Interplay between Ions, the Cytoskeleton, and Cell Wall Properties during Tip Growth^,

Abstract

Tip growth is a focused and tightly regulated apical explosion that depends on the interconnected activities of ions, the cytoskeleton, and the cell wall.

Tip growth is a specialized type of growth utilized by only a few cell types in the plant kingdom. Nevertheless, these cell types perform vital functions for plants. Pollen tubes are necessary for seed plant reproduction, root hairs are vital for nutrient acquisition, and protonemata comprise the entire juvenile moss plant. Tip growth requires exquisite coordination of intracellular cargo delivery and simultaneous weakening of the apical cell wall. This process is precarious, with imbalances resulting in cell rupture or cessation of growth. In this Update, we provide context and highlight recent advances in our understanding of how ions, the cytoskeleton, and properties of the cell wall are interconnected during tip growth across three tip-growing systems: pollen tubes, root hairs, and protonemata. We propose that further mechanistic insights are likely to come from studying the commonalities among these systems, particularly focusing on the molecular basis of the communication between cell wall status and intracellular cargo delivery.

Plant cells are surrounded by a cell wall that imposes both physical restrictions to shape as well as a balance to turgor pressure. However, most undifferentiated plant cells are small and uniform. During differentiation, plant cells expand, which occurs when loosening events allow the wall to yield to the existing turgor pressure, discussed further in this issue (Cosgrove, 2018). The momentary decrease in water potential results in an influx of water, causing expansion. As new cell wall material is deposited to account for this expansion, the increase in size becomes irreversible, and thus the cell has grown. Modifications to cell wall viscosity promoting expansion can result from a number of different processes. For example, cell wall-loosening enzymes may be activated in or delivered to specific regions. Alternatively, the secretion of flexible cell wall material could be spatially regulated.

One extreme form of cell wall patterning is known as tip growth. In tip-growing cells, the zone of reduced cell wall viscosity is focused at the apex of the growing cell. To grow, the cell must balance the delivery of wall material with the retrieval of excess membrane, as the delivery of excess flexible wall material at the cell tip could result in lysis. Tip growth occurs across plant lineages in cell types such as pollen tubes, root hairs, moss protonemata, and bryophyte and algal rhizoids. Although tip growth mechanisms have been generalized as equivalent in pollen tubes, root hairs, and moss protonemata, it is likely that significant differences exist in order for these cells to carry out their divergent functions.

Of the three cell types, pollen tubes are the most specialized and short-lived cell, in that their aim is to deliver the sperm cells to the ovule. By transmitting location information about the ovule from outside the pollen tube to the intracellular growth machinery, pollen tubes grow rapidly toward the female gametophyte and burst upon entry. Root hairs are the site of nutrient and water uptake for the plant. In contrast to pollen tubes, root hairs are longer-lived and must integrate many environmental cues (moisture, nutrient content, soil texture, etc.). Root hairs internalize this suite of information to maximize the efficacy of their growth. In contrast to pollen tubes and root hairs, moss protonemata have a colonization role. They germinate from the spore and race against other plants to acquire moisture, nutrients, and an optimal position for photosynthesis. This tip-growing cell continues to undergo cell divisions as the plant develops, reaching a point in which the entire juvenile plant, persisting on the scale of months, consists entirely of protonemata. Thus, moss protonemata are longer-lived than both pollen tubes and root hairs and integrate substantially more environmental information during their growth and development. Even with such divergent functions, mechanistically describing tip growth in each system promises to illuminate fundamental processes conserved throughout plant evolution as well as derived processes that may have evolved in each cell type.

Ultimately, the key to understanding tip growth will require understanding a complex interaction network; the reader is directed to recent reviews that expand upon a variety of facets of tip growth not covered here (Mendrinna and Persson, 2015; Hepler, 2016; Damineli et al., 2017; Michard et al., 2017; Stephan, 2017). In this Update, we speculate that further mechanistic advances will come from identifying deeply conserved tip growth mechanisms that decipher signaling events between the cell wall and the cytoplasm. Thus, we will focus on advances in the understanding of cytoskeletal organization, intracellular ion gradients, and the cell wall itself, all key aspects involved in tip growth signaling. A link between wall status and cytoplasmic activity is critical because, in many ways, tip growth resembles a controlled explosion: the cell wall must weaken exactly enough to allow for expansion but not so much that the cell ruptures. Indeed, there are many conditions under which this balance is lost, resulting in either growth arrest or tip bursting. We have compiled a list of agents and mutants that uncouple various aspects of tip growth with similar results (Table I). While not an exhaustive list, Table I demonstrates the wide-ranging conditions that disrupt the precarious balancing act between intracellular and extracellular events during tip growth. By identifying the molecular targets of agents and cellular deficiencies in mutants leading to growth arrest or tip blowout, it may be possible to understand the molecular regulation of this controlled explosion (see Outstanding Questions).

Table I. Agents and mutants abrogating tip growth.

RNAi, RNA interference; N/A, not applicable.

