Plant material

Commercial varieties of Agata and Fianna potatoes were tested in this study. These varieties contrast in growth habit and tuber quantity. Seed tubers for both varieties, healthy and apparently pathogen and pest free were obtained from the municipality of Zaragoza, Puebla, México.

Seeds were sown in 4-L capacity black polyethylene bags and kept in a greenhouse. The bags were covered with anti-aphid mesh to control pests and diseases until production. Three seeds were placed into perlite at 10 cm depth. Bags were kept in a controlled-environment chamber (GCA model 815 Freas; USA) at 20±1 °C, 50 % relative humidity, and in complete darkness. Ten days after sowing (DAS), tubers with vegetative shoots (Figure 1) were sampled. Sampled shoots were placed in a cooler (6±1 °C) for dissection in the laboratory. Sampling continued until 30 DAS. Mother tubers were washed three times with distilled water to remove substrate particles. From each tuber, vegetative buds were extracted with a size 11 scalpel and a stereoscopic microscope (OLYMPUS model SZ-CTV; USA).

Figure 1 A) Seed tuber with vegetative shoots of S. tuberosum L. var. Agata 15 days after sowing; B) Vegetative shoot; and C) Longitudinal dissection from the stem of a vegetative shoot stained with 0.05 % acid fuchsine (stereoscopic microscope, Carl Zeiss^® SZ60). CA: cataphyll covering the stolon bud (Y), RA: adventitious roots, and TS: underground stem. Orientation of the growing apex is marked by the black arrow. 

Samples of buds in four stages of stolon development were extracted from buds at the basal zone of the stem with a stereoscopic microscope (Carl Zeiss model SZ60; Germany): E1) flattened and latent stolon bud at initial stage; E2) swollen and still latent stolon bud; E3) sprouted stolon bud, also called stolon primordia, in which it starts its elongation and development; and E4) stolon in early formation stage.

Two hundred mg of buds (containing almost no meristematic tissue) were placed in 2 mL Eppendorf tubes. This quantity was obtained with 50 buds per variety and in each development event. Microtubes containing the buds were kept at -20 °C (Puffer Hubbard model IUF1821, USA) until processing.

Buds at stages E1, E2, E3 and E4 were extracted by longitudinal dissection (with a metal blade; Guillette^®) of axillary buds from underground stems from both varieties. Buds were stained with fuschine acid (0.05 %)

Extraction of plant growth regulators (plant hormones)

Before extraction, samples were lyophilized for 12 hours (Labconco model Freezone 4.5; USA). Ten mg from each lyophilized bud type were placed in a plastic microtube and 500 μL of extraction solution added as per Pan’s protocol (2010). The extraction solution contained 2-propanol HPLC grade (J.T. Baker^®), deionized water and reagent grade concentrated HCl (J.T. Baker^®) in a 2:1:0.002 ratio. Microtubes were agitated at 100 rpm for 30 minutes using a vortex (Scientific Industries model G560; USA) and kept at 6±1 °C. After agitation, 1000 μL of HPLC-grade dichloromethane were added to each microtube. The mixture was agitated for 30 min at 100 rpm and then centrifuged at 716 g for 10 min (Hettich, model EBA 21; USA). The lower organic phase was extracted with an Eppendorf micropipette (USA) and placed in a clean microtube. Using nitrogen gas (high purity grade; Infra^®), the solvent was completely evaporated, and 500 μL of HPLC-grade filtered methanol were added (J.T. Baker^®) to each microtube. These microtubes were stored until chromatographic analysis.

Analytic standard preparation

Quantitative standards for this research were 3-gibberellic acid (GA3, 90 % purity), indole-3-acetic acid (IAA, 98 % purity), abscisic acid (ABA, 98.5 % purity), kinetin (KIN, 98 % purity), 6-(^γ,^γ-dimethylalylamino)-purine (2-iP, 98% purity) and 6-benzylaminopurine (BAP, 99 % purity) (Sigma-Aldrich^®). For each standard, dilutions at 0.01, 0.1 and 1.0 mg mL^-1 were prepared in HPLC-grade methanol (J.T. Baker^®).

Chromatographic analysis

An Agilent HPLC model 1100 coupled to an UV-Vis detector was used for detection. The column used for separation was a 4.6(75 mm Rx/SB-C8 column (model 866953-906; Agilent). The mobile phase contained an 80:20 mixture of solutions A and B. Solution A included HPLC-grade acetonitrile (J.T. Baker^®) and triflouroacetic acid spectrophotometric grade (Sigma-Aldrich^®) in a 1:0.001 ratio. Solution B was a 1:0.001 mixture of deionized water and spectrophotometric-grade trifluoroacetic acid (Sigma-Aldrich^®). Both solutions were filtered using 0.45 μm HV-type Millipore^® membranes (adapted from ^{Pan et al., 2010}).

Flow rate for the mobile phase was 2 mL min^-1. Three 20 μL replications of every extract were injected into the HPLC using a manual injector (Rheodyne^®, model 755). Readings were taken at three wavelengths, and the reading showing the highest absorption was chosen. The working wavelengths were 206 nm for gibberellins, 254 for auxins and abscisic acid (^{Harborne, 1994}), and 280 for cytokines. Peak area was recorded as milli-absorbance units per second (mAU s^-1) (^{Ortiz and Flórez, 2008}). The identity for each fraction was determined by comparison with retention time and maximum absorption wavelength for the standards. Calibration curves were also calculated for each standard.

