Morphological studies

Seed collection and germination 

Twenty-one (21) rice land races were obtained from local rice farmers in Benue and Cross River States, Nigeria (Supplementary Table 1). The rice seeds were soaked in water for 24 h to break seed dormancy and thereafter incubated in Petri dishes containing wet filter paper for 7 days for germination. 

Preparation of nutrient solution

The modified Hoagland’s solution described by Hoagland and Arnon (1950) was used for the present study with slight modification. The main adjustment made was addition of dipotassium hydrogen phosphate (K2HPO4) in different concentrations in place of monopotassium dihydrogen phosphate (KH2PO4) (Supplementary Table 2).

Transplanting and treatment application

Three (3) seedlings per landrace were transplanted into 50 ml labeled Falcon tubes containing water and kept for two days for acclimatization. The Falcon tubes were wrapped with foil paper to protect them from direct sunlight. After two days, water was drained from the labeled Falcon tubes and replaced with the modified Hoagland solutions containing 0.00 mM, 0.4 mM and 0.8 mM K2HPO4 (zero phosphate, normal phosphate and excess phosphate, respectively). The experiment was a 21 X 3 factorial, laid out using the Completely Randomized Design. At 6 weeks of age, the seedlings were transplanted to more spacious Polyvinyl Chloride (PVC) pipes with the same treatment solutions continued (Supplementary Figure 1). In all cases, the treatment solution was changed every four days to prevent a build-up of ions released by the plants. 

Data collection and analysis of morphological parameters

At ten weeks after planting, data were collected in triplicates on plant height, number of tillers, root length, shoot and root dry mass in all the 21 landraces at the 3 phosphate concentrations. The means from these were used in estimating Phosphorus Deficiency Tolerance Index (PDTI) or Phosphorus stress factor for each landrace in all the traits using the formula: Trait under sufficient Pi - Trait under deficient Pi / Trait under sufficient Pi, that is (Trait under PiN - trait under Pi-/ trait under PiN) (Chankaew et al., 2019; Irfan et al., 2020).

Molecular studies

DNA extraction and amplification of Pup markers

DNA was extracted from the twenty-one rice landraces using modified cetyltrimethylammonium bromide (CTAB) method as described by Uyoh et al. (2019). Four (4) gene-based Pup markers selected from the improved set of Pup-1 markers developed by Chin et al. (2011) based on stable and conserved protein-coded genes verified through gene expression and sequencing data, were used in this study. The reaction was constituted in a total of 50 µL made up of 40.75 µL distilled water, 2.5 µL 10X Ex Taq buffer (1X), 2.5 µL MgCl2 (1.5 mM), 1 µL deoxynucleoside triphosphate (dNTPs) (200 µM), 1 µL forward primer (0.2 µM), 1 µL reverse primer (0.2 µM), 0.25 µL Ex Taq polymerase enzyme (1.25 units), and 1 µL DNA (100 ng/µL). The polymerase chain reaction (PCR) conditions were as follows: Initial denaturation at 95°C for 30 s, 35 cycles of denaturation at 95°C for 20 s, annealing at 55°C for 55 s, extension at 68°C for 1 min, followed by a final extension at 72°C for 5 min and holding at 4°C. The amplified products were separated by electrophoresis in 1.3% agarose gel at 100 V for 30 min, stained with ethidium bromide and visualized under UV trans-illuminator.

RNA extraction and cDNA synthesis

To assess gene expression level and responsiveness to P, reverse transcriptase-PCR (RT-PCR) analyses were done using roots and shoots from rice samples germinated from the same seed stock used previously for the morphological analyses. They were grown in Falcon tubes for 2 weeks with (PiN) or without P (Pi-) as described previously. RNA was extracted from the roots and shoots of selected samples (Phillipine, Ayange and Obasanjo) using the E.Z.N.A.® Total RNA kit (Omega Bio-tek), according to the manufacturer’s protocol manual. The landraces were selected randomly from those that showed PCR amplicons for all genes tested. The final RNA was eluted with 40 µL diethyl pyrocarbonate (DEPC) water and stored at -80°C. cDNA was synthesized using LunaScript™ RT SuperMix kit (New England BioLabs Inc.) according to the instructions in user manual. Five µL of the synthesized cDNA was used for PCR analysis to detect  the base pairs of the phosphate uptake gene using the primers OsPupK04-1 and OsPupK05-1. Ubiquitin gene was used as internal control to check the quality of the cDNA using the primers OsUbq_F and OsUbq_R (Supplementary Table 3). The reaction was constituted in a total of 50 µL volume containing 5 µL of 10X Ex Taq buffer (1x), 1 µL dNTPs (200 µM), 1 µL of 10 µM forward primer (0.2 µM), 1 µL of 10 µM reverse primer (0.2 µM), 0.25 µL of Ex Taq polymerase enzyme (1.25 units), 5 µL of cDNA, and 36.75 µL of nuclease-free water. PCR amplification was performed as follows: Initial denaturation set at 94°C for 5 min, followed by 30 cycles of denaturation at 94°C for 45 s, annealing at 52°C for 1 min, extension at 72°C for 1 min, final extension at 72°C for 7 min. After amplification, 10 µL of the PCR products were resolved on 1% agarose gel stained with ethidium bromide.

