Arabidopsis lonely guy (LOG) multiple mutants reveal a central role of the LOG-dependent pathway in cytokinin activation


(fax +81 45 503 9609; e-mail


Cytokinins are phytohormones that play key roles in the maintenance of stem cell activity in plants. Although alternative single-step and two-step activation pathways for cytokinin have been proposed, the significance of the single-step pathway which is catalyzed by LONELY GUY (LOG), is not fully understood. We analyzed the metabolic flow of cytokinin activation in Arabidopsis log multiple mutants using stable isotope-labeled tracers and characterized the mutants’ morphological and developmental phenotypes. In tracer experiments, cytokinin activation was inhibited most pronouncedly by log7, while the other log mutations had cumulative effects. Although sextuple or lower-order mutants did not show drastic phenotypes in vegetative growth, the log1log2log3log4log5log7log8 septuple T-DNA insertion mutant in which the LOG-dependent pathway is impaired, displayed severe retardation of shoot and root growth with defects in the maintenance of the apical meristems. Detailed observation of the mutants showed that LOG7 was required for the maintenance of shoot apical meristem size. LOG7 was also suggested to play a role for normal primary root growth together with LOG3 and LOG4. These results suggest a dominant role of the single-step activation pathway mediated by LOGs for cytokinin production, and overlapping but differentiated functions of the members of the LOG gene family in growth and development.


Cytokinins are a class of phytohormones originally identified as positive regulators of cell proliferation (Miller et al., 1955). Recent genetic studies in Arabidopsis cytokinin-deficient and -insensitive mutants confirmed that cytokinins play key roles in various processes of plant growth and development, including apical dominance, senescence, and meristem maintenance (Werner et al., 2003; Higuchi et al., 2004; Nishimura et al., 2004; Miyawaki et al., 2006; Riefler et al., 2006). The major natural cytokinins in plants are N6-prenylated adenines such as N6-(Δ2-isopentenyl)adenine (iP), trans-zeatin (tZ), cis-zeatin (cZ), and dihydrozeatin (DZ), which structurally vary at the prenyl side chain (Mok and Mok, 2001; Sakakibara, 2006). Cytokinins also occur as sugar conjugates and form nucleotides, nucleosides, and glucosides (Sakakibara, 2006). Generally, the nucleobases are the active forms, while other forms are less active or inactive (Sakakibara, 2006). In Arabidopsis thaliana, iP and tZ are the major active cytokinins, and the three cytokinin receptors, AHK2, AHK3, and AHK4/CRE1 (AHK4), bind the nucleobases as ligands (Romanov et al., 2006; Stolz et al., 2011).

The first step of the cytokinin biosynthetic pathway is catalyzed by adenosine phosphate-isopentenyltransferase, which transfers a prenyl moiety from dimethylallyl diphosphate to adenosine phosphate, producing the iP nucleotide (Kakimoto, 2001; Takei et al., 2001). The iP nucleotide is converted to the tZ nucleotide by CYP735A which hydroxylates the trans-end of the prenyl side chain (Takei et al., 2004). Biologically active nucleobases such as iP and tZ are generated from the corresponding nucleotides through ‘cytokinin-activation pathways’ (Sakakibara, 2006; Kurakawa et al., 2007; Zhao, 2008; Kamada-Nobusada and Sakakibara, 2009). The active cytokinins may be inactivated by cytokinin oxidase (CKX)-mediated irreversible degradation (Houba-Herin et al., 1999; Morris et al., 1999; Mok and Mok, 2001; Werner et al., 2006) or by conjugation catalyzed by cytokinin glycosyltransferases (Martin et al., 1999, 2001; Mok and Mok, 2001; Hou et al., 2004). They also may be converted to the nucleotides by enzymes of the purine salvage pathway including adenine phosphoribosyltransferase (Moffatt et al., 1991) and adenosine kinase (von Schwartzenberg et al., 1998; Kwade et al., 2005).

Although the general outline of cytokinin metabolism has been established, the process of cytokinin activation is not fully understood. A two-step and a single-step pathway have been proposed (Kamada-Nobusada and Sakakibara, 2009) (Figure S1). In the two-step pathway, it is postulated that nucleotides are converted to nucleosides by nucleotidase (Chen and Kristopeit, 1981a) and to nucleobases by nucleosidase (Chen and Kristopeit, 1981b), but the corresponding genes have not been identified. On the other hand, cytokinin riboside 5′-monophosphate phosphoribohydrolase, an important enzyme of the proposed single-step pathway that is encoded by LONELY GUY (LOG), has been identified through the analysis of rice (Oryza sativa) lonely guy (log) mutants that are deficient in the maintenance of shoot meristems (Kurakawa et al., 2007). In Arabidopsis, nine LOG genes (LOG1 to LOG9) were predicted as rice LOG homologs, and seven LOG proteins (LOG1 to LOG5, LOG7, and LOG8) showed phosphoribohydrolase activity although the values of the specificity constants (kcat/Km) varied (Kuroha et al., 2009). The expression patterns of LOG genes are spatially and quantitatively differentiated but overlap in various tissues during plant development (Kuroha et al., 2009), suggesting redundant functions of members of the LOG gene family. Analyses of the loss-of-function mutant log3log4log7 suggested that LOG genes control cytokinin activity during normal development in Arabidopsis (Kuroha et al., 2009). However, it still is possible that active cytokinins (i.e. nucleobases) are produced by the two-step pathway. Thus, the degree to which the proposed LOG-dependent pathway of cytokinin activation regulates cytokinin metabolism and thus controls plant growth and development remains unclear. In addition, the possible functional differentiation in the Arabidopsis LOG gene family remains an open question.

