ABC Transporters’ Interactome: Mapping the Human ABCG2 & ABCG1 Protein-Protein
Interactions
Kristen Lee
Institute of Parasitology
McGill University
Montréal, Québec
April 2021
A thesis submitted to McGill University in partial fulfillment of the requirements of the degree
of Master of Science
©Kristen Lee, 2021 TABLE OF CONTENTS
ABSTRACT ...... 2
ABRÉGÉ...... 2
ACKNOWLEDGEMENTS ...... 4
CONTRIBUTIONS TO ORIGINAL KNOWLEDGE ...... 6
CONTRIBUTION OF AUTHORS...... 6
LIST OF ABBREVIATIONS ...... 7
INTRODUCTION ...... 9
METHODOLOGY ...... 38
RESULTS ...... 46
DISCUSSION ...... 72
SUMMARY AND CONCLUSION ...... 81
REFERENCES ...... 84
Page 1 of 110
ABSTRACT
ABC transporters play essential roles of detoxification and selective transport of substances in humans. ABCG2 is a half-transporter mainly expressed at the plasma membrane in many healthy tissues including tissue barriers and the gastrointestinal tract. Determining protein-protein interactions of ABCG2 can elucidate a concrete mechanism of regulation (e.g., activation and inhibition) in diseased and healthy states. Moreover, differential tissue regulation of ABCG2 can fully exploit ABCG2’s native roles and broad substrate specificity. Mutations and alterations of expression have been linked to pathologies such as Alzheimer’s, diabetes, cancer, genetic disorders and many infections. Understanding signaling pathways through protein interactions can uncover new therapeutics and drug targets against such ailments. This project uses a proximity-dependent biotinylation approach (BirA) to identify two types of ABCG2 interacting proteins: direct interactors and those within a 10Å sphere. ABCG2 fused with BirA at its N-terminus was stably transfected into HEK293F (an embryonic cell line) and several clones were functionally characterized. Biotinylated proteins, using the latter HEK-BirA-huABCG2 transfectants, identified by mass spectrometry were compared to those interacting with ABCG1 (using HEK-
BirA-ABCG1 transfectants), an endosomal transporter that is involved in lipid homeostasis.
Proteins that interacted with ABCG2 and ABCG1 were analyzed and compared for differential subcellular localizations and cellular pathways.
ABRÉGÉ
Les transporteurs ABC sont essentiels pour la désintoxication et le transport sélectif des substances dans les cellules humaines. ABCG2 est un demi-transporteur exprimé principalement dans la membrane cytoplasmique de nombreux tissus sains, tels que les barrières tissulaires et le système digestif. En déterminant les interactions protéines-protéines de ABCG2, cela nous permet
Page 2 of 110 d’élucider les mécanismes d’activation et d’inhibition dans les cas de maladies et dans les cas normaux. Les mutations et les altérations dans l’expression de génétiques sont liées aux maladies telles que la maladie d'Alzheimer, le diabète, le cancer, les anomalies génétiques et plusieurs infections. La découverte d’une nouvelle voie de signalisation peut dévoiler de nouvelles thérapeutiques et cibles contres ces maladies. Ce projet utilise l'approche BioID (avec BirA) pour identifier des interactions protéines-protéines avec ABCG2 dans une sphère d'interaction de 10Å.
BirA fusionné au N-terminus de ABCG2 a été transfecté dans HEK293F (un ligne cellulaire embryonnaire) et plusieurs clones stables ont été caractérisés fonctionnellement. Les protéines, qui sont biotinylées par HEK-BirA-huABCG2 et identifiées par la spectrométrie de masse ont été comparés à ceux qui interagissent avec ABCG1 (en utilisant HEK-BirA-huABCG1), un demi- transporteur dans la même famille de ABCG2 qui reste à l’endosome et qui transporte le cholestérol. Ces protéines qui interagissent avec ABCG2 et ABCG1 ont été analysées et comparées pour déterminer leurs emplacements dans la cellule et leurs voies cellulaires.
Page 3 of 110
ACKNOWLEDGEMENTS
I would like to offer my absolute gratitude to Professor Elias Georges, my project investigator, for granting me this amazing opportunity in being able to contribute to the growing field of biochemistry and health sciences. Each experience in his lab has challenged me to think critically and independently as well as contribute to my maturity as a person. I thank every one of my lab members, namely: Dr. Fadi Baakdah, Dr. Georgia Limniatis, and Zahra Sahaf for training me upon my arrival at the lab as well as offering me advice during stressful times. I will never forget laughing till my stomach hurt because of Fadi’s dad jokes (and constantly reminding him my name is Kristen with an E) or discussing key issues with Georgia (ie. stress shopping and eating cheese because a result made no sense). I thank Haritha Menon for supporting me as we worked through the stress of our M.Sc. projects together and pushed each other to stay consistent at the gym. My lab mates have been like family to me and I am immensely grateful to have shared these past few years with them and call them friends.
Special thanks to my committee member Professor Sébastien Faucher not only for being an active member of my committee but also training me during my undergraduate degree as one of his interns. Additional thanks to the Fonds de recherche du Quebec – Nature et technologies
(FRQNT) for providing me with the scholarship to financially support me through my degree. I would also like to thank the entirety of the Institute of Parasitology at McGill, namely fellow students: Nisha Ramamurthy, Rishi Rajesh, Dr. Mark Kaji, Jennifer Noonan, Maude Dagenais,
Jysiane Cardot, Moti Sobat, Elizabeth Siciliani, Cathy Shang Kuan and Eyre Nomi. I will always have nightmares from the horror movie Rishi, Nisha and Mark dragged me to see but will still cherish the time we spent together. Thank you to Dr. Norma Bautista-Lopez, Victoria Muise,
Kathy Keller, Linhua Zhang, Mike Massé, Shirley Mongeau and Serghei Dernovici.
Page 4 of 110
I offer my thanks to Kyle Roux for graciously providing us with the pcDNA3.1 mycBioID and pcDNA3.1 MCS-BirA(R118G)-HA vectors. I also would like to thank Chris Hicks for providing me with technical support and ensuring I did not lose my data when my computer crashed several times.
My deepest gratitude to my best friend, Caroline Estrada (M.Sc.), for illustrating the diagrams of our fusion protein. She has been my rock throughout my academic career, and I am honoured to have her in my life.
I want to thank the rest of my amazing friends outside of the university who provided stability and encouragement throughout my degree. I extend my thanks to my college biology teachers: Dr. Shireef Darwish for helping me realize my passion for molecular biology, and Dr. J.
P. Parkhill for offering the advice of never giving up in the pursuit of higher education despite the hurdles you may face.
Finally, I would like to offer my most sincere gratitude to my amazing parents, Maria
Scordilis and Kenny Lee, who have influenced the woman and scientist I am today. Without them
I would have not been encouraged to push myself to my limits to strive for success. They have taught me to fight against all odds, never give up on my dreams and stay true to myself.
Page 5 of 110
CONTRIBUTIONS TO ORIGINAL KNOWLEDGE
The current knowledge on these transporters is limited to studies demonstrating knocking down or amplifying expression of proteins and pathways to alter expression of ABCG1 or ABCG2.
Many studies try to explore resolution techniques involving using inhibitors of these ABCG transporters that have low toxicity, however, most have not been approved for use in human trials.
Understanding possible differences in the interactions between these transporters in healthy versus diseased states, such as cancer, could better isolate therapies that target solely the transporters involved in disease and not hinder their native functions in healthy tissues.
In this thesis we provide a list of proteins hypothesized to interact and therefore take part in a mechanism or pathway involved in mitigating ABCG1 and ABCG2 expression and activity.
CONTRIBUTION OF AUTHORS
The experiments conducted for the project of this thesis were carried out under the oversight of Professor Elias Georges, who supervised the experimental design and is involved in the editing of this thesis and any subsequent manuscripts. The construction of vectors with the N- terminal BirA tag and all of the primers were designed by Zahra Sahaf, also under Professor
Georges’ supervision.
Page 6 of 110
LIST OF ABBREVIATIONS
ABC ATP-binding cassette
ABCB1 ATP-binding cassette transporter, family B, member 1
ABCC ATP-binding cassette transporter, family C
ABCD3 ATP-binding cassette transporter, family D, member 3
ABCG1 ATP-binding cassette transporter, family G, member 1
ABCG2 ATP-binding cassette transporter, family G, member 2
ATP Adenosine triphosphate
BCRP Breast cancer resistant protein/ABCG2
BirA Bifunctional ligase/repressor c2, c4, c5, c15, c16, c20 Clones 2, 4, 5, 15, 16 and 20 (respectively)
DMEM Dulbecco’s modified eagle’s medium
DTT Dithiothreitol
EDTA Ethylenediaminetetraacetic acid
FBS Fetal bovine serum
G418 Geneticin®/G418 disulfide
GOI Gene of interest
HA Hemagglutinin
HEK293F Human embryonic kidney cells 293F
HDL High-density lipoprotein huABCG1 Human ABCG1 huABCG2 Human ABCG2 mAB Monoclonal antibody
Page 7 of 110
MS Mass spectrometry/mass spec
NBD Nucleotide binding domain
PBS Phosphate buffered saline
PCR Polymerase chain reaction
POI Protein of interest
PVP-40 Polyvinylpyrrolidone
ROS Reactive oxygen species
SDS Sodium dodecyl sulfate
TMD Transmembrane domain wt Wild type
Page 8 of 110
INTRODUCTION
ABC transporters
ATP-binding cassette (ABC) transporters are a family of proteins tasked to shuttle substances across cellular membranes and against their electrochemical gradients. These substrates are vast and include lipids, amino acids, peptides, anions, acids, drugs and other native and xenobiotic substrates [1]. Primarily discovered in bacteria, this highly conserved family of proteins relies on ATP hydrolysis for energy to carry out transport [2, 3]. A comparison of these ATP- driven transporters’ protein sequences led to their grouping, as they all shared an ATP-binding cassette, as well as highly conserved Walker A, B and C motifs in the nucleotide binding domain
(NBD) [4-6]. Eventually, a link was established between the ABC transporters identified in bacterial nutrient uptake and those involved in drug resistant phenotypes in mammalian cells [7].
Forty-eight ABC transporters have been recognized in humans which have been classed into seven subfamilies according to genetic similarities (ABCA, ABCB, ABCC, ABCD, ABCE, ABCF and
ABCG) [4, 8, 9].
Despite their discovery in pathologies, ABC transporters are responsible for vital functions in normal human cells. In healthy tissues, these transporters play protective, secretory and selective roles through their trafficking abilities. They tend to line tissues and organs, acting as barriers and permitting selective transport to ensure proper physiology in their respective locations. For instance, ABCC7 lines the epithelial mucosa and regulates chloride efflux that drives mucous secretion in select tissues [10, 11]. Indeed, the mucosal epithelium of the respiratory tract are lined with a variety of ABC transporters with ABCA5, ABCA13 and ABCC5 predominantly expressed
[10].
Page 9 of 110
One of the main roles of certain ABC transporters is tissue protection. This is done by removing toxic host metabolites and preventing foreign compounds from entering said tissues.
These substances will diffuse or enter cells through endocytosis or other transporters and be immediately effluxed back into the extracellular space. Such compounds include many drugs, which often leads to inefficient treatment of patients with neurological disorders or diseases in tissues housing the transporters. While initially discovered as one of the causative agents of multidrug resistance in tumor cells, many ABC transporters were later found to be expressed at the surface of endothelial cells lining the blood-brain barrier including ABCB1, ABCC2, ABCC4,
ABCC5 and ABCG2 [12, 13]. ABCB1 and ABCG2 play the biggest roles in drug efflux. In fact, a suppression of ABCB1, also coined P-glycoprotein 1 (P-gp), expression can cause normally non- neurotoxic drugs such as ivermectin to accumulate in the brain and become lethal to the host [14].
ABCG2 is involved in the removal of multiple drugs and metabolites such as beta-amyloids from brain tissue [15, 16]. ABCC4 has been shown to mediate what crosses into the cerebral spinal fluid as well as limit the entry of drugs such as methotrexate and raltitrexed into the brain [12, 13, 17].
Moreover, ABCB9, ABCC9 and ABCD3 have been identified in rat cerebral micro vessels [18] .
Similarly, ABC transporters line other tissue barriers as those in the placenta, mammary glands and blood-testes barrier. There they not only serve a protective function, but offer selective transport of nutrients to the tissues in question [19-23]. Nutrient uptake and secretory roles mitigated by ABC transporters are also seen in the digestive tract, hepatic tissues and kidneys [10,
21, 24-28].
Intriguingly, the protective role is extended as well to ABC transporters’ involvement in the immune system. ABCA1 and ABCG1, which transport cholesterol and other lipids, do so in
Page 10 of 110 immune cells such as dendritic cells and macrophages and show involvement in the activation of the inflammasome [29].
ABC transporters’ role in diseases
As they are required for many vital biological functions in the human body, ABC transporters are linked to various pathologies. ABCB1 was first discovered in drug resistant
Chinese hamster ovary (CHO) cells, and its lack of expression was evident in the non-resistant wild type (wt) cells [30]. Further studies revealed increased ABCB1 expression, through gene amplification and mutation, to positively correlate with an increased resistance to many cancer therapeutics [31, 32]. Many other ABC transporter overexpression phenotypes have been common in cancer cells including but not excluded to ABCC1 and ABCG2 [13, 16, 30, 33-38].
Perhaps one of the most known non-cancerous diseases acquired by ABC transporter dysfunction is cystic fibrosis. ABCC7 acts as an ATP-dependent ion channel for chlorine, and its mutation leads to the development of cystic fibrosis. The resultant dysfunction leads to an interruption of water movement to mucous membranes, compromising the individual’s immune system along with other pathologies. This has led it to be coined the cystic fibrosis transmembrane conductance regulator (CFTR) [10, 11, 39].
Pathologies linked to these transporters extend past dysfunctions with the host and have been witnessed in many infections. Evidence suggests natural inflammatory response to HIV-1 infection induces a downregulation of ABCB1 and increase in ABCC1 expression [40, 41]. Since the anti-retroviral drugs have been identified as substrates to many ABC transporters, AIDS treatment is limited as the drugs are removed at the blood-brain and blood-testes barriers leaving infected cells residing there intact [21, 22, 40, 41]. While an immune response can alter ABC
Page 11 of 110 transporter expression, the reverse is also witnessed, where ABCA1 and ABCG1 null mice show induction of the inflammasome and increased lymph node size, a feature often seen in lupus patients [29] .
As mentioned previously, ABC transporters play a role in mucosal epithelia of the respiratory tract. As such alterations in expression and function have been shown to amplify or suppress respiratory diseases. For instance, consistent use of tobacco tends to significantly favor heightened levels of ABCB6, ABCC1 and ABCC3 and possibly lower ABCA13. Increased disease severity of asthmatic patients has been linked to ABCC2 downregulation and a possible upregulation of ABCC1 and downregulation of ABCC4 and ABCA13 [10] .
O-linked β-N-acetylglucosamine (O-GlcNAc) protein modifications through increased availability of blood glucose during hyperglycemia have been demonstrated in vitro to upregulate expression of ABC proteins ABCB1 and ABCG2. Thus, their roles in cancer drug resistance are amplified, feeding into the proliferation of cancerous cells. This increase in expression was found to be associated with O-GlcNAc modifications of GLI transcription factors belonging to the
Hedgehog pathway [42]. Further studies have implicated the Hedgehog pathway in ABC transporter activity modulation, showing binding sites for GLI transcription factors in ABCA2,
ABCB1, ABCB4, ABCB7, ABCC2, ABCG1 and ABCG2 promoter sequences, as well as a decrease of anti-cancer drug resistance in tumors with GLI transcription factor inhibition [43, 44].
Although, most if not all ABC transporters have been shown to be involved in many diseases, this thesis will center around the human variants of two: ABCG2, a protein mainly expressed at the plasma membrane of cells lining tissue barriers effluxing drugs and toxins; and its endosomal family member ABCG1, famous for its role in cholesterol transport.
Page 12 of 110
ABCG2
Human ABCG2 was identified in the late 1990s as a major contributor of multidrug resistance in tumors to clinically relevant chemotherapeutics [45]. Initially, separate discoveries led to the protein being coined with various names. Its original discovery in persistent breast cancer tissues prompted its name as the ‘breast cancer resistance protein’ (BCRP), while its ability to efflux cancer drug mitoxantrone labeled it as a mitoxantrone resistance (MXR) protein [45, 46] .
Later studies have since linked ABCG2 overexpression to multidrug resistance phenotypes in many aggressive cancers [13, 33, 36, 37, 44, 45, 47]. Despite this, the protein exists in healthy cells to serve functions vital to biological processes including excretion, selective transport, and protection through removal of toxic xenobiotics. Moreover, Western blot detection and immunohistochemistry have revealed localization of ABCG2 to favor the plasma membrane, unlike many other ABC transporters that are situated within vesicular membranes [48].
As an ATP-dependent protein, it uses ATP hydrolysis to drive these functions. Although its transport is more complex than the simple presence of a ligand and ATP, the basic mechanism has been established with cryo-electron microscopy and other imaging techniques.
ABCG2 structure and proposed mechanism
The basic structure of human ABCG2 has been long established since the early 2000s, which consisted of aspects of the structure found in other ABC transporters. They are all integral membrane proteins with two NBDs that bind an ATP each and allow their hydrolysis, and two transmembrane domains (TMDs). Within the NBDs are highly conserved motifs (Walker A and
Walker B) [1-8]. The differences then lie with the number of passes in the TMD and certain amino acid sequences, which in part determine which substrates are ligands to the transporters [49].
Page 13 of 110
In general, ABC transporters require two NBDs and two transmembrane domains to be functional, yet some like ABCG2 are ‘half-transporters’ and thus function as a dimer, each unit containing one NBD and one TMD. ABCG2 has been depicted with six TMDs, with a relatively short C-terminus and longer N-terminus. There are many important amino acid residues, to which post-translational modifications or mutations may alter ABCG2’s function and expression in the cells and thus induce any possibility for disease onset (Figure 1). Because of commonalities between ABCG2 and ABCG5 and ABCG8, a lot of the missing data on ABCG2’s structure was filled out by known information on the latter two proteins [37, 50]. Recent advancements in technology, however, have uncovered new aspects of the structure and function of ABCG2.
A B
Figure 1: A) basic structure of the ABCG2 half transporter [51]. A simplified two-
dimensional representation of human ABCG2 as a homodimer in the plasma membrane,
as depicted in 2008. The amino acids indicated by the arrows above have been
demonstrated to be essential as mutations affect the function and expression. B) 3D model
Page 14 of 110
of ABCG2 in a plasma membrane with key amino acids involved in gliomas pointed out
[13].
For instance, in vivo and in vitro, ABCG2 has been known to dimerize to achieve its active state as a transporter in a mammalian system [52]. Interestingly, contrary to this belief, recent evidence from correlation spectroscopy, photon counting and photobleaching indicate GFP-
ABCG2 might prefer a tetrameric form in live mammalian cells when expressed at the plasma membrane regardless of substrate presence [53].
Some of the amino acids detrimental to structure and function have been confirmed in recent years. Through a combination of single-particle cryo-electron microscopy and the use of
5D3 antibodies complementary to the extracellular linker region between the two homodimers, a more resolved 3D structure and low-resolution model of the NBD was established. In this study by Taylor et al. (2017), Glu211 in Walker B was demonstrated as essential to function as a point mutation to glutamine completely inhibited ABCG2 efflux of estrone-3-sulfate, a known substrate
[54].
