The Salix SmSPR1 Involved in Light-Regulated Cell Expansion by Modulating Microtubule Arrangement

Light signaling and cortical microtubule (MT) arrays are essential to the anisotropic growth of plant cells. Microtubule-associated proteins (MAPs) function as regulators that mediate plant cell expansion or elongation by altering the arrangements of the MT arrays. However, current understanding of the molecular mechanism of MAPs in relation to light to regulate cell expansion or elongation is limited. Here, we show that the microtubule-associated protein SPR1 is involved in light-regulated directional cell expansion by modulating microtubule elongation in Salix matsudana. Overexpression of SmSPR1 in Arabidopsis results in right-handed helical orientation of hypocotyls in dark-grown etiolated seedlings, whereas the phenotype of transgenic plants was indistinguishable from those of wild-type plants under light conditions. Phenotypic characterization of the transgenic plants showed reduced anisotropic growth and left-handed helical MT arrays in etiolated hypocotyl cells. Protein interaction assays revealed that SPR1, CSN5A (subunits of COP9 signalosome, a negative regulator of photomorphogenesis), and ELONGATED HYPOCOTYL 5 (HY5, a transcription factor that promotes photomorphogenesis) interacted with each other in vivo. The phenotype of Arabidopsis AtSPR1-overexpressing transgenic lines was similar to that of SmSPR1-overexpressing transgenic plants, and overexpression of Salix SmSPR1 can rescue the spr1 mutant phenotype, thereby revealing the function of SPR1 in plants. Highlight Function of microtubule-associated protein SPR1 is directly related to light, and crucial to the balance of tubulin polymerization


Introduction
For the spr1 mutant of the Arabidopsis complement test and AtSPR1 overexpression assay, SmSPR1 159 and AtSPR1 cDNA was amplified and introduced into pDONR221 via BP reaction and to 160 pEarleyGate104 of Gateway vectors via LR recombinase (Invitrogen) (Earley et al., 2010). All primers 161 are listed in Supplemental Table S4. The resulting constructs were transformed into Arabidopsis using 162 Agrobacterium tumefaciens (GV3101) via Arabidopsis floral dip method as described elsewhere (Zhang 163 et al., 2006). The homozygous T3 seedlings were used for further analyses.

164
Phenotypic analysis 165 The spiral phenotypes of seedlings were observed using an Ultra depth of field microscope (Leica 166 DVM6) equipped with CCD (PLANAPO FOV 12.55). For measuring length and width of hypocotyl and 167 root, the relevant parameter was measured using ImageJ (http://rsb.info.nih.gov/ij/). Hypocotyls of 168 five-day-old seedlings were fixed with 50% FAA and then embedded in spr resin (SPI). A series of  For drug treatment, wild and transgenic seeds were grown on 0.8% agar-solidified 1/2 MS vertically 173 oriented plates for 7 days with or without specific concentration of Propyzamide (Sigma-Aldrich). To 174 detect the morphology of the cells of 7-day-old seedlings, etiolated hypocotyls and roots were soaked in 7 10 μM PI (Sigma-Aldrich). Zeiss LSM510 confocal microscope (with 543 nm diode laser, and an 176 emission band of 560 to 690 nm) was used for images collecting. 177 For measurement of MT arrays, SmSPR1 was over-expressed in 35S: GFP-TUB6 background and 178 detected on a Zeiss LSM510 confocal microscope. The orientation of cortical MTs in epidermal cell was 179 measured at upper regions. Measurements were performed using ImageJ (http://rsb.info.nih.gov/ij/). 180 Microtubules with clear visible were selected for measurements in each cell (n ≥ 25cells). The procedure 181 was performed as previously described (Liu et al., 2013).  Yeast Two-Hybrid analysis 189 The CDS of SmCSN5A, SmCOP1 and SmHY5 were constructed on the yeast two-hybrid prey vector 190 pGADT7, respectively. SmSPR1 was constructed on the bait vector pGBKT7. The primers were listed 191 in Supplemental Table S6. The bait vector of SmSPR1 was transformed with SmCSN5A, SmCOP1 and 192 SmHY5 respectively into the yeast strain AH109 as instructions for Matchmaker  Systems 3 (Clontech). Yeast Two-Hybrid analysis were performed following Yeast Protocols Handbook 194 (Clontech). Transformed yeast cells were separately spread to 2D synthetic deficiency medium (SD-TL:

218
Cloning and Expression Pattern of the SmSPR1 Genes 219 A total of six SmSPR1 genes were identified according to the sequences of the Arabidopsis SPR1 220 family genes (Nakajima et al., 2006;Sedbrook et al., 2004). These SmSPR1s were then isolated from S. 221 matsudana using PCR-based approaches with gene-specific primers (Supplemental Table S1). We 222 named these Salix SPR1 genes as SmSPR1 and SmSPR1-LIKE genes (SmSPR1_L1-SmSPR1_L5) based 223 on their amino acids sequence identity and phylogenetic relationship with Arabidopsis SPR1 family 224 genes ( Fig. 1A). All Salix and Arabidopsis SPR1s were classified into three classes (designated to Class 225 I-III): Class I included Salix SPR1, SPR1_L3, SPR1_L4, with Arabidopsis SPR1 and SPRL2; Class II 226 consisted of Salix SPR1_L1 and SPR1_L2 and Arabidopsis SPR1L3, SPR1L4 and SPR1L5; and Class 227 III comprised Salix SPR1_L5. Salix SPR1 and its homolog sequences shared N-and C-terminal regions 228 except SmSPR1_L5, and the outgroup position of SmSPR1_L5 may be due to the absent of conserved 229 C-terminal region (Fig. 1B). Highly conserved repeat amino acids sequences were observed at the N-230 and C-termini in SmSPR1, L1, L2, and L4, with the consensus motif being GGG/DQ/SSSLG/DY/FLFG 231 ( Fig. 1B). At the C-terminal of this conserved motif, the PGGG sequence is present in many mammalian 232 MAPs and is a conserved binding sequence of microtubules (Nakajima et al., 2004).

233
To determine SmSPR1 expression level at different tissues, we conducted a qRT-PCR-based 234 tissue-specific transcript abundance analysis of five tissues (shoot tips, xylem, phloem, leaves, and roots) 9 from three trees with gene-specific primers, and the transcript expression level of six Salix SPR1 genes 236 are shown in Fig. 1C. Class I SPR1 gene had the highest expression level and was detected at almost 237 equal levels in all tissues tested. SPR1_L3 and SPR1_L4, which also belong to Class I, were expressed in 238 all tissues, with a moderate transcript level compared to SPR1. The expression levels of Class II 239 SPR1_L1 and SPR1_L 2 were significantly lower than those of the Class I genes, and extremely low 240 transcript levels were observed in the roots of SPR1_L2 and SPR1_L5. The relative expression levels of 241 Class I SPR1 members were higher than those of Class II and III SPR1 members, suggesting the Class I 242 SPR1 family genes, especially SmSPR1, is the major gene in Salix, similar to the results of Arabidopsis 243 AtSPR1, which also had a predominant transcript level in all tissues tested (Nakajima et al., 2004).

244
To further obtain details on the tissue specific expression pattern of SmSPR1, we generated transgenic 245 tobaccos (Nicotiana tabacum) of P SmSPR1 : GUS which were then stained for GUS for GUS activity test.

246
Strong GUS activity was observed at the internodes of stems, including phloem, cambium, and xylem, 247 but not in epidermal cells (Fig. 1D). Midveins at each internode were also stained for GUS activity, and 248 a similar expression pattern was obtained as that observed in the stems; GUS staining was also observed 249 in vascular tissues, which included strong GUS staining in the phloem, moderate GUS staining of the   (Furutani et al., 2000). This model has been used to explain the spiral phenotype of spr1 (Nakajima 486 et al., 2004;Nakajima et al., 2006;Sedbrook et al., 2004). Overexpression of SmSPR1 was also 487 observed in the helix phenotype; the results of semi-thin section assays and PI staining experiments 488 showed isotropic growth and expansion of cells in transgenic plants (Figs. 3B, D). However, the 489 hypocotyls of SmSPR1 overexpressing transgenic lines did not become shorter (Fig. 2). The shortening 490 of hypocotyls is one of the main characteristics of the above model, which is discordant with our results 491 and cannot explain the SmSPR1 overexpression helix phenotype. According to the model,