TICB 1586 No.

of Pages 15

Trends in Cell Biology

Review

Cholesterol Handling in Lysosomes

and Beyond
Ying Meng,1,3 Saskia Heybrock,2,3 Dante Neculai,1 and Paul Saftig2,*,@

Lysosomes are of major importance for the regulation of cellular cholesterol ho- Highlights
meostasis. Food-derived cholesterol and cholesterol esters contained within li- Cholesterol is delivered to lysosomes by
poproteins are delivered to lysosomes by endocytosis. From the lysosomal receptor-mediated delivery of low den-
sity lipoproteins (LDLs) from the extracel-
lumen, cholesterol is transported to the inner surface of the lysosomal membrane
lular space.
through the glycocalyx; this shuttling requires Niemann–Pick C (NPC) 1 and
NPC2 proteins. The lysosomal membrane proteins lysosomal-associated mem- In lysosomes, cholesterol esters are hy-
brane protein (LAMP)-2 and lysosomal integral membrane protein (LIMP)-2/ drolyzed and subject to binding to spe-
cialized proteins such as NPC2.
SCARB2 also bind cholesterol. LAMP-2 may serve as a cholesterol reservoir,
whereas LIMP-2, like NPC1, is able to transport cholesterol through a The transport of cholesterol to the limiting
transglycocalyx tunnel. Contact sites and fusion events between lysosomes membrane of the lysosome requires that
and other organelles mediate the distribution of cholesterol. Lysosomal choles- the dense layer of carbohydrates has to
be overcome by hydrophobic tunnel sys-
terol content is sensed thereby regulating mammalian target of rapamycin com- tems found in NPC1 as well as in LIMP-
plex (mTORC)-dependent signaling. This review summarizes our understanding 2/SCARB2.
of the major steps in cholesterol handling from the moment it enters the lyso-
some until it leaves this compartment. At the cytosolic side of the limiting mem-
brane of the lysosome, protein interac-
tions between lysosomal membrane
Lysosomes Control Cholesterol Homeostasis proteins and organelle membrane pro-
Lysosomes are cellular organelles involved in degradation processes, membrane flow, and teins mediate membrane contact sites
and transfer of cholesterol.
cell signaling [1,2]. Deficiency of lysosomal proteins is associated with human diseases af-
fecting the visceral and nervous systems. In many cases, storage of undigested or improp- The lysosomal cholesterol efflux is
erly transported substrates causes tissue dysfunction and cell death [3]. Niemann–Pick sensed and translated to the regulation
type C disease (NPCD) (Box 1) is an example of such a fatal disorder associated with a of cell proliferation and autophagy.

pathological accumulation of cholesterol and secondary storage products in lysosomes

[4,5].

Cholesterol, as a major component of cell membranes, can either be generated endogenously or

can mainly be acquired by cells via uptake of low-density lipoproteins (LDLs; see Glossary)
(Figure 1) by LDL-receptor (LDLR)-mediated endocytosis [6]. Intracellular cholesterol distribution
and homeostasis can be followed biochemically and in cell-based assays using labelled choles-
terol/lipid compounds (Table 1). The LDL-embedded cholesterol follows the endocytic route to
reach lysosomes, where it is liberated. The processes leading to the delivery of cholesterol to
the limiting membrane of the lysosome, which separates the acidic lumen from the cytosol,
remained enigmatic for a long time. Only recently, new insights were obtained after having re- 1
Department of Cell Biology, School of
solved the structures of lysosomal membrane proteins directly involved in this cholesterol
Basic Medical Sciences, Zhejiang
transport process. The identification of hydrophobic tunnel domains in these proteins revealed University, Hangzhou, Zhejiang, China
2
how cholesterol can be transported [7,8] (Figure 2). Lysosomes can physically interact with the Biochemisches Institut, Christian-
Albrechts-Universität Kiel, Kiel, Germany
endoplasmic reticulum, peroxisomes, trans-Golgi network, and mitochondria, mediating the 3
These authors contributed equally
dispersal of cholesterol within the cell [9,10]. Cholesterol levels are sensed at the lysosomal
membrane [11,12]. Cholesterol sensing by lysosomes regulates cellular signaling through the
*Correspondence:
mammalian target of rapamycin complex (mTORC), as well as cell proliferation and autophagy. [email protected] (P. Saftig).
@
Here, we summarize the most current view of how cholesterol is handled in lysosomes and Twitter: @SaftigPaul (P. Saftig).

Trends in Cell Biology, Month 2020, Vol. xx, No. xx https://doi.org/10.1016/j.tcb.2020.02.007 1

© 2020 Elsevier Ltd. All rights reserved.
Trends in Cell Biology

Box 1. Lysosomal Acid Lipase (LAL) Deficiency and NPCD

Glossary
Mutations in the genes encoding NPC1, NPC2, and LAL can cause cholesterol storage in lysosomes leading to systemic
Glycocalyx: a membrane domain rich
and neurological symptoms in patients. NPC1/2 (OMIM #277220 – NPC1, OMIM#607625 – NPC2) are rare autosomal
in glycoproteins and glycolipids. At the
recessive disorders with an impaired activity of NPC1 and NPC2, respectively [5]. More than 300 disease-causing muta-
cell surface it contributes to cell–cell
tions in the NPC1 gene are described. NPC2 mutations are less frequent and only 5% of all cases of NPC are due to NPC2.
recognition, whereas at the inner surface
In both cases, unesterified cholesterol is the primary lysosomal storage product. However, storage of sphingomyelin, bis-
of lysosomes this dense layer of
(monoacylglycero)phosphate, free sphingoid bases, and glycosphingolipids has been described [80]. The storage may
carbohydrates is thought to be
also cause abnormal lysosomal calcium levels [81]. NPCD is a rare disorder that is characterized by liver, lung, spleen,
protective against lysosomal hydrolase
and progressive central nervous pathology and neurodegeneration, usually followed by premature death of the patient.
attack.
Foam cell infiltrates, ballooned neurons with stored lipids, meganeurite formation, neuroaxonal dystrophy, and neuronal
Low-density lipoproteins (LDLs):
death are cellular hallmarks of NPCD. Although the majority of reports favor the role of NPC2 and NPC1 in the sequential
involved in carrying lipids (fatty acids,
transfer of cholesterol from the lysosomal lumen to the limiting membrane, there is also a discussion that cholesterol ac-
triglycerides, and esterified and
cumulation affects cellular vesicle trafficking, further accelerating the cellular pathology and contributing to secondary stor-
unesterified cholesterol) to cells.
age. The lysosomal luminal part of NPC1 is also used by the Ebola virus for its intracellular entry [82]. NPCD can be treated
Through apoprotein B-100 they can be
with the small iminosugar misglustat reducing the accumulation of glycosphingolipid storage. Drugs that aim to stabilize
taken up by cells by LDLR through
the lysosomal membrane inducing the lysosomal stress response as well as cholesterol complexing cyclodextrins are un-
receptor-mediated endocytosis.
dergoing clinical trials for NPCD.
Lysosomal membrane proteins: a
class of around 200 proteins residing in
LAL deficiency is designated depending on the disease phenotype Wolman disease (WD) or cholesteryl ester storage dis-
the limiting membrane of the lysosome.
ease CESD (OMIM #278000) [83]. In WD there is a complete absence of acid lipase activity and a rapidly progressing dis-
At the luminal side they can contribute to
ease by failure to thrive, diarrhea, calcification of adrenals, hepatosplenomegaly, and death by 6–8 months. By contrast, in
the formation of a dense glycan layer. At
CESD a partially active enzyme leads to a more attenuated disease characterized by liver enlargement, short stature, ane-
the cytosolic side they can be involved in
mia, pain, and gastrointestinal bleeding. The reduction in LAL leads to cholesteryl ester and triglyceride storage, which is
cell signaling, fusion events, membrane
most pronounced in hepatocytes, intestine, macrophages as well as adrenal glands.
contact formation, and motility of
lysosomes. They can act as a scaffold or
transporter of proteins.
speculate how cholesterol reaches the limiting membrane of lysosomes. We also refer to useful re- Macroautophagy: is a central cellular
cent reviews discussing the fate and transport of intracellular cholesterol [12–14]. pathway where the cell’s own material is
enwrapped by autophagosomes.
Autophagosomes fuse with lysosomes
The Lysosome as a Lipid Turnover Location wherein the cellular contents are
Lysosomes are usually small digestive organelles with variable shape and a surrounding degraded by acid hydrolases.
Mammalian target of rapamycin
membrane that contains heavily glycosylated membrane proteins [1]. As a result, a dense layer, complex (mTORC): a huge protein
glycocalyx, with a thickness of about 8 nm at the inner (luminal) surface of the membrane is complex involved in nutrient sensing. It
formed [15]. The role of the glycocalyx is to protect the lysosomal membrane from degradation acts as a kinase and can be activated at
the cytosolic surface of the lysosomal
by the luminal acid hydrolases [16]. Consequently, mechanisms had to evolve to allow for efficient
membrane. mTORC controls protein
transfer of metabolites produced in the lysosomal lumen through the glycocalyx. Inside lyso- synthesis and autophagic degradation.
somes, an array of about 60 lysosomal hydrolases and some accessory proteins mediate the hy- Resistance-nodulation-division
drolysis of different types of macromolecules [3]. Lipids are also degraded here and a defective (RND) proteins: large bacterial efflux
transporters involved in the
catabolism of lipids and other macromolecules causes typical lysosomal storage diseases. Lyso-
detoxification and export of virulence
somes and late endosomes, although representing different maturation steps in the endocytic factors. They have a large substrate
compartment share a number of features including major lysosomal membrane proteins and a spectrum ranging from metal ions,
hydrolase-rich lumen. In most cases we use the term late endosomes/lysosomes (LELs) to drugs, to hydrophobic and amphiphilic
compounds.
refer to both organelles.

