Arabidopsis thaliana genotypes used are WT (Columbia-0), vps45-3 (TILLING mutant), vps45-3 COM (complemented line), WT-EYFP-RabF2a, vps45-3-EYFP-RabF2a. All genotypes were grown at 22°C either on soil in growth chambers or on sterile nutrient media under light racks. Soil-grown plants were kept in long day (16 hr light/8 hr dark) conditions. For growth on nutrient media, seeds were surface sterilized in 33% (v/v) bleach and 0.1% (v/v) Triton X-100 (Thermo Scientific, AAA16046AP) solution for 10 minutes and washed with sterile water at least five times. After two days of cold treatment in the dark, the seeds were plated on solid ¹/₂-strength Murashige-Skoog (MS) medium with vitamins (Caisson Labs, MSP09), 1% (w/v) sucrose (IBI scientific, IB37160), 2.4 mM 2-morphinolino-ethanesulfonic acid pH 5.7 (Sigma-Aldrich, M3671) and 0.8% (w/v) Phytoagar (Caisson Labs, PTP01). vps45-3 was generated by TILLING (Colbert et al., 2001) and mutants identified using forward primer 5’ -TGGCGTTGAAACGAAGACCTGTCA-3’ and reverse primer 5’-GAGCAGGACTTGGCTTGCAATGGT-3’ as described (de Vere et al., 2015). The point mutation introduces a novel MseI restriction site in a 998bp or 587bp gDNA or cDNA region respectively. Homozygous point mutants were identified by PCR amplification of this 998bp region containing the novel restriction site followed by digestion with MseI restriction enzymes at 37 °C for 1 hr. Upon gel electrophoresis, WT gDNA results in 3 bands of 422, 333 and 243 bp while vps45-3 gDNA results in 4 bands of 422, 243,213 and 120 bp. Restriction digestion of amplified cDNA results in two bands of 328bp and 259 bp for vps45-3 cDNA while the WT cDNA lacks this site and thus results in a single 587 bp band.

Complementation of vps45-3 mutants was performed by introducing a binary vector containing the VPS45 coding sequence driven by the VPS45 endogenous promoter described in (Zouhar et al., 2009). Plants were transformed using Agrobacterium tumefaciens by the floral dip method (Clough and Bent, 1998). Complemented lines were identified by resistance to hygromycin (30 mg L⁻¹) and MseI restriction digestion as described above. Homozygous transformant lines were identified by appearance of all three bands in the restriction profile and resistance to hygromycin in subsequent progeny of the primary transformants.

EYFP- RabF2a (Preuss et al., 2004) constructs were generously provided by Dr. Erik Nielsen. All constructs were introduced into Arabidopsis by the floral dip method (Clough and Bent, 1998). Transgenic plants were selected on kanamycin and were imaged by confocal microscopy using a YFP filter at excitation and emission wavelengths of 488 nm and 528 nm respectively.

Flowers were collected from Arabidopsis plants 1 to 2 weeks after bolting and dehydrated at room temperature for at least 2 hrs. Pollen was germinated on an agar medium containing 18% sucrose (IBI Scientific, IB37160), 0.01% boric acid, 1 mM MgSO₄, 1 mM CaCl₂, 1 mM Ca(NO₃)₂, and 0.5% agar, pH 7 (Li et al., 1999) at room temperature for 12 hours. It was then examined under a Zeiss MacroZoom light microscope (Carl Zeiss Inc., Jena, Germany) and photographed with a 35 mm camera. Pollen tube lengths were measured as the distance from the pollen grain to the pollen tube tip, using segmented line and length measurements with the ImageJ software (Schneider et al., 2012). Average length and standard deviations for 100 pollen tubes were calculated for 3 independent biological replicates, n = 100.

Five-day-old seedlings were mounted on a slide and imaged using a Zeiss AxioImager microscope (Carl Zeiss Inc., Jena, Germany) with a 20X objective with bright field and differential interference contrast (DIC). Root hair length quantification was carried out by using segmented line and length measurements with the ImageJ software. Average length and standard deviations among at least 100 root hairs were calculated. Root hairs were imaged in the root elongation zone and neighboring cells in the early maturation zone, while excluding the older maturation zone cells.