As with structure, basic revelations about the mechanism of transport of ABCG2 are known. Simplified, ABCG2 works as the other members of its family. Two ATPs are hydrolyzed in the two NBDs, providing the energy required to shuttle two substrates across a membrane.
Taylor et al. (2017) was one of many to confirm this, providing additional information about some amino acids involved in the process. Using an antibody, 5D3-Fab, to induce an open-in state, cavity
1 or the substrate binding domain is seen to be composed of hydrophobic domains TM2 and TM5a from each monomer. The two complimentary leucine residues (L554) create the leucine plug separating the cavity 1 from cavity 2 which is made of EL3. In the open-in state, cavity 1 has a high affinity for the substrate while cavity 2 has a lower affinity and is blocked off by the plug.
Page 15 of 110
Cavity 1 consists of 6 hydrophobic residues: TM2’s F432 and F439 and TM5a’s L539, I543, V546, and M549. The larger EL3 domain allows for cavity 2 to exist in ABCG2 but not ABCG5 and
ABCG8 [54].
Other differences have been revealed to exist between ABCG2 and other ABC transporters, such as those in subfamily B, despite their homology. ABCG2’s transmembrane helices and loops are shorter and thus the NBD is closer to the membrane and the overall structure is smaller. An inward cavity is created by a shift in domains TM2 and 5a in each subunit. Motif C2 does not seem to take part in ATPase activity directly, as it lies 20 Å and 45 Å away from Walker-A and signature motifs respectively [54, 55].
The NBDs are linked by their TM1a domains by a linker with a high charge. Connections between the two monomers are due to EL3, the extracellular loop which consists of 3 cysteines
(C592, C608, C603) that are responsible for two intramolecular and one intermolecular disulfide bonds, respectively with their partnering cysteines in the other monomer. Furthermore, an N- glycosylation site exists at Asn596 [54, 56-58]. Interestingly the intramolecular disulfide bonds are critical to drug transport, however C603 is not [54].
Recently, the transporter has been witnessed to open and close like a switch, activated by the binding of a substrate. When in a relaxed state in the membrane with no ATP nor substrate bound, the transmembrane helices keep the transporter off, with the NBDs of each transporter apart. This conformation of ABCG2 is similar to when it is bound by an inhibitor [47]. At this point, ABCG2’s substrate binding domain is still open to the intracellular space, where its first cavity is exposed and cavity2 is blocked by the leucine plug. A substrate binding brings the partnering NBDs closer together and induces a conformational change to an open-out state [54] .
Cryo-electron microscopy has revealed one ligand, such as estrone-3-sulfate, can bind to the
Page 16 of 110 substrate binding domain and be transported by an ABCG2 homodimer one at a time [59]. The conformational change induced by this interaction shifts the substrate to cavity 2 which has very little affinity to the substrate. A proceeding ATP hydrolysis to ADP attracts the conformational change back to an open-in state, by collapsing cavity 2 through a series of alterations of the conformation of the transmembrane domains, pushing the substrate out [54, 59].
While it seems as if the substrates presence switches ABCG2 to an open and closed conformation, what drives it into an active and inactive state still remains unknown. However, regulation at the post-transcriptional and post-translational levels have been revealed.
There are many transcriptional factors and pathways thought to be involved in manipulating ABCG2 expression. For instance, as stated before the GLI transcription factors in the Hedgehog pathway can induce expression of the transporter when glycosylated prompting their activity and likewise a decrease in expression and function with their inhibition [42-44]. Ligands of PPARα have also shown to upregulate ABCG2. In their presence, there was a decreased accumulation of mitoxantrone in cerebral microvessel endothelial cells, emulating the blood-brain barrier. Contrarily, inhibitors of PPARα and treatment with small iRNAs against PPARα significantly reduced expression of the transporter [60]. Hormones, such as naturally produced estrogen, also decrease ABCG2 concentration at the plasma membrane [61] .
At the post-translational level, residue N596’s ability to be glycosylated is essential for ubiquitonation and thus is key for ABCG2’s structural maintenance as disruption leads to degradation [62] . As mentioned before any alteration in expression or function opens a wide array of possible pathologies as natural function becomes altered.
ABCG2 in diseases
Page 17 of 110
Cancer
One of ABCG2’s most known functions is its ability to efflux drugs making them less bioavailable to tissues. This has contributed to preventing the resolution of certain diseases.
Perhaps one of its most famous roles is in the case of aggressive cancers, where overexpression of
ABCG2 grants them immunity to many clinically relevant chemotherapeutics. In fact, ABCG2 was originally discovered in MCF-7 breast cancer cells, giving it one of its many names: Breast
Cancer Resistance Protein or BCRP [45]. Throughout the years, many other multidrug resistant cancers demonstrated overexpression of ABCG2. For example, it has been seen in colon cancer cells resistant to SN38, a metabolite to irinotecan and analogue to camptothecin [33].
The overexpression of this multidrug transporter has cancer patients unable to combat their disease. Expression at the protein and RNA levels has a direct correlation with patient survivability and resolution in response to chemotherapy, seen in many studies on a multitude of ABCG2 expressing cancers. Many of these cancers are known to be aggressive in nature, overcoming extensive treatments such as pediatric acute megakaryoblastic leukemia, clear cell renal cell carcinoma and salivary adenoid cystic carcinoma [63-65].
Hyperglycemia, which is a hallmark of diabetes, is also known to feed cancers. Cancerous cells utilize mass amounts of glucose to aid in their rapid proliferation, metastasis and angiogenesis. The relative expression of ABCG2 increases significantly in SUM1315 and
SUM159 breast cancer cells exposed to hyperglycemic conditions (25mM glucose) compared to non-hyperglycemic states (5mM glucose) [42]. Since ABCG2 requires ATP to function, a spike in its expression and activity is plausible when blood glucose levels are high, allowing for the generation of more ATP.
Page 18 of 110
Hyperglycemia in diabetes
As mentioned previously, ABCG2 expression can be altered with a change in blood glucose. Blood sugars fluctuate naturally after digestion of carbohydrates but also in an immune response to infection or trauma. Diabetes is highlighted by the inability to manage hyperglycemia either through decreased sensitivity of cells to insulin or the decreased availability of insulin.
Increased blood glucose can promote ABCG2 expression in cancer cells through GLI transcription factors in the Hedgehog pathway [42]. Elements of the blood-brain barrier have been shown to be altered in response to hyperglycemia in vitro and poorly managed diabetes often leads to secondary pathologies related to the nervous system such as Alzheimer’s disease [66, 67]. A trend of upregulation and increased ABCG2 activity has been noted in studies exposing ABCG2 expressing cells to hyperglycemic conditions. A twelve-hour exposure to high serum glucose levels rendered a significant upregulation of ABCG2 activity in human cerebromicrovascular endothelial cells
[68]. In vivo studies show an increase in ABCG2 expression at the blood-brain and blood-cerebral spinal fluid barriers in diabetes-induced rats, but a decrease of mRNA levels in hepatocytes in obese rats mimicking conditions favoring the development of type II diabetes [69, 70].
Clinical research on women with type 1 diabetes and gestational diabetes shows well managed diabetes does not impact ABCG2 expression in hyperglycemic events at the placental barrier. This study suggests that expression would change if these women were not regulating their blood sugar, however follow-up studies on this hypothesis have yet to be tested [71].
Gout/hyperuricemia
With poorly controlled diabetes, comes a multitude of secondary health complications.
Hyperuricemia has been seen to develop in patients with chronic hyperglycemia [72, 73]. It is the
Page 19 of 110 failure to excrete uric acid, leading to its accumulation in the serum [74]. As a uric acid transporter,
ABCG2 is commonly associated with these pathologies, and thus mutations and certain variants rendering the protein dysfunctional but still expressed are often found in patients with hyperuricemia. Chronic hyperuricemia results in the development of gout. To date, Q126X,
Q141K and certain polymorphisms of ABCG2 have been identified in patients suffering from this disease [72, 73, 75-79].
Alzheimer’s disease
Apart from gout, preliminary findings show a specific polymorphism in ABCG2 (C421A) has been seen in populations with Alzheimer’s disease [80]. An involvement of ABCG2 in
Alzheimer’s disease has been established in in vivo, in vitro and clinical studies. A hallmark feature of Alzheimer’s is an accumulation of free radicals and metabolites, especially beta-amyloids [81].
ABCG2 effluxes metabolites upon their entry across the blood-brain barrier as a protective mechanism. Some of ABCG2’s substrates are beta-amyloid peptides, which have been demonstrated in in vitro studies in human cells, in vivo studies in mouse brains and clinical studies in Alzheimer’s disease patients. In situ brain perfusion in mouse brains showed ABCG2 knock-in of ABCB1 deficient mice prevented the accumulation of beta-amyloids [82]. Human endothelial cells treated with inhibitors of ABCG2 showed apical to basolateral accumulation of Aβ(1-40)
[83]. It seems as if ABCG2 works in concert with ABCG4 to prevent beta-amyloid concentration
[82]. Strangely, beta-amyloid peptides also seem to interact with ABCG2, which ends with the suppression of ABCG2’s function [15, 84]. Despite this, Alzheimer’s disease patients show an upregulation of expression at the transcriptional and translational levels [15].
The protection ABCG2 offers to the brain in this disease also extends to preventing the accumulation of reactive oxygen species (ROS), which are generated in response to the
Page 20 of 110 inflammatory reaction brought upon by beta-amyloid accumulation [85]. In general, inflammation instigates the production of cytokines which have been proven to regulate ABCG2 and other ABC transporters’ expression at the plasma membrane of many cells, such as IL-β, IL-6 and TNF-α
[86]. Alzheimer’s disease, cancer, diabetes and many other ailments regularly activate inflammatory response in a variety of tissues, and thus they may compromise the stability of many tissue barriers as well as other native functions mediated by ABCG2 mentioned below.
ABCG2 role in physiological processes
Despite its pertinent role in disease pathology, ABCG2 is also essential for maintaining various processes vital for healthy physiology. It serves a protective function in limiting which xenobiotic compounds may enter a tissue, deals with waste management to prevent accumulation of toxic metabolites created by said tissue, and mitigates nutrient uptake and secretion.
ABCG2 is present in various organ systems such as the nervous (CNS, cerebellum, cortex, frontal lobe), cardiovascular, respiratory, digestive (colon, small intestine, stomach, duodenum), reproductive (testis, ovaries, mammary glands, placenta), urinary, endocrine and lymphatic systems (thymus, kidneys, liver, adrenals, spleen) [4, 8, 12, 13, 16, 18-23, 30, 33, 34, 74, 79, 87-
94] (Figure 2). Maliepaard et al. (2001) were one of the first pioneers in identifying human
ABCG2 expression in various native tissues with monoclonal antibody staining. They established expression in placental syncytiotrophoblasts, colon and small intestine epithelium, canicular membrane of the liver, and breast tissue lobules and ducts. ABCG2 was also interestingly found to be in the endothelium of venous but not arterial capillaries of breast, colon, small intestine, placental and liver tissues. The expression was confirmed at the transcriptional and protein level through mRNA analysis and immunohistochemistry [90].
Page 21 of 110
Figure 2: Expression of human ABCG2 in tissues according to reads per kilobase per
million reads placed (RPKM) in RNA-seq experiment. Tissue samples were collected
from 95 individuals. Figure retrieved from NCBI [26].
One of the endogenous functions of ABCG2 includes mediating secretion of substances.
Such is this case in cells within the mammary glands, kidneys and digestive tract where selective transport takes place.
Mammary gland and milk secretion
Although it has been found in breast cancer cells, studies show ABCG2 to be expressed and functional in non-cancerous breast tissue, more specifically in the mammary gland. Upon lactation, immunohistochemistry shows ABCG2 to be upregulated significantly in both mouse and human mammary gland cells [95]. A supplement of prolactin in T-47D human breast cancer cells mediates an increase of ABCG2 expression. This process seems to be influenced by the induction of the JAK2/STAT5 pathway by prolactin in which STAT5 initiates ABCG2 transcription [96]. In
Page 22 of 110 the mammary gland, ABCG2 may lead to secretion of nutrients but also to the concentration of several toxins that are being excreted from the mother’s system into the milk [95, 97, 98].
Excretory system and selective transport: kidneys, liver and gastrointestinal tract
Kidneys are responsible for removing urate from the serum into the bladder for urination.
In renal proximal tubule cells, ABCG2 has been demonstrated to excrete urate [99]. In fact, dysfunctional ABCG2 has been linked to the hyperuricemia phenotype in populations, where the mutated variant Q126X is expressed [72-79]. A controlled experiment in human and rat ABCG2 expressing vesicles found internalization of uric acid when ATP was available. Furthermore,
ABCG2 null mice were incapable of clearing uric acid after oxonate treatment, as more plasma uric acid was seen in their intestinal lumen compared to in the wild type mice [76].
Moreover, ABCG2 is expressed in the liver, contributing further to the excretory system.
Evidence shows that it is responsible for helping the liver excrete xenobiotics and toxins like protoporphyrin IX and chlorophylls by concentrating them in the organ [100]. A lack of proper
ABCG2 expression in hepatocytes has even been seen to cause accumulation of protoporphyrin
IX in hepatocytes compromising their mitochondria [92]. Conversely, in the case of erythropoietic protoporphyrin, ABCG2 is detrimental to the liver, allowing toxic accumulation of protoporphyrin
IX, and thus in this disease a downregulation of the transporter would be ideal [101]. ABCG2 is capable of transporting numerous drugs and bile acids allowing for bile excretion from the liver
[24, 27].
In mice, those with a bile duct ligation, mimicking bile obstruction, had a significant increase in ABCG2 expression in the intestinal epithelium, showing an increase in effort to excrete substances to the bile which were accumulating in the mice [34]. Uric acid is also transported by
Page 23 of 110
ABCG2 across the intestinal epithelium, which initiates the excretion process [25]. Thus, selective transport of substances across the gastrointestinal tract via ABCG2 are vital in the excretory system as well.
Selective transport in the intestinal tract also serves as a defense against external compounds. Our innate immunity not only consists of serum-based mechanisms of defense but of mechanical ones of the gut. Aside from physical barriers such as the pH levels of stomach acid and commensal organisms encompassing the gut microbiome, transporters like ABCG2 prevent xenobiotics and toxins from entering the mammalian body. For instance, toxic chlorophyll derivatives are one of the many substrates of the transporter [100]. Drugs like imatinib are removed to prevent their oral toxicity. This has been demonstrated in mice where an increased expression of ABCG2 in the gastrointestinal tract had a decreased accumulation of ingested imatinib [34].
Other carcinogens are also banned from oral absorption, mediated by ABCG2’s presence in the gut [102].
The blood-brain barrier
ABCG2 is important in protecting neurological tissue and lines the blood-brain barrier, being 20 times more abundant in expression in brain microvessels than within the cortex itself [94,
103]. It is also fairly more abundant at this barrier than ABCB1 [94]. The discovery of its expression at the blood-brain barrier allowed for an investigation of its native function. Here,
ABCG2 has been shown to expel many metabolic waste products that may cause neurological damage if concentrated in the tissues. As previously mentioned, these products include serum Aβ and other beta-amyloids largely responsible for the development of Alzheimer’s disease, as well
ROS and other xenobiotics [15, 82, 83, 104]. As one of the components of the blood-brain barrier,
ABCG2 contributes to many failed treatments of neurological disorders as it decreases delivery
Page 24 of 110 efficiency of many drugs to the brain [13, 21, 40, 105]. Similarly, ABCG2 activity with estrogen has been shown to aid in intracellular accumulation of glutathione that aids in neuroprotection of brain cells during ischemic strokes [61].
Reproductive tissues: the blood-testes barrier and ovaries
Like the blood-brain barrier, ABCG2 is expressed in and protects reproductive tissues. As such, it also limits success of drug mediated therapies, most notably preventing anti-retroviral drugs from reaching HIV infected tissues through the blood-testes barrier [22]. The sensitive tissues house spermatogenesis, where they may be largely affected by xenobiotics and metabolites.
In late stages of spermatogenesis, ABCG2 expression rises, which seems to correlate with cholesterol removed from sperm [106]. Likewise, ABCG2 protects the ovaries against harmful substances that might impinge on reproductive health like cyclophosphamide which causes oxidative stress and inflammation [107].
Development: the placenta and stem cells
Of the many tissue barriers known to express ABCG2 is the placenta, which shields a developing fetus against external substance and waste product accumulation that might terminate the pregnancy or result in congenital defects. It often works alongside other ABC transporters like
ABCB1 in this defensive mechanism and can prevent drugs taken by the mother from entering the placenta. It also effluxes waste produced by the mother or fetus. In vitro placental ABCG2 eliminates bile acids, as it does in the liver, such as cholic acid, taurocholic acid, taurolithocholic acid-3-sulfate, cholylglycylamidofluorescein and glycocholic acid [24].
The placenta is one of tissues expressing ABCG2 the most [108]. Indeed there has been evidence supporting defects in ABCG2 increases the possibility of congenital disorders or death
Page 25 of 110 of the fetus in mothers who consume drugs known to be ABCG2 substrates during pregnancy
[109]. Chorioamnionitis, or inflammation of the placenta, induces preterm delivery. Studies show it increases placental ABCG2 which may contribute to immunomodulation in response to infection
[20, 110]. Interestingly, preterm deliveries showed a decrease in ABCG2 protein expression, while pregnancies carried out to term had an increase in protein expression despite the downregulated mRNA levels [111]. The expression of ABCG2 will also change according to the stages of pregnancy as corticoids and estrogen mediate it [61, 104, 112].
ABCG2 is not restricted to selective transport in the placenta in the developmental process of embryos, but also during embryonic stem cell generation where it selects which cues stem cells are exposed to, ultimately affecting their maturation [89]. For example, multiple mouse hemopoietic stem cells have demonstrated high expression of ABCG2 which plummeted upon differentiation. When ABCG2 expression was maintained, stem cells differentiated to bone marrow cells and were halted from further maturation [87]. Here, it is hypothesized that ABCG2’s function is to protect the stem cells against xenobiotics and oxidative stress [89]. Interestingly stem cell differentiation into erythrocytes is also upregulated by an increased ABCG2 expression where it reduces toxic substances like intracellular protoporphyrin IX, protecting the cells and transports heme [113-115].
This native function in stem cells not only makes ABCG2 a good candidate as a stem cell biomarker but opens avenues to explore for regeneration and growth of cells that do not normally regenerate such as cardiac progenitor cells [116]. This stems from the established endogenous role played in skeletal muscle regeneration as ABCG2 deficient mice seem to lack this ability [93].
ABCG2’s known substrates
Page 26 of 110
Drugs
ABCG2 has many known substrates consisting of both native metabolites and compounds and xenobiotics. The initial discovery of this breast cancer resistance protein was coupled with the newfound information that ABCG2 is able to efflux many common cancer therapeutics out of multidrug resistant tumor cells [45]. Recognition of substrates and inhibitors might be attributed to the strength of binding affinity and size of compound [59].
One of the most known pharmaceutical substrates of ABCG2 is mitoxantrone. Early studies showed recognition of mitoxantrone resistant cell lines lacking expression of known effluxers of the drug such as ABCB1 and ABCCs, which were later confirmed to express ABCG2 [46]. Cells expressing ABCG2 shuttle mitoxantrone out at the point of the plasma membrane conferring their resistance [117]. Mitoxantrone has thus become a popular substrate in studies on ABCG2 activity and function and confirming its expression [21, 45, 46, 48, 50, 56, 58, 60, 86, 89, 90, 113, 117].
