Newly Formed Vacuoles in Root Meristems of Barley and Pea Seedlings Have Characteristics of both Protein Storage and Lytic Vacuoles

Plant cells are considered to possess functionally different types of vacuole in the same cell. One of the papers cited in support of this concept is that of Paris et al. (2001 Cell 85: 563-572) who reported that protein storage and lytic vacuoles in root tips of barley and pea seedlings were initially separate compartments which later fused to form a central vacuole during cell elongation. We have reinvestigated the situation in these two roots using immunogold electron microscopy as well as immunofluorescence microscopy of histological sections. Using antisera generated against the whole protein of α− TIP as well as specific C-terminal TIP peptide antisera against α -, γ−, and δ -TIP, together with antisera against the storage proteins barley lectin, and pea legumin and vicilin, we were unable to obtain evidence for separate vacuole populations. Instead, our observations point to the formation of a single type of vacuole in cells differentiating both proximally and distally from the root meristem. This is a hybrid type vacuole containing storage proteins and having both α - and γ - TIPs, but not δ -TIP, in its tonoplast. As cells differentiate towards the zone of elongation their vacuoles are characterized by increasing amounts of γ− TIP, and decreasing amounts of α− TIP.


INTRODUCTION
Most differentiated plant cells are dominated by the vacuole, an organelle that by storing sugars, inorganic ions, organic acids and secondary metabolites serves to generate and regulate cell turgor (De, 2000). The various functions of the vacuole are highly tissue-specific (Marty, 1999;Tomos et al., 2000), whereby a distinction is generally made between lytic vacuoles (LV) which harbor hydrolytic enzymes, and protein storage vacuoles (PSV) in which nonenzymic proteins accumulate. PSV are, however, compound organelles since they contain specific hydrolytic enzymes required for storage protein processing (Müntz and Shutov, 2002), as well as crystalline inclusions (Jiang 4 seems to vary from cell type to cell type, especially in leaves (Flückiger et al., 2003). These differences may well relate to vacuolar pH with some vacuoles being neutral and others acidic (Swanson et al., 1998;Diwu et al., 1999;Epimasko et al., 2004).
The most frequently cited paper in support of what is sometimes termed the "two-vacuole hypothesis" is a paper from Paris et al. (1996), which deals with the identification of different vacuole populations in root tips. Since roots have meristems whose cells have no or only small vacuoles, followed distally by a zone of elongation which is functionally correlated with vacuolar expansion, this is a classical organ for studies on vacuole biogenesis and vacuole development (Marty, 1999;De 2000). Until publication of the paper of Paris et al. (1996), vacuoles in the primary root -being vegetative tissuewere generally considered to be of the lytic (LV) type. Using α and γ IP antisera as membrane probes, and antibodies against barley lectin and aleurain as content markers, Paris et al. (1996) demonstrated the presence of PSV in addition to LV in the roots of barley and pea. According to these authors, the two vacuole types were separate compartments in the immediate post-meristematic cells and later fused with one another to form the large central vacuole as the cells differentiate upwards in the root.
We have reinvestigated the distribution of TIPs and content markers in the roots of barley and pea using immunofluorescently labeled sections from Steadman wax embedded samples, and immunogold electron microscopy. In contrast to Paris et al. (1996), whose observations were made on squash preparations where cell position in the root cannot be accurately determined, our results do not support the existence of two initially separate vacuole populations. Instead, post-meristematic cells both in the calyptra and upwards in the root form only one type of vacuole: a PSV. As differentiation proceeds, α TIP is gradually replaced by γ TIP in the membrane of the enlarging vacuole.
The production of a PSV rather than a LV as the primary vacuole type in postmeristematic cells is most likely related to seed-specific signals still being dominant in the meristem of the primary root.

