Rapid Movement of ‘Candidatus Liberibacter asiaticus’ on ‘Hamlin’ Sweet Orange and ‘Swingle’ Citrumelo Trunks

Huanglongbing (HLB), caused by phloem-limited ‘Candidatus Liberibacter asiaticus’ (CLas) and other closely related species, is the primary limiting factor of production in most citrus regions of the world (Arratia-Castro et al. 2014; Bové 2006; Chen et al. 2010; Teixeira et al. 2008). In Florida, citrus production declined 74% for all citrus varieties compared with pre-HLB production (USDA 2021). Before 2004, CLas was only present in Asia, where the disease was first described in 1919 (Reinking 1919). In the Americas, HLB was reported first in Brazil in 2004 (Texeira et al. 2005), followed by the United States in 2005 (Halbert 2005).

Characteristic canopy symptoms of HLB are blotchy mottle, sectored yellowing, leaf drop, twig dieback, preharvest fruit drop, and low fruit quality. Fruit quality reduction includes small misshapen fruit, reduced sugar content, and bitter flavors (Bassanezi et al. 2009; Bové 2006; Halbert and Manjunath 2004). A less obvious early symptom of disease is a 30 to 50% root loss observed before canopy symptom development (Johnson et al. 2014). This root loss corresponds to rapid infection of the root system shortly after initial canopy infection by psyllids or grafting with infected tissue (Graham et al. 2013; Johnson et al. 2014) and shortened lifespan of HLB-affected roots (Pulici et al. 2022). Field and greenhouse observations demonstrated infections throughout the plant root system, whereas the canopy infection remained sectored in one or a few branches (Gast and Russo 2014; Johnson et al. 2014; Louzada et al. 2016).

After infection, CLas is transported systemically throughout the phloem tissue following the source-sink movement of phloem sap and symplastic connections (Kim et al. 2009; Raiol-Junior et al. 2021a; Su 2008). The primary movement of water and organic compounds in most vascular tissue is vertically through the tree, connecting branches with roots (Vasconcellos and Castle 1994). However, phloem movement is not limited to terminal sinks but also occurs in lateral sinks (De Schepper et al. 2013). Therefore, CLas could move in both ways, vertically and horizontally, in the phloem of infected trees (Gast and Russo 2014). The highly pleomorphic morphology of CLas, with spherical and rod shapes observed in multiple citrus species, might contribute to possible lateral movement (Achor et al. 2020; Shokrollah et al. 2010; Su 2008).

The unique interactions within the tree are known to occur at the graft union and the crown. The graft union offers a unique site for development of vascular tissue as two genetically distinct trees connect their phloem to become one tree (Reuther et al. 1968; Williamson et al. 1992). The crown is a site where unique interactions occur as vascular tissue differentiates between root and trunk (Reuther et al. 1968). This initial alignment and fusion of the vascular tissue could play a role in vertical and lateral movement as all carbohydrates are funneled through this union. Although bacteria may cross laterally through the phloem cells, it is unknown if CLas movement is influenced by the distance of the inoculum source or the anatomical structures through which the vascular tissue passes (Achor et al. 2020; Su 2008).

Splitting tree roots and trunks can physically separate any lateral phloem connections, allowing for the study of lateral movement through the vasculature of the trunk. Inoculating trees split up to different heights through the crown and graft union should provide insight into whether specific tissues or vertical distance is important in the down and lateral movement of CLas based on differential detection of bacteria in the dual root systems. Better understanding of the initial distribution of CLas within a citrus tree could help assess new strategies to control HLB and give insight into targeted therapies. The objectives of this study were (i) to investigate movement of CLas between sieve tubes and (ii) to understand if the seasons can affect the speed of CLas movement.

MATERIALS AND METHODS

Plant materials and rhizoboxes

One hundred and twenty trees composed of 60 ‘Swingle’ citrumelo (seedling) trees and 60 ‘Hamlin’ sweet orange grafted on ‘Swingle’ citrumelo trees were used in each experimental trial. Seedling trees were raised from seed grown in greenhouses at the Citrus Research and Education Center, Lake Alfred, and were approximately 2 years old. The grafted trees were purchased from Southern Citrus Nurseries, Dundee, Florida, were approximately 3 years old, and were larger than the seedling trees.