In fact, most of the known substrates have been drugs effluxed by ABCG2 in an attempt to resolve a disease, many of them being tyrosine kinase inhibitors. Of these include methotrexate, topotecan, doxorubicin, rosuvastatin, acyclovir, cimetidine, sunitinib, grepafloxacin and sulfasalazine [45,
95, 118-123].
Native compounds and metabolites
ABCG2 effluxes many native compounds and metabolites. Glutathione is involved in redox homeostasis and thus an important factor in cell death. It has been identified as one of
ABCG2’s substrates which has linked the transporter to oxidative stress and allowing brain cell death following ischemic stroke [61, 124]. Bile acids and uric acid are known substrates, involving the transporter in hyperuricemia and gout formation if mutated or its expression is altered [73, 77-
Page 27 of 110
79]. Beta-amyloids, cholesterol, protoporphyrin and heme are also natural substrates to ABCG2
[15, 83, 101, 106, 125].
ABCG1
Unlike its counterpart ABCG2, ABCG1 is an intracellular transporter that mainly is expressed in endosomal vesicles [126, 127]. Like ABCG2, ABCG1 is a half-transporter and works as a homodimer, with two ABCG1 subunits coming together to form a full transporter. The structure of ABCG1 is not as resolved as that of ABCG2, but similarly has six transmembrane domains and a nucleotide binding domain in its N-terminus in the lumen of the cell or organelle it is expressed on [128, 129] (Figure 3 and 4).
Figure 3: Basic structure of ABCG proteins. Simplified schematic of ABCG transporters
as depicted in the cytoplasmic membrane, with the N- and C- termini facing the lumen of
the cell [128] .
Page 28 of 110
Figure 4: Structure of ABCG1 with each domain highlighted. ABCG1 represented by
its domains, with a small C-terminus, similarly seen on ABCG2, and six transmembrane
domains. ABCG1 (+12) refers to the isoform illustrated [130].
Initially, also named ABC8, ABCG1 was found to be an intracellular transporter of cholesterol and phospholipids in human macrophages [131]. Since then, it has also been identified as a general sterol transporter and is known for lipid homeostasis, working alongside ABCA1 [127,
132]. This function is detrimental to cholesterol homeostasis of many cells, trafficking the sterol to high-density lipoprotein (HDL) so it may be exported into the plasma. Its disruption thus leads to an accumulation of cellular cholesterol and reduced HDL concentration in the plasma. It is also an important player in controlling lipid levels in tissues at a larger scale [133].
In general, ABCG1 is an intracellular protein. It is mainly expressed in the late endosome but may be trafficked to the plasma membrane [134]. Research indicates plasma membrane
ABCG1 is most likely mediated by actin filaments [135]. In pancreatic β-cells, ABCG1 is localized in the Golgi apparatus and endosomal recycling complex [136]. In these cells, it actually serves to help the formation of insulin granules for insulin secretion, preventing the degradation of these granules [137].
ABCG1 is regulated in a variety of tissues such as the adrenals, appendix, bone marrow, central nervous system, gastrointestinal tract, fat, heart, kidneys, liver, lungs, leukocytes,
Page 29 of 110 mammary glands, pancreatic β-cells, placenta, skin, spleen, and male reproductive tissues [126,
130, 138-152]. In most cases this is because of its expression in local tissue macrophages residing there. Modulation of lipid and sterol composition through ABCG1 regulates the immune system and overall physiology. Its expression defines many aspects of immune cells, for example ABCG1 null mice mount a pro-inflammatory response to lipopolysaccharides (LPS) [139].
ABCG1 in diseases
As it is expressed in immune cells such as macrophages, ABCG1 is involved in the immune system activating the NLRP inflammasome. Dendritic cell ABCA1/ABCG1 deficient mice have demonstrated a lupus-like phenotype having enlarged lymph nodes and mounting T-cell activation and differentiation [29]. A lack of ABCG1 expression in alveolar macrophages, along with subdued ABCA1 expression prompt inflammation in pulmonary granulomas in mice [153].
Conversely, an overexpression of ABCG1 is seen in macrophages of patients with Tangier disease, where serum HDL is abnormally low and intracellular cholesterol is high [154]. In these diseases, as well as in atherosclerosis, ABCG1 seems to work in tandem with ABCA1 for pathogenesis. In atherosclerosis, ABCG1 and ABCA1 expression is subdued in macrophages, leading to the formation of ‘foam cells’. Studies have attempted to boost the expression of these transporters to reduce cholesterol accumulation in these macrophages and reverse atherosclerosis [141, 155-157].
As with other ABC transporters, ABCG1 upregulation has been seen in cancer. For instance, it is a feature of many lung cancers and induces lung cancer cell proliferation in vitro
[158]. Mice with MB49-bladder carcinoma and B16-melanoma had a reduced tumor growth when
ABCG1 was knocked out. Activation of NF-κB and cytotoxic responses to the tumor cells were also amplified [159]. Inversely, ABCG1 expression reduction in T-cells seems to promote cell division, suggesting intracellular cholesterol levels induce lymphocyte proliferation, and ABCG1
Page 30 of 110 is a negative regulator of this pathway [160]. This protein has even been identified in breast cancer cells, at a higher expression level than that of normal breast tissue [161]. ABCG1 overexpressing cancerous cells are also granted drug resistance, as was seen with other ABC transporters, such as with etoposide and doxorubicin in OST-EC50 cells [162].
Moreover, other diseases may be linked to the transporter. Methylation of ABCG1 at
CpG13, CpG14 and CpG15 increase the risk of type II diabetes in a rural Chinese population [163].
Retina damage was observed in mice where ABCG1 and ABCA1 function was interrupted in retinal pigment epithelial cells [149].
Why ABCG2 and ABCG1 cannot be inhibited to alleviate related diseases
Early studies on multidrug resistant cancers overexpressing ABCG2 explored the possibility of inhibiting ABCG2 to allow chemotherapeutics to target tumor cells. For example,
Fumitremorgin C and analogue Ko143 have been shown to inhibit ABCG2 in cancerous cells, however, in a live model, the drug targets the protein in a non-specific fashion, making it capable of causing adverse side effects. Moreover, it is an inhibitor of ABCB1 and ABCC1 [164-167].
Mairinger et al. (2018) have shown the inhibitor to be able to target multiple ABCG2 expressing tissues [168]. Other inhibitors used in in vitro and in vivo animal studies have included selonsertib, voruciclib, linsitinib, ulixertinib and other drugs as potent ABCG2 silencers [164, 169-172].
Vindesine administered orally alongside methotrexate actually increased chances of kidney injury in a clinical trial of patients with hematological malignancies [173].
Depending on the expression level and location of expression of ABCG2, this protein is both beneficial and detrimental to human health. As mentioned before the functions in normal physiology are absolutely vital to human biology and thus functional transporters are necessary to
Page 31 of 110 prevent morbidity and mortality. While overexpression of ABCG2 may contribute to the onset of many diseases such as cancer, an untargeted inhibition of ABCG2 can bring upon pathologies that are just as fatal.
Such is the case with ABCG1, as shown in knock-out and deficiency studies where mice and even humans developed a range of pathologies from altered expression and function in immune response, retina damage, atherosclerosis and may be more susceptible to type II diabetes
[141, 149, 153-157, 163]. It could even induce glucocorticoid deficiency and lead to other pathologies downstream [143].
Despite definitive evidence of ABC transporter involvement in vital physiological functions, studies involving ABCG2 and ABCG1 disruption of function and expression continue to proceed in hopes of discovering novel therapeutics. For instance, lipid-saporin nanoparticles have been exploited recently, however, without promise of direct delivery to tumorous cells without affecting healthy tissues, the therapeutic approach can induce many complications [174].
Established protein-protein interactions of ABCG2 and ABCG1
While not many direct protein interactions have been identified with ABCG2, this protein’s expression has been associated with a few known pathways. ABCG2 protein expression levels seem to be mitigated by activity of GLI transcription factors involved in the Hedgehog pathway
[42-44] As such, it is expected that some of the proteins interacting with ABCG2 are linked to this pathway. Interestingly, ABCG1 as well as ABCA2, ABCB1, ABCB4, ABCB7 and ABCC2 also have binding sites for GLI transcription factors [44]. Inhibition of the Rho-associated protein kinases or the ROCK2/moesin/β-catenin pathway with fasudil decreases ABCG2 translocation to the surface [175]. ROCKs are involved in cytoskeletal arrangements and hypothesized to permit
Page 32 of 110 chemotherapy resistance [176]. Growth hormone receptor knockdowns lead to silencing of the mTOR/AKT pathway as well as decreased ABCG2 expression and promotion [177]. Silencing the
PI3/AKT signaling reduces ABCG2 upregulation in the presence of uric acid [25]. The JAK/STAT pathway also influences ABCG2 expression, where its suppression in turn suppresses ABCG2.
This pathway has been affected by suppressing growth hormone receptors and SOX9 [48, 96, 177,
178]. As previously mentioned, defense against hypoxia seems to be mediated by ABCG2, by interacting with the MAPK/ERK pathway [91].
Some proteins have been identified to mitigate ABCG2 expression. Mouse knockout studies indicate enhancer of zeste homolog 2 (EZH2) to have tumor-suppressor functions by also reducing ABCG2 expression [179]. TGF-beta induces Smad2/3 complex formation to bind transcription sites in ABCG2 silencing the gene, ultimately reducing cell proliferation in 2MLN- dnTβRII cells [180]. E2F1 transcription factor promotes ABCG2 expression [181]. As mentioned before the inflammasome also seems to affect ABCG2 or be affected by it. Inhibitors of TLR4-
NLRP3 inflammasome, reduce expression at the plasma membrane [25].
Few proteins have been established as directly interacting with ABCG2, and most are hypothesized to associate with each other, with no functional link established (Figure 5). To date, fellow ABC transporters ABCC1 and ABCC2 have been shown to associate with ABCG2, being involved in several similar pathways. For example, ABCC1 expression is also maneuvered by the
Hedgehog pathway and it shares many substrates with ABCG2 such as glutathione [28, 182, 183].
These two transporters, along with ABCB1 are the three most prevalent in multidrug resistant cancers so possible interactions are similar induction mechanisms are plausible. For the second member in class C, ABCC2 serves an excretory function in the hepatobiliary system as does
ABCG2, which may link it in several pathways [184].
Page 33 of 110
Figure 5: Proteins known to associate with ABCG2 to date (Retrieved from string- db.org). Current protein network of known interactions with human ABCG2.
Figure 6: Proteins known to interact with ABCG1 to date. (Retrieved from string- db.org)
Page 34 of 110
ABCG1 also has no known interacting proteins to date (Figure 6). A study on bladder cancer linked CYP27A1 to the transporter, demonstrating its upregulation of ABCG1 resulting in decreased levels of intracellular cholesterol [185]. Several transcription factors involved in sterol transport such as sterol-regulatory element binding protein 2 (SREBP-2), have been found upregulated when ABCG1 was overexpressed. Interestingly, liver X receptor also induces ABCG1 expression [150]. Protein kinase A is capable of stimulating ABCG1 mediated cholesterol efflux, through a direct interaction with a PKA consensus motives within the protein [130].
A
Page 35 of 110
B
Figure 7: Schematic representation of BirA-huABCG2 fusion protein. A) The 2D
monomer integrated within a bilayer B) Diagram of the fusion protein as a
homodimer with the sphere of interaction illustrated. Components and labeling radius
are not proportionally accurate. Diagrams illustrated and graciously provided to us by
Caroline Estrada.
As many of the interacting proteins are not known and a proper mechanism has not been established, the aim of this project is to identify these interacting proteins using BirA, a 35kDa protein that promiscuously covalently binds biotin to lysine residues that are within a 10 nm radius
[186, 187]. Previous research on protein-protein interaction has been conducted using this proximity-dependent biotinylation approach [187-190]. BirA will be fused to our protein of interest ABCG2 and ABCG1 which will be used as an internal control for subcellular localization purposes (Figure 7). The fusion protein will be expressed in human embryonic kidney cells (HEK)
Page 36 of 110
293F. The unspecific nature of this method means that results are broad and do not confirm direct interaction between these ABC transporters and the biotinylated proteins. However, it a sufficient preliminary step to identifying possible players in the mechanisms of action of each protein in their natural and diseased states.
While the focus of this project remains on ABCG2, ABCG1’s interacting proteins will also be explored with aims to use this protein as an internal control.
Since disruptions to ABCG1 and ABCG2’s expression and function are linked to many physiological disorders and diseases, treating diseases like cancer through suppression by drugs is not a favorable solution. Thus, exploring the interacting proteins that may be involved in the pathways that regulate the transporters’ active and inactive states may reveal new drug targets that can selectively inhibit ABCG2 and ABCG1 involved in pathologies, without disrupting natural function.
Page 37 of 110
METHODOLOGY
Cloning
Cloning began with a PCR amplification of the GOI with the appropriate restriction sites at each end. Digestion of both the vector and inserts took place to create sticky ends. Primers were designed and ordered from IDT. HuABCG2 and huABCG1 were ligated into pcDNA3.1 mycBioID (N-terminal constructs) between NotI (GCGGCCGC) and AflII (CTTAAG) restriction sites, and pcDNA3.1 MCS-BirA(R118G)-HA (C-terminal constructs) between NheI (CTAGC) and AflII (CTTAAG). pcDNA3.1 mycBioID (Addgene plasmid # 35700 ; http://n2t.net/addgene:35700 ; RRID:Addgene_35700) and pcDNA3.1 MCS-BirA(R118G)-HA
(Addgene plasmid # 36047 ; http://n2t.net/addgene:36047 ; RRID:Addgene_36047) were a gift from Kyle Roux [187]. We used the following primers for the N-terminal constructs:
F.HuABCG1.NotI (AGCACGGCGGCCGCCGATGGCCTGTCTGATGGCC),
R.S.HuABCG1.AflII (AGCACGCTTAAGTTACCTCTCTGCCCGGATTTTG),
F.HuABCG2.NotI (AGCACGGCGGCCGCCGATGTCTTCCAGTAATGTCGAAGT), and
R.S.HuABCG2.AflII (AGCACGCTTAAGTTAAGAATATTTTTTAAGAAATAACAATTTC).
For the C-terminal constructs we used the following primers: F.HuABCG1.NheI
(AGCACGGCTAGCGCCACCATGGCCTGTCTGATGGCC), R.HuABCG1.AflII
(AGCACGCTTAAGCGCCTCTCTGCCCGGATTTTG), F.HuABCG2.NheI
(AGCACGGCTAGCGCCACCATGTCTTCCAGTAATGTCGAAGT), and R.HuABCG2.AflII
(AGCACGCTTAAGCGAGAATATTTTTTAAGAAATAACAATTTC).
Gel extraction and PCR purification were performed using EZ-10 Column Kits (BioBasic; BS354,
BS364) to isolate the purified plasmids. Constructs were then cloned into TOP10 Escherichia coli cells through transformation via heat shock. Transformed colonies were selected on LB agar
Page 38 of 110 supplemented with 200µg/mL ampicillin (BioBasic; AB0028), as each plasmid confers ampicillin resistance (Figure 8). Randomly selected colonies underwent PCR amplification (denaturation temperature: 94°C, annealing temperature: 59°C, extension temperature: 72°C, number of cycles:
32) using the primers aforementioned to confirm they contained the GOI. Positive colonies were determined through gel electrophoresis and then sent for Sanger sequencing (Genome Quebec) to ensure no mutations took place.
A
Page 39 of 110
B
Figure 8: Vector diagram of A) pcDNA3.1 mycBioID used for the N-terminal
constructs and B) pcDNA3.1 MCS-BirA(R118G)-HA used for the C-terminal
constructs.
Cell culture and maintenance
Adherent HEK293F cells were used as the mammalian vector to express the fusion proteins. Cells were maintained in DMEM-high glucose (Gibco; 11965092) with 10% FBS
(Wisent Inc; 098150) in T25 flasks (Biobasic; SP81136) at 37°C with 5% CO2 mimicking physiological conditions and trypsonized with 0.3% trypsin (1:250 porcine pancreas; Bioshop;
Page 40 of 110
TRP004.50). Cells were frozen in FBS with 10% DMSO (Sigma; D8418-100ML) at -80°C or in liquid nitrogen.
Transfecting into HEK293F
Plasmid extraction was performed with the EZ-10 Column Kit (Biobasic; BS414) on E. coli with BirA-huABCG1 and BirA-huABCG2 as per the manufacturer’s instructions. All plasmids, including empty vectors, were linearized with MluI (New England Biolabs; R0198S).
We plated 400 000 cells per well in a 6 well plate (Falcon®; 353046) 18 hours before transfection.
We then mixed Lipofectamine® 2000 (Invitrogen; 11668-019) along with Opti-MEM™ (Gibco;
11058021) and the plasmid extract according to manufacturer’s instructions. The media was removed from the cells in the wells and replaced with DMEM without the addition of FBS. The lipofectamine-Opti-MEM-plasmid mixture made before was added dropwise with a 1mL pipette over the cells and left to incubate for 45h in 37°C with 5% CO2.
After the 45h incubation in the transfection media, media was changed to DMEM + 10%
FBS with 0.8mg/mL of G418 (Biobasic; GDJ958). Both plasmids contain a neomycin resistance gene, allowing us to use G418 to select for positive clones at a concentration of 0.8mg/mL (Figure
8). Cells were then trypsonized and transferred to a T25 flask, with 10µL put into 10cm cell culture dishes (Progene®; 229621) for selection. When colonies of HEK293F were visible, media was removed, and clones were picked by scraping with a 10µL pipette tip (Vertex; 4135N00) and deposited into a 48 well plate (VWR®; 734-2326) to grow.
Western blot analysis
Proteins were extracted using a RIPA buffer (50mM TrisHCl pH 7.4, 150mM NaCl, 1mM
EDTA, 0.25% Na-deoxycholate, 1% igepal) with 1x complete protease inhibitor. A BCA protein
Page 41 of 110 assay kit (Thermo Scientific; 23225) was used to quantify protein concentration and 20ng of each protein was resuspended with 5x Laemmli Buffer (250mM TrisHCl pH 6.8, 346.7mM SDS,
0.8mg/mL bromophenol blue, 500mM DTT, 30% glycerol). Proteins were run on an SDS-PAGE and then transferred onto PVDF Immobilion-P Transfer membrane (0.45µm pore size; Millipore;
IPVH00010).
N-terminal expression in HEK293F was confirmed using antibodies against ABCG1 and
ABCG2 as well as the C-myc tag located in the vector. Antibodies used include ABCG2 antibody
(H-70) (Santa Cruz biotechnology; sc-25821; 1/1000 dilution) later replaced by ABCG2 antibody
(B-1) (Santa Cruz Biotechnology; sc-377176; 1/1000 dilution), anti-ABCG1(Abcam; ab52617;
1/1000 dilution), anti-c-myc (9E-10) (University of Iowa; 9E10; 1/4 dilution), anti-HA (ATCC;
1/4 dilution), anti-tubulin (University of Iowa; 12G10; 1/100 dilution), and streptavidin-HRP
(Abcam; ab7403; 1/40 000 dilution).