Treatment/Mutant	Cellular Consequence	Gene Name	Growth Consequence	System	Reference
Latrunculin B	Actin depolymerization	N/A	Swelling, loss of polarized growth	Pollen tubes, root hairs, protonemata	Gibbon et al. (1999); Ketelaar et al. (2003);Harries et al. (2005)
Calcium ionophore	Equalizes calcium concentrations	N/A	Apical wall thickening, growth reorientation	Root hairs, pollen tubes	Malho and Trewavas (1996); Monshausen et al. (2008)
Oryzalin	Microtubule depolymerization	N/A	Loss of directional growth	Root hairs, moss protonemata	Doonan et al. (1988); Bibikova et al. (1999)
Lanthanum	Calcium channel and pump blocker	N/A	Bursting	Pollen tubes, root hairs	Monshausen et al. (2008)
Cyanide	ATP synthesis inhibition	N/A	Apical wall thickening	Pollen tubes	Winship et al. (2016)
Propidium iodide	Competes with calcium in cell wall	N/A	Bursting	Pollen tubes	Rounds et al. (2011)
EGTA/EDTA	Low extracellular calcium	N/A	Bursting	Pollen tubes, root hairs	Monshausen et al. (2008)
Acidic growth medium	Low extracellular pH	N/A	Bursting	Root hairs	Monshausen et al. (2008)
Yariv agent	Arabinogalactan protein blocking	N/A	Expansion stops, deposition continues	Pollen tubes, moss protonemata	Roy et al. (1999); Lee et al. (2005)
apg knockdown/knockout	Aberrant extracellular signaling	ARABINOGALACTAN	Shorter cells	Pollen tubes, moss protonemata	Lee et al. (2005); Levitin et al. (2008)
cngc mutants	Aberrant calcium signaling	CYCLIC NUCELOTIDE-GATED CHANNELS	Shorter, deformed cells	Pollen tubes, root hairs	Gao et al. (2016);Zhang et al. (2017)
glr1.2 mutant	Aberrant calcium signaling	GLU RECEPTOR-LIKE	Deformed cells	Pollen tubes	Michard et al. (2011)
exo70C mutant	Faster growth rate, thin cell wall	EXOCYTOSIS 70C2	Bursting	Pollen tubes	Synek et al. (2017)
cog3 and cog8 single mutants	Aberrant Golgi morphology, vesicle trafficking	CONSERVED OLIGOMERIC GOLGI 3 and 8	Bursting	Pollen tubes	Tan et al. (2016)
anx1,anx2 double mutant	Aberrant cell wall formation	ANXUR 1 and 2	Bursting	Pollen tubes	Boisson-Dernier et al. (2009)
LePRK RNAi/overexpression	Aberrant extracellular signaling	POLLEN RECEPTOR KINASE	Knockdown: shorter pollen tubes, bursting; overexpression: ballooning of tip	Pollen tubes	Gui et al. (2014)
lrx1,lrx2 double mutant	Aberrant extracellular signaling	LUCINE-RICH REPEAT EXTENSIN 1 and 2	Rupture after initiation	Root hairs	Baumberger et al. (2003)
prx44 mutant	Aberrant extracellular signaling	PEROXIDASE44	Bursting	Root hairs	Kwon et al. (2015)
kjk/csld3 mutant	Aberrant cell wall formation	KOJAK/cellulose synthase-like D3	Rupture after initiation	Root hairs	Favery et al. (2001)
bup mutant	Loss of germination plaque	BURSTING POLLEN (Golgi-localized glycosyltransferase)	Bursts upon germination	Pollen tubes	Hoedemaekers et al. (2015)
ext18 mutant	Aberrant cell wall formation	EXTENSIN18	Bursting	Pollen tubes	Choudhary et al. (2015)
reb1 mutant	Altered sugar metabolism	ROOT EPIDERMAL BULGER1	Ballooning	Root hairs	Andème-Onzighi et al. (2002)
ROP RNAi	Aberrant intracellular signaling	Rho/Rac of Plants (small GTPase)	Loss of polarized growth, loss of cell adhesion	Moss protonemata	Burkart et al. (2015)
ROP overexpression	Aberrant intracellular signaling	Rho/Rac of Plants	Swelling, branching	Pollen tubes, root hairs	Wu et al. (2001); Jones et al. (2002)

THE CYTOSKELETON

Within the cell, both the actin and microtubule cytoskeletons contribute to tip growth. As we discuss the structure of these cytoskeletons in tip-growing cells, it is important to recall that each array is decorated with, and regulated by, microtubule- and actin-binding proteins. Thus, their localization (Table II) during tip growth provides insights into the mechanism of cytoskeletal organization and dynamics.

Table II. Localization of cytoskeleton-binding proteins.

MTs, Microtubules; PM, plasma membrane.