We generated various higher-order log mutants (up to septuple) in which we monitored cytokinin activation using stable isotope-labeled tracer, and characterized the mutants’ morphological and developmental phenotypes in Arabidopsis. Our results demonstrated that the LOG-dependent pathway plays a dominant role in cytokinin activation, both metabolically and developmentally. Detailed observation of the mutants suggested that LOG7 is the most important LOG gene for the maintenance of the shoot apical meristem (SAM) and normal primary root growth in Arabidopsis.


Establishment of a stable isotope tracer method to analyze the cytokinin-activation pathway

In order to assess the significance of the LOG-dependent pathway in the metabolic conversion from cytokinin nucleotides to active nucleobases, we first developed a method to monitor the metabolism of cytokinins in Arabidopsis seedlings using a stable isotope-labeled tracer. Due to secretion of acid phosphatases from Arabidopsis roots (Trull and Deikman, 1998) and the low permeability of the plasma membrane, it is assumed that cytokinin riboside 5′-monophosphate, the substrate of LOG, is not directly absorbed by the plant. Since it was demonstrated that cytokinin riboside is absorbed and rapidly converted into the nucleotide (Singh et al., 1988; Jameson, 1994; Letham, 1994), we adopted iP riboside (iPR) as exogenously supplied tracer.

We prepared the stable isotope-labeled iPR, whose molecular mass exceeds that of authentic iPR by 15 Da [iPR(+15); hereafter, values in parenthesis show the difference in molecular mass between the labeled compound and the authentic one]. It is worth noting that both the adenine and ribose moieties were labeled. Therefore, we could discriminate between primarily formed iPR 5′-phosphates (iPRPs) and secondarily formed iPRPs as iPRPs(+15) and iPRPs(+10), respectively, due to the addition of endogenous ribose through the salvage pathway (Figure S1).

The roots of Arabidopsis seedlings grown for 20 days on MGRL-agar plate were submerged in water containing 1 μm iPR(+15) and incubated for 15, 30, 60, 120, and 180 min. The labeled cytokinins and conjugates in whole seedlings were quantified by liquid chromatography-mass spectrometry (LC-MS). iPR(+15) was absorbed and accumulated up to around 5 pmol g−1 fresh weight (FW), and this concentration did not change significantly during the experimental period (Figure 1). iPRPs(+15) was detected at 15 min and accumulated to more than 100 pmol g−1 FW, more than 10-times higher than other metabolites including iP(+10) and iP-7-N-glucoside(+10) [iP7G(+10)]. This finding suggested that a large proportion of the absorbed iPR(+15) was converted to iPRPs(+15). The tZ riboside 5′-phosphates(+15) [tZRPs(+15)], which is produced from iPRPs(+15) by CYP735A, was detected at 15 min and plateaued at 60 min. tZ riboside(+15) [tZR(+15)] accumulated with similar kinetics as tZRPs(+15). The concentration of iP(+10) rose and leveled off in a similar pattern as those of iPRPs(+15), while tZ(+10) increased similarly as tZR(+10), tZR(+15), and tZRPs(+15) did. On the other hand, five derivatives that could be produced from iP(+10) or tZ(+10), namely iP7G(+10), iPR(+10), iPRPs(+10), tZ-7-N-glucoside(+10) [tZ7G(+10)], and tZRPs(+10), continued to accumulate during the experimental period. This suggested that iP(+10) and tZ(+10) are formed continuously through the two-step and/or single-step activation pathways, and then are transformed into the conjugates. Labeled O-glucosides, cZ-type and DZ-type cytokinins were not detected in this experimental condition. We concluded that the metabolic flows occurring during cytokinin activation could be monitored by our tracer method.

Figure 1.

 Changes in the accumulation levels of stable isotope-labeled cytokinins and their conjugates in Arabidopsis seedlings.
Arabidopsis seedlings grown for 20 days were incubated with 1 μm [1013C,515N]N6-(Δ2-isopentenyl) adenine riboside [iPR(+15)] for 15, 30, 60, 120, and 180 min. The labeled cytokinins and conjugates were analyzed by LC-MS. Each data point represents the mean ± standard deviation (SD) of three experimental replicates.

