Cypress Canker Disease (CCD), also known as cypress bark canker, is a pandemic caused by the invasive necrotrophic fungus Seiridium cardinale (Wagener) Sutton and Gibson (Order: Xylariales, Family: Amphisphaeriaceae) affecting many Cupressaceae (Danti et al., 2013a). A native of California (Wagener, 1928, 1939; Della Rocca et al., 2011, 2013), this fungal pathogen has spread all around the world mainly through the commerce of infected plants (Della Rocca et al., 2019); this has subsequently caused severe loss and significant limitations to the cultivation of cypresses (genus Hesperocyparis and Cupressus), especially Cupressus sempervirens L. (Graniti, 1998; Danti and Della Rocca, 2017).

A genetic improvement program of C. sempervirens for resistance to CCD has been carried out in Europe since the 1970s (Raddi and Panconesi, 1981; Danti et al., 2011), due to this species' valuable role in the landscape and its multifunctionality, especially in Mediterranean countries (Xenopoulos et al., 1990; Della Rocca et al., 2007; Farahmand, 2020). As a result, a series of canker-resistant varieties of C. sempervirens have been patented (Panconesi and Raddi, 1990; Danti et al., 2006, 2013b).

In the breeding program of cypress genotypes, it was essential to study the metabolic, molecular, and anatomical mechanisms which underlie the resistance, or susceptibility, to the fungal disease. While the ultrastructural modifications in the phloem, cambium, phloem parenchyma, and ray parenchyma colonized by S. cardinale were described by Mutto and Panconesi, 1987, it has been observed that resistance of the common cypress to CCD relies on the tree's ability to produce a barrier that compartmentalizes the bark tissues invaded by the pathogen by building up a ligno-suberized boundary zone (LSZ), which is then followed by the formation of a necrophylactic periderm (NP) (Ponchet and Andreoli, 1990; Spanos et al., 1999). This is the result of the activation of metabolic pathways leading to the synthesis of suberin in bark tissues (Danti et al., 2018). Another host defense reaction against the pathogen (easily visible to the naked eye) is the more or less abundant emission of resin from the cankered bark (Danti et al., 2013a). This was found to be due to increased production of terpenoids and a qualitative change in their mixture in the bark tissues around the S. cardinale infection point (Achotegui-Castells et al., 2015, 2016).

Recently, Della Rocca et al. (2021) focused on infection-associated anatomical changes in the phloem of C. sempervirens. Specifically, the dynamic accumulation of polyphenolic parenchyma cells (PP cells) and the time course of phloem axial resin duct (PARD)-like structures that form during the initial phase of the infection were documented at the histological level using light microscopy. Notably, a faster and more intense reaction, from a quantitative point of view, occurred a few days after inoculation in a CCD-resistant cypress genotype compared to the susceptible one, in terms of the number of PP cells and PARD-like structures. Danti et al. (2018) showed that the host responses in terms of expression of suberin biosynthesis-related genes differ between bark wounding and bark wounding coupled with pathogen inoculation.

This work intends to increase the understanding of anatomical and cytological changes in the reaction in the bark tissues of common cypress induced by infection with the pathogen S. cardinale. Our study investigated the differentiation and genesis of PP cells and PARD-like structures through observations using Light (LM), Fluorescence (FM), and Transmission Electron Microscopes (TEM). Particular attention was paid to the synthesis of polyphenols in the PP cells and of resin in the PARD-like structures. LM investigations employed different dyes to clarify whether PARD-like structures actually synthesize terpenoids inside them and consequently can be considered as induced resin ducts.

This work investigated the genesis and the evolution of PP cells and PARD-like structures in Cupressus sempervirens in the phloem tissues after inoculation with the fungal bark pathogen S. cardinale. Their developments were examined at intervals, from 7 to 45 days after infection with the pathogen, at both anatomical and cytological levels via LM, FM, and TEM using different dyes and techniques. In Della Rocca et al. (2021), the involvement of both PARD-like structures and PP cells in the response of C. sempervirens to the necrotrophic bark pathogen S. cardinale had been found in both canker-resistant and -susceptible clones. This study evidenced a faster phloem response of the canker-resistant clone in terms of the number of PP cells and PARD-like structures formed 5 and 12 days after infection compared to the susceptible clone.