Lysosomal lipid handling and more specifically the efflux machinery of cholesterol in this compart-
ment are key to the understanding of the regulation of endogenous cholesterol synthesis and ex-
ogenous cholesterol uptake. In lysosomes the LDL-derived cholesteryl ester and triglycerides are
hydrolyzed to cholesterol and fatty acids by lysosomal acid lipase (LAL) [17]. Macroautophagy
is an independent cellular pathway that can also deliver endogenous cholesterol from cellular
membranes to lysosomes through fusion with autophagosomes. The endocytic delivery and au-
tophagy pathway generate intraluminal membrane-rich vesicles from where lipids, including cho-
lesterol, require transport to the limiting membrane of the lysosome. It is important to note that
cholesterol can not only immediately be subject to the mentioned lysosomal egress pathway: it
may be modified in the lysosomal lumen by the acid hydrolase β-glucocerebrosidase, which
can also act as a glucosyltransferase. Although the (patho)physiological relevance of such

2 Trends in Cell Biology, Month 2020, Vol. xx, No. xx

Trends in Cell Biology

Nutrition

LDL

LDLR
LAMP2

NPC1
EE
NPC2

G H+

Saps
LE LAL
M H+
S2P H+
S1P
H+
nSREBP2

LD LIMP-2
ACAT SREBP2 H+
SCAP LY
ER INSIG H+
P

HMGCR

LDLR, HMGCR Autophagosome

Nu

Trends in Cell Biology

Figure 1. The Endocytic Pathway Delivers Cholesterol to Lysosomes. Cholesterol derived from nutritional sources is
found in lipoproteins, which include low-density lipoproteins (LDLs). They are recognized in the majority of cell types by LDL
receptors (LDLRs) triggering endocytosis and formation of early endosomes (EEs), from where LDLR is recycled to the
plasma membrane. After maturation into late endosomes (LEs), acidification and further maturation into lysosomes, LDL is
degraded and cholesteryl esters are found in the lumen of the lysosome (LY). Acid lipase (LAL) hydrolyzes these esters
and generates free cholesterol. After binding to lysosomal-associated membrane protein 2 (LAMP-2), Niemann–Pick C2
(NPC2), or possibly saposins (Saps), it can be transferred to NPC1 and lysosomal integral membrane protein 2 (LIMP-2),
respectively. Transfer of cholesterol through hydrophobic tunnel cavities spanning the glycocalyx facilitates delivery of the
lipid to the lysosomal-limiting membrane. Membrane contact sites mediate cholesterol delivery to other organelles
including the endoplasmic reticulum (ER) and peroxisomes (P). Through the activity of acetyl-CoA acetyltransferase 1
(ACAT1), cholesterol esters are formed and stored in lipid droplets (LDs). At the ER, cholesterol is sensed by the sterol
regulatory element-binding protein cleavage-activating protein (SCAP) and the insulin-induced gene protein (INSIG),
leading to transport of sterol regulatory element-binding protein (SREBP2) to the cis-Golgi (G) where it is processed by
Site1 and Site2 proteases (S1P and S2P), which lead to the liberation of the nuclear (n) fragment of SREBP2 activating
nuclear gene expression of the LDLR and the 3-hydoxy-3-methylglataryl-coenzyme-A reductase (HMGCR). The turnover
of intracellular membranes by autophagy also leads to delivery of cholesterol to lysosomes. For reasons of clarity this
figure only illustrates possible distribution routes of cholesterol after export from LYs as indicated by green arrows. Thus, it
may appear that this is an unidirectional pathway, however it should be noted that the flux of cholesterol is multidirectional.
Dashed arrows indicate possible (but not yet validated) links. Abbreviations: M, mitochondrion; Nu, nucleus

glucosylated cholesterol and its cellular transport pathways have not yet been documented, its
occurrence in mammalian cells and lysosomes has been unequivocally demonstrated [18,19].

Management of Free Cholesterol in the Lumen of Lysosomes

As mentioned above, cholesterol is produced from the LDL-associated cholesterol esters by the
activity of LAL. Deficiency of this central cholesterol ester-handling enzyme can lead to Wolman
disease (WD; OMIM ID: 278000) and cholesterol ester storage disease [20] (CESD; OMIM ID:
278000) (Box 1). CESD is a rare lysosomal storage disease caused by a variety of mutations of
the LIPA gene coding for LAL [21].

Trends in Cell Biology, Month 2020, Vol. xx, No. xx 3

Trends in Cell Biology

Table 1. Selected and Commonly Used Biochemical and Cell-Based Assays to Monitor Intracellular Cholesterol Distribution and Cellular Cholesterol
Homeostasis [84,95,96, and references therein].
Type of assay Principle Advantage and disadvantage Refs
Filipin staining Polyene antibiotic as a fluorescent detergent binds Advantage: [85]
selectively to free cholesterol (not to cholesterol esters); has Can be used for qualitative assays to monitor free
also been used in electron microscopy; used to study cholesterol distribution; can be seen in
lysosomal storage disorders freeze-fracture micrographs

Disadvantage:
May also bind to gangliosides; not necessarily a
quantitative method; cholesterol may redistribute
during long incubations, easily bleached
Dehydroergosterol (DHE) Fluorescent naturally occurring derivative of cholesterol Advantage: [86]
used for fluorescence microscopy and trafficking studies in Similar distribution in cells; can be used in
mammalian cells. different types of imaging and live cell microscopy

Disadvantage:
Quantitative differences in the transport of DHE,
fluorescence dependent on microenvironment;
bleaching and low brightness
Cholestatrienol The closest fluorescent derivative of cholesterol used for Advantage: [87]
trafficking studies and in the study of membrane Incorporated and transported similar to
biophysics cholesterol

Disadvantage:
Easily bleached
Bacterial toxins (e.g., Used for electron microscopy and light microscopy Advantage: [88]
perfringolysin O) conjugated Well suited for state-of-the art microscopy as well
with fluorophore as live cell imaging

Disadvantage:
No binding to cholesterol esters, not good to
analyze interorganellar dynamics. Binds only to
cholesterol-rich membranes.
BODIPY–cholesterol Fluorescent derivative of cholesterol used in cellular Advantage: [84,89,90]
trafficking studies Brightness and good resolution in microscopy,
can be used to follow lysosome delivery of
LDL–cholesterol

Disadvantage:
Partial mistargeting in cells
NBD-cholesterol (esters) Fluorescent derivative of cholesterol used in biophysical Advantage: [91]
assays and trafficking studies Use in fluorescence microscopy

Disadvantage:
Bleaching, different membrane orientation
TLC, cholesterol oxidase Classical density enrichment or immunoisolation of Advantage: [92–94]
assays and mass organelles followed by biochemical assays Quantitative readout
spectrometric detection of
cholesterol after subcellular Disadvantage:
fractionation Organelles usually not pure and cholesterol
cannot be followed dynamically
SREBP2-nuclear fragment SREBP2 is liberated after cholesterol depletion and Advantage: [10]
detection activates transcription of sterol responsive Allows to study overall effect in cellular cholesterol
genes like LDLR and HMG-CoA reductase. Immunoblot sensing/regulation after genetic and
analysis of SREBP2 nuclear fragment in the pharmacological interventions
presence/absence of LDL-cholesterol
Disadvantage:
Does not measure cholesterol levels directly, but
indicates free cholesterol levels in ER

4 Trends in Cell Biology, Month 2020, Vol. xx, No. xx

Trends in Cell Biology

NPC2 SapA

Cholesterol
binding
soluble proteins

NPC1 LIMP2 PTCH1

Cholesterol
binding
membrane
proteins

Trends in Cell Biology

Figure 2. Niemann–Pick C (NPC) 1 Protein, Lysosome Integral Membrane Protein 2 (LIMP-2), and Protein
Patched Homolog 1 (PTCH1) Are Cholesterol Transporters. Surface representations of the tunnels of NPC1,
LIMP-2 luminal domain, and PTCH1 are presented as a longitudinal cut through the body of the protein (sliced solid
shown in pink). The ribbon representations are also shown for reference. Cholesterol molecules present in the tunnels of
LIMP-2 luminal domain and PTCH1 are shown in blue. The lipid bilayer is schematically illustrated in grey colors. NPC2
and saposin A (SapA) are also depicted to illustrate their role in handing over cholesterol to the transmembrane proteins.