FM4-64 staining was modified from (Dettmer et al., 2006). To test bulk endocytosis, 4-day-old seedlings were transferred to MS liquid medium containing 4 μM FM4-64 (Invitrogen, T3166) for 2 min and subsequently washed twice for 30 s each time in 0.5× MS liquid medium before visualization. For analyzing arrival of FM4-64 at Brefeldin A (BFA) bodies, 4-day-old seedlings were transferred to 0.5× MS liquid medium containing 35 μM BFA (Sigma-Aldrich, B7651) for an hour followed by a 10 min treatment with 4 μM FM4-64 plus 35 μM BFA and two subsequent washes of 30 s each. The root tips were visualized using a Leica SP5 confocal laser scanning microscope (Leica Microsystems, Wetzlar, Germany) at the Iowa State University Roy J Carver High Resolution Microscopy Facility, using a 63× oil immersion objective lens and excitation and emission wavelengths of 558 and 734 nm. Images were acquired under identical conditions for both genotypes with equal exposure, scan frequency and line average settings. A total of 15 seedlings from at least three independent replicates were observed for each treatment and genotype.

0.8 g of WT and vps45-3 7-day-old seedlings were collected and ground in 1 mL of cold extraction buffer (0.3 M Sucrose, 0.1 M Tris-HCl, 1 mM EDTA, pH 7.5), with protease inhibitor cocktail (Roche). This was followed by centrifugation at 2,800g for 5 minutes at 4^oC. A 100 μl portion of the supernatant was kept as the total protein fraction and the rest of the supernatant was transferred to a new tube followed by centrifugation at 13,000g for 30 minutes at 4^oC. The pellet was resuspended in 100 μl of extraction buffer and represents the P13 fraction, and the supernatant was transferred to ultra-centrifuge tubes and centrifuged at 100,000g for 30 minutes at 4^oC. The supernatant was transferred to a new tube and represents the SUP fraction, and the pellet was resuspended in 100 μl of extraction buffer and represents the P100 fraction.

Protein fractions were dissolved in SDS loading buffer and analyzed by immunoblotting using the indicated antibodies (Zouhar et al., 2009).

Immunoprecipitation was done as previously described (Bassham et al., 2000) using antibodies also previously described (Zouhar et al., 2009). 5 grams of 4 - to 6 - week-old Arabidopsis leaves were ground in 15 ml cold extraction buffer (0.3 M Sucrose, 0.1 M Tris-HCl, 1 mM EDTA, pH 7.5) with protease inhibitor cocktail (Roche, 11836153001). The crude extract was passed through Miracloth to remove debris, followed by centrifugation at 1000g for 5 min at 4°C. To dissolve membrane proteins, 0.5% Triton X-100 (v/v) was added to the supernatant, followed by rocking at 4°C for 2-3 hours. The protein extract was then transferred to ultra-centrifuge tubes followed by centrifugation at 100,000g to pellet the insoluble material. The supernatant was transferred to new 15 mL tubes and anti-SYP41 (1:200) antibodies were added to the samples followed by 2 hours rocking at 4°C. Protein A Sepharose CL-4B (Sigma-Aldrich, GE17-0780-01) was prepared according to the manufacturer’s protocol. The samples were further rocked with 50 μl suspended prepared protein A Sepharose overnight at 4°C. Beads were collected by centrifugation at 200g for 5 minutes at 4°C and washed 3 times with PBS buffer with 0.1% (v/v) Triton X-100. Proteins were eluted in SDS loading buffer (62.5 mM Tris-HCl (pH 6.8), 2% (w/v) sodium dodecyl sulfate, 25% (v/v) glycerol, and 0.01% bromophenol blue). Eluted proteins were analyzed by immunoblotting using the indicated antibodies.

Fluorescein diacetate (FDA) staining was performed as previously described (Saedler et al., 2009). Seedlings were submerged in a solution of 40 μg FDA in water for 5 min, and then mounted on a slide. Confocal microscopic images of root hairs were obtained using a Leica confocal microscope (Leica Microsystems, Wetzlar, Germany) using 63× oil immersion objective lens after excitation of the dye at 488 nm and emission was detected between 520 and 560 nm.

MDY-64 staining was performed as described (Scheuring et al., 2015). Seedlings were submerged in a solution of 0.25 μM MDY-64 (Invitrogen, Y7536) in 0.5X liquid MS medium for 5 min. The seedlings were then rinsed in 0.5X liquid MS medium and mounted on a slide. Confocal images of root hairs were obtained using a 63× oil immersion objective lens after excitation of the dye at 451 nm using an Ar/Kr laser, and emission was detected at 497 nm.

Five-day-old seedlings were transferred to a slide and imaged using a Zeiss AxioImager microscope (Carl Zeiss Inc., Jena, Germany) with a 40X objective with differential interference contrast and confocal microscope using an EYFP-specific filter.

