanks.] So, here we see the production of bound hydroxylamine with Acetobacter Over a period of three hours no bound hydroxylamine forms with ammonium nitrate. But with molecular nitrogen. Its production can be measured after just one hour. The same is true with nitrate. It is therefore … it is here only molecular nitrogen. So, they are parallel experiments. But we also did experiments with nitrate, and we can see here that after just one hour the product of bound hydroxylamine with nitrate begins. And the same is true for molecular nitrogen. After just three hours production also starts with ammonium … It is therefore unlikely that the hydroxylamine is formed during aerobic nitrogen fixation and nitrate reduction by oxidation of the initially formed ammonia. Because the reaction occurs more slowly with ammonia. Findings by Csaky and me argue against the assertion that the hydroxylamine is completely alien to cells possibly as a product of an ancillary reaction. We observed, in aerated cultures of Torula yeast using nitrate as a nitrogen source, that the organically bound NO group reaches its maximum value within just 10 to 15 minutes, to fall again rapidly. So, we see here that it is a question of Torula utilis. And we found very significant production of bound nitrogen, bound hydroxylamine after five minutes or so. Here we have the peak after ten minutes. And then the bound hydroxylamine falls. With ammonium sulphate we saw during this time no oxine production. We worked here with normal Torula yeast and with so-called low-nitrogen yeast, yeast cells with a low nitrogen content. It is very easy to deplete nitrogen from the cells of microorganisms. We have worked a lot with these depleted cells. Our interpretation of these findings is that after a brief period of adaptation the cell utilises the bound hydroxylamine by reductively converting it to amino groups. So that is the explanation: that we initially see no production, that the cells are not adapted. Observations to date have led me to the assumption that the first phase of aerobic nitrogen fixation is an oxidative step. And that the oxidation product of nitrogen is then reduced by the same pathways as nitrate. So, we have molecular nitrogen here and nitrate here, nitrate is reduced via nitrite and HNO2 and then via hydroxylamine to ammonia. And here, in anaerobic nitrogen fixation I assume that nitrogen is first oxidized. An oxide is formed – we do not know what kind of oxide, N2O or NO – which is then reduced, so that nitrate reduction and nitrogen production proceed along the same pathways. Hydroxylamine, which forms as an intermediate product of reduction, would be reduced for the most part at the ammonia stage, but to a certain extent would also be capable of forming oximes or hydroxamic acids. We also, to clarify the mechanism underlying the production of hydroxylamine, investigated the possible production of oxime and hydroxamic acid in the anaerobic production of nitrogen by butyric acid bacilli. If hydroxylamine is produced, we can rule out primary oxidation of nitrogen during nitrogen production. So, Bloom assumed that a water molecule binds to molecular nitrogen and then the reduction followed, and according to Wieland, the reaction should proceed in such a way that nitrogen is first reduced to hydrazine, after which a water molecule binds to this molecule. And if that is the case, hydroxylamine production should also occur with anaerobic bacilli. In our numerous experiments with Clostridium botulinum under anaerobic conditions with molecular nitrogen as a nitrogen source, we never found even traces of bound hydroxylamine. So, reaction pathways 1 and 2 are unlikely. The absence of bound hydroxylamine in Clostridium cultures strongly supports the theory that anaerobic nitrogen fixation involves the reduction of nitrogen to ammonia without intermediate oxygen-containing products. No experimental findings contradict this theory. Nitrogen fixation by aerobic and anaerobic organisms therefore occurs along different pathways. This theory is also supported by the different effect of hydrogen gas on aerobic and anaerobic nitrogen fixation. Wilson and coworkers found some years ago that hydrogen specifically prevents nitrogen fixation in Acetobacter as well as in Nostoc algae and the root nodules of legumes. Together with Hakala, by contrast, I have found no prevention of nitrogen fixation by hydrogen in butyric acid bacilli. Of particular interest in this context is the observation by Elfolk and me, that the fixation of molecular nitrogen, which is definitely oxidative and is inhibited by hydrogen, is absent during exposure to ultrasound. A conceivable explanation is that the molecular hydrogen successfully competes with molecular nitrogen for oxygen. This may also be the case in anaerobic biological nitrogen fixation. Butyric acid bacilli release a large portion of the nitrogen bound to them into the nutrient solution. In our experiments this portion was as high as 60 percent, though usually the values are between 30 and 40 percent. In the solution we found aspartic acid and glutamic acid in addition to their amides, plus alpha-alanine and valine, but no other free amino acids. Ammonia was always added to the solution. The nitrogen released