Changes in the metabolic status of the cytokinin-activation pathway in Arabidopsis log mutants

To understand how each LOG contributes to cytokinin activation, we prepared a series of multiple log mutants from the log1-2, log2-1, log3-1, log4-3, log5-1, log7-1, and log8-1 single mutants (Kuroha et al., 2009), and subjected them to tracer experiments together with the single mutants. We omit the allele numbers for these mutants in the following description. Arabidopsis seedlings were incubated in water containing 1 μm iPR(+15) for 60 min, and the labeled cytokinins and conjugates were quantified.

In the single mutants, the accumulation levels of iPR(+15), iPRPs(+15), iP(+10), and iP7G(+10) were not significantly changed compared to the wild type, except for that of iPRPs(+15) in log7 (Figure 2a). The level of tZRPs(+15) was significantly increased in log1, log3, log4, log7, and log8, and that of tZR(+15) was increased in log7 and log8. On the other hand, tZ(+10) showed decreased accumulation in log5 and log7, while tZ7G(+10) was lowered in log7 (Figure 2a). Among the single mutants, log7 showed the most pronounced changes. When we measured the concentrations of authentic cytokinins in the mutant seedlings, significant but only modest increases of cytokinin nucleotides were found in log7 as well as the other log mutants (Figure S2), showing that the levels of cytokinins and their conjugates are being kept stable.

Figure 2.

 Changes in the metabolism of iPR(+15) in the seedlings of wild type and log mutants.
(a) Comparison between wild type and log single mutants.
(b) Comparison between wild type and log multiple mutants. Arabidopsis wild type (WT) and log mutant seedlings grown for 20 days were incubated with 1 μm [1013C,515N]N6-(Δ2-isopentenyl) adenine riboside [iPR(+15)] for 60 min before labeled cytokinins and conjugates were analyzed by LC-MS. The numbers under each column represent the log mutants. Error bars represent the standard deviation (SD) of three biological replicates. Data sets marked with single and double asterisk differ significantly from the wild type as assessed by Student’s t-test at P < 0.05 and P < 0.01, respectively.

As LOGs are assumed to function redundantly (Kuroha et al., 2009), we examined iPR(+15) metabolism in higher-order mutants. The accumulation of tZRPs(+15) was significantly enhanced in log1log8, log2log7, and log4log7, as well as in all triple and quadruple mutants we examined. On the other hand, levels of tZ(+10) and tZ7G(+10) were decreased in log5log8, log2log7, log4log7, log2log4log7, log3log4log7, log1log2log4log7, and log1log3log4log7 (Figure 2b). The accumulation of tZR(+15) in the mutants showed a similar pattern as tZRPs(+15). In log1log2log4log7 and log1log3log4log7, levels of both iP(+10) and iP7G(+10) were significantly decreased.

We further analyzed quintuple (log1log2log3log4log7) and sextuple (log1log2log3log4log5log8 and log1log2log3log4log5log7) mutants. Although a severe inhibition of the conversion from tZRPs(+15) to tZ(+10) was found in log1log2log3log4log7 and log1log2log3log4log5log7, only a moderate inhibition was observed in log1log2log3log4log5log8. Taken together, these results indicated that LOG1, LOG2, LOG3, LOG4, LOG5, LOG7, and LOG8 contributed redundantly to cytokinin activation in vivo, and that the contribution of LOG7 was most important on the whole seedling level. In our experiments, the effects of log mutations on the conversion from nucleotide to nucleobase were more pronounced in tZ-type than in iP-type cytokinins (Figure 2b). This situation is probably due to the application of an excessive amount of iPR(+15).

Time-course analysis of the metabolic flow of cytokinin activation in the log1log2log3log4log7 quintuple mutant

To further characterize the LOG-dependent pathway, we performed time-course analysis of iPR(+15) metabolism in the log1log2log3log4log7 quintuple mutant. Although there were no differences at 5 min in the accumulation of iPRPs(+15) and tZRPs(+15) between wild type and mutant, a much larger amount of these compounds accumulated in the mutant in a time-dependent manner (Figure 3). While the accumulation of tZR(+15), the precursor of tZ(+10) in the two-step pathway, was also increased in the quintuple mutant, the levels of tZ(+10) were drastically reduced. The accumulation of iP(+10), iP7G(+10), and tZ7G(+10) also was retarded in the mutant. These results strongly suggest that the conversion of iPRPs to iP and that of tZRPs to tZ are mainly mediated by the LOG-dependent activation pathway rather than by the two-step pathway.

Figure 3.