Protein	Localization	Species/System	Reference
Fimbrin	1) Apical and subapical F-actin structures	1) Lily pollen tubes	1) Su et al. (2012), α-LlFIM1 antibody
	2) Apical filaments	2) Arabidopsis pollen tubes	2) Zhang et al. (2016), pFIM5::FIM5-GFP
Villin	1) AtVLN3, thick subapical filaments	1) Arabidopsis root hairs	1) van der Honing et al. (2012), pAtVLN3::VLN3-GFP
	2) AtVLN2, thick subapical filaments, fine apical filaments	2 and 3) Arabidopsis pollen tubes	2) Qu et al. (2013), pAtVLN2::VLN2genomic-GFP
	3) AtVLN5, thick subapical filaments, fine apical filaments		3) Qu et al. (2013), pAtVLN5::VLN5genomic-GFP
Actin Depolymerizing Factor (ADF)/Cofilin	1) Apical and subapical actin structures	1) Tobacco pollen tubes and grains	1) Chen et al. (2002), Lat52::NtADF-GFP
	2) Apical structures	2) Lily pollen tubes	2) Lovy-Wheeler et al. (2006), α-lily ADF antibody
	3) Cytoplasmic	3) P. patens protonemata	3) Augustine et al. (2008), α-PpADF polyclonal antibody
Profilin	Cytoplasmic	Lily pollen grains and tubes	Vidali and Hepler (1997), α-profilin polyclonal antibody
Actin Interacting Protein1	1) Apical F-actin fringe	1) Lilium spp. pollen tubes	1) Lovy-Wheeler et al. (2006), α-lily AIP antibody
	2) Cytoplasmic	2) P. patens protonemata	2) Augustine et al. (2011), AIP-mCherry locus integration
Class I Formin	LlFH1, apical vesicles, PM	Lily pollen tubes	Li et al. (2017), transient overexpression of LlFH1-GFP
Class II Formin	PpFor2a, apex, not on PM	P. patens protonemata	Vidali et al. (2009b), For2a 3xGFP locus integration
Class VIII Myosin	PpMyo8A, cortical puncta with apical accumulation, spindle and phragmoplast midzone	P. patens protonemata	Wu and Bezanilla (2014), pMaizeUbiquitin::Myo8A-3xGFP, in Δmyo8 a,b,c,d,e
Class XI Myosin	1) PpMyo11a cytosolic with apical enrichment	1) P. patens protonemata	1) Vidali et al. (2010), pMaizeUbiquitin::3xGFP-Myo11a
	2) AtMyo11k, apical cloud	2) Arabidopsis root hairs	2) Park and Nebenführ (2013), pMyo11K::YFP-Myo11K
Kinesin	1) KIND1a,b, plus ends of MTs focused on the tip	1) P. patens protonemata	1) Hiwatashi et al. (2014), KIND1a/1b-cerulean locus integration
	2) ARK1, cytoplasmic, plus ends of MTs	2) Arabidopsis root hairs	2) Eng and Wasteneys (2014), pARK1::ARK1-GFP
Actin Related Protein2/3 (ARP2/3) complex	PpARPC4, apically enriched cap	P. patens protonemata	Perroud and Quatrano (2006), 2xYFP locus integration
Brick1 (brk1)	PpBRK1, apically enriched cap	P. patens protonemata	Perroud and Quatrano (2008), 3xYFP locus integration
Rho/Rac of Plants (ROP)	1) AtROP2, apical dome	1) Tobacco pollen tubes	1) Kost et al. (1999), LAT52::GFP-AtROP2
	2) ROP1, apical dome	2) Arabidopsis root hairs	2) Jones et al. (2002), 35S::GFP-ROP2
	3) AtROP1-PM, apical enrichment, cytoplasm	3) Tobacco pollen tubes	3) Hwang et al. (2005), LAT52::RIC4ΔC-GFP
	4) PpROP2-PM, enrichment at cross walls	4 and 5) P. patens protonemata	4) Ito et al. (2014), HSP::cerulean-PpROP2
	5) PpROP4, apical PM and cross walls	5) P. patens protonemata	5) Burkart et al. (2015), pROP4::GFP-ROP4coding sequence, locus replacement
ROP Activating Proteins (RopGAPs)	NtRopGAP1, subapical PM	Tobacco pollen tubes	Klahre and Kost (2006), Lat52::YFP-NtRopGAP1
ROP Interacting CRIB-Containing Protein (RIC)	1) AtRIC 3,10, cytosolic	1 to 4) Tobacco pollen tubes	1) Wu et al. (2001), Lat52::GFP-RIC n
	2) AtRIC 1, 6, 7, apical PM, cytosolic		2) Wu et al. (2001), Lat52::GFP-RIC n
	3) AtRIC9, PM		3) Wu et al. (2001), Lat52::GFP-RIC 9
	4) AtRIC 2, 4, 5, apical PM		4) Wu et al. (2001), Lat52::GFP-RIC n
ROP Guanine Dissociation Inhibitors (RopGDIs)	NtRopGDI2, cytosolic	Tobacco pollen tubes	Klahre et al. (2006), Lat52::YFP-NtRopGDI2
ROP Guanine Exchange Factors (RopGEFs)	1) AtRopGEF1, apical PM	1 to 3) Tobacco pollen tubes	1) Gu et al. (2006), Lat52::GFP-RopGEF1
	2) AtRopGEF 8, 9, 14, apical PM, cytoplasm		2) Gu et al. (2006), Lat52::GFP-RopGEF n
	3) AtRopGEF12, cytoplasm		3) Gu et al. (2006), Lat52::GFP-RopGEF12
	4) PpRopGEF3, cytoplasmic apical enrichment	4) P. patens protonemata	4) Ito et al. (2014), HSP::PpRopGEF3-cerulean
Microtubule-Associated Protein18	1) Shank PM, inverted cone at tip	1) Arabidopsis pollen tubes	1) Zhu et al. (2013), pAtMAP18::AtMAP18-GFP
	2) Shank of PM, apical cytoplasm	2) Arabidopsis root hairs	2) Kang et al. (2017), pAtMAP18::AtMAP18-mCherry
End-Binding1	1) Cortical puncta, apical accumulation	1) P. patens protonemata	1) Hiwatashi et al. (2014), PpEB1b-mCitrine locus integration
	2) Cytoplasmic puncta	2) Arabidopsis root hairs	2) Eng and Wasteneys (2014), 35S::AtEB1b-GFP

Actin Cytoskeleton

Of the two cytoskeletal elements in plants, the actin cytoskeleton, which is discussed further in this issue (Szymanski and Staiger, 2018), is essential for tip growth in angiosperm pollen tubes, root hairs, and moss protonemata. To understand how actin contributes to cell expansion, it is important to determine the subcellular architecture of the F-actin arrays. However, across these systems, visualizing F-actin has been challenging. For example, in fixed cells, fluorescent phalloidin, a widely used probe for F-actin in many cell types, does not readily stain plant cells (discussed by Olyslaegers and Verbelen, 1998). Furthermore, phalloidin only interacts with a subset of F-actin arrays (Nishida et al., 1987) and stabilizes F-actin (Cooper, 1987). Thus, phalloidin often leads to an increase in bundled F-actin (Vidali et al., 2009a). Additionally, to date, GFP-actin fusions have not been successful in plants. Thus, for live-cell imaging, researchers have used fluorescent actin-binding probes, such as Lifeact (Riedl et al., 2008), talin (Kost et al., 1998), and the actin-binding domain of fimbrin (Sheahan et al., 2004). However, these probes also generate artifacts, particularly when highly expressed (Vidali et al., 2009a; van der Honing et al., 2011).