into solution, in percent of total bound nitrogen, remained on the same order throughout the entire experiment, indicating a kind of secretion and not, for example, cellular autolysis. Wilsson Boris and coworkers have confirmed the large quantity of secretion and, using N-15 as an indicator, made the important observation that the latter is predominantly found in excreted ammonia, which is a product of nitrogen fixation and is not produced by the amination of amino acid. So, it is a process of anaerobic nitrogen fixation. In this context it should be mentioned that we observed copious secretion of nitrogen compounds from leguminous root nodules in the 1920s. We determined that the excreted nitrogen was almost consistently amino nitrogen and that it was mainly incorporated in aspartic acid. We are evidently dealing here with amino acids that the host plant normally receives from the root nodules as the products of nitrogen fixation. If, in some circumstances, the host plant is unable to utilise these amino acids in sufficient quantity, they are excreted in the solid substrate. This phenomenon has of course considerable practical significance, because it also provides non-legumes with access to atmospheric nitrogen as a nitrogenous nutrient. Unfortunately, the factors on which this secretion depends are not currently known in detail. So, this is a sterile culture system, we have in this case no bacteria. Here we have a very effective bacterial strain, clover bacteria. We can see that the first … so, here we have no increase in nitrogen. This small plant contains approximately 8 mg of nitrogen, and this occurs in the seeds. And barley contains around one milligram of nitrogen, which is also found in the seeds. But here, when a pea plant is inoculated and has nodules, the barley grows in this case rather vigorously. But we do not have control over this phenomenon, and we very often obtain negative results. The enzyme system involved in nitrogen fixation is still unknown; because carbon monoxide inhibits nitrogen fixation, it was suspected that haem iron is part of the enzyme system. Molybdenum, according to Bortels, Anderson and Mulder, is essential for nitrogen fixation as well as nitrate reduction. Moreover, in greater quantities than with these. It is interesting to note that the plants thrive without molybdenum if they are fertilised with ammonium nitrogen. Namely peas and other legumes. But not if they are grown with nitrate or molecular nitrogen. Nitrogen fixation and the respiration of aerobic microorganisms proceed in parallel. This has made it difficult to identify the enzyme system involved in nitrogen fixation. Wilson draws attention particularly to the occurrence of hydrogenase, an enzyme that catalyses the dissociation of water molecules to hydrogen atoms, in nitrogen-fixing microorganisms and believes it likely that between the hypothetical nitrogenase and hydrogenase a parallelism exists. Also, by no means do all microorganisms contain hydrogenase capable of fixing nitrogen. The role of hydrogenase in nitrogen fixation is therefore unclear. In the past twelve years, however, observations have been made with regard to the pigment that occurs in legume root nodules and its relationship to nitrogen fixation. Observations that shed light on the machinery at work in symbiotic nitrogen fixation. The red pigment present in the root nodules of legumes was identified as haemoglobin in 1939 by the Japanese researcher Kubo. Our laboratory confirmed the haemoglobin-like nature of the pigment and showed that it is essential for nitrogen fixation. We found that inactive root nodules exhibiting no nitrogen fixation also contain no haemoglobin, which, on the other hand, is always present in inactive bacterial nodules. The ability of active nodules to fix nitrogen appears to be dependent to a remarkable degree on their haemoglobin content. Here we have an experiment series conducted by our laboratory in which we demonstrated the relationship between nitrogen fixation and haemoglobin content. So we have here - there are nine plants in each experiment - we inoculated the seeds with various strains of clover bacteria, in this case with A1, and then with A2, etc. And these strains differ in activity. So, here we see nitrogen fixation with various strains. Here we see approximately the maximum fixation with these strains, here we have weaker strains. And this strain, HO8, is completely inactive. We have studied this strain closely yet have never observed nitrogen fixation with it. Moreover, the nodules formed by this bacterium are free of haemoglobin. They contain no trace of haemoglobin. We determined the haemoglobin content and it is zero in this case. Excuse me. So here we have: nitrogen fixation is zero and haemoglobin content is also zero. But here we have haemoglobin in the other nodules, and we see a certain parallelism between nitrogen fixation and haemoglobin content. The parallelism is not complete. For example, here we see that we have somewhat more haemoglobin than here but significantly more nitrogen fixation. Nevertheless, by and large you can say that a certain parallelism exists. These experiments were carried out together with my coworkers Dr Erkama and Mrs Linkola. Thus, for the first time, a