 Time-course analysis of the accumulation of iPR(+15) metabolites in wild type and log1log2log3log4log7 mutant.
Arabidopsis wild type (open circles) and log1log2log3log4log7 mutant (closed circles) seedlings grown for 20 days were incubated with 1 μm [1013C,515N]N6-(Δ2-isopentenyl) adenine riboside [iPR(+15)] for 5, 15, 30, and 60 min before the labeled cytokinins and conjugates were analyzed by LC-MS. Error bars represent the standard deviation (SD) of six biological replicates. Data sets marked with single and double asterisk are significantly different from the wild type as assessed by the Student’s t-test at P < 0.05 and P < 0.01, respectively.

Morphological phenotypes of log multiple mutants

Previously we had found that log2log7 and log3log4log7 are less sensitive to cytokinin nucleoside in lateral root formation, and that log3log4log7 shows increased adventitious root formation and decreased flower number (Kuroha et al., 2009). However, other phenotypes typical of cytokinin deficiency were not observed in these mutants. To better understand the role of the LOG-dependent pathway in growth and development, we generated a log1log2log3log4log5log7log8 septuple mutant. We confirmed T-DNA insertions and examined the expression of LOG genes in the septuple mutant because in some log mutants, the T-DNA is inserted in an intron (Figure S3; Kuroha et al., 2009). Reverse-transcription polymerase chain reaction (PCR) with saturated reaction cycles detected significant amounts of LOG1 transcript in the septuple mutant (Figure S3b). Further analysis with quantitative reverse-transcription PCR revealed that the accumulation level of LOG1 transcript in the multiple mutants containing log1 was about 20% of that in the wild type at a maximum (Figure S3c). In addition, trace amount of LOG8 transcript was also detected in the septuple mutant (Figure S3b). As no other loss-of-function mutants of LOG1 and LOG8 were available, we investigated the available septuple mutant as a LOG-impaired mutant.

We analyzed the morphological phenotypes of log multiple mutants in various developmental stages. There were no aberrant phenotypes under standard growth conditions in any double mutant that we generated (log1log3, log1log4, log1log5, log1log8, log2log7, log3log4, log3log5, log3log7, log3log8, log4log7, log5log8). In addition to the previously reported phenotype in the log3log4log7 mutant in which adventitious root formation and flower number were affected (Kuroha et al., 2009), careful observation revealed that log3log4log7 and the higher-order mutants containing log3log4log7 show short primary root (Figure 4).

Figure 4.

 Root phenotype of log multiple mutants.
(a) Thirteen-day-old seedlings of wild type and the log multiple mutants grown on 1/2 MS-agar medium.
(b) Length of primary root at 8-day-old seedlings grown on 1/2 MS-agar medium. The numbers under each column represent the log mutants. WT, wild type. compL7, T3 homozygous log1234578 mutant transformed with pBG-LOG7. Error bars represent the standard deviation (SD) ( 14). Data sets marked with asterisk differ significantly from the wild type as assessed by Student’s t-test at P < 0.01.

Furthermore, triple, quadruple, quintuple, and sextuple mutants that contained log7 developed fewer rosette leaves and showed a delayed bolting time (Table 1). These changes may have resulted from a prolonged plastochron as the leaf numbers at the bolting stage were barely affected, and because the size of the SAM was reduced in log1log2log3log4log7 in comparison to the wild type (Werner et al., 2003; Miyawaki et al., 2006) (Figure 5a). It is noteworthy that no prolonged plastochron was observed in log1log2log3log4log5log8 (Table 1). On the other hand, although no significant change in leaf number was observed in the log7 single mutant, the meristem size was smaller than in the wild type (Figure 5b,c). These results suggest that LOG7 plays a more important role for the SAM function than other LOG members.

Table 1.   Leaf formation and flowering in log mutants
LineNumber of rosette leavesaBolting time (DAG)Number of leaves at bolting time
At 15 DAGAt 22 DAG
  1. Means ± standard deviation (SD) (n ≥ 13). Data sets marked with a single and double asterisk are significantly different from the wild type as assessed by Student’s t-test at P < 0.05 and P < 0.01, respectively.

  2. aBolting time was defined as the time when the inflorescence reached 0.5 cm length.