Confirmation of function: clonogenic and cytotoxicity assays
Clonogenic and cytotoxicity assays were performed using mitoxantrone and topotecan
(Sigma; M6545, T2705), respectively. These were conducted on clones of HEK293F BirA- huABCG2 as well as HEK293F transfected with the empty vector and the HEK293F wt. For huABCG2-BirA, transfection did not prove successful. Clonogenic assays measured percentage of cell growth according to a seven-day incubation in various concentrations of mitoxantrone
(0.78nM - 100nM). Cells were seeded at 1500 cells per well in a 48 well plate (VWR®; 734-2326) in biological and technical triplicates for each concentration. Cytotoxicity assays were performed on each cell line in the same manner as clonogenics, however, they proceeded in a 96 well plate
(Biobasic; SP41207) with a starting cell count of 1000 cells per well. The incubation in topotecan
(1.95nM - 1000nM) lasted five days. Both drugs were dissolved in DMSO. Analysis of cell growth
Page 42 of 110 was obtained through resazurin fluorescent assay. Resazurin (Sigma; R7017-5G) dissolved in PBS was added to each well for a final concentration of 0.0125% over a period of 10-12h at 37°C.
Fluorescence was measured using the Synergy H4 hybrid spectrophotometer from BioTek® and
Gen5™ microplate reader and image software at Ex530/ Em590. Percent growth was calculated by subtracting wells with 100% death (highest concentration) and dividing by 100% growth (cells with added DMSO). Results were plotted using GraphPad Prism version 8.4.2 (679), which was also used to generate approximate IC50 values. At least 2 technical replicates were performed for each drug assay.
Confirmation of function: Fluorescence confocal microscopy with Hoechst 33342 dye
Hoechst 33342 dye was diluted in PBS to a final concentration of 10mg/mL. Five-thousand cells were seeded per well in a 96 black well plate (Fisher Costar; 7200565). Two days after seeding, the media was removed carefully and 100µL of 10µg/mL Hoechst was added to cover the bottom of each well and incubated for 10 minutes, protected from light. The dye was removed and replaced with 1x HBSS and let sit for 30 minutes. Cells were then imaged via fluorescent microscopy (Zeiss LSM710 laser scanning confocal microscope) with Airyscan and the appropriate program-integrated fluorescence wavelengths.
Proximity-dependent biotinylation
Cells were seeded at 1 500 000 per T75 flask (Biobasic; SP81186). The next day the media was switched with that containing 50µM biotin (Sigma; B4639-IG) and the cells were left to incubate for 48 hours. Upon completion of incubation, cells were rinsed with ice cold 1x PBS five times and then detached with mechanical force and suspended in 3mL of ice-cold PBS. The cells were spun down and resuspended in 800µL RIPA buffer (50mM Tris pH 7.4, 500mM NaCl, 5mM
Page 43 of 110
EDTA, 1mM DTT, 04% SDS) with 1x complete protease inhibitor (Thermoscientific; 87786).
Along with vortexing, 200µL of 20% Triton-X was added as well as 1mL of 50mM Tris pH 7.4 and 4µL of DNAse I (Thermoscientific; #90083). The protein extracts were then incubated with
100µL streptavidin agarose (Millipore; 69203-5ML) at 4°C, rotating overnight. They were then washed with 2% SDS then a second wash buffer (0.1% deoxycholate , 1% Triton X-100, 500mM
NaCl, 1mM EDTA, and 50mM HEPES, pH 7.5), a third (250mM LiCl, 0.5% NP-40 , 0.5% deoxycholate, 1mM EDTA, and 10mM Tris, pH 8.1) and a final wash buffer (50mM Tris, pH 7.4, and 50mM NaCl). We removed 40µL of the 200µL bead slurry and dried the beads by sucking up the excess liquid with 27-gauge needle (Sigma -Aldrich®; Z192384-100EA) attached to a 1mL syringe (BD; 309659) and then vortexed them in 30µL 1x Laemmli buffer with 100µM biotin to release the proteins. The released or extracted proteins were resolved on SDS-PAGE and silver stained.
Mass spectrometry analysis
Beads were then resuspended in 50mM ammonium bicarbonate (Bioshop; AMC107.500) and dried using a 27-gauge needle and syringe, sealed with parafilm in a 1.5mL Eppendorf tube and sent to The Hospital for Sick Children in Toronto for standard protease digestion according to their protocols and mass spectrometry analysis.
Data analysis
Results were analyzed on Scaffold v4.11 and Cytoscape-3 v3.8 [191, 192]. Proteins identified had a unique peptide count above 2 and were eliminated if not listed in all three replicates with a unique peptide count higher than cells not treated with biotin. Proteins were then analyzed based on function and localization in the cell using the data sent to us from The Hospital for Sick
Page 44 of 110
Children. Function and localization were obtained from the data provided on Scaffold. Proteins selected for subsequent experiments were further analyzed using data provided on Uniprot for size and function [193].
Page 45 of 110
RESULTS
Cloning of huABCG1 and huABCG2 into the BirA vectors. HuABCG2 full-length cDNA was cloned into both the N- and C-terminal vector backbones in TOP10 E. coli cells. The selected colonies were sequenced to confirm the absence of introduced mutations in the GOI, promoter, BirA and C-myc or HA tags. Following the confirmation of BirA-huABCG2 and huABCG2-BirA fusion constructs by DNA sequencing, the constructs were transfected into
HEK293F cells (see Methods) and their expression confirmed or analyzed by Western blot. The same was carried out and proved true for huABCG1.
Western blot detection to confirm expression of BirA-huABCG2 and huABCG2-BirA in mammalian cells. Total cell extracts from HEK293F BirA-huABCG2 and huABCG2-BirA were resolved on SDS-PAGE, transferred to a PVDF membrane and probed with antibodies against human ABCG2, as well as a C-myc monoclonal antibody (mAb) for the N-terminal constructs and anti-HA mAB for the C-terminal constructs. Three clones were selected based on expression level: clone 2 (c2), clone 15 (c15), and clone 20 (c20). Figure 9A shows the presence of a polypeptide approximately ~110kDa in size in all three clones of the HEK293F-BirA- huABCG2 transfectants (clones 2, 15 and 20). The molecular mass of 108.3kDa (~110kDa) is the expected size of the fusion protein given the full-length size of huABCG2 is approximately 72kDa,
BirA is 35.1kDa and C-myc is 1.2kDa. The HA tag has a mass of approximately 1.1kDa for the
C-terminal constructs. No signal for the BirA-huABCG2 polypeptide was seen in either HEK293F non-transfected cells or HEK293F cells transfected with the empty vector N-terminal BirA vector
(NBirA empty) (Figure 9A). To confirm the loading of cell extracts, the same PVDF membrane was re-probed with an anti-alpha-tubulin (anti-tubulin) mAb. Figure 9B shows the presence of
~55kDa polypeptide representing alpha-tubulin in all five lanes, as expected. Moreover, although
Page 46 of 110 equal amounts (20µg) of total cell lysates were loaded in the SDS-PAGE per well, c2 seems to have a weaker signal for anti-tubulin, suggesting a lesser amount of protein loaded for this sample.
To further confirm the identity of the ~110kDa polypeptide as the full-length construct encoding
BirA, huABCG2 and the C-myc peptide tag, the same cell extracts shown in Figure 9A were resolved on SDS-PAGE, transferred to PVDF membrane and probed with anti-C-myc mAb.
Figure 9B shows the presence of the same 110 kDa polypeptide in all three cell lysates of
HEK293F transfected with BirA-huABCG2 (c2, c15 and c20), but not in wt HEK293F or NBirA empty vector. As expected, however, the gel lane containing resolved proteins from HEK293F cells transfected with NBirA empty vector revealed ~37kDa polypeptide which encodes BirA sequence plus the C-myc tag at its N-terminal. It should be stated that a total of seven positive clones were generated for BirA-huABCG2 when tested with the huABCG2 and C-myc antibodies and the three highest expressors, clones 2, 15 and 20, were chosen for subsequent experiments
(refer to Figure 9B).
In addition, it was not possible to obtain any stable clones expressing huABCG2 with BirA fused to its C-terminus (results not shown). The latter is not unexpected as we have previously attempted to fuse a different fragment to the C-terminus of huABCG2 and were not successful.
Page 47 of 110
A
135 kDa
Probed with anti-ABCG2 H70 100 kDa
75 kDa
63 kDa Probed with anti-tubulin
48 kDa
B
135 kDa
100 kDa
75 kDa
63 kDa Probed with anti-cmyc
48 kDa
35 kDa
75 kDa
63 kDa Probed with anti-tubulin
48 kDa
Figure 9: Western blot analysis of N-terminal BirA tagged huABCG2 clones
transfected into HEK293F cells. A) Probed against ABCG2 and tubulin as a loading
control. B) Probed against C-myc and tubulin. Negative controls are the empty
pcDNA3.1mycBioID vector stable transfectants and untransfected wt HEK293F cells.
Page 48 of 110
Assessing whether BirA-huABCG2 can efflux known substrates and confer drug resistance. Having confirmed the expression of the BirA-huABCG2 fusion protein in HEK293F cells as stable transfectants, it was of interest to demonstrate the fusion protein was functional with respect to drug resistance and efflux. Mitoxantrone and topotecan, known substrates, were used to evaluate the function of ABCG2. Untransfected HEK293F and the empty N-terminal vector showed similar intolerance to mitoxantrone, as growth slowed significantly at 0.195nM and an
IC50 of 0.8643nM and 0.7885nM respectively. Clone 2 had an IC50 of 13.77nM and clone 15 had an IC50 of 6.42nM. Clone 20, which is the lowest expressor of the three clones, had an IC50 of
2.599nM (Figure 10A). In response to topotecan, the wt cell line and empty N-terminal vector transfectant had an IC50 of 23.21nM and 23.42nM, respectively. Clone 2, clone 15 and clone 20 had an IC50 of 174.4nM, 125.9nM and 102.9nM respectively (see Figure 10B).
A
Page 49 of 110
B
Figure 10: Confirmation of function in three BirA-huABCG2 expressing HEK293F
clones assessed by their percent growth in the presence of cytotoxic drugs compared
to wt HEK293F and HEK293F expressing the empty N-terminal BirA vector (NBirA). A)
Percent growth in response to mitoxantrone. B) Percent growth in response to
topotecan. Each data point is depicted as an average of three replicates. Graphs generated
on Graphpad Prism v8.4.0.
The ability of BirA-huABCG2 clones to mediate efflux was then evaluated using a fluorescent drug substrate: Hoechst 33342. Figure 11 shows accumulated Hoechst dye in HEK293F cells transfected with empty pcDNA3.1mycBioID (Figure 11A), and HEK293F-BirA-huABCG2 clones 2, 15 and 20 (Figures 11B, 11C and 11D respectively). Cells transfected with the empty vector retained the fluorescent dye within their nuclei, while cells expressing BirA-huABCG2 (all three clones) effluxed the dye as is seen in Figure 11.
Page 50 of 110
A B
C D
Figure 11: Fluorescence microscopy of BirA-huABCG2 expressing HEK293F clones
when incubated with Hoechst 33342 dye compared to HEK293F containing the empty
vector. Function is assessed by the ability to efflux Hoechst dye and prevent its
accumulation within the cell after an incubation of 10-15 minutes. On the left is the
fluorescent, and the right is the brightfield image A) HEK293F transfected with the
empty vector (NBirA empty vector). B) HEK293F BirA-huABCG2 c2. C) HEK293F
BirA-huABCG2 c15. D) HEK293F BirA-huABCG2 c20. Images taken using a Zeiss
LSM710 Laser Scanning Confocal Microscope.
Assessing biotinylation activity of BirA-huABCG2. Having established a functional
ABCG2 transporter in HEK-BirA-huABCG2 transfectants, we then determined the capacity or functionality of BirA by examining its biotinylation activity on ABCG2 and neighboring proteins in our three clones. The cells were subjected to an incubation of 24 hours in 50µM biotin with
DMEM + 10% FBS to assess biotinylation from BirA. Results shown in Figure 12A indicate the empty vector transfected cells and the three clones displayed functional BirA proteins that
Page 51 of 110 biotinylated other cell proteins. Western blot analysis probed with streptavidin-HRP showed a distinctive biotinylation pattern for each cell line (Figure 12A). The three BirA-huABCG2 clones displayed the same pattern, with a very concentrated band between 100kDa and 135kDa. Cells transfected with the empty vector show a unique pattern to the clones. Wild type HEK293F displayed two faint bands. A separate probing with an antibody against ABCG2 on the same PVDF membrane confirms the maintained expression of BirA-huABCG2 in the clones. Equivalently concentrated bands appear in each lane when probed with the tubulin antibody indicating equal protein loading in each well (Figure 12B).
Page 52 of 110
A
B
Figure 12: Western blot analysis of N-terminal BirA tagged huABCG2 clones in
HEK293F cells, the empty vector stable transfectants and untransfected HEK293F
incubated in the presence of 50µM of biotin for 24h.
Determining the ideal biotinylation incubation time for BirA-huABCG2. To determine optimal biotinylation times, HEK-BirA-huABCG2 clone 2 was subjected to incubation in the absence of biotin or presence of 50µM of biotin for different time periods of 6h, 12h, 24h and 48h.
Page 53 of 110
Untransfected HEK293F and the NBirA empty vector transfectants underwent the same treatment.
The cells were all plated and protein extracted at the same time for the Western blot and probed with streptavidin-HRP. Figure 13A shows that a longer incubation time in biotin leads to higher concentration and variety of biotinylated proteins in cells expressing NBirA empty vector and
BirA-huABCG2. These results are consistent with increased biotinylation and not increased cell lysate loading as evident by the signal from the anti-tubulin probing signal (Figure 13B). The 48- hour incubation in biotin seemed better to obtain as many proteins as possible to be identified in the mass spec and thus was the time point chosen.
A
B
Figure 13: Western blot analysis of various incubation times of 50µM biotin with
BirA-huABCG2 c2, empty N-terminally tagged BirA vector and wt HEK293F.
Measured at 0h, 6h, 12h, 24h and 48h, where 0h represents no addition of biotin. A) Probed
with streptavidin conjugated with horse radish peroxidase (streptavidin-HRP) to
visualize proteins biotinylated by BirA. B) Probed with anti-tubulin as a loading control.
Page 54 of 110
Affinity capture of biotinylated proteins by BirA-huABCG2 shown by silver stain and
Western blot detection. Having established the optimal biotinylation time to identify biotinylated proteins by mass spectrometry, next it was important to demonstrate the pull-down of biotinylated proteins using covalently linked streptavidin to Sepharose beads. To this end, all three clones of
BirA-huABCG2 and the HEK293F wt were incubated in 50µM biotin for 48 hours in T75 flasks, followed by a protein extraction. Elution from three technical replicates (#5, #6 and #7) extracted from the streptavidin agarose beads for clone 2 and 15 were visualized on SDS-PAGE following a silver stain of the gel alongside their respective total protein extract for replicate #5 and a biotin- free control (Figure 14). The wt HEK293F cells incubated with biotin were also resolved on SDS-
PAGE (Figure 14). A difference in protein concentration was witnessed even though the starting material for each experiment was constant, demonstrating the difficulties in separating biotin from streptavidin. The elution, total protein, and proteins not bound to the beads were also run on a
Western blot and probed with streptavidin-HRP demonstrating biotinylated proteins were attained in the sample (See Figure 15). The results in Figure 14 and 15 show protein patterns similar to that of the Western blot in Figure 12 indicating successful capture and elution of biotinylated proteins by the streptavidin-beads.
Page 55 of 110
Figure 14: Silver stain results for protein extracts of HEK BirA-huABCG2 clones 2 and 15, incubated in 50µM of biotin for 48h. Gel shows extract of biotinylated wt
HEK293F. For both clones an extract of unbiotinylated cells eluted from streptavidin-beads serves as a control (labeled NO biotin), followed by 3 technical replicate extracts from biotinylated cells eluted from the beads (Elu #5, #6 and #7) and a total protein extract from one of the replicates (replicate #5) before the affinity capture (TOT).
Page 56 of 110
Figure 15: Western blot analysis of HEK293F BirA-huABCG2 clones and wild type
HEK293F incubated in the presence of 50µM of biotin for 48 hours. Blot probed with
streptavidin conjugated with horse radish peroxidase to visualize proteins biotinylated by
BirA. Shown is a total protein extract before the affinity capture (TOT), an extract of
unbiotinylated cells eluted from streptavidin-beads serves as a control (labeled NO biotin)
and an extract from biotinylated cells eluted from the beads (Elu). The final lane is an old
sample serving a positive control for streptavidin-HRP.
Identifying proteins interacting with HEK293F BirA-huABCG2 c15 through mass spectrometry. Clone 15 displayed a higher protein concentration than the other two clones and tied with clone 2 as the highest expressors for BirA-huABCG2. Thus, streptavidin agarose beads from two biological replicates (#5 and #6) of this clone, as well as the biotin-free control were selected to be sent for mass-spectrometry analysis (MS). The MS results were obtained as a
Page 57 of 110
Scaffold file. We looked at the exclusive unique peptide counts (EUPC) and removed any proteins that had an EUPC equal to or less than that of the control; these criteria had to be met for both replicates. Scaffold automatically only showed proteins with a EUPC greater than 2 with 98% accuracy. A total of 513 proteins met the criteria. We conducted a third biotin incubation of BirA- huABCG2 clone 15 for 48hours, extracted the proteins and incubated them in the streptavidin agarose beads and sent the beads for a second round of MS alongside its respective unbiotinylated control. The resultant proteins underwent the same analysis process. A total of 616 proteins met the criteria previously stated. The protein lists from the two mass spectrometry rounds, containing all three technical replicates, were compared and 188 common proteins were found.
We used Cytoscape, a networking platform, to generate the desired ball and string model to represent the interactions between ABCG2 and the 188 identified proteins (Figure 16). We further analyzed the proteins based on subcellular localization, provided to us on Scaffold. The MS information on Scaffold also included biological process and molecular function related to each protein. We focused our analysis on localization, looking at proteins expressed at the plasma membrane and nucleus. Note that the proteins are all expressed in more than one location and as a combination of the different locations classified on Scaffold as the Golgi, cell, cell junction, cell part, cytoplasm, cytosckeleton, ER, endosome, extracellular region, intracellular organelle, macromolecular complex, membrane, membrane-enclosed lumen, mitochondrion, nucleus, organelle, organelle membrane, organelle part, plasma membrane and ribosome. We deduced 88 proteins to localize at the plasma membrane and 86 in the nucleus with 33 expressed at both. Forty- seven proteins were expressed in other locations. It is interesting to see that while ABCG2 is mainly plasma membrane bound, more than half of the identified proteins are not.
Page 58 of 110
We also noted the biological pathway each protein belonged to. This was a combination of biological adhesion, biological regulation, cell killing, cellular component organization/biogenesis, cellular process, developmental process, establishment of localization, growth, immune system process, localization, locomotion, metabolic process, multiorganism process, multicellular organismal process, pigmentation, reproduction, reproductive process, response to stimuli, rhythmic process, signaling and viral process. The most prominent pathways seemed to be biological regulation, cell component, localization and its establishment, metabolic process multicellular organismal process and signaling. In addition, the molecular function of each protein recognized was a combination of binding, catalytic activity, molecular function, structural molecular activity and transporter activity. Solely IGF2R had molecular transducer activity as well as a function.