The trees were divided into three groups separated by the height of the split. The grafted trees had the tap root and trunk split to below the crown (Group 1 or Graft Root); split to approximately 5 cm above the crown (Group 2 or Graft Low); and split through the graft union to approximately 24 cm above the crown (Group 3 or Graft High). The seedling trees were similarly divided with tap root and trunks split to “Seed Root,” “Seed Low,” and “Seed High” (Fig. 1). Seed High were split to approximately 16 cm above the crown. Seed High were split lower than the Graft High due to the size differences of the trees. Roots and trunk were split with a sterilized knife, and each half was wrapped with grafting tape to speed healing and prevent the separated sides from fusing and rejoining. Trees were given at least 3 months to heal, keeping the two root systems separate in different conical pots. Conical pot dimensions for seedling trees were approximately 21 cm in depth with a diameter of 4 cm. Pots used for grafted trees were 25 by 6.5 cm, respectively.

Download as PowerPoint

For both trials, rhizoboxes were used for observation of root growth and minimal-disturbance root sampling. Rhizoboxes were constructed with a wooden frame (52 cm height × 41 cm width × 3 cm depth) with clear acrylic (52 cm length × 41 cm width) sheets screwed onto the flat width side of the wood. The rhizobox frame included a wood divider (41 cm height × 3 cm width × 3 cm depth), keeping the root systems separated. Rhizoboxes were designed for easy sampling of roots without disturbing the root system by removing and replacing one acrylic panel using screws (1.5 cm size) on each piece of wood. A thin layer of clear silicon caulking adhesive was added between the acrylic and wood to prevent soil loss. The bottom piece of the frame was drilled with six holes and wrapped in landscape cloth to allow water drainage while preventing soil loss. Candler fine sand soil obtained from a grove site in Polk County on the “Ridge” with a small particle size, typical for the region (USDA 1985), was used to simulate field root-growing conditions. The soil was autoclaved twice to reduce risk of contamination by Phytophthora spp. and nematodes. The rhizoboxes were kept in a wooden box to exclude light from the roots and suppress algal growth.

Trees were transplanted to the rhizoboxes at 3 and 9 months after root splitting for the first and second trials, respectively. Plants were given 1 week to establish in the rhizoboxes and produce new root growth. Plants were watered three to four times weekly and received a slow-release fertilizer application, Harrell's 18-5-10 (Harrell's, Lakeland, FL). Budwood from healthy (controls - CLas(-)), and CLas-infected (HLB) ‘Pineapple’ sweet orange trees were used for graft inoculation into one side of each tree 5 cm above the split trunk (Fig. 1). The side that received the graft was designated the “inoculated side.” Five trees per split group served as healthy controls.

Trees were grown inside a greenhouse, and environmental conditions were not recorded. However, air temperature and relative humidity were recorded at the Lake Alfred weather station (station ID 330), which is part of FAWN (Florida Automated Weather Network) and located 0.65 km away from the greenhouse. Air temperature varied from 7.1 to 35.9°C, and relative humidity varied from 60.1 to 90.3% (averages 23.3°C and 84.2%, respectively) during the first trial and varied from 7.6 to 36.6°C and 63.4 to 88.4% (averages 22.4°C and 76.5%, respectively) during the second one (Fig. 2).

Download as PowerPoint

In the first trial, roots were sampled from the inoculated and noninoculated sides weekly for the initial 18 weeks and then biweekly. All 240 root samples, 120 from each side, were DNA extracted every 4 weeks. If samples tested CLas(+), intermediate time points were extracted to determine the time of initial CLas presence. In the second trial, trees were sampled from both sides weekly for 15 weeks, and then biweekly. The roots were DNA extracted every 6 weeks, and when they tested positive, intermediate samples were extracted to determine the time of initial CLas presence. The time frame for group extractions was increased to reduce the sample processing required. This sampling pattern reduced the total number of DNA extractions and qPCR assays required without compromising the resolution. Colonization of roots by CLas was confirmed when two consecutive samples were positive by qPCR, which in this study was considered CLas positive when the CLas population was ≥10 cells g⁻¹ of root tissue.