Arabidopsis VPS45 is essential for plant growth and development, as homozygous null mutants are inviable (Zouhar et al., 2009). As an alternative approach to determine the physiological roles of VPS45, a point mutation in VPS45 was recovered by a TILLING approach (Colbert et al., 2001) and designated as vps45-3. The mutation is a C-to-T substitution at the 851^st nucleotide of the VPS45 coding sequence, leading to a serine-to-phenylalanine substitution at the 284^th amino acid position ( Figure 1A ). This mutation introduces a new MseI restriction site in the vps45-3 coding sequence, allowing differentiation between mutant and wild-type alleles ( Figure 1B ).

vps45-3 plants were dwarfed, with highly reduced sizes of many organs ( Figures 1C–E , S1 ). similar to the previously reported VPS45 RNAi lines (Zouhar et al., 2009), although less severe. To confirm that the observed phenotype results from mutation of VPS45, we introduced the VPS45 cDNA driven by the native VPS45 promoter into the vps45-3 mutant to generate complementation lines (vps45-3 COM) and assessed the plant phenotype. Growth and organ size defects of vps45-3 plants were ameliorated by complementation with the VPS45 transgene ( Figures 1C-E , S1 ), confirming that the defects were caused by mutation of the VPS45 gene and suggesting that the VPS45-3 protein has reduced function. The vps45-3 plants were fertile and produced flowers and viable seeds without any noticeable abnormalities. This confirms that VPS45 is important for plant growth and that the vps45-3 mutant is valuable to further analyze the function of VPS45.

We reasoned that the dwarf phenotype of vps45-3 plants could be due to reduced cell size, similar to RNAi plants. To analyze this possibility, 4-day-old seedling roots of WT, vps45-3 and vps45-3 COM were stained with propidium iodide (PI), which stains cell walls (Scheuring et al., 2015), and imaged by confocal microscopy. Cell size was significantly reduced in vps45-3 compared to WT and vps45-3 COM roots ( Figures 2A–D ). This was further visualized by using an agar imprinting method (Mathur and Koncz, 1997) to analyze the hypocotyl and root cells ( Figure S2 ). Thus, a single ser-to-phe change in VPS45 causes cell expansion defects.

We observed that vps45-3 had a shorter main root compared to WT and had root hair defects ( Figures 3A-C ). Mutant root hairs were significantly shorter, wider and wavy compared with the root hairs of WT ( Figures 3D, E ). Complementation of the mutant rescued the root hair elongation defects ( Figures 3F, G ) confirming that the point mutation causes the observed root hair defects and that VPS45 is required for root hair cell expansion.

Root hairs and pollen tubes undergo tip growth, in contrast to other cells which undergo diffuse growth (Mathur and Hülskamp, 2001). Tip growth involves development of apical-basal polarity of the endomembrane system and rapid secretion at the tip of the developing root hairs and pollen tubes (Cole and Fowler, 2006; Rounds and Bezanilla, 2013; Šamaj et al., 2006). To determine if vps45-3 has general defects in tip growth, we analyzed the growth of pollen tubes in vitro. Pollen from WT and vps45-3 plants was plated onto pollen germination medium (Li et al., 1999) and incubated overnight to allow germination ( Figures 3H-J ). While no differences could be seen in the extent of germination between pollen from WT and mutant plants ( Figure 3K ), pollen tubes were significantly shorter in vps45-3 compared to WT, and this defect was rescued in vps45-3 COM lines ( Figure 3L ). These data suggest that VPS45 may play a role in tip growth in Arabidopsis.

Our results indicate that vps45-3 plants are dwarf with significant reduction in organ sizes, have defects in root hairs and pollen tubes, and show cell expansion defects. This suggests that the substitution of the serine to a phenylalanine, i.e. a polar to non-polar substitution, affects VPS45 function. Unlike a previously described VPS45 point mutant (ben2) (Tanaka et al., 2013), amino acid sequence alignment showed that the substituted amino acid in vps45-3 is not conserved across different organisms ( Figure S3 ), and that this amino acid may be important for VPS45 function in plants only.

VPS45 has been implicated in regulating the stability and localization of its cognate SNARE complex (VTI12/SYP41/SYP61), as the levels of SYP41 were reduced in parallel to the levels of VPS45 in RNAi-silenced lines (Zouhar et al., 2009). To test the stability of both the SM protein and the SNARE complex we carried out subcellular fractionation of organelles from WT and vps45-3 seedlings, followed by immunoblotting of different fractions with VPS45, SYP41 and SYP61 antibodies. The amount of SYP41, SYP61 and VPS45 was similar in WT and vps45-3 in all fractions ( Figure 4A ), suggesting that both VPS45 and the SNARE proteins are stable in the mutant and that VPS45-3 can still associate with membranes.