Page 59 of 110
Figure 16: Ball and string diagram of 188 proteins identified to interact with BirA-
huABCG2 c15 according to mass spectrometry divided according to subcellular
localization. Eighty-eight proteins expressed in the plasma membrane (indicated in green),
86 in the nucleus (shown in pink) with 33 in common (green surrounded by pink), and 47
in other subcellular locations (blue). Generated with Cytoscape v3.8 [192].
Of these proteins, thirty-seven have been chosen to confirm expression through Western blot detection. They have been selected based on literature searches and percent coverage given to us on Scaffold. These proteins include: FASN (fatty acid synthase), TUBB (tubulin beta chain), EEF2
Page 60 of 110
(elongation factor 2), ANKRD26 (ankyrin repeat domain-containing protein 26), WWOX (WW domain-containing oxidoreductase), ENO1 (alpha-enolase), LRP2 (low-density lipoprotein receptor-related protein 2), ITGB1 (integrin beta-1), SEMA4C (semaphoring-4C), DSC2
(desmocollin-2), ROCK1 (rho-associated protein kinase 1), TJP1 (tight junction protein ZO-1),
IGF2R (cation-independent mannose-6-phosphate receptor), SLC4A7 (isoform 6 of Sodium bicarbonate cotransporter 3), PRDX1 (peroxiredoxin-1), EPHA2 (ephrin type-A receptor 2),
PAK2 (serine/threonine-protein kinase PAK 2), SLC25A5 (ADP/ATP translocase 2), SCAMP1
(secretory carrier-associated membrane protein 1), WASF2 (Wiskott-Aldrich syndrome protein family member 2), JPH1 (junctophillin-1), EMD (emerin), MARCKS (myristoylated alanine-rich
C-kinase substrate), EGFR (epidermal growth factor receptor), STX7 (syntaxin-7), VDAC2
(voltage-dependent anion-selective channel protein 2), CAD (CAD protein), EZR (ezrin),
NUMBL (numb-like protein), SLC12A7 (solute carrier family 12 member 7), STK38
(serine/threonine-protein kinase 38), SLC30A1 (zinc transporter 1), TRAP1 (heat shock protein
75 kDa), SNX9 (sorting nexin-9), YWHAQ (14-3-3 protein theta), TSC1 (hamartin), and SNX3
(sorting nexin-3). It should be noted that these are preliminary results which may change following more detailed data analysis.
Confirming expression of BirA-huABCG1 through Western blot detection. Having identified proteins that are within a 10nm radius sphere of BirA-huABCG2, which is localized to the cell membrane, it was of interest to compare such set of potentially interacting proteins (direct or within short proximity) to another ABC protein (huABCG1) normally localized to endosomal compartments. Hence, we used the same methods to clone BirA-huABCG1 as we did with BirA- huABCG2 into HEK293F cells. The constructs were sent for sequencing confirming no mutations in the gene sequence. Protein extracts from BirA-huABCG1 clones, the HEK293F wt and the
Page 61 of 110 stable population of the empty BirA vector were analyzed by Western blotting, using antibodies raised against ABCG1 and c-myc tag specific mAb (Figure 17A and 17B). Total protein extracts from six clones were examined by Western blot probed initially with anti-C-myc mAb. The results in Figure 17B show a polypeptide band migrating slightly above 75kDa, with clone 4, 5 and 16
(c4, c5 and c16) expressing the highest levels. The protein extract from cells transfected with the empty vector showed a band running approximately 37kDa in size, corresponding to the soluble
BirA protein (Figure 17B).
A
Page 62 of 110
B
Figure 17: Western blot analysis of N-terminal BirA tagged huABCG1 clones
transfected into HEK293F cells. A) Probed against huABCG1and tubulin as a loading
control. B) Probed against C-myc and tubulin. Negative controls are the empty
pcDNA3.1mycBioID vector stable transfectants and untransfected HEK293F cells.
Confirming BirA-huABCG1’s biotinylation activity and optimal incubation period for biotinylation of proteins. As was conducted for BirA-huABCG2 for the biotinylation time course, the same approach was applied to HEK-BirA-huABCG1. As with Bir-huABCG2 clones in Figure 13, BirA-huABCG1 clone 4 underwent differential biotin incubations, followed by protein extractions resolved on SDS-PAGE and proceeded by Western blot detection with streptavidin-HRP. The resulting images show biotinylation patterns fainter than that of BirA- huABCG2 incubated in biotin for 24 hours. In addition, the pattern shows less protein diversity
(Figure 18).
Page 63 of 110
Figure 18: Differential biotinylation of HEK293F BirA-huABCG1 according to time
incubated in biotin. Western blot analysis of N-terminal BirA tagged huABCG1 clone 4
transfected in HEK293F cells, HEK293F wt and the empty vector stable transfectants in
the presence of 50µM of biotin for 0h, 6h, 12h, 24h and 48h; and BirA-huABCG2 clone 2
incubated in the same concentration of biotin for 24h.
Affinity capture of biotinylated proteins by BirA-huABCG1 shown by silver stain and
Western blot detection. Following the assessment of optimal biotin incubation time, we proceeded with a 48-hour incubation as done with BirA-huABCG2, then immediately conducted a total protein extract, incubated the crude protein with streptavidin agarose beads and eluted a portion. Elutions of one replicate for BirA-huABCG1 clone 5 and two replicates for clone 16, as well as the third replicate for BirA-huABCG2 clone 15, followed by an elution for the unbiotinylated controls for each clone, were run as a Western blot and probed with streptavidin-
Page 64 of 110
HRP (Figure 19). The BirA-huABCG1 clones show less proteins biotinylated that BirA- huABCG2. It is also important to note that the total protein extract was run in Figure 18, and only eluted proteins from beads were run in Figure 19.
Figure 19: Western blot analysis of N-terminal BirA tagged huABCG1 clones
transfectants in HEK293F cells, the empty vector stable transfectants and
untransfected HEK293F incubated in the presence of 50µM of biotin for 24 hours.
Probed with streptavidin conjugated with horse radish peroxidase to visualize proteins
biotinylated by BirA.
To ensure enough proteins were eluted from the beads, the rest of the elution was run on a silver stain, using the same methods as was conducted for BirA-huABCG2 (Figure 20).
Page 65 of 110
Figure 20: Silver stain results for protein extracts of HEK BirA-huABCG1 clones 5
and 16, incubated in 50µM of biotin for 48h. For BirA-huABCG1 clones (c5 and c16)
and BirA-huABCG2 c15 a total protein extract (TOT) is run, followed by 2 technical
replicate extracts from biotinylated cells eluted from the beads (Elu #8 and #9) and an
extract from the unbiotinylated control (NO biotin).
The silver stain in Figure 20 consists of the total crude protein extract, two elutions and the unbiotinylated control for BirA-huABCG1 clone 5, BirA-huABCG1 clone 16 and BirA- huABCG2 clone 15. This silver stain does not show a distinct pattern between the biotinylated and unbiotinylated controls for the BirA-huABCG1 clones; however, the Western blot images from
Page 66 of 110
Figures 18 and 19 indicate a difference in proteins biotinylated. This silver stain does conclude there is enough protein in each sample to send for mass spectrometry.
Identifying proteins interacting with HEK293F BirA-huABCG1 c16 through mass spectrometry. As with BirA-huABCG2, the BirA-huABCG1 results for 2 replicates for clone 16 were received on Scaffold and data was transferred to Cytoscape (Figure 21). A total of 93 proteins were identified by mass spectrometry. We analyzed three subcellular locations for proteins identified to associate with ABCG1: the endosome, nucleus and plasma membrane. Our thought was to explore proteins that would lie within the same localization as ABCG1 and ABCG2.
Twenty-nine of these proteins were identified to localize in the endosome, 25 can be expressed at the plasma membrane, 57 are expressed at the nucleus and 24 are expressed in other locations. As with the data we generated for ABCG2, all these proteins can be expressed in multiple subcellular localizations and are not limited to the ones indicated in Figure 21.
Page 67 of 110
nucleus
Figure 21: Ball and string diagram of 93 proteins identified to interact with BirA-
huABCG1 c16 according to mass spectrometry. A) Of the 93, 29 proteins are expressed
in the endosome (yellow), 25 in the plasma membrane (green), 18 found in both the plasma
membrane and endosome (yellow with green outline) and 57 in the nucleus (pink and
isolated and grouped by the pink bracket). The remaining 24 proteins are localized in other
regions. Generated using Cytoscape v3.8 [192].
Comparing suspected interacting proteins with both huABCG1 and huABCG2 to assess commonalities. A quick analysis of common possible interactors with ABCG1 and ABCG2 revealed a total of 32 proteins (Figure 22). Analysis of subcellular localization indicates 24 can be expressed in the nucleus (RPS3A, RPS2, TXNL1, PCBP1, PRDX1, RPS16, RAI14, CACYBP,
TCP1, ZC3HAV1, NUDC, TPR, DHX9, CCT8, PRKDC, RANBP2, EPB41L2, EEF2, RPS3,
Page 68 of 110
WWOX, AHNAK, AFDN, SLC25A5 and RAN), 12 at the plasma membrane (IRS4, FASN,
DSG2, CD2AP, CIP2A/KIAA1524, EPB41L2, EEF2, RPS3, WWOX, AHNAK, AFDN,
SLC25A5), one at the endosome (RAN) and three expressed at other locations (COPG2, VCPIP1 and EIF4A1). As with BirA-huABCG2, this list of proteins may change following more detailed data analysis.
In terms of function, we divided the proteins into 12 categories, independent of those generated of
Scaffold and dependent of more specific information confirmed by Uniprot. The categories were: signaling pathways (IRS4, CD2AP, RAN, RPS3A), metabolic process (IRS4, FASN, WWOX,
PRDX1, SLC25A5,TXNL1) , nuclear translocation (RANBP2, TPR), mitosis/cell proliferation
(EPB41L2, VCPIP1, NUDC), viral replication and immune response (ZC3HAV1, PCBP1), tumorigenesis (CIPA2A, WWOX), cell differentiation (AHNAK, NUDC), cell adhesion/junctions
(DSG2, AFDN, RAI14), transcription and translation (EEF2, EIF4A1, RPS3, DHX9, RPS2,
RPS16, RPS3A), repair activity (CCT8, TCP1, PRKDC), vesicular transport (COPG2) and protein degradation (CACYBP).
Page 69 of 110
A
B
Figure 22: Ball and string diagram of proteins identified to interact with both ABCG2
(purple) and ABCG1 (orange). A) Interacting proteins are grouped based on
Page 70 of 110
subcellular localization where 24 are expressed at the nucleus (pink), 12 at the plasma
membrane (green), one endosome (yellow) and three in other locations (blue). A colored
outline indicates the protein also belongs to the subcellular localization represented by that
color. B) Interacting proteins are grouped according to function including signaling
pathways (yellow), metabolic processes (red), nuclear translocation (light pink), mitosis
(fuchsia), viral replication and immune response (indigo), tumorigenesis (royal blue), cell
differentiation (sky blue), cell adhesion and junctions (turquoise), transcription and
translation (green), repair activity (lime), vesicular transport (black) and protein
degradation (gray).
ABCG1 and ABCG2 do not interact with each other according to our data, however, do interact with themselves.
Page 71 of 110
DISCUSSION
Initially, our objective was to construct N- and C-terminal BirA tagged ABCG1 and
ABCG2 fusion proteins. However, it was not possible to isolate C-terminal BirA tagged ABCG2 likely due to the small size of the C-terminal domain which has 4 amino acids exposed within the cytoplasm, and contains essential lysine residues [194]. Haider et al. (2011) also found tagging
ABCG2 at the C-terminus with full or half-length YFP interrupted cellular trafficking, causing most of the proteins to rest intracellularly in endosomes. They thus hypothesize that the C-terminus contains a signal sequence mandatory for integration of the protein to the cellular surface, which has been suggested by other studies, especially in the last 4 residues [52, 195, 196]. We also hypothesized that the short length of the C-terminus on ABCG2 coupled with a fusion to the BirA protein causes steric hinderance, thus preventing proper integration into the membrane and possibly blocking expression. This was seen in our previous attempt to express a previous construct
ABCG2-GFP. There is yet conclusive evidence explaining how ABCG2 inserts itself into the membrane. Based on the concept of the sphere of interaction, we do not require a C-terminal BirA fusion protein as well as our N-terminal construct to tag all of the interacting proteins.
Our functional assays conducted for BirA-huABCG2 proved the clones selected were capable of effluxing mitoxantrone, topotecan and Hoechst 33342 (Figures 10 and 11). For the drug assays, we also tested the function of ABCG2 against doxorubicin and methotrexate. The clones however, showed no difference in growth percentage compared to the wt cells and the empty vector transfectant. It is possible that the variant of ABCG2 we have cloned is not resilient to these drugs, or that HEK293F cells do not permit ABCG2 to effectively efflux doxorubicin or methotrexate.
Page 72 of 110
Interestingly, topotecan and mitoxantrone had opposing effects on clone 2 and 15 (Figure
10). For mitoxantrone, clone 15 was clearly the most efficient at clearing the drug allowing for better cell growth and a higher IC50. However, in the case of topotecan, clone 2 had a higher IC50.
This trend was consistent in the other technical replicates conducted. For both drugs, clone 20 showed significantly lower efflux efficiency than the other clones, as seen by the lower IC50s. This coincides with clone 20 growing the slowest of the three clones during cell culture. We hypothesize that BirA-huABCG2 may have been inserted in a gene involved in proliferation during stable transfection in clone 20, causing the slower growth rate. The Western blots probed against ABCG2 show an almost equal expression of the protein of interest between clones 2 and 15 and a fainter band for clone 20, supporting this theory (Figure 9).
In our detection of BirA-huABCG1, we visualized a band running slightly above 75kDa when probing with anti-ABCG1 and anti-cmyc (See Figure 17). As both antibodies detected this band, this suggested this was most likely our desired fusion protein of interest: c-myc-BirA- huABCG1. Interestingly the expected molecular weight should be the combined 1.1kDa cmyc tag,
35kDa BirA and 72kDa huABCG1 protein, and as such the band detected runs below the expected molecular weight. Other studies have indicated ABCG1 running at a lower molecular weight of
67kDa, which seems to be due to being expressed in CHO cells and THP-1 macrophages [197-
202]. As such, we hypothesized the difference in molecular weight to be attributed to the cell line the protein is expressed in. A deviation in molecular weight of proteins on an SDS-PAGE is also often due to post-translational modifications, folding and charges [203].
Issues arose attempting to visualize proteins eluted from beads incubated with total protein extract from HEK293F BirA-huABCG1 clones that underwent biotinylation. We hypothesized that since Western blots probed with streptavidin-HRP showed a more muted signal for proteins
Page 73 of 110 biotinylated for huABCG1 compared to huABCG2, this was the reason for the lack of proteins seen in a silver stain. Western blot detection is more precise in solely detecting biotinylated proteins without picking up the background that might be detected by silver stains. Due to the faintness of the bands, we decided to concentrate the number of cells by 3, extracting protein from
3 T75 flasks, each with 1 500 000 cells, instead of one. This resulted in some visualization of proteins as seen in Figures 19 and 20.
Almost half of the proteins identified to interact with BirA-huABCG2 clone 15 are known to be expressed at the plasma membrane. This is interesting knowing ABCG2 localizes mainly to the membrane but is also known to traffic intracellularly and is expressed in organelles such as the mitochondria [204]. With our data, 15 proteins were identified as being expressed in the mitochondria: RAI14, FASN, WWOX, RPS3, DDX3X, RANBP2, PRDX1, SLC25A5, TUFM,
VDAC2, RACK1, TRAP1, ACSL3, FLVCR1, and YWHAQ. While we have tested the subcellular our specific huABCG2 isomer tagged with an N-terminal GFP, which has been shown to be restricted to the plasma membrane, our cytotoxicity and dye efflux results are consistent with plasma membrane subcellular localization. Moreover, using the identical cDNA clones of huABCG2 and huABCG1, tagged with GFP at their N-termini, we showed the two proteins localize to the plasma membrane and endosomal fractions, respectively (data not included in this thesis).
Almost all of the proteins are expressed at the cytoplasmic and/or cytoskeleton, excluding
24: DSG2, GPRIN1, CTNND1, SLITRK5, CRYBG3, GPRIN3, CPNE8, EPHA2, FAM171A2,
KLRG2, UNC5C, FAM171A1, NECTIN2, SLC12A7, FAM171B, SLC6A15, MPP6, SLITRK3,
TBC1D22B, RASAL2, EPB41L1, SLC4A7, WIPI2, and CORO1C. These proteins are known to
Page 74 of 110 localize in other areas such as organelles or membranes and thus we speculate the interaction would likely be when ABCG2 traffics to the membrane, or if it is expressed in an organelle.
Our mass spectrometry data revealed a total of 188 proteins as possible interactors with
ABCG2. These proteins were considered based on being present in all three technical replicates and having an EUPC higher than that of the unbiotinylated control. Interestingly, a little less than half of the identified proteins are expressed at the membrane and a similar amount are expressed in the nucleus. The subcellular localizations listed in Figures 16 and 22 are not excluded to these locations, rather are one of the possible areas. They also do not represent the exact location during their interaction with ABCG2.
As previously mentioned, String-db recognizes 10 proteins to possibly associate with
ABCG2 (PROM1, CD44, MRPS7, ABCC1, ABCC2, KIT, SLCO1B1, SLC22A8, SLC2A9,
SLC22A12) and 10 for ABCG1 (PPARG, SREBF2, SCARB1, LDLR, NR1H2, APOE, CETP,
APOA1, NR1H3 and LCAT). None of the proteins identified by String-db or Uniprot were revealed in our mass spec samples, save for the ABC transporters interacting with themselves.
This may be due to the String-db list of associating proteins being based upon word searches in the current literature, known as text mining. Most of the papers cited by String-db mention ABCG2 or ABCG1 and the possible interactor, but no experimental evidence to prove the two interact. It seems a link was established between the associated protein and ABCG2 or ABCG1 when both proteins experienced a similar change in expression when a treatment was applied to their host cell. The lack of direct functional evidence and lack of support that our data provides to the associated proteins listed by String-db does not dismiss their interaction, however. It is possible that the proximity-dependent biotinylation may not tag these proteins with biotin as they may be a part of a signaling pathway that is further than the 10nm radius of interaction. Thus, these proteins
Page 75 of 110 could be indirect interactors with ABCG2 or ABCG1. It is also possible that some of the proteins may not have exposed lysine residues and thus are not tagged by BirA. Finally, some interacting proteins may be missed as the biotinylation did not occur within a snapshot moment while ABCG2 or ABCG1 is in an active or inhibited state, and thus certain pathways and interactions may not be activated.
Despite String-db not corresponding with our results, proteins involved in notable pathways were identified. An extensive literature search has linked pathways to be involved with regulating ABCG2 expression, and vice versa. Beta-catenin/ABCG2 signaling pathway has been demonstrated in the literature, more specifically mitigating of ABCG2 expression through suppression of the ROCK2/moesin/β-catenin pathway [175, 205]. Moreover, the Wnt/β-catenin pathway has been shown to lead to positively induce ABCG2 expression [206]. Interestingly,
CTNND1 (Catenin delta-1) has been identified as a possible interactor, and is an oncogene [207].