Many of the issues arising from using F-actin probes stem from the fact that the binding of these probes to F-actin differs and often is affected by proteins that may be associated with the actin cytoskeleton (Table II). Thus, depending on the probe used and the relative amount of the probe, different subcellular structures can be observed. For example, in Arabidopsis (Arabidopsis thaliana) pollen tubes, fluorescent phalloidin staining revealed F-actin occupying the whole tip region, while Lifeact-GFP revealed pollen tubes with relatively little actin at the tip (Zhang et al., 2016). In root hairs, Ketelaar et al. (2003) visualized F-actin with antibodies or phalloidin. While both methods produced similar results in general, the details of the actin structure were quite different. In moss protonemata, F-actin imaging with fluorescent phalloidin (Vidali et al., 2009a) and Talin-GFP (Finka et al., 2007; Perroud and Quatrano, 2008), revealed thick apical bundles and a prominent apical accumulation. However, the expression of Lifeact-GFP at noninhibitory levels labeled finer actin filaments and provided more detail in the apical structures (Vidali et al., 2009a). Given the species-specific variation and the inherent challenges of imaging F-actin arrays, one must consider the conclusions made within the context of the visualization method and plant species used.

Despite differences in F-actin structures between the probes used and the species of cell type examined (Stephan, 2017), it is nevertheless possible to recognize an emergence of commonalities. For example, most pollen tubes exhibit a fringe of short longitudinally oriented actin microfilaments located at the cell cortex near the cell tip, and long actin bundles are found along the length of the cell (Lovy-Wheeler et al., 2005; Fig. 1A). In growing pollen tubes expressing Lifeact-GFP, the fringe was observed to rapidly remodel and move forward with growth (Rounds et al., 2014). Additionally, when pollen tube growth was inhibited with cyanide, the apical fringe degraded more rapidly than the actin cables in the shank of the tube (Winship et al., 2016). In pollen grains, Vogler et al. (2015) showed an F-actin accumulation 180° from the future germination site. This accumulation translocates to the shank after growth initiates. Using a pollen-specific promoter, Actin3, to drive the expression of Lifeact-GFP, Jásik et al. (2016) reported fine F-actin at the tip and dynamic actin along the shank of germinating pollen tubes. Studies such as these demonstrate that pollen tubes have at least two distinct arrays: an apical array that is very dynamic and a subapical array that tends to have more bundled F-actin.

Figure 1.

Apical cytoskeletal structures and ion gradients. A, Pollen tube, modeled after lily. Cortical F-actin (orange) forms a fringe overlapping with the OH⁻ gradient and the outer reaches of the calcium gradient. Microtubules (blue) form an inverted cone that just intersects the actin fringe. B, Root hair, modeled after Arabidopsis. Short F-actin filaments fill the apex, overlapping with the highest levels of calcium and OH⁻. As both gradients taper off, actin bundles and microtubules are found toward the rear of the cell. C, Moss chloronemal tip, modeled after P. patens. Apical actin is enriched along the entire dome, and fine cortical actin is present along the length of the cell. Contiguous cytoplasmic microtubules originating from the nucleus (N) point toward the cell tip. D, Moss caulonemal tip, modeled after P. patens. The apical actin is a highly focused spot and also the target of polymerizing cytoplasmic microtubules.

F-actin structures in growing root hairs share aspects of pollen tube structures. Actin bundles are prominent along the shank of the root hair but do not invade the tip region (Fig. 1B). Rather than a fringe, fine actin filaments appear at the tip (Ketelaar, 2013). As with the apical actin structures in pollen tubes, the fine actin filaments in the apex of root hairs are dynamic, as they are more sensitive to depolymerizing drugs than actin cables in the shank (Ketelaar et al., 2003). In addition to vesicle guidance to the tip, it is thought that this fine F-actin functions to keep larger organelles out of the tip region (Emons and Ketelaar, 2009), thus maintaining the apical clear zone (Ketelaar, 2013). Using F-actin as a filter is reminiscent of what has been observed in pollen tubes (Hepler and Winship, 2015).

Moss protonemata have two tip-growing cell types: chloronemata and caulonemata. Chloronemata are distinguished from caulonemata by growing more slowly and possessing perpendicular cell plates. Because caulonemata grow more quickly than chloronemata, F-actin architecture in living tip-growing cells has been analyzed predominantly in caulonemata. Moss protonemata exhibit an enrichment of F-actin at the very apex and fine cortical actin filaments along the length of the cell (Vidali et al., 2009a; Fig. 1, C and D). Bundles are apparent but are few and reside toward the rear of the cell, with subapical cells exhibiting more bundled F-actin. While an apical F-actin enrichment occurs in both cell types, their structures differ depending on the cell type. In chloronemata, apical F-actin localizes over the whole apical dome (Vidali et al., 2009a; Fig. 1C), whereas in caulonemata, F-actin exists as a focused apical spot (Vidali et al., 2010; Fig. 1D). Both the apical F-actin aggregation and the fine cortical F-actin are very dynamic, remodeling on the order of seconds (Vidali et al., 2010). Similar to the apical actin in root hairs, these structures in moss have been shown to be involved in vesicle guidance (Bibeau et al., 2017).

