Symbiosis: at the Forefront of Chemical Ecology
Chemical ecology studies the structure and function of naturally occurring chemicals that mediate interactions between living organisms as well as their interactions with their environment. During the last few decades, the study of symbiotic nitrogen fixation has developed into a frontier research area in chemical ecology. It has become evident that the symbiotic partners communicate with each other through a vast range of compounds, that is, signals and effectors for sensing and affecting each other’s physiology and development. These chemical interactions are required for all steps of symbiosis; for the initial interactions in the soil between the legume plants and their Rhizobium bacterium partners; and for the formation of root nodules, infection of the plant cells and conversion of bacteria in the symbiotic cells to nitrogen-fixing bacteroids. The latter are not only capable of reducing the airborne nitrogen to ammonia, but also of supporting the growth of their host plant.
I have been working in this exciting field of research for more than 30 years, and my passion for symbiosis is as strong as it was at the beginning. I am fascinated by the secret chemical language of the symbiotic partners, how this dialogue results in the development of a new plant organ, and how differentiation of the plant cells and the bacteria are coordinated.
Symbiosis appears to be full of miracles – of unexpected discoveries that guided me into unexplored territories resulting in fundamental discoveries in plant development, revealing similar mechanisms in the differentiation of plant cells and bacteria and extending my research towards therapeutic applications.
In addition to its beauty and excitement, research on symbiotic nitrogen fixation is also of key importance for the future of mankind. It is clear that by the middle of this century we will have to double crop production in order to cope with the rapid increase of human population on our planet. Plants need nitrogen for their growth, but soils contain limited and usually insufficient amounts of nitrogen sources. Thus, adequate agricultural production can only be realized by an excessive use of nitrogen fertilizers. Plants, however, utilize only 45% of these fertilizers, while the rest remains in the soil, severely damaging the environment by polluting groundwater, surface water and the air, with a negative impact on health, biodiversity and the climate. In contrast to nitrogen fertilizers, symbiotic nitrogen fixation does not harm the environment as the bacteria transport the ammonia directly to their host.
A great future challenge is to implement the mechanism of symbiotic nitrogen fixation in non-legume plants. This could vastly reduce the use of nitrogen fertilizers, contributing to a more sustainable planet. To this aim we obviously have to work out all of the details of the Rhizobium-legume symbiosis. Despite the impressive advances over the past few decades, there is still a long way ahead of us. In a shorter term, however, it seems feasible to improve nitrogen fixation in existing systems as its effectiveness varies greatly among legumes with good fixers, like alfalfa, and weaker ones, such as bean or soybean.
The Initial Molecular Dialogue between the Host Plant and Its Rhizobium Partner
I started my research career with a fellowship in the Institute of Biochemistry at the Biological Research Centre (BRC) of the Hungarian Academy of Sciences in Szeged. This was in the early age of molecular biology. During this time, our Institute was the first one in the Central and East European countries where the methods of genetic engineering had been introduced. Originally, I was more interested in microbial genetics, which was a research topic in the Institute of Genetics. However, as a consequence of my marriage with Adam Kondorosi, who had just started as a group leader in the Institute of Genetics at the BRC, I was not allowed to work in the same institute, even though I was in a different group than the one led by Adam. At that time, Adam was probably the best Rhizobium geneticist in the world, and soon he was surrounded by highly talented and motivated young researchers and students from Hungary as well as abroad. Their five o’clock tea time was reserved for lively scientific discussions and brainstorming. Sometimes I joined them, but at the dinner table we discussed the new results from our laboratories daily, and I became quite fascinated by the Rhizobium research.
The fundamental research questions of the early 1980s dealt with how the bacteria can provoke development of a new plant organ, the root nodules; how the plant recognizes its Rhizobium partner out of billions of microbes around the root system; and how the Rhizobium strain-legume host specificity is achieved. In 1982, Adam got a one-year invitation to supervise and strengthen the Rhizobium research in the Max-Planck Institute für Züchtungsforschung (MPIZ) in Cologne. The MPIZ director, Jeff Schell, invited me as well, and for the first time we were officially allowed to work together and could take advantage of our complementary expertise. The most pressing question was which and how many Rhizobium genes were required for nodulation. The global competition was extremely intensive in the field, but by jointly tackling the problem, our «symbiosis» proved to be extremely effective. Returning to Szeged I worked as before in the Institute of Biochemistry, but I continued my collaboration with Adam and his group, and published the first nodulation genes in 1984, and in 1985 their DNA sequences as well. We demonstrated that one set of the genes was conserved in all rhizobia – we named them common nodulation genes – while the other set was required for host specificity.
