Post-translational Modifications of Transporters

Abstract

Drug transporter proteins are critical to the distribution of a wide range of endogenous compounds and xenobiotics such as hormones, bile acids, peptides, lipids, sugars, and drugs. There are two classes of drug transporters– the solute carrier (SLC) transporters and ATP-binding cassette (ABC) transporters –which predominantly differ in the energy source utilized to transport substrates across a membrane barrier. Despite their hydrophobic nature and residence in the membrane bilayer, drug transporters have dynamic structures and adopt many conformations during the translocation process. Whereas there is significant literature evidence for the substrate specificity and structure-function relationship for clinically relevant drug transporters proteins, there is less of an understanding in the regulatory mechanisms that contribute to the functional expression of these proteins. Post-translational modifications have been shown to modulate drug transporter functional expression via a wide range of molecular mechanisms. These modifications commonly occur through the addition of a functional group (e.g. phosphorylation), a small protein (e.g. ubiquitination), sugar chains (e.g. glycosylation), or lipids (e.g. palmitoylation) on solvent accessible amino acid residues. These covalent additions often occur as a result of a signaling cascade and may be reversible depending on the type of modification and the intended fate of the signaling event. Here, we review the significant role in which post-translational modifications contribute to the dynamic regulation and functional consequences of SLC and ABC drug transporters and highlight recent progress in understanding their roles in transporter structure, function, and regulation.

1. Introduction

Transporter proteins are integral membrane proteins critical to the uptake, distribution, and excretion of endogenous compounds and xenobiotics such as nutrients, hormones, bile acids, peptides, lipids, sugars, and drugs (1). They can be broadly categorized into two superfamilies: the SoLute Carrier (SLC) transporters comprising over 400 integral membrane proteins subdivided into 50+ families and the ATP-Binding Cassette (ABC) superfamily, consisting of 7 families (ABCA through ABCG). Membrane transporters are inherently hydrophobic and large transmembrane domains span the cellular membrane (2, 3). Despite their hydrophobicity, transporters have dynamic structures, and can adopt conformations that are not readily captured via X-ray crystallography. Therefore, specific proteoforms identified biochemically and computationally are often relied upon for mechanistic insight into the function of these proteins (4-6). Post-translational modifications (PTMs) are regulators of these structural events and are critical for the transporters’ structure, function, and regulation within the confines of the lipid environment. The hydrophilic loops and termini that face the intracellular milieu may be accessible for these modifications if the exposed amino acid side chains are solvent accessible, providing a mechanism in which the chemical nature of an amino acid can be altered (7).

While there are 400+ types of PTMs that have been identified to date, the most common variants that are known to play a role (or, those that have been actively investigated) in the regulation of transporters include phosphorylation, glycosylation, and ubiquitination. Biochemically, these modifications diversify the nature of the amino acid peptide-backbone or side chain through the addition of small chemical groups (e.g. phosphates), lipids (e.g. palmitic acid), carbohydrates (e.g. mannose), small proteins (e.g. SUMO), among other entities (4-11). Most, if not all, eukaryotic proteins undergo PTMs, and the likelihood for a given modification to occur is driven by the amino acid sequence, the structural and chemical constraints of the protein surface, and the availability of the necessary protein machinery and precursors to facilitate the modification (Figure 1) (8, 10). For transporters, this likelihood is further complicated by lipid-protein interactions driven by the buried transmembrane domains, which are obligatory to the functional expression of the protein. As such, solvent accessible residues located within the hydrophilic loops and termini of the membrane transporter often contain canonical consensus motifs (Table 1) that serve as recognition sites for not just the PTM, but also for the required adaptor proteins and enzymes needed to facilitate the modification (7, 12-16). While common consensus motifs are conserved across the proteome, there also are non-canonical recognition motifs that have been identified under physiologically relevant contexts (17).

Figure 1. Localization of Common PTMs on a General Transporter Structure.

Post-translational modifications occur at specific locations on a transporter. Most modifications described in the literature for transporters face the intracellular milieu, and modified amino acid residues must be solvent accessible prior to modification. Palmitoylation occurs closer to the lipid-protein interface of transmembrane domain regions. N-glycosylation is an extracellular facing modification.

Table 1. Examples of PTM Consensus Motifs.

Post-translational modifications often occur at specific recognition signals within the amino acid sequence of target protein. Where some modifications, such as N-glycosylation, occur within a strict motif, other modifications like phosphorylation occur at sequences which are specific recognition sites for a given kinase or phosphatase. Other modifications, like palmitoylation, have yet to be strictly defined.

PTM	Common Consensus Motif(s)
Serine/Threonine	RRXS; [R/K]XX[S/T]; RXRXX[S/T]
Phosphorylation*
Tyrosine Phosphorylation*	X[D/E]YX; YIYGSFK; EXIYXXPX
N-Glycosylation	N-X-S/T (X≠P or D)
Palmitoylation	Poorly defined, required thiolate anion
Ubiquitination	PEST rich regions
SUMOylation	ψKXE (ψ=hydrophobic amino acid)

^*Motifs are kinase specific

X= any amino acid

PTMs have been shown to influence transporter kinetics, both directly and indirectly (18). They do not just regulate the innate structure-function relationship driven by a transporter’s global architecture, but rather are also able to regulate this relationship down to the resolution of the structural events occurring at a residue level. From this level, PTMs are able to modulate a transporter’s function, expression, efficiency, structure, fate, interactions, and more. Additionally, the complexity of the response of PTM modulation is a function of the extent of modifications a transporter is able to accept and the availability of the environmental cues to signal for the post-translational event (Table 2) (7, 9, 17). This results in an exponential increase in specific “species” of a given transporter, referred to as proteoforms (19), which can provide deeper mechanistic understanding in transporter biology.

Table 2. Examples of Commonly Accepted Modifications at Amino Acid Residues.

Post-translational modifications occur only at specific residues that are capable of accepting the addition. Some residues can accept multiple types of modifications.

PTM	Common Consensus Motif(s)	Chemical Structure Examples
Serine/Threonin e Phosphorylation *	RRXS; [R/K]XX[S/T]; RXRXX[S/T]
Tyrosine Phosphorylation *	X[D/E]YX; YIYGSFK; EXIYXXPX
N-Glycosylation	N-X-S/T (X≠P or D)
Palmitoylation	Poorly defined, required thiolate anion
Ubiquitination	PEST rich regions
SUMOylation	ψKXE (ψ=hydrophobi c amino acid)

^*Motifs are kinase specific

X= any amino acid

The consequence of PTM modulation has been investigated for a wide range of SLC and ABC transporters, including but not limited to those described in the sections below. The purpose of this review is to provide a broad overview of the roles of PTMs in regulating transporters in higher vertebrates and humans, and consequently is not intended to be a comprehensive list of all of the PTMs identified to-date. Consequently, a diverse set of examples was selected from the literature to support the overarching themes in the post-translational regulation of ABC and SLC transporters (Table 3).

