"Citrus Huanglongbing is a pathogen-triggered immune disease". How did we get there?

"Citrus Huanglongbing is a pathogen-triggered immune disease". How did we get there?

Now our manuscript has been accepted, it is time for a retrospective in our recent discovery that citrus Huanglongbing (HLB, also known as citrus greening) is a pathogen-triggered immune disease. The work on HLB has been a struggle since I started to work on it in 2007. As a bacteriologist and plant pathologist, we can not do the most basic and favorite thing: making a mutant of the pathogen Candidatus Liberibacter asiaticus (CLas). My thinking of how the HLB pathogen causes damage to citrus has been influenced greatly by other phloem pathogens such as Phytoplasma. Phytoplasma, a member of Mollicutes, secretes effectors to manipulate plant symptoms via the Sec pathway. Does CLas secrete effectors to cause HLB symptoms? If so, how? We know CLas does not contain the type II secretion system that is usually required to translocate effectors across the bacterial outer membrane, whereas Phytoplasma is without the outer membrane. Maybe CLas can use the flagellar type III secretion system? Outer membrane vesicles? Without genetic mutation tools at hand, we have applied approaches to test whether CLas uses pathogenicity factors to cause HLB symptoms: 1) we have expressed CLas putative Sec-dependent effectors using a TMV expression vector and expressed them in Nicotiana benthamiana; 2) we have conducted transgenic expression in citrus. Surprisingly, multiple effector candidates caused strong symptoms when expressed using the TMV vector. Are they all pathogenicity factors? The symptoms are so similar to viral diseases. I have decided to wait for our transgenic results. Even more surprisingly, none of the effector candidates causing symptoms in N. benthamiana caused symptoms in transgenic citrus, Arabidopsis, or N. tabacum. There is one special occasion that one of the transgenic citrus plants showed some abnormalities in the early stage, which we could not repeat and remains a mystery (maybe caused by Phytophthora effect on the roots in our greenhouse). We did all these with the mindset that plant diseases are caused by pathogenicity factors as we have been familiar with. None of the effector candidates worked. Maybe we missed the right one? We tested, 5, 10, 15, 20…still nothing. We were stuck in this stalemate situation with no light at the end of the tunnel for a long time.

In 2016, Annual Review of Phytopathology invited me to write a review on HLB. I reviewed all published HLB data thoroughly and noticed many of the early responses of HLB are related to immune responses, such as gene expression of PR genes, callose deposition……Is it possible that immune responses are the cause? It can’t be. I could not find any previous examples that immune responses are responsible for plant diseases. Don’t be crazy. I told myself. But, maybe we should give a try. Wenxiu and Zhiqian tested ROS production first in leaves. Intriguingly, the H2O2 concentration induced by HLB is comparable to that in Kumquat undergoing HR like symptoms induced by Xanthomonas citri subsp. citri. We then used an elegant method developed by Gitta lab to extract exudates of the phloem-enriched bark tissues. Wow, I was amazed by the high concentrations of H2O2 in HLB positive plants, which are way above the concentration that kills animal cells. Is it possible that CLas causes overproduction of reactive oxygen species (ROS) to kill phloem cells? Diann, who is one of the best TEM experts, and I analyzed hundreds of TEM pictures of phloem tissues. We noticed that cell death is among the earliest events after CLas infection, and it happens first in companion cells. Sheo further looked at young flushes. He observed ROS and cell death are the earliest events after CLas enters the newly emerged tissues, in addition to other immune responses such as PR gene expression and callose deposition. I felt that we have found the key for how CLas causes damages to citrus. Luckily, we have been working with citrus tissue culture and protoplast. Xiaoen found that the H2O2 concentrations detected in the phloem tissues are enough to kill citrus protoplast cells or tissue culture cells. Once we have got these results, I was so excited and stayed awake for several nights. The other data were pretty naturally generated including testing antioxidants and immunoregulator gibberellin led by Jinyun, and the large scale RNA-seq analyses by Jin. We have found the causative relationship that CLas causes systemic and chronic overproduction of ROS that cause cell death of phloem tissues, which in turn causes HLB symptoms. This is the first pathogen-triggered immune disease of plants. My colleagues and I were extremely thrilled by this conceptual advance about HLB and in plant pathogen field in general. We sent our manuscript to Science with high hope, rejected. We sent it to Nature, rejected, but Nature Editors had some positive comments: would have considered it had we have some genetic evidence. To make the long story short, we appreciate that Nature Communication Senior Editor César Sánchez and the reviewers recognized the value of our work, and helped us significantly improve the manuscript. With their suggestions, we further understood the contribution of RBOH genes to HLB development and how gibberellin suppresses HLB. We also thank Dr. Steve Lindow for critical reading of our manuscript during revisions.

I shall say more than half of the work were conducted during the pandemic although the work really started more than 10 years ago. I am so grateful that my colleagues (Wenxiu Ma, Zhiqian Pang, Xiaoen Huang, Jin Xu, Sheo Shankar Pandey, Jinyun Li, Diann S. Achor, Fernanda N.C. Vasconcelos, Connor Hendrich, Yixiao Huang, Wenting Wang, Donghwan Lee, Daniel Stanton) were able to move the work forward diligently during the difficult time. We are continuing our journey to explore the genetic determinants for HLB being a pathogen-triggered immune disease and understand the signaling pathway. We have finally been able to generate homozygous or biallelic citrus mutant at decent efficacies (up to 89%) to make it possible. I look forward to further uncovering the mystery of HLB and developing solutions to HLB together with my colleagues.