Distribution and Functions of the Novel Membrane-Spanning ... · 5 this thesis, except where it is necessary to name the individual orthologues for historical purposes. 1.4.1 Expression

Distribution and Functions of the

Novel Membrane-Spanning Four-

Domains, Subfamily A Member

HCA112

Wendy Parker, B. Biotechnology (Hons) (University of Adelaide)

A thesis submitted for the degree of Doctor of Philosophy

School of Molecular and Biomedical Science

Faculty of Sciences

The University of Adelaide

Adelaide, South Australia, Australia

July, 2009

1

Chapter 1: Introduction

1.1 Preface

This project commenced as a study of the function of what was believed to be a Leydig

cell-specific protein and for this reason, a short review of the testis and Leydig cells is

presented. However, the protein (HCA112) was found to have much wider expression than

just Leydig cells. Furthermore, at the time of commencing the study (and presently), the

international literature on HCA112 and orthologous proteins consisted of only two

publications. Very little is known about its functions, although more is known about some

other members of the gene family to which it belongs. The introduction of this thesis

contains, therefore, a review of this gene family. However, it is necessary to also review

areas of cell biology that have become relevant during the course of the study, and some

other proteins that have attracted attention because they may (or may not!) have functional

relationships with HCA112.

1.2 Background to the study

The original motivation behind this study was to identify specific genes that might be

involved in the differentiation of Leydig cells, regulation of testosterone synthesis and the

pathogenesis of male infertility. To this end, a microarray study of gene expression was

performed, comparing RNA harvested from the testis of rats whose Leydig cells had been

destroyed by the Leydig cell-specific cytotoxin ethane dimethanesulphonate (EDS) with

testicular RNA from control rats. For the duration of the experiment, physiological levels

of testosterone were maintained in the EDS-treated rats from a silastic tubing testosterone

implant. The purpose of the hormone replacement was mainly to prevent a reactive

increase in luteinising hormone (LH) release from the anterior pituitary, which could

induce differentiation of new Leydig cells, and partly to maintain an androgen-dependent

phenotype in the testis in the absence of Leydig cells (Ivell, unpublished 2003). Expression

of approximately 200 genes was altered by ablation of Leydig cells, based on statistical

differences in differential transcription levels between the EDS-treated and control groups.

Of these, approximately half were known to be involved in Leydig cell functions and half

2

were previously undescribed, at least in the testis. Quantitative reverse transcriptase

polymerase chain reaction (RT-PCR) revealed that transcripts from ten previously un-

described genes were significantly less abundant in the EDS-treated group compared to the

controls. This suggested that in the testis, expression of these genes is confined to Leydig

cells. One of these ten genes was GS188/KEG2, which has a human orthologue named

hepatocellular carcinoma-associated antigen 112 (HCA112). This gene was chosen for

further characterisation and it is the subject of the present study.

1.3 The testis

In the male reproductive system, the testes perform two major functions – generation of the

male gametes (sperm) and production of androgens (the predominant one being

testosterone). These functions are organised into two respective compartments: the

seminiferous tubules (and their collecting system) and the supporting interstitial tissue. The

seminiferous tubules contain somatic cells (Sertoli cells) and the developing germ cells

(spermatogonia and developing sperm), and are surrounded by peri-tubular cells. Sertoli

cells have an intimate physical relationship with the differentiating germ cells, providing

nourishment, protection and growth factors for maturing sperm, as well as assisting in the

removal of cytoplasm as the gametes mature (Russell and Clermont, 1976; Cheng and

Mruk, 2002; Mruk and Cheng, 2004). Tight junctions between adjacent Sertoli cells form

the blood-testis barrier, which provides a degree of immunological privilege for the post-

meiotic and hence potentially ‘foreign’ germ cells (Russell and Clermont, 1976; Russell,

1979; Fujisawa, 2001; Mruk and Cheng, 2004; Walker and Cheng, 2005). The interstitial

region surrounds the seminiferous tubules and contains the androgen synthesising Leydig

cells (Haider, 2004). Also located in the interstitium are blood and lymphatic vessels,

nerves, fibroblasts and bone marrow-derived cells that include monocytes, macrophages,

lymphocytes, plasma cells and mast cells (Russell, 1996). Of the blood-derived cells,

macrophages are the most abundant. There is one macrophage to every four Leydig cells in

the testis (Bergh, 1987), while other blood-derived cell types are infrequent.

3

1.3.1 Leydig cells and testosterone production

Leydig cells exist in clusters surrounding blood vessels in the interstitium of the testis

(Russell, 1996). They have structural features characteristic of steroid-secreting cells,

including a large cytoplasm/nucleus ratio, large amounts of smooth endoplasmic reticulum

(SER), a prominent Golgi apparatus, mitochondria that have characteristic tubular or

finger-shaped cristae, and the presence of lipid droplets (Russell, 1996).

Testosterone production by Leydig cells is under the control of hormones secreted by

the anterior pituitary. The hypothalamus produces gonadotropin-releasing hormone

(GnRH) in a pulsatile pattern, and this stimulates pulsatile secretion of LH and follicle

stimulating hormone (FSH) by the gonadotrophin-producing cells of the anterior pituitary.

In turn, LH binds to its receptors on Leydig cells, stimulating pulses of testosterone

synthesis (Haider, 2004).

Cholesterol, the precursor of testosterone synthesis, is obtained by Leydig cells from

lipoproteins in the circulation, and the cells can also synthesise cholesterol de novo

(Freeman and Rommerts, 1996). Binding of LH to its G-protein coupled receptor in Leydig

cells, activates adenylate cyclase, leading to the production of the intracellular second

messenger cyclic adenosine monophosphate (cAMP). This results in activation of

testosterone synthesis by initiating transport of cholesterol from cellular stores (e.g. the

plasma membrane or lipid droplets) to the outer membrane of the mitochondria, via a

mechanism that involves cytoskeletal elements (actin and tubulin) and sterol carrier

proteins (Papadopoulos et al., 1990; Hall, 1991). LH stimulation of Leydig cells also

induces transcription of the cytosolic steroidogenic acute regulatory protein (StAR) and

increased activity of steroidogenic enzymes 17�-hydroyylase/C17-20 lyase (P450c17),

cytochrome P450 side chain cleavage (P450scc) and 3�-hydroxysteroid dehydrogenase

(3�-HSD). The protein StAR is responsible for the transfer of cholesterol from the outer to

the inner mitochondrial membrane (Stocco, 1999). Once in the mitochondria, the enzyme

P450scc, located on the matrix side of the inner mitochondrial membrane, catalyses the

hydroxylation of cholesterol (removing the 6 carbon side chain at C21) to form

pregnenolone (Haider, 2004). Pregnenolone is then transferred to the membrane of the

SER where all subsequent steps in androgen synthesis take place. First, pregnenolone is

converted to progesterone by the enzyme 3�-HSD. Progesterone is then hydroxylated at

4

C17 to form 17�-hydroxyprogesterone, and the side chain of C17 is then cleaved by the

enzyme P450c17, producing androstenedione. Finally, the ketone at C17 is reduced to an

alcohol by type 3 17�-hydroxysteroid dehydrogenase (17�-HSD), giving rise to

testosterone (Payne and O'Shaughnessy, 1996; Payne and Hales, 2004), see Figure 1-1 for

cartoon representation of enzyme localisation.

1.4 Distribution and function of HCA112 and LR8

HCA112 was identified by Wang and colleagues, who used auto-antibodies from patients

with hepatocellular carcinoma to screen a cDNA expression library with an aim to identify

genes associated with liver cancer (Wang et al., 2002). The gene encoding HCA112 has 7

exons and spans approximately 4 kb of DNA on human chromosome 7q36. An

orthologous gene, named GS188 or KEG2, was identified in mouse kidney by Nakajima et

al. during a screen to identify genes induced in the proximal tubules during experimental

proteinuria (Nakajima et al., 2002). A gene named LR8 (see below) is located in the

immediate vicinity of the HCA112 orthologue in all species that have been examined. The

predicted amino acid sequence of HCA112 has approximately 30% amino acid sequence

identity with the gene product of LR8, while searches of protein databases do not reveal

any other proteins that have comparable similarity. HCA112 and LR8 also share predicted

topologies. HCA112 is predicted to have four strongly hydrophobic domains, each long

enough to pass through the plasma membrane, intracellular N- and C-termini and unevenly

sized extracellular loops. The resulting four-transmembrane (TM) topology is similar to

that predicted for LR8 (Louvet et al., 2005), as well as members of a newly described

family of small, four TM domain proteins named the membrane-spanning 4-domains,

subfamily A (MS4A) (Ishibashi et al., 2001; Liang et al., 2001; Liang and Tedder, 2001).

The close chromosomal proximity of HCA112 and LR8, the amino acid sequence similarity

of HCA112 and LR8, and similarities in the predicted structures and topologies of the two

proteins, suggests that HCA112 and LR8 may have arisen from gene duplication.

Furthermore, they may share some aspects of function and regulation.

For convenience and clarity, HCA112 and LR8 and their orthologues in species other

than human will be referred to as ‘HCA112’ and ‘LR8’, respectively, for the remainder of

5

this thesis, except where it is necessary to name the individual orthologues for historical

purposes.

1.4.1 Expression pattern and potential function of HCA112

Using RT-PCR, Wang et al. reported that HCA112 was transcribed in human kidney, lung

and pancreas, but not in brain, heart, liver, placenta, skeletal muscle, colon, ovary,

peripheral blood leukocytes, prostate, small intestine, spleen, testis or thymus (Wang et al.,

2002). This is similar to the pattern of expression of a mouse orthologue, named GS188,

determined by Nakajima and colleagues using Northern blot analysis. These workers

identified high levels of GS188 mRNA in kidney and lung, and to a lesser extent in spleen,

liver and colon. GS188 transcripts were not detected in heart, brain or skeletal muscle

(Nakajima et al., 2002). As mentioned above, Nakajima et al. found that the proteinuria

that follows administration of large amounts of bovine serum albumin (BSA) induces

transcription of GS188 in kidney proximal tubule cells. The site of induction was

determined using RT-PCR analysis of RNA harvested from laser-micro-dissected proximal

tubules, and by in situ hybridisation. These findings lead the authors to hypothesise that

GS188 might play a role in the immune response that leads to tubulo-interstitial damage in

this model (Nakajima et al., 2002).

1.4.2 Expression pattern and potential function of LR8

LR8 was identified first in humans by Lurton and colleagues, who believed the transcript

was expressed specifically by a subset of lung fibroblasts (Lurton et al., 1999). A rat

orthologue named TORID (which stands for tolerance-related and induced transcript) was

identified as a cDNA in 2005. Transcripts of TORID were found to be over-expressed

specifically in tolerated heart allografts during a study of tolerance induced by donor-

specific blood transfusion (Louvet et al., 2005). Expression of TORID mRNA was

analysed in normal rat tissues by real-time RT-PCR, and the highest expression was

observed in the spleen, liver, lung, lymph nodes, and colon, and to a lesser extent in the

bone marrow, aorta, thymus and intestine. Lower levels of the transcript were detected in

peripheral blood mononuclear cells, skeletal muscle, brain, testis, kidney and heart (Louvet

et al., 2005). Although fibroblasts are components of most tissues, it seems unlikely that

TORID (LR8) expression is limited to fibroblasts.

6

Louvet et al. proposed that TORID was transcribed preferentially in lymphoid tissues

and they went on to analyse its expression in leukocyte populations (Louvet et al., 2005).

Expression of the gene was found to be highest in peritoneal macrophages and dendritic

cells (DCs), specifically CD4+ DCs. A rabbit polyclonal antibody produced against

recombinant TORID was used to investigate expression of TORID in rat spleen and lymph

node. Expression of the protein was observed in the splenic red pulp and to a lesser extent

in the periarteriolar lymphoid sheath (PALS, T-cell-rich area), the marginal zones and the

B-cell follicles. At a cellular level, most of the B-cells and T-cells were TORID-negative.

Nevertheless, the TORID-positive cells all expressed CD45 indicating that they were bone-

marrow derived, although these workers were unable to link expression of TORID to a

particular cell type (Louvet et al., 2005). Cells in the B-cell follicles labelled with anti-

TORID antibodies have a reticular pattern, which is suggestive of expression by follicular

dendritic cells (FDCs) or tingible body macrophages.

At a subcellular level, TORID was found to exhibit a peri-nuclear ring-like staining

pattern, which the authors suggested indicated localisation to the nuclear envelope (Louvet

et al., 2005). In rat peritoneal macrophages, rat DCs and human immature monocyte-

derived DCs, levels of TORID mRNA were reduced in response to stimulation with LPS,

CD40L or TNF� plus polyl:C, respectively. This result suggests that expression of TORID

may be related to the activation or maturation of DCs and macrophages. Furthermore,

over-expression of a TORID-GFP fusion protein in bone marrow–derived DCs (BMDCs)

using recombinant adenovirus, resulted in reduced production of IL-12p40, TNF� and IL-

10 in comparison with BMDCs transduced with control adenovirus. Together, the results

suggest strongly that TORID plays a role in maturation of DCs (Louvet et al., 2005).

An LR8 orthologue, named Clast1, has been described in mice (Maeda et al., 2006).