The biophysical properties of the virtually water-insoluble cholesterol molecule make it necessary
for the lipid released inside the lysosome to be either: (i) quickly incorporated into pre-existing
membranes; (ii) immediately engaged by specific proteins with hydrophobic sterol-binding do-
mains; or (iii) derivatized to more water-soluble compounds such as glucosyl-β-D-cholesterol
(GlcChol). Using liquid chromatography–mass spectrometry (LC-MS)/MS and 13C-labeled
GlcChol as an internal standard, lysosomal cholesterol was quickly glucosylated into GlcChol
by acid β-glucocerebrosidase and this reaction was dependent on the availability of local choles-
terol acceptors [18]. The cholesterol glucosylation changes its physicochemical properties by
making it more water soluble and possibly favoring the formation of GlcChol micelles. GlcChol
may be, although less abundant, better suited for transport and an attractive intermediate in
the transport pathway. After transport GlcChol may be reconverted back to inert cholesterol [18].

However, most of the free cholesterol is immediately bound by proteins with hydrophobic sterol-
binding domains to help transport it out of lysosomes. NPC2, a small, 132-amino-acid soluble
lysosome luminal protein plays a critical role in lysosomal cholesterol egress via its cholesterol-
binding pocket [22]. NPC2 deficiency leads to lysosomal cholesterol accumulation, causing

Trends in Cell Biology, Month 2020, Vol. xx, No. xx 5

Trends in Cell Biology

NPC2 disease, a fatal neurovisceral disorder. X-ray crystallographic studies revealed the struc-
ture of NPC2 in complex with cholesterol sulfate, an analog that binds with greater apparent af-
finity than cholesterol. Interestingly, a lysosomal sulfohydrolase isolated from lysosomal
membranes of human placenta was described to cleave cholesterol sulfate. Although the phys-
iological role of cholesterol sulfate is still enigmatic, a contribution to lysosomal export processes
was speculated [23].

Lately, other proteins have been shown to play a role in the regulation of lysosomal cholesterol
export, such as the abundant and ubiquitous LAMP-1 and 2. Notably, yeast do not have
LAMP proteins. LAMP-1/2 are type 1 transmembrane proteins with a large, heavily glycosylated
luminal domain, a transmembrane domain, and a C-terminal cytoplasmic tail. LAMP proteins
were shown to bind cholesterol directly with high affinity, in a manner similar to NPC1, that buries
the cholesterol 3β-hydroxyl group within their luminal domain [24,25]. Furthermore, LAMP pro-
teins have been shown to be involved in the LEL cholesterol egress pathway [26,27]. Cells lacking
LAMP-1/2, as well as LAMP-2-deficient mouse liver, showed prominent storage of lysosomal
cholesterol [26]. This finding seems somewhat surprising given the fact that the glycosylation of
LAMPs contributes to the dense carbohydrate layer at the inner surface of lysosomes, possibly
allowing cholesterol to freely diffuse to the limiting membrane. However, it has been realized
that other lysosomal membrane proteins carry extensive glycosylation which still maintain a sig-
nificant barrier.

Recently, a role in lysosomal cholesterol egress was also suggested for saposin A (Figure 2), a
member of the lysosomal saposin family of small, nonenzymatic sphingolipid-activator proteins
(SAPs, saposins A, B, C, and D), which possess strong lipid binding/solubilizing properties
(i.e., cholesterol). SAPs are lysosomal cofactors that work together with hydrolases, playing sig-
nificant roles in sphingolipid catabolism. Saposin A forms lipoprotein complexes that share some
similarities with high-density lipoprotein (HDL) particles, although they are significantly smaller
[28]. Experiments have demonstrated that they can associate with the luminal domain of LIMP-
2 (also known as SCARB2) – another abundant integral protein of the limiting membrane of
lysosomes – (Figure 3) to deliver cholesterol to the hydrophobic tunnel of LIMP-2 [8]. These
new findings, which are explained in more detail below, reveal that LIMP-2 works in parallel
with NPC proteins to mediate a slower mode of lysosomal cholesterol export [8].

How Does Cholesterol Shuttle to the Limiting Lysosomal Membrane?

When cholesterol binds to NPC2, its 8-carbon isooctyl side chain is recognized and buried deep
within the hydrophobic pocket of NPC2 [22]. NPC2 then hands over the cholesterol molecule to
the amino terminal domain (NTD) of NPC1 and possibly LAMP-2 (see below) with the 3β-hydroxyl
group of cholesterol now buried within the binding pocket of NPC1 and LAMP-2, respectively.
Given this scenario it seems likely that NPC2 (which binds cholesterol in the opposite orientation
[22]) delivers cholesterol to both LAMPs and NPC1. The NPC1-bound cholesterol is thought to
be delivered through the channel within the NPC1 molecule that spans the entire glycocalyx
layer (Figure 2) to the limiting lysosomal membrane [29]. Recently, the 2.4-Å resolution crystal
structure of the complex composed of the NPC1 middle luminal domain together with NPC2 re-
vealed more details of their overall structural organization. Following this model, NPC2 accepts
free cholesterol and transfers it to the sterol-binding pocket of NPC1 through a cholesterol trans-
fer tunnel formed between NPC2 and the NTD of NPC1 [24].

NPC1, a multimembrane-spanning protein, is at the center of this transfer process. NPC1 defi-
ciency or its functional impairment leads to a dramatic lysosomal build-up of unesterified choles-
terol (among other lipids, such as glycolipids) in multiple tissues including the central nervous

6 Trends in Cell Biology, Month 2020, Vol. xx, No. xx

Trends in Cell Biology

3
EE/LE EE/LE
5 1
4 3 EV71/CVA16
EE/LE 1 LE Phosphatidylserine
containing vesicle
5
LY 2 Phosphatidylserine
H+
2 1
LIMP-2 LDL particle
5
NPC1
H+ 2 Cholesterol

LD LDLR
1
LIMP-2

Nucleus
βGC

AP3

Trends in Cell Biology

Figure 3. Lysosome Integral Membrane Protein 2 (LIMP-2): a Multifunctional and Abundant Membrane Protein
of the Lysosomal Membrane. LIMP-2 resides in the limiting membrane of lysosomes. It plays important roles in the
trafficking of β-glucocerebrosidase (βGC), lipid uptake, and lysosomal cholesterol export: (1) adaptor protein 3 (AP3)
trafficking route of LIMP-2 to lysosomes. The delivery of LIMP-2 to lysosomes (in part via the plasma membrane) is
mediated by AP3, which binds to the C terminal of LIMP-2. (2) LIMP-2 binds βGC in the endoplasmic reticulum (ER) and
directs it to the lysosome where receptor and ligand dissociate. (3) The binding of lipids to a LIMP-2 monomer promotes
the formation of LIMP-2 dimer which interacts at the cell membrane with extracellular lipid vesicles. The hydrophobic
cavities within the LIMP-2 ectodomain dimer tether the bound lipid vesicles and cell membrane and facilitate lipid
exchange. (4) LIMP-2 also acts as a cellular receptor for enterovirus 71 (EV71) and coxsackievirus A16 (CVA16). EV71 and
CVA16 are pathogens in humans causing hand, foot, and mouth disease. (5) LIMP-2 acts as a cholesterol transporter
mediating the cholesterol transport out of the lysosome through its hydrophobic tunnel. Abbreviations: EE, early
endosome; LD, lipid droplet; LE, late endosome; LY, lysosome.

system, liver, and spleen [30,31] (Box 1). In recent years, the structure of human NPC1 has been
gradually revealed by the work of several groups who crystallized either the full length or different
fragments of the protein [24,32,33]. The middle one of the three luminally oriented domains
termed MLD of NPC1 was demonstrated to interact with NPC2, allowing the latter to participate
in the ‘hydrophobic hand-off’ of cholesterol to the NTD of NPC1 [33]. From there, cholesterol is
thought to be delivered to the sterol-sensing domain (SSD) of NPC1, which consists of trans-
membrane domains 3–7 [32,34]. The overall domain architecture of NPC1 is shared by members
of the Patched protein family. Crystal structures of Patched recently revealed a hydrophobic SSD
within the protein that might serve to transport lipids [35,36] (Figure 2).