Detection of Storage Proteins and TIPs in Barley and Pea Roots
We isolated total membrane fractions from 1 mm segments sequentially excised from barley and pea root tips, separated the proteins by SDS-PAGE and carried out Western blotting with TIP antisera and antibodies against barley lectin (for barley roots) and vicilin and legumin (for pea roots). In barley roots, both α− and γ-TIPs were found to be present in all root tip segments ( Fig. 1 A, B). For the detection of α-TIP, polyclonal antibodies directed against the whole protein (Johnson et al., 1989) as well as the C-terminal peptide (Jauh et al., 1998(Jauh et al., , 1999 were used. Similarly for γ−TIP a peptide antibody (Fig. 1 B;Jauh et al., 1998), and a whole protein antibody (VM 23, Maeshima, 1992; data not shown) were used, and give identical results. However, δ-TIP which, together with α-TIP, is typical of the boundary membrane of protein bodies from pea cotyledons ( Fig. 1 E), was not detected by the δ-TIP peptide antiserum of Jauh et al. (1998) in barley root tips. Barley lectin was also found to be present in all root tip segments ( Fig. 1 C). αand γ-TIPs were detected in all segments of the pea root tip ( Fig. 1 E), as were vicilin and legumin -the principal storage globulins of pea seeds. Interestingly, these globulins were not proteolytically processed into mature storage proteins, as they are in developing cotyledons. Thus, only prolegumin polypeptides with molecular mass around 60 kDa (Hinz et al., 1999) were detected in root tips of pea.

Cellular Distribution of Barley Lectin and TIPs in Barley Roots
Barley lectin is not uniformly expressed throughout the root tips of germinating barley, neither in 3 day nor 10 day old roots ( whereas the immunostaining pattern for α-TIP was similar to that for barley 6 lectin in the non-calyptra portion of the root tip, only the cells surrounding the columella of the calyptra stained positively for α-TIP ( Fig. 2 E-H). In root tips from older seedlings the α−TIP signal is particularly strong in the cells of the meristem (Suppl. Fig.1 B). In addition, the cells of the central cylinder are also labeled, but more weakly. The distribution of γ-TIP in 3 day old roots was remarkably similar to that of barley lectin: all cells of the calyptra, and only the rhizodermis and outermost cortex cells reacting positively ( Fig. 3 A -C). In comparison, in 10 day old roots, cells of the central cylinder gave a weak signal for γ-TIP, but there was almost no signal in the rhizodermis (Suppl. Fig.   1 C).

Cellular Distribution of Vicilin and TIPs in Pea Roots
In root tips of 3 day old pea seedlings, labeling with vicilin antibodies was seen in the meristem, cortex and extreme tip of the calyptra, but neither in the central cylinder nor the rest of the calyptra (Fig. 4 B). In 10 day old root tips, vicilin and α−TIP were no longer detected (Suppl. Fig. 1 D, E). In pea root tips, γ−TIP labeling was observed in the outermost cells of the root cap, and the innermost layers of the cortex above the meristem, while the meristem itself, the majority of the calyptra and the rhizodermis were without signal ( Fig.   5 A -C). In 10 day old root tips, the γ-TIP signal resembled that of 3 day old root tips, with the strongest signal being in the innermost cells of the cortex (Suppl. Fig. 1 F).

Evidence for Separate Vacuole Populations
We have performed IEM on barley root tips with antibodies generated against the storage protein barley lectin as a content marker, and with antibodies against αand γ-TIP. We selected cells in those areas which stained immunofluorescence, also contained vicilin-positive storage protein aggregates ( Fig. 8 A). These vacuoles also labeled positively with both α− and γ-TIP antibodies ( Fig. 8 B). In double-immunogold labeled sections we could not find vacuoles which were labeled exclusively by only one TIP antibody.