Maeda et al. showed that levels of Clast1 transcripts were high in kidney, colon, lung and

liver, while expression was lower, but detectable, in cerebellum, cerebrum, thymus, heart,

spleen, and intestine. The distribution of Clast1 transcription is similar, therefore, to that

described for TORID. In Clast1 gene knockout (GKO) mice, 65% of offspring showed

severe ataxia by two weeks of age. When pairs of asymptomatic Clast1-GKO mice were

mated, approximately 18% of the progeny showed severe ataxia. Ataxia appeared

sporadically in homozygous Clast1-GKO mice, but not in heterozygotes. The authors

7

found that the cerebellum in Clast1-GKO mice with severe ataxia was small in size and

had severely aberrant lobulation, loss of the internal granule cell layer, and disorganised

Purkinje cells. Because Clast1 mRNA was expressed in the cerebellar granule cells of

normal adult mice and in granule cell precursors, the authors hypothesised that Clast1 is

required for normal development of these cells (Maeda et al., 2006).

1.5 Membrane-spanning 4-domains, subfamily A (MS4A) proteins

The predicted four TM topology of HCA112 is similar to that of LR8 and to proteins of the

large, newly described protein family membrane-spanning 4-domains, subfamily A

(MS4A). The founding members of MS4A are the B-cell-specific surface protein CD20,

the high affinity IgE Fc receptor � chain (Fc�RI�) and a haematopoietic-cell-specific

protein HTm4 (haematopoietic cell specific transmembrane 4). At least 23 other members

of the MS4A family have been described in humans, mice and rats. The functions of most

members of the MS4A family remain unknown, but some appear to have signalling or ion

channel functions.

The family is characterised by the small size of the proteins (approximately 250 amino

acids), the presence of four TM domains, and amino acid identity, especially in the first

and second TM domains. The genes encoding MS4A members show similar intron / exon

organisation and, with the exception of HCA112 and LR8, they are clustered on

chromosome 11q12 (Ishibashi et al., 2001; Liang and Tedder, 2001). In addition, common

amino acid motifs have been identified in the TM domains of MS4A family members

(Ishibashi et al., 2001; Liang and Tedder, 2001). The motifs VLGAIQIL (Ishibashi et al.,

2001), LGAXQI and LSLG (Liang and Tedder, 2001) have been identified within the first

TM domain, the second TM domain contains the motifs GYPFWG (Ishibashi et al., 2001;

Liang and Tedder, 2001) and FIISGSLS (Liang and Tedder, 2001), while the third TM

domain contains the motifs SLX2NX2 and SX3AX2G (Liang and Tedder, 2001). A further

conserved feature of MS4A proteins is that the predicted second extracellular loop is rich

in tyrosine and proline residues (Ishibashi et al., 2001). The functional significance of

these common amino acids is unknown.

As noted by Louvet and colleagues, HCA112 and LR8 have 10-20% amino acid

identity overall with members of the MS4A, they are of similar size, they share predicted

8

topologies and there is a high degree of identity in the predicted amino acid sequences

within the first three TM domains. HCA112 and LR8 have intron / exon organisations

similar to genes encoding the MS4A proteins. This prompted the authors to suggest that

despite the different chromosomal localisation of the genes encoding HCA112/LR8 and

the MS4A proteins, the similarities in protein structure and gene layout indicate all of these

genes have evolved from a common ancestor, and that HCA112 and LR8 are legitimate

members of the MS4A family of proteins (Louvet et al., 2005).

1.5.1 CD20

One of the most extensively studied members of the MS4A family is CD20 (formerly

know as B1). CD20 is a 33-36 kDa non-glycosylated phospho-protein whose exact

function is unknown. The molecule is thought to serve as a calcium channel, to initiate

intracellular signals, and to modulate cell growth and differentiation (Tedder and Engel,

1994). As a founding member of the MS4A family, CD20 shares the characteristic

properties of this protein family. The molecule is expressed exclusively by pre-B cells and

mature B-cells, but it is lost on differentiation of the cells into plasma cells. CD20 is used

as a prime target for near-selective monoclonal antibody (mAb)-based treatment of mature

B-cell malignancies and for a range of disorders associated with reactive B-cells (including

virus-associated lympho-proliferation disorders and some autoimmune conditions) (von

Schilling, 2003).

1.5.1.1 Function of CD20

CD20 is involved in regulating B-cell activation, cell-cycle progression and TM Ca2+

conductance. It has been reported that CD20 associates with itself and with other proteins,

including major histocompatibility complex class II (MHC-II) molecules, CD40 (Leveille

et al., 1999), and recently with the novel MS4A family member MS4A8B/L985P (Deans et

al., 2008). Using fluorescence energy transfer, CD20 has been shown to be in close

proximity with MHC class I (MHC-I) molecules and with the 4 TM proteins CD37, CD53,

CD81 and CD82 (Szollosi et al., 1996). Experiments using biochemical cross-linking of

cell surface molecules, followed by immunoprecipitation, have shown that CD20 forms

homo-dimeric and homo-tetrameric complexes (Bubien et al., 1993).

9

Due to the structural similarity of CD20 with some known ion channels, Bubien and

colleagues investigated the role of CD20 in plasma membrane ion conductance and

regulation of intracellular calcium. They found that transfection of Daudi B

lymphoblastoid cells with CD20 resulted in increased TM calcium conductance, and that

this was enhanced by binding of certain anti-CD20 mAbs (Bubien et al., 1993; Kanzaki et

al., 1997). Depending on the specific epitope of the mAb, the alteration in calcium

conductance induced was either acute (mAb 1F5), or apparent only after long-term

exposure (anti-Bla mAb) (Bubien et al., 1993). As discussed above, CD20 associates with

the newly described MS4A family member MS4A8B/L985P (Deans et al., 2008), and it

has been shown that these proteins form a calcium and strontium permeable store-operated

calcium channel (SOC) that is located in lipid rafts (Li et al., 2003; Bangur et al., 2004;

Deans et al., 2008). It is thought that MS4A8B/L985P may be able to compensate

functionally for CD20, as CD20 GKO mice appear to have normal immune responses

(Uchida et al., 2004; Deans et al., 2008).

Binding of specific mAbs to CD20 has been shown to cause enhanced phosphorylation

of the molecule (Tedder and Schlossman, 1988), induction of oncogene expression (c-myc

and B-myb) (Smeland et al., 1985; Golay et al., 1992), and increased expression of CD18,

CD58 and MHC-II molecules (Clark and Shu, 1987; White et al., 1991; Tedder and Engel,

1994). Most anti-CD20 mAbs inhibit progression of mitogen-stimulated B-cells from the

G1 phase of the cell cycle into the S/G2+M stages (Tedder et al., 1985; Tedder et al.,

1986). However, one anti-CD20 mAb (1F5) is capable of activating resting tonsilar B-cells

and driving them into the G1 phase of the cell cycle (Smeland et al., 1985). It is thought

that the regulation of cell-cycle in B-cells by CD20 may involve increases in intracellular

calcium concentration by the SOC and/or ion channel function of the molecule (Tedder

and Engel, 1994). Another MS4A protein HtM4 (see below) might interact with CD20 and

contribute to cell cycle modulation by CD20.

1.5.1.2 Targeting CD20 as an anti-tumour treatment

The chimeric anti-CD20 mAb, rituximab, is used routinely as the treatment for some B-cell

malignancies. It has been shown to have anti-tumour properties, but the mechanism of

action in vivo is still controversial. Complement-mediated lysis and antibody-dependent

10

cell-mediated cytotoxicity (ADCC) are thought to be involved (Harjunpaa et al., 2000).

However, rituximab has also been shown to exert apoptotic or anti-proliferative effects

directly (Maloney et al., 2002), but this remains contentious. It has been established,

however, that rituximab induces only modest levels of apoptosis unless there is extensive

cross-linking (Shan et al., 1998; Ghetie et al., 2001). The apoptotic / anti-proliferative

effects of rituximab on CD20+ B-cells are thought to require translocation of CD20 into

lipid rafts in the plasma membrane (Unruh et al., 2005), which is followed by increased

intracellular calcium levels and downstream apoptotic signalling (caspase activation)

(Janas et al., 2005) that involves the SRC-family of kinases (Deans et al., 1998; Deans et

al., 2002).

1.5.2 Fc�RI�

The second of the founding members of the MS4A family is Fc�RI�, the common �

subunit of two multi-subunit plasma membrane immunoglobulin common chain (Fc)

receptors, Fc�RI (high affinity IgE) and Fc�RIII (low affinity IgG) (Ra et al., 1989a). The

� subunit has been studied most extensively in the context of the IgE Fc receptor Fc�RI,

and is a candidate susceptibility gene for human atopic allergies (Shirakawa et al., 1994).

Tetrameric Fc�RI is expressed on mast cells and basophils, while a trimeric form is

expressed on monocytes, Langerhans cells and DCs in humans that does not contain

Fc�RI�. When the Fc�RI-bound IgE antibody on mast cells and basophils is cross-linked

by multivalent antigen, intracellular signalling leads to secretion of pre-formed histamine,

synthesis of leukotrienes, as well as synthesis and secretion of range of cytokines that

include IL-4, IL-6, TNF� and GM-CSF. These mediators lead to local inflammation, with

recruitment and activation of polymorphonuclear leukocytes, including eosinophils (Galli

and Costa, 1995; Turner and Kinet, 1999).

The Fc�RI is expressed at the cell surface predominantly as a tetrameric (���2)

complex. However, in humans the � subunit is not obligatory for expression, and trimeric

(��2) complexes are also present (Bieber et al., 1996). The � subunit contains two

immunoglobulin-type domains and binds the Fc portion of IgE with high affinity. As a

prototypic member of the MS4A family, the � subunit has four TM-domains and

cytoplasmic tails, with the carboxy terminus containing a signal-transducing motif called

11

the immuno-receptor tyrosine-based activation motif (ITAM). The � subunit contains a

single TM domain with an ITAM in its cytoplasmic tail, and is present in the receptor as a

disulphide-linked homo-dimer (Turner and Kinet, 1999). The ITAM has a consensus

sequence of [D/E]XXYXXLX7-11YXXL[L/I], where the tyrosine residues are sites of

tyrosine kinase phosphorylation (Reth, 1989; Cambier, 1995). When phosphorylated, the

tyrosine residues in the � and � subunit ITAMs interact with signalling molecules (protein

tyrosine kinases) through their SRC homology 2 (SH2) domains. The ITAMs of the � and

� subunits vary slightly in their sequence (�– DRVYEELNIYSATYSEL, �-

DGVYTGLSTRNGGETYETL) and this appears to confer different functions (see below).

The ITAM of the � subunit binds the SRC family kinases LYN and FYN, whereas the

ITAM of the � subunit binds the kinase SYK (Turner and Kinet, 1999; Kraft and Kinet,

2007).

1.5.2.1 Signalling via the Fc�RI

The Fc�RI is activated by cross-linking induced when receptor-bound IgE is cross-liked by

multivalent antigen. The resulting aggregation of the � subunit initiates a complex

intracellular signalling pathway, resulting ultimately in the effector functions summarised

above. Two Fc�RI-initiated signalling pathways have been described - a primary pathway

and a complementary pathway that regulates mainly Fc�RI-induced mast cell

degranulation. Signalling via both the primary and the complementary pathways results in

degranulation, eicosanoid production (ERK and MAPK signalling) and cytokine

production (gene transcription). However, only the primary pathway, which originates

mainly from the activities of the protein tyrosine kinase LYN, results in calcium

mobilisation.

After cross-linking of Fc�RI, relay of information depends on interaction of the �

subunit with the � and � subunits, possibly via its Ig domains (Donnadieu et al., 2000a),

because the cytoplasmic domain of the � subunit lacks motifs with which to interact with

signalling targets (Alber et al., 1991). This interaction leads to activation of LYN, which is

bound to the � subunit under resting conditions. Activated LYN phosphorylates tyrosine

residues in the � and � subunit ITAMs, the latter recruits the tyrosine kinase SYK, and this

is also phosphorylated and activated by LYN (El-Hillal et al., 1997). LYN also

12

phosphorylates ITAMs (� and � subunit) on neighbouring Fc�RI molecules, resulting in

signal amplification (Turner and Kinet, 1999; Kraft and Kinet, 2007). LYN- and SYK-

dependent intracellular signalling follows, which involves adapter proteins called linker for

activation of T-cells (LAT) (Saitoh et al., 2000), SH2-domain-containing leukocyte protein

of 76 kDa (SLP76) (Pivniouk et al., 1999) and Bruton’s tyrosine kinase (BTK) (Kawakami

et al., 2000). Phospholipase C�1 is activated, generating the second messengers inositol-

1,4,5-triphosphate (I(1,4,5)P3) and diacylglycerol (DAG). These are responsible for the

release of calcium from intracellular stores (ER) and activation of various protein kinase C

(PKC) isoforms, respectively (Lin et al., 1996; Dombrowicz et al., 1998; Turner and Kinet,

1999; On et al., 2004). Calcium release from the endoplasmic reticulum (ER) results in

store depletion and STIM1-mediated opening of plasma membrane store-operated calcium

channels (SOC), leading to an influx of extracellular calcium (Peinelt et al., 2006).

A study using SYK-deficient mast cells showed that these cells fail to degranulate,

synthesise leukotrienes, or secrete cytokines upon stimulation through Fc�RI,

demonstrating that SYK is essential for Fc�RI signalling (Costello et al., 1996). LYN-

deficient bone marrow-derived mast cells, on the other hand, are able to undergo normal

degranulation and production of cytokines, although calcium mobilisation is impaired in

these cells compared with control cells. These findings suggest that an alternative pathway

for degranulation exists in these cells that is not dependent on LYN (Nishizumi and

Yamamoto, 1997). It has been reported that this alternative pathway of IgE-mediated

signalling through Fc�RI involves SYK, FYN and the adaptor protein growth-factor-

receptor-bound protein 2 (GRB2)-associated binding protein 2 (GAB2) (Gu et al., 2001;

Parravicini et al., 2002). When Fc�RI is cross-linked, GAB2 is phosphorylated by FYN,

leading to activation of phosphoinositide 3-kinase (PI3K) and production of

phosphatidylinositol 3,4,5-trisphosphate (PIP3). PIP3 recruits PI3K-dependent kinase 3-

phosphoinositide-dependent protein kinase 1 (PDK1), which leads to activation of protein

kinase C� (PKC�) and degranulation of the mast cell. FYN and LYN can also activate

sphingosine kinase, which also leads to an influx of extracellular calcium (Kraft and Kinet,

2007).