  3. DAG, days after germination.

Wild type7.7 ± 0.512.9 ± 0.623.7 ± 1.014.0 ± 1.1
log3log47.6 ± 0.512.7 ± 0.524.0 ± 0.914.1 ± 0.8
log3log77.5 ± 0.512.9 ± 0.723.9 ± 1.213.9 ± 1.2
log2log77.7 ± 0.512.6 ± 0.523.8 ± 0.913.9 ± 1.1
log4log77.6 ± 0.512.1 ± 0.6**24.1 ± 1.413.7 ± 1.3
log2log4log77.2 ± 0.4**11.7 ± 0.7**24.8 ± 1.3*13.8 ± 1.3
log3log4log76.2 ± 0.4**10.4 ± 0.9**26.7 ± 1.2**14.4 ± 0.9
log1log2log4log77.1 ± 0.5**12.1 ± 0.5**25.8 ± 1.1**14.5 ± 0.9
Wild type7.8 ± 0.412.2 ± 0.722.0 ± 0.912.8 ± 1.0
log1log2log3log4log5log87.8 ± 0.412.3 ± 0.622.9 ± 0.912.6 ± 0.6
log1log2log3log4log76.3 ± 0.6**9.5 ± 0.5**25.7 ± 1.3**13.4 ± 0.8
log1log2log3log4log5log76.1 ± 0.4**9.2 ± 0.6**26.3 ± 1.7**13.4 ± 0.6
Figure 5.

 Shoot apical meristem (SAM) size of 11-day-old seedlings of wild type and log mutants grown on 1/2 MS-agar medium. The diameters of SAMs of wild type (WT, mean = 92 μm, n = 7) and log12347 (mean = 67 μm, n = 9), and that of WT (mean = 85 μm, n = 14) and log7 (mean = 68 μm, n = 6) are shown as box plots in (a) and (b), respectively. (c) Comparison of leaf formation and flowering in WT and log7. Means ± standard deviation (SD) (n ≥ 13). Bolting time was defined as the time when the inflorescence reached 0.5 cm length. DAG, days after germination.

In contrast to the modest phenotypes in the lower-order log mutants, the LOG-impaired log1log2log3log4log5log7log8 septuple mutant displayed a severe retardation of shoot and root growth and development (Figures 4 and 6a,b). The septuple mutant set flowers and seeds (Figure 6c), and the seeds were larger than in the wild type (Figure 6d,e). Such seed size increase is known from other cytokinin-deficient mutants (Werner et al., 2003; Miyawaki et al., 2006). The growth retardation of the septuple mutant was rescued by transformation with the genomic region of wild-type LOG7 (Figures S4 and 4b). It is notable that the cytokinin receptor triple mutant ahk2ahk3ahk4, which exhibits the severest phenotype among cytokinin-insensitive and -impaired mutants characterized so far, shows defects in shoot and root apical meristems (Higuchi et al., 2004; Nishimura et al., 2004; Riefler et al., 2006). Therefore, we examined the SAMs and found their sizes markedly reduced in the septuple mutant (Figure 6f–h). The size of the root apical meristem was also decreased (Figure 6i,j). The roots of septuple mutants had a reduced diploid (2C) content (Figure 6k) and reduced cell numbers within vascular bundle (Figure 6l–o), which also had been found in the ahk2ahk3ahk4 triple mutant (Higuchi et al., 2004; Nishimura et al., 2004; Mähönen et al., 2006). The levels of cytokinin nucleotides were increased more than threefold, whereas those of nucleobases and glucosides were decreased to about one-third in the septuple mutant (Tables 2 and S1). These results strongly suggested that active cytokinins produced in the LOG-dependent pathway play a critical role for the maintenance of shoot and root meristems.

Figure 6.

 Morphological phenotypes of the log1log2log3log4log5log7log8 septuple mutant.
(a, b) Eleven-day-old seedlings of wild type (a) and the log septuple mutant (b) grown on 1/2 MS-agar medium.
(c) Terminal stature of wild type (WT) and the septuple mutant (1234578) plants grown on soil.
(d, e) Seeds of wild type (d) and the septuple mutant (e). (f, g) Shoot apical meristems of 11-day-old seedlings of wild type (f) and the septuple mutant (g). (h) Box plot of the SAM diameter of wild type (mean = 85 μm, n = 14) and the septuple mutant (1234578, mean = 46 μm, n = 13).
(i, j) Cellular organization of the root meristem visualized by propidium iodide-staining in 8-day-old seedlings of wild type (i) and the septuple mutant (j) grown on the MS-agar medium. The white arrowhead indicates the position at which cell expansion in epidermal cell layers initiates.
(k) Flow cytometric analysis of the ploidy of 4-day-old seedling roots of wild type (open columns) and the septuple mutant (filled columns). Results are expressed in ratios of tetraploid (4C) and octaploid (8C) to diploid (2C). Error bars represent the standard deviation (SD) of three biological replicates.
(l–o) Transverse sections of 4-day-old root of wild type (l, m) and log septuple mutant (n, o). Red arrowheads indicate endodermal cells. Scale bars: (a, b), 5 mm; (d, e), 1 mm; (f, g), 20 μm; (i, j, l, n), 50 μm; (m, o), 10 μm.