clear chemical difference was found between active and inactive root nodules. The haemoglobin of leguminous nodules, which I have shortened to leghaemoglobin for the sake of brevity, has now become a focus of attention. We have investigated this substance from various approaches and now know the following about it: The pigment readily dissolves in water from digested root nodules. Precipitated with ammonium sulphate, a preparation is obtained with a saturation level of 60 to 75 percent which contains approximately 0.26 percent iron. Electrophoresis separates it into two components, whereby the more rapid component has an isoelectric point of 4.4 and the slow component an isoelectric point of 4.7. The former, namely the component with IP 4.4, is, in our opinion, pure leghaemoglobin. It has the same iron content as blood haemoglobin, around 0.34 percent, but has a molecular weight of about 17,000, compared to 68,000 for blood haemoglobin. So the molecular weight is the same as that of myoglobin. The haem group is identical to the analogous group in blood haemoglobin. By contrast, the amino acid composition of leghaemoglobin differs fundamentally for that of both blood haemoglobin and myoglobin. The histidine content of leghaemoglobin is just under three percent, whereas blood haemoglobin and myoglobin contain over twelve percent histidine. Leghaemoglobin contains isoleucine, which is absent from blood haemoglobin, etc. The IP of leghaemoglobin with 4.4 is lower than at… In urine 900 kg of nitrogen was excreted, in solid fertiliser 650 kg. Then, of nitrogen in urine, the plants benefited approximately: 70 percent. And it cannot be higher, you also have … [inaudible, 00:29:58]. And of the nitrogen in solid fertiliser 30 percent. So, a total of 830 kg. Nitrogen fixation supplied the soil and plants with over 2000 kg of nitrogen. It must be higher, because we have so far had nitrogen losses. But it cannot be calculated, because we do not know how high it is. And the crop, namely the feed purity, so and so much. So, here we now see how much my farm produces in comparison to the average production in Finland. This calculation covers three years: from 1948 to 1950. So here we see that the cultivated area in percent of farmland on my farm is approximately the same as in Finland in general. The differences are the same - cereal on 18 percent and in Finland somewhat less, around 16 percent. Then oats, 18 percent, here somewhat more, 21 percent. Potatoes, more on my farm, around eight percent; here five percent. And then clover fields, around 40 percent and here slightly more than 40 percent. And here we see the soil. So, cereal, summer wheat is simpler, only summer wheat on my farm. And oats: 3500, on average 3,600. Potatoes 22,000 and here over 15,000. And then, yes, these clover-timothy meadows: 4,000 feed units and here 1400 on average. It is not actually clover fields but mostly timothy-grass on average in Finland. And then milk production: per hectare, I get around 2,300 kg of milk compared to 1,000 kg milk per hectare on average. A comparison of this harvest provides clear proof that biological nitrogen fixation really can be used on such a scale, that it is able to meet the demands of intensive farming in which emphasis is placed on animal production. This requires a suitable sequence of crops with the intensive cultivation of legumes, especially clover or alfalfa, and their effective preservation, this is very important, even at a relatively early growth stage where protein content and digestibility are high. By continuing to study biological, and especially symbiotic, nitrogen fixation and the factors that affect it in conjunction with plant cultivation should be possible in the course of time to obtain progressively better results. Sadly, little work has been done in this field. Nitrogen fixation may also prove successful in promoting forest growth in nitrogen-poor soil. In the Nordic countries, the alder offers good opportunities in this respect. I am going to show you several pictures that illustrate how well alders with nodules grow without nitrogen or nitrogenous compounds. So, this is a one-year-old alder in quartz sand. And here we have the corresponding plant with ammonium nitrate, and, it is not growing very well. And that is what we observed - it grows very poorly under these conditions. But the thing is, this alder was inoculated via the air so that nodules often form here. And here we have no nitrogen nutrient and no inoculation during growth. This we see it after two years: Here is spruce, grown alone in quartz sand, and here with alder. Here we see it after five years, and you can see that this five-year-old spruce is growing vigorously. This is after nine years. So there are very large …, so the alder is already very large. In this experiment an attempt was made to gather fallen alder foliage, which is very nitrogen-rich, from the sand. The spruce therefore received its nitrogen nutrient in this experiment mainly via the alder roots. If the alder leaves in autumn are left untouched on the surface of the sand , as happens every year in nature when they lay under the trees, a layer of black, nitrogen-rich soil forms above the sand, and the spruces or pines receive significantly more nitrogen nutrient than from the alder roots alone. In Finland, field experiments are currently underway with mixed cultures of conifers and alders. Biological nitrogen fixation is, in addition to carbon dioxide assimilation, a process of fundamental importance for all life on our planet. All measures to promote the fixation of molecular nitrogen are therefore apt to increase crop yields, boost protein production, in order to create improved opportunities for a rapidly growing human population. Unfortunately, the chemical industry is only interested in the nitrogen industry, whereas research into the enormous possibilities that nature offers us in the form of biological nitrogen fixation …