Table 3. Summary of Literature References.

Type of PTM	Role of Modification	Transporter Protein	Identified Residue(s)	Referenl ce(s)
Phosphory lation	On/Off Switch	DMT1; SLC22A2	S43	Seo 2016
		OCT1-3; SLC22A1-3	Y362 (OCT2)	Sprowl 2016; Minematsu 2011; Sauzay 2016
		SERT; SLC6A4		Annamalai 2012; Singhai 2017
		ASBT; SLC10A2	Multiple tyrosines	Annaba 2012;
		MATE1; SLC47A1		Sauzay 2016
		OAT1; SLC22A6
		OATP1B1; SLCO1B1
	Subcellular Fate	GLYT2; SLC6A5		de-Juan-Sanz 2011; de-Juan Sanz 2013
		DAT; SLC6A3		Zhu 1997; Melikian 1999; Hong; 2013
		OAT1; SLC22A6		Li 2013; Zhang 2008; Zhang 2013
		OATP1B1; SLCO1B1		Hong 2015
		GLUT1/4; SLC2A1/4		Foley 2011, Lee 2015
		MRP1; ABCC1	S871, 915, 930,961	Ambadipudi 2017; Olsen 2010; Shukalek 2016
		CERP; ABCA1	S/T	Wang 2005; Wang 2007;
	Cell Surface Trafficking	P-gp; ABCB1	S661, S667, S671, S683	Chambers 1994; Germann 1996; Goodfellow 1996; Xie 2018
	Multimerization	BCRP; ABCG2	T362	Xie 2008
	Confirmational Change/Gating Mechanism	CFTR; ABCC7	S660, S737, S795, S813	Cheng 1991; Zhang 2017
N- glycosylati on	Protease Protection	PEPT1; SLC15A1	N50 (six other)	Stelzl 2016; Stelzl 2017; Zhang 2002; Boll 1994; Wuensch
		ASBT; SLC10A2
		P-gp; ABCB1		Xie 2018
	Status Altered in Diseased States	OATP1B1/3; SLCO1B1/3		Clarke 2017
		NTCP; SLC10A1	N5, N11	Clarke 2017; Appelman 2017
	Microdomain Partitioning	GLUT2; SLC2A2		Ohtsubo 2013
	Substrate Binding	MRP4; ABCC4	N746; N754	Miah 2016
		MRP1; ABCC1	N19/N23 (cross talk with Y290/S921)	Shukalek 2016
Acetylatio n	On/Off Swtich	CACT; SLC25A20	K148, K157, K170, K244	Giangregorio 2017; Rauh 2013; Kim 2006; Palmieri 2015
		CIC; SLC25A1		Giangregor io 2018; Palmieri 2015
	Unknown	ABCD3		Lundby 2012
		ABCB7/8		Lundby 2012
		ABCA5	K614, K620, K1438	Lundby 2012
		BCRP; ABCG2	K171	Lundby 2012
Palmitoyla tion	Crosstalk with other PTMs	DAT; SLC6A3	rC580 with S7	Rastedt 2017; Moritz 2015; Foster 2017
	Lipid Raft Clustering	NCX; SLC8A1		Reilly 2015; Hilgemann 2013
	Trafficking to PM	CERP; ABCA1	C3, C23, C1110, C1111	Singaraja 2009
		ABCG1	C26, C250, C390, C402	Gu 2013
Ubiquitina tion	Degradation via Proteosome	ASBT; SLC10A2		Chen 2012; Chothe 2014
		OAT1/3; SLC22A6/8	K48 (OAT1)	Li 2013; Zhang 2013; Xu 2016a; Xu 2016b;
		SERT; SLC6A4		Mouri 2012; Mouri 2016
		EAAT2; SLC1A2		Zhang 2017;
		GLYT2; SLC6A5		de-Juan-Sanz 2011; de-Juan Sanz 2013
		DAT; SLC6A3		Sorkina 2006
		CERP; ABCA1		Hsieh 2014
		P-gp; ABCB1		Hsieh 2014; Pandzic 2017
	Degradation via Lysosome	CERP;ABCA1		Mizuno 2011
		MRP2; ABCC2		Aida 2013
	Endocytosis and Recylcing	BSEP; ABCB11		Aida 2013
SUMOylati on	Maintain Response of Intracellular Pools	EAAT2; SLC1A2	K580	Foran 2014; Foran 2011; Gibb 2007
		GLUT4; SLC2A4		Sadler 2013; Giorgino 2000;
				Silveirinha 2013;
		GLUT1; SLC2A1		Gioginio 2000
		NCX3; SLC8A3		Cuomo 2016; Lee 2009
	Surface Expression Modulation	MRP2;ABCC2	rK949	Minami 2008
		CFTR; ABCC7		Ahner 2016; Ahner 2013; Gong 2016

2. Phosphorylation

2.1. Background

Protein phosphorylation is a reversible PTM that affects an estimated one third of the total human proteome. Protein phosphorylation involves the addition of a terminal phosphate group to a free hydroxyl on the side chains of primarily serine and threonine residues, and to a lesser extent on tyrosine residues. The addition of the phosphate group is catalyzed by kinases, which are specific to the residue being modified (e.g. tyrosine kinase) and to the specific signal cascade initiated (e.g. the tyrosine kinase, Fyn). Protein phosphorylation is reversible, and protein phosphatases are integral to this reaction (7-9, 20)

2.2. SLC Transporters

2.2.1. Phosphorylation as Functional On/Off Switch

Phosphorylation can act as a functional regulator of SLC transporters. These proteins often have residues that can respond to a set of signals and cycle between a phosphorylated/dephosphorylated state to turn transport activity on or off. While often this type of mechanism is coupled to adjacent modifications, it also can have reciprocity with PTMs on another protein or transporter.

The divalent metal transporter (DMT1, also known as natural resistance-associated macrophage protein 2 (NRAMP2) or divalent cation transporter 2 (DCT2); SLC11A2) is an iron transporter in the intestine with numerous isoforms and subcellular distribution patterns that differ among tissues. It has been observed that DMT1 is upregulated during iron deficiency, and that the ubiquitination-proteasome pathway also influences its expression and degradation. DMT1 appears to be both functionally and basally phosphorylated at S43 and is inhibited by a wide range of cannabinoids, in stably expressed HEK293T cells. The classical cannabinoid Δ⁹-THC has anti-inflammatory effects which are modulated by the cannabinoid receptor type 2 and result in the dephosphorylation of the transporter protein. S43 of DMT1 thus appears to be vital to this regulatory mechanism (21).