How cholesterol is then delivered to either the luminal or cytosolic face of the lysosomal mem-
brane remains unclear. Li and colleagues speculated that the sterol molecule is transferred to
the luminal leaflet of the membrane and is flipped to the cytoplasmic leaflet thereafter, either
with the assistance of NPC1 or spontaneously [33]. Recently, new insight was gained regarding
the mechanisms by which NPC1 is able to transfer cholesterol through the glycocalyx and into the
limiting membrane. Using single-particle cryoelectron microscopy and crystallography, Winkler

Trends in Cell Biology, Month 2020, Vol. xx, No. xx 7

Trends in Cell Biology

and coworkers determined the structure of the yeast counterparts Ncr1 and Npc2 [7]. The Ncr1
protein contains an intramolecular tunnel that connects the different domains of the protein,
namely the NTD and the SSD. Using sterol transfer assays, the authors elucidated the specific
roles of each domain in the transfer of cholesterol. In agreement with previous reports [29],
Winkler and colleagues [7] described how the N-terminal domain of the Ncr1 protein accepts
cholesterol from Npc2, from where it is shuttled to the membrane-embedded tunnel
consisting of the C-terminal (CTD) and MLD domains, finally arriving at the SSD shelf, thereby en-
abling the lipid to bypass the glycocalyx. While it could be envisaged that the sterol passes through
the tunnel by passive diffusion, the data from Winkler et al. [7] pointed towards an active transport
mechanism. Intriguingly, the driving force behind this translocation was proposed to consist of a
proton-relay network leading to pincer-like movement of the sterol through the tunnel, based on
the proximity of two likely protonated, conserved acidic amino acids at positions Asp631 and
Glu1068 as well as His1072. Mutational studies revealed significantly reduced cholesterol transfer
efficiency when these amino acids were substituted [7]. Intriguingly, other members of the re-
sistance-nodulation-division (RND) superfamily, like MmpL3 and HpnN, also utilize proton
relay networks for active transport of hopanoid molecules, trehalose monomycolate, and phos-
phatidylethanolamine and tuberculosis drug targets, respectively [37–39].

Li and Pfeffer showed that the luminal domain of LAMPs also tightly binds to the NPC1 NTD and
NPC2 proteins, suggesting that these abundant lysosome membrane proteins may also accept
cholesterol from NPC2 and subsequently deliver it to NPC1. While there is no evidence for a direct
lipid export function of LAMPs, it has been suggested that they serve as a reservoir on the limiting
lysosome membrane for cholesterol extracted from intralysosomal membranes prior to NPC1-
mediated export [24]. A future challenge will be to determine exactly how and when cholesterol
is delivered to LAMP-2 or NPC1, and to establish whether cholesterol binding interferes with
the other reported roles of LAMP proteins, such as chaperone-mediated autophagy, fusion of ly-
sosomes with phagosomes and autophagosomes, and lysosomal stability [1].

Surprisingly, another player emerged recently on the cholesterol-export stage. Two independent
crystal structures of LIMP-2 – one in complex with cholesterol – revealed a hydrophobic tunnel
within the luminal domain that traverses the entire length of the elongated region of the protein
that resides in the lysosomal lumen [40,41] (Figure 2) that has been also found in the other mem-
bers of class B scavenger receptors, CD36 and SR-B1 [40,42,43]. In the case of SR-B1, this tun-
nel is used to deliver cholesterol from HDL to the plasma membrane (PM) [40,44]. The CD36
tunnel is needed for steroid pheromone detection by their receptors in olfactory cilia delivery in
flies [45]. These observations strongly suggest that LIMP-2 could participate in lysosomal choles-
terol egress. In vitro studies using cross-clickable lipid probes have suggested that LIMP-2 pref-
erentially binds cholesterol. Besides, LIMP-2 sterol transport ability was shown by using the
purified LIMP-2 luminal domain tethered to liposomes, and by using cells expressing a PM-
targeted LIMP-2 chimera. Endogenous LIMP-2's role in lysosomal cholesterol egress was dem-
onstrated by following the trafficking of LDL-derived BODIPY–cholesterol to lipid droplets, as well
as analyzing the processing of the sterol-regulatory transcription factor SREBP2. These experi-
ments revealed a defect in LDL-derived cholesterol homeostasis and trafficking in absence of
LIMP-2 or when its tunnel is blocked [8]. The biochemical assessment of the cholesterol esterifi-
cation in the presence or absence of LIMP-2 confirmed the role of LIMP-2 in cholesterol transport.
This newly discovered route of lysosomal cholesterol egress seems to operate independently of
the NPC1 pathway [8]. However, both pathways share the ability to transfer the hydrophobic cho-
lesterol molecule through the lysosomal glycocalyx (Figure 2). It remains to be elucidated whether
the LIMP-2 pathway specifically serves to support lysosomal cholesterol export in specialized cell
types, such as Schwann cells, or under certain nutrient situations, since it is striking that the rate

8 Trends in Cell Biology, Month 2020, Vol. xx, No. xx

Trends in Cell Biology

of cholesterol export by LIMP-2 is markedly slower than that of NPC1 [8]. It also remains an open
question if the function of LIMP-2 as a chaperone for the acid glucocerebrosidase (Figure 3) is
coupled with its sterol transport function. In this regard, glucosylated cholesterol may be a prod-
uct of glucocerebrosidase activity, in which case glucosylated cholesterol could possibly be a
more natural substrate for the LIMP-2 tunnel.

Models of Cholesterol Transfer from the Lysosomal Lumen to the Lysosomal

Membrane
The need to bypass the protective polysaccharide layer lining the luminal side of the lysosomal
membrane is met by the tunnel structures present in both NPC1 and LIMP-2 (Figure 2). Both pro-
teins feature domains that can penetrate through the glycocalyx to enable cholesterol loading into
the tunnel. While the molecular makeup of NPC1 and LIMP-2 is distinctly different, the cavities in
both seem to consist of a continuous hydrophobic environment where lipids such as cholesterol
do not have fixed binding sites [7,40,41]. Based on the presence of the above-mentioned poten-
tial proton relay system in Ncr1, Winkler et al. proposed that the sterol is moved through the tun-
nel towards the SSD by sequential protonation and deprotonation of the charged residues. After
transitioning, cholesterol is thought to dissociate from the SSD/tunnel and to partition into the
outer leaflet of the lysosomal membrane to exert its function there or to be transferred to other or-
ganelles such as the endoplasmic reticulum by lipid-transport proteins [7].

From the Lysosomal Membrane to the Endoplasmic Reticulum (ER) and Other
Organelles
It is now becoming clear that many, possibly all, organelles communicate by means of interorganellar
tethering, thereby maintaining proximity between apposing membranes to ensure cell homeostasis.
Membrane contact sites (MCSs), domains of close membrane connections, play important roles in
cell physiology, since these membrane junctions execute a multitude of functions in lipid synthesis,
interorganellar lipid transport, calcium regulation, signaling, autophagy, endosome fission, mitochon-
drial scission, retrograde transport from endosomes to the Golgi, and sorting and degradation of at
least one endocytic cargo (epidermal growth factor receptor; EGFR) [46–49]. The MCSs between
the endosomes and other organelles including the ER, peroxisomes, and PM, are conduits of bidirec-
tional sterol exchange, which can vary depending on the cell type and physiological state [47,50].
Thus, exogenous cholesterol carried by LDL (Figure 1) is rapidly removed from the lysosome and is
found in several acceptor compartments (i.e., PM, ER, and Golgi) that require the putative cholesterol
carriers, NPC1 and LIMP-2 [8]. Disruption or inhibition of these MCSs leads to cholesterol accumula-
tion in LELs, which is associated with neurological diseases [51,52]. Due to their frequent, transient
and dynamic nature, the molecular composition of the MCS along the endocytic pathway remains
poorly characterized. Yet, it has been shown that EGFR-negative populations of MCSs are tethered
to the ER by VAMP-associated proteins (VAPs), whereas EGFR-containing multivesicular
endosomes/bodies (MVBs) are tethered by annexin-A1 and its calcium-dependent ligand
S100A11 [53]. Important components of these MCSs are members of the highly conserved
oxysterol-binding proteins (OSBP/ORP; OSBP1L, ORP5, and ORP6), steroidogenic acute regulatory
protein (StAR)-related lipid-transfer proteins (START) protein superfamily (STARD3), and LAMs family
of lipid transporters (GRAMD1B), which can mediate cholesterol transfer and hence regulate choles-
terol homeostasis [54]. OSBP/ORPs contain a conserved oxysterol-binding-related domain (ORD),
two phenylalanines-in-an-acid tract (FFAT) motifs and pleckstrin-homology (PH) domains, which
can bind to phosphatidylinositol polyphosphates (PIPs). STARD3 contains a MLN64 N-terminal
(MENTAL) domain, an FFAT-like motif, and a C-terminal START domain. LAM proteins contain a
GRAM domain and a StART-like domain [55]. OSBP-related protein 1L (ORP1L), OSBP, ORP5,
ORP6, and GRAMD1B together with NPC1 and NPC2 and LIMP-2 contribute to the LDL–cholesterol
egress from endocytic organelles to the ER. Under low-cholesterol conditions, cholesterol can be