Studies have shown CTNND1 suppression to be correlated with the inhibition of beta-catenin and the Wnt/β-catenin signaling pathway [207-209]. Thus, a feedback loop may exist where ABCG2, which is upregulated by Wnt/β-catenin, especially during tumorogenesis, may interact with
CTNND1, one of the mediators of this pathway.
Another pathway of interest is the AKT pathway, often referred to as the PI3k/mTOR/AKT pathway, which has been shown to induce ABCG2 expression [177, 210, 211]. In our results, we identified RICTOR (rapamycin insensitive companion of mTOR) as a possible interactor with
ABCG2. Rapamycin has indeed been demonstrated as an ABCG2 and an mTOR inhibitor [212].
Moreover, we identified TSC1 (hamartin) as a suspected interactor, which is involved in TOR regulation via signaling pathways [213, 214].
Page 76 of 110
Previous reports have shown that inhibiting Rho-associated protein kinases or the
ROCK2/moesin/β-catenin pathway prevent ABCG2 trafficking to the cell surface [175, 176, 215].
Our results identified ROCK1 (Rho associated protein kinase 1) and ARHGAP21 (Rho associated
GTPase-activating protein 21) as possible associating proteins. ROCK1 tends to be involved in cytoskeletal remodeling and has been shown to induce proliferation and metastasis of a variety of cancerous cells, even being involved with AKT [216-218]. Given this role, ABCG2 and ROCK1 may work in similar pathways during tumorigenesis, interacting to give some cancer cells their aggressive and resilient phenotypes. ARHGAP21 is involved in a multitude of pathways, and as a
GTPase activating protein, promotes the function of other proteins [219, 220]. Thus, even though
ABCG2 does not hydrolyze GTP, it may interact with ARHGAP21 to invoke the activation of other proteins down the line.
The Hedgehog pathway has been attributed to mediating ABCG2 mRNA levels and some of its transcription factors have been shown to bind ABCG2 motifs [42-44, 182, 221]. However, none of the proteins we identified are a part of this pathway.
ABCG2 is known for its expression at the blood-brain barrier [12, 14, 16, 18, 21, 40, 59,
82, 86, 94, 104, 105, 168] . As such, some of the suspected interactors are involved in neurogenesis and brain development like NUDC, NUMBL, and SEMA4C. LRP2 which transports leptin across the blood-brain barrier was also labeled in our data set. The tight junctions that make up many tissue barriers are made of many membrane proteins. Of our data, we found many proteins involved with adhesion including, but not limited to, DSC2 (desmocollin-2), TJP1 (tight junction protein), and EPHA2 (ephrin type-A receptor 2). The functions of each of these proteins were obtained from Uniprot [193].
Page 77 of 110
Similarly, our results for ABCG2 reveal a few proteins involved in tumorigenesis to be suspected interactors. These include: WWOX, and PRDX1 which may play a role in tumor necrosis. As ABCG2 is often found in multidrug resistant cancer cells, ABCG2 expression may induce some tumor related pathways that these interactors could be involved in, or they may influence ABCG2’s function.
Of our list of proteins, VDAC2 was also labeled as a potential interactor. VDAC2, also known as the voltage-dependent anion-selective channel protein 2, allows entry of small hydrophilic molecules. If one of the substrates that is internalized into the cell through VDAC2, corresponds to substrates effluxed by ABCG2, it is possible that the two might be localized in close proximity in the membrane. This could be an evolutionary advantage to have ABCG2 nearby for immediate removal of a possibly toxic substance upon entry of the cell. It would thus be interesting if ABCG2 and other ABC transporters were dispersed across the membrane in lipid rafts that contain other transporters that may internalize some of their ligands.
Ninety-three proteins were identified as possible associators with ABCG1 and were mostly endosomal or nuclear (Figure 21). Of the 93 proteins identified to interact with ABCG1, the ABC transporter ABCD3 was identified. ABCD3 is mainly known as a bile-acid and fatty acid transporter and is expressed at the peroxisome [222, 223]. Our data extrapolated from Scaffold also identified ABCD3 to be expressed broadly across the cell at the cytoplasm, intracellular organelles, membrane, membrane enclosed lumen, mitochondria, organelle, organelle membrane and organelle part. Despite a lack of evidence suggesting ABCD3 interacts with ABCG1, both are capable of fatty acid transport although in different subcellular locations: the peroxisome and endosome, respectively.
Page 78 of 110
As demonstrated in Figure 22, 32 proteins were identified as potential interactors with both ABC transporters evaluated in this study. The functions of each of the proteins were obtained from Uniprot [193].
As 24 of these proteins belong to the nucleus, many of the functions are related to transcription and translation, such is the case for EEF2, EIF4A1, RPS3, DHX9, RPS2, RPS16 and
RPS3A. We can also see suspected interactors involved in exporting proteins from the nucleus, where they play a role in forming the nuclear pore complex such as RANBP2 and TPR. COPG2 is involved in transport with clathrin-coated vesicles between the ER and Golgi apparatus. These interactions thus may occur during the processing of ABCG2 and ABCG1 in the nucleus and their export to the desired subcellular locations. Moreover, proteins involved with mitosis and cytoskeletal rearrangement like EPB41L2, VCPIP1 and NUDC were also labeled with biotin for both ABC transporters. As these proteins are present in the nucleus and cytoskeleton, it is likely they were within the zone of interaction during cell proliferation.
NUDC is also a protein involved in neurogenesis, which is aligned with ABCG2’s expression in neural tissues. It is possible that ABCG2’s protective role at the blood-brain barrier may be mitigated by NUDC in some way, however, it is important to note that these proteins were derived from HEK293F, an embryonic kidney cell line.
Six proteins were identified as being involved in metabolic processes. Fatty acid synthase
(FASN) as its name suggests, is involved in de novo generation of long chained fatty acids. ABCG1 is a regulator of lipid homeostasis [152]. This also may link ABCG1 with ABCD3. WWOX (WW domain-coating oxidoreductase) is an oxidoreductase also involved in apoptosis and tumorogenesis. PRDX1protects cells against oxidative stress but may also be involved in tumor necrosis. SLC25A5 transports cytoplasmic ADP into the mitochondria and in exchange sends out
Page 79 of 110
ATP into the cytoplasm. It is possible that this protein’s function may be activated by ABCG1 or
ABCG2 when they need ATP to carry out their function, or the levels of cytoplasmic ATP and
ADP drive this function to maintain homestasis. TXNL1 is a thioredoxin protein. Insulin receptor substrate 4 (IRS4), is involved in insulin induced effects on the cell. It has tyrosine kinase activity and is involved in the IGF1R and AKT signaling pathways. As a player in multiple signaling pathways, it may mediate interactions with other proteins in pathways down the line.
More proteins known to be involved in signaling pathways include CD2AP and RAN.
RPS3A is a ribosomal protein, however, it may regulate erythropoiesis by mitigating DDIT3, a transcription factor, and thus may be involved in signaling pathways altering DDIT3’s expression.
EPB41L2, NUDC and VCPIP1 are involved in mitosis and most likely interact with or come into the vicinity of ABC transporters during cell replication.
DSG2 and AFDN, RAI14 are proteins involved in cell junction and adhesion as mentioned in the previous section. While it makes sense that they would interact with ABCG2, a protein expressed at the cellular surface, a link with ABCG1, an endosomal protein that is not known to be expressed at tissue barriers, has yet to be deduced.
Page 80 of 110
SUMMARY AND CONCLUSION
Proximity-dependent biotinylation is an approach that can provide experimental evidence for protein-protein interactions. These include both direct interactions involving the protein of interest and the interactors coming into direct contact, as well as proteins that may be in the vicinity as part of a signaling pathway. Our results have provided preliminary data as to which proteins are suspected to interact with human ABCG1 and human ABCG2. This offers potential identification of mediators of activation and suppression of ABCG1 and ABCG2, which could be further explored to reveal their implications in native and diseased functions. This can provide the foundations to establish mechanisms of action and full signaling pathways involving these proteins in pathogenesis, which may prelude the development of drugs to target members of these signaling pathways, preventing inhibition of the ABC transporters’ vital roles in physiology.
Our data was collected by using this proximity-dependent biotinylation approach developed by Kyle Roux et al. to fuse BirA, an E. coli biotin ligase to the N-terminus of human
ABCG1 and ABCG2 [187]. Both constructs were successfully expressed in HEK293F a mammalian expression vector.
In HEK293F, we demonstrated three clones of BirA-huABCG2 have functional transporter activity using drug assays and incubation with Hoechst 33332 dye. The activity level correlated with various expression levels of BirA-huABCG2 for each clone. For our BirA-huABCG1 clones, we have yet to assess the function of ABCG1 as it will involve cholesterol photolabeling.
Although we have previously demonstrated that the subcellular localization of the specific isoforms of ABCG2 and ABCG1 we work with are restricted to the plasma membrane and endosomal compartments, respectively, using GFP fusion constructs, we have yet to verify this with our BirA-constructs using confocal microscopy.
Page 81 of 110
A total of 188 and 93 suspected interacting proteins were identified by mass spectrometry for ABCG2 and ABCG1, respectively. Of these proteins, 32 were found to be in common between our ABC transporters of interest, despite their distinct subcellular localizations. Most of these proteins were nuclear and might be involved in the processing of these proteins.
Using BirA to label potential interacting proteins has its limitations, which are attributed to the promiscuous nature of this biotin ligase. Many proteins may be labeled at random due to traffic within the cytoplasm and cytoskeleton, the fluidity of membranes (both cytoplasmic and endosomal) and possible Brownian motion within the cell. As such, validation of protein-protein interaction must proceed, including more in depth literature searches concerning the roles of each suspected interactor to see if any relation to ABCG1 or ABCG2 can be made. Subsequent experiments may involve the confirmation of direct interaction using an overlapping peptide approach or co-immunoprecipitation. Unfortunately, these methods will not confirm if indirect interactions are present between the ABC transporters and the suspected interactors, as an interaction might be mediated by intermediate proteins. However, a reduction in of the labeling period may be carried out to reduce randomness, by decreasing incubation time with biotin.
Additionally, to ensure the suspected interactors are indeed interacting with our proteins of interest, we may introduce a ligand of ABCG2 and ABCG1to induce transport activity during the labeling period. This would also identify possible activators of these ABC transporters’ activities and pathways the activation may induce. Similarly, introducing a repressor of each transporter may reveal pathways that silencing may induce.
Moreover, expressing the proteins in different cell lines may allow us to explore mediators and pathways that may be unique to ABCG1 and ABCG2 in diseased versus normal functions.
Page 82 of 110
This project brought on preliminary data that requires further exploration but has uncovered potential proteins that may be key players in ABCG2 and ABCG1 regulation. While some knockdown studies and inhibition and activation experiments have linked the involvement of pathways to regulating ABC transporters, direct experimental evidence is lacking. This information can bring upon a plethora of information on transporters that can allow us to engineer better therapeutics to target the transporters involved in pathologies without compromising their native functions.
Page 83 of 110
REFERENCES
1. Higgins, C.F., ABC transporters: from microorganisms to man. Annu Rev Cell Biol,
1992. 8: p. 67-113.
2. Theodoulou, F.L. and I.D. Kerr, ABC transporter research: going strong 40 years on.
Biochem Soc Trans, 2015. 43(5): p. 1033-40.
3. Berger, E.A. and L.A. Heppel, Different mechanisms of energy coupling for the shock-
sensitive and shock-resistant amino acid permeases of Escherichia coli. J Biol Chem,
1974. 249(24): p. 7747-55.
4. Dean, M., A. Rzhetsky, and R. Allikmets, The human ATP-binding cassette (ABC)
transporter superfamily. Genome Res, 2001. 11(7): p. 1156-66.
5. Higgins, C.F., et al., A family of related ATP-binding subunits coupled to many distinct
biological processes in bacteria. Nature, 1986. 323(6087): p. 448-50.
6. Hyde, S.C., et al., Structural model of ATP-binding proteing associated with cystic
fibrosis, multidrug resistance and bacterial transport. Nature, 1990. 346(6282): p. 362-
365.
7. Walker, J.E., et al., Distantly related sequences in the alpha- and beta-subunits of ATP
synthase, myosin, kinases and other ATP-requiring enzymes and a common nucleotide
binding fold. Embo j, 1982. 1(8): p. 945-51.
8. Fukuda, Y. and J.D. Schuetz, ABC transporters and their role in nucleoside and
nucleotide drug resistance. Biochem Pharmacol, 2012. 83(8): p. 1073-83.
9. Cheepala, S., et al., Cyclic nucleotide compartmentalization: contributions of
phosphodiesterases and ATP-binding cassette transporters. Annu Rev Pharmacol
Toxicol, 2013. 53: p. 231-53.
Page 84 of 110
10. Aguiar, J.A., et al., The impact of cigarette smoke exposure, COPD, or asthma status on
ABC transporter gene expression in human airway epithelial cells. Scientific Reports,
2019. 9(1): p. 153.
11. Riordan, J.R., et al., Identification of the cystic fibrosis gene: cloning and
characterization of complementary DNA. Science, 1989. 245(4922): p. 1066-73.
12. Obermeier, B., R. Daneman, and R.M. Ransohoff, Development, maintenance and
disruption of the blood-brain barrier. Nat Med, 2013. 19(12): p. 1584-96.
13. Wijaya, J., Y. Fukuda, and J.D. Schuetz, Obstacles to Brain Tumor Therapy: Key ABC
Transporters. Int J Mol Sci, 2017. 18(12).
14. Schinkel, A.H., et al., Disruption of the mouse mdr1a P-glycoprotein gene leads to a
deficiency in the blood-brain barrier and to increased sensitivity to drugs. Cell, 1994.
77(4): p. 491-502.
15. Xiong, H., et al., ABCG2 is upregulated in Alzheimer's brain with cerebral amyloid
angiopathy and may act as a gatekeeper at the blood-brain barrier for Abeta(1-40)
peptides. J Neurosci, 2009. 29(17): p. 5463-75.
16. Zhang, W., et al., The expression and functional characterization of ABCG2 in brain
endothelial cells and vessels. Faseb j, 2003. 17(14): p. 2085-7.
17. Kanamitsu, K., et al., Investigation of the Importance of Multidrug Resistance-Associated
Protein 4 (Mrp4/Abcc4) in the Active Efflux of Anionic Drugs Across the Blood-Brain
Barrier. J Pharm Sci, 2017. 106(9): p. 2566-2575.
18. Al Feteisi, H., et al., Identification and quantification of blood-brain barrier transporters
in isolated rat brain microvessels. J Neurochem, 2018. 146(6): p. 670-685.
Page 85 of 110
19. Bloise, E., et al., ATP-binding cassette transporters in reproduction: a new frontier. Hum
Reprod Update, 2016. 22(2): p. 164-81.
20. do Imperio, G.E., et al., Chorioamnionitis Induces a Specific Signature of Placental ABC
Transporters Associated with an Increase of miR-331-5p in the Human Preterm
Placenta. Cell Physiol Biochem, 2018. 45(2): p. 591-604.
21. Hoque, M.T., et al., Raltegravir permeability across blood-tissue barriers and the
potential role of drug efflux transporters. Antimicrob Agents Chemother, 2015. 59(5): p.
2572-82.
22. Whyte-Allman, S.K., et al., Xenobiotic Nuclear Receptors Pregnane X Receptor and
Constitutive Androstane Receptor Regulate Antiretroviral Drug Efflux Transporters at
the Blood-Testis Barrier. J Pharmacol Exp Ther, 2017. 363(3): p. 324-335.
23. Yagdiran, Y., et al., ABC- and SLC-Transporters in Murine and Bovine Mammary
Epithelium--Effects of Prochloraz. PLoS One, 2016. 11(3): p. e0151904.
24. Blazquez, A.G., et al., Characterization of the role of ABCG2 as a bile acid transporter
in liver and placenta. Mol Pharmacol, 2012. 81(2): p. 273-83.
25. Chen, M., et al., Soluble uric acid increases PDZK1 and ABCG2 expression in human
intestinal cell lines via the TLR4-NLRP3 inflammasome and PI3K/Akt signaling pathway.
Arthritis Res Ther, 2018. 20(1): p. 20.
26. Fagerberg, L., et al., Analysis of the human tissue-specific expression by genome-wide
integration of transcriptomics and antibody-based proteomics. Mol Cell Proteomics,
2014. 13(2): p. 397-406.
Page 86 of 110
27. Hernández Lozano, I., et al., Measurement of Hepatic ABCB1 and ABCG2 Transport
Activity with [(11)C]Tariquidar and PET in Humans and Mice. Mol Pharm, 2020. 17(1):
p. 316-326.
28. Zheng, L.-s., et al., Vandetanib (Zactima, ZD6474) Antagonizes ABCC1- and ABCG2-
Mediated Multidrug Resistance by Inhibition of Their Transport Function. PLOS ONE,
2009. 4(4): p. e5172.
29. Westerterp, M., et al., Cholesterol Accumulation in Dendritic Cells Links the
Inflammasome to Acquired Immunity. Cell Metab, 2017. 25(6): p. 1294-1304.e6.
30. Juliano, R.L. and V. Ling, A surface glycoprotein modulating drug permeability in
Chinese hamster ovary cell mutants. Biochim Biophys Acta, 1976. 455(1): p. 152-62.
31. Kartner, N., et al., Daunorubicin-resistant Chinese hamster ovary cells expressing
multidrug resistance and a cell-surface P-glycoprotein. Cancer Res, 1983. 43(9): p.
4413-9.
32. Ling, V., et al., Multidrug-resistance phenotype in Chinese hamster ovary cells. Cancer
Treat Rep, 1983. 67(10): p. 869-74.
33. Candeil, L., et al., ABCG2 overexpression in colon cancer cells resistant to SN38 and in
irinotecan-treated metastases. Int J Cancer, 2004. 109(6): p. 848-54.
34. Kawanishi, T., et al., Bile Duct Obstruction Leads to Increased Intestinal Expression of
Breast Cancer Resistance Protein With Reduced Gastrointestinal Absorption of Imatinib.
J Pharm Sci, 2019. 108(9): p. 3130-3137.
35. Lin, H., et al., KDM5c inhibits multidrug resistance of colon cancer cell line by down-
regulating ABCC1. Biomed Pharmacother, 2018. 107: p. 1205-1209.
Page 87 of 110
36. Spitzwieser, M., et al., Promoter methylation patterns of ABCB1, ABCC1 and ABCG2 in
human cancer cell lines, multidrug-resistant cell models and tumor, tumor-adjacent and
tumor-distant tissues from breast cancer patients. Oncotarget, 2016. 7(45): p. 73347-
73369.
37. Szakács, G., et al., Targeting multidrug resistance in cancer. Nature Reviews Drug
Discovery, 2006. 5(3): p. 219-234.
38. Yamada, A., et al., ABCC1-Exported Sphingosine-1-phosphate, Produced by Sphingosine
Kinase 1, Shortens Survival of Mice and Patients with Breast Cancer. Mol Cancer Res,
2018. 16(6): p. 1059-1070.
39. Gadsby, D.C., P. Vergani, and L. Csanády, The ABC protein turned chloride channel
whose failure causes cystic fibrosis. Nature, 2006. 440(7083): p. 477-83.
40. Ronaldson, P.T., Y. Persidsky, and R. Bendayan, Regulation of ABC membrane
transporters in glial cells: relevance to the pharmacotherapy of brain HIV-1 infection.
Glia, 2008. 56(16): p. 1711-35.