The organic cation transporter (OCT2; SLC22A2) is also regulated by a PTM on/off switch, but via a tyrosine residue both in vitro and in vivo. OCT2 phosphorylation is mediated by the Src family kinase, Yes1, and Y362 was predicted to be the critical tyrosine (22, 23). Dephosphorylation of OCT2 diminished transport without altering membrane expression. Structural models suggest that the negative charge from the phosphotyrosine assists the binding of cations to the adjacent substrate-binding pocket (23, 24). Clinically, FDA approved tyrosine kinase inhibitors (TKIs), including dasatinib, reduced OCT2 function in both cells and in kidney tissue via Yes1 inhibition. This mechanism may serve a clinical benefit for patients by decreasing OCT2-mediated transport of platinum-based medications into the kidney thereby minimizing toxicity. However, TKIs also appear to have the potential to influence many transporters, such as the multidrug and toxin extrusion transporter 1 (MATE1; SLC47A1), organic anion transporter 1 (OAT1; SLC22A6), organic anion transporting polypeptide 1B1 (OATP1B1; SLCO1B1), and organic cation transporters 1 and 3 (OCT1/3 ; SLC22A1/3), in which the resulting clinical implications are unresolved (22, 24).

Additionally, the serotonin transporter (SERT; SLC6A4) (25, 26) and apical sodium-dependent bile acid transporter (ASBT; SLC10A2) (27) are tyrosine phosphorylated and regulated by Src family kinases, suggesting that TKIs may have some off-target risk (22-24, 28-30). Both ASBT and SERT have reduced pTyr status and reduced function during enteropathogenic E. coli (EPEC) infections (25-27). For ASBT, this mechanism of inhibition is attributed to the bacterial T3SS effector molecules which activate Tyrosine Specific Protein Phosphatases (PTPases) (27). For SERT, T3SS did not play a role in the dephosphorylation of the transporter, but rather resulted from an increased association with Src-homology-2 (SH2) domain containing PTPases (26).

Together, these studies demonstrate that a vast array of divergent signals ranging from the indirect action of endogenous receptors, pharmacological agents, to bacterial virulence factors have an influence on SLC transporter PTMs and result in the regulation of transporter functional expression.

2.2.2. Regulation of Phosphorylation by Protein Kinase C (PKC) Influences the Subcellular Fate of SLCs

Most of the literature regarding SLC post-translational regulation is related to the promiscuous role of PKC in the endocytosis and trafficking of internalized transporters. Additionally, PKC is often the initiator of signaling cascades that alter many other types of PTMs, some of which are mentioned in subsequent sections of this review. Often, PKC-induced internalization involves the ubiquitination of the target SLC transporters residing within lipid rafts, leading to the internalization of the membrane. The combined signals from multiple residues also have been implicated in transporter sorting mechanism. Transporters observed to be regulated in this manner include the sodium- and chloride-dependent glycine transporter (GLYT2; SLC6A5) (31, 32), dopamine transporter (DAT; SLC6A3) (33-35), organic anion transporter 1 (OAT1; SLC22A6) (36-38), organic anion transporting polypeptide 1B1 (OATP1B1; SLCO1B1) (39), and glucose transporters 1 and 4 (GLUT1/4; SLC2A1/4) (40, 41), among many others. The role of PKC in the regulation of drug transporters has been reviewed in great detail by Fardel and colleagues (42).

2.3. ABC Transporters

2.3.1. Phosphorylation regulates ABC-mediated drug resistance

Multidrug resistance (MDR) is a major concern in the overall efficacy of a wide range of pharmacological agents. The ABC transporters, P-glycoprotein (P-gp; ABCB1), Multidrug Resistance-associated Protein 1, (MRP1; ABCC1) and Breast Cancer Resistance Protein (BCRP; ABCG2), are the main contributors to-date in regards to MDR. They have broad tissue expression patterns and are robustly expressed in cancer cells. MRP1 has two linker domains (L0 and L1) and within these disordered regions reside numerous PTM sites. T249 was shown to be a phosphorylation site contributing to transporter function. Inhibition of Casein kinase 2α (CK2α) led to dephosphorylation of the transporter and decreased efflux of doxorubicin and estradiol 17β-D-glucuronide, unrelated to changes in protein stability or localization. Furthermore, CK2α was co-immunoprecipitated with MRP1 and in vitro kinase assays suggested that the kinase directly phosphorylates the transporter (43). In addition to threonine phosphorylation, the in vitro phosphorylation statuses of serines in L1 were shown by Ambadipudi and Georges (2017) to modulate MRP1 association with tubulin, whereby substitution at S871, S915, S961 with phospho-mimetic aspartic acid blocked the association (44). Further characterization is warranted to determine the kinases and phosphatases obligatory to this interaction, as well as to determine if tubulin association alters the surface expression of the transporter.

Interestingly, P-gp has been shown to bind tubulins with a high affinity to two regions in the linker-domain enriched with serines, yet it is unknown whether serine phosphorylation status is vital to the interaction (45). It is worth noting that within this region, S661 is phosphorylated by PKC (46-50), and S683 by PKA. S667 and S671 are phosphorylated by both kinases (50). In addition, S683 may be phosphorylated by the serine/threonine kinase Pim-1, which appears to selectively phosphorylate complex glycosylated P-gp to maintain surface expression (50). Together, multiple signaling pathways appear to modulate the same residues in P-gp, dynamically altering the fate of the transporter. It is worth noting that BCRP is also phosphorylated by Pim-1, however, phosphorylation at T362 does not alter the glycosylation status, but rather promotes the multimerization and surface translocation of the transporter (51).

2.3.2. Phosphorylation Influences ABC-transporter gating mechanisms

The cystic fibrosis transmembrane conductance regulator (CFTR; ABCC7) is an atypical transporter, which functions as a chloride channel. Phosphorylation is an influential regulator in the transport gating mechanisms of CFTR. When the regulatory domain is non-phosphorylated, it has been shown to block the dimerization of the two nucleotide binding domains (NBDs). The resulting structure is conformationally distinct from the phosphorylated transporter, and as such, the NBD can act as a gate to the translocation channel (52). Protein kinase A has previously been show to phosphorylate S660, S737, S795, and S813, whereby modification at any of the four sites maintains Cl⁻ conductance (53). The four phosphorylation sites appear to be degenerate: no one site is essential for channel activity, and phosphorylation at one site alone is sufficient for regulation of Cl⁻ channel activity. In this context, it suggests that CFTR’s channel-like behavior is in part due to post-translationally mediated conformational changes.