Resin production is one of the main mechanisms of defense in conifers (Krokene et al., 2008a). Conifer oleoresin is a complex mixture of volatile mono- (C10) and sesquiterpenes (C15), as well as non-volatile diterpene and resin acids. It accumulates at the wound site to counteract invaders (insects and fungi) in several ways: through some antimicrobial activity, by flushing the wound, and by sealing the injury (Phillips and Croteau, 1999; Trapp and Croteau, 2001; Zulak and Bohlmann, 2010).

As reported in Krokene et al. (2008a), Cupressaceae do not have pre-formed resin structures, while induced structures appear following a stimulus to the xylem (Sequoia, Sequoiadendron, Metasequoia) or the phloem (Cupressus, Hesperocyparis, Chamaecyparis, Cryptomeria). C. sempervirens develops PARD-like structures as a reaction to the infection of the bark canker agent S. cardinale (Moriondo, 1972; Ponchet and Andreoli, 1990; Della Rocca et al., 2021); its main external symptom is more or less intense resinosis by the cankered bark (Danti et al., 2013a). Although the association between resinosis and infection by pathogenic fungi has been reported (Madar et al., 1995; Spanos et al., 1999; Achotegui-Castells et al., 2015, 2016; Danti et al., 2018), the recent work by Della Rocca et al. (2021) has not definitively clarified the resin production site, perhaps because of the short duration of the experiment. For this reason, one of the goals of this research was to investigate in depth the hypothesis that PARD-like structures are the true sites where resin is produced by the host, whether as a reaction to a fungal pathogen invading the bark tissues or to wounds. Our observation, at first, indicated a quick reaction by the host to the infection/wound signal, after only 7 days from the injury stimulus. The PARD-like structures were produced not only at the IP but also 5 cm above and below the inoculation/wound point, thereby showing a “signal speed” of almost 1 cm per day (at least in the first phase of the stimulus). This is the first case reported of phloematic resin structure being remotely induced in the first days after a “traumatic event” (either pathogen infection or wound) in Cupressaceae. The presence of resin ducts in C. sempervirens barks affected by S. cardinale was previously observed in various works (Moriondo, 1972; Ponchet and Andreoli, 1990; Spanos et al., 1999; Achotegui-Castells et al., 2015, 2016; Danti et al., 2018). Madar and Liphschitz (1989) also reported the formation of traumatic resin ducts in C. sempervirens bark upon challenge by S. cardinale and Diplodia cupressi. The formation of resin ducts in the bark after fungal inoculation was also reported in C. obtusa (Yamanaka, 1989; Suto, 1998; Kusumoto and Suzuki, 2003; Fujii et al., 2018) and in C. macrocarpa, Cryptomeria japonica, Sequoiadendron giganteum, after stem treatment with methyl jasmonate (Hudgins and Franceschi, 2004). However, none of these studies reported the formation of traumatic resin ducts in the bark at a distance from the inoculation or wounding site.