DNA extraction and qPCR

For root DNA extraction, a 25-mg sample of root tissue was chopped to approximately 2 mm, using a new razor blade for each sample, and placed in a 2.0-ml Conical Screw Cap Tube (Fisher Scientific). A sterile 5-mm stainless steel bead (Qiagen) was added to each tube and placed in a −80°C freezer for at least 2 h. Samples were then ground in a Tissuelyzer II (Qiagen) with a frequency of 30 Hz for 30 s. Samples were ground twice, with blocks rotated to ensure an even grind. DNA was extracted with the PowerSoil DNA extraction kit (Mo Bio). The kit protocol was modified to where instead of transferring the sample to a PowerSoil Bead Tube, the contents (garnets and buffers) of the PowerSoil Bead Tube were added to a 2.0-ml Conical Screw Cap containing the ground sample to limit sample loss. After C1 buffer was added, tubes were vortexed for 10 s instead of 10 min, and the manufacturer's protocol was continued. After extraction, samples were kept in a −20°C freezer for short-term storage until PCR was run and at −80°C for long-term storage.

Real-time qPCR was performed using previously published primers: forward primer CQULA04F 5′-TGGAGGTGTAAAAGTTGCCAAA-3′, reverse primer CQULA04R 5′-CCAACGAAAAGATCAGATATTCCTCTA-3′, and probe CQULAP10 5′-ATCGTCTCGTCAAGATTGCTATCCGTGATACTAG-3′ (Wang et al. 2006) on an Applied Biosystems 7500 Fast Real Time PCR system using HotStarTaq Master Mix Kit DNA polymerase (Qiagen), ROX passive dye (Bio-Rad), and template DNA. Reactions used a total volume of 20 μl of the following: 2× Taq DNA Polymerase Master Mix (Qiagen) 8.25 μl, sample DNA 1.0 μl, ROX Passive Dye (Bio-Rad) 0.8 μM, forward primer 0.8 μM, reverse primer 0.8 μM, probe 0.2 μM, and H2O was added to bring the total volume to 20 μl. All samples were run in duplicate. CT values were converted to CLas genomes using a standard curve of the pLBA2 plasmid (Trivedi et al. 2009) kept in a 50% glycerol stock solution ranging from 101 to 106 copies of target DNA, equivalent to the number of bacteria. Thermocycling conditions were (i) 2 min at 50°C, (ii) 10 min at 95°C, and (iii) 50 cycles of 95°C for 15 s, followed by 45 s at 60°C.

Statistical analysis

For data on the time of detection and weeks between first detection of CLas on the inoculated side and noninoculated side, no statistical analysis was performed. Data on the quantity of bacteria were transformed logarithmically to meet assumptions of normality and then analyzed by two-way (distance of inoculum source × anatomical structures) or one-way (anatomical structures; time of detection) ANOVA using PROC GLM (SAS v. 9.3). Means separation was performed using LSD with P ≤ 0.05.

RESULTS

Detection of CLas in split root trees

To observe if the distance of the inoculum source and anatomical structures can affect CLas movement, as well as the effects of seasonality on the speed of this initial movement into the trees, Hamlin sweet orange grafted on Swingle citrumelo trees and Swingle citrumelo (seedling) trees were graft-inoculated with CLas-infected or healthy (control) budwoods in two seasons, summer and winter. Roots from each side of the split root system were then collected for CLas detection and quantification by qPCR.

In the first trial, plants were inoculated during the summer (July), and 43 of the 90 HLB-inoculated trees tested qPCR positive for CLas on at least one side (Fig. 3A, C, and E). CLas(+) trees were detected at 5 weeks postinoculation (wpi) for the grafted trees and 4 wpi in the seedling trees. Thirty-two trees tested CLas(+) on the inoculated side first, of these, two remained CLas(-) on the noninoculated side until the end of the experiment (40 wpi). Eight trees tested positive on the noninoculated side followed by the inoculated side. In these trees, five remained CLas(-) on the inoculated side until 40 wpi. Three trees tested positive on both sides at the same time (Fig. 3A, C, and E). From the 43 CLas(+) trees, 34 tested CLas(+) on both sides within 6 weeks after first detection (summer season) (Fig. 4). Of the initial 120 trees (HLB and control trees), 14 died.

Download as PowerPoint

Download as PowerPoint

In the second trial, plants were inoculated during the winter (November). From the initial 90 HLB-inoculated trees, 36 trees tested CLas(+) on at least one side (Fig. 3B, D, and F). The bacterium was detected at 14 wpi in grafted trees and 16 wpi in seedling trees. Nineteen trees were CLas(+) on the inoculated side first; of these trees, nine remained CLas(-) on the noninoculated side until 36 wpi. Twelve trees tested positive on the noninoculated side first. Of these trees, nine remained CLas(-) on the inoculated side until the end of the experiment, at 36 wpi. Five trees tested positive on both sides at the same time (Fig. 3B, D, and F). Of the 36 trees CLas(+), 15 tested CLas(+) on both sides within 6 weeks after first detection (spring season) (Fig. 4). No trees died during this replication.