3. N-glycosylation

3.1. Background

More than half of the mammalian proteome and nearly all SLC transporters are glycosylated, with N-glycosylation predominating. N-glycosylation is the process of enzymatically adding an oligosaccharide to extracellular asparagine residues and is often signaled by the motif N-X-S/T, where X is any amino acid except proline (P) or aspartic acid (D). N-glycosylation is a highly variable PTM, which is impacted by the availability of the sugar monomers at the time of protein folding and processing. Addition of the first sugar chain to an asparagine residue occurs within the endoplasmic reticulum (ER) from the coordination of a preassembled substrate scaffold and an oligosaccharyltransferase. The chain undergoes subsequent trimming while in the ER by exoglycosidases. Properly folded proteins are transported to the golgi apparatus, where the remaining glycan chain is trimmed down to the mannose-rich core and rebuilt by resident glycosyltransferases. The resulting product falls into three broad categories: mannose-rich core glycosylated, complex glycosylated, and hybrid glycosylated proteins (7). While highly studied in vitro, the complex interactions as a consequence of glycosylation regulation are lacking. Critically, biochemical assays often fail to capture the structural diversity of glycan chains resulting from the random interactions during the glycan processing stages.

Pedersen and colleagues recently extensively reviewed the role of glycosylation in modulating SLC transporters and highlighted the limitation of over-interpreting glycan data between the various cellular contexts. Among these variables are observable glycosylation differences between species, organs, carbohydrate accessibility, metabolism, and more (54).

3.2. N-Glycosylation as a Core PTM of SLC Transporters

3.2.1. N-glycosylation May Protect SLCs from Proteases

The peptide transporter 1 (PEPT1; SLC15A1) is responsible for the uptake of di- and tripeptides and has a wide range of apparent molecular weights in various organs and species. This variability is attributed to differences in the glycosylation pattern (55-58). In intestinal tissues, murine PEPT1 (mPept1) was demonstrated to be N-glycosylated by predominantly complex-type glycans, which contributed to nearly 1/3 of the protein’s overall apparent molecular weight of ~95 kDa. In contrast to the small intestine, colonic mPept1 was observed to be ~105 kDa, whereas kidney mPept1 was ~75 kDa (56, 59). Using site-directed mutagenesis and protein expression in Xenopus laevis oocytes, six putative glycosylation sites were kinetically characterized, and N50 was identified as a critical residue for the efficiency of peptide transport. N50 lies close to the membrane surface in extracellular loop 1 (EL1) and when glycosylated is hypothesized to have a negative steric effect on both the substrate binding-site and on the transport mechanism. Therefore, it is expected that deglycosylation at this residue, or alterations in the glycosylation pattern, will positively modulate PEPT1 function (56). However, further investigation suggested that the fundamental role of N-glycosylation at N50 is to protect the extracellular loops from protease degradation in oocytes, and further studies are needed to determine whether this mechanism is consistent in other expression models (57). This finding may mechanistically explain the variability in function between species and tissues (57).

3.2.2. N-Glycosylation of SLCs Change in Disease

Non-alcohol fatty liver disease (NAFLD) is a disease characterized by fat accumulation in the liver. Aggravated cases of NAFLD often lead to the development of non-alcoholic steatohepatitis (NASH), typified by excess fat, inflammation, and liver damage. These conditions have been shown to not only cause altered metabolism in patients (60), but also shown to alter the glycosylation status of SLC transporters in the liver (61). Gene array data and subsequent western blot analysis of human liver donors indicate that the expression of non-glycosylated organic anion transporting polypeptides, OATP1B1/3 (SLCO1B1/3), OATP2B1 (SLCO2B1), and non-glycosylated Na⁺-taurocholate co-transporting polypeptide (NTCP; SLC10A1) was increased in NASH as a result of altered gene expression of the necessary enzymes in the ER and Golgi (60). Interestingly, NTCP mRNA is upregulated in NASH, yet down regulated in morbidly obese NASH patients, whereas in both patient subsets, transporter protein levels were diminished. Together, this suggests that non-glycosylated NTCP may be destined to the ER for degradation, and that there may be confounding transcriptional mechanisms with comorbid pathologies. In support of this hypothesis, a recent study found that NTCP deglycosylation, or mutagenesis at N5 and N11, promoted rapid endocytosis and degradation in the lysosome (62). Thus, it would be interesting to determine if there is a difference in handling in the ER/golgi of the mutated protein that facilitates its membrane targeting in lieu of a glycan chain, or if the non-glycosylated NTCP in NASH patients trafficks to the membrane.

3.2.3. N-Glycosylation May Regulate Microdomain Partitioning of SLC transporters

An intriguing feature of some multi-antennary N-glycans (complex) is the enrichment of Gal-Glc-NAc moieties on the terminal ends, which serve as recognition sites for a class of lectins named galectins. This weak interaction creates a structural lattice between the glycan and the lectin that alters the lateral movement of surface proteins. This lattice formation occurs between the glycan of the facilitative glucose transporter 2 (GLUT2; SLC2A2) and galectin9, anchoring the transporter within non-lipid raft domains in the pancreas in order to facilitate glucose uptake. In pancreatic beta cells, altered glycosylation or disruption of the glycan lattice on GLUT2 was shown to lead to a shift in the transporter into lipid raft domain where GLUT2 function is significantly diminished. It is hypothesized that this mechanism is integral for glucose sensing and insulin secretion. In support, a high-fat diet alters the glycan pattern of GLUT2, potentially implicating this glycan-galectin lattice interaction in the development of type 2 diabetes (63).

3.3. ABC Transporters

3.3.1. Glycosylation influences the stability of Surface expressed ABC transporters

As previously highlighted, mature P-gp glycoform is selectively modulated via phosphorylation by Pim-1. It appears that this mechanism may enable the complete glycosylation of immature glycoforms of P-gp and promote surface translocation. While the phospho-signal appears to prevent the degradation of the immature glycoform, N-glycosylation appears to be important for stability of the protein in the membrane. Xie and co-workers (50) hypothesized that Pim-1 inhibitors may improve drug resistance by preventing the maturation of P-gp’s glycan chain and promoting the degradation of newly synthesized P-gp protein.