Regarding the rapidity of the formation of phloematic resin ducts upon a traumatic event, a timing similar to that observed on C. sempervirens in this work is that reported for C. obtusa by Yamanaha (1989) and by Kusumoto and Suzuki (2003). In these two studies, the presence of phloematic resin structures was observed 7–15 days after wounding. All the other studies do not allow a direct comparison because they were not based on time series observations but on the detection of phloem resin ducts after arbitrary intervals of time, after infection, wounding, or treatment with methyl jasmonate. In this study, a lower number of PARD-like structures was also observed in the wounded ramets compared to the inoculated ones. Similar results were obtained by Luchi et al. (2005) in Pinus nigra, although in that case, they were xylematic traumatic resin ducts (TRDs). In that study, the non-aggressive fungal canker agent, Diplodia scrobiculata, artificially inoculated into the stem, induced in 12 days a higher number of TRDs per surface unit compared to those caused by mock inoculations (with sterile agar). In addition, Luchi and fellow authors observed that TRD numbers declined as the sampling distance increased from the IP. One might however speculate that the anatomical modifications (i.e., the induction of resiniferous structures) at the phloematic level (as in C. sempervirens) might be a little faster than at the xylematic level (as in Pinus nigra). The remote induction of TRDs has been also observed in Picea abies by Franceschi et al. (2000), Nagy et al. (2000) and Krekling et al. (2004). Surprisingly, the serious canker agent Sphaeropsis sapinea was unable to induce TRD in the xylem of P. nigra (Luchi et al., 2005). This might indicate that the induction of TRDs may be a result of a transmittable signal that may relate to perception by the host of fungus or fungal activity.

Due to the degeneration of some phloem elements and ray parenchyma, in this study, we also observed, 45 days after inoculation, a sort of tangential coalescence of neighboring PARD-like structures, which thus formed larger cavities. Coalescence of neighboring PARD-like structures had already been described by Suto in Chamaecyparis obtusa 1 month after inoculation with the fungal pathogen Cistella japonica (2005). A similar process was found in Thuja plicata by Cleary and Holmes (2011), although with the difference that in T. plicata, elliptical zones of phellem-like cells in degeneration and necrosis were observed around the longitudinal phloematic resiniferous ducts, which does not occur in the phloem of Cupressus.

During the activity of PARD-like epithelial cells, different phases could be observed at the same moment in adjacent cells, which showed a different electron-dense cytoplasm. The increased electron density was probably due to a greater number of ribosomes involved in protein synthesis. Similarly, in Picea abies, the epithelial cells of xylem TRD exhibited a greater cytoplasmic density, a high number of plastids, and enlarged nuclei, as observed by Nagy et al. (2000). Apart from the vacuole, the organelle mostly involved in the synthesis of terpenes was the chloroplast which was almost surrounded by a smooth ER; therefore, the formation of terpenes was to be related to the interaction between chloroplast and endoplasmic reticulum. The produced terpenes tended to get out of the cytoplasm by passing through the plasmalemma apparently without the need for specific transport mechanisms. Terpenes in the form of gray bodies were also observed between the plasmalemma and epithelial cell wall. We propose here a possible model of terpenoid formation (Figure 6) starting from plastids, with four sequential stages (Figures 6A–D), based on the observation of gray bodies via TEM. The involvement of the ER in the modification of terpenes had already been proposed by Berthelot et al. (2012), even if without any evidence of an ER-plastid spatial colocalization, while Mehrshahi et al. (2013) showed the importance of ER in modifying the terpenoid tocopherol (with fatty acid desaturation) produced by chloroplasts. The constant presence of mitochondria even in the advanced stages of PARD-like epithelial cell maturation shows that the production of terpenes is an active mechanism that occurs with ATP consumption. Finally, in PARD-like structures, there are also cellular residues derived from the degeneration of fibers and phloem cells, that are being lysed to allow the enlargement of the PARD-like structure lumen via wall digestion. The involvement of mitochondria in the last phases of PCD, in general, and in plants, in particular, has been often observed as a consequence of an active PCD process (Adrain and Martin, 2001; Yao et al., 2004; Brighigna et al., 2006).

The content of the PARD-like structure in C. sempervirens was confirmed to be lipophilic (via Sudan red and Fluorol Yellow positivity) and more specifically composed of terpenoids, as showed by NADI staining positivity and the autofluorescence with UV leading to blue light emission, which may be due to diterpenes and/or oleoresins (Donaldson and Williams, 2018; Donaldson, 2020). About the nature of the terpenes contained in the PARD-like structure, Li et al. (2021) tried to link a specific intensity of NADI staining to a particular class of molecules belonging to terpenes. The topic has not yet been well explored, however, in Cupressus, the NADI staining intensity here observed in the PARD-like structures appeared to be more similar to that corresponding to terpenes similar to (E)-β-farnesene found in pyrethrum, even if the exact identification of the specific molecule is not possible with histochemistry.