Trends in Cell Biology, Month 2020, Vol. xx, No. xx 9

Trends in Cell Biology

moved from ER to LELs via distinct MCSs formed by ORP1L, OSBP, and STARD3 interaction with
the conserved ER membrane VAPs through their FFAT motifs [50].

ER–LEL Contact Sites

How LDL-derived cholesterol reaches the ER from LELs is an area of active investigation, but it
probably involves redundant and parallel pathways (Figure 4). Whereas a fraction of LDL-
derived cholesterol is delivered to the ER via the PM, there is strong supporting evidence for direct
transport of approximately 30% of the LDL–cholesterol from the endosomes to the ER [9,56,57].
It has also been demonstrated that cholesterol can affect the composition of the ER–LEL MCSs,
which in turn can affect LEL traffic and localization. Thus, when cellular cholesterol levels are high,

VAPB PI4P
OSBP
?
ER
NPC1 Peroxisome
VAPB
LIMP-2

Saposins
SYT7
ORP1L H+
NPC2

NPC1 Lysosome
ORP5
NPC2
NPC2
H+ STARD3

NPC1
GramD1b

PIPs?
STARD3
VAPB Mitochondrion

Cholesterol

Ptdlns(4,5)P2

Trends in Cell Biology

Figure 4. Lysosome–Organelle Contact Sites Mediate Intracellular Cholesterol Transfer. Contact sites of
lysosome with the endoplasmic reticulum (ER) are conduits of cholesterol transfer to and from lysosomes. The lysosomal
cholesterol egress is likely mediated by oxysterol-binding protein (OSBP)–VAMP-associated protein (VAP) [in exchange
with phosphatidylinositol 4-phosphate (PI4P), OSBP-related protein 1L (ORP1L)–VAP, ORP5–Niemann–Pick (NP)C1, and
GRAM domain containing 1B (GRAMD1B)–NPC1]. By contrast, StAR-related lipid transfer domain-3 (STARD3)–VAP can
mediate ER to lysosome cholesterol transport. The role of lysosome integral membrane protein 2 (LIMP-2) in establishing
membrane contact sites is still unclear. Contacts with peroxisomes via synaptotagmin VII (SYT7) and phosphatidylinositol
4,5-bisphosphate [PtdIns(4,5)P2] mediate NPC1-dependent cholesterol transfer. Also, lysosomal cholesterol can be
delivered to the mitochondrion via STARD3. Within the lysosome saposins or NPC2 can be involved in binding and the
transfer of cholesterol to NPC1 or LIMP-2.

10 Trends in Cell Biology, Month 2020, Vol. xx, No. xx

Trends in Cell Biology

LELs accumulate in the perinuclear region of cells. By contrast, when cellular cholesterol levels are
low, LELs accumulate at the cell periphery [58]. This LEL redistribution is controlled by the
cholesterol-sensing protein ORP1L, a key regulator of endosomal transport and maturation,
which can undergo cholesterol-dependent conformational changes and hence form
cholesterol-dependent interactions between endosomes and the ER or microtubule network.
ORP1L localizes at the LELs and contains an ORD that has been shown to be capable of binding
oxysterols, cholesterol, and phospholipids in vitro. In high cholesterol conditions, ORP1L
changes its conformation, allowing it to interact with RAB7 by virtue of its N-terminal ankyrin mo-
tifs [59]. The subsequent association of ORP1L, RAB7, and RILP (RAB7-interacting lysosomal
protein) with dynactin subunit p150Glued and βIII-spectrin [58] on LELs ultimately results in fur-
ther dynein motor recruitment and hence increased perinuclear accumulation of tethered LELs.
In low-cholesterol conditions, ORP1L acquires a conformation such that its FFAT domain can
freely interact with the ER VAPs, resulting in its dissociation from the dynein motor complex,
thus reducing centripetal motion [60].

Besides regulating vesicle trafficking, ORP1L has been implicated in the bidirectional transfer of cho-
lesterol from ER to LELs and vice versa [53,61]. Using ORP1L knockout or NPC1-inhibited cells, it
was shown that LDL–cholesterol egresses from LELs to the ER through an ORP1L-mediated mech-
anism that requires its sterol-, phospholipid-, and VAP-binding activities. Subsequently, it was
shown that this process is also dependent on PI(4,5)P2/PI(3,4)P2-phosphoinositides [53]. By
contrast, in conditions of low-cholesterol levels, de novo synthesized cholesterol is imported to
the LELs via ORP1L–VAP at specialized MCSs [61]. A rare heterozygous truncation mutant of
ORP1L in humans causes low circulating HDL levels, which in turn leads to a reduction of the
cholesterol efflux from the cell in order to maintain essential membrane functions [62].

OSBP was previously shown to be localized at the ER and at the trans-Golgi where it is thought to
transfer ER-derived cholesterol to the Golgi in exchange for phosphatidylinositol 4-phosphate
(PtdIns4P) [63,64], and was recently implicated at contacts between the ER and LEL as well
[65]. At the contact sites between the ER and LELs, OSBP together with VAPA and VAPB deliver
cholesterol from ER to lysosomes in exchange for PtdIns4P. This exchange can affect their actin-
dependent motility [66]. In addition, the OSBP–VAP tethering complex is unique among VAP-
interacting proteins, which is necessary for mTORC1 activation under low-cholesterol
conditions [65].

By contrast, other family members such as ORP5 and ORP6 have been implicated only in LEL to
ER cholesterol transport. The ER tail-anchored protein ORP5 mediates cholesterol efflux to the
ER in a NPC1-dependent manner [67]. Immunoprecipitation experiments have revealed that
ORP5 interacts with NPC1. Consequently, ORP5 depletion results in cholesterol accumulation
in LELs, consistent with an NPC-like phenotype [67].

Overexpression of ORP6 results in its association with the early endolysosomal network and ER,
which depends on its FFAT motif and PH domain. Silencing of ORP6 leads to increased filipin
staining and a fused endocytic network, indicating a role of ORP6 in the organization of early
endosomes and cholesterol delivery to the ER [68].

GRAMD1B is another interesting player in mediating cholesterol transfer between membranes. It

is not only targeted to ER–PM contact sites to mediate the transfer of HDL-derived cholesterol
from the PM to ER [55] but it is also targeted to cholesterol-rich organelles within the endocytic
pathway [9]. Thus, GRAMD1B interacts with NPC1 at ER–lysosome contact sites in the presence
of high LDL–cholesterol in the endocytic pathway to stabilize the interorganellar association for

Trends in Cell Biology, Month 2020, Vol. xx, No. xx 11

Trends in Cell Biology

transport of cholesterol to the ER. The cholesterol transport function of GRAMD1B may involve
phosphoinositides as a counter substrate, as it was described for OSBP at the ER-Golgi
interface.

The START-domain protein STARD3 mediates cholesterol transfer from ER to LELs. STARD3
overexpression is known to promote the formation of MCS at the ER-endocytic organelle. As a
result, endosomes are covered by the ER where the distance between the ER and endosome
membranes is less than 10 nm [69,70]. Besides, overexpression of STARD3 in CHO and HeLa
cells results in accumulation of unesterified cholesterol in endosomes at the expense of PM cho-
lesterol, which is dependent on its FFAT motif and START domain [70]. However, the more phys-
iologically relevant function might be to support intraluminal vesicle formation since STARD3 was
shown to be required for the biogenesis of intraluminal vesicles, in a manner analogous to MCS
mediated by ORP1L and annexin A1 [47].

