Effects of ‘Candidatus’ Liberibacter Asiaticus on the Root System of Poncirus trifoliata Hybrids as a Rootstock for ‘Valencia’ Scion

Abstract

The symptoms of huanglongbing (HLB), a disease caused by the bacterium Candidatus Liberibacter asiaticus (CLas), are visible on the canopy of citrus plants. They include mottling of leaves followed by dropping and lopsided fruits with premature dropping. Loss in phloem functionality and degradation of the root system are also HLB symptoms with a severe impact on plant growth and production. Some Citrus relatives, such as Poncirus trifoliata and its hybrids, have shown more tolerance to HLB disease and low titers of CLas compared to Citrus species, but little is known about the effects of CLas on their root system. In this study, we investigated the effects of CLas-infected ‘Valencia’ scion on the citrandarin IAC3222 (a hybrid between P. trifoliata and Sunki mandarin) used as rootstock as well as interstock between ‘Valencia’ and Swingle citrumelo rootstock. At 13 months post-inoculation, the cycle threshold values (C_T) for CLas in the infected scion samples indicated a high CLas titer (from 15.9 to 22.7) regardless of the rootstock variety or interstock used. However, no CLas-positive samples were detected in the roots of IAC3222 (C_T ranging from 37.9 to 40.0), in contrast to all Swingle roots (C_T ranging from 27.9 to 31.3). Both root volume and mass were reduced in IAC3222 compared to uninfected ‘Valencia’ scion, suggesting that scion infection damages roots, regardless of whether they are contaminated or not by CLas. The damage to the root system of IAC3222 was significantly less severe than that of the Swingle rootstock. Multivariate hierarchical analysis considering all evaluated parameters clustered the CLas-infected plants grafted on IAC3222 together with the non-inoculated plants. We concluded that the IAC3222 rootstock was less affected by the CLas-infected scion compared to the Swingle rootstock and is a promising rootstock to minimize the HLB effect on plant development.

1. Introduction

Grafting in citrus has allowed growers to vegetatively propagate the most desirable characteristics of both scion and rootstock varieties for centuries. However, the right scion and rootstock combination is essential to achieve high orchard productivity and longevity [1]. As the rootstock influences scion development, yield, fruit quality, and tolerance to biotic and abiotic stress [2,3], the choice of the rootstock variety is as important as the scion [4]. The citrus industry in Brazil has traditionally relied on a few rootstocks, and the switch from one predominant variety to another was mainly motivated by biotic stress. For example, the first commercial citrus plantations in Brazil were based on ‘Caipira’ sweet orange (Citrus sinensis L.) but due to its susceptibility to Phytophthora sp., this rootstock was changed to ‘Sour orange’ (Citrus aurantium L.) [5]. Later, the susceptibility of ‘Sour orange’ to Citrus Tristeza Virus (CTV) motivated the use of ‘Rangpur lime’ (Citrus reticulata hybrid) as the main rootstock variety [6,7]. Recently, ‘Rangpur lime’ has been replaced by ‘Swingle citrumelo’ as a consequence of its susceptibility to Citrus Sudden Death disease (in Northern Sao Paulo State) [8]. Additionally, the ‘Swingle citrumelo’ is chosen for its ability to provide better quality scion fruits for the production of no-frozen concentrate juice [6].

The huanglongbing (HLB) disease, caused by the bacterium ‘Candidatus Liberibacter asiaticus’ (CLas), is the main limitation of citrus production worldwide. In addition to visible symptoms on the scion (e.g., mottling of leaves followed by dropping, lopsided fruits with premature fruit drop, and plant stunting), the root system (mainly fibrous roots) of the rootstock is significantly reduced in CLas-infected plants [9,10]. Such a decrease in the citrus root system has been hypothesized as the cause of HLB symptoms in the leaves, but the results are controversial [11]. According to Johnson et al. [9], the density of fibrous roots was reduced in the citrumelo Swingle rootstock before the development of foliar HLB symptoms and was directly associated with CLas in roots. However, increased starch content in leaves [12,13,14], reduced photosynthetic activity [15], and increased callose deposition in the phloem [16,17] are symptoms frequently reported for HLB-infected plants, which together result in an imbalance of carbohydrates between the source and sink tissues [18,19]. Keeley et al. [15] speculated that those effects could be a consequence of CLas rather than strictly systemic responses to infection. We hypothesize that rootstocks more resilient to CLas infection (i.e., hosting low bacterium titers in the root and higher fibrous root density) can be an important tool to reduce damage caused by HLB.