In both trials, there was a similar number of CLas(+) grafted and seedling trees at the end of the experiment, at least on one side of the split root: 24 and 21 grafted trees (in the first and second trials, respectively) and 19 and 15 seedling trees in the first and second trials, respectively (Fig. 3). However, plants infected in the summer had 84% CLas(+) on both sides, whereas trees infected in the winter had 50% CLas(+) on both sides (Fig. 3).

Trees had similar CLas populations. However, the CLas population increased from the initial time point of detection (Table 1).

TABLE 1 Populations of ‘Candidatus Liberibacter asiaticus’ (CLas) on inoculated and noninoculated sides of the split root system of ‘Hamlin’ sweet orange grafted on ‘Swingle’ citrumelo trees and 'Swingle' citrumelo (seedling) trees at the first and second time of detection

	Inoculated sidex				Noninoculated sidex
	First detection		Second detection		First detection		Second detection
Treatment	ny	Log ± SEz	ny	Log ± SEz	ny	Log ± SEz	ny	Log ± SEz
First replication
Graft high	7	2.10 ± 0.36	7	3.77 ± 0.29	6	2.26 ± 0.25	5	2.26 ± 0.25
Graft low	7	2.08 ± 0.43	7	3.11 ± 0.42	8	2.35 ± 0.20	8	3.01 ± 0.23
Graft root	7	2.05 ± 0.29	7	2.98 ± 0.32	9	2.02 ± 0.23	9	2.54 ± 0.39
Seed high	7	1.88 ± 0.29	7	2.86 ± 0.51	7	1.86 ± 0.19	7	2.46 ± 0.23
Seed low	5	1.93 ± 0.19	5	2.57 ± 0.15	5	1.50 ± 0.08	5	1.90 ± 0.16
Seed root	6	2.28 ± 0.07	6	2.57 ± 0.48	6	1.95 ± 0.24	6	2.40 ± 0.34
P		0.9646		0.3006		0.1588		0.0728
Graft	21	2.07 ± 0.20 B	21	3.28 ± 0.20 A	23	2.20 ± 0.13 aB	23	2.76 ± 0.24 A
Seed	18	2.03 ± 0.13 B	18	2.68 ± 0.25 A	18	1.79 ± 0.11 bB	18	2.28 ± 0.15 A
P		0.8474		0.0646		0.0276		0.1193
		≤0.0248				≤0.0148
Second replication
Graft high	6	2.72 ± 0.58	6	3.19 ± 0.46	2	2.22 ± 0.93	2	2.86 ± 0.10 b
Graft low	8	2.53 ± 0.27	8	2.96 ± 0.39	8	2.70 ± 0.31	8	2.93 ± 0.44 b
Graft root	4	2.45 ± 0.42	4	2.58 ± 0.61	5	2.59 ± 0.43	5	2.81 ± 0.51 b
Seed high	4	2.03 ± 0.53	4	4.31 ± 0.52	5	2.37 ± 0.58	5	4.44 ± 0.12 a
Seed low	3	2.15 ± 0.46	3	3.25 ± 0.48	3	2.24 ± 0.61	3	4.58 ± 0.76 a
Seed root	2	1.52 ± 0.25	2	2.54 ± 0.41	4	2.01 ± 0.32	4	2.88 ± 0.29 b
P		0.7141		0.2895		0.8916		0.0368
Graft	18	2.57 ± 0.23	18	2.95 ± 0.16	15	2.60 ± 0.23	15	2.88 ± 0.28 b
Seed	9	1.96 ± 0.27 B	9	3.56 ± 0.36 A	12	2.22 ± 0.28 B	12	3.95 ± 0.30 aA
P		0.1224		0.1851		0.2988		0.015
		0.0029				0.0003

^xSides of split roots from of ‘Hamlin’ sweet orange grafted on ‘Swingle’ citrumelo trees and ‘Swingle’ citrumelo (seedling) trees.

^yNumber of individual CLas(+) trees on each side of the root system.