In the following years I worked primarily on the regulation of nodulation genes. We have shown that the 23 nodulation genes in Rhizobium meliloti (later renamed as Sinorhizobium) were organized in 6 transcriptional units and co-regulated by a conserved cis-element, the nod-box, which was the binding site for the transcriptional activator NodD protein. The nodulation genes are silent in Rhizobium cultures, but can be activated by flavonoids or isoflavonoids excreted from the seeds and roots of their host legumes. These plant flavonoids/isoflavonoids represent the first signalling molecules in the Rhizobium-legume interactions. While their direct interaction with NodD has not been yet confirmed even today, we demonstrated that flavonoids are required for the strong binding of NodD to the nod-box, and thus for the expression of nodulation genes.
In 1987, the Centre National de la Recherche (CNRS) decided to create a modern Plant Science Institute in Gif-sur-Yvette, France. An international search committee selected Adam as the director of the future institute, and offered me a group leader position. This was a great – but unexpected – opportunity in our life. It was, however, not an easy decision. The work had advanced very well in Hungary, and was highly recognized on an international level. Following up on our first stay in Cologne, we maintained an excellent collaboration with the MPIZ. Moreover, our daughter was born in 1988, when the CNRS offer came. Setting up a brand new institute and starting a new life with a baby in a new home country was highly challenging. We accepted the CNRS offer on the condition that we could maintain the Szeged laboratory. The CNRS agreed, and created a twinning program with the Biological Research Centre in Szeged, generously supporting the exchange of young scientists and their work in Szeged. As a scientific director (DR2 and then DR1 level), I was one of the four initial group leaders at the Institut des Sciences Végétales. At that time many labs – including the ones in Gif-sur-Yvette, Szeged and Cologne – were studying the functions of nodulation genes. I was involved in the discoveries made in the early 1990s demonstrating that nodulation genes are involved in the synthesis of host-specific bacterial signals, the Nod factors. In 1991, French chemists had first identified the structure of a Nod factor from Rhizobium meliloti as a sulfated lipochitooligosaccharide molecule. A year later, we reported the existence of a family of Nod factors from the same bacterium. Afterwards Nod factors were identified from many other Rhizobium species, each having the same chitooligosaccharide backbone, but carrying different substitutions on the terminal sugar residues. The chemical structure of the Nod factors confirmed our original finding and classification of nodulation genes since the core Nod factor structure was encoded by the common nodulation genes, while the substitutions – required for the host-specific interactions – were provided by the host-specific nodulation genes. Application of Nod factors on the host plant root induced cell division in the root cortex and development of the nodule primordium. On the root surface, in the root hairs and epidermis, the Nod factors were required for bacterial infection of the plant cells. Subsequently, the mechanisms of Nod factor perception and signal transduction had to be elucidated. By the beginning of this century, as a result of the concerted worldwide efforts of the scientific community, the Nod factor receptors were finally identified in the model legumes Medicago truncatula and Lotus japonicas, as were elements of the signal transduction pathway.