LEL–Peroxisome Contact Sites

Recently, peroxisomes were identified as a site for cholesterol egress from LELs, via LEL–
peroxisome contact sites [10,71]. This work provided evidence that the contact sites between
these organelles were dependent on the presence of NPC1, synaptotagmin VII, and phos-
phatidylinositol 4,5-bisphosphate (in peroxisomes) (Figure 4). Importantly, MCSs were enhanced
by cholesterol in LEL and their disruption caused cholesterol accumulation in these organelles.

LEL–Mitochondrion Contact Sites

Except for Saccharomyces cerevisiae mitochondrion–vacuole contact sites, mitochondrion–
melanosome contact sites, and mitochondrion–lysosome contact sites had not been described
until recently [72]. Thus, it was shown that mitochondrion–lysosome contacts form dynamically
in healthy untreated HeLa cells (Figure 4). Contact formation is promoted by active GTP-bound ly-
sosomal RAB7, while untethering was mediated by recruitment of the RAB7 GTPase-activating
protein TBC1D15 to the mitochondrion by FIS1 in order to drive RAB7 GTP hydrolysis and thereby
release contacts [72]. Notably, lysosome contacts do not represent sites of bulk protein transfer or
mitochondrial degradation, either directly through mitochondrial-derived vesicles fusing with
lysosomes or indirectly through mitophagy [72]. Therefore, it has been proposed that
mitochondrion–lysosome contacts may function as platforms for exchange of metabolites,
such as fatty acids or calcium. In this regard, it has been observed that in NPC1- or
GRAMD1B-deficient cells the number of lysosome–mitochondrion MCSs is increased and is
dependent on the late endosomal sterol-binding protein STARD3 [65]. Consequently, it has
been proposed that lysosome–mitochondrion MCSs may offer an alternative route to clear
cholesterol from the endocytic pathway in a STARD3-dependent manner [73]. This mechanism
would account for the elevated cholesterol observed in the mitochondria of NPC1-deficient or
GRAMD1B-deficient cells [74].

Cholesterol and mTORC1 Signaling

Recent studies have demonstrated the essential role of cholesterol in the activation of mTORC1
kinase in a cell-autonomous manner [50,65,75,76]. When activated, mTORC1 is dynamically as-
sociated with the lysosomal membrane, from where it can initiate programs that upregulate ana-
bolic processes and downregulate autophagy [65]. For its activation, mTORC1 interacts with the
activated Rag GTPases in an amino-acid- and cholesterol-dependent manner [65,77]. The
amino-acid- and cholesterol-sensing mechanisms of Rag GTPases depend partly on SLC38A9
[75], a multipass amino acid permease, and OSBP which is thought to establish a lysosomal cho-
lesterol pool essential for mTORC1 activation [65]. Thus, SLC38A9 is not only a transporter for
essential amino acids [78] but it also possesses both a putative cholesterol-interacting motif

12 Trends in Cell Biology, Month 2020, Vol. xx, No. xx

Trends in Cell Biology

known as cholesterol recognition amino acid consensus (CRAC) and the inverse motif, known as Outstanding Questions
CARC [79]. Under low-cholesterol conditions, mTORC1 cannot interact with the Rag GTPases Are other lipids (such as fatty acids,
and remains inactive in the cytosol [50]. By contrast, the lysosomal recruitment of mTORC1 – in- sphingosins) present in lipoproteins
dependently of the acute mTORC1 in response to amino acids – can be induced under some cir- and ending up in lysosomes also
specifically transported and how are
cumstances (e.g., in cells treated with LDL–cholesterol) by a Rag-GTPase-dependent these further utilized or sensed by the
mechanism which involves both cholesterol-binding NPC1 and OSBP [65]. This has been ele- cell?
gantly shown in NPC1-deficient cells where OSBP drives aberrant accumulation of cholesterol
Is there an involvement of LIMP-2/
in LELs [65]. This, in turn, contributes to dysregulated mTORC1 signaling and autophagy,
SCARB2 in establishing and maintain-
which can be suppressed by genetic and chemical inhibition of OSBP. Thus, through their ability ing lysosome/organelle contact sites?
to exchange cholesterol, ER–lysosome MCSs arise as signaling nodes that coordinately regulate Does ER stress, lipid loading, or lyso-
mTORC1 signaling and autophagy [50]. somal dysfunction modulate these
contact sites and the transfer of lipids
between them?

Concluding Remarks and Future Perspectives How does cholesterol pass through
Having in part resolved the cellular regulation of cholesterol homeostasis during the last two de- the hydrophobic tunnel of LIMP-2? Is
a gradient involved similar to SR-BI or
cades, it is becoming increasingly clear that lysosomes are at the center of this important regula- are amino acids within the tunnel re-
tory network. This is illustrated by human diseases where lysosomal cholesterol handling and quired for transport?
export are deficient. Research studying the pathophysiology of these diseases – but also recent
How are lysosomal cholesterol and
protein crystallization studies – have demonstrated an important role of the multispanning lyso-
lipid processes regulated in specific
somal membrane NPC1 in the transport of cholesterol from the lumen of the lysosome to its lim- tissues and cell types? Is the
iting membrane. Only recently, the functions of other lysosomal membrane proteins such as lysosomal lipid transport pathway
LAMPs and LIMP-2/SCARB2 in mediating sterol transport out of lysosomes were revealed. different if, for example, brain and liver
are being compared?
How these processes are regulated and (whether they) communicate with each other, and in
which cells and tissue they are of importance is still an open question. It has, however, been re- How is cholesterol (and other lipids)
solved that the dense layer of carbohydrates at the inner surface of the lysosomal membrane can handled in lysosome-related organ-
be bridged by the elongated hydrophobic intramolecular cavities within NPC1 and LIMP-2. Once elles such as melanosomes?

cholesterol or cholesterol derivatives reach the outer leaflet of the limiting lysosomal membrane a How far are lysosomal cholesterol export
complex cytosolic machinery of cholesterol-binding proteins, lysosomal proteins, and organellar- pathways involved in plasma membrane
resident membrane proteins mediate the contact of lysosomes with other organelles, thereby repair mediated by lysosomes and in
the repair of damaged lysosomes?
possibly contributing to the transfer of cholesterol between their respective membranes. Similar
to the situation of amino acid export from lysosomes, lysosomal cholesterol export is sensed How are cholesterol derivatives
by the transmembrane protein, SLC38A9 [11] and the intraluminal cholesterol status is thereby transported and what is their (patho)
translated into activation of the mTORC pathway and control of cell proliferation and autophagy. physiological role?

How much is an asymmetry of

Open questions (see Outstanding Questions) in the field range from the substrate specificity of the membrane lipids and cholesterol in
lysosomal lipid transporters; how these lipids are recognized and handed over; how these trans- lysosomes linked to lipid transport
pathways? Do these specific lipids and
port processes are regulated at a cell type and tissue-specific level; how lysosomal cholesterol/
lipid transport pathways contribute
lipid levels control cellular behavior under disease conditions; and how the complex machinery to the maintenance of lysosomal
between lysosomes and recipient organelles is molecularly determined and regulated. membrane asymmetry?

Acknowledgments Does lipid transport-dependent lyso-

We thank Johannes Aerts, Leiden University (The Netherlands) and Sergio Grinstein and Gregory D. Fairn, University of Toronto somal inside-out signaling but also
outside-in signaling affect cellular func-
(Canada) for critical reading and commenting on our manuscript. This work was supported by the Deutsche
tions and the functioning of lysosomes?
Forschungsgemeinschaft (DFG; FOR2625), Natural Science Foundation of China (31770938) and the Key Program of
Zhejiang Provincial Natural Science Foundation of China (LZ16C050001).