In summary, all three tip-growing cells possess longitudinally oriented actin cables in their shanks. However, apical actin structures differ. Pollen tubes, at least in lily (Lilium longiflorum) and tobacco (Nicotiana tabacum), possess a cortical actin fringe, while moss caulonemata maintain a focused spot of F-actin. Root hairs, in contrast, have an array of fine actin at their tips. Importantly, though, all three tip structures are dynamic, which contrasts with the relatively stable F-actin in the subapical regions. Critically, it is the dynamic apical populations that are essential for tip growth (Vidali et al., 2001; Vazquez et al., 2014).

Microtubule Cytoskeleton

In contrast to the essential nature of F-actin, microtubule-depolymerizing drugs have little effect on angiosperm pollen tube germination and growth in vitro, suggesting that microtubules are not required. However, in both root hairs (Bibikova et al., 1999) and protonemata (Doonan et al., 1988), microtubules likely play a role in maintaining the orientation of polarity. When treated with microtubule-depolymerizing drugs, both cell types no longer grow straight. Instead, the cells tend to change direction and fork, generating two or more outgrowths (Doonan et al., 1988; Bibikova et al., 1999; Ketelaar et al., 2003; Sieberer et al., 2005; Finka et al., 2007).

Regardless of their contribution to tip growth, striking microtubule arrays are evident in pollen tubes as well as root hairs and moss protonemata (Fig. 1). As opposed to F-actin visualization, which relies on actin-binding reagents, imaging the microtubule cytoskeleton is most commonly accomplished by labeling microtubules directly using antibodies to tubulin in fixed material or translational fusions of tubulin with a fluorescent protein in live material. As a result, microtubule imaging is less prone to the artifacts that plague F-actin visualization.

In pollen tubes, microtubules are oriented parallel to the long axis of the cell. In both lily and tobacco pollen tubes, these long cortical microtubules are often coaligned with fine filaments, which appear to be F-actin based on immunogold labeling (Lancelle and Hepler, 1991). Additionally, lily pollen tubes possess a funnel array of microtubules at the interface between the apical clear zone and the shank (Lovy-Wheeler et al., 2005; Fig. 1A). In root hairs, microtubule organization depends on the stage of growth. Generally, root hairs have two populations of microtubules: cortical and endoplasmic (cytoplasmic). Cortical microtubules are observed in both elongating and mature root hairs (Fig. 1B). Whether cortical microtubules are visualized at the very apex is dependent on the method of fixation (Sieberer et al., 2005). The polarity of polymerizing cortical microtubules, however, is dependent on growth status. Growing root hairs have cortical microtubules that grow away from the nucleus and, thus, have baseward and tipward trajectories. Conversely, mature root hairs have exclusive tipward polymerization (Ambrose and Wasteneys, 2014). Similarly, endoplasmic microtubules are only observed in growing root hairs (Sieberer et al., 2005) and emanate from the nucleus (Van Bruaene et al., 2004; Fig. 1B). In moss protonemata, there are also two populations of microtubules: cytoplasmic and cortical. In the apical cell, the polarity of cytoplasmic microtubules between the nucleus and the cell apex is tipward, with an apical focus of microtubule plus ends (Hiwatashi et al., 2014). In contrast, the cortical microtubule array consists of shorter microtubules that lack polarity (Burkart et al., 2015; Nakaoka et al., 2015).

Because microtubules appear to be required for directionality in root hairs and protonemata, one plausible hypothesis is that they spatially restrict specific F-actin populations mediating outgrowth (see Outstanding Questions). In support of this hypothesis,Hiwatashi et al. (2014) discovered that two microtubule-based motors, KINESIN FOR INTERDIGITATED MICROTUBULES1a (KIND1a) and KIND1b, regulate microtubule organization and protonemal growth. The kind1a,b knockout plants exhibit protonemata that do not grow straight but still have tipward-growing cytoplasmic microtubules. However, the apical microtubules are disorganized, suggesting that KIND1a and KIND1b focus apical microtubules in protonemata (Hiwatashi et al., 2014). Interestingly, in wild-type cells, cytoplasmic microtubules focus on the same location occupied by the apical actin spot (Fig. 1, C and D), suggesting that microtubules may interact directly with the apical F-actin structure. Myosin VIII is an actin-based motor protein that links microtubules to actin during cell division in protonemata (Wu and Bezanilla, 2014; Table II). Protonemata completely lacking myosin VIII (ΔmyoVIIIa,b,c,d,e) grow slowly and less straight than the wild type, particularly when grown under nutrient-deficient conditions (Wu and Bezanilla, 2014). Together, these studies suggest that an active link between actin and microtubules is critical for directionality during tip growth.

Studies of MICROTUBULE ASSOCIATED PROTEIN18 (AtMAP18) also support a role for microtubule-actin interactions in tip growth. MAP18 binds both microtubules and F-actin (Wang et al., 2007; Zhu et al., 2013). In addition, MAP18 has calcium-dependent F-actin-severing activity (Zhu et al., 2013). Importantly, MAP18 knockdown pollen tubes cannot grow straight (Zhu et al., 2013), reminiscent of root hairs and protonemata treated with oryzalin (Doonan et al., 1988; Bibikova et al., 1999). Furthermore, F-actin arrays are altered in map18 pollen tubes: map18 pollen tubes no longer exhibit fine apical F-actin; instead, F-actin bundles invade the tip. AtMAP18-GFP localized to the plasma membrane and to an apical inverted cone pattern in map18 knockdown pollen tubes (Zhu et al., 2013; Table II). It is plausible that the location of MAP18 is determined by microtubules, and where microtubules interact with F-actin, the latter may be severed depending on the presence of calcium (Fig. 1).