3.3.2. Glycoform-specific implications to ABC-transport mechanism

Multidrug resistance-associated protein 4 (MRP4; ABCC4) has two N-glycosylation sites in the first extracellular loop of the second multi-spanning domain (MSD). While in close proximity to one another, N754 and N746 have opposing roles in regulating the transport of MRP4. Specifically, N746 appears to inhibit transport of prostaglandin E₂, as mutation at the site increased transport. On the other hand, N754 mutation decreased function suggesting that this specific glycan site promotes transport. Intriguingly, mutations did not affect the transport of estradiol glucuronide, suggesting that there may be conformational changes resulting from glycosylation that does not alter the binding mechanism of all substrates in a similar fashion (64).

The multidrug resistance-associated protein 1 (MRP1; ABCC1) also is modulated by glycosylation, which appears to influence the phosphorylation-status of the transporter, such that it has the capacity to shift from a high capacity/low affinity transporter to a low capacity/high affinity one. MRP1 has three glycosylated sites at N19, N23, and N1006. N19 and N23 are found in the first MSD and dual modification at these residues promotes higher affinity transport. When the glycoprotein was also phosphorylated at T920 and S921 within the linker region between NBD1 and MSD2, transport shifted to higher capacity (64). Together, studies with MRPs suggest that N-glycosylation can influence the binding affinity and overall transport mechanisms, yet further studies are required to determine if this regulatory mechanism is a result of small conformational shifts or via more complex mechanisms such as PTM crosstalk.

4. Acetylation

4.1. Background

Lysine acetylation (Table 2) is a reversible modification that occurs on the ε-amino group of intracellular lysines. It has been implicated as a dynamic post-translational response to changes in metabolism, and modified sites may be conserved between species. Lundby and colleagues (65) recently carried out an extensive proteomic analysis to define tissue and cellular acetylation patterns. Intriguingly, they found over 4,000 unique acetylated proteins in 16 different rat organs. Approximately 15% of the identified proteins were mitochondrial or plasma membrane associated, with transporter proteins being the highest fraction of acetylated proteins in the latter pool. Additionally, the highest percentage of acetylated proteins in the brain was membrane resident proteins. Together, this indirectly suggests that the potential for a transporter to be modified will be dependent on its tissue expression pattern and subcellular localization (65).

4.2. SLC Acetylation as an On/Off Switch for Mitochondrial SLC Transporters

Within the mitochondrial proteome, acetylation has been identified as a common PTM occurring both enzymatically and non-enzymatically. Non-enzymatic acetylation occurs in the presence of high concentrations of acetyl-CoA (7). In the liver mitochondria, the carnitine-acylcarnitine transporter (CACT; SLC25A20) and the citrate transporter (CIC; SLC25A21) have opposing roles in the β-oxidation pathway feedback mechanism (66). For both transporters, acetylation of intracompartment-facing lysine residues functions as an on/off switch for the transporter (66, 67). CACT (SLC25A20) is inhibited by acetylation (66), and CIC (SLC25A21) is activated by acetylation (67). For CACT (SLC25A20), putative lysine residues were first identified with mass spectrometry (68, 69) and coupled with homology modeling and functional studies; K148, K157, K170, and K244 were identified as the likely modification sites (66).

4.3. ABC transporter acetylation is a relatively unexplored PTM

The proteomics data by Lundby and co-workers suggest multiple ABC transporters may be acetylated; yet the functional significance is unknown for most. Two unique peptides with three acetylated residues were attributed to ABCA5 at K614, K620, and K1438. The mitochondrial transporters ABCB7, ABCB8, and peroxisomal transporter ABCD3, were also shown to be acetylated along with some uncharacterized proteins. In addition, BCRP was modified at K171, a residue that was conserved in the human sequence and located in the first multi-spanning domain (65).

5. Lipidic Modifications

5.1. Background

Palmitoylation involves the enzymatic addition of the C16 saturated fatty acid, palmitic acid, to a free cysteine that lies adjacent to the protein-lipid interface (13, 14). Palmitoylation differs greatly from the relatively static nature of similar lipid modifications, such as myristoylation, in that it is highly dynamic. It has been observed that the reversal of protein palmitoylation can occur at variable rates from minutes to days depending on the protein target (70, 71).

5.2. SLC Transporters

5.2.1. Palmitoylation -Phosphorylation Cross-talk Regulates SLC Transporters

The palmitoylation status of the dopamine transporter (DAT; SLC6A3) is reciprocally regulated by phosphorylation. Rat DAT (rDat) is palmitoylated at C580 (and potentially at a secondary location) and has been shown to increase rDat transport efficiency without modulating rDAT surface expression (72). The increase in transport efficiency appears to be mechanistically linked to the inverse decrease in V_max resulting from PKC-mediated phosphorylation at S7 (72-74). Interestingly, pS7 is enriched in lipid rafts (73, 74), whereas the enrichment of palmitoylated DAT remains unresolved, but may suggest that palmitoylation is an insufficient signal for the lateral partitioning of DAT (75). Additionally, long-term inhibition of DAT palmitoylation drives the subsequent degradation of phosphorylated DAT, suggesting that there is a time-dependent component for the reciprocal regulation of the transporter. This also points to the greater role of palmitoylation in modulating transporter stability and expression levels (72).

5.2.2. Palmitoylation Potentiates SLC Clustering in Lipid Rafts

In contrast to the role of palmitoylation in modulating SLC stability, the electrogenic Na⁺/Ca²⁺ exchanger 1 (NCX1; SLC8A1) is functionally unaffected by its palmitoylation status directly. In ventricle muscles, 60% of the total pool of NCX1 is palmitoylated, and only during metabolic stress does transport become affected along with the depletion of intracellular Ca²⁺. Transport inactivation occurs with the palmitoylated transporter, and results in transporter clustering within lipid raft domains. This crowding exacerbates the innate curvature of the microdomain and triggers ‘Massive Endocytosis’ (MEND) (76, 77). MEND is a physiological phenomenon in which up to 70% of the plasma membrane is endocytosed during a single event (76). PKC appears to be activated during this process, but whether this is related to the previously discussed mechanism of DAT modulation is unknown.

Curiously, detergents encourage the clustering of palmitoylated proteins in lipid rafts, potentiating the MEND event (78). Along those lines, bile acids have been shown to solubilize non-raft domains thus leading to a stabilization of lipid rafts and of raft-resident proteins (79). It is worth noting that clustering and enrichment of SLC transporters within lipid rafts has been observed as referenced herein, suggesting that studies are warranted to determine if palmitoylation serves as a protective mechanism for SLC transporters exposed to high concentrations of bile acids. If implicated, it may suggest a novel on/off mechanism for transporters lining the intestinal tract pre- and post-prandially whereby a) palmitoylated SLC transporters selectively partition to lipid rafts; b) bile acids stabilize lipid rafts; c) lipid raft proteins cluster to increase signaling and/or transporter efficiency; and d) SLC-mediated transport is terminated via MEND resulting in membrane and transporter turnover.