13

1.5.2.2 Role of the � subunit of Fc�RI in signalling

In rodents, the � subunit appears to be essential for expression of the Fc�RI and for its

normal function (Turner and Kinet, 1999). However, the � subunit it is not essential in

humans, as a number of human cell types (including mast cells and basophils) express

Fc�RI either with or without the � subunit, and both forms appear to function normally

(Bieber et al., 1996; Dombrowicz et al., 1998). Polymorphisms in Fc�RI� have been

associated with atopic traits in humans (Shirakawa et al., 1994; Lee et al., 2008), although,

these mutations (I181L, V183L and E237G) do not seem to have any effect on Fc�RI

function in vitro (Donnadieu et al., 2000a). To understand the role that the � subunit plays

in regulating the IgE-associated allergic response, Lin and colleagues performed

experiments in cell lines (NIH 3T3 and U937) expressing human Fc�RI as either ���2

tetramers or ��2 trimers, and also associated cytoplasmic signalling molecules

(Scharenberg et al., 1995). After stimulation via IgE, cells expressing the Fc�RI tetramers

were found to have 5- to 7-fold higher LYN-dependent � subunit tyrosine phosphorylation,

approximately 20-fold higher SYK kinase activation (tyrosine phosphorylation) and five-

fold higher levels of cytosolic calcium than cells expressing trimers. These workers

concluded that Fc�RI� functions as a signal amplifier and that the � dimer functions as an

autonomous activation molecule (Lin et al., 1996). To further investigate the function of

the � subunit, Dombrowicz et al. generated transgenic mice expressing humanised trimeric

and tetrameric Fc�RI molecules. Their results paralleled those found by Lin and

colleagues. Furthermore, the study showed that the � subunit also plays an amplification

role in vivo. Transgenic mice were injected three times with humanised IgE, followed by

challenge with antigen (intravenous). Fc�RI ���2-expressing mice showed greater systemic

anaphylaxis (measured by temperature drop) than Fc�RI ��2-expressing mice

(Dombrowicz et al., 1998). Together, these studies show that Fc�RI� enhances LYN-

dependent phosphorylation of tyrosine residues in both the � subunit and SYK, and this

results in amplification of downstream signalling, including calcium mobilisation, mast

cell degranulation and systemic anaphylaxis.

As mentioned above, the � and � subunit ITAM sequences are slightly different. The �

subunit ITAM fits the consensus, whereas the � subunit ITAM differs in two ways – the

spacer region contains a third tyrosine residue and it is also one amino acid shorter than the

14

consensus. This variation in ITAM sequence is shown to be the reason why Fc�RI� cannot

bind SYK (On et al., 2004). The canonical tyrosine (Tyr219) is responsible for the

amplification effect of the � subunit, due to its ability to recruit LYN (On et al., 2004) (see

above). Mutation of this residue (Y219F) leads to reduced receptor-associated LYN, and

reduction in calcium mobilisation, mast cell degranulation and cytokine synthesis

(Furumoto et al., 2004; On et al., 2004). The non-canonical tyrosine of the � subunit ITAM

(Tyr225), on the other hand, has an inhibitory effect on Fc�RI-signalling (Furumoto et al.,

2004; On et al., 2004). Mutation of this residue (Y225F) results in reduced

phosphorylation of the negative regulator named SH2-domain-containing inositol-5-

phosphatase (SHIP (Huber et al., 1998)), and enhanced cytokine synthesis and secretion.

SHIP and LYN are still recruited to the mutant receptor (Furumoto et al., 2004). Thus,

Tyr225 of Fc�RI� is involved in the phosphorylation and activation of SHIP, which

functions to dephosphorylate and inactivate (PIP3) (generated by PI3K) (Rohrschneider et

al., 2000), leading to decreased activity of protein kinase B, protein kinase C, p38 and

extracellular signal related kinase (ERK). This results in reduced NF-�B activity and

reduced levels of IL-6 mRNA and protein (Kalesnikoff et al., 2002). It has been suggested

that the amplifier/inhibitor functions of Fc�RI� might be a negative-feedback mechanism

that occurs during high-intensity stimulation of Fc�RI (Xiao et al., 2005).

1.5.2.3 Regulation of synthesis and cell surface expression of the Fc�RI

Non-covalent association of the newly synthesised �, � and � subunits of the Fc�RI occurs

co-translationaly in the ER (Fiebiger et al., 2005). This association supports core

glycosylation of the � subunit, a process that is required for proper folding, signal-peptide

cleavage, export from the ER and subsequent cell surface expression (Letourneur et al.,

1995b; Fiebiger et al., 2005). Following core glycosylation of the � subunit in the ER, the

complex containing the immature � subunit is transferred from the ER to the Golgi for

terminal glycosylation, followed by transport of the mature receptor complex to the cell

surface. The � subunit appears to be essential for export of the receptor complex from the

ER. This is because the � subunit contains an ER retention motif in its cytoplasmic tail,

and masking of this by the � subunit is necessary to allow export and maturation of the

complex (Letourneur et al., 1995a).

15

1.5.2.4 Role of Fc�RI� in receptor expression

As discussed above, human mast cells and basophils express both tetramer (���2) and

trimer (��2) forms of Fc�RI, while only the trimeric form is expressed by monocytes,

cutaneous Langerhans cells, and dermal and peripheral blood DCs (Bieber et al., 1992;

Wang et al., 1992; Maurer et al., 1994; Osterhoff et al., 1994). It has been noted that cell

surface Fc�RI expression by cells that express the trimer is 10- to 100-fold lower than by

cells that express the tetramer (Gounni et al., 1994; Sihra et al., 1997; Kita et al., 1999). To

investigate the reason for this difference in receptor density, Donnadieu et al. used Fc�RI

��2 (trimer)- and ���2 (tetramer)-expressing cell lines, and they found that the � subunit

accelerates processing and maturation of the nascent � subunit, thus leading to greater cell

surface expression (Donnadieu et al., 2000b). This, together with the finding that the �

subunit is required for cell surface expression of Fc�RI in rodents but not humans (Ra et

al., 1989b), suggests that the rodent � subunit extracellular domain may contain an

additional ER retention motif which, in this case, is masked by the � subunit (Blank et al.,

1991).

A splice variant of Fc�RI� that has antagonistic function has been identified and named

�T. This transcript retains the fifth intron, resulting in a truncated form with a 16-amino

acid sequence replacing the fourth TM domain and the carboxy terminus (Donnadieu et al.,

2003). �T inhibits � chain maturation by competing with the full length � subunit for

binding to nascent � subunit in the ER. This has the effect of sequestering the Fc�RI

complex away from the normal maturation pathway, resulting in accumulation of ��T�2

Fc�RI in the ER of transfected cells, and thus low levels of cell surface expression. In

human basophils, � and �T are co-expressed in variable proportions, thus it has been

hypothesised that the ratio of full length to truncated Fc�RI� might influence susceptibility

to allergic disorders (Donnadieu et al., 2003).

1.5.3 HTm4

The third of the founding members of the MS4A family of proteins is haematopoietic cell

specific transmembrane-4 (HtM4). Similar to Fc�RI� and CD20, HtM4 is expressed

primarily by haematopoietic cells of myeloid and lymphoid origin (Adra et al., 1994;

Hulett et al., 2001). However, it is expressed also in the developing mouse brain and in the

16

adult mouse nervous system (Kutok et al., 2005). HtM4 is thought to modulate G1-S cell

cycle transition in haematopoietic cells, as over-expression of the protein in synchronised

U937 cells can cause cell cycle arrest (Donato et al., 2002). Cell cycle progression is

regulated firstly by the sequential activation and inactivation of the cyclin-dependent

kinases (cdks) by phosphorylation and dephosphorylation, and secondly by activatory and

inhibitory proteins (such as p21 and p27) (Harper et al., 1993). Binding of inhibitory

proteins to various cyclin-cdk complexes is sufficient to arrest the cell cycle. Conversely,

inactive monomeric cdk can be activated via association with a specific cyclin, with

concurrent threonine phosphorylation (e. g. threonine 160 in cdk2). Full activation of cdk2

also requires dephosphorylation of threonine 14 and tyrosine 15. Threonine 160 of cdk2 is

phosphorylated by CAK (cdk-associated kinase), and dephosphorylation by KAP (cdk-

associated serine/threonine phosphatase) is critical for inactivation of cdk2 (Morgan,

1997). It has been shown that expression of KAP slows the G1 phase cell cycle progression

in transfected HeLa cells, and that aberrant KAP transcripts can be detected in some

hepatocellular carcinomas (Matsuda, 2008). HTm4 interacts directly with KAP via its C

terminus, and exogenous expression of HTm4 leads to dephosphorylation of cdk2 and

arrest of the cell cycle at the G0/G1 phase (Donato et al., 2002; Kutok et al., 2005).

Recently, it has been reported that HTm4 regulates cdk2 activity, and hence cell cycle

progression, in a dual fashion. Firstly, binding of HTm4 to the KAP-cdk2-cyclin A

complex prevents interaction of cyclin A with cdk2. Secondly, binding of the HTm4 C

terminus to KAP tyrosine 141 facilitates access of KAP to cdk2 threonine 160, which

accelerates dephosphorylation and inactivation of cdk2 (Chinami et al., 2005).

1.5.4 Novel MS4A members

In addition to the three founding members discussed above, the MSA4 family includes at

least 20 other predicted proteins in humans and rodents. Two papers have described

cloning of several cDNAs encoding predicted human and mouse MS4A proteins, all of

which share amino acid similarity, chromosomal localisation, intron / exon organisation

and amino acid motifs with the canonical MS4A proteins (Ishibashi et al., 2001; Liang and

Tedder, 2001). There is discrepancy between the two reports with respect to distribution of

the transcripts in tissues, which may be due to use of Northern blot versus PCR,

respectively (Ishibashi et al., 2001; Liang and Tedder, 2001). However, both groups found

17

that the newly identified MS4A gene transcripts were expressed in a wide variety of

tissues, with most expressed lymphoid tissues (Ishibashi et al., 2001; Liang and Tedder,

2001). Several of these predicted proteins have been investigated.

Expression of MS4a4B/Chandra was reported first in T helper type 1 cells, but not in T

helper type 2 cells (Venkataraman et al., 2000). Liang and Tedder reported that

MS4a4B/Chandra transcripts were expressed in thymus, spleen, peripheral lymph nodes,

bone marrow, liver, kidney, heart, colon and lung, but not in testis or brain (Liang and

Tedder, 2001). Recently, expression of MS4a4B/Chandra protein was explored in a range

of cells that have haematopoietic origin. The protein was found to be expressed in

immature thymocytes, mature single-positive thymocytes and in peripheral naïve T-cells,

but it was absent during intermediate stages of thymocyte development. MS4a4B/Chandra

was also expressed by natural killer (NK) cells but not by B-cells (Xu et al., 2006). At the

cellular level, the protein has a topology similar to that of CD20 and Fc�RI�. It was

expressed on the cell surface, with intracellular N- and C-termini, and when expressed in a

T-cell hybridoma line it was associated constitutively with lipid raft microdomains. In

normal T-cells, however, the protein only became enriched in rafts after T-cell activation

(Xu et al., 2006). Over-expression of MS4a4B/Chandra in primary CD4+ T-cell blasts

results in enhanced T-cell receptor (TCR)-induced production of Th1 cytokines (Xu et al.,

2006), which is similar to the role of Fc�RI� in Fc�RI signalling. As MS4a4B/Chandra is

highly regulated during T-cell development, it has been proposed that the protein may be

an important regulator of T-cell differentiation and/or function (Xu et al., 2006), which is

similar to the function proposed for LR8; regulation of DC maturation in rats.

Another MS4A member, MS4A8B/L985P, is expressed in a variety of lymphoid and

non-lymophoid tissues (Liang and Tedder, 2001), and is reported to be expressed at high

levels by tumour cells in small cell carcinoma of the lung (Bangur et al., 2004). In normal

human lung, expression of the protein is restricted to ciliated bronchiolar epithelium

(Bangur et al., 2004). Deans and colleagues reported that MS4A8B/L985P is expressed in

B-cells, where it associates with CD20 at the plasma membrane (Deans et al., 2008). Over-

expression of MS4A8B/L985P in a human B-cell line has been shown to result in an

enhanced calcium influx following B-cell receptor (BCR) stimulation. Interestingly,

stimulation via the BCR also leads to rapid down-regulation of MS4A8B/L985P (Deans et

18

al., 2008). These findings suggest that MS4A8B/L985P may have a similar function to

CD20, possibly accounting for the relatively normal immune phenotype observed in CD20

knockout mice (Deans et al., 2008).

Finally, TETM4 is reported to be a testis-specific member of the MS4A family (Hulett

et al., 2001). It shares with CD20 and Fc�RI� the features of small size, predicted

topology, amino acids motifs, and chromosomal localisation. Although, no function has yet

been attributed to the protein (Hulett et al., 2001).

1.6 Other four transmembrane domain proteins

In addition to the MS4A protein family, there are several other four TM domain-containing

protein families. None of these have significant amino acid sequence similarities with

HCA112 or the MS4A family, although they are instructive in developing hypotheses for

the function of HCA112. Several of these are discussed below.