Table 2.   Changes in the cytokinin concentrations in the log septuple mutant
  1. Cytokinin (CK) concentrations in 11-day-old Arabidopsis seedlings of wild type and log1log2log3log4log5log7log8 (log1234578) grown on MGRL-agar medium are shown in pmol g−1 FW; the percentages of each cytokinin type in the total cytokinin are shown in parenthesis. Means ± standard deviation (SD) (n = 3). The complete data set is presented in Table S1.

Wild type12.9 ± 0.3 (19.4)2.8 ± 0.3 (4.3)1.3 ± 0.3 (1.9)49.4 ± 2.2 (74.4)
log123457853.1 ± 2.3 (63.4)7.7 ± 2.5 (9.1)0.50 ± 0.17 (0.6)22.4 ± 1.7 (26.8)


Although it has been postulated that two biochemical pathways are involved in cytokinin activation (Kurakawa et al., 2007; Kuroha et al., 2009), the contribution of LOG proteins to cytokinin activation and developmental control have remained unclear. Our present study provides metabolic and morphological data that highlight the critical role of the LOG-dependent pathway of cytokinin production, especially for the maintenance of apical meristems in Arabidopsis.

In spite of the severe cytokinin-deficient phenotypes in the log septuple mutant, the mutant set flowers and seeds (Table 2 and Figure 6c–e). Given that ahk2ahk3ahk4 showed similar but more severe phenotypes that often failed to set seeds (Higuchi et al., 2004; Nishimura et al., 2004; Riefler et al., 2006), it seems that minimum amounts of active cytokinins required for seed development are still produced in the log septuple mutant. In fact, cytokinin nucleobases and the glucosides were detected in the mutant (Tables 2 and S1). There are several possible explanations. First, this study did not consider LOG6 and LOG9 because no functional LOG6 and LOG9 mRNAs have been detected in Arabidopsis so far (Kuroha et al., 2009). However, it is possible that LOG6 and LOG9 transcripts are alternatively spliced in certain circumstances, resulting in the production of functional LOGs. Second, as log1 and log8 are knockdown mutants (Figure S3b,c), the remaining LOG activity possibly produces sufficient amounts of cytokinin nucleobases in the septuple mutant. Third, we cannot rule out that active cytokinins are formed by the two-step and/or other pathways in the reproductive growth stage. Finally, cytokinin degradation may be regulated to maintain homeostasis of active cytokinins. In fact, the expression level of some CKX genes in the log septuple mutant was decreased compared to wild type (Figure S5).

In young seedlings of the log1log2log3log4log7 mutant, the level of tZRPs(+15) was significantly elevated, and those of tZ(+10) and tZ7G (+10) were markedly reduced in time-course tracer experiments (Figure 3). Moreover, tZR(+15) accumulated in a similar pattern as tZRPs(+15). Notably, tZR(+15) is the product of dephosphorylation of tZRPs(+15). Thus, our results indicate that the conversion rate from the cytokinin nucleoside to the nucleobase in the two-step pathway is much slower than the LOG-catalyzed conversion from the nucleotide to the nucleobase. These lines of evidence suggest that cytokinin activation is predominantly mediated by the LOG-dependent pathway in the vegetative growth stage.

Our tracer experiments suggested that LOG7 is the major contributor to cytokinin activation on the whole plant level. However, we should be cautious in interpreting the results because the labeled iPR was applied exogenously to the root. Considering that LOG7 has a relatively high specificity constant (kcat/Km) and that the LOG7 expression level is higher than that of other LOGs in the root (Kuroha et al., 2009), the activity of LOG7 might be somewhat overestimated. Nonetheless, the results from our morphological studies are in line with our conclusion from tracer experiments that LOG7 plays a prominent role in cytokinin activation.

In root, multiple mutants containing log3log4log7 developed short primary root (Figure 4), indicating that cooperation of these 3 LOGs are necessary for normal root growth. Among the three, LOG7 must play an important role because log1log2log3log4log5log8 sextuple mutant showed milder phenotype than log3log4log7 triple mutant (Figure 4). In our previous study, the phenotype was not observed in log3log4log7 root (Kuroha et al., 2009). This discrepancy may be due to the difference of growth medium: the MS salt was used in this study while the MGRL was used in previous study. In the log multiple mutants containing log3log4log7, primary root growth was cumulatively inhibited and finally terminated in the septuple mutant (Figure 4). Termination of primary root growth was reported in some higher-order mutants of cytokinin signaling, such as ahk2ahk3ahk4, ahp1ahp2ahp3ahp4ahp5 and arr1arr10arr12 (Higuchi et al., 2004; Nishimura et al., 2004; Hutchison et al., 2006; Riefler et al., 2006; Argyros et al., 2008; Ishida et al., 2008). On the other hand, lower-order cytokinin signaling mutants, ipt multiple mutants, and CKX-overexpression lines showed enhanced root elongation (Werner et al., 2001, 2003; Higuchi et al., 2004; Nishimura et al., 2004; Hutchison et al., 2006; Miyawaki et al., 2006; Riefler et al., 2006; Dello Ioio et al., 2007; Argyros et al., 2008; Ishida et al., 2008; Matsumoto-Kitano et al., 2008), indicating that, although cytokinin acts to inhibit root elongation at relatively high revels, a basal level of cytokinin is required for root meristem maintenance. Together, our results suggest that local activation of cytokinin by LOG-dependent pathway might be critical for supplying the basal level of cytokinin to the root meristem. Consistently, in transgenic Arabidopsis lines harboring LOG promoter:β-glucoronidase chimeric genes, LOG3 and LOG4 were expressed in root procambium and LOG7 was expressed in root elongation zone (Kuroha et al., 2009).