The Crucial Role of the Cell Cycle and Genome Amplification in the Growth and Maturation of Symbiotic Cells
After the discovery of Nod factors, I was curious to know how these bacterial signals can reactivate the cell cycle in the differentiated root cortical cells, and how this local cell proliferation leads to the development of a complex nodule structure. In an almost unique effort in the field, I studied how the plant cell cycle is activated and regulated at different stages of nodule development. These studies resulted in the identification and characterization of several new cell cycle genes, and had a significant impact on plant cell cycle research. In the nodules, the Rhizobium-infected symbiotic plant cells grow gradually, reaching a size that is eighty times higher than the diploid cells in the nodule meristem. I was fascinated by this extraordinary cell growth. By 1999, with my team I discovered the molecular mechanism, showing that cell growth is mediated by repeated endoreduplication cycles –duplication of the genome without mitosis and cell division. We demonstrated that the formation of highly polyploid symbiotic cells is essential for the development of nitrogen-fixing nodules, which provided the first evidence for the biological significance of polyploidy in plant development. The constant development and presence of polyploid cells in the nodules allowed us to discover a key conserved regulator of the cell cycle, the CCS52A protein, which controls the switch between the mitotic and endoreduplication cycles, and consequently, cell proliferation and cell differentiation. After the discovery of CCS52A in Medicago we extended our studies to Arabidopsis, and during the period 2000-2004 we clarified the mode of action and the regulatory functions of the CCS52A proteins. Our discoveries represented a key contribution to the field of plant developmental biology and cell cycle research. Our latest work from the past year confirmed the importance of growing ploidy levels in gene regulation as changes in the epigenome, including chemical modifications of DNA and histone proteins as well as ploidy-dependent accessibility of DNA correlated with specific expression patterns of symbiotic genes. Empty nodules, devoid of bacteria, might undergo only a single endoreduplication cycle but never reach the ploidy levels of the nitrogen-fixing symbiotic cells, indicating that bacterial signals might be required for induction of repeated endoreduplication cycles in the host cell. The nature of these chemical signals represents another frontier in symbiotic chemical ecology, and remains to be discovered.
Host-Dependent Reproductive Fate and Morphological Diversity of Endosymbionts
From 2004 onwards, an old question brought research on the bacterium partner back to my attention. It has been known for 150 years that the morphology of bacteria shows wide diversity in the nodules of different legumes. The reason, however, was unknown. In certain legumes, like soybean or Lotus japonicus, the nitrogen-fixing bacteroids are similar in size and form to the free-living bacteria, while in pea, alfalfa or vetch, the nitrogen-fixing bacteria are elongated or even branched cells of a very large size. The first observation on these huge elongated structures in the nodules dates from the seventeenth century, when Marcello Malpighi’s studies on plants led him to believe that the nodules contain worms. In 1888, Beijerinck was the first to describe the nodules as containing bacteria. From the beginning of the twentieth century, many publications dealt with the different shapes and sizes of bacteroids, their cultivability, and the question of what their form can represent. Pfeiffer (1928) considered the elongated-branched structures to be true bacteroids that could be induced by specific chemical and physical properties of certain compounds. Cappalletti (1926) believed that the special forms of the bacteroids represent reactions against the plant immune system. Due to inconsistent data that sometimes arise from technical problems, the fate of bacteria in various legume nodules and the molecular mechanisms of the diverse morphology of bacteroids have remained unsolved, even in the twenty-first century. When I studied the function of CCS52A in fission yeast, the expression of CCS52A provoked the formation of large, elongated polyploid yeast cells that were remarkably similar to nitrogen-fixing bacteroids in M. truncatula nodules. Although it has been a generally accepted view that bacteria do not differentiate, and indeed they do not do it in the culture medium, we discovered a remarkable multistep differentiation process of S. meliloti bacteria in the M. truncatula nodule cells. S. meliloti bacteria infect the smallest and youngest nodule cells below the meristem and multiply in the young, growing symbiotic cells until they reach a certain cell density. Then the bacterial cell division stops. The bacterial cells begin to elongate and grow progressively to huge elongated and branched nitrogen-fixing bacteroids in parallel to the growth of the host cell, occupying and entirely filling the host cytoplasm in the matured symbiotic cells.