Morphological studies and phosphate deficiency tolerance indices (PDTI)

The mean responses of the 21 rice landraces in terms of shoot length, tiller number, root length, wet and dry biomass under Pi+, PiN, Pi- conditions, along with the calculated PDTI values are presented in Table 1. The mean shoot length ranged from 20.3 cm in Togo to 51.33 cm in OC for those treated with PiN; it ranged from 19.5 cm in Togo to 49.43 cm in Election 3 for those given deficient P (Pi-) and 17.71 cm in Variety 1.4 to 56.33 cm in Election 3 for those given excess P (Pi+). The mean tiller number ranged from 2.00 in Variety 1.4 to 5.33 in Sedi 1, Election 2 and Dantala Mass under PiN; it ranged from 1.00 in variety 1.4 to 3.67 in Sedi1 and Achancha for those treated with Pi-; and from 1.33 in Variety 1.4 to 5.03 in Sidi 3 for those given Pi+. Wet biomass ranged from 0.76 g in Togo to 3.99 g in Election 3 for those treated with PiN; those treated with Pi- gave a range from 0.69 g in variety 45 to 2.37 g in Election 3 and for those treated with Pi+, it ranged from 0.46 g in Variety 1.4 to 2.91 g in Obasanjo. Dry biomass ranged from 0.25 g in Varieties 1.4 and 45 to 0.59 g in Ayange for those treated with PiN; it ranged from 0.22 g in Variety 1.4 and Election 2 to 0.38 g in Variety 44 for those treated with Pi-; for those treated with Pi+ it ranged from 0.22 g in Togo to 0.49 g in Election 3. About 90% of the landraces showed increased root length in Pi- condition compared with the values for PiN and Pi+. Based on the mean PDTI values obtained, the 21 landraces can be seen to fall into 3 groups (Table 2): (a) those with negative PDTI values ranging from -0.29 to -2.18, indicating that their mean performance under Pi- was better than that under PiN; (b) those with low positive values (≤ 1.50) signifying that their mean performance under PiN and Pi- were quite similar and (c) those with high positive values (>1.5) indicating their mean performance under PiN was much higher than at Pi-.

Validation of Pup-1 linked markers in the rice landraces

Four Pup-1 linked markers were validated on the 21 rice landraces (Figure 1). The presence of any of the PCR products in a land race was taken as an indication that the land race possessed the Pup-1 gene (Table 3). At least, one of the four Pup-1 genes was confirmed to be present in each of the 21 landraces, but none of the genes was present in all samples.

Expression of Pup genes in the rice landraces

Expression of 2 phosphate uptake genes (OsPupKO4-1 and OsPupKO5-1) under zero (Pi-), normal (PiN) and excess (Pi+) phosphate levels was tested in 3 out of the 21 rice land races using RT-PCR. The 3 land races were selected randomly from those that showed PCR products for all the 4 genes tested. Expression profiles of the 2 genes and the control (Ubiquitin gene) are given in Figure 2a, b and c. The highest expression of OsPupKO5-1 gene was seen in the shoot of Ayange under normal and excess phosphate conditions (Figure 2aii). The gene was also strongly expressed in the root of this landrace under excess phosphate but poorly expressed under normal and zero phosphate levels (Figure 2ai). It was poorly expressed in the shoots of Philippine landrace but showed slightly better expression in the roots. For Obasanjo landrace, there was no expression of this gene in the shoots, but it was strongly expressed in the roots under low and normal phosphate levels. 

The 21 rice landraces evaluated in this study revealed different responses under Pi+, PiN and Pi- conditions. Tiller number, wet and dry biomass of shoot and root were generally higher when the plants were raised under sufficient P than when raised under deficient P, except in a few (5) landraces. On the other hand, all the landraces except Togo showed increased root length to different levels under deficient P condition, an adaptation for increased nutrient uptake under stressful situations. Wang et al. (2015) reported that different crops show divergent morphological and physiological responses to low P availability, including specific traits of root morphology and root exudation which enhance their P-uptake capacity under low-P conditions. Nirubana et al. (2020) also observed a wide range of varied responses to P deficiency for all the traits they studied in the 30 rice genotypes from India. They reported increased mean performance of root length and enzyme activity under P- condition compared to P+ condition. Vejchasarn et al. (2016) observed an increase in root hair length and density and a reduction in tiller number and shoot length under low P availability in all the 15 rice accessions they studied. 