Although no Citrus genotype is completely resistant to CLas, published works have shown different degrees of susceptibility among Citrus species [20]. Poncirus trifoliata, a close relative and sexually compatible with Citrus species, has been reported as tolerant to HLB and capable of hosting low CLas titers [21,22]. This characteristic has been genetically transmitted as some citrandarins (hybrids of Citrus sunki x P. trifoliata cv Rubidoux) remain uninfected after CLas inoculation through natural infection [23] or grafting [24,25,26]. However, the effects of HLB on the root system of citrandarins used as a rootstock remain to be investigated. Tolerant genotypes used as interstock between the susceptible scion and the rootstock led to positive effects on scion vigor (e.g., tree growth rate and canopy volume) [27]. Here, we used the citrandarin hybrid IAC3222 as the rootstock for ‘Valencia’ sweet orange and evaluated CLas colonization (in the rootstock and scion) and root volume and mass. IAC3222 was chosen based on previous results showing an 80% decrease in the number of CLas-infected IAC3222 plants during the experimental period [24], as well as its horticultural advantages in the experimental trials conducted by our team [28].

2. Materials and Methods

The experiment was conducted at the Citrus Center Sylvio Moreira, Agronomic Institute (IAC), in Cordeirópolis, São Paulo, Brazil from February 2019 to September 2021. Briefly, seeds of rootstock varieties were germinated in fine granulometry pine-bark substrate. After selection, nucellar seedlings (10–15 cm in height) were transplanted into 15 L plastic vessels containing medium granulometry pine-bark substrate and grown until grafting. The nursery plants were managed following the recommended technique [8], and CLas inoculation was performed by grafting two infected budsticks onto the main branch of each nursery plant. To avoid natural infections after grafting, plants were maintained in a greenhouse with temperatures not exceeding 30 °C and were watered twice daily (totaling 300 mL of water per plant/day). Macro (N:P₂O₅:K₂O) and micronutrients (Zn, Mn, B) were added every 15 days. The treatments consisted of citrandarin hybrid IAC3222 as a rootstock (T1), Swingle citrumelo as a rootstock (T2), IAC3222 as interstock (a piece of stem—5 cm—inserted between the rootstock and scion using two graft unions) and Swingle citrumelo as a rootstock (T3), and IAC3222 as interstock (30 cm) and Swingle citrumelo as a rootstock (T4). ‘Valencia’ sweet orange was grafted as a scion in all treatments (Figure 1). We analyzed a total of six repetitions (2 inoculated with non-infected buds (control) and 4 inoculated with infected buds) per treatment.

Thirteen months after inoculation, 10 leaves from the canopy and fibrous roots were randomly harvested from each plant. The volume of the root system (estimated using the conventional method of water displacement by Bernardi et al. [29]), root dry mass, and main stem diameter (5 cm above substrate level) were also determined. Both the detection and quantification of CLas were performed using real-time quantitative PCR (qPCR) based on the protocol outlined by Li et al. [30] and described in detail by Cavichioli et al. [24]. DNA for the qPCR analysis was extracted from 300 mg of fresh tissues (leaf petioles from the scions or fibrous roots from the rootstocks) and isolated as described by Murray and Thompson [31].

The data were compared using the Scott–Knott test, with a significance level set at 5%. The horticultural indices of rootstocks (root system volume and weight, and main stem diameter) as well as the threshold cycle (CT) values obtained by qPCR, were analyzed using the multivariate technique of the hierarchical agglomerative clustering method proposed by Ward [32], with cluster arranged hierarchically according to Murtagh and Legendre [33].

3. Results and Discussion

The citrandarin hybrid of P. trifoliata X C. sunki (IAC3222) used as a scion has remained uninfected by CLas through natural transmission for over 8 years in an area with a high incidence of HLB and its insect vector Diaphorina citri [23]. Previous investigations have not found evidence supporting the natural P. trifoliata repellence of D. citri nymphs or adults [34], which suggests that IAC3222 could be naturally resistant or at least tolerant to CLas. Recently, our group demonstrated through top-grafting of IAC3222 onto HLB-diseased citrus plants hosting high CLas titers that the number of grafted sprouts infected with the bacteria decreased by 80% over 13 months [24]. Taken together, the citrandarin IAC3222 genotype has shown genetic resistance under natural infection conditions and high tolerance to CLas under a stressful experimental setup with constant pathogen flux from the source plant to the top-grafted citrandarin. Additionally, IAC3222 as a rootstock for the ‘Valencia’ scion in field trials with HLB incidence has shown less disease severity (grade 2) and horticultural advantages such as drought resistance and higher juice quality (e.g., °brix = 10.3, °brix: acidity ratio = 13.29, and juice weight = 103.2 g) and fruit size (242 g/fruit) compared to Rangpour lime (C. x limonia) rootstocks [28].

Our results reinforce the severe reduction in rootstock root volume (42% in T1 (IAC3222 as rootstock) to 70% in T2 (Swingle citrumelo) 13 months after inoculation) when the scion was infected by CLas, with a similar trend for the root dry weight (

Table 1. Impact of ‘Candidatus Liberibacter asiaticus’-infected ‘Valencia’ orange plant on the root system.