Comment

Virtanen was a Finish biochemist, being educated in physical chemistry, bacteriology and enzymology. In more than 50 years of research activity, he worked on problems connected to the dairy industry, agriculture and human nutrition. In the present talk he focuses on an issue which is still highly topical in the context of global nutrition: pathways of nitrogen uptake by plants. Nitrogen, phosphorous (via phosphate) and potassium (via potash) are the elements soils are commonly fertilized with to enhance crop yield. However, nitrogen stands out of this trio, because contrary to phosphorous and potassium, it is contained in the air we breathe, and not in a small quantity: the nitrogen in our entire atmosphere has a weight of approximately 4 quadrillion tons. At the same time, in 2010, approximately 135 million tonnes of nitrogen were industrially converted to fertilizers [1]. This nitrogen mostly comes from atmospheric, elementary nitrogen, which is converted to ammonia (NH3) by reaction with elementary hydrogen. This reaction is done by the so-called Haber-Bosch method, which is connected to a series of Nobel Prizes: Fritz Haber (Chemistry) in 1918 for the invention of the method. Carl Bosch and Friedrich Bergius (Chemistry) in 1931 for general technical improvements allowing an industrial scale-up and Gerhard Ertl (Chemistry) in 2007, inter alia for suggesting a complete theoretical explanation of the reaction mechanism. The Haber-Bosch method is still used today. However, it is so energy-hungry that it accounts for 1-2 % of the global energy consumption. If plants could obtain nitrogen directly from the air, a great deal of energy and money could be saved. Unfortunately, plants themselves are unable to comply: elementary nitrogen is very stable and difficult to metabolize – as opposed to the nitrogen contained in nitrate or ammonium salts, for instance. Still, by symbiotically associating to bacteria, certain plants, the legumes (including important crops like beans, peas, lentils, soy and peanuts), have found a way to access aerial nitrogen. In his talk, Virtanen describes experiments aimed at elucidating the extent and mechanism of this aerial nitrogen uptake. He begins by mentioning that such research is complicated by the fact that the enzymes involved in nitrogen fixation may not be studied outside living cells (in fact, this only became possible in 1960, 8 years after the talk). He then formulates what he believed to be the major question of the time with respect to nitrogen fixation: is the first step of its mechanism oxidative (leading to NO species) or reductive (leading to NH species)? After a review of the current knowledge he makes the “educated guess” that the first step of aerobic nitrogen fixation is oxidative, while the first step of anaerobic nitrogen fixation is reductive. Today we know that there is merely one, reductive way of aerial nitrogen fixation. However, other remarks of Virtanen on the character of the enzyme system, namely that it is likely to contain iron and molybdenum, have proven correct.
In the last part of the talk, Virtanen mentions his discovery of leghemoglobin, a hemoglobin like macromolecule, which is synthesized by legumes as a reaction to the colonization of their roots by nitrogen fixing bacteria. He points out that the production of leghemoglobin is tightly associated with the plant’s ability to fix nitrogen. And in fact, the role of leghemoglobin in the symbiosis of plants and bacteria was later found to be the scavenging of oxygen, which would otherwise inhibit the enzymes responsible for the nitrogen fixation process.
Eventually, Virtanen stresses the importance of introducing legumes to agricultural crop cycles, as they can significantly reduce the need for fertilization. In fact, Virtanen mentions good experiences he made with legumes on his own little seaside farm in Finland, which he had bought to test his research results in practice. In view of a steadily increasing human population, legumes could help to create better living conditions by ensuring food supply, he says. But he also mentions that the chemical industry naturally has a high interest in selling nitrogen fertilizers. Looking at the development of nitrogen fertilizer sales in the 60 years after Virtanen’s talk (they increased approximately ten-fold), it sure seems as if Virtanen’s suggestions of 1952 would have deserved more attention.

David Siegel

[1] Food and Agriculture Organization of the United Nations (FAO), Current world fertilizer trends and outlook, 2010.

Lindau Mediatheque

Artturi Virtanen