The PP cells have been identified as another form of defense of C. sempervirens as well as a potential proxy of resistance to CCD (Della Rocca et al., 2021), demonstrating that in C. sempervirens, new PP cells can originate from albuminous cells through the accumulation of phenolic substances in the vacuole, as noted by Fahn (1990), Krekling et al. (2000), and Krokene et al. (2008b) in Picea abies. Their origin, however, had never been investigated before at an ultrastructural level. Although Della Rocca et al. (2021) observed a decrease of PP cells during the formation of the PARD-like structures, probably because some PP cells had lost their vacuolar polyphenol content, similarly to what was observed in Picea abies by Nagy et al. (2000). In our study, we observed a few albuminous cells starting a differentiation process 5 days after infection, and then forming what would then become the PARD-like epithelial cells. These cells were very active, with some rounded plastids that formed intra-plastid lipid bodies. Other albuminous cells showed polyphenol accumulations in the middle of the vacuole. TEM observations clarified the ultrastructure of the PP cells in C. sempervirens, better distinguishing their genesis. The albuminous cells exhibited low electron-dense cytoplasm; the plastids were located at the periphery of the cytosol, in contact with peroxisomes and mitochondria. In these cells, the plastids were particularly active, showing thylakoids arranged into grana, more plastoglobules (defined as in Van Wijk and Kessler, 2017), and with less accumulated starch compared to PP cells during formation. The constitutive PP cells showed a large vacuole that occupied almost the entire cell volume in which electron-dense material was identified as polyphenols. In the periphery of these cells, the cytoplasm was very dense and we recognized dictyosomes (whose function is to secrete vesicles, which generally contain carbohydrates or glycoproteins, often outside the cell) and plastids with plastoglobules. Plastids are involved in the production of polyphenol precursors, and the key enzyme of their biosynthesis has been localized on the chloroplast starch granules (Grundhöfer et al., 2001). TEM observations showed that the plastids observed in the PP cells of the common cypress contained large starch granules, similarly to what Krekling et al. (2000) observed in PP cells of Picea abies. The same authors also detected the presence of lipids, hypothesizing that both compounds constituted an energy reserve for a rapid synthesis of defense compounds (Krokene et al., 2008a). Within PP cells, some peroxisomes, probably involved in lipid metabolism, were also visible. The newly induced PP cells showed vacuoles with a higher electron density and granular material inside compared to the constitutive PP cells. In more advanced stages of maturation, instead, PP cells showed an increased amount of polyphenols in the vacuoles, suggesting that the host defense response against the infection could be related to the total quantity of polyphenols accumulated around the affected area.

In C. sempervirens, both PP cells and resins producing cells in PARD structures can be induced as a response to an attack by a pathogen. Both these cell types develop out of albuminous or parenchyma cells in the secondary phloem layer of the stem's secondary structure.

Plastids appear to be involved in the development of the PP cells, with a phase of starch accumulation and a later stage in which the majority of the cells are filled by a vacuole almost completely occupied by polyphenols.

The resin-producing cells surrounding the PARDs secrete terpenoids toward a large lumen formed by programmed cell death of albuminous cells and parenchyma cells around sclerenchymatous fibers. Terpenoids are produced inside the cells after the ER surrounds the plastids, then the secretion occurs directly through the plasma membrane toward the lumen.

This study showed the production of terpenoids in the lining cells surrounding the lumens which formed in the phloem of C. sempevirens following both infections with S. cardinale or mechanical injury. These findings suggest that the structures previously defined as PARD-like can be considered traumatic resin ducts of the bark (BTRD).

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

GD, AP, and RD conceived the idea and designed the methodology. CT and AP performed the microscope observations. SB and IP carried out the inoculations, sampling, and preparation of samples for microscope observations. GD, AP, RD, SB, and SM analyzed the data and wrote the manuscript. All authors contributed critically to the drafts and gave final approval for publication.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.