The banding pattern of the four Pup markers used in the present study suggests that these genes, especially those of OsPupK29-1 and OsPupK20-2 which showed up in over 95% of the landraces are reasonably conserved in the 21 upland rice landraces. Previous reports also indicated that Pup-1 gene family was highly conserved in stress-adapted rice accessions (Chin et al., 2010, 2011). Heuer et al. (2009) added that Pup-1 locus increases P uptake under adverse conditions rather than increasing internal P-use efficiency. Chankaew et al. (2019) however reported that the presence or absence of the tolerant allele at the Pup-1 locus only showed a slight relationship with the tolerance indices in Thai indigenous upland rice germplasm. Moreover, some lines expressed high tolerance without the Pup-1- linked gene product. These lines, they suggested, may carry unknown P-deficient tolerance genes. Sarkar et al. (2011) reported that haplotyping of Pup-1-K42 markers showed the expected 918 bp amplification in 9 out of the 31 genotypes studied, but among them, only 3 genotypes showed higher P-uptake and dry-matter weight in P-limiting condition. Thus, the gene may be present but not expressed. This differential expression of the Pup-1 genes was also observed in the present study. Even though at least one of the genes was present in all the 21 landraces, their expression levels varied with the landrace and the phosphate concentration in the medium. The highest expression of OsPupK05-1 gene was seen in the roots and shoots of Ayange under excess (Pi+) and normal phosphate (PiN) rather than under Pi- when it is most needed. This perhaps resulted in the better performance of this land race under PiN compared to that at deficient P. Moreover, OsPupK04-1 gene was weakly expressed in the roots and shoots of this landrace under deficient P. OsPupK05-1 gene was well expressed in the root of Obasanjo under normal (PiN) and deficient (Pi-) levels which also explains the enhanced performance of Obasanjo seen under PiN compared to Pi- (the medium had zero P). It was also observed that Phillipine had the longest root under low P relative to the length under normal P, as reflected in the PDTI value of -1.8 for this trait. However, this did not reflect in the overall plant biomass for this landrace as all the other traits performed better under normal P level. Furthermore, OsPupK04-1 gene was well expressed in the shoot of Philippine under excess and normal phosphate which explains the better performance of this land race under PiN compared with Pi-. The average tolerance index for this landrace was, however, negative because of the very long root produced under Pi-. Since this OsPupKO4-1 gene was not expressed under Pi- condition in this landrace, the possibility of another phosphate uptake gene operating under this stress condition cannot be ruled out. 

The present report showed that only 3 out of the 21 landraces (Variety 44, Ankulyan and Dantala Mass) had higher dry-matter weight in P-limiting conditions which supports the report by Chankaew (2019) that the presence of the tolerant allele at the Pup locus does not have significant correlation with the tolerance indices. The choice of OsPupK04-1 and OsPupK05-1 for gene expression analysis in the present study was based on previous report of their ubiquitous expression in both roots and shoots of P- and P+ rice accessions by Chin et al. (2011) unlike OsPup20 which is restricted to roots with higher transcript abundance under P deficient conditions. Chin et al. (2011) observed that most of the Pup markers were not informative when tested in a wider range of rice germplasm and that OsPup-K29 was unstable. This marker, however, gave very consistent bands in 95% of the landraces used in the present study even though some were not expressed.

The study indicated that although low phosphate concentration causes adverse effects on the growth and productivity of these rice plants, the presence and adequate expression of phosphate deficiency tolerance genes in such plants improves adaptation to the stress by producing much longer roots in search of phosphate in the growth medium. Three genotypes namely Dantala Mass, Ankulyan, and Variety 44 may be regarded as tolerant to phosphate deficiency as they each had mean negative PDTI value with better performance in desirable roots and shoots traits under zero phosphate, in addition to possessing at least 3 of the 4 Pup 1 genes used. Also, the landraces that had good biomass with PDTI values close to zero (which means they performed well in both P-deficient and P-sufficient media) such as Achancha and which invariably also contained the Pup-1 gene(s), may be further explored along with the other three for possible inclusion in the breeding of this crop for P deficiency tolerance.

The authors have not declared any conflict of interests.

The authors are indirectly grateful to Bill and Melinda Gates Foundation as well as Nigerian Tertiary Education Trust Fund (TETFund) for equipping the laboratory used for the molecular aspect of the study.