Interestingly, MAP18 localizes similarly in root hairs (Table II), and MAP18 knockdown plants have root hairs half the length of wild-type plants (Kang et al., 2017). To investigate MAP18’s function in root hairs, Kang et al. (2017) tested for genetic and physical interactions between MAP18 and Rho/Rac of Plants2 (ROP2), a small GTPase considered to be a master regulator of tip growth in root hairs (Jones et al., 2002). Kang et al. (2017) found that MAP18 and ROP2 are in the same genetic pathway and that they interact physically with MAP18, preferentially binding the inactive form of ROP2 and thereby promoting ROP2 activity. Kang et al. (2017) went on to show that MAP18 competes with the ROP2 guanine dissociation inhibitor for binding to ROP2. From these and earlier studies (Wang et al., 2007; Zhu et al., 2013), Kang et al. (2017) suggest that MAP18 functions through ROP2 to reinforce polarity and independently of ROP2 by modulating the cytoskeleton.

IONS

Modulation of the many processes and structures involved in tip growth, including the cytoskeleton, is achieved not only by regulator proteins but also by the cytoplasmic environment. Here, we highlight local ion concentrations, in particular calcium and protons, as drivers for these processes (Hepler, 2005, 2016; Michard et al., 2017). To that end, a tip-focused calcium gradient is particularly prominent in growing pollen tubes, where the high point localizes to the extreme apex of the pollen tube, directly appressed to the apical plasma membrane (Fig. 1A). Root hairs have a similarly tip-focused, albeit shallower, calcium gradient (Fig. 1B). To date, calcium in moss protonemata remains understudied. In an early report on dark-grown caulonemata injected with a calcium-sensitive fluorescent dye, Tucker et al. (2005) observed UV (340 and 380 nm light)-stimulated waves associated with both the apex and the base of apical cells. However, Tucker et al. (2005) did not observe a tip-focused calcium gradient, as seen in growing pollen tubes and root hairs.

In root hairs and pollen tubes, many channels both on the plasma membrane as well as on organelles are likely to be involved in establishing and maintaining the apical calcium gradient. Patch-clamp assays have identified stretch-activated channels in pollen tubes (Dutta and Robinson, 2004), and their involvement in generating the apical calcium gradient seems plausible as a product of turgor-driven deformation of the apical plasma membrane. More recently, the Cyclic Nucleotide-Gated Channel18 (CNGC18) in Arabidopsis was implicated in tip growth, since CNGC18-null plants are male sterile (Gao et al., 2016; Table I). Investigating further, cngc18 pollen tubes were shown to have irregular calcium oscillations and have difficulty finding the ovule. Additionally, root hairs of CNGC14 knockdown and knockout plants are shorter and have altered calcium oscillation profiles (Zhang et al., 2017). Another class of transporters, the Glu receptor family (GLR), plays a role in modulating the calcium signal in pollen tubes. Atglr1.2-1 pollen tubes are deformed, and the amplitudes of calcium oscillations are diminished compared with the wild type (Michard et al., 2011; Table I). In addition to stretch-activated, CNG, and GLR channels, pharmacological assays on pollen suggest that F-actin (Wang et al., 2004) and voltage-regulated (Wu et al., 2011) calcium channels play a role in calcium homeostasis. Further study is needed to determine what complement of channels is required to establish the calcium gradients during tip growth (see Outstanding Questions).

The calcium gradient may modulate growth by stimulating secretion (Winship et al., 2016) and/or regulating F-actin structures through actin-binding proteins, which we focus on here. Two well-established actin-binding proteins with calcium dependence are villin/gelsolin and profilin (Table II), which are prime candidates for playing a key role in establishing different actin arrays in root hairs and pollen tubes. While the actin-binding and bundling activities of both the gelsolin repeat domains and the C-terminal headpiece domain (Finidori et al., 1992) are calcium independent (Khurana et al., 2010), villin’s actin-severing activity is calcium dependent (Huang et al., 2004). Villin’s localization is primarily cytosolic, but it appears to bind to filaments along the shank of root hairs (van der Honing et al., 2012; Table II). It seems likely that apical villin would likely sever, not bundle, F-actin, given the tip-focused calcium gradient (Monshausen et al., 2008). In support of this, vln2vln5 pollen tubes have an enlarged apical actin structure (Qu et al., 2013). Taken together, villin is thought to be a major regulator of actin structures (Huang et al., 2015).

Profilin binds to monomeric actin in a calcium-dependent manner (Kovar et al., 2000) and inhibits nucleation and polymerization. Pollen tubes of profilin mutants have aberrant (Qu et al., 2017) or absent (Liu et al., 2015) actin fringes. Transient knockdown of profilin with RNAi in moss demonstrated that protonemal growth is dependent on profilin (Vidali et al., 2007). Indeed, profilin knockdown plants are significantly less than half the size of controls, and cells grow isotropically (Vidali et al., 2007). The profilin-actin dimer is the substrate for formins: actin nucleators and elongators (Goode and Eck, 2007; Table II). Formins, which are not known to have calcium dependency, will nevertheless be affected by calcium concentrations because they require profilin-actin as a substrate. One can imagine, then, that profilin’s binding activity modulates the structure of the actin array via its interactions with various cellular nucleators of actin polymerization (Suarez et al., 2015).