5.3. ABC transporters

There are numerous signaling mechanisms for the effective targeting of transporters to the plasma membrane, including lipid modifications like palmitoylation. For the lipid transporter, cholesterol efflux regulatory protein (CERP; ABCA1), its palmitoylation status is vital for its rapid (de)-association in the plasma membrane, and consequently for its functional status. Four cysteine residues total are palmitoylated; two in the N-terminal region – C3 and C23– and two in the linker region –C1110 and C1111. Removal of the lipid modification at any of the residues decreased the efflux of phospholipids and cholesterol by ~50% and decreased CERP surface expression by ~90%. This suggests that appropriate targeting to the membrane requires palmitoylation at all four residues. Aside from its efflux activity, CERP is also vital for the lipidation of apolipoprotein A1 (Apo-A1). Curiously, a loss of palmitoylation status did not completely abrogate lipid efflux from the cells, and some studies suggest that CERP may lipidate Apo-A1 from intracellular pools (80).

In concert with CERP, the ABCG family of transporters also is involved in cholesterol transport, yet appears to be affected differently by palmitoylation. ABCG1 is highly expressed in macrophages alongside of CERP and is implicated in the efflux of cholesterol onto lipidated lipoproteins, serving a complicated role in the progression of atherosclerosis and foam cell formation (81). ABCG1 is palmitoylated in its N-terminal domain at C26, C150, C311, C390, and C402. However, unlike ABCA1, where multiple palmitoylation residues are critical to expression, only C311 appears to have a functional role for ABCG1. Depalmitoylation at C311 reduced cholesterol efflux by altering trafficking of the transporter from the ER to the plasma membrane. In contrast to the cholesterol transporter (ABCG1), other ABCG family members are differentially regulated by palmitoylation. ABCG5, while it has been shown to be palmitoylated, does not appear to be functionally impacted by depalmitoylation. ABCG8 on the other hand, is not palmitoylated (81).

6. Ubiquitination

6.1. Background

The ubiquitin-proteosome pathway is a physiological system for the signaling and subsequent degradation of unstable proteins via the proteasome. Ubiquitin (Ub), a small, 8 kDa polypeptide, is covalently linked to accessible lysine residues via a three-step process. Initiation of the reaction is carried out by a ubiquitin-activating enzyme (E1). Following activation, Ub structurally associates with a carrier protein (E2) which, coupled with a ubiquitin-protein-ligase (E3), modifies the target protein. While this mechanism requires an available lysine residue, most ubiquitinated proteins are conjugated to multiple Ub polypeptides driven by the presence of seven lysine residues on Ub itself. Therefore, a target protein can be mono-, multi-, or poly-ubiquitinated (82-85).

6.2. SLC Transporters

6.2.1. Ubiquitin as a Degradative Signal for SLC Transporters

It has been observed that inflammation within the ileum signals a phosphorylation-dependent activation of the ubiquitin-proteosome pathway. This degradative pathway is vital for short-lived proteins like the apical sodium-dependent bile acid transporter (ASBT; SLC10A2). Rat Asbt (rAsbt) requires JNK-mediated phosphorylation at S335 (pS335) and T339 (pT339) to be ubiquitinated and consequently degraded (82). Contrary to this finding, our laboratory found that ubiquitination of rAsbt occurred in lieu of pS335 and pT339 following treatment with resveratrol (RSV) (86). RSV, a phytoalexin found in red grapes, has been extensively studied for a wide range of promising health benefits. Among its many attributes, it is well known to have anti-inflammatory properties that have been characterized in animal models of gastrointestinal inflammation. This contrast between the JNK-mediated and phospho-independent modulation of rAsbt ubiquitination suggests that that there are diverse signaling mechanisms that target transporters to the same degradative fate (87-91).

In line with this hypothesis is the existence of two predominant families of E3 ubiquitination ligases in mammals: HECT and RING. Both families have been structurally resolved and shown to have different binding mechanisms to Ub potentially explaining how ubiquitination may occur via alternative signaling events (92). Many SLC transporters are ubiquitinated by HECT ligases, and specifically by the subfamily Nedd4. Nedd4 ligases contain domains with a conserved pair of tryptophans that recognize the structural motif (L/P)PxY, and which are implicated in PKC-induced ubiquitination mechanisms. Organic anion transporters 1 and 3 (OAT1/3; SLC22A6/8) both are signaled for Nedd4-2 ubiquitination by PKC, resulting in the internalization of the transporters and a consequential decrease in their function (36, 38, 85, 93, 94).

6.2.2. Ubiquitination Status of SLC Transporters as a Clinical Biomarker

Adaptor proteins and their relative expression can influence the ubiquitination of SLC transporters. For instance, the serotonin transporter (SERT; SLC6A4) binds to a protein called MAGE-D1, facilitating the modification of SERT with Ub. Mechanistically, a decrease in MAGE-D1 expression, such as is the case in manic depressive disorder (MDD), leads to a decrease in ubiquitinated SERT and a subsequent increase in SERT membrane expression, thus promoting hyposerotinergic conditions (95-97). While, the administration of sertaline and imipramine improved the depressive behavior in MAGE-D1 knockout mice, it did not modulate the increased stability of SERT (96). Additionally, fluvoxamine-resistant MDD patients had a global decrease in the Ub-status of SERT. This resulted in the upregulation of SERT in leukocytes and lymphoblasts signifying their potential as a novel biomarker for the efficacy of MDD therapies (96, 97).

6.2.3. Modulating the Ubiquitination Status of SLC Transporters Therapeutically

Therapeutically, knock-down or inhibition of E3 ligases has been proposed as a means to modulate the excitatory amino acid transporter 2 (EAAT2; SLC1A2) to ameliorate Parkinson’s Disease (PD) after it was observed that EAAT2 was modified by Ub through cooperation with Caveolin-1 in the PD mouse model (98). Complicating that finding, it has been shown that increased Ub expression has a dose-dependent effect on motor function, and that the availability of free Ub is the rate-limiting step in the ubiquitination-dependent degradation of proteins. Modest increases in the free Ub pool, restored the balance and improved motor function in wild-type mice, but as the availability increased, motor function declined (99). As numerous SLC transporters like SERT (SLC6A4) (96, 97), the glycine transporter GLYT2 (SLC6A5) (31, 32), and the dopamine transporter DAT (SLC6A3) (100) are ubiquitinated within the brain, the overall homeostasis between ubiquitination and deubiquitination may be fundamental to restoring balance in neurodegenerative disorders, suggesting that deubiquitinating enzymes may be potential therapeutic targets.