To solve the problem as to whether bacteroid development involves genome amplification, we examined the morphology, size, DNA content, reproducing ability and membrane permeability of various rhizobia in free-living and symbiotic states. Compared to the bacterium cultures, the S. meliloti bacteroids were 5 to 10 times longer and showed extensive amplification of the entire bacterial genome with endoreduplication cycles, which also revealed in prokaryotes a positive correlation between the DNA content and cell size. Moreover, these bacteroids displayed increased membrane permeability, and were non-cultivable. Therefore, this bacteroid differentiation is irreversible and terminal. In contrast, the rhizobia, which are similar in free-living cultures and symbiosis, like those in Lotus or soybean nodules, do not show alteration in their DNA content or membrane permeability, and do not lose their ability to resume growth. Consequently, their fate is reversible between the free-living and symbiotic states. Subsequently, we demonstrated that the reproductive fate and morphological diversity of the endosymbionts are controlled by the legume host. A few years later, Oono and Denison compared the symbiotic efficiency of swollen (terminally differentiated) versus non-swollen (reversible) rhizobial bacteroids and concluded that swollen bacteroids are more efficient in nitrogen fixation. It is obviously to the benefit of the host plant to guide the terminal differentiation of bacteroids. The host-dependent fate of bacteroids raised the question: ”What are the plant factors?”
Plant Effectors of Terminal Bacteroid Differentiation: The Extraordinary Complexity of Nodule-Specific-Secreted Peptides
We predicted that the plant factors provoking terminal bacteroid differentiation would be present in M. truncatula but absent in L. japonicas, where the fate of bacteroids is reversible. Comparing the gene expression in M. truncatula and L. japonicus nodules resulted in the discovery of an astonishing complexity of novel secreted symbiotic peptides that were only expressed in M. truncatula and not in L. japonicus. More than 700 M. truncatula genes code nodule-specific cysteine-rich (NCR) peptides with exclusive expression in the S. meliloti-infected symbiotic cells. A symbiotic cell expresses 500 to 600 NCRs during its maturation, but different sets at the early, intermediary and late stages, and in correlation with the ploidy levels and in line with their presumed function in bacteroid differentiation. In 2010. we proved that these peptides are the plant effectors of bacteroid differentiation. Blocking the delivery of these secreted peptides to the bacteroids abolished their differentiation while expressing NCRs in L. japonicus nodules induced features of terminal bacteroid differentiation. The NCR genes are small, and with a few exceptions, are composed of two exons. The first exon encodes the signal peptide while the second one does so with the mature peptide, which is most frequently 30–-50 amino acids long and contains either 4 or 6 cysteines at conserved positions. The signal peptide is relatively conserved, but the amino acid sequence of the processed mature peptides is highly divergent. Due to differences in amino acid composition, the isoelectric points of the NCR peptides vary from 3.2 to 11.25; with roughly equal numbers of anionic and cationic peptides. While the NCR family is unique, their structure has a certain similarity to antimicrobial peptides, especially to plant defences. The latter are also secreted, cysteine-rich peptides, but they have 8 cysteines and a different signal peptide. Thus it appears that the various early – and mostly forgotten – assumptions originating from the beginning of the past century, which predicted the involvement of certain compounds with specific chemical and physical properties and reactions against the plant immune system, are surprisingly valid.
As M. truncatula belongs to the inverted repeat-lacking clade (IRLC) of legumes where bacteroids undergo terminal differentiation, we expected similar forms of bacteroids in these legumes. However, in selecting 10 legumes representing different subclades of the IRLC, we found that the size and shape of the bacteroids were unexpectedly different, with swollen, elongated, spherical and branching morphologies. Identifying NCRs expressed in these nodules revealed significant differences in the complexity of the NCR families. The simplest, less altered, swollen morphology correlated with the presence of only 7 NCRs, the spherical form with more than 200, and the elongated and branched morphology with 300 – 600 NCRs, respectively. NCR families with fewer members were found to be composed of acidic and neutral peptides, while the large NCR families also contained positively charged cationic peptides and the drastic morphological alterations of bacteroids correlated with their more recent evolution.
For the NCR functions, it was hard to imagine that out of the 700 NCR genes, spread in the M. truncatula genome by gene duplication, some peptides would have unique essential functions. But they exist. We identified NCR peptides with unique characters by screening in plant mutants that were unable to form nitrogen-fixing nodules. Such a peptide is NCR169, for example. When NCR169 is absent, the differentiation of the bacteroids does not reach their final nitrogen fixation stage. An unexpected function of a few other NCR peptides was discovered in collaboration with US teams. These peptides control the bacterial strain selectivity in the nodules, which revealed a further control level of bacterium strain – host specificity.