Pollen tubes and root hairs also possess proton or pH gradients. In growing lily pollen tubes, there is an alkaline band a few micrometers behind the tip, with a small acidic patch at the tip itself (Feijó et al., 1999; Lovy-Wheeler et al., 2006). It is noteworthy that tobacco H⁺-ATPase1-GFP, an enzyme responsible for creating proton gradients, localizes to the plasma membrane of the shank and an apical cytoplasmic inverted cone in tobacco pollen tubes (Certal et al., 2008), fitting well with the proton gradient (Fig. 1). Although pH gradients are difficult to image because of the rapid diffusion of protons, their importance to tip growth is apparent because modulating intracellular pH can change the direction of growth (Hu et al., 2017). Imaging both the cytosolic and extracellular environments of root hairs also has revealed localized gradients (Monshausen et al., 2007). While the pH gradient is not as steep as in pollen tubes, the gradient is present nonetheless and is coupled with root hair development (Bibikova et al., 1998).

The calcium and pH gradients present at the cell apex of pollen tubes and root hairs are not static but oscillate. Indeed, analyses of these oscillations explore how ion gradients mechanistically link to tip growth. Initial studies on pollen tubes demonstrated that calcium gradient oscillations followed the oscillations in growth rate (Messerli et al., 2000; Cárdenas et al., 2008). By contrast, the apical alkaline band oscillation preceded the growth rate oscillation, virtually matching the oscillation in secretion of new wall material (Lovy-Wheeler et al., 2006). Conversely, studies on the extracellular pH in root hairs indicate that an elevation in alkalinity follows the peak in growth rate (Monshausen et al., 2007). While it is unclear how these studies fit together, the work on pollen tubes provides support for the idea that changes in pH have a close relationship with the process of growth. To explore the connection between calcium, pH gradients, and growth in pollen tubes, Winship et al. (2016) reversibly inhibited pollen tube growth with cyanide (Table I) and then followed the return of the calcium and pH gradients. Although the calcium gradient appeared before growth restarted, it was noticeably behind that of the alkaline band and the process of secretion itself. Indeed, the alkaline band reemerged even before the cyanide had been removed (Winship et al., 2016). These results provide impetus to the notion that pH changes are fundamentally involved in tip growth.

Again, pH regulation of tip growth is thought to affect the actin cytoskeleton. In particular, the elevated pH (7.4), as found in the alkaline band, would stimulate the turnover of F-actin in the fringe in pollen tubes because Actin Depolymerizing Factor (ADF) optimally fragments F-actin at this pH (Bamburg, 1999; Hepler, 2016; Table II). It is additionally noteworthy that, in lily pollen tubes, ADF localizes to the region of the alkaline band and to the actin fringe (Lovy-Wheeler et al., 2006). Activation of ADF activity helps explain the constant turnover of the actin fringe as it keeps pace with elongation. Interestingly, recent work has uncovered a new calcium-ADF interaction. ADF1 from Malus domestica binds to and severs actin in a calcium-dependent manner (Yang et al., 2017), suggesting that dual regulation of ADF may be important for fine-tuning actin fringe turnover. In contrast to pollen tubes, ADF in moss protonemata is cytosolic (Augustine et al., 2008; Table II). However, ADF is absolutely essential for tip growth and F-actin dynamics in protonemata (Augustine et al., 2011), reinforcing the key role that dynamic actin plays in tip growth.

While calcium oscillations in pollen tubes appear to have a single periodicity, quantification of calcium oscillations via light sheet microscopy and spectral analyses in root hairs revealed two concurrent periodicities (Candeo et al., 2017). The mechanism behind the concurrent periodicities in root hairs is not known (see Outstanding Questions). Two methods to generate different periodicities could include variable entry or variable sequestration. For entry, CNGC14 is a prime candidate for being one component of that oscillation profile, as cngc14 root hairs have altered oscillation profiles (Zhang et al., 2017). Mechanisms of sequestration are not as well understood. That being said, pollen tubes with suppressed expression of calreticulin, a resident endoplasmic reticulum protein with calcium sequestration properties, had aberrant cytosolic calcium levels (Suwińska et al., 2017), suggesting a role for calreticulin in calcium homeostasis. How calcium enters the endoplasmic reticulum of pollen tubes, however, is unclear. These works highlight the need to study the mechanism of calcium recycling in tip-growing systems (see Outstanding Questions).

CELL WALL

Plant cell shape, and thus growth, are both confined and defined by the cell wall. To understand how to change the cell wall, for example achieving a local wall weakening at the cell tip for tip growth, one must consider its composition. Tip-growing cells have polymers found in most plant cells, such as cellulose, hemicellulose, and pectins. However, immunochemical data show that the amounts and distribution of these polymers differ from typical plant cells, with cellulose being sparse, while pectin is considerably enriched at the tip in pollen tubes, root hairs, and protonemata (Bosch and Hepler, 2005; Berry et al., 2016). For pollen tubes, mounting evidence increasingly supports the view that pectins, either as deposited directly or following enzymatic modification, play a major role in controlling the rigidity of the wall (Caffall and Mohnen, 2009).