41. Ronaldson, P.T., T. Ashraf, and R. Bendayan, Regulation of multidrug resistance protein
1 by tumor necrosis factor alpha in cultured glial cells: involvement of nuclear factor-
kappaB and c-Jun N-terminal kinase signaling pathways. Mol Pharmacol, 2010. 77(4): p.
644-59.
42. Das, S., et al., O-GlcNAcylation of GLI transcription factors in hyperglycemic conditions
augments Hedgehog activity. Laboratory Investigation, 2019. 99(2): p. 260-270.
43. Singh, R.R., et al., ABCG2 is a direct transcriptional target of hedgehog signaling and
involved in stroma-induced drug tolerance in diffuse large B-cell lymphoma. Oncogene,
2011. 30(49): p. 4874-4886.
Page 88 of 110
44. Po, A., et al., Hedgehog-GLI signalling promotes chemoresistance through the regulation
of ABC transporters in colorectal cancer cells. Scientific Reports, 2020. 10(1): p. 13988.
45. Doyle, L.A., et al., A multidrug resistance transporter from human MCF-7 breast cancer
cells. Proc Natl Acad Sci U S A, 1998. 95(26): p. 15665-70.
46. Miyake, K., et al., Molecular cloning of cDNAs which are highly overexpressed in
mitoxantrone-resistant cells: demonstration of homology to ABC transport genes. Cancer
Res, 1999. 59(1): p. 8-13.
47. Orlando, B.J. and M. Liao, ABCG2 transports anticancer drugs via a closed-to-open
switch. Nature Communications, 2020. 11(1): p. 2264.
48. Rocchi, E., et al., The product of the ABC half-transporter gene ABCG2
(BCRP/MXR/ABCP) is expressed in the plasma membrane. Biochem Biophys Res
Commun, 2000. 271(1): p. 42-6.
49. Biemans-Oldehinkel, E., M.K. Doeven, and B. Poolman, ABC transporter architecture
and regulatory roles of accessory domains. FEBS Lett, 2006. 580(4): p. 1023-35.
50. Ozvegy, C., et al., Functional characterization of the human multidrug transporter,
ABCG2, expressed in insect cells. Biochem Biophys Res Commun, 2001. 285(1): p. 111-
7.
51. Polgar, O., R.W. Robey, and S.E. Bates, ABCG2: structure, function and role in drug
response. Expert Opinion on Drug Metabolism & Toxicology, 2008. 4(1): p. 1-15.
52. Haider, A.J., et al., Dimerization of ABCG2 Analysed by Bimolecular Fluorescence
Complementation. PLOS ONE, 2011. 6(10): p. e25818.
Page 89 of 110
53. Wong, K., et al., Plasma membrane dynamics and tetrameric organisation of ABCG2
transporters in mammalian cells revealed by single particle imaging techniques. Biochim
Biophys Acta, 2016. 1863(1): p. 19-29.
54. Taylor, N.M.I., et al., Structure of the human multidrug transporter ABCG2. Nature,
2017. 546(7659): p. 504-509.
55. Locher, K.P., A.T. Lee, and D.C. Rees, The E. coli BtuCD structure: a framework for
ABC transporter architecture and mechanism. Science, 2002. 296(5570): p. 1091-8.
56. Wakabayashi, K., et al., Identification of cysteine residues critically involved in
homodimer formation and protein expression of human ATP-binding cassette transporter
ABCG2: a new approach using the flp recombinase system. J Exp Ther Oncol, 2006.
5(3): p. 205-22.
57. Kage, K., T. Fujita, and Y. Sugimoto, Role of Cys-603 in dimer/oligomer formation of the
breast cancer resistance protein BCRP/ABCG2. Cancer Sci, 2005. 96(12): p. 866-72.
58. Henriksen, U., et al., Identification of intra- and intermolecular disulfide bridges in the
multidrug resistance transporter ABCG2. J Biol Chem, 2005. 280(44): p. 36926-34.
59. Manolaridis, I., et al., Cryo-EM structures of a human ABCG2 mutant trapped in ATP-
bound and substrate-bound states. Nature, 2018. 563(7731): p. 426-430.
60. Hoque, M.T., K.R. Robillard, and R. Bendayan, Regulation of Breast Cancer Resistant
Protein (BCRP) by Peroxisome Proliferator-Activated Receptor Alpha (PPARα) in
Human Brain Microvessel Endothelial Cells. Molecular Pharmacology, 2012: p.
mol.111.076745.
Page 90 of 110
61. Shin, J.A., et al., Repression of adenosine triphosphate-binding cassette transporter
ABCG2 by estrogen increases intracellular glutathione in brain endothelial cells
following ischemic reperfusion injury. Neurobiol Aging, 2018. 66: p. 138-148.
62. Nakagawa, H., et al., Disruption of N-linked glycosylation enhances ubiquitin-mediated
proteasomal degradation of the human ATP-binding cassette transporter ABCG2. Febs j,
2009. 276(24): p. 7237-52.
63. Campbell, P.K., et al., Identification of a novel, tissue-specific ABCG2 promoter
expressed in pediatric acute megakaryoblastic leukemia. Leuk Res, 2011. 35(10): p.
1321-9.
64. Li, W., et al., Expressions of ABCG2, CD133, and podoplanin in salivary adenoid cystic
carcinoma. Biomed Res Int, 2014. 2014: p. 132349.
65. Reustle, A., et al., Characterization of the breast cancer resistance protein
(BCRP/ABCG2) in clear cell renal cell carcinoma. Int J Cancer, 2018. 143(12): p. 3181-
3193.
66. Prasad, S., et al., Diabetes Mellitus and Blood-Brain Barrier Dysfunction: An Overview. J
Pharmacovigil, 2014. 2(2): p. 125.
67. Sajja, R.K., S. Prasad, and L. Cucullo, Impact of altered glycaemia on blood-brain
barrier endothelium: an in vitro study using the hCMEC/D3 cell line. Fluids Barriers
CNS, 2014. 11(1): p. 8.
68. Sajja, R.K. and L. Cucullo, Altered glycaemia differentially modulates efflux transporter
expression and activity in hCMEC/D3 cell line. Neuroscience letters, 2015. 598: p. 59-65.
Page 91 of 110
69. Zhang, L., et al., Diet-induced obese alters the expression and function of hepatic drug-
metabolizing enzymes and transporters in rats. Biochem Pharmacol, 2019. 164: p. 368-
376.
70. Reichel, V., et al., P-glycoprotein and breast cancer resistance protein expression and
function at the blood-brain barrier and blood-cerebrospinal fluid barrier (choroid
plexus) in streptozotocin-induced diabetes in rats. Brain Res, 2011. 1370: p. 238-45.
71. Anger, G.J., A.M. Cressman, and M. Piquette-Miller, Expression of ABC Efflux
transporters in placenta from women with insulin-managed diabetes. PLoS One, 2012.
7(4): p. e35027.
72. Kushiyama, A., et al., Linking uric acid metabolism to diabetic complications. World J
Diabetes, 2014. 5(6): p. 787-95.
73. Yan, D., et al., Association between serum uric acid related genetic loci and diabetic
kidney disease in the Chinese type 2 diabetes patients. J Diabetes Complications, 2016.
30(5): p. 798-802.
74. Ichida, K., et al., Decreased extra-renal urate excretion is a common cause of
hyperuricemia. Nat Commun, 2012. 3: p. 764.
75. Dehghan, A., et al., Association of three genetic loci with uric acid concentration and
risk of gout: a genome-wide association study. Lancet, 2008. 372(9654): p. 1953-61.
76. Hosomi, A., et al., Extra-renal elimination of uric acid via intestinal efflux transporter
BCRP/ABCG2. PLoS One, 2012. 7(2): p. e30456.
77. Matsuo, H., et al., ABCG2 dysfunction causes hyperuricemia due to both renal urate
underexcretion and renal urate overload. Sci Rep, 2014. 4: p. 3755.
Page 92 of 110
78. Matsuo, H., et al., Common defects of ABCG2, a high-capacity urate exporter, cause
gout: a function-based genetic analysis in a Japanese population. Sci Transl Med, 2009.
1(5): p. 5ra11.
79. Takada, T., et al., ABCG2 dysfunction increases serum uric acid by decreased intestinal
urate excretion. Nucleosides Nucleotides Nucleic Acids, 2014. 33(4-6): p. 275-81.
80. Fehér, Á., et al., Association between the ABCG2 C421A polymorphism and Alzheimer's
disease. Neurosci Lett, 2013. 550: p. 51-4.
81. Retz, W., et al., Free radicals in Alzheimer's disease. J Neural Transm Suppl, 1998. 54:
p. 221-36.
82. Do, T.M., et al., ABCG2- and ABCG4-mediated efflux of amyloid-β peptide 1-40 at the
mouse blood-brain barrier. J Alzheimers Dis, 2012. 30(1): p. 155-66.
83. Tai, L.M., et al., P-glycoprotein and breast cancer resistance protein restrict apical-to-
basolateral permeability of human brain endothelium to amyloid-beta. J Cereb Blood
Flow Metab, 2009. 29(6): p. 1079-83.
84. Shubbar, M.H. and J.I. Penny, Effect of amyloid beta on ATP-binding cassette
transporter expression and activity in porcine brain microvascular endothelial cells.
Biochim Biophys Acta Gen Subj, 2018. 1862(10): p. 2314-2322.
85. Shen, S., et al., ABCG2 reduces ROS-mediated toxicity and inflammation: a potential
role in Alzheimer's disease. J Neurochem, 2010. 114(6): p. 1590-604.
86. Poller, B., et al., Regulation of BCRP (ABCG2) and P-glycoprotein (ABCB1) by
cytokines in a model of the human blood-brain barrier. Cell Mol Neurobiol, 2010. 30(1):
p. 63-70.
Page 93 of 110
87. Zhou, S., et al., The ABC transporter Bcrp1/ABCG2 is expressed in a wide variety of
stem cells and is a molecular determinant of the side-population phenotype. Nat Med,
2001. 7(9): p. 1028-34.
88. Smith, P.J., et al., ABCG2-associated resistance to Hoechst 33342 and topotecan in a
murine cell model with constitutive expression of side population characteristics.
Cytometry A, 2009. 75(11): p. 924-33.
89. Rosen, M.B., et al., ATP Binding Cassette Sub-family Member 2 (ABCG2) and
Xenobiotic Exposure During Early Mouse Embryonic Stem Cell Differentiation. Birth
Defects Res, 2018. 110(1): p. 35-47.
90. Maliepaard, M., et al., Subcellular localization and distribution of the breast cancer
resistance protein transporter in normal human tissues. Cancer Res, 2001. 61(8): p.
3458-64.
91. Liu, W.H., et al., ABCG2 protects kidney side population cells from
hypoxia/reoxygenation injury through activation of the MEK/ERK pathway. Cell
Transplant, 2013. 22(10): p. 1859-68.
92. Lin, Y.H., et al., Protoporphyrin IX accumulation disrupts mitochondrial dynamics and
function in ABCG2-deficient hepatocytes. FEBS Lett, 2013. 587(19): p. 3202-9.
93. Doyle, M.J., et al., Abcg2 labels multiple cell types in skeletal muscle and participates in
muscle regeneration. J Cell Biol, 2011. 195(1): p. 147-63.
94. Dauchy, S., et al., ABC transporters, cytochromes P450 and their main transcription
factors: expression at the human blood-brain barrier. J Neurochem, 2008. 107(6): p.
1518-28.
Page 94 of 110
95. Jonker, J.W., et al., The breast cancer resistance protein BCRP (ABCG2) concentrates
drugs and carcinogenic xenotoxins into milk. Nat Med, 2005. 11(2): p. 127-9.
96. Wu, A.M., et al., Induction of multidrug resistance transporter ABCG2 by prolactin in
human breast cancer cells. Mol Pharmacol, 2013. 83(2): p. 377-88.
97. van Herwaarden, A.E., et al., Breast cancer resistance protein (Bcrp1/Abcg2) reduces
systemic exposure of the dietary carcinogens aflatoxin B1, IQ and Trp-P-1 but also
mediates their secretion into breast milk. Carcinogenesis, 2006. 27(1): p. 123-30.
98. García-Lino, A.M., et al., Transporters in the Mammary Gland-Contribution to Presence
of Nutrients and Drugs into Milk. Nutrients, 2019. 11(10).
99. Woodward, O.M., et al., Identification of a urate transporter, ABCG2, with a common
functional polymorphism causing gout. Proceedings of the National Academy of Sciences
of the United States of America, 2009. 106(25): p. 10338-10342.
100. Jonker, J.W., et al., The breast cancer resistance protein protects against a major
chlorophyll-derived dietary phototoxin and protoporphyria. Proc Natl Acad Sci U S A,
2002. 99(24): p. 15649-54.
101. Wang, P., et al., The essential role of the transporter ABCG2 in the pathophysiology of
erythropoietic protoporphyria. Sci Adv, 2019. 5(9): p. eaaw6127.
102. van Herwaarden, A.E., et al., The breast cancer resistance protein (Bcrp1/Abcg2)
restricts exposure to the dietary carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5-
b]pyridine. Cancer Res, 2003. 63(19): p. 6447-52.
103. Cooray, H.C., et al., Localisation of breast cancer resistance protein in microvessel
endothelium of human brain. NeuroReport, 2002. 13(16): p. 2059-2063.
Page 95 of 110
104. Petropoulos, S., W. Gibb, and S.G. Matthews, Breast cancer-resistance protein (BCRP1)
in the fetal mouse brain: development and glucocorticoid regulation. Biol Reprod, 2011.
84(4): p. 783-9.
105. Traxl, A., et al., Inhibition of ABCB1 and ABCG2 at the Mouse Blood-Brain Barrier with
Marketed Drugs To Improve Brain Delivery of the Model ABCB1/ABCG2 Substrate
[(11)C]erlotinib. Mol Pharm, 2019. 16(3): p. 1282-1293.
106. Scharenberg, C., et al., ABCG2 is expressed in late spermatogenesis and is associated
with the acrosome. Biochem Biophys Res Commun, 2009. 378(2): p. 302-7.
107. Brayboy, L.M., et al., Multidrug resistance transporter-1 and breast cancer resistance
protein protect against ovarian toxicity, and are essential in ovarian physiology. Reprod
Toxicol, 2017. 69: p. 121-131.
108. Langmann, T., et al., Real-time reverse transcription-PCR expression profiling of the
complete human ATP-binding cassette transporter superfamily in various tissues. Clin
Chem, 2003. 49(2): p. 230-8.
109. Ellfolk, M., et al., Placental transporter-mediated drug interactions and offspring
congenital anomalies. British journal of clinical pharmacology, 2020. 86(5): p. 868-879.
110. Evseenko, D.A., et al., The ABC transporter BCRP/ABCG2 is a placental survival factor,
and its expression is reduced in idiopathic human fetal growth restriction. Faseb j, 2007.
21(13): p. 3592-605.
111. Petrovic, V., et al., Maternal bacterial infections impact expression of drug transporters
in human placenta. Int Immunopharmacol, 2015. 26(2): p. 349-56.
112. Lye, P., et al., Glucocorticoids modulate multidrug resistance transporters in the first
trimester human placenta. J Cell Mol Med, 2018. 22(7): p. 3652-3660.
Page 96 of 110
113. Zhou, S., et al., Bcrp1 gene expression is required for normal numbers of
side population stem cells in mice, and confers relative protection to mitoxantrone in
hematopoietic cells in vivo. Proceedings of the National Academy of
Sciences, 2002. 99(19): p. 12339-12344.
114. Zhou, S., et al., Increased expression of the Abcg2 transporter during erythroid
maturation plays a role in decreasing cellular protoporphyrin IX levels. Blood, 2005.
105(6): p. 2571-6.
115. Zong, Y., et al., Expression of mouse Abcg2 mRNA during hematopoiesis is regulated by
alternative use of multiple leader exons and promoters. J Biol Chem, 2006. 281(40): p.
29625-32.
116. Sereti, K.I., et al., ATP-binding cassette G-subfamily transporter 2 regulates cell cycle
progression and asymmetric division in mouse cardiac side population progenitor cells.
Circ Res, 2013. 112(1): p. 27-34.
117. Homolya, L., et al., Mitoxantrone is expelled by the ABCG2 multidrug transporter
directly from the plasma membrane. Biochim Biophys Acta, 2011. 1808(1): p. 154-63.
118. Ando, T., et al., Involvement of breast cancer resistance protein (ABCG2) in the biliary
excretion mechanism of fluoroquinolones. Drug Metab Dispos, 2007. 35(10): p. 1873-9.
119. Kitamura, S., et al., Involvement of multiple transporters in the hepatobiliary transport of
rosuvastatin. Drug Metab Dispos, 2008. 36(10): p. 2014-23.
120. Mizuno, T., et al., Impact of genetic variation in breast cancer resistance protein
(BCRP/ABCG2) on sunitinib pharmacokinetics. Drug Metab Pharmacokinet, 2012. 27(6):
p. 631-9.
Page 97 of 110
121. Safar, Z., et al., ABCG2/BCRP: variants, transporter interaction profile of substrates and
inhibitors. Expert Opin Drug Metab Toxicol, 2019. 15(4): p. 313-328.
122. Vlaming, M.L., et al., Functionally overlapping roles of Abcg2 (Bcrp1) and Abcc2
(Mrp2) in the elimination of methotrexate and its main toxic metabolite 7-
hydroxymethotrexate in vivo. Clin Cancer Res, 2009. 15(9): p. 3084-93.
123. Zaher, H., et al., Breast cancer resistance protein (Bcrp/abcg2) is a major determinant of
sulfasalazine absorption and elimination in the mouse. Mol Pharm, 2006. 3(1): p. 55-61.
124. Brechbuhl, H.M., et al., Glutathione transport is a unique function of the ATP-binding
cassette protein ABCG2. J Biol Chem, 2010. 285(22): p. 16582-7.
125. Leimanis, M.L. and E. Georges, ABCG2 membrane transporter in mature human
erythrocytes is exclusively homodimer. Biochemical and Biophysical Research
Communications, 2007. 354(2): p. 345-350.
126. Sturek, J.M., et al., An intracellular role for ABCG1-mediated cholesterol transport in
the regulated secretory pathway of mouse pancreatic beta cells. J Clin Invest, 2010.
120(7): p. 2575-89.
127. Tarling, E.J. and P.A. Edwards, ATP binding cassette transporter G1 (ABCG1) is an
intracellular sterol transporter. Proceedings of the National Academy of Sciences, 2011.
108(49): p. 19719.
128. Tarr, P.T., et al., Emerging new paradigms for ABCG transporters. Biochim Biophys
Acta, 2009. 1791(7): p. 584-93.
129. Dergunov, A.D., et al., Significance of Cholesterol-Binding Motifs in ABCA1, ABCG1,
and SR-B1 Structure. J Membr Biol, 2019. 252(1): p. 41-60.
Page 98 of 110
130. Gelissen, I.C., et al., Protein kinase A modulates the activity of a major human isoform of
ABCG1. J Lipid Res, 2012. 53(10): p. 2133-40.
131. Klucken, J., et al., ABCG1 (ABC8), the human homolog of the Drosophila white gene, is
a regulator of macrophage cholesterol and phospholipid transport. Proc Natl Acad Sci U
S A, 2000. 97(2): p. 817-22.
132. Wang, N., et al., ATP-binding cassette transporters G1 and G4 mediate cellular
cholesterol efflux to high-density lipoproteins. Proceedings of the National Academy of
Sciences of the United States of America, 2004. 101(26): p. 9774.