Identification and classification of stolon buds

Latency loss and sprouting in the stolon bud was detected at stage E3 (Figure 2).

Figure 2 Longitudinal sections of buds taken from axillary stolons of underground stems of S. tuberosum L. var. Agata, stained with acid fuchsine (0.05 %), and observed under a microscope (10 X, Carl Zeiss^®). E1) Stolon bud at initial, flattened and latent stage, with remains of the cataphyll (CA) covering the stolon bud (Y) and adventitious roots (RA); the axis of the underground stem and the orientation of the growth apex is marked by the arrow. E2) Stolon bud at a swollen but still latent stage. E3) Sprouted stolon bud or stolon primordia (RE) when stolon is clearly elongating and developing. E4) Stolon (E) at an early stage of development 

The extraction and analysis method in this research was specifically designed to detect and quantify hormones. Thus, in each tested wavelength, each fraction should correspond to compounds with similar physical and chemical (size and charge) characteristics to the standards (Figure 3 and Table 1). In this research, no cromatographic fraction at any wavelength overlapped; thus, every peak from each of the eight fractions obtained was considered independent (Table 2).

Figura 3 HPLC chromatograms for analytical standards. A) GA3 at 206 nm. B) AIA, ABA, IBA and ANA at 254 nm. C) CK, 2-iP and BAP at 280. 

Fractions F-1 and F-8 agreed with auxins, and their maximum absorption was at 254 nm. Fractions F-2, F-3 and F-4 agreed with cytokines, which absorb maximally at 280 nm. At 206 nm, fractions F-5, F-6 and F-7 matched gibberellins (Table 2).

The eight fractions changed concentration as stolon development stages progressed in both varieties (Figure 4). Four of the eight fractions appeared to be directly associated with stolon sprouting stages from E2 to E3, where E3 is the key event in the tuber formation. The fractions involved in this response were F-3, F-4 and F-8.

Figure 4 Changes in concentration (±standard deviations) in the fractions extracted from stolons at four stages of development. Varieties Agata (---) and Fianna ((). 

In both varieties, fraction F-7, a gibberellin-type compound, decreased sharply by E3 (less than 1000 mAU s^-1), and it accounted for 17 % of its content in the latent bud at E1 (Figure 4); F-8, an auxin-type compound, was expressed in E1 and E2 but not in E3. This behavior suggests that both growth regulators (F-7 and F-8) act as inhibitors of sprouting for stolon buds in the varieties Agata and Fianna, and probably in other potato varieties.

Detectable hormones during formation of stolons in H. tuberosus L. were GA3 and zeatin. The first hormone content increased initially and then decreased as tuber induction and growth continued. Zeatin content increased during all stages, and its lowest content occurred during stolon initiation (^{Li et al., 2017}).

According to ^{Taiz and Zeiger (2002)}, low auxin levels promote bud development, and an increase in its concentration inhibits sprouting. In contrast, GA1 at low concentrations promotes stem elongation stems and high concentrations tend to drastically reduce their growth rate.

An increase in the number of vegetative shoots in potato might be achieved faster by supplying GA3 at low concentrations; this behavior shows the important role gibberellins play in potato development (^{Alexopoulos et al., 2007}; ^{Salimi et al., 2010}). ^{Abdala et al. (2000)} showed by comparison of potato organs that gibberellins influence stolon development in stages following sprouting. These authors obtained increased levels of jasmonates, GA1 and GA3 in sprouted and developing stolons, and such levels were higher during tuber formation.

Contrary to reports by other researchers, the results in our study show that the gibberellin-type fraction (F-7) participated as a repressor of sprouting of stolon buds (Figure 4). Fraction F-7 did not match GA3, as the analytical standard for this hormone had different retention time (Table 2 and Figure 4).

In contrast, in Fianna potato fractions F-3 and F-4 of the cytokinin-type, were not present in the developmental stages E1 and E2 but did show in E3 and E4. These results may indicate that the two fractions promote stolon germination only in some genotypes. When the stolon broke latency during stage E3 and E4, fractions F-3 and F-4 did not appear, thus suggesting that in this variety these cytokines are unnecessary to induce stolon sprouting. These results coincide with those of ^{Ortiz and Flórez (2008)}; they pointed out that the interaction genotype(cytokinins was not significant between the potato varieties in the study after detection and quantification of different kinds and concentrations of hormones.

It is then possible to assume that stolon sprouting in both varieties of potatoes takes place when gibberellin F-7 decreases and auxin F-8 is absent, because these two growth regulators, inhibit or suppress stolon sprouting. It can also be inferred that cytokinins participation in stolon sprouting is required only in some potato genotypes, and that the interaction of cytokinin(genotype is associated with habits of expression of stolons. This can be appreciated in var. Agata: stolon induction was early and abundant compared to Fianna. In the latter variety, this process occurred gradually and at a later stage.

The results in this study contribute new knowledge on the process of stolon initiation in potato, although more research is necessary to identify the chemical structure of the compounds described here, as well as to carry out bio-essays to determine their role.