Sorting signals for PIN1 trafficking and localization

ABSTRACT

PIN-FORMED (PIN) family proteins direct polar auxin transport based on their asymmetric (polar) localization at the plasma membrane. In the case of PIN1, it mainly localizes to the basal (rootward) plasma membrane domain of stele cells in root meristems. Vesicular trafficking events, such as clathrin-dependent PIN1 endocytosis and polar recycling, are probably the main determinants for PIN1 polar localization. However, very little is known about the signals which may be involved in binding the μ-adaptin subunit of clathrin adaptor complexes (APs) for sorting of PIN1 within clathrin-coated vesicles, which can determine its trafficking and localization. We have performed a systematic mutagenesis analysis to investigate putative sorting motifs in the hydrophilic loop of PIN1. We have found that a non-canonical motif, based in a phenylalanine residue, through the binding of μA(μ2)- and μD(μ3)-adaptin, is important for PIN1 endocytosis and for PIN1 traffcking along the secretory pathway, respectively. In addition, tyrosine-based motifs, which also bind different μ-adaptins, could also contribute to PIN1 trafficking and localization.

PIN-FORMED (PIN) family proteins are auxin efflux carriers which are characterized by their asymmetric (polar) localization in plant cells. In Arabidopsis, this family includes 8 members, all of them characterized by a central hydrophilic loop which separates 2 hydrophobic domains, each one with 5 predicted transmembrane regions.¹ Depending on the length of the central hydrophilic loop PIN proteins can be classified into 2 subfamilies. Type 1 or long PINs (including PIN1, PIN2, PIN3, PIN4 and PIN7) contain a long hydrophilic loop and localize to the plasma membrane. In contrast, type 2 or short PINs have a shorter (PIN6) or strongly reduced (PIN5 and PIN8) hydrophilic loop and localize at the endoplasmic reticulum.¹ The predominant localization of long PINs to the plasma membrane and their polarity suggest that their central hydrophilic loop should contain signals for their trafficking and localization, although cell-type specific factors may also be important.²

In contrast to animal epithelial cells, which are characterized by apical and basal domains, separated by the so-called tight junctions, 4 polar domains (apical, basal, outer lateral and inner lateral) have been identified in plant epidermal cells (for an excellent review see ³). Therefore, PIN proteins need a polarized vesicular transport to reach specific plasma membrane domains, including endocytosis and polar recycling,⁴ processes which may involve clathrin-coated vesicles. Clathrin-coated vesicles are formed at the plasma membrane and have been shown to be involved in endocytosis.^5-7 They also form at the trans-Golgi network (TGN), but their precise role in post-Golgi trafficking is far from being clear (for an excellent review see ⁸). Sorting within clathrin-coated vesicles requires clathrin adaptor complexes (APs), which recognize sorting signals in cargo proteins via their medium (μ) subunit.⁹ The Arabidopsis genome encodes subunits (adaptins) of 4 types of putative AP complexes (AP-1 to AP-4) including 5 medium subunits, named μA (μ2), μB1 (μ1–1), μB2 (μ1–2), μC (μ4), and μD (μ3).¹⁰ Several clathrin adaptors (including AP-1A, AP-1B and AP-4) have been shown to mediate basolateral polarity in epithelial cells and neurons.^11-13 A lot of recent reports have investigated the function of clathrin adaptor proteins in plant cells, including their possible contribution to the polar localization of PIN1 ¹⁴ and PIN2.¹⁵ However, very little is known about the signals which are recognized by μ-adaptins in plants. PIN1 endocytosis, which may take place preferentially at lateral cell sides,¹⁶ is clathrin-dependent ⁵ and presumably requires the AP-2 complex. PIN1 endocytosis signals may be important for the polar localization of PIN1. Indeed, it has been reported that signals involved in basolateral sorting in epithelial cells often overlap with those involved in endocytosis, including tyrosine-based motifs but also other non-canonical motifs.^11-12 However, the signals involved in PIN1 endocytosis have not been identified yet.

The central hydrophilic loop of PIN1 contains several tyrosine residues which could match the consensus YXXΦ motif (where Y is tyrosine, X is any amino acid and Φ a bulky hydrophobic residue) and could thus be involved in binding μ-adaptins (Fig. 1). In addition, it also contains a non-canonical motif based in a phenylalanine residue, very similar to one of the endocytosis signals of the mannose 6-phosphate receptor MPR46.^17-18 In a recent publication, we have performed a systematic mutagenesis analysis to investigate the contribution of these residues to μ-adaptin binding and PIN1 trafficking and localization.¹⁹ In this article we show new data with another mutant involving tyrosine 480 (Fig. 2). Interestingly, the residues analyzed showed certain degree of specificity in binding to different μ-adaptins. Phenylalanine165 bound specifically μA(μ2)- and μD(μ3)-adaptin, but not μB1- or μC-adaptin ¹⁹(Table 1). While tyrosines 328 and 394 could bind all μ-adaptins, tyrosine 260 bound μB(μ1)- and μD(μ3)-adaptin with much more efficiency than μA(μ2)-adaptin and did not bind μC(μ4)-adaptin ¹⁹ (Table 1). As shown in Fig. 2A, tyrosine 480 bound μB(μ1)- and μC(μ4)-adaptin, but not μA(μ2)- or μD(μ3)-adaptins (which bound to the C-terminal portion of the hydrophilic loop of PIN1 independent of tyrosine 480). It is possible that some of the tyrosine residues analyzed are functionally redundant in μ-adaptin binding. These would explain why single mutants in some of these residues do not affect steady-state localization of PIN1.¹⁹ Tyrosine 480 was previously proposed to be important for PIN1 localization, since a triple mutation in the motif containing this tyrosine (NPNSY to NSLSL) caused the accumulation of PIN1 at the ER.²⁰ We have now generated a PIN1:GFP mutant with a single mutation in tyrosine 480 and found that it partially accumulated at intracellular structures similar to those found in the PIN1:GFP-F165A mutant (Fig. 2B). However, we have not further analyzed the identity of these intracellular structures. In any case, this tyrosine residue, which is found in all long PINs, seems to be important for PIN1 trafficking and localization.