We were particularly interested to find those NCRs that induce the first steps of bacteroid differentiation, which starts with cell division arrest and the elongation of bacteroids. Expression of NCR247 was detected exactly in those nodule cells where bacteria stop dividing and begin to elongate. We were able to demonstrate that these small cationic peptides enter the bacteria, and we identified their bacterial targets. The NCR247 peptide, due to its physicochemical properties, has an extremely strong protein-binding ability, and is capable of interacting with many bacterial proteins, including FtsZ, a conserved regulator of the bacterial cell cycle. The binding of NCR247 to FtsZ inhibits polymerization of FtsZ, which is essential for septum formation and the division of bacteria. Another peptide, NCR035, produced in the same symbiotic cell, interacts directly with the septum of cultured bacteria, indicating that the plant host has evolved several NCR peptides to attack the same biological pathway at different points, thus making it impossible for the bacteria to escape from the plant governance and to regain their cell-division ability. NCR247 has many other bacterial targets, and affects gene expression and protein translation, particularly by its interaction with numerous ribosomal proteins and by altering the composition of the bacterial ribosomes.
NCR Peptides: Opening to Application
In 2006, based on a French-Hungarian intergovernmental agreement, there was a call for the creation of a new biotechnology institute in Szeged. With the permission and support of the CNRS, I applied, and became the director of the BAYGEN Institute, while also maintaining my group leader position at CNRS. For several years I travelled between France and Hungary weekly, and supervised the work at both locations.
BAYGEN belonged to the Bay Zoltan Applied Research Foundation. Therefore, the research in our institute was more application-oriented. NCR peptides represent a treasure trove of novel, hitherto unexplored biological activities, and as we knew that certain NCRs can inhibit bacterial cell division, we investigated their antimicrobial activities in order to develop new antibiotics from them. In 2010, based on our preliminary results, I applied for an Advanced Grant of the European Research Council (ERC) with «SymBiotics», a project exploring the role of peptides in symbiosis and beyond, in therapies and agriculture. In December 2010, official notification from the ERC of my having been awarded the grant arrived, but a few days later the Hungarian government decided to close down all research-foundations in Hungary, including my host organization. Consequently, I had no host institution for my prestigious ERC grant. Fortunately, ERC grants are portable. Nevertheless, almost a year passed before I could start my ERC research project in the Biological Research Centre of the Hungarian Academy of Sciences in Szeged. It was during these turbulent times, in January 2011, that my husband and my source of inspiration passed away. In the years that followed, the ERC provided me with optimal conditions, scientific freedom and encouragement to conduct my research. My ERC project resulted in breakthrough discoveries in symbiosis, and we also developed excellent new antibiotics candidates from NCR peptides that effectively kill pathogenic bacteria and fungi, including those which cannot be eliminated with classic antibiotics. A further significant advantage of the tested NCRs and their derivatives is that they are non-toxic for human cells. The vast value of symbiotic peptides remains to be explored, and I anticipate that the Balzan research project will contribute a great deal to this field.
The Balzan Prize and Research Project
I am extraordinarily honoured by the Balzan Prize, which acknowledges my scientific field and research community, and particularly the hard, excellent work of all my talented co-workers and students. Without Adam, I probably would not have known the beauty and excitement of research on symbiosis. We enjoyed our research and inspired each other. From my team, I would mention my closest colleague and friend, Peter Mergaert, who joined my group as a young postdoc from Belgium and led my former research group in Gif-sur-Yvette from 2013 onwards.
I wish that I could share this honour today with Adam and with my mother, who was always proud of me and highly supportive of my scientific endeavours. Fortunately, my daughter Fanny, her husband Andras and their lovely children will be around me in the years to come.
I look forward to the Balzan Research Project, and have high expectations for its success and impact. The Balzan Research Project aims to obtain insights in first-order questions concerning the structure-function relationships and the regulation of NCR genes. The project I am leading will involve a very motivated international team of young researchers and my close co-worker, Dr. Gabriella Endre.
As I already mentioned, freedom is particularly important for scientists, since it is indispensable for creativity and for great discoveries in research. To receive a prize with the name of Eugenio Balzan, an exceptional person, who stood for democratic values and freedom, means a great deal to me.