Tulip axillary bud growth in the yearly bulb growth cycle
The natural growth cycle of tulip bulbs was monitored (Fig. 1a, b) with special emphasis on the growth of the axillary buds (Fig. 1c, d). The planted bulbs, of size 9–11 cm in perimeter, had on average five bulb scales and a tunica, six axillary buds (one axillary bud per bulb scale) and an apical shoot bearing a floral primordium (hence also called floral bud) (Fig. 1a). The bulbs were dug out at different time points during the growth season, and growth of their axillary buds was measured (Fig. 1c).
As stated by other researchers (De Hertogh and Le Nard 1993b; Le Nard and de Hertogh 1993; Rees 1968), we also found that not all axillary buds had the same growth capacity, neither the same growing behaviour (Fig. 1a–c). The outermost axillary bud (H) resembled very much an apical bud in the sense that both experienced shoot elongation, sprouting and bulb renewal (Fig. 1a). But contrary to the description of Rees (Rees 1968), which indicated that all axillary buds continue to grow during winter, although at a low rate, we did not detect growth in mid-located (C, D and E) buds during that period (Fig. 1c, from December till February, and Online resource OSM1). Once the apical bud of the mother bulb sprouted and its leaves unfolded and turned dark green (Fig. 1a), a boost in growth of A, B and H axillary buds was observed (Fig. 1c, March). Nevertheless, the growth of the mid-located buds resumed mildly, resembling a mild dormant state. By the end of the growing season (July), all axillary buds made a tunica (moment often referred to as summer dormancy), A, B and about half of the H daughter bulbs formed a reproductive apical meristem (Fig. 1d), and their A and B buds (hence, grand-daughter buds) arose later during the storage period.
The halt in growth of the mid-located buds (C, D and E buds in this study) during winter (December till February) and slow growth in early spring (March till May) suggested a differential control of axillary bud outgrowth in tulip buds, which might be determined by different levels of bud dormancy. The term dormancy as a synonym to temporary growth arrest has been redefined several times through the years mainly because one definition does not apply to all species. The concept of dormancy in geophytes is still controversial since growth of, for example, the apical bud of tulip bulbs is not arrested after its initiation (Okubo 2012). For practical reasons, we refer in this text to dormancy as the lack of sustained growth. The TB1/BRC1 TCP transcription factor gene has been used in several dicot and monocot species as a marker to assess bud dormancy (Nicolas and Cubas 2016). To have an extra parameter to assess dormancy in tulip bulbs, we then isolated the putative TB1/BRC1 transcript in Tulipa gesneriana and tested its role as a bud dormancy marker.
Identification and gene expression of a tulip TB1-like TCP transcription factor
We had previously identified several TCP domain-containing transcripts of tulip and lily, among which, one lily (Lilium oriental) transcript revealed a close sequence homology with Arabidopsis BRC1 (TCP18) (Moreno-Pachon et al. 2016). Using primers based on the lily TB1 (LoTB1) sequence, a 406-bp fragment (including the entire basic helix-loop-helix domain called TCP domain) of a potential tulip TgTB1 transcript was isolated (Online resource OSM2). It has been reported that TB1/BRC1 proteins share two specific amino acids in the basic region of the domain, which distinguishes them from the other class II TCP transcriptions factors (Martín-Trillo and Cubas 2010). Thus, we aligned the potential tulip TgTB1, lily LoTB1, Arabidopsis AtBRC1 and rice OsTB1 protein domain sequences and found a high degree of sequence similarity, as well as the presence of the two TB1/BRC1-specific amino acid features in the isolated tulip sequence at the expected position (Fig. 2a).
TB1/BRC1 transcripts have been detected in the meristem and leaf primordia of axillary buds of maize and in pro-vascular tissues of buds bearing flowers in Arabidopsis (Aguilar-Martínez et al. 2007; Hubbard et al. 2002). To corroborate the true TB1/BRC1 nature of our identified TgTB1 transcript, the expression of TgTB1 was quantified in different organs of the bulb (axillary buds, floral bud, scales and basal plate) at the end of the storage period (prior to planting). Among all tissues tested, TgTB1 expression was only significantly detected in the axillary buds. Moreover, active A and H buds had the lowest expression, while highest expression was found in the dormant D buds (Fig. 2b). Subsequently, in situ hybridization was carried out on D buds of bulbs at the end of the storage period, and TgTB1 transcripts were detected in the pro-vascular tissue just below the central meristematic region (Fig. 2c).
To investigate further the role of TgTB1 as a marker for dormancy in axillary buds, we assessed the expression of TgTB1 in daughter bulbs having a so-called Springpartij phenotype. “Springpartij” refers to a phenomenon experienced in some tulip bulbs where there seems to be no control in outgrowth of axillary meristems contained in the mother bulb, including the axillary buds of the axillary buds (hence, grand-daughter buds). As a consequence, many daughter bulbs of more or less the same size are formed from a single mother bulb at the end of the growing season (Fig. 3A). In general, this spontaneously occurring phenomenon cannot be reverted, and in the next growing season the phenotype will be repeated in each of the daughter bulbs. In line with the proposed function of TgTB1 in repressing axillary meristem outgrowth, we found overall low expression of TgTB1 in ‘Springpartij’ buds (Fig. 3b) and did not observe significant differences in TgTB1 expression among the axillary buds, as it was found in normal buds (Figs. 2b, 3b).