References
1. Saftig, P. and Klumperman, J. (2009) Lysosome biogenesis and 3. Marques, A.R.A. and Saftig, P. (2019) Lysosomal storage disor-
lysosomal membrane proteins: trafficking meets function. Nat. ders - challenges, concepts and avenues for therapy: beyond
Rev. Mol. Cell Biol. 10, 623–635 rare diseases. J. Cell Sci. 132, Published online January 16,
2. Ballabio, A. and Bonifacino, J.S. (2020) Lysosomes as dynamic 2019. https://doi.org/10.1242/jcs.221739
regulators of cell and organismal homeostasis. Nat. Rev. Mol. 4. Patterson, M.C. and Walkley, S.U. (2017) Niemann-Pick disease,
Cell Biol. 21, 101–118 type C and Roscoe Brady. Mol. Genet. Metab. 120, 34–37

Trends in Cell Biology, Month 2020, Vol. xx, No. xx 13

Trends in Cell Biology

5. Wheeler, S. and Sillence, D.J. (2019) Niemann-Pick type C dis- 30. Carstea, E.D. et al. (1997) Niemann-Pick C1 disease gene:
ease: cellular pathology and pharmacotherapy. J. Neurochem. homology to mediators of cholesterol homeostasis. Science
Published online October 14, 2019. https://doi.org/10.1111/ 277, 228–231
jnc.14895 31. Rosenbaum, A.I. and Maxfield, F.R. (2011) Niemann-Pick type C
6. Brown, M.S. and Goldstein, J.L. (1986) A receptor-mediated disease: molecular mechanisms and potential therapeutic ap-
pathway for cholesterol homeostasis. Science 232, 34–47 proaches. J. Neurochem. 116, 789–795
7. Winkler, M.B.L. et al. (2019) Structural insight into eukaryotic 32. Gong, X. et al. (2016) Structural insights into the Niemann-Pick
sterol transport through Niemann-Pick type C proteins. Cell C1 (NPC1)-mediated cholesterol transfer and Ebola infection.
179, 485–497.e18 Cell 165, 1467–1478
8. Heybrock, S. et al. (2019) Lysosomal integral membrane 33. Li, X. et al. (2016) Clues to the mechanism of cholesterol transfer
protein-2 (LIMP-2/SCARB2) is involved in lysosomal cholesterol from the structure of NPC1 middle lumenal domain bound to
export. Nat. Commun. 10, 3521 NPC2. Proc. Natl. Acad. Sci. U. S. A. 113, 10079–10084
9. Hoglinger, D. et al. (2019) NPC1 regulates ER contacts with 34. Li, X. et al. (2016) Structure of human Niemann-Pick C1 protein.
endocytic organelles to mediate cholesterol egress. Nat. Proc. Natl. Acad. Sci. U. S. A. 113, 8212–8217
Commun. 10, 4276 35. Gong, X. et al. (2018) Structural basis for the recognition of Sonic
10. Chu, B.B. et al. (2015) Cholesterol transport through Hedgehog by human Patched1. Science Published online
lysosome-peroxisome membrane contacts. Cell 161, August 10, 2018. https://doi.org/10.1126/science.aas8935
291–306 36. Qi, X. et al. (2018) Structures of human Patched and its
11. Castellano, B.M. et al. (2017) Lysosomal cholesterol activates complex with native palmitoylated sonic hedgehog. Nature
mTORC1 via an SLC38A9-Niemann-Pick C1 signaling complex. 560, 128–132
Science 355, 1306–1311 37. Kumar, N. et al. (2017) Crystal structures of the Burkholderia
12. Thelen, A.M. and Zoncu, R. (2017) Emerging roles for the lyso- multivorans hopanoid transporter HpnN. Proc. Natl. Acad. Sci.
some in lipid metabolism. Trends Cell Biol. 27, 833–850 U. S. A. 114, 6557–6562
13. Pfeffer, S.R. (2019) NPC intracellular cholesterol transporter 1 38. Zhang, B. et al. (2019) Crystal structures of membrane transporter
(NPC1)-mediated cholesterol export from lysosomes. J. Biol. MmpL3, an anti-TB drug target. Cell 176, 636–648.e13
Chem. 294, 1706–1709 39. Su, C.C. et al. (2019) MmpL3 is a lipid transporter that binds tre-
14. Luo, J. et al. (2019) Intracellular cholesterol transport by sterol halose monomycolate and phosphatidylethanolamine. Proc.
transfer proteins at membrane contact sites. Trends Biochem. Natl. Acad. Sci. U. S. A. 116, 11241–11246
Sci. 44, 273–292 40. Neculai, D. et al. (2013) Structure of LIMP-2 provides functional
15. Wilke, S. et al. (2012) Crystal structure of the conserved domain insights with implications for SR-BI and CD36. Nature 504,
of the DC lysosomal associated membrane protein: implications 172–176
for the lysosomal glycocalyx. BMC Biol. 10, 62 41. Conrad, K.S. et al. (2017) Lysosomal integral membrane protein-2
16. Kundra, R. and Kornfeld, S. (1999) Asparagine-linked oligosac- as a phospholipid receptor revealed by biophysical and cellular
charides protect Lamp-1 and Lamp-2 from intracellular studies. Nat. Commun. 8, 1908
proteolysis. J. Biol. Chem. 274, 31039–31046 42. Pepino, M.Y. et al. (2014) Structure-function of CD36 and impor-
17. Zhang, H. (2018) Lysosomal acid lipase and lipid metabolism: tance of fatty acid signal transduction in fat metabolism. Annu.
new mechanisms, new questions, and new therapies. Curr. Rev. Nutr. 34, 281–303
Opin. Lipidol. 29, 218–223 43. Canton, J. et al. (2013) Scavenger receptors in homeostasis and
18. Marques, A.R. et al. (2016) Glucosylated cholesterol in immunity. Nat. Rev. Immunol. 13, 621–634
mammalian cells and tissues: formation and degradation 44. Acton, S. et al. (1996) Identification of scavenger receptor SR-BI
by multiple cellular beta-glucosidases. J. Lipid Res. 57, as a high density lipoprotein receptor. Science 271, 518–520
451–463 45. Gomez-Diaz, C. et al. (2016) A CD36 ectodomain mediates in-
19. Akiyama, H. et al. (2013) Cholesterol glucosylation is catalyzed sect pheromone detection via a putative tunnelling mechanism.
by transglucosylation reaction of beta-glucosidase 1. Biochem. Nat. Commun. 7, 11866
Biophys. Res. Commun. 441, 838–843 46. Prinz, W.A. (2014) Bridging the gap: membrane contact sites in
20. Sun, Y. et al. (2014) Reversal of advanced disease in lysosomal signaling, metabolism, and organelle dynamics. J. Cell Biol.
acid lipase deficient mice: a model for lysosomal acid lipase de- 205, 759–769
ficiency disease. Mol. Genet. Metab. 112, 229–241 47. Ridgway, N.D. and Zhao, K. (2018) Cholesterol transfer at
21. Fasano, T. et al. (2012) Lysosomal lipase deficiency: molecu- endosomal-organelle membrane contact sites. Curr. Opin.
lar characterization of eleven patients with Wolman or Lipidol. 29, 212–217
cholesteryl ester storage disease. Mol. Genet. Metab. 105, 48. Gallo, A. et al. (2016) Endoplasmic reticulum-plasma membrane
450–456 associations: structures and functions. Annu. Rev. Cell Dev. Biol.
22. Xu, S. et al. (2007) Structural basis of sterol binding by NPC2, a 32, 279–301
lysosomal protein deficient in Niemann-Pick type C2 disease. 49. Gatta, A.T. and Levine, T.P. (2017) Piecing together the patch-
J. Biol. Chem. 282, 23525–23531 work of contact sites. Trends Cell Biol. 27, 214–229
23. Roszek, K. and Gniot-Szulzycka, J. (2008) Cholesterol sulphate 50. Thelen, A.M. and Zoncu, R. (2017) Emerging roles for the lyso-
sulphohydrolase of human placenta lysosomal membrane. some in lipid metabolism. Trends Cell Biol. 27, 843–860
J. Steroid Biochem. Mol. Biol. 110, 48–55 51. Maxfield, F.R. (2014) Role of endosomes and lysosomes in
24. Li, J. and Pfeffer, S.R. (2016) Lysosomal membrane glycopro- human disease. Cold Spring Harb. Perspect. Biol. 6, a016931
teins bind cholesterol and contribute to lysosomal cholesterol 52. Platt, F.M. et al. (2012) Lysosomal storage disorders: the cellular
export. Elife Published online September 24, 2016. https://doi. impact of lysosomal dysfunction. J. Cell Biol. 199, 723–734
org/10.7554/eLife.21635.001 53. Eden, E.R. et al. (2016) Annexin A1 tethers membrane contact
25. Hulce, J.J. et al. (2013) Proteome-wide mapping of cholesterol- sites that mediate ER to endosome cholesterol transport. Dev.
interacting proteins in mammalian cells. Nat. Methods 10, Cell 37, 473–483
259–264 54. Horenkamp, F.A. et al. (2018) Molecular basis for sterol
26. Schneede, A. et al. (2011) Role for LAMP-2 in endosomal cho- transport by StART-like lipid transfer domains. EMBO J. Pub-
lesterol transport. J. Cell. Mol. Med. 15, 280–295 lished online March 15, 2018. http://dx.doi.org/10.15252/
27. Eskelinen, E.L. et al. (2004) Disturbed cholesterol traffic but embj.201798002
normal proteolytic function in LAMP-1/LAMP-2 double- 55. Sandhu, J. et al. (2018) Aster proteins facilitate nonvesicular
deficient fibroblasts. Mol. Biol. Cell 15, 3132–3145 plasma membrane to ER cholesterol transport in mammalian
28. Popovic, K. et al. (2012) Structure of saposin A lipoprotein discs. cells. Cell 175, 514–529.e20
Proc. Natl. Acad. Sci. U. S. A. 109, 2908–2912 56. Underwood, K.W. et al. (1998) Evidence for a cholesterol trans-
29. Kwon, H.J. et al. (2009) Structure of N-terminal domain of NPC1 port pathway from lysosomes to endoplasmic reticulum that is
reveals distinct subdomains for binding and transfer of independent of the plasma membrane. J. Biol. Chem. 273,
cholesterol. Cell 137, 1213–1224 4266–4274