In lily and tobacco pollen tubes, pectins are secreted in an oscillatory manner, with the increase of pectin anticipating the increase in growth rate (McKenna et al., 2009; Rounds et al., 2011). These observations have given rise to the idea that the intercalation or intussusception of wall material weakens the existing wall and allows for stress relaxation at the locus of secretion (Hepler et al., 2013). But further results in tobacco show that pectin methyl esterase (PME) is secreted with the same kinetics as the bulk pectin (McKenna et al., 2009). Thus, PME secretion anticipates growth. It is thought that PME increases the presence of free carboxyl residues, thereby enhancing calcium cross-linking and rigidifying the apical wall (Bosch and Hepler, 2005). This activity can explain the rigidity of the pollen tube shank cell wall. However, at the tube tip, a PME-induced stiffening would be contrary to the local weakening of the wall. An explanation for this paradox may come from work in nonpollen tube systems, where PME may weaken the existing cell wall (Peaucelle et al., 2011, 2012, 2015; Braidwood et al., 2014). The proposed model suggests that PME demethoxylates pectin, rendering the wall components more accessible to degrading enzymes, which, in turn, weakens the wall. Additionally, if PME removes methoxy esters in a block-wise manner, the resulting wall would be more tightly cross-linked by calcium. However, if the PME activity results in random demethoxylation, then the wall would not necessarily be strongly cross-linked. Regardless of the manner of demethoxylation, enzymatic cleavage results in the release of a proton, which locally acidifies the wall and contributes to cell wall loosening and extension. It is also possible that there could be a parallel increase in the PME inhibitor (Röckel et al., 2008), which could negate the esterase activity. Whether PME is active or inactive at the very apex is still unclear. Nevertheless, it is clear that the presence of PME correlates with local weakening, leading to apical cell wall extension.

Glycoproteins also may contribute substantially to cell wall integrity and, thus, polarized growth. Extensin18 (EXT18), a Hyp-rich glycoprotein, appears to direct pollen tube structure and growth (Choudhary et al., 2015). EXT18-null mutants are less fertile, which may result in part from reduced pollen tube growth but also from an increase in pollen tube bursting (Table I). The bursting phenotype lends support to the idea that EXT18 provides structural support to the growing pollen tube at the tip or is involved in the communication between the wall and the cytoplasm.

More attention has been given to arabinogalactan proteins (AGPs), which have been shown to modulate growth in different tip-growing cells (Tan et al., 2012). AGPs consist of branched chain carbohydrates linked to a peptide backbone, which may be attached to the plasma membrane through a glycosylphosphatidyl inositol anchor. While ubiquitous, seemingly involved in growth across plant cell types, their specific functions are poorly understood (Ellis et al., 2010). To elucidate the functions of AGPs, many studies have used Yariv phenylglycoside, a dye that binds and stains AGPs as well as blocks their function. At low concentrations (1 μm), Yariv reagent blocks cell extension in pollen tubes (Roy et al., 1998, 1999; Mollet et al., 2002) and moss protonemata (Lee et al., 2005; Table I). Interestingly, the Yariv reagent inhibits cell extension in pollen tubes but not secretion. An electron microscopic analysis revealed that Yariv-treated cells build up wall material in the apical periplasmic space (Roy et al., 1999). In contrast to other inhibitors such as caffeine or cyanide, which degrade the apical calcium gradient, pollen tubes treated with Yariv have only mildly diminished calcium levels at their tip (Roy et al., 1999; Table I). Continued secretion and the presence of the calcium gradient suggest that these activities are likely linked.

Genetic approaches also have implicated AGPs in tip growth. In Arabidopsis, RNAi of AGP6 and AGP11 resulted in plants with reduced fertility, which is partly due to impaired pollen tube growth (Levitin et al., 2008; Table I). Similarly, moss protonemata AGP1 knockout plants have reduced protonemal cell length (Table I). In root hairs, AGPs are reduced markedly in the Arabidopsis reb1-1 mutant. The root epidermal trichoblasts of reb1-1 exhibit morphological defects, including bulging and disorganized cortical microtubules (Andème-Onzighi et al., 2002; Driouich and Baskin, 2008; Table I). Given that REB1 is part of a gene family involved in sugar metabolism, it seems likely that REB1 is involved in the synthesis of AGPs rather than regulating microtubule organization directly. However, compromised AGPs at the plasma membrane and in the cell wall may affect the organization of cortical microtubules (Andème-Onzighi et al., 2002), ultimately affecting the direction of cell expansion (Ketelaar et al., 2003).

While it is clear that AGPs are a major player in cell wall expansion, the underlying mechanism is less straightforward. AGPs could control the assembly or incorporation of new material into the matrix of the cell wall. Alternatively, given that AGPs bind calcium with high affinity, periplasmic AGPs could serve as a calcium capacitor, creating a significant calcium reservoir (Lamport and Várnai, 2013; Lamport et al., 2014). Given that calcium ions are essential for cell wall integrity, it is not clear how the AGP/calcium capacitor modulates the wall. Perhaps the capacitor competes for calcium throughout the wall and works together with the stressed pectin residues to weaken the wall (Boyer, 2009). Alternatively, AGPs might directly facilitate the intercalation of new pectin monomers deep into the wall matrix, where its bonding activities will weaken existing wall bonds (Ray, 1992; Hepler et al., 2013). If the AGP calcium reservoir also is available for intracellular signaling, then it may serve to link wall status to the cytoplasm.

CONCLUSION

In this Update, we have called attention to advances in our understanding of tip growth in pollen tubes, root hairs, and moss protonemata. To establish a mechanistic understanding of tip growth is to elucidate how intracellular events cause a local weakening of the apical cell wall, a prerequisite for turgor-driven cell expansion. The weakening, of course, must be balanced with the deposition of new wall material, or else tip growth fails (Table I). As discussed, localized ion gradients almost certainly play pivotal roles in different processes. For example, the turnover of F-actin in the apex, thought to occur through the ion-dependent activities of actin-binding proteins (Table II), might facilitate the dynamics of exocytic and endocytic vesicles (Bibeau et al., 2017; see Outstanding Questions). It is unsurprising, then, that dynamic ion gradients and cytoskeletal elements occur in the same place: the site of polarized expansion (Fig. 1). Moving forward, it is imperative to understand how the cell wall signals to the intracellular machinery responsible for its expansion (see Outstanding Questions). By continuing to investigate both the similarities and differences across tip-growing systems, we gain deep insights into fundamental and derived processes that sum together in a tip-growing plant cell.