In aerial organs, LOG7 is expressed in a restricted region in the upper part of the SAM (Yadav et al., 2009). As the spatial expression pattern of rice LOG (Kurakawa et al., 2007) is similar, it appears that the production of active cytokinins in this specific region is important for SAM maintenance. The effects on the size of the SAM in log1log2log3log4log7 and log7 were comparable, whereas the plastochron was significantly prolonged in the quintuple mutant only (Figure 5 and Table 1). A cooperative function of multiple LOGs could be involved in the determination of the plastochron.

In addition to LOG7, phenotypic difference between log1log2log3log4log5log7log8 and log1log2log3log4log5log7 (Figure 4) also suggest a role of LOG8 for producing active cytokinins for normal growth. At present, it is difficult to mention the importance of other LOGs because we do not have all combinations of log mutants.

The expression patterns of LOG genes are spatiotemporally differentiated (Kuroha et al., 2009). Although a series of mutants excluding log7 mutation did not show aberrant phenotypes, some morphological changes might locally occur in specific cells or tissues in these mutants. Detailed comparative studies of such mutants would enable us to reveal a specific role of each LOG for plant development.

Recent studies suggest that a localized cytokinin response acts as a positional cue for patterning the expression of the homeodomain transcription factor WUSCHEL (WUS) in the SAM, and that this cue requires a local concentration peak of cytokinin and/or a locally increased cytokinin perception capability (Gordon et al., 2009). In fact, an enlarged WUS domain in the SAM was induced by increased cytokinin concentrations in the ckx3ckx5 double mutant (Bartrina et al., 2011). Furthermore, the regulation of WUS expression is precisely controlled by the location of cytokinin receptors and cytokinin signaling is highest in the center of the SAM (Gordon et al., 2009). Thus, the site of cytokinin perception may be different from the site of cytokinin production in the SAM. Further studies will provide new insights into the biological significance for meristem function of the local activation of cytokinins by LOGs.

Experimental Procedures

Plant material and growth conditions

Arabidopsis thaliana ecotype Columbia (Col-0) was used as the wild-type plant. T-DNA insertional mutants of log1-2, log2-1, log3-1, log4-3, log5-1, log7-1, and log8-1 had been identified in our previous study (Kuroha et al., 2009). Multiple log mutants were obtained by crossing and identified by PCR with gene- and T-DNA-specific primers (Kuroha et al., 2009). For tracer labeling experiments, wild type and a series of log mutants were grown on 1.5% (w/v) agar plates with MGRL salt (Fujiwara et al., 1992) supplemented with 1% (w/v) sucrose at 22°C under fluorescence illumination at an intensity of 70 μmol m−2 s−1 with a photoperiod of 16 h light per 8 h dark. For phenotypic analysis, plants were grown on 1% agar plates with half-strength MS salt (Murashige and Skoog, 1962) plus 1% sucrose for 7 days and then transferred to soil (Supermix A; SAKATASEED,

Synthesis of labeled iPR

iPRMP(+15) was synthesized from [1013C,515N]-adenosine 5′-monophosphate and dimethylallyl diphosphate with recombinant Tzs (Sugawara et al., 2008) and purified using MonoQ HR 5/5 columns (Amersham Biosciences, as described previously (Takei et al., 2003). iPRMP(+15) was dephosphorylated by alkaline phosphatase to form iPR(+15). The quantity and purity were checked with an LC-MS system (model 2695/ZQ2000MS; Waters) on an Octadecylsilyl column (Supersphere RP-select B, 4 × 250 mm; Merck, (Takei et al., 2001, 2004). The mass spectrum of the synthesized iPR(+15) is shown in Figure S6.