6.3. ABC Transporters

6.3.1. The Ubiquitin-Proteosome Pathway Regulates the fate of ABC Transporters

Similar to SLC transporters, ubiquitination is often an obligatory step for the degradative fate of internalized ABC transporters. In vitro, in cancer cell lines, P-gp degradation is promoted by inhibition of MAPK pathways and regulated by the ubiquitin-proteasome pathway leading to enhanced resistance to chemotherapeutics. Coupled to traditional molecular biology techniques, MALDI-TOF mass spectrometry analysis led to the identification of an E2 ligase and E3 protein complex that promotes the ubiquitination of P-gp in cancer cell lines. Immunopurification of a C-terminal dual epitope labeled transporter, followed by trypsin digestion and MALDI-TOF analysis resulted in the identification of approximately two dozen candidate proteins. Subsequent immunoprecipitation studies suggested that the protein, FBX015, interacted with P-gp as an E3 ligase component. Further immunoprecipitation studies, revealed that Ube2r1 is the E2 ligase which is involved with the ubiquitination of P-gp in cancer cell lines. Knockout of either Ube2r1 or FBX015 resulted in the decrease of ubiquitinated P-gp and an increase in P-gp surface expression, further implicating their role in modulating P-gp functional expression. It is worth noting that the E2 ligase, which interacts with P-gp indirectly through association with the E3 ligase, was not immunopurified and detected with MALDI-TOF analysis. The investigators hypothesized that this was a result of the C-terminal peptide being of a similar weight as Ube2r1, yet it likely is due to a combination of factors including the purification strategy, the presence of a dual epitope label in the C-terminal region, as well as the digestion and analytical conditions (101).

6.3.2. The Ubiquitin-mediated fate of ABC Transporters Can Follow Many Cellular Paths

While the ubiquitin-mediated degradative pathway traditionally is considered to involve the proteasome, some transporters appear to utilize Ub as a sorting signal from within endosomal compartments and/or as a signal for lysosomal degradation. Specifically, for the lipid and cholesterol transporter, CERP, degradation was mediated via the lysosome and is reliant on a group of proteins broadly grouped together as ‘Endosomal Sorting Complex Required for Transport’ (ESCRT), proteins. In HuH-7 and Hela cells expressing CERP, the transporter was ubiquitinated at the cell surface. The ubiquitination status determined the sorting of the modified protein to endosomal compartments. Specifically, Ub-CERP directly associated with HRS, an ESCRT-0 protein, which acts as the first step to recognize and promote endosomal association (102).

Subsequent studies suggested that the bile salt export pump (BSEP; ABCB11) might also be degraded through the ESCRT-lysosomal pathway. At the canicular membrane, BSEP and the multidrug resistance-associated protein 2 (MRP2, ABCC2) are efflux transporters for both endogenous and exogenous substrates from the liver. Both transporters have been shown to be degraded in a ubiquitin-dependent manner, albeit via divergent pathways. Using surface biotinylation and a mutant Ub construct lacking two critical glycine residues that are necessary to bind trafficking protein complexes, Aida colleagues demonstrated that the internalization of BSEP followed ubiquitination and a clathrin-mediated pathway. While the exact mechanism and machinery for the internalization pathway was not discerned in its entirety, the investigators concluded that Ub was likely the initial signal for internalization into the endosome, rather than the signal for degradation for BSEP. Conversely, the same investigators found that MRP2 is internalized through a dynamin-dependent process to early endosomes irrespective of the Ub status of the transporter. Within the endosomal compartment however, ubiquitin serves as the sorting signal for the lysosomal degradation (103).

6.3.3. Ubiquitination may be a Separate Event from the degradative fate of ABC transporters

While the ESCRT pathway provides one distinct avenue for the Ub signaled fate of transporters, the mechanism driving the initial modification of lysine by Ub adds an extra layer of complexity. For CERP, cholesterol not only serves as a substrate and a regulator of the fluidity of the membrane environment, but also appears to modulate the transporter post-translationally. High cholesterol concentrations were shown to promote CERP degradation through the ESCRT-lysosomal pathway by disrupting a protein-protein interaction between CERP and the nuclear receptor isoform, LXRβ, at the cell surface. The disruption resulted in a less stable cell surface protein that was susceptible to modification by the Ub by E3 ligases (104).

Curiously, ABCG1, a transporter often working in synergy with CERP is stabilized and protected from Ub by cholesterol binding. Its degradation occurs via the proteasome and mutation at Y667, a vital residue within a cholesterol recognition region known as a CRAC/CARC motif, exacerbated the degradative fate. Through in silico modeling, the investigators determined that Y667 contributed sterically to the binding of the substrate, cholesterol (105). Mechanistically however, how this contributes to the susceptibility of the apo-transporter to be post-translationally modified remains to be resolved. As many transporters contain numerous CRAC/CARC motifs, the divergent effects of cholesterol as both a regulator of ubiquitination and as a putative modification itself suggests that there is a layer of complexity in the regulation of transporters that remains to be elucidated.

7. SUMOylation

7.1. SUMOylation as a Gatekeeper of SLC Transporter Intracellular Pools

The expression of the EAAT2 (SLC1A2) is not just influenced by ubiquitination, but also is influenced by the closely related PTM, SUMOylation (100, 106-108). Like ubiquitin, SUMO (Small Ubiquitin-like MOdifier) is a reversible modification that involves the enzymatic addition of a small protein SUMO to an accessible lysine (109). EAAT2 is modified at K580 under normal physiological conditions causing internalization of the transporter. Unlike Ub, modification by SUMO does not cause degradation, but rather facilitates the formation of an intracellular holding pool that is available for the rapid response to central nervous system signaling. Curiously, while this process appears vital in the context of normal physiology, SUMO-EAAT2 is cleaved by caspase3 in a mice model of amyotrophic lateral sclerosis (ALS, Lou Gehrig’s disease), leading to the accumulation of the cleaved product in the spinal cord (106-108).