DISCUSSION
The classical notion of the vacuole as being a multifunctional single organelle in the plant cell (Wink, 1993), has gradually been superseded over the last decade by a multi-compartment concept with different vacuole types coexisting in the same cell (Robinson and Rogers, 2000 , 2007;Sanmartin et al., 2007). Especially important for the development of the multivacuolar concept has been the discovery of vacuole-specific isoforms of tonoplast intrinsic proteins (TIPs, Johnson et al., 1989;Höfte et al., 1992;Ludevid et al., 1992). These allow lytic-type vacuoles to be distinguished from vacuoles in which proteins are stored, especially in seeds.
Actually, the TIP isoforms are very similar and differ only in their C-terminal cytoplasmic tails, against which specific peptide antibodies have been generated (Jauh et al., 1998;1999). In immunofluorescence studies, together A key paper in support of the multivacuole concept is that of Paris et al.

Plant materials
Barley (

Antibodies
The antisera and their dilutions used for (Western blotting) and

Immunofluorescence labeling
Sections on glass-slides were incubated at RT with 100 µl each of the following: Single antibody labeling HM20-embedded sections on grids were blocked with 3% (w/v) BSA in PBS (

Double antibody labeling
Since all the primary antisera were generated in rabbits, a double labeling procedure was performed by first gently placing an HM20-embedded section, directly after cutting, onto a drop of blocking solution. Most of the drop was removed afterwards with a fine syringe, and the next solution added by carefully under-layering it, still allowing the section to float. After the incubation and washing regime (described below), the section was gently transferred to a Formvar-coated grid by sliding the grid under the section so that the upper (non-labeled) side of the section was now accessible for a new round of labelling with another primary antibody (also as below).
Labeling was done using similar solutions as above: Blocking in block buffer for 1 x 2 min, then for 1 x 30 min, primary antibody incubation for 1h, washing in wash buffer 1 x 2 min, then for 3 x 10 min, secondary antibody incubation for 1 h, washing in wash buffer for 3 x 5 min and in H 2 O for 3 x 5 min. The sections were post-stained in 3% (w/v) uranyl-acetate in H 2 O for 5 min, washed in H 2 O for 3 x 1 min, then incubated in 0.3% (w/v) lead-citrate in 1 M NaOH for 5 min, washed in H 2 O for 4 x 1 min. After drying, samples were examined in a Philips CM10 transmission electron microscope.

Cryosectioning
Cryosections were prepared, and labeled with primary and secondary antibodies exactly as described in Pimpl et al. (2000).

Protein extraction, SDS-PAGE and Western blotting
Total membrane protein fractions from barley and pea roots.
The homogenate was passed through one layer of Miracloth and centrifuged at 1000 xg in a Sorvall HB 4 rotor for 10 min. The pellet was discarded and the supernatant was centrifuged at 12.000g for 20 min in the same rotor. The pellet was saved and the supernatant centrifuged at 100.000g for 1h in a Sorvall TFT 50.38 rotor. The supernatant was discarded and the pellets were resuspended in homogenisation buffer using a glass homogenizer. Protein concentrations were measured according to Lowry (Lowry et al., 1951) and protein samples for SDS-PAGE were precipitated with methanol/chloroform (Wessel and Flügge, 1984).

Isolation of protein bodies from pea cotyledons.
Pea seeds were harvested 4 weeks after flowering, the testa was removed and the seeds weighed. The isolation procedure was according to Hohl et al. (1996) andHinz et al. (1999) in a Sorvall HB4 rotor for 10 min. The pellet (material collected from the top of the cushion) was resuspended in homogenisation buffer and re-centrifuged, this step was then repeated once more. The pellet was saved; protein concentration determination and sample preparation for SDS-PAGE was done as described above.

Isolation of integral membrane proteins from barley leaves.
Leaves  in the cells of the calyptra, the rhizodermis and the outermost cells of the cortex.
Inset: high magnification of a single calyptra cell. The lectin signal is more intense at the rims of the vacuoles, as seen in EM sections (Fig. 6). E -H) α-TIP has a similar distribution to barley lectin, except that it is absent from the innermost cells of the calyptra. E, H) α−TIP total protein antiserum (Johnson et al., 1989) G) α-TIP peptide antiserum (Jauh et al., 1998). Arrowhead points to the meristem. Bars = 100 µm