Figure 1.

Sequence of the hydrophilic loop of the PIN1 auxin efflux carrier. Residues analyzed for μ-adaptin binding are highlighted in red and their position in the sequence with an asterisk. The motifs involved in μ-adaptin binding containing these residues are shown in a grey box. Putative ER export signals (containing acidic residues, in blue) are shown in a yellow box.

Figure 2.

Characterization of a PIN1 mutant in tyrosine 480. (A) Binding of the receptor binding domain of Arabidopsis μ-adaptins, with a N-terminal (His)₆-tag, to the C-terminal part of the cytosolic loop of PIN1 (residues 403–482) or to a mutant version in tyrosine 480 (Y480A), both fused with GST, using pull-down experiments and Western blotting with His antibodies (as described in 19). (B) CLSM of primary roots of 4-days-old seedlings expressing a PIN1:GFP-Y480A mutant in pin1 background. Note the accumulation of this mutant in intracellular structures at stele cells. Scale bar: 10 μm.

Table 1.

μ-adaptin binding to different residues in the hydrophilic loop of PIN1.

Residue	μ-adaptin binding
F165	μA(μ2), μD(μ3)
Y260	μB(μ1) > μD(μ3) > μA(μ2)
Y328	μA(μ2), μB(μ1), μC(μ4), μD(μ3)
Y394	μA(μ2), μB(μ1), μC(μ4), μD(μ3)
Y480	μB(μ1), μC(μ4)

We found that mutation of phenylalanine 165 led to a significant decrease in endocytosis, consistent with this residue (which binds μA(μ2)-adaptin) being important for PIN1 endocytosis.¹⁹ To our knowledge, this is the first endocytosis signal reported for the cytosolic loop of PIN1. However, the PIN1:GFP-F165A mutant was still efficiently internalized, which suggests that other residues in the cytosolic tail of PIN1 may also participate in PIN1 endocytosis. Indeed, μA(μ2)-adaptin also bound to tyrosines 260, 328 and 394 (Table 1). This situation is similar to that of the cytosolic tail of mannose 6-phosphate receptor MPR46, which contains 3 internalization signals, one based on a similar phenylalanine residue, a tyrosine (YXXϕ) motif and a dileucine motif. A single mutation in the critical phenylalanine did not cause a significant decrease in endocytosis. However, when one or 2 of the other internalization signals were also mutated, mutation of the phenylalanine caused an additive negative effect on endocytosis.¹⁷ On the other hand, this motif was shown to bind AP2 but not AP1. This is consistent with our data, which showed that the motif containing phenylalanine 165, which is present in all long PINs, can only bind μA(μ2)- and μD(μ3)-adaptins, but not μB(μ1)- or μC(μ4)-adaptin.

In addition, phenylalanine 165 may also be important for trafficking along the secretory pathway, since the PIN1:GFP-F165A mutant accumulated in intracellular structures containing ER markers in root stele cells.¹⁹ Since the localization of PIN1-GFP in a μ2-adaptin mutant was very similar to that in wild-type background, we hypothesized that the lower binding of μ2-adaptin cannot be the reason for the accumulation of the PIN1:GFP-F165A mutant in intracellular structures. In contrast, PIN1-GFP accumulated in big intracellular structures in a μ3-adaptin mutant,¹⁹ similar to those found in mutants of other subunits of the AP-3 complex,^21-22 suggesting that the AP-3 complex is essential for PIN1 trafficking and localization. However, the precise function of AP-3 in PIN1 trafficking still needs to be elucidated.

Using a gnl1 mutant expressing BFA-sensitive GNL1 and BFA treatment we have also found that transport of PIN1 to the plasma membrane takes place via the Golgi apparatus, presumably via COPII vesicles.¹⁹ Indeed, the hydrophilic loop of PIN1 contains putative ER export signals, in particular 3 di-acidic motifs ((D/E)x(D/E)) which could be involved in COPII binding (Fig. 1). However, the contribution of these motifs to ER exit and transport of PIN1 to the plasma membrane has not been analyzed yet.

In summary, analysis of the sorting signals in the central hydrophilic loop of PIN1 may shed some light on its trafficking pathways and on the function of μ-adaptins and clathrin adaptor complexes in plant cells.

Disclosure of potential confllicts of interest

No potential conflicts of interest were disclosed.