Treatments	Diameter of the Main Stem (mm)		Root System Volume (cm3)		Root System Dry Weight (g)		CT Values 1
	Infected	Healthy	Infected	Healthy	Infected	Healthy	Scion	Roots
T1	19.5 ± 2.1 a	21.0 ± 1.8 a	222.5 ± 36.5 a	381.5 ± 95.5 * a	154.1 ± 18.3 a	238.5 ± 6.9 * a	19.0 ± 2.6 a	37.4 ± 0.9 a
T2	18.1 ± 1.8 a	20.5 ± 1.5 * a	154.5 ± 48.8 b	509.5 ± 3.2 * a	86.8 ± 18.5 b	243.0 ± 4.3 * a	19.3 ± 0.7 a	31.3 ± 0.5 b
T3	16.9 ± 3.2 b	20.5 ± 0.5 * a	163.7 ± 38.5 b	364.5 ± 10.8 * a	72.8 ± 12.5 b	218.8 ± 8.9 * a	22.7 ± 4.3 a	27.9 ± 0.9 c
T4	14.4 ± 1.9 b	16.2 ± 0.5 * b	106.5 ± 46.7 b	307.0 ± 15.7 * a	63.2 ± 11.6 b	129.1 ± 7.5 * b	15.9 ± 0.4 a	30.5 ± 1.4 b

Data are means ± standard deviation. Significant differences (Scott–Knott test at 5% probability) are shown by letters (a,b,c) among the treatments and by asterisks (*) between infected and healthy plants in the same treatment. T1 = Rootstock citrandarin IAC3222, T2 = Rootstock Swingle citrumelo, T3 = Interstock IAC3222 (5 cm) and rootstock Swingle citrumelo, T4 = Interstock IAC3222 (30 cm), and rootstock Swingle citrumelo. ¹ Samples with cycle threshold (CT) values < 32 were considered CLas-positive. All samples from the non-inoculated plants had CT values ranging between 34.8 and 40 (undetermined).

). This significantly lower damage to the root system of IAC3222 (T1) could explain the lack of difference in the diameter of the main stem between CLas-infected and non-infected treatments. However, for treatments T2 and T3, the diameter of the main stem measured 5 cm above the substrate surface was significantly larger in non-infected plants (Table 1). The lack of difference between infected and non-infected T4 plants could be attributed to the negative effect of the interstock length (30 cm) on plant development. In contrast to Swingle citrumelo roots, regardless of the presence of interstocks (T2, T3, and T4), CLas was not detected in the roots of IAC3222 (T1) (Table 1). The average CLas titer in the ‘Valencia’ scion was generally high (CT values ranging from 19 to 22.7) at 13 months post-inoculation, with no significant difference among treatments (Table 1). This suggests that the lack of detection of bacteria in the IAC3222 rootstock did not reflect the actual titer in the scion. Graham and co-workers [35] showed no correlation between CLas quantification in the scion and the root system but they demonstrated that the bacterium colonizes the roots, serving as a source of inoculum for the scion. Regardless of whether the reduced root system density was a cause or effect of HLB symptoms in the leaves [11], the density of fibrous roots was reduced in the tested rootstocks. Specifically, for IAC3222, this decrease was not associated with CLas in the roots (Table 1), in contrast to what was hypothesized by Johnson et al. [9] using Swingle citrumelo as the rootstock. We supposed that the damage to scion physiology (e.g., accumulation of starch in the leaves [12,13,14], reduced photosynthetic activity [15], increased callose deposition in the phloem [16,17], and imbalance of carbohydrates between the source and sink [18,19]) could explain the reduced root density observed in IAC3222.

Taking together all the evaluated parameters and using the Euclidean distance to measure similarity among treatments [32], we observed that citrandarin IAC3222 as the rootstock of CLas-infected plants was clustered together with non-infected plants (Figure 2). This further corroborates that CLas in the scion has a lower impact on the IAC3222 rootstock. Recently, Dutt et al. [27] demonstrated that the use of HLB-tolerant pummelo as an interstock between Swingle citrumelo and ‘Valencia’ enhanced the scion response against Clas compared to the Swingle control (no interstock). However, these conditions did not impair the multiplication of the bacteria in the scion. Here, the interstock IAC3222 (P. trofilita hybrid, tolerant to CLas) did not contribute to increasing the root volume or mass of the Swingle citrumelo rootstock. One possible explanation is that the sap flow to the rootstock was reduced as a consequence of poor vascular union between these genotypes [36].

Here, we provide evidence of a lesser effect of CLas in the root system of the citrandarin IAC3222 used as rootstock for the ‘Valencia’ scion containing a high population of bacteria. Furthermore, we could not detect CLas in the fibrous roots of that hybrid. We hypothesize that the use of a CLas-tolerant rootstock such as the citrandarin IAC3222, in combination with management strategies (e.g., the application of biostimulants [37], bactericidal molecules [38,39], and HLB-tolerant interstocks [36]) may enhance the plant response to CLas. The management of HLB requires a set of actions that begins with the optimal combination of scion/rootstock and vector population reduction, in addition to cultural practices such as adequate nutritional programs for the specific rootstock variety and other strategies to boost plant health.