14 Trends in Cell Biology, Month 2020, Vol. xx, No. xx

Trends in Cell Biology

57. Neufeld, E.B. et al. (1996) Intracellular trafficking of cholesterol 77. Sancak, Y. et al. (2010) Ragulator-Rag complex targets
monitored with a cyclodextrin. J. Biol. Chem. 271, 21604–21613 mTORC1 to the lysosomal surface and is necessary for its acti-
58. Johansson, M. et al. (2007) Activation of endosomal dynein mo- vation by amino acids. Cell 141, 290–303
tors by stepwise assembly of Rab7-RILP-p150(Glued), ORP1L, 78. Wyant, G.A. et al. (2017) mTORC1 Activator SLC38A9 is re-
and the receptor beta III spectrin. J. Cell Biol. 176, 459–471 quired to efflux essential amino acids from lysosomes and use
59. Johansson, M. et al. (2005) The oxysterol-binding protein protein as a nutrient. Cell 171, 642–654.e12
homologue ORP1L interacts with Rab7 and alters functional 79. Fantini, J. and Barrantes, F.J. (2013) How cholesterol interacts
properties of late endocytic compartments. Mol. Biol. Cell 16, with membrane proteins: an exploration of cholesterol-binding
5480–5492 sites including CRAC, CARC, and tilted domains. Front. Physiol.
60. Rocha, N. et al. (2009) Cholesterol sensor ORP1L contacts the Published online February 28, 2013. https://doi.org/10.3389/
ER protein VAP to control Rab7-RILP-p150(Glued) and late fphys.2013.00031
endosome positioning. J. Cell Biol. 185, 1209–1225 80. Vanier, M.T. (2010) Niemann-Pick disease type C. Orphanet.
61. Zhao, K. and Ridgway, N.D. (2017) Oxysterol-binding protein- J. Rare Dis. 5, 16
related protein 1L regulates cholesterol egress from the endo- 81. Lloyd-Evans, E. et al. (2008) Niemann-Pick disease type C1 is a
lysosomal system. Cell Rep. 19, 1807–1818 sphingosine storage disease that causes deregulation of lyso-
62. Motazacker, M.M. et al. (2016) A loss-of-function variant in somal calcium. Nat. Med. 14, 1247–1255
OSBPL1A predisposes to low plasma HDL cholesterol levels 82. Carette, J.E. et al. (2011) Ebola virus entry requires the choles-
and impaired cholesterol efflux capacity. Atherosclerosis 249, terol transporter Niemann-Pick C1. Nature 477, 340–343
140–147 83. Pericleous, M. et al. (2017) Wolman’s disease and cholesteryl
63. Mesmin, B. et al. (2013) Insights into the mechanisms of sterol ester storage disorder: the phenotypic spectrum of lysosomal
transport between organelles. Cell. Mol. Life Sci. 70, 3405–3421 acid lipase deficiency. Lancet Gastroenterol. Hepatol. 2,
64. Goto, A. et al. (2016) Oxysterol-binding protein activation 670–679
at endoplasmic reticulum-Golgi contact sites reorganizes 84. Wustner, D. et al. (2016) Potential of BODIPY-cholesterol for
phosphatidylinositol 4-phosphate pools. J. Biol. Chem. 291, analysis of cholesterol transport and diffusion in living cells.
1336–1347 Chem. Phys. Lipids 194, 12–28
65. Lim, C.Y. et al. (2019) ER-lysosome contacts enable cholesterol 85. Schroeder, F. et al. (1971) Fluorometric evidence for the binding
sensing by mTORC1 and drive aberrant growth signalling in of cholesterol to the filipin complex. J. Antibiot. (Tokyo) 24,
Niemann-Pick type C. Nat. Cell Biol. 21, 1206–1218 846–849
66. Dong, R. et al. (2016) Endosome-ER contacts control actin nu- 86. McIntosh, A.L. et al. (2008) Fluorescence techniques using
cleation and retromer function through VAP-dependent regula- dehydroergosterol to study cholesterol trafficking. Lipids 43,
tion of PI4P. Cell 166, 408–423 1185–1208
67. Du, X.M. et al. (2011) A role for oxysterol-binding protein-related 87. Schroeder, F. et al. (1988) Fluorescence properties of
protein 5 in endosomal cholesterol trafficking. J. Cell Biol. 192, cholestatrienol in phosphatidylcholine bilayer vesicles. Biophys.
121–135 Chem. 32, 57–72
68. Ouimet, M. et al. (2016) miRNA Targeting of oxysterol-binding 88. Johnson, B.B. et al. (2017) Mechanistic insights into the
protein-like 6 regulates cholesterol trafficking and efflux. cholesterol-dependent binding of perfringolysin O-based probes
Arterioscler. Thromb. Vasc. Biol. 36, 942–951 and cell membranes. Sci. Rep. 7, 13793
69. Alpy, F. et al. (2013) STARD3 or STARD3NL and VAP form a 89. Wustner, D. et al. (2011) Quantitative assessment of sterol traffic
novel molecular tether between late endosomes and the ER. in living cells by dual labeling with dehydroergosterol and
J. Cell Sci. 126, 5500–5512 BODIPY-cholesterol. Chem. Phys. Lipids 164, 221–235
70. Wilhelm, L.P. et al. (2017) STARD3 mediates endoplasmic 90. Holtta-Vuori, M. et al. (2016) Use of BODIPY-cholesterol (TF-Chol)
reticulum-to-endosome cholesterol transport at membrane con- for visualizing lysosomal cholesterol accumulation. Traffic 17,
tact sites. EMBO J. 36, 1412–1433 1054–1057
71. Costello, J.L. et al. (2017) ACBD5 and VAPB mediate membrane 91. Sparrow, C.P. et al. (1999) A fluorescent cholesterol analog
associations between peroxisomes and the ER. J. Cell Biol. 216, traces cholesterol absorption in hamsters and is esterified
331–342 in vivo and in vitro. J. Lipid Res. 40, 1747–1757
72. Wong, Y.C. et al. (2018) Mitochondria-lysosome contacts regu- 92. Wolfman, L. and Sachs, B.A. (1964) Separation of cholesterol
late mitochondrial fission via RAB7 GTP hydrolysis. Nature 554, and demosterol by thin-layer chromatography. J. Lipid Res. 5,
382–386 127–128
73. Charman, M. et al. (2010) MLN64 mediates egress of cholesterol 93. Lange, Y. (1992) Tracking cell cholesterol with cholesterol oxi-
from endosomes to mitochondria in the absence of functional dase. J. Lipid Res. 33, 315–321
Niemann-Pick Type C1 protein. J. Lipid Res. 51, 1023–1034 94. Gardner, M.S. et al. (2017) Simultaneous quantification of free
74. Balboa, E. et al. (2017) MLN64 induces mitochondrial dysfunc- cholesterol, cholesteryl esters, and triglycerides without ester hy-
tion associated with increased mitochondrial cholesterol con- drolysis by UHPLC separation and in-source collision induced
tent. Redox Biol. 12, 274–284 dissociation coupled MS/MS. J. Am. Soc. Mass Spectrom. 28,
75. Castellano, B.M. et al. (2017) Cholesterol sensing lysosomal 2319–2329
cholesterol activates mTORC1 via an SLC38A9-Niemann-Pick 95. Solanko, K.A. et al. (2015) Fluorescent sterols and cholesteryl
C1 signaling complex. Science 355, 1306–1311 esters as probes for intracellular cholesterol transport. Lipid
76. Wang, S. et al. (2015) Lysosomal amino acid transporter Insights 8, 95–114
SLC38A9 signals arginine sufficiency to mTORC1. Science 96. Maxfield, F.R. and Wustner, D. (2002) Intracellular cholesterol
347, 188–194 transport. J. Clin. Invest. 110, 891–898

Trends in Cell Biology, Month 2020, Vol. xx, No. xx 15