In vivo labeling tracer experiments

Arabidopsis seedlings were grown aseptically on MGRL plates for 20 days, and three plantlets were transferred to each well in dishes (Micro Plate 12well/Flat Bottom, Asahi Glass; IWAKI, containing 600 μl water with 1 μm iPR(+15). After incubation at 22°C under illumination, the seedling was washed twice with deionized water, weighed, and immediately frozen in liquid nitrogen. Stable isotope-labeled and authentic cytokinins were analyzed as described elsewhere (Kojima et al., 2009). Monitored ion transitions for the quantification of labeled cytokinin species were as follows: 351.2/214.2 for iPR(+15), 346.2/214.2 for iPR(+10), 367.2 > 230.2 for tZR(+15), 362.2 > 230.2 for tZR(+10), 214.2 > 146.2 for iP(+10), 230 > 146.1 for tZ(+10), 376.2 > 214.2 for iP7G(+10), and 392.2 > 230.2 for tZ7G(+10). Other conditions for MS analysis were as described previously (Kojima et al., 2009).

Complementation analysis

pBG-LOG7, which covers the genomic region of the LOG7 gene as well as the putative promoter region (Kuroha et al., 2009), was introduced into the log1log2log3log4log5log7log8 septuple mutant by the floral dip method (Clough and Bent, 1998). T2 and T3 lines were used for complementation analysis.

Ploidy measurement

Ploidy levels were determined from whole root grown for 4 days using the ploidy analyzer PA-I and the reagents (Partec, Görlitz, previously described (Ishida et al., 2010). Nuclei isolated from at least 10 plants were used for each ploidy measurement.

Histological observation

For observation of SAM, wild type and log mutants were grown on soil (Supermix A) for 11 days after germination, and then fixed and stained as detailed in previous studies (Kurakawa et al., 2007). For observation of cell arrangement in root vascular bundle, the Arabidopsis seedlings were grown for 4 days after germination, and then fixed, embedded in Technovit 7100 (Heraeus Kulzer, Sections were prepared with Leica RM2165 (Leica Microsystems, and stained with toluidine blue-O as previously described (Kiba et al., 2004). Cross sections were examined with an Olympus BX-51 microscope (Olympus,

RT-PCR analysis

Total RNA was prepared using the RNeasy Plant Mini Kit with RNase-free DNase I (Qiagen). cDNA was synthesized using SuperScript III RT (Invitrogen) with oligo(dT)12–18 primers. The primers used for PCR were shown in (Kuroha et al., 2009). Quantitative RT-PCR was performed on a StepOnePlus Realtime PCR system (Applied Biosystems, by monitoring the amplification with SYBR Premix ExTaq II (Takara, The primers for quantitative RT-PCR were as follows: 5′-GCTGCTATCGAACTCGGAAC-3′ and 5′-GCGACCTCCATTGAAAACAG-3′ for LOG1, 5′-AGTGATGCTGCCATCGAACT-3′ and 5′-CCCATCAATCCAACACTTCC-3′ for LOG8, 5′-GCTTGGACAGTTTGGCATAAT-3′ and 5′-TGGTCCCTTGAAAATGCAGA-3′ for CKX1, 5′-GAGGAACGTTGTCGAATGGT-3′ and 5′-GTTTAGCTGTCGCGAGCATG-3′ for CKX2, 5′-ACAGTCGGTGGGACGTTATC-3′ and 5′-CCTAACACCGCGAAGAAAAG-3′ for CKX3, 5′-GTTTAGACACGGCCCTCAGA-3′ and 5′-CAATCCTGGCCCTCGTTAT -3′ for CKX4, 5′-TGGTATCATCACCAGAGCTAGG-3′ and 5′-ACTCGGCGTCTTGAGTGAACTC-3′ for CKX5, 5′-AGATCACCAAGAACTACCACG-3′ and 5′-TCCGTTGTAAAGACCGATGTC-3′ for CKX6, 5′-ATTCCGACATGGACCACAGA-3′ and 5′-GTCGCTGTTCTGCCTCTTTG-3′ for CKX7, and 5′-GAAGTTCAATGTTTCGTTTCATGT-3′ and 5′-GGATTATACAAGGCCCCAAA-3′ for Ubiquitin 10.

Accession Number

Sequence data from this article can be found in the Arabidopsis Genome Initiative of GenBank/EMBL databases under the following accession numbers: Arabidopsis sequences: LOG1, At2g28305; LOG2, At2g35990; LOG3, At2g37210; LOG4, At3g53450; LOG5, At4g35190; LOG6, At5g03270; LOG7, At5g06300; LOG8, At5g11950; LOG9, At5g26140; CKX1, At2g41510; CKX2, At2g19500; CKX3, At5g56970; CKX4, At4g29740; CKX5, At5g21482; CKX6, At1g75450; CKX7, At3g63440; UBIQUITIN 10, At4g05320 and ACT2, At3g18780.


This work was supported by a Grant-in-Aid for Scientific Research on Innovative Areas (No. 21114005) and for Scientific Research (B) (No. 21370023) from the Ministry of Education, Culture, Sports, Science and Technology, Japan. H. Tokunaga was supported by Scientific Research for Plant Graduate Students from Nara Institute of Science and Technology.