133. Kennedy, M.A., et al., ABCG1 has a critical role in mediating cholesterol efflux to HDL
and preventing cellular lipid accumulation. Cell Metab, 2005. 1(2): p. 121-31.
134. Neufeld, E.B., et al., Cellular Localization and Trafficking of the Human ABCG1
Transporter. Biology (Basel), 2014. 3(4): p. 781-800.
135. Pandzic, E., et al., The ATP binding cassette transporter, ABCG1, localizes to cortical
actin filaments. Scientific Reports, 2017. 7(1): p. 42025.
136. Harris, M.T., et al., Reinterpretation of the localization of the ATP binding cassette
transporter ABCG1 in insulin-secreting cells and insights regarding its trafficking and
function. PLoS One, 2018. 13(9): p. e0198383.
137. Hussain, S.S., et al., Control of insulin granule formation and function by the ABC
transporters ABCG1 and ABCA1 and by oxysterol binding protein OSBP. Mol Biol Cell,
2018. 29(10): p. 1238-1257.
138. Aye, I.L., et al., Placental ABCA1 and ABCG1 transporters efflux cholesterol and protect
trophoblasts from oxysterol induced toxicity. Biochim Biophys Acta, 2010. 1801(9): p.
1013-24.
Page 99 of 110
139. Baldán, A., et al., Loss of ABCG1 results in chronic pulmonary inflammation. J Immunol,
2008. 180(5): p. 3560-8.
140. Baldán, A., et al., Deletion of the transmembrane transporter ABCG1 results in
progressive pulmonary lipidosis. J Biol Chem, 2006. 281(39): p. 29401-10.
141. Cheng, H.Y., et al., Loss of ABCG1 influences regulatory T cell differentiation and
atherosclerosis. J Clin Invest, 2016. 126(9): p. 3236-46.
142. Gelissen, I.C., et al., Expression and stability of two isoforms of ABCG1 in human
vascular cells. Atherosclerosis, 2010. 208(1): p. 75-82.
143. Hoekstra, M., et al., ATP-binding cassette transporter G1 deficiency is associated with
mild glucocorticoid insufficiency in mice. Biochim Biophys Acta Mol Cell Biol Lipids,
2019. 1864(4): p. 443-451.
144. Jiang, Y.J., et al., Regulation of ABCG1 expression in human keratinocytes and murine
epidermis. J Lipid Res, 2010. 51(11): p. 3185-95.
145. Lou, J., et al., ABCA1 and ABCG1 expression in the small intestine of type 2 diabetic
rats. Lab Med, 2014. 45(1): p. 17-24.
146. Mani, O., et al., Identification of ABCA1 and ABCG1 in milk fat globules and mammary
cells--implications for milk cholesterol secretion. J Dairy Sci, 2011. 94(3): p. 1265-76.
147. Palme, N., et al., Immunolocalization of the cholesterol transporters ABCA1 and ABCG1
in canine reproductive tract tissues and spermatozoa. Reprod Domest Anim, 2014. 49(3):
p. 441-7.
148. Ranalletta, M., et al., Decreased atherosclerosis in low-density lipoprotein receptor
knockout mice transplanted with Abcg1-/- bone marrow. Arterioscler Thromb Vasc Biol,
2006. 26(10): p. 2308-15.
Page 100 of 110
149. Storti, F., et al., Impaired ABCA1/ABCG1-mediated lipid efflux in the mouse retinal
pigment epithelium (RPE) leads to retinal degeneration. Elife, 2019. 8.
150. Tarr, P.T. and P.A. Edwards, ABCG1 and ABCG4 are coexpressed in neurons and
astrocytes of the CNS and regulate cholesterol homeostasis through SREBP-2. J Lipid
Res, 2008. 49(1): p. 169-82.
151. Wang, N., et al., ATP-binding cassette transporters G1 and G4 mediate cholesterol and
desmosterol efflux to HDL and regulate sterol accumulation in the brain. Faseb j, 2008.
22(4): p. 1073-82.
152. Wei, H., et al., ABCG1 regulates mouse adipose tissue macrophage cholesterol levels
and ratio of M1 to M2 cells in obesity and caloric restriction. J Lipid Res, 2015. 56(12):
p. 2337-47.
153. McPeek, M., et al., Alveolar Macrophage ABCG1 Deficiency Promotes Pulmonary
Granulomatous Inflammation. Am J Respir Cell Mol Biol, 2019. 61(3): p. 332-340.
154. Lorkowski, S., et al., Expression of the ATP-binding cassette transporter gene ABCG1
(ABC8) in Tangier disease. Biochem Biophys Res Commun, 2001. 283(4): p. 821-30.
155. Chistiakov, D.A., et al., Mechanisms of foam cell formation in atherosclerosis. J Mol
Med (Berl), 2017. 95(11): p. 1153-1165.
156. Jiang, T., et al., Leonurine Prevents Atherosclerosis Via Promoting the Expression of
ABCA1 and ABCG1 in a Pparγ/Lxrα Signaling Pathway-Dependent Manner. Cell
Physiol Biochem, 2017. 43(4): p. 1703-1717.
157. Ren, K., et al., Mangiferin promotes macrophage cholesterol efflux and protects against
atherosclerosis by augmenting the expression of ABCA1 and ABCG1. Aging (Albany
NY), 2019. 11(23): p. 10992-11009.
Page 101 of 110
158. Tian, C., et al., ABCG1 as a potential oncogene in lung cancer. Exp Ther Med, 2017.
13(6): p. 3189-3194.
159. Sag, D., et al., The cholesterol transporter ABCG1 links cholesterol homeostasis and
tumour immunity. Nat Commun, 2015. 6: p. 6354.
160. Armstrong, A.J., et al., ATP-binding cassette transporter G1 negatively regulates
thymocyte and peripheral lymphocyte proliferation. J Immunol, 2010. 184(1): p. 173-83.
161. Pan, H., et al., Expression of LXR‑β, ABCA1 and ABCG1 in human triple‑negative breast
cancer tissues. Oncol Rep, 2019. 42(5): p. 1869-1877.
162. Roundhill, E.A., S. Jabri, and S.A. Burchill, ABCG1 and Pgp identify drug resistant, self-
renewing osteosarcoma cells. Cancer Lett, 2019. 453: p. 142-157.
163. Qie, R., et al., Association of ABCG1 gene methylation and its dynamic change status
with incident type 2 diabetes mellitus: the Rural Chinese Cohort Study. Journal of Human
Genetics, 2020.
164. Allen, J.D., et al., Potent and specific inhibition of the breast cancer resistance protein
multidrug transporter in vitro and in mouse intestine by a novel analogue of
fumitremorgin C. Mol Cancer Ther, 2002. 1(6): p. 417-25.
165. Rabindran, S.K., et al., Reversal of a novel multidrug resistance mechanism in human
colon carcinoma cells by fumitremorgin C. Cancer Res, 1998. 58(24): p. 5850-8.
166. Rabindran, S.K., et al., Fumitremorgin C reverses multidrug resistance in cells
transfected with the breast cancer resistance protein. Cancer Res, 2000. 60(1): p. 47-50.
167. Weidner, L.D., et al., The Inhibitor Ko143 Is Not Specific for ABCG2. J Pharmacol Exp
Ther, 2015. 354(3): p. 384-93.
Page 102 of 110
168. Mairinger, S., et al., Influence of breast cancer resistance protein and P-glycoprotein on
tissue distribution and excretion of Ko143 assessed with PET imaging in mice. European
journal of pharmaceutical sciences : official journal of the European Federation for
Pharmaceutical Sciences, 2018. 115: p. 212-222.
169. Gupta, P., et al., Voruciclib, a Potent CDK4/6 Inhibitor, Antagonizes ABCB1 and
ABCG2-Mediated Multi-Drug Resistance in Cancer Cells. Cell Physiol Biochem, 2018.
45(4): p. 1515-1528.
170. Ji, N., et al., Selonsertib (GS-4997), an ASK1 inhibitor, antagonizes multidrug resistance
in ABCB1- and ABCG2-overexpressing cancer cells. Cancer Lett, 2019. 440-441: p. 82-
93.
171. Ji, N., et al., Ulixertinib (BVD-523) antagonizes ABCB1- and ABCG2-mediated
chemotherapeutic drug resistance. Biochem Pharmacol, 2018. 158: p. 274-285.
172. Zhang, H., et al., Linsitinib (OSI-906) antagonizes ATP-binding cassette subfamily G
member 2 and subfamily C member 10-mediated drug resistance. Int J Biochem Cell
Biol, 2014. 51: p. 111-9.
173. Huang, C., et al., Coadministration of vindesine with high-dose methotrexate therapy
increases acute kidney injury via BCRP, MRP2, and OAT1/OAT3. Cancer Chemother
Pharmacol, 2020. 85(2): p. 433-441.
174. Zhang, G.N., et al., Lipid-Saporin Nanoparticles for the Intracellular Delivery of
Cytotoxic Protein to Overcome ABC Transporter-Mediated Multidrug Resistance In
Vitro and In Vivo. Cancers (Basel), 2020. 12(2).
Page 103 of 110
175. Zhang, X., et al., Fasudil increases temozolomide sensitivity and suppresses
temozolomide-resistant glioma growth via inhibiting ROCK2/ABCG2. Cell Death &
Disease, 2018. 9(2): p. 190.
176. Pranatharthi, A., C. Ross, and S. Srivastava, Cancer Stem Cells and Radioresistance:
Rho/ROCK Pathway Plea Attention. Stem Cells Int, 2016. 2016: p. 5785786.
177. Arumugam, A., et al., Silencing growth hormone receptor inhibits estrogen receptor
negative breast cancer through ATP-binding cassette sub-family G member 2.
Experimental & Molecular Medicine, 2019. 51(1): p. 1-13.
178. Yang, H., et al., Extracellular ATP promotes breast cancer invasion and chemoresistance
via SOX9 signaling. Oncogene, 2020. 39(35): p. 5795-5810.
179. Kawabata, K.C., et al., High expression of ABCG2 induced by EZH2 disruption has
pivotal roles in MDS pathogenesis. Leukemia, 2018. 32(2): p. 419-428.
180. Ehata, S., et al., Transforming growth factor-β decreases the cancer-initiating cell
population within diffuse-type gastric carcinoma cells. Oncogene, 2011. 30(14): p. 1693-
1705.
181. Rosenfeldt, M.T., et al., E2F1 drives chemotherapeutic drug resistance via ABCG2.
Oncogene, 2014. 33(32): p. 4164-4172.
182. Ding, J., et al., Hedgehog signaling pathway affects the sensitivity of hepatoma cells to
drug therapy through the ABCC1 transporter. Laboratory Investigation, 2017. 97(7): p.
819-832.
183. Nasr, R., et al., Molecular analysis of the massive GSH transport mechanism mediated by
the human Multidrug Resistant Protein 1/ABCC1. Scientific Reports, 2020. 10(1): p.
7616.
Page 104 of 110
184. Kranz, J., et al., The role of the efflux carriers Abcg2 and Abcc2 for the hepatobiliary
elimination of benzo[a]pyrene and its metabolites in mice. Chem Biol Interact, 2014.
224: p. 36-41.
185. Liang, Z., et al., CYP27A1 inhibits bladder cancer cells proliferation by regulating
cholesterol homeostasis. Cell Cycle, 2019. 18(1): p. 34-45.
186. Lambert, J.P., et al., Proximity biotinylation and affinity purification are complementary
approaches for the interactome mapping of chromatin-associated protein complexes. J
Proteomics, 2015. 118: p. 81-94.
187. Roux, K.J., et al., A promiscuous biotin ligase fusion protein identifies proximal and
interacting proteins in mammalian cells. Journal of Cell Biology, 2012. 196(6): p. 801-
810.
188. Kim, J., et al., Use of in vivo biotinylation to study protein–protein and protein–DNA
interactions in mouse embryonic stem cells. Nature Protocols, 2009. 4(4): p. 506-517.
189. Mehus, A.A., R.H. Anderson, and K.J. Roux, BioID Identification of Lamin-Associated
Proteins. Methods Enzymol, 2016. 569: p. 3-22.
190. Schweingruber, C., et al., Identification of Interactions in the NMD Complex Using
Proximity-Dependent Biotinylation (BioID). PLoS One, 2016. 11(3): p. e0150239.
191. Searle, B.C., Scaffold: a bioinformatic tool for validating MS/MS-based proteomic
studies. Proteomics, 2010. 10(6): p. 1265-9.
192. Shannon, P., et al., Cytoscape: a software environment for integrated models of
biomolecular interaction networks. Genome Res, 2003. 13(11): p. 2498-504.
193. Consortium, T.U., UniProt: a worldwide hub of protein knowledge. Nucleic Acids
Research, 2018. 47(D1): p. D506-D515.
Page 105 of 110
194. Khunweeraphong, N., T. Stockner, and K. Kuchler, The structure of the human ABC
transporter ABCG2 reveals a novel mechanism for drug extrusion. Scientific Reports,
2017. 7(1): p. 13767.
195. Cui, J., et al., New ABCG2 homology model reveals importance of C-terminus in protein
trafficking. The FASEB Journal, 2017. 31(S1): p. 603.24-603.24.
196. Takada, T., H. Suzuki, and Y. Sugiyama, Characterization of polarized expression of
point- or deletion-mutated human BCRP/ABCG2 in LLC-PK1 cells. Pharm Res, 2005.
22(3): p. 458-64.
197. Gelissen, I.C., et al., ABCA1 and ABCG1 synergize to mediate cholesterol export to
apoA-I. Arterioscler Thromb Vasc Biol, 2006. 26(3): p. 534-40.
198. Hori, N., H. Hayashi, and Y. Sugiyama, Calpain-mediated cleavage negatively regulates
the expression level of ABCG1. Atherosclerosis, 2011. 215(2): p. 383-91.
199. Hsieh, V., et al., Cellular cholesterol regulates ubiquitination and degradation of the
cholesterol export proteins ABCA1 and ABCG1. J Biol Chem, 2014. 289(11): p. 7524-36.
200. Kobayashi, A., et al., Efflux of sphingomyelin, cholesterol, and phosphatidylcholine by
ABCG1. J Lipid Res, 2006. 47(8): p. 1791-802.
201. Mauerer, R., S. Ebert, and T. Langmann, High glucose, unsaturated and saturated fatty
acids differentially regulate expression of ATP-binding cassette transporters ABCA1 and
ABCG1 in human macrophages. Experimental & Molecular Medicine, 2009. 41(2): p.
126-132.
202. Ogura, M., et al., Proteasomal inhibition promotes ATP-binding cassette transporter A1
(ABCA1) and ABCG1 expression and cholesterol efflux from macrophages in vitro and in
vivo. Arterioscler Thromb Vasc Biol, 2011. 31(9): p. 1980-7.
Page 106 of 110
203. Matsumoto, H., H. Haniu, and N. Komori, Determination of Protein Molecular Weights
on SDS-PAGE. Methods Mol Biol, 2019. 1855: p. 101-105.
204. Kobuchi, H., et al., Mitochondrial localization of ABC transporter ABCG2 and its
function in 5-aminolevulinic acid-mediated protoporphyrin IX accumulation. PLoS One,
2012. 7(11): p. e50082.
205. Wang, Z., et al., Caveolin-1 mediates chemoresistance in breast cancer stem cells via β-
catenin/ABCG2 signaling pathway. Carcinogenesis, 2014. 35(10): p. 2346-56.
206. Chen, B., et al., Upregulation of miR-199a/b contributes to cisplatin resistance via
Wnt/β-catenin-ABCG2 signaling pathway in ALDHA1(+) colorectal cancer stem cells.
Tumour Biol, 2017. 39(6): p. 1010428317715155.
207. Yamada, N., et al., Tumor-suppressive microRNA-145 targets catenin δ-1 to regulate
Wnt/β-catenin signaling in human colon cancer cells. Cancer Lett, 2013. 335(2): p. 332-
42.
208. Cao, N., et al., MicroRNA-298 represses hepatocellular carcinoma progression by
inhibiting CTNND1-mediated Wnt/β-catenin signaling. Biomed Pharmacother, 2018.
106: p. 483-490.
209. Tang, B., et al., Overexpression of CTNND1 in hepatocellular carcinoma promotes
carcinous characters through activation of Wnt/β-catenin signaling. J Exp Clin Cancer
Res, 2016. 35(1): p. 82.
210. Takada, T., et al., Regulation of the cell surface expression of human BCRP/ABCG2 by
the phosphorylation state of Akt in polarized cells. Drug Metab Dispos, 2005. 33(7): p.
905-9.
Page 107 of 110
211. Wang, L., N. Lin, and Y. Li, The PI3K/AKT signaling pathway regulates ABCG2
expression and confers resistance to chemotherapy in human multiple myeloma. Oncol
Rep, 2019. 41(3): p. 1678-1690.
212. Hegedüs, C., et al., PI3-kinase and mTOR inhibitors differently modulate the function of
the ABCG2 multidrug transporter. Biochem Biophys Res Commun, 2012. 420(4): p.
869-74.
213. Huang, J. and B.D. Manning, The TSC1-TSC2 complex: a molecular switchboard
controlling cell growth. Biochem J, 2008. 412(2): p. 179-90.
214. Jiang, M., et al., The TSC1-mTOR-PLK axis regulates the homeostatic switch from
Schwann cell proliferation to myelination in a stage-specific manner. Glia, 2018. 66(9):
p. 1947-1959.
215. Jing, W., et al., KD025, an anti-adipocyte differentiation drug, enhances the efficacy of
conventional chemotherapeutic drugs in ABCG2-overexpressing leukemia cells. Oncol
Lett, 2020. 20(6): p. 309.
216. Du, W., et al., miR-335-5p inhibits TGF-β1-induced epithelial-mesenchymal transition in
non-small cell lung cancer via ROCK1. Respir Res, 2019. 20(1): p. 225.
217. Hu, C., et al., ROCK1 promotes migration and invasion of non‑small‑cell lung cancer
cells through the PTEN/PI3K/FAK pathway. Int J Oncol, 2019. 55(4): p. 833-844.
218. Liang, H., et al., LncRNA DANCR promotes cervical cancer progression by upregulating
ROCK1 via sponging miR-335-5p. J Cell Physiol, 2019. 234(5): p. 7266-7278.
219. Lazarini, M., et al., ARHGAP21 is a RhoGAP for RhoA and RhoC with a role in
proliferation and migration of prostate adenocarcinoma cells. Biochim Biophys Acta,
2013. 1832(2): p. 365-74.
Page 108 of 110
220. Rosa, L.R.O., et al., ARHGAP21 as a master regulator of multiple cellular processes. J
Cell Physiol, 2018. 233(11): p. 8477-8481.
221. Chua, V.Y., et al., Activity of ABCG2 Is Regulated by Its Expression and Localization in
DHT and Cyclopamine-Treated Breast Cancer Cells. J Cell Biochem, 2016. 117(10): p.
2249-59.
222. Ferdinandusse, S., et al., A novel bile acid biosynthesis defect due to a deficiency of
peroxisomal ABCD3. Hum Mol Genet, 2015. 24(2): p. 361-70.
223. Violante, S., et al., Peroxisomes can oxidize medium- and long-chain fatty acids through
a pathway involving ABCD3 and HSD17B4. Faseb j, 2019. 33(3): p. 4355-4364.
Page 109 of 110