^zValues are the mean Log₁₀ of colony-forming units per gram of tissue ± standard error. Different lower-case letters in the same column are significantly different from each other at P ≤ 0.05. Different capital letters in the same row are significantly different from each other at P ≤ 0.05.

View as image HTML

DISCUSSION

The phloem, the principal sugar conductive tissue in plants, interconnects the whole-plant level by long-distance transport (De Schepper et al. 2013) and is also the known pathway of CLas within the tree (Raiol-Junior et al. 2021a). The phloem connects not only the plant body but also CLas structures into decentralized organs, such as leaves, roots, stem, and branches. This determines if the vertical distance of the inoculum source, as well as different anatomical structures (grafted and seedling trees), can affect the speed of the CLas movement within the plant, which is important to understand the process of host plant colonization.

The time for CLas to be detected in the roots was not affected by the distance of the inoculum source nor seedlings or grafted trees. It was similar among all combinations in both trials (Fig. 3). Vertical is the most common movement in vascular trees (De Schepper et al. 2013). Recent studies suggested that the vertical movement is also the primary direction of CLas (Graham et al. 2013; Johnson et al. 2014; Pulici et al. 2022) as observed in Hamlin sweet orange grafted on Cleopatra mandarin rootstock or grapefruit trees on sour orange rootstock (Braswell et al. 2020) and in Valencia sweet orange grafted onto sour orange (Louzada et al. 2016). That corroborates our results, as we observed most of the plants testing CLas(+) on the inoculated side first (Fig. 3).

Additionally, bacterial titer was similar among all combinations and on both sides. However, the CLas population increased from the first to the second time of detection (Table 1). These results suggest that after CLas is present in the roots, the bacterium needs time for multiplication, supporting the hypothesis that roots act as a CLas reservoir (Johnson et al. 2014; Pulici et al. 2022).

The environmental conditions when the plants were infected affected the speed of vertical and lateral CLas movement. Trees inoculated in the summer, which had higher temperature (average of 28°C during the first 10 wpi), had fast and uniform movement; 48% of HLB-inoculated trees became infected. CLas started to be detected in the roots 4 wpi and could be detected on the inoculated and noninoculated sides in 84% of CLas(+) trees. Plants inoculated in the winter (average temperature of 19°C in the first 10 wpi) had a slow and uneven movement. The plants had a lower infection rate (40%), and CLas started to be detected in the roots 14 wpi (spring season) and could be detected on both sides of the root system in only 50% of CLas(+) trees.

Environmental conditions, in particular temperature, are important factors for plant development and can impact flushing dynamics and the extension growth of roots (Hall and Albrigo 2007; Schneider 1952). Temperature also influences CLas movement inside the plant, which is preferential to newly developed citrus sinks (Hilf and Luo 2018; Lemoine et al. 2013; Raiol-Junior et al. 2021a). This was observed in our results: Warmer temperatures shorten the time for CLas to be detected in the roots.

CLas movement is also highly influenced by the seasonality of the emission of new flushes (roots and shoots) (Pulici et al. 2022; Raiol-Junior et al. 2021b). Citrus trees have an extended dormant period in the winter, during which there is neither shoot nor root growth (Schneider 1952). Dormancy could delay CLas colonization, providing an explanation as to why CLas was only detected in the roots during the spring season, one of the major shoot and root extension times in Florida (Hall and Albrigo 2007; Zekri et al. 2008). This hypothesis, however, requires further investigation.

In this work, we attempted to gain insight into how CLas moves within the plant, a poorly understood phenomenon that has only recently been receiving attention. The data demonstrated strong effects of seasonality on this movement. Summer and spring seasons not only appear to be the major root growth extension times (Hall and Albrigo 2007) but also the peak for CLas movement independent of the distance of the inoculum source or plants’ anatomical structures. We evaluated graft-inoculated (rhizobox) plants under greenhouse conditions, where Diaphorina citri (CLas vector) pressure is low or absent and reinfection events are not as common as in the field, which could result in different CLas responses. Both hypotheses warrant further investigation into environmental conditions under both greenhouse and field conditions to better define how each of these factors affects pathogen movement.

ACKNOWLEDGMENTS

The authors thank Sara Commerford, Diane Bright, Harrison Davis, Tony McIntosh, Jian Wu, and Alex Dunn for technical assistance. This research is part of James Orrock's thesis.