To further understand the potential effect of the point mutation, we used homology modeling to fit the predicted Arabidopsis VPS45-3 protein sequence onto the crystal structure of c6MX1 from Chaetomium thermophilum, with a confidence score of 100.0 among all available protein structures. The mutation is in domain 3a of the protein ( Figure 4B ) and in close proximity to a region that is important for VPS45 interaction with other proteins (Eisemann et al., 2020).

Since binding of a SM protein to its cognate SNARE is required for SNARE complex function (Furgason et al., 2009; Shanks et al., 2012), we assessed whether defects in the mutant might be caused by altered interaction of VPS45-3 with SYP41 and SYP61. To test the interaction of VPS45-3 with SYP41 and SYP61, SYP41 was immunoprecipitated from WT or vps45-3 plants using anti-SYP41 antibodies, and co-immunoprecipitation of VPS45 and SYP61 was assessed by immunoblotting ( Figure 4C ). The amount of VPS45 and SYP61 that co-precipitated with SYP41 was equivalent in WT and vps45-3 mutant. This implies that the point mutation does not affect the interaction of VPS45-3 with SYP41 and SYP61 ( Figure 4C ).

Given that other vps45 mutants have endocytic defects (Tanaka et al., 2013), we assessed whether endocytosis and recycling are affected in the vps45-3 plants. This could explain the cell expansion phenotype owing to slower trafficking at the TGN during cell expansion (Gendre et al., 2015), particularly during root hair growth. We stained root cells with a FM4-64, a lipophilic styryl dye that is used as an endocytic tracer (Dettmer et al., 2006). The uptake of the FM4-64 was similar in both vps45-3 and WT cells ( Figure 5A ), suggesting no major effect of the vps45-3 mutation on internalization from the plasma membrane. We also tested whether membrane arrival at the TGN was affected in the mutants. We stimulated formation of TGN-endosomal aggregates by treating roots of 4-day old seedlings with Brefeldin A (BFA), a vesicle trafficking inhibitor, followed by incubation with FM4-64 and imaging with confocal microscopy. Labeling of the BFA compartments with the dye occurred at similar times in mutant and WT, suggesting that BFA body formation and arrival of membrane cargo at the TGN is not affected in the mutant ( Figure 5B ).

RabF2a is a Rab GTPase that localizes to early endocytic compartments in plants (Ueda et al., 2001; Preuss et al., 2004). We examined the distribution of RabF2a in root hairs of both WT and vps45-3. We transformed plants with an EYFP-RabF2a construct and observed three independently transformed lines by confocal microscopy. A similar distribution of EYFP-RabF2a was evident in both WT and vps45-3 root hairs, with small punctate structures spread along the length of the root hair as previously reported (Preuss et al., 2004) ( Figures 5D–F ). The organization of the endosomal system therefore appears to be intact in vps45-3. Taken together, these data suggest that endocytosis and membrane trafficking from the PM to the TGN are unaffected in vps45-3 mutants in both tip-growing and diffusely-growing cells.

The root hair and pollen tube phenotypes suggest that the vps45-3 mutation might cause polarized tip growth defects. Polarized tip growth involves targeted deposition of cell wall and membrane material at the cell apex, and turgor pressure is a driving force for cell expansion via uptake of water into the vacuole (Cosgrove, 1993; Mendrinna and Persson, 2015). We reasoned that the root hair abnormality observed in vps45-3 could be due to vacuole defects that disrupt tip growth. To test this, we analyzed root hair vacuoles by staining root hairs of five day old seedlings with the tonoplast marker MDY-64 (Scheuring et al., 2015), and imaged them using confocal microscopy.

In elongating WT and vps45-3 COM root hairs, the vacuole was seen to occupy most of the cell, whereas in vps45-3, the root hair cell was filled with cytoplasm, with numerous small vacuoles visible ( Figure 6 ; Movies M1 , M2 ). Staining root hairs with fluorescein diacetate (FDA), which labels the cytoplasm, leaving the unstained vacuole visible, supported the idea that vps45-3 roots hairs have defects in vacuole morphology, as they appeared to have increased cytoplasmic staining ( Figure S4 ). The changes in vacuolar morphology in root hairs therefore correlate with the cell expansion defects observed in vps45-3.