In a similar fashion, SUMOylation acts as a sorting mechanism for the facilitative glucose transporter 4 (GLUT4; SLC2A4). While GLUT4 is minimally expressed on the plasma membrane of adipocytes, under normal conditions and in response to insulin signaling GLUT4 readily shuttles to the surface and increases glucose transport significantly. This dynamic response results from storage of GLUT4 in specific vesicles that have been shown to increase the half-life of the transporter considerably. SUMOylation is thought to be the sorting signal to these specialized vesicles (110). Additionally, a proteomic analysis was carried out to derive the global SUMOylation changes during hypoxia (111). This study identified numerous metabolic associated proteins, including the closely related glucose transporter, GLUT1 (SLC2A1), which had increased SUMOylation, and experimental evidence suggests that modification leads to a decrease in levels of GLUT1 surface expression, in contrast to GLUT4 (112-114). It was hypothesized that perhaps this may be a mechanism for a cell to preferentially express GLUT4 to potentiate insulin responsiveness (114). Hypoxia-driven increases in SUMOylation might serve as a mechanism to increase stability of the transporter, increase glucose sequestering in the brain, and minimize ischemia-related consequences (112). SUMO1’s neuroprotective effects also translate to NCX3 (SLC8A3) in the ipsilateral temporoparietal cortex (115). Ischemia preconditioning leads to an increased SUMOylated NCX3 and silenced the resident excitatory neurons, preventing appreciable damage to this region (115, 116).

7.2. ABC transporters

7.2.1. The fate of SUMOylation influences the surface expression of ABC transporters

The linker region of the multidrug resistance-associated protein (MRP2; ABCC2), found between the nuclear binding domain 1 and the last membrane spanning domain region, contains numerous putative PTM binding sites, including a ΨKXE motif (where Ψ is any hydrophobic amino acid residue, X any residue). This lysine motif is obligatory for Ubc9 (a ubiquitin conjugating enzyme) binding and the subsequent conjugation of SUMO to MRP2. Mutation of the critical lysine residue within this consensus motif abolished SUMO binding and altered the expression of MRP2. The decrease in expression however, did not appear to be due to altered protein stability, and the mechanism for the decrease in expression remains to be resolved. It also is not clear as to whether the expression of the surface resident transporter is altered with de-SUMOylation, limiting mechanistic insight into the role of SUMOylation in regulating MRP2 (117).

Mammalian SUMOylation can occur via any of the multiple SUMO paralogs at the consensus motif ΨKXD/E (118) however there are differences in the signaled fate of the modified transporter. CFTR (ABCC7) is another SUMOylated ABC transporters and has been discussed by Ahner and colleagues in the context of SUMO paralogs (119-121). CFTR has been shown to have three putative sites located in the NBD regions and the C-terminal region in the wild-type protein as well as in the clinically relevant ΔF508 mutant, commonly found in cystic fibrosis patients (120). Between the two CFTR forms, there appears to be divergence in the SUMOylation patterns. The ΔF508 isoform binds SUMO-1 and SUMO-2/3, of which binding of SUMO-2 to NBD1 appears to be enhanced by HSP70, and modification promoted subsequent ubiquitination and proteasomal degradation, thereby providing a mechanism by which misfolded ΔF508-CFTR is rapidly degraded upon translation, leading to the clinical manifestation of cystic fibrosis. HSP70 did not have an effect on SUMOlyating wild-type CFTR in vitro (120, 121).

8. Conclusions

While the modulation of signaling mechanisms is readily recognized as a critical regulatory facet for the functional integrity of homeostatic processes, there is still limited information on how they influence transporter proteins. ABC and SLC transporters are vital to the efficient tissue handling of endogenous and exogenous solutes. The two families combined represent over 500 functional proteins that are integral in almost every single homeostatic process in the human body. The two families have general structural differences. ABC transporters have multispanning transmembrane domains (TMD) connected by nuclear binding domains and linker regions, and SLC transporters do not have NBDs but contain intracellular loops bridging one TMD to another.

PTMs have been found in all of the described regions, although there likely is a mechanistic preference for a given modification to occur in one region versus another. Phosphorylation for instance, appears to heavily crosstalk with other PTMs, and the influence and relative magnitude of the intended fate is often determined by the specific modification location. This is complicated by the fact that a single residue can be modified by different modifications. Lysine residues for instance can be acetylated, SUMOylated, or ubiquitinated. Additionally, in the case of phosphorylation, a single residue can be modified by numerous different kinases and phosphatases, yet the consequence of the modification may be entirely different. We do not understand the general mechanisms for how the ultimate fate of a signaling event is determined; regardless, this again highlights the likely importance of PTM crosstalk in transporter regulation.

The fate of the transporter following a post-translational event also does not always lead to an extreme consequence, such as degradation. There is significant data for a variety of modifications involving transient effects. GLUT4 for instance is (de)-SUMOylated to rapidly respond to insulin signals and translocate from intracellular pools to the cell surface. This occurs rapidly to adapt to the metabolic needs of the system. Additionally, post-translational signals can dictate the sub-membrane fate and trafficking of many transporters. This is critical in interpreting transporter expression, such that surface expression does not always equate to functional expression. An example of this is with the Na⁺/Ca2⁺ exchanger 1, NCX1 (SLC8A1), where palmitoylation is capable of shifting the SLC transporter into lipid raft domains. Similar observations have been seen with PKC-signaled fates for SLC transporters, and N-glycosylation of GLUT2.

It important to recognize that there also seems to be a general bias in the literature for studying PTMs of transporters. While we tried to choose a diverse set of transporters for this review, generally from our observations a significant amount of the literature appears to focus on specific SLC neurotransporters, such as DAT and the ABC transporter, CFTR. While this yields impressive mechanistic insight into those particular transporters, there is limited information still to extrapolate roles for a specific PTM to a general transport regulatory mechanism. The extent in which PTMs regulate transporters is far from being fully appreciated, yet it cannot be expected that every proteoform will have a unique and clinically relevant phenotype. However, those that do must be considered not only in the context of human health but also in the context of the safety and efficacy of therapeutic interventions. While transporters play significant roles in the pharmacokinetics and pharmacodynamics of drugs, the capacity of drugs to modulate off-target signaling pathways is typically only recognized during development. The indirect implications of a pharmacological agent on the endogenous regulation of transporters may be significant in characterizing off-target effects. This is exemplified by the use of kinase inhibitors clinically. In the case of MDR, kinase inhibitors provide promise for increasing the availability of a therapeutic in a target tissue, yet we do not fully understand how this impacts transporters for endogenous substrates. In the case of OCTs and OATs, it suggests that these inhibitors affect transporters at a cellular level, but what the consequences are in vivo is less certain. Despite the need for understanding the risks, as more information accumulates regarding the consequence of modulating PTMs, transporter proteoforms may prove to be viable targets and/or biomarkers for a profound range of clinical manifestations.

Glossary

Abbreviations:

ABC

ATP-binding cassette

SLC

solute carrier

PTM

post-translational modification

NBD

nucleotide binding domain

PKC

protein kinase C