To study the dynamics of TgTB1 expression in tulip axillary buds and its correlation with outgrowth, five physiological states of the growth cycle (during storage, at planting time, before anthesis, after anthesis and at lifting) were studied in A and D axillary buds. Those buds were chosen because of their contrasting behaviour in growth during the growing season (Fig. 1c), and their different TgTB1 expression at storage (Fig. 2b).
The results indicated that during the storage period (time previous to bulb planting), D buds did not significantly grow, while A buds experienced a linear growth increase (Fig. 4ai). Regarding the TgTB1 relative expression during this time, there was a dramatic upregulation in D buds by the end of the storage period (late storage time), while TgTB1 decreased in A buds (Fig. 4aii). Figure 4bii indicates that after the bulb is planted, TgTB1 relative expression in D buds decreases also. However, the relative expression of TgTB1 in both A and D buds becomes similar only at the moment of lifting at the end of the growing season (Fig. 4bii).
Following the idea of TB1 expression as a marker for bud dormancy, we expected the downregulation of TgTB1 in D buds towards the end of the growth season to result in significant bud growth. From Fig. 4bi, it can be observed that limited, although significant growth was accomplished in D buds in this last period. A possible explanation for the limited growth of D buds could be the lack of a good vascular connection. We discarded this possibility, because both A and D buds showed a main vasculature connecting them with the basal plate (Online resource OSM3).
In summary, TgTB1 expression in the selected axillary buds was inversely correlated with their growth pattern during storage (r
2 = − 0.9) and during the growing season, however, to a lower extent (r
2 = − 0.56). When using TgTB1 as a marker to assess dormancy, it can be said that D and not A buds enter a dormancy state during storage, and this dormancy is only broken in late spring after planting and growth during winter. However, although D buds are freed from a molecular imposed dormancy in spring, their growth is still limited and not supported.
The role of sucrose in modulating tulip axillary bud outgrowth and TgTB1 expression
Sugars are stored in the bulb scales as starch and translocated to the sink organs based on their demand (Ho and Rees 1976). Nevertheless, axillary buds obtain most of the carbon from the leaves of the apical bud once it sprouts in early spring (Ho and Rees 1975). Taking into account the reported role of sucrose to break bud dormancy (Barbier et al. 2015; Mason et al. 2014) and the observed correlation of TB1/BRC1 with axillary bud growth (Aguilar-Martínez et al. 2007; Braun et al. 2012; Martín-Trillo et al. 2011; Takeda et al. 2003), we reasoned that mid-located buds might not get enough sucrose to break dormancy in early spring, probably because of competition with the already active buds (e.g. A bud).
To test whether making sucrose equally available to the buds could break dormancy and promote growth in D buds, we cultured them in vitro with or without sucrose in the medium. A buds were used as control. The buds were excised from mother bulbs stored at 20 °C for three months followed by 4 °C for another three months. These temperature pre-treatments recreated the storage and cold period of their natural growth cycle, resembling the early spring time of tulip growth cycle (Fig. 1). As expected, sucrose enhanced the growth in both buds, although to a much lesser extend in D buds (Fig. 5). Without sucrose both A and D buds lost weight and this was lethal for the majority of the smaller D buds. Hence, sucrose was fundamental for the survival of D buds, which is an indication that the buds are able to uptake the sucrose that was present in the medium. TgTB1 expression in dormant D buds was downregulated after the sucrose treatment, as expected. However, it was intriguing that D buds did not grow significantly in the last 3 weeks of the experiment where sucrose was supplemented. As sugars as an energy source do not seem to be limiting, we speculated that other factors, such as hormones, might also play a role in controlling the growth capacity of axillary tulip buds.
Hormonal modulators of TgTB1
Rameau et al. proposed a general model in which TB1/BRC1 integrates multiple pathways that control axillary bud outgrowth (Rameau et al. 2015). In this model TB1/BRC1 is activated by strigolactone (SL) while repressed by sucrose, cytokinin (CK) and gibberellic acid (GA). However, gibberellins can induce axillary shoot elongation instead of biomass gain in tulip [revised in (Okubo 2012)]. To investigate the role of SL and CK in the regulation of axillary bud outgrowth in tulip, we tested the combined effect of sucrose and CK (BAP) in the outgrowth of dormant D buds, since we found that sucrose is required for the survival of this type of buds. The effect of SL (GR24) was tested in the active A buds using mannitol as an osmotic control (Henry et al. 2011).
The results showed that the combination of sucrose + BAP downregulated TgTB1 expression faster than sucrose on its own in dormant D buds (Fig. 6aii). Nonetheless, after 6 weeks of culture, both the treatments of sugar and sugar + BAP showed the same capacity to induce bud growth (Fig. 6ai) and downregulated TgTB1 to the same level (Fig. 6aii). With regard to A buds, they experienced an upregulation of TgTB1 expression upon culture with mannitol and strong increase in TgTB1 expression with the combination mannitol + GR24 (Fig. 6bii), indicating that SL can modulate TgTB1 expression in tulip bulbs. Moreover, for both treatments, mannitol and mannitol + GR24, axillary bud outgrowth (Fig. 6bi) was negatively correlated with the level of TgTB1 expression (Fig. 6bii).