1.6.1 Tetraspanin proteins

The tetraspanins are a large family of cell surface proteins that have been identified in

many species - from schistosomes to humans (Hemler, 2005). In humans, 33 members

have been identified, which include CD9, CD37, CD53, CD63, CD81/TAPA-1,

CD82/KAI1, CD151, and uroplakin. Tetraspanin proteins form several specific molecular

complexes that have been implicated in a variety of cellular functions, which include

oocyte fertilisation, susceptibility to infection by mammalian and plant parasites,

metastasis of cancer cells, and cell-cell interactions in the central nervous system and the

immune system (Hemler, 2001; Hemler, 2005; Levy and Shoham, 2005). Most of these

cellular functions involve protein-protein interactions, often with integrins, in a molecular

network in the plasma membrane known as the tetraspanin web or tetraspanin-enriched

microdomains (TEMs). TEMs appear to be important in organising functional multi-

molecular complexes (Hemler, 2005; Levy and Shoham, 2005; Le Naour et al., 2006; Min

et al., 2006). As in the case of MS4A proteins, tetraspanins are small, four TM domain-

containing proteins, with unevenly sized extracellular loops. Several additional features are

conserved within the tetraspanin family of proteins. These include conservation of three

polar residues in the TM domains (TM1, N; TM3, E; TM4, E), the presence of a CCG

19

motif and an additional 2-6 cysteine residues in the second extracellular domain, and five

intracellular cysteine residues, which are potentially palmitoylated (Levy and Shoham,

2005). In addition to palmitoylation, tetraspanin proteins are modified post-translationally

by the addition of several N-glycans (Andre et al., 2007). CD63 is chosen for further

discussion as a proteotypic tetraspanin, and because it has some interesting parallels with

HCA112, the protein investigated herein.

1.6.1.1 CD63

CD63 was the first tetraspanin to be characterised. The molecule is also known as

Lysosome Associated Membrane Protein 3 (LAMP-3), LIMP-3, platelet glycoprotein 40

(Pltgp40), and melanoma antigen 491 (ME491). A recent review has highlighted the

intracellular trafficking of this tetraspanin and suggested that the molecule plays a role in

the intracellular transport of other proteins (Pols and Klumperman, 2009). CD63 is

expressed on the cell-surface, but the majority is found within late endosomes, multi-

vesicular bodies, lysosomes and shed secretory vesicles known as exosomes (Mantegazza

et al., 2004; Pols and Klumperman, 2009). The cytoplasmic carboxy-terminal domain of

CD63 contains a consensus YXXØ motif (GYEVM) which is responsible for its

interaction with clathrin adapter protein (AP) complexes AP-2 and AP-3, and its

involvement in clathrin-dependent endocytosis (Rous et al., 2002; Peden et al., 2004; Pols

and Klumperman, 2009) (see below for discussion on amino acid motifs involved in

endocytosis). In addition to interacting with clathrin adaptor complexes, CD63 interacts

with various other proteins and modulates their expression, playing a role in antigen

presentation, HIV-infection, tumour cell motility and the process of metastasis (Pols and

Klumperman, 2009).

Immature and mature monocyte-derived dendritic cells (DCs) express CD63 on their

surfaces, in early endosomes, MHC-II-enriched compartments (MIICs) and lysosomes

(Mantegazza et al., 2004). Co-immunoprecipitation indicates that CD63 interacts with

dectin-1, a �-1,3 glycan receptor that is involved in the phagocytosis of yeast by immature

DCs. After surface labelling of CD63, the internalised antibody-bound molecule is found

surrounding phagocytosed yeast cells. However, CD63 does not appear to be involved in

the endocytosis of FITC-dextran or latex beads by these cells (Mantegazza et al., 2004).

20

The data suggests that this tetraspanin protein has a role in capture, processing and/or

presentation of antigen, possibly in the context of MHC-II molecules, by immature DCs

(Mantegazza et al., 2004).

Another interesting interaction of CD63 is with the heteromeric ion pump H, K-

ATPase, which is responsible for the secretion of acid by gastric parietal cells. CD63

interacts with the �-subunit of H, K-ATPase in transfected COS-7 cells, where it is found

to induce redistribution of this subunit from the cell surface to CD63-containing

intracellular compartments (Duffield et al., 2003). The redistribution of the �-subunit of H,

K-ATPase is dependent on the ability of CD63 to interact with the clathrin adapter

complexes AP-2 and AP-3, thus facilitating clathrin-mediated endocytosis. This finding

suggests that CD63 acts as an adapter between this particular partner and the endocytic

machinery of the cell (Duffield et al., 2003).

That CD63 may have a more general role in directing the intracellular trafficking of

partner proteins is suggested by a recent study on the chemokine receptor CXCR4. In

addition to its role as a receptor for certain chemokines, CXCR4 facilitates entry of HIV-1

into T-lymphocytes. A screen to identify novel HIV-1 entry blockers, has identified an N-

terminal deletion mutant of CD63 that can suppress cell surface expression of CXCR4

(Yoshida et al., 2008). Expression of this mutant tetraspanin, which lacks the first 2 TM

and cytoplasmic domains but retains the C-terminal GYEVM motif, results in mis-

targeting of CXCR4 to late endosomes and lysosomes. Depletion of endogenous CD63 by

RNAi, however, enhanced cell surface expression of CXCR4. Together, these results

suggest that a normal function of CD63 is to reduce the level of CXCR4 at the cell-surface,

and that this capacity is enhanced by deletion of the N-terminal portion of the protein

(Yoshida et al., 2008).

In several types of cancer (including, melanoma, lung adenocarcinoma, ovarian, breast

and colon cancers), a correlation has been observed between reduced expression of CD63

and increased malignancy, metastasis, invasiveness and/or tumour growth (Sordat et al.,

2002; Jang and Lee, 2003; Sauer et al., 2003; Kwon et al., 2007; Zhijun et al., 2007). The

role CD63 plays in cancer progression is yet to be identified. One possibility is through its

effects on levels of expression of important chemokine receptors (see above). Another

mechanism may involve association of CD63 with integrins, which are important

21

mediators of cell interactions with the extracellular matrix. CD63 could influence the

activity and stability of integrins via interaction with other tetraspanins within TEMs.

Alternatively, CD63 may be involved in the endocytosis of integrins, and thus modulate

their cell surface expression (Pols and Klumperman, 2009).

1.6.2 Other 4 transmembrane-domain containing proteins

Many ligand- and voltage-gated ion channels are multimeric complexes that consist of 4

TM protein subunits. An example is the AMPA-type glutamate receptors (AMPAR),

which consist of 4 closely related subunits, each containing four TM-like domains.

However, the second TM does not traverse the membrane, instead jutting in and out. The

result is that the N- and C-termini of these molecules are on opposite sides of the plasma

membrane. The extracellular domains form the ligand-binding site, while the TM regions

form the channel (Shepherd and Huganir, 2007).

Another family of 4 TM proteins includes the tight-junction localised protein claudin.

The exact functions of proteins in this family are unknown, but they are thought to be

structural components of tight-junction strands and they may be involved in carcinogenesis

(Morita et al., 1999; Kondo et al., 2008).

1.7 HCA112 in renal proximal tubules

As mentioned previously, Nakajima et al. reported that HCA112 expression was increased

in the mouse kidney proximal tubule in response to experimental proteinuria (Nakajima et

al., 2002). Mice given repeated intraperitoneal injections of bovine serum albumin (BSA)

developed some of the characteristic renal changes associated with pathological

proteinuria, including, increased glomerular permeability, tubular changes and interstitial

infiltration of macrophages and T lymphocytes. In proteinuria induced by protein overload,

proximal tubule cells exhibit increased transcription of genes encoding enzymes that are

involved in the degradation of the re-absorbed protein. There is also an increase in

expression of genes that encode vasoactive and inflammatory cytokines (Nakajima et al.,

2002). These findings raise the possibility that HCA112 could be involved in the

inflammatory process and/or processes related to the re-absorption of albumin from the

glomerular filtrate.

22

1.7.1 Re-absorption of protein from the glomerular filtrate

Each day up to 8100 mg of albumin is filtered from the plasma by the glomeruli. However,

only about 30 mg is excreted daily in the urine. Thus, more than 99% of the filtered

albumin is re-absorbed along the nephron (reviewed by (Gekle, 1998)). Investigations have

shown that protein re-absorption takes place by endocytosis, exclusively by cells lining the

proximal tubule (Straus, 1964; Ericsson, 1965; Graham and Karnovsky, 1966; Carone et

al., 1979; Tojo and Endou, 1992). Albumin binds to the high molecular weight receptors

megalin and cubulin, and possibly others, expressed on the apical membrane of proximal

tubular epithelial cells (Zhai et al., 2000; Verroust and Kozyraki, 2001). The albumin is

then internalised by clathrin-dependent receptor-mediated endocytosis and it is directed to

the endosomal and lysosomal compartments where it is degraded to amino acids. The

membrane-bound receptors, on the other hand, are recycled back to the cell surface

through recycling endosomes (Gekle, 1998).

1.7.2 Interstitial inflammation in proteinuria

Rodents injected with large daily doses of albumin develop ‘overload proteinuria’, and this

is accompanied by interstitial inflammation and fibrosis. These pathological changes

appear to result from the effects of excessive uptake of proteins by the epithelial cells of

the proximal tubules. Droplets containing re-absorbed proteins accumulate in the

cytoplasm of the proximal tubule cells and this is associated with increased transcription of

genes encoding vasoactive, inflammatory, and fibrogenic molecules (Zoja et al., 2003).

Several molecular mechanisms have been identified that lead to these changes. High

concentrations of proteins (either de-lipidated or lipid-enriched albumin, IgG and

transferrin) can induce increased expression of endothelin-1 by proximal tubular cells in

vitro (Zoja et al., 1995). This peptide is involved in kidney injury through its ability to

stimulate renal cell proliferation and production of extracellular matrix, and its action as a

chemoattractant for monocytes. Expression of nuclear factor kappa B (NF-�B) is also up-

regulated in a dose-dependent manner in response to albumin and IgG (Zoja et al., 1998).

This transcription factor up-regulates the transcription of various chemokines, including

monocyte chemoattractant protein-1 (MCP-1), RANTES, IL-8 (Hoffmann et al., 2002),

and membrane-bound and soluble fractalkine (CX3CL1) (Bhavsar et al., 2008). These

23

cytokines are responsible for the recruitment of monocytes, macrophages and T

lymphocytes to the kidney interstitium, where they are involved in inflammation and

interstitial fibrosis (reviewed by (Zoja et al., 2003; Eddy, 2004)).

Complement is activated in the proximal tubule under proteinuric conditions (Morita et

al., 1997). Deposits of complement components C3 and the C5b-9 membrane attack

complex (MAC) are found on the surface of proximal tubule cells, and C3 is also

detectable within subapical and lysosomal compartments of the cells in kidneys from

animals with overload proteinuria (Biancone et al., 1994; Abbate et al., 1999).

Complement deposition leads to cytoskeletal alterations in proximal tubular epithelial

cells, generation of reactive oxygen species, and synthesis of pro-inflammatory cytokines

such as IL-6 and TNF-� (David et al., 1997). Proximal tubular epithelial cells are able to

synthesise complement components, including C3, in vitro. Moreover, expression of C3

mRNA and secretion of C3 is enhanced following exposure of proximal tubular epithelial

cells to total serum proteins or transferrin in vitro (Zoja et al., 2003).

Interstitial fibrosis is another result of overload proteinuria. This involves inflammatory

cells such as macrophages and lymphocytes, as well as accumulation of myofibroblasts. At

least some of these changes appear to be a consequence of fibrogenic factors produced by

the proximal tubular epithelial cells. Furthermore, inflammatory macrophages and

lymphocytes secrete vasoactive products (e.g. endothelin-1 and angiotensin II), products

which impair extracellular matrix degradation, as well as a range of cytokines that includes

transforming growth factor � (TGF-�) and platelet-derived growth factor (PDGF), which

can stimulate the transformation of interstitial cells into myofibroblasts. Interestingly, re-

absorption of excessive amounts of protein by proximal tubule cells stimulates

transcription of the gene encoding TGF-� (Zoja et al., 2003).

1.8 Endocytosis

As HCA112 is up-regulated in kidney proximal tubules cells during periods of excessive

albumin and/or protein re-absorption (Nakajima et al., 2002) and this involves endocytosis

of intact protein molecules (Straus, 1964), it is appropriate to review the process of

endocytosis briefly.

24

Endocytosis is a process by which cells internalise ligand-bound small molecules,

macromolecules, and particles, and target them to specific cytoplasmic organelles.

Mammalian cells use this process for many different cellular functions, including nutrient

uptake, migration, receptor signalling and down-regulation, cholesterol homeostasis,

maintenance of cell polarity, neurotransmission and phagocytosis. Conversely, endocytosis

is exploited to gain entry into the cell by viruses, bacteria and toxins.

Endocytosis is a blanket term that encompasses several mechanisms employed by cells

to take up external materials. These can be classified broadly into phagocytosis (‘cell

eating’) and pinocytosis (‘cell drinking’), the latter including clathrin-dependent receptor-

mediated endocytosis and clathrin-independent endocytosis. Generally, all of these

processes result in delivery of the contents of endocytic vesicles to lysosomes for

degradation. Membrane components, on the other hand, undergo complex sorting events

for delivery to their target destinations, often recycling back to the plasma membrane.

Correct trafficking of proteins is a complex process that depends on properties of the

internalised molecules, the properties of receptors in receptor-mediated endocytosis (e.g.

internalisation motifs), and characteristics of the organelles that carry them (e.g. size,

shape, luminal pH and lipid composition).

1.8.1 Phagocytosis

Phagocytosis refers to the internalisation of large particles (>0.5 µm diameter), and it can

occur in many cells types, but most importantly in specialised cells of the immune system,

such as macrophages, monocytes and neutrophils. In the case of opsonised particle (bound

with IgG or complement), phagocytosis involves binding to specific receptors, such as IgG

Fc receptors, or complement receptors, some of which are integrins. In other cases, binding

involves components of the particle, such as mannose residues to specific mannose lectin

receptors (Mellman et al., 1983; Wright and Silverstein, 1983; Ezekowitz et al., 1991;

Isberg and Tran Van Nhieu, 1994). An example of the phagocytic process is IgG-

opsonised particles bound to IgG Fc receptor (Fc�R IIIA, CD16). After binding, a

phagocytic signal is transduced, which involves recruitment of SRC-family tyrosine

kinases to the ITAM present in the cytoplasmic domain of � subunit of the receptor (Cox et

al., 1996; Greenberg et al., 1996). This causes polymerisation of actin locally at the site of

25

contact with the particle, pseudopod extension and engulfment of the particle into a

membrane-bound phagosome (Greenberg et al., 1990). The phagosome then fuses with

vesicles in the endocytic pathway in a complex and still poorly understood process. The

ingested particle is then digested in late endosomes and phago-lysosomes by hydrolytic

enzymes, and some intrinsic membrane proteins are recycled back to the plasma

membrane. In the endocytic pathway of specialised antigen presenting cells, some peptides

derived from the ingested particle are loaded onto MHC-II molecules for presentation to T-

cells at the cell surface (Mukherjee et al., 1997).

1.8.2 Pinocytosis / endocytosis

Pinocytosis, often referred to as ‘endocytosis’, involves the uptake of extracellular fluid

and macromolecules that may be either in the fluid phase, or bound specifically or non-

specifically to the plasma membrane. All cells are capable of endocytosis by a process that

involves formation of small (>0.2 µ diameter) vesicles at the plasma membrane. In most

mammalian cells under normal conditions, uptake of receptor-bound ligands and

extracellular fluid involves the formation of clathrin-coated vesicles. However, caveolae

and actin-based mechanisms can also be involved in clathrin-independent endocytosis.

1.8.2.1 Clathrin-mediated endocytosis

Clathrin-dependent endocytosis occurs at specialised plasma membrane domains known as

coated-pits, which contain a non-random selection of surface membrane proteins

(including the transferrin and low-density lipoprotein (LDL) receptors). Mammalian cells

use the process to internalise extracellular fluid (fluid-phase endocytosis) and ligands

bound to receptors. A similar process is used also to capture and transport proteins from

the trans-Golgi network (TGN) to the cell surface (Mukherjee et al., 1997). The major

component of coated vesicles is clathrin (Keen et al., 1979), which was first identified by

Pearse in 1975 (Pearse, 1975). The protein functions as a trimer, known as a triskeleton,

which is composed of three heavy chains (180 kDa) and three light chains (30-35 kDa) that

are joined at a central vertex (Fotin et al., 2004). The triskeleton polymerises to form a

polyhedral lattice, or cage, which serves as a mechanical scaffold for the vesicle. Clathrin

is unable to bind directly to the membrane. The connection between the clathrin scaffold

26

and the membrane is mediated by adaptor proteins that bind directly to clathrin and to

protein components of the membrane.

In addition to mediating binding of clathrin to the membrane, adaptor protein

complexes (APs) and accessory proteins are involved in selection/recruitment of TM

proteins (cargo) into vesicles (discussed below), recruitment of cytoplasmic proteins to the

membrane and the process of pit invagination. The specificity of cargo selection and

subcellular destination is mediated by the location of the coat-associated proteins within

the endocytic system. Plasma membrane clathrin coats contain AP-2 and various accessory

factors. TGN and endosomal coats contain AP-1 and/or the monomeric adaptors Golgi-

localised, �-ear-containing, ADP-ribosylation factor-binding proteins (GGAs) GGA1,

GGA2, and GGA3. AP-3 is also found on clathrin coats associated with endosomes

(reviewed by (Bonifacino and Traub, 2003)).

The process of clathrin-mediated endocytosis occurs as follows. Polymerisation of

clathrin results in the formation of a clathrin coated-vesicle that is attached to the

membrane by a narrow neck. The vesicle pinches off to form a cytoplasmic vesicle, in a

process that involves the GTPase dynamin. The clathrin-coated vesicle must then be un-

coated in order for it to fuse with other vesicles and thus deliver its cargo. This process

involves the constitutively expressed heat shock proteins auxilin and heat shock 70 kDa

protein 8 (HSPA8), which bind to clathrin and disassemble the clathrin lattice in an ATP-

dependent manner (Brodsky et al., 2001; Owen et al., 2004; Wilbur et al., 2005; Young,

2007). Once un-coated, vesicles can then fuse with early or sorting endosomes, where

sorting takes place. The acidic pH of the early endosome causes dissociation of some

receptor-ligand complexes. Commonly, the ligand is transported to late endosomes and

lysosomes for degradation, and the receptor is returned to the plasma membrane, via

recycling endosomes, where it can undergo additional rounds of endocytosis, as

exemplified by the LDL receptor (reviewed by (Mukherjee et al., 1997)).

Certain receptor-ligand pairs take other routes. In the case of transferrin and its

receptor, it is the bound iron that is released in endosomes rather than transferrin itself,

which remains bound to the receptor throughout the endocytic pathway. Transferrin-bound

receptor is then recycled back to the plasma membrane via recycling endosomes, following

the same route as the LDL receptor. Binding of some ligands, such as hormones and

27

growth factors to their receptors, causes down-regulation of the receptors by hastening

their internalisation and delivery to lysosomes (Mukherjee et al., 1997). Figure 1-2

illustrates the pathways involved in clathrin-dependent endocytosis and trafficking from

the TGN in a non-polarised cell.

1.8.2.2 Endocytic sorting

Endocytic sorting is a complex system. It involves sorting signals located in the

cytoplasmic domains of the TM proteins (cargo) that are targeted to endosomes, lysosomes

and related organelles, and the molecular machinery that recognises these signals. Several

sorting-signal motifs have been identified that are recognised with fine specificity by

specific clathrin adaptor protein complexes and clathrin-associated proteins. These motifs

fall into two major classes: tyrosine-based (NPXY and YXXØ, where Ø represents a bulky

hydrophobic residue) and di-leucine-based ([DE]XXXL[LI] and DXXLL). The adaptor

protein complex AP-2 (present at the plasma membrane) binds to the majority of sorting-

signals, whereas others (including AP-1, AP-3, AP-4, disabled 2 (Dab2) and GGAs) bind

only one or two classes of motifs (summarised in Table 1-1). AP-2 is a heterotetramer,

consisting of 4 subunits - � and �2 adaptins, µ2 and �2, which binds both the terminal

globular domain of the clathrin heavy chain and the sorting signal motif in the cargo

protein (Bonifacino and Traub, 2003).

Two tyrosine-based motifs have been identified - NPXY and YXXØ. The NPXY motif

is recognised by AP-2 and the accessory protein Dab2, and is present in several type I

integral membrane proteins, including the LDL receptor, megalin, the insulin receptor, and

integrin � and �-amyloid precursor protein families (Bonifacino and Traub, 2003). The

NPXY motif has been shown to mediate only rapid internalisation of motif-containing

proteins and not other intracellular sorting events (Bonifacino and Traub, 2003). NPXY is

the minimal motif shared by these proteins and, while it is essential for rapid endocytosis

(Chen et al., 1990), it may not always be sufficient. In the LDL receptor, for example, a

phenylalanine residue at two positions amino-terminal to the asparagine residue is also

required for rapid internalisation. Furthermore, while chimeric transferrin receptors

expressing only NPVY fail to undergo rapid endocytosis, insertion of FDNPVY rectifies

the normal rate of internalisation (Collawn et al., 1991).

28

YXXØ sorting signals are found in a wide variety of TM proteins, such as the

transferrin receptor, asialoglycoprotein receptor, CI-MPR, CD-MPR, lysosomal membrane

proteins LAMP-1 and LAMP-2, and TGN proteins such as TGN38 and furin. The YXXØ

motif is recognised by AP-1 and AP-2, and is the minimal motif essential for the rapid

internalisation and sorting of these proteins. Substitution of the tyrosine residue results in

aborted function, however other factors also contribute to the strength and specificity of the

signal. These include the identity of the X residues and the bulky hydrophobic residue (Ø),

the nature of other residues flanking the motif, as well as the position of the motif within

the cytoplasmic domain of the protein (Bonifacino and Traub, 2003). A glycine residue

preceding the critical tyrosine residue is involved in lysosomal targeting, but not

endocytosis (Harter and Mellman, 1992). Similarly, lysosomal-sorting signals often have

acidic residues at the X positions (Rous et al., 2002). The distance of the YXXØ motif

from the TM domain and the carboxy terminus is important for sorting of TM proteins to

lysosomes, TGN, endosomes, late endosomes or recycling endosomes (Bonifacino and

Traub, 2003).

The di-leucine based motif [DE]XXXL[LI] is present in single- and multispanning-

membrane proteins, and it is recognised by AP-1, AP-2 and AP-3 adaptors (Bonifacino and

Traub, 2003; Sorkin, 2004). The [DE]XXXL[LI] sorting signal is essential for rapid

internalisation and sorting of the TM proteins mentioned below, which have a variety of

subcellular destinations. Substitution of the leucine/isoleucine residues for alanine has been

shown to abrogate all functions of the motif (Letourneur and Klausner, 1992). This sorting

signal is involved in the serine phosphorylation-dependent down-regulation of CD4 and

the CD3-� chain by T-cells, where phosphorylation induces rapid internalisation and

lysosomal targeting of these molecules. Other proteins contain constitutively active forms

of this motif and they are targeted mainly to endosomes and lysosomes (e.g., NPCI and

LIMP-II), as well as specialised endosomal-lysosomal compartments such as endocytic

antigen-processing compartments, synaptic dense-core granules, and melanosomes

(Bonifacino and Traub, 2003). The differences in endosomal sorting involving this motif

are influenced by the position of the motif in the protein relative to the TM domains and

the carboxy and amino termini (Bonifacino and Traub, 2003).

29

DXXLL sorting signals are present in several plasma membrane receptors and other

proteins (e.g., CI-MPR and CD-MPR, sortilin, the LDL-receptor related proteins LRP3 and

LRP10, and �-secretase), where they are involved in the cycling of the respective proteins

between the TGN and endosomes. Unlike the other tyrosine- and di-leucine based sorting

motifs, DXXLL does not bind to AP complexes, but instead is bound by GGAs. In all

examples described thus far, the DXXLL motif is located one to two residues from the

carboxy-terminus of the cargo protein. This location seems to be critical for binding to

GGAs, as addition of a few amino acids to the carboxyl end of the cargo protein blocks

interaction with the adaptor proteins (Doray et al., 2002; Doray et al., 2008).

Ubiquitination of proteins also functions as an intracellular sorting signal. Ubiquitin

(Ub) is a 76 amino acid polypeptide that can be attached covalently by Ub-protein ligases

(e.g., Nedd4) to proteins via an isopeptide bond between the C-terminal glycine of Ub and

a lysine residue within the substrate protein. The addition of a single Ub to a substrate

protein is called mono-ubiquitination. The addition of one Ub to each of several lysine

residues in a substrate is called multiple mono-ubiquitination. Poly-ubiquitination occurs

when a single lysine residue in a substrate is bound by Ub, which itself is ubiquitinated on

one of its seven lysine residues (and so on), thus forming a poly-Ub chain. It is well known

that proteins that are poly-ubiquitinated are targeted to the 26S proteasome for degradation.

Mono-ubiquitination and multiple mono-ubiquitination, on the other hand, have been

implicated recently in endocytosis of plasma membrane proteins, while mono-

ubiquitination may be involved also in the sorting of proteins in multivescicular bodies

(Haglund et al., 2003). The role of (multiple) mono-ubiquitination in endocytosis is still

unclear, but it has been shown that the down-regulation of several receptor tyrosine kinases

(including platelet derived growth factor (PDGF) and epidermal growth factor receptors)

involves mono-ubiquitination, followed by endocytosis and traffic to lysosomes for

degradation. The process involves Ub-interacting motif (UIM)-containing Ub binding

proteins (Ub-receptors, such as Eps15, epsins) that are localised throughout the endocytic

pathway and recognise Ub-bound cargo proteins and also bind to AP-2, thus interacting

with the clathrin-dependent pathway (Haglund et al., 2003). Ubiquitination is a dynamic

process and de-ubiquitination of cargo proteins by de-ubiquitinating enzymes results in

recycling of the TM cargo proteins back to the plasma membrane (Haglund et al., 2003).

30

1.8.2.3 Caveolin/lipid raft-dependent endocytosis

Clathrin-independent endocytosis includes endocytosis mediated by glycolipid rafts and by

caveolae, in addition to the constitutive pinocytic pathway. Glycolipid rafts are found in

detergent-insoluble, low-density membrane fractions that are rich in cholesterol,

sphingolipids and glycosyl-phosphatidylinositol (GPI)-anchored proteins. Biochemically,

lipid rafts are defined by their resistance to extraction in non-ionic detergents (e.g., Triton

X-100) at low temperature. Caveolae are cholesterol- and sphingolipid-rich flask-shaped

50-80 nm diameter invaginations of the plasma membrane. They partition with detergent-

resistant lipid raft fractions when solubilised membranes are separated based on their

density in sucrose gradients, and they are identified by their association with caveolin-1

(Nabi and Le, 2003). Caveolae are, therefore, a subset of biochemically defined lipid rafts.

Caveolae and raft domains are important sites of cholesterol uptake and efflux from

cells (Fielding and Fielding, 1997; Fielding and Fielding, 2001). They have been

implicated also in the internalisation of sphingolipids and sphingolipid-binding toxins

(cholera toxin and shiga toxin), GPI-anchored proteins, extracellular ligands (folic acid,

albumin, autocrine motility factor (AMF) and growth hormone), certain receptors (e.g., IL-

2 receptor), several non-enveloped viruses (Simian virus 40, Polyoma virus), and bacteria

(Nabi and Le, 2003). Their role in endocytosis, however, remains controversial. There is

evidence that caveolae and raft domains both mediate clathrin-independent, dynamin-

dependent and cholesterol-sensitive internalisation, which are thought to represent

essentially equivalent routes, that involve membrane invaginations and intracellular

vesicles termed ‘caveolar invaginations’ and ‘caveolar vesicles’, respectively (Nabi and

Le, 2003).

Ligand sorting into caveolae/raft domains occurs at the plasma membrane, leading to

cargo internalisation into distinct intracellular vesicles for delivery to independent

destinations such as the ER or Golgi. Caveolar invaginations form at the plasma membrane

and bud off in a dynamin-dependent manner, thus forming caveolar vesicles or

caveosomes. These vesicular structures are pH neutral, and they are negative for the

protein markers of endosomes, lysosomes, the TGN, Golgi apparatus and the ER

(Pelkmans and Helenius, 2002). Caveolar invagination is facilitated when the actin

cytoskeleton is disrupted as a result of tyrosine kinase signalling, and is inhibited by kinase

31

inhibitors (such as staurosporine and genistein) and enhanced by phosphatase inhibitors

(such as okadaic and vanadate) (Nabi and Le, 2003; Lajoie and Nabi, 2007). Contrary to

previous belief, caveolin-1 is now thought to be a negative regulator of caveolar/raft-

dependent endocytosis because caveolin-1-GFP is highly immobile at the plasma

membrane. It appears that the protein does not induce raft invagination, but rather

stabilises the association of invaginated rafts with the plasma membrane, thus retarding

their dynamin-dependent budding and detachment (Nabi and Le, 2003; Hommelgaard et

al., 2005).

The best characterised instance of entry via caveolae/raft-dependent mechanisms is that

for Simian virus 40 (SV40) (Pelkmans and Helenius, 2002). After the virus particles bind

to MHC-I molecules on the cell-surface, they diffuse laterally along the plasma membrane

until becoming trapped in caveolae, which are linked to the actin cytoskeleton. Within

caveolae, the virus particles activate a local tyrosine kinase-based signalling cascade,

which results in de-polymerisation of the local actin cytoskeleton and recruitment of

dynamin II to the site of internalisation. Dynamin II facilitates budding of the vesicle, and

thus the virus plus dynamin II and caveolin-1 (but not MHC-I molecules) is endocytosed

into intracellular vesicles (Pelkmans and Helenius, 2002).

1.9 Development of hypotheses about possible functions of HCA112

The overview of the literature regarding HCA112 and closely related molecules presented

above raises a number of hypotheses about possible functions of the molecule. These are

detailed below.

1.9.1 HCA112 may be involved in intracellular protein trafficking

Firstly, studies by Nakajima and co-workers have shown that HCA112 expression is

increased in mouse kidney proximal tubule cells during experimental overload proteinuria

(Nakajima et al., 2002). This condition is associated with protein re-absorption by the

epithelium of the proximal tubules, via the process of endocytosis. A possible consequence

of excessive protein re-absorption is expression of a number of other genes by the

epithelial cells, including some encoding pro-inflammatory cytokines and fibrogenic

32

growth factors. The increased transcription of HCA112 that accompanies proteinuria

suggests that HCA112 might be involved in endocytosis and degradation of re-absorbed

proteins by the tubular epithelial cells. Alternatively, it could be involved in the

inflammatory process in the renal parenchyma, or its expression could be regulated by one

or more of the pro-inflammatory cytokines that are produced by the epithelial and/or

inflammatory cells.

Findings reported by Louvet et al. indicate that transcription of LR8 is down-regulated

in immature DCs and macrophages following activation (Louvet et al., 2005). This

observation suggests that expression of LR8 might be linked to the functional state of DCs,

and perhaps with the phagocytic and macro-endocytic functions associated with immature

DCs. This hypothesis is strengthened by the finding that the topologically similar

tetraspanin protein CD63 is involved in phagocytosis of yeast by immature DCs

(Mantegazza et al., 2004). In addition, CD63 is reported to interact with the clathrin

adapter complexes AP-2 and AP-3, and to be involved in the intracellular trafficking of a

number of proteins, including integrins, H, K-ATPase and CXCR4 (Duffield et al., 2003;

Yoshida et al., 2008; Pols and Klumperman, 2009). The MS4A protein Fc�RI� also plays a

role in the intracellular trafficking of other Fc�RI subunit proteins, and its expression

enhances both Fc�RI signalling and cell surface expression (Donnadieu et al., 2000b).

Together, these data lead to the hypothesis that HCA112 might be involved in intracellular

protein trafficking.

1.9.2 HCA112 might be involved in ion channel or calcium signalling activity

As the closely related MS4A proteins CD20 and MS4A8B/L985P have been shown to

regulate plasma membrane calcium conductance (Bubien et al., 1993; Deans et al., 2008)

and store operated calcium entry in B-cells (Li et al., 2003), it is reasonable to hypothesise

that the function of HCA112 might also be related to the cellular handling of calcium or

other ions. Furthermore, the related molecule Fc�RI� is involved in intracellular calcium

signalling. The ITAM present in the intracellular tail of Fc�RI� binds SRC family tyrosine

kinases, leading to phospholipase C�1 signalling, release of calcium from intracellular

stores, and activation of isoforms of PKC (Lin et al., 1996; Dombrowicz et al., 1998).

33

Thus HCA112, like its MS4A family member, might be involved in intracellular calcium

signalling.

1.9.3 HCA112 might function in specialised plasma membrane microdomains

Both CD20 and the tetraspanin proteins function in specialised plasma membrane

microdomains. Li and colleagues have found that the involvement of CD20 in store

operated calcium entry in B-cells requires its localisation in cholesterol- and sphingolipid-

rich membrane microdomains (Li et al., 2003). Furthermore, the tetraspanin proteins have

been shown to interact with other tetraspanins and associated proteins, facilitating their

positioning into plasma membrane microdomains known as tetraspanin-enriched

microdomains (TEMs) or the tetraspanin web. These specialised microdomains play an

important role in many cellular processes, functioning as organisers of multi-molecular

complexes that are similar conceptually to lipid rafts, but with a different lipid and protein

composition (Hemler, 2005; Levy and Shoham, 2005; Le Naour et al., 2006; Min et al.,

2006). As both CD20 and the tetraspanin proteins have structural similarities with

HCA112, it is reasonable to hypothesise that HCA112 might have functions that include

organisation within lipid rafts or facilitation of function within a multi-molecular

complex(es).

1.10 Objectives of the study

The primary focus of this study was to examine the expression, subcellular localisation and

function of HCA112. HCA112 is a newly discovered protein and, as such, very little is

known about it. First, this study details bioinformatics analysis of HCA112 and the

expressed gene product. Epitope-tagged HCA112 constructs were generated for use in

investigating the subcellular localisation of the molecule in transfected cells. Antibodies

against HCA112 have not been available previously, and investigation of the tissue

distribution of the molecule has, therefore, been very difficult. The next step in this project

was to produce an anti-HCA112 polyclonal antibody and to use it to perform an extensive

tissue survey.

Due to the similarity of HCA112 with CD20 and Fc�RI�, which modulate intracellular

calcium levels, patch clamping and fluorescence microscopy employing the calcium

34

binding fluorophore fura-2/AM were used to investigate a role of HCA112 in regulating

plasma membrane ion conductance, store-operated calcium entry and intracellular calcium

levels. The results of these experiments suggest that HCA112 does not affect these

processes. Proteomics analysis of co-immunoprecipitates was then used to identify

potentially interacting proteins, with the expectation that the results might implicate

HCA112 as part of a known molecular complex or process. Several proteins were

identified, and immunoprecipitation/Western blot was used to confirm the physical

interaction between HCA112 and some of these proteins, and immunofluorescence

confocal microscopy was used to examine their subcellular localisation.

35

Figure 1-1. Location of the enzymes involved in testosterone synthesis in Leydig cells.

Adopted from (Diemer et al., 2003).

36

Figure 1-2. The mechanism of clathrin-dependent endocytosis and trafficking.

Clathrin-coated pits formed at the plasma membrane invaginate and pinch off to form endocytic vesicles. These vesicles un-coat in the cytoplasm and fuse with early endososmes, where receptors and their ligands are sorted to various intracellular destinations. Internalised membrane components can be recycled back to the plasma membrane from early endosomes or the late recycling compartment. Some membrane proteins, on the other hand, are associated with membranes that invaginate into the endosome to form multivesicular bodies (MVBs), which are ultimately delivered to lysosomes for degradation. Lysosomal enzymes and other molecules are delivered from the Golgi apparatus to endosomes and the plasma membrane by means of clathrin-mediated or other types of membrane transport. The intravesicular pH of early endosomes is between 6.0–6.5, while that of late endosomes and lysosomes is approximately 4.5–5.5. Adopted from (Sorkin and von Zastrow, 2002)

37

Table 1-1. Endosomal/lysosomal sorting signals.

Amino acid residues are designated according to the single letter code as follows: A, alanine; C, cysteine; D, aspartic acid; E, glutamic acid; F, phenylalanine; G, glycine; H, histidine; I, isoleucine; K, lysine; L, leucine; M, methionine; N, asparagine; P, proline; Q, glutamine; R, arginine; S, serine; T, threonine; V, valine; W, tryptophan, and Y, tyrosine. X stands for any amino acid and Ø stands for an amino acid residue with a bulky hydrophobic side chain. Adopted from (Bonifacino and Traub, 2003)

Abbreviations: PTB, phosphotyrosine-binding; Dab2, disabled-2; AP, adaptor protein; VHS, domain present in Vps27p, Hrs and Stam; GGAs, Golgi-localised, �-ear-containing, ARF-binding proteins; PACS-1, phosphofurin acidic cluster sorting protein 1; TIP47, tail-interacting protein of 47 kDa; SHD1, Sla1p homology domain 1; UBA, ubiquitin associated; UBC, ubiquitin conjugating; UIM, ubiquitin interaction motif.

38

Chapter 2: Materials and Methods

2.1 Solutions

FACS wash buffer PBS, 1% foetal calf serum (FCS) (v/v), 10 mM NaN3

FACS fixative solution PBS, 0.1% formalin (v/v), 111 mM D-glucose, 10 mM NaN3

GTS (SDS-PAGE running buffer) 192 mM glycine, 25 mM Tris, 0.1% SDS (w/v) (pH 8.3)

PBS (phosphate-buffered saline) 1.06 mM KH2PO4, 155.17 mM NaCl, 2.97mM Na2HPO4-7H2O (pH 7.4)

SDS-PAGE sample buffer (5x) (or ‘SDS reducing buffer’) 62.5 mM Tris-HCl (pH 6.8), 10% glycerol (v/v), 2% SDS (w/v), 0.05% �-mercaptoethanol

(v/v), 0.006% bromophenol blue (w/v)

TAE (tris-acetate-EDTA) 40 mM Tris-HCl, 20 mM acetic acid, 1 mM EDTA

TBS (tris-buffered saline) 100 mM Tris-HCl, 150 mM NaCl (pH 7.4)

TBST (tris-buffered saline containing 0.1% Tween-20) 100 mM Tris-HCl, 150 mM NaCl, 0.1% Tween-20 (v/v) (pH 7.4)

39

TNE buffer 25 mM Tris-Cl (pH 7.5), 150 mM NaCl, 5 mM EDTA

Trypsin-EDTA 0.25% Trypsin (w/v), 0.53 mM EDTA-4Na

Western transfer buffer 25 mM Tris, 192 mM glycine, 20% methanol (v/v), pH 8.3

Western blot stripping buffer 62.5 mM Tris-HCl (pH 6.8), 2% SDS, 0.7% �-mercaptoethanol (v/v)

SDS-PAGE components Resolving gel Stacking gel

(0.375 M Tris, pH 8.8) (0.125 M Tris, pH 6.8)

12% 4%Acrylamide/Bis (30% stock)* 4.0 ml 1.3 ml

Distilled water 3.35 ml 6.1 ml

1.5 M Tris-HCl, (pH 8.8) 2.5 ml -

0.5 M Tris-HCl, (pH 6.8) - 2.5 ml

10% SDS (w/v) 100 µl 100 µl

10% ammonium persulfate (APS) (w/v)$ 50 µl 50 µl

TEMED+ 5 µl 10 µl

Total 10 ml 10 ml

* Acrylamide/bis solution (37.5:1) was purchased from Bio-Rad.

$ APS was purchased from Amersham Biosciences. Immediately after dissolving in

distilled water, aliquots were frozen and stored at -20oC until required.

+ TEMED was purchased from Sigma.

40

2.2 Animals

2.2.1 Mice

C57BL/6 mice were purchased from Laboratory Animal Services and held at the Medical

School Animal House, University of Adelaide. They were provided with unlimited access

to water and rodent chow.

2.2.2 Rats

Six week old Albino Wistar rats were purchased from Laboratory Animal Services (colony

derived from rats at the Animal Resource Centre, Perth). The rats were housed at the

Medical School Animal House, University of Adelaide. They were provided with

unlimited access to water and rodent chow.

2.3 Cell culture techniques

Table 2-1. Cell Lines used in this study

Name Origin Reference

COS-7 African green monkey kidney epithelial ATCC# CRL-1651

H4IIE Rat hepatoma ATCC# CRL-1548

HEK-293 Human embryonic kidney epithelial ATCC# CRL-1573

L929 Mouse connective tissue fibroblast ATCC# CRL-1

MA-10 Mouse tumor Leydig cell (Ascoli, 1981)

2.3.1 General cell culture medium

COS-7, H4IIE cells and L929 cells were grown in Dulbecco’s Modified Eagle’s Medium

(DMEM) (Gibco Cell Culture, Invitrogen), supplemented with 10 units/ml penicillin

(Sigma), 10 µg/ml streptomycin (Sigma) and 10% heat inactivated foetal calf serum (FCS)

(Gibco BRL, Invitrogen).

MA-10 cells were grown in a 1:1 mixture of DMEM (containing 1 g/l d-glucose, 4 mM

l-glutamine, 25 mM HEPES buffer, 4 mg/l pyroxene HCl and 110 mg/l sodium pyruvate;

Gibco Cell Culture, Invitrogen) plus nutrient mixture F-12 HAM (with NaHCO3, without l-

glutamine; Sigma, Australia), supplemented with 7.5% horse serum (Sigma), 2.5% FCS

41

(Gibco Cell Culture, Invitrogen), 10 units/ml penicillin (Sigma) and 10 µg/ml streptomycin

(Sigma).

HEK-293 cells were maintained in a 1:1 mixture of DMEM (Gibco cell culture,

Invitrogen) and F-12 HAM nutrient medium containing L-Glutamine (Sigma, Australia),

supplemented with 2 mM L-glutamine, 10 units/ml penicillin (Sigma), 10 µg/ml

streptomycin (Sigma) and 10% heat inactivated FCS (Gibco BRL, Invitrogen).

2.3.2 Maintenance of cell cultures

Cells were grown at 37ºC with 5% CO2 in a humidified environment in 0.2 µm vented-cap

tissue flasks (25 or 75 ml), trays (6-, 24-, 48- or 96-well) or dishes (30, 60, 100 or 150 mm)

(Falcon, Becton Dickinson, Franklin Lakes, NJ, USA) containing culture medium as

indicated. All manipulations were carried out in a laminar flow unit (Gelman Sciences

Australia) using sterile equipment to ensure sterility. Cultures were split every 2-3 days,

depending on the speed of growth. Adherent cells were first washed with sterile PBS,

followed by addition of enough trypsin-EDTA (Gibco cell culture, Invitrogen) to cover the

bottom of the flask. After incubation at 37ºC for two minutes, the cells were dislodged and

resuspended in warm culture medium containing 10% FCS to inactivate trypsin activity.

The cells were then replated at a dilution of 1:2-1:10 of the original culture.

2.3.3 Cryopreservation

Cells were harvested at log phase using trypsin/EDTA (if adherent) followed by

centrifugation at 1000 x g. The cell pellet was then resuspended in culture medium

containing 10% DMSO (filter sterilised) and added to a cryotube (Lat Tek Nunc, Roskilde,

Denmark). The tubes were transferred to a freezing chamber (Nalgene, Rochester, NY,

USA) containing room temperature isopropanol. The chamber was then placed in a -80oC

freezer, to achieve a cooling rate of approximately -1oC/minute. Frozen cells were stored in

liquid nitrogen until required.

2.3.4 Thawing frozen cell lines

Cryotubes (Lat Tek Nunc, Roskilde, Denmark) containing frozen cells were thawed in a

37ºC water bath and transferred to a sterile tube. Ten ml of warm culture medium was

42

added drop-wise and the suspension centrifuged at 1000 x g to harvest the cells. The

supernatant was removed, and the cells were washed in 10 ml of warm culture medium

before being transferred to cell culture flasks at a cell density of approximately 2 x 105

cells/ml.

2.3.5 Transfection of cell lines

Cells were transfected using Fugene-HD (Roche), according to manufacturer’s directions.

Briefly, the cells were plated into 35 mm tissue culture dishes 16-24 hours prior to

transfection, to be approximately 70% confluent on the day of transfection. Plasmid DNA

was diluted in DMEM containing no additives before adding Fugene-HD. After incubation

at room temperature for 15-60 minutes the mixture was added drop-wise to the cells. The

transfected cells were used in experiments 24-48 hours after transfection. A Fugene-

HD:DNA ratio of 3:1 (µl:ug) was used for transfection of COS-7, MA-10, L929, H4IIE

and HEK-293 cells.

2.3.6 Production of cell lines with stable expression of transfected genes

Cells stably expressing HCA112 were produced by transfection using Fugene-HD,

followed by selection of co-expression of a resistance marker. Forty eight hours after

transfection, the relevant selecting antibiotic was added to the cell culture media. The

concentration of antibiotic used was the minimum concentration required to kill 100% of

untransfected cells. Following selection, cells were maintained in antibiotic at a lower

concentration, which nevertheless killed most untransfected cells. The following

concentrations of puromycin (Sigma) were used: COS-7: selection 7 µg/ml, maintenance 5

µg/ml; HEK-293: selection 2 µg/ml, maintenance 0.5 µg/ml; L929: selection 6 µg/ml,

maintenance 5 µg/ml.

2.3.7 TNF� stimulation of MA-10 cells

Cells were grown in tissue culture dishes for 24 hours prior to stimulation. Fresh medium

containing the appropriate concentration of recombinant rat TNF� (RND Systems) was

added and culture continued under normal conditions for 24 hours.

43

2.4 Nucleic acid and recombinant DNA techniques

2.4.1 Oligonucleotides used in this study

See Table 2-2 for the sequence of oligonucleotides used in this study.

2.4.2 Synthetic oligonucleotides

All synthetic oligonucleotides were purchased from GeneWorks (Adelaide, South

Australia) (see Table 2-2). The concentrations of the synthesised oligonucleotides were

calculated using the following formula and an average mononucleotide MW of ~ 330

daltons:

Oligonucleotide concentration (µM) = Concentration (mg/ml) x 106)

Length (nucleotides) x mononucleotide MW

2.4.3 RNA extraction

RNA was extracted from cultured cell lines and tissues using TRIzol (Invitrogen)

according to the manufacturer’s instructions.

2.4.4 Estimation of RNA concentration

To estimate the concentration of total RNA in a sample, the optical density (OD) was

measured at 260 nm using a UV spectrophotometer (Eppendorf). RNA concentration was

calculated using the following formula:

RNA concentration (µg/ml) = OD260 x dilution factor x 40

RNA purity was estimated by the OD260 /OD280 ratio, with a minimum accepted ratio of

1.80. To ensure the RNA sample was not degraded, 1 µl of total RNA was mixed with

Formaldehyde Loading Dye (Ambion), heated for 10 minutes at 70oC and subjected to

agarose gel electrophoesis (see Section 2.4.12).

2.4.5 Preparation of cDNA from mRNA by reverse transcription

RNA was reverse transcribed to produce cDNA using SuperScript II RNaseH- Reverse

Transcriptase (Invitrogen, Carsbad, CA, USA). One microlitre of oligo dT (Invitrogen)

was added to 2 µg of RNA in 11 µl of RNase-free water (Takara) and heated to 70oC for 5

44

minutes. The mixture was allowed to cool at room temperature for 10 minutes and the

following were added and mixed: 4 µl first strand buffer, 2 µl 0.1 M DTT, 1 µl Superscript

(all from Invitrogen) and 1 µl 10 mM dNTP (Sigma). The mixture was then incubated at

42oC for 90 minutes, after which 80 µl of RNase free water was added and the cDNA

stored at -20oC until required.

2.4.6 Polymerase chain reaction (PCR)

All thermal cycling, including PCR and sequencing reactions, were performed using a

PTC-100 Programmable Thermal Controller (MJ Research, Cambridge, MA, USA). Five

micromolar stocks of each primer were prepared in MilliQ water from original stocks.

Each reaction consisted of 12.5 µl ExTaq Premix (Takara), 1 µl forward primer (5 µM), 1

µl reverse primer (5 µM) and 1 µl template DNA (diluted plasmid DNA [~5 ng] or

undiluted synthesized cDNA, as specified). The following PCR program was used for most

PCR amplifications, with minor variations in annealing temperatures (step 3), extension

times (step 4) and cycle number (step 5) as appropriate: Step 1, 95°C for 10 min; Step 2,

95°C for 30 secs; Step 3, 58°C for 30 secs; Step 4, 72°C for 30 secs; Step 5, repeat steps 2-

4 (x25); Step 6, 72°C for 5 mins; Step 7, 4°C until end.

2.4.7 DNA sequencing

Plasmid DNA was sequenced by the addition of 1 µl of plasmid DNA (~300 ng), 1 µl of

BigDye Terminator v.3 (Applied Biosystems), 1 µl of oligonucleotide (5 µM), 3 µl of

BigDye Buffer and 12 µl of MilliQ water to a PCR tube. After mixing by vortexing and

brief centrifugation, samples were subjected to thermal cycling according to the following

program: Step 1, 96°C for 30 seconds; Step 2, 50°C for 15 seconds; Step 3, 60°C for 4

minutes; Step 4, repeat steps 1-3 (x 25); Step 5, 4°C for 15 mins.

To precipitate extension products, samples were transferred to 1.5 ml microcentrifuge

tubes, along with 5 µl of 125 mM EDTA. Precipitation was achieved by vortexing briefly

with 60 µl of 100% ethanol, followed by incubation at room temperature for 15 minutes.

Tubes were then centrifuged at 20,800 x g for 25 minutes at 4oC and the supernatant was

removed. The pellets were washed by the addition of 60 µl of 70% ethanol, brief vortexing

and then centrifugation at 20,800 x g for 5 minutes. After careful aspiration of the

45

supernatants, the samples were dried in a vacuum centrifuge for 5 minutes. The samples

were then delivered to the DNA sequencing facility at the Institute of Medical and

Veterinary Science (IMVS) for nucleotide sequencing using a model 3700 automated DNA

sequence analyser (Applied Biosystems, Foster City, CA).

2.4.8 Preparation of competent bacteria for transformation

Unless noted otherwise, a single colony of E. coli, strain DH5�, was grown overnight in 10

ml Luria Bertani (LB) broth at 37oC with shaking. Five ml of overnight culture was sub-

cultured into pre-warmed 100 ml LB broths and grown at 37oC until an optical density ~

0.6 at 600 nm was reached. The culture was then cooled on ice for 15 minutes before

centrifugation at 4000 x g for 7 minutes at 4oC. The bacterial pellet was resuspended in 40

ml of cold 100 mM MgCl2, followed by centrifugation at 4000 x g for 7 minutes at 4oC.

The resulting cell pellet was resuspended in 4 ml of cold CaCl2 and incubated on ice for 1

hour. Glycerol was added to a final concentration of 13% and the competent cell

suspension was dispensed into 200 µl aliquots and stored at -80oC until required.

2.4.9 Transformation of competent bacteria with plasmid DNA

To transform competent bacteria, competent cells were first defrosted on ice, before adding

plasmid DNA and incubating on ice for 5 minutes. The bacteria were then ‘heat-shocked’

at 42oC for 30 seconds before incubation on ice for a further 2 minutes. SOC medium (800

µl) was added and the bacteria were incubated at 37oC for 45 minutes, before

centrifugation at 20,800 x g for 1 minute. The supernatant was removed and the pellet was

resuspended in 100 µl of saline (0.85% NaCl) and spread onto Luria agar plates containing

the appropriate antibiotic to select for transformed bacteria (100 µg/ml ampicillin, 50

µg/ml kanamycin, 50 µg/ml chloramphenicol, 75 µg/ml streptomycin). Plates were

incubated overnight at 37oC to detect resistant colonies.

2.4.10 Preparation of plasmid DNA

For all types of plasmid DNA preparations, single colonies of bacteria were used to

inoculate LB broths (containing appropriate antibiotic for plasmid selection), and these

were incubated at 37oC overnight. The following kits were used to isolate DNA according

to the manufacture’s instructions: QiaPrep Spin Miniprep Kit (Qiagen); PureYield Plasmid

46

MidiPrep System (Promega, Madison WI, USA); and EndoFree Plasmid Maxi Kit

(Qiagen). For diagnostic purposes, single colonies were isolated from Luria agar plates and

resuspended in 50 µl MilliQ water. The samples were further mixed by vortexing, followed

by heating for 5 minutes at 95oC to lyse the bacterial cells and release the plasmid DNA.

2.4.11 Restriction digestion of plasmid DNA

Restriction endonucleases were purchased from New England Biolabs (Beverly, MA,

USA) and used at a working concentration of approximately 10 units/µl. For diagnostic

digests of plasmid DNA using a panel of restriction enzymes, 1 µl of DNA (approximately

0.2 µg of DNA) was mixed with 1 µl of restriction enzyme(s), 1 µl of the appropriate 10x

buffer and MilliQ water to a final volume of 10 µl. The samples were incubated at 37oC for

1-2 hours to allow digestion to take place and analysed by agarose gel electrophoresis (see

below). For preparation of fragments for ligation, 10 µl of plasmid DNA was mixed with 1

µl the appropriate restriction enzyme(s), 2 µl of the appropriate 10x buffer and MilliQ

water to a final volume of 20 µl, and the samples were incubated at 37oC for 1-2 hours.

2.4.12 Agarose gel electrophoresis

Agarose gels were prepared by dissolving DNA grade agarose (Progen, Heidelberg,

Germany) to a concentration of 1% (w/v) in TAE buffer using a microwave oven. After

cooling to approximately 55oC, the gels were cast in horizontal EasyCast Mini Gel

Electrophoresis Systems (Owl Separation Systems, Portsmouth, NH, USA), according to

the manufacturer’s instructions. Samples to be electrophoresed were mixed with loading

buffer (Trackit, Invitrogen) at a ratio of 5:1. Appropriate volumes of sample and DNA

ladder (100 bp, Invitrogen or 1 kb, New England Biolabs) were then loaded into wells and

subjected to electrophoresis until the desired separation was achieved. Gels were stained

for 10 minutes with either Gel Red Nucleic Acid Gel Stain (Biotium) (3x in water) or

ethidium bromide (2,7-Diamino-10-ethyl-9-phenyl phenanthridinum bromide) (Sigma)

(0.2% w/v) in TAE buffer and de-stained for 5 minutes in water. DNA was visualised

using a UV transilluminator (UVP, Inc.).

47

2.4.13 Extraction of DNA from gels

After electrophoresis, DNA was extracted from bands in agarose gels for downstream

applications such as cloning. The required band was excised using a scalpel blade and the

DNA purified using the QIAquick Gel Extraction Kit (Qiagen), according to the

manufacture’s instructions.

2.4.14 Annealing of oligonucleotides

In order to insert DNA encoding a small peptide into a restriction site in a plasmid (e.g. an

epitope tag into the middle of a protein sequence), complementary oligonucleotides were

designed which encoded the peptide of interest plus single-stranded overhangs to

complement the insertion site in the digested plasmid. The single-stranded oligonucleotides

were first annealed together and then ligated into the appropriately digested destination

plasmid. To anneal the single-stranded oligonucleotides to form a double stranded ‘insert’,

the oligonucleotides encoding the top and bottom strands were resuspended in MilliQ

water to a concentration of 200 µM. The oligonucleotides were then mixed in annealing

buffer (10 mM Tris-HCl, pH 8.0, 1 mM EDTA, pH 8.0, 0.1 M NaCl) at a 1:1 ratio to give

a final concentration of 50 µM, heated to 95oC for 4 minutes and allowed to cool to room

temperature over 5-10 minutes. This resulted in a 50 µM solution of double stranded

‘insert’. The ‘insert’ was subjected by agarose gel electrophoresis to check for successful

annealing before ligation with an appropriately digested plasmid (described below).

2.4.15 Site directed mutagenesis using overlap extension PCR

Overlap extension PCR allows site-directed mutagenesis in the middle of a coding

sequence without using mega-primers. Complementary oligonucleotide primers, encoding

the desired mutation and PCR are used to generate two DNA fragments with overlapping

ends. These fragments are combined and used as template in a subsequent PCR reaction

using primers which amplify the entire coding region. The overlapping ends anneal,

allowing the 3’ overlap of each strand to serve as a primer for the 3’ extension of the

complementary strand, as depicted in Figure 2-1, below (Ho et al., 1989).

48

Figure 2-1. Overlap extension PCR

Cartoon depicting process of site-directed mutagenesis using overlap extension PCR. Adopted from (Ho et al., 1989).

The first PCR (1) amplifies fragment ‘AB’ using the forward external primer (a) and

the internal reverse primer (b), which encodes the mutation of interest. The second PCR (2)

amplifies fragment ‘CD’ using the internal forward primer (c), which encodes the mutation

of interest, and the external reverse primer (d). The two PCR reactions generate two

fragments which contain the desired mutation, which are mixed and used as the template

in a third PCR using the external forward and reverse primers (a and d). The external

primers are designed such that they have an annealing temperature similar to that of the

internal primers and hence the two fragments.

2.4.16 DNA ligation

Digested ‘insert’ DNA and similarly digested ‘backbone’ plasmid DNA were mixed at a

molar ratio of 3:1 and made up to a final volume of 17 µl. The mixture was added to a thin

walled PCR tube, followed by 2 µl of 10x ligation buffer and 1 µl T4 DNA ligase (New

England Biolabs, Beverly, MA, USA). The mixture was then incubated at either 15oC for 2

hours or overnight at 4oC. Ligation products were transformed into chemically competent

bacteria, as described above, and plated onto Luria agar plates containing the appropriate

49

selection antibiotic. As a control for ‘backbone’ plasmid self-ligation, ‘insert’ DNA was

replaced with MilliQ water and ligation performed as described.

2.4.17 Generation of expression vectors

Several expression vectors were utilised during this study and are described herein. The

sequence of the insert cDNA in all vectors was confirmed by automated DNA sequencing

after cloning (see above).

2.4.17.1 pcDNA3-HCA112-cmyc and pcDNA3-HCA112-HA

To generate c-terminal cmyc (EQKLISEEDL) and influenza haemagglutinin (HA)

(YPYDVPDYA) epitope tagged HCA112 constructs, HCA112 was PCR-amplified from

the RZPD clone IRAVp968F0530D6 (RZPD GmbH, Germany) using the forward primer,

which incorporates a Kozac consensus sequence (bold); 5’-

GGATCCCACCATGTCCACAGACATGGAGACTGCAG-3’ and either of the reverse

primers; 5’-

GAATTCCTACAGATCTTCTTCAGAAATAAGTTTTTGTTCGATCACAGCTGCACCCA

GCAG-3’ encoding cmyc (italicised) or 5’-

GAATTCCTAAGCGTAGTCTGGGACGTCGTATGGGTAGATCACAGCTGCACCCAG

CAG 3’ encoding HA (italicised). The products were cloned into pGEM-T Easy Vector

(Promega) to yield pGEM-T Easy-HCA112-cmyc and pGEM-T Easy-HCA112-HA. The

inserts were then excised with BamHI and EcoRI (sites underlined in primers) and inserted

into the respective sites in pcDNA3 (Invitrogen), generating pcDNA3-HCA112-cmyc and

pcDNA3-HCA112-HA.

2.4.17.2 pEF-HCA112-cmyc-IRES-puro6 and pEF-HCA112-HA-IRES-puro6

To generate c-terminal c-myc or HA epitope tagged HCA112 constructs featuring a

puromycin resistance cassette, the coding region was sub-cloned from pcDNA3-HCA112-

cmyc and pcDNA3-HCA112-HA, respectively, into pEF-IRES-puro6 (a generous gift from

Dr Daniel Peet, University of Adelaide, Australia), using the restriction sites BamHI and

NotI.

50

2.4.17.3 pEGFP-HCA112

To generate an EGFP-HCA112 fusion construct, HCA112 was amplified by PCR from the

pCMVSport6 plasmid containing the RZPD clone IRAVp968F0530D6 (RZPD GmbH,

Germany) using the following primers; 5’-

CTCGAGCTATGTCCACAGACATGGAGACT-3’ and 5’-

GGATCCCTAGATCACAGCTGCACCC-3’. The PCR product was ligated into pGEM-T-

Easy (Promega) and then sub-cloned in-frame into pEGFP-C1 (Clontech Laboratories,

Palo Alto, CA), utilising the XhoI and BamHI restriction sites (underlined in primer).

2.4.17.4 pEGFP-HCA112-HA

To generate an EGFP-HCA112 fusion protein with a HA tag on the predicted extracellular

side on the protein, HCA112 was first sub-cloned from pEGFP-HCA112 into pEF-IRES-

puro6 using the BamHI and XhoI restriction sites, generating pEF-HCA112-IRES-puro6.

The following oligonucleotides (encoding HA (italicised) and featuring NdeI overhangs)

were annealed and ligated into the NdeI restriction site (located between the third and forth

transmembrane domains of the HCA112 protein) in pEF-HCA112-IRES-puro6; 5’-

TATGACTACCCATACGACGTCCCAGACTACGCTTCA-3’ and 5’-

TATGAAGCGTAGTCTGGGACGTCGTATGGGTAGTCA-3’. Following ligation,

HCA112-HA was sub-cloned from pEF-IRES-puro6 back into pEGFP-C1, utilising the

BamHI and XhoI restriction sites. The resulting plasmid is named pEGFP-HCA112-HA

and the fusion protein expressed is designated EGFP-HCA112-HA.

2.4.17.5 pEF-HCA112-HA (EC)-IRES-puro6

It was next decided to generate a construct encoding HCA112 with an extracellular HA tag

(HCA112-HA (EC)) that utilised a plasmid with a puromycin resistance cassette (pEF-

IRES-puro6) so stably transfected cells could be generated. To achieve this, HCA112-HA

(EC) was amplified using PCR from the plasmid pEGFP-HCA112-HA using the forward

primer, which incorporates a Kozac consensus sequence (bold), 5’-

GGATCCCACCATGTCCACAGACATGGAGACTGCAG-3’; and the reverse primer, 5’-

GGATCCCTAGATCACAGCTGCACCC-3’. The PCR product was ligated into pGEM-T-

Easy (Promega) and then sub-cloned into pEF-IRES-puro6 utilising the BamHI restriction

51

site (underlined in primer), generating the construct pEF-HCA112-HA (EC)-IRES-puro6.

The fusion protein expressed is designated HCA112-HA (EC), where EC is the

abbreviation of extracellular.

2.4.17.6 pET-21a-HCA112140-191

To generate an isopropyl-�-D-thiogalactoside��IPTG) inducible construct which could be

used to express a 6xHis tagged HCA112 peptide (amino acids 140-191) in bacteria, the

following primers (which contain NheI and XhoI restriction sites (underlined)), were used

to amplify the peptide from pEF-HCA112-HA-IRES-puro6; 5’-

TATACATATGGCTAGCATGATCGTTATTGGGTCTCGTG-3’ and 5’-

GGTGGTGGTGCTCGAGGCTTGTGTAGTATATGCACAAGG-3’. The PCR product

was digested with NheI and XhoI and ligated into the respective restriction sites of pET-

21a (Novagen, Madison, WI).

2.4.17.7 pEF-HCA112-HA (EC)�2-56-IRES-puro6 and pEF-HCA112-HA (EC)�2-

60-IRES-puro6

To generate mammalian expression constructs encoding N-terminal truncation mutants of

HCA112 featuring an extracellular HA tag, HCA112-HA (EC) was amplified by PCR from

pEGFP-HCA112-HA using either of the forward primers: 5’-

CTCGAGCACCATGAGCAGCAGAGTGCTGGT-3’ or 5’-

CTCGAGCACCATGCTGGTGGCCTCCTG-3’ and the reverse primer: 5’-

GGATCCCTAGATCACAGCTGCACCC-3’, to generate PCR products encoding

HCA112-HA (EC)�2-56 or HCA112-HA (EC)�2-60, respectively. The forward primers

feature a Kozak consensus sequence (bold) upstream of the mutated start codon. The PCR

product was ligated into pGEM-T-Easy (Promega) and then sub-cloned into pEF-IRES-

puro6, utilising the XhoI and BamHI restriction sites (underlined in primer).

2.4.17.8 pEF-HCA112-HA (EC)�223-244-IRES-puro6

To generate a mammalian expression construct encoding a C-terminal truncation mutant of

HCA112 featuring an extracellular HA tag, HCA112-HA (EC) was amplified by PCR from

pEGFP-HCA112-HA using the forward primer: 5’-

CTCGAGCACCATGTCCACAGACATGGAGACT-3’ which incorporates a Kozak

52

consensus sequence (bold) and the reverse primer: 5’-

GGATCCCTAGATGTAGACACATACAGGGGT-3’, to generate PCR products encoding

HCA112-HA (EC)�223-244. The reverse primer features a BamHI restriction site

(underlined) downstream of the mutated stop codon. The PCR product was ligated into

pGEM-T-Easy (Promega) and then sub-cloned into pEF-IRES-puro6, utilising the XhoI

and BamHI restriction sites (underlined in primer).

2.4.17.9 pEF-HCA112-HA (EC)LL238AA-IRES-puro6

To generate a mammalian expression construct encoding an LL to AA mutation at amino

acids 238 and 239 of HCA112 featuring an extracellular HA tag, HCA112-HA (EC) was

amplified by PCR from pEGFP-HCA112-HA using the forward primer: 5’-

CTCGAGCACCATGTCCACAGACATGGAGACT-3’ which incorporates a Kozak

consensus sequence (bold) and the reverse primer: 5’-

GGAATCCCTAGATCACAGCTGCACCCGCCGCTTTCTT -3’, which incorporates the

LL to AA mutation (italicised), to generate PCR products encoding HCA112-HA

(EC)LL238AA The PCR product was ligated into pGEM-T-easy (Promega) and then sub-

cloned into pEF-IRES-puro6, utilising the XhoI and BamHI restriction sites (underlined in

primer).

2.4.17.10 pEF-HCA112-HA (EC)YIWKRFF222AIAAAAA-IRES-puro6

Overlap extension PCR was used to generate a mammalian expression construct encoding

HCA112, featuring an extracellular HA tag with a Y222IWKRFF to AIAAAAA mutation.

Fragment one (~700 bp) was amplified by PCR from pEGFP-HCA112-HA using the

‘external’ forward primer: 5’- CTCGAGCACCATGTCCACAGACATGGAGACT-3’

which incorporates a Kozak consensus sequence (bold) and the ‘internal’ reverse primer:

5’- GTTTCCGCCTTTGTGGCAGCTGCTGCCGCGATGGCGACACATAC -3’, which

incorporates the YIWKRFF222AIAAAAA mutation (italicised). Fragment two (~100 bp)

was amplified using the same template and the ‘internal’ forward primer 5’-

GTATGTGTCGCCATCGCGGCAGCAGCTGCCACAAAGGCGGAAAC-3’, which

incorporates the YIWKRFF222AIAAAAA mutation (italicised), and the ‘external’ reverse

primer 5’-GGATCCCTAGATCACAGCTGCACCC-3’. The two fragments were mixed

and used as the template (1µl of each) in a third PCR using the ‘external’ forward and

53

reverse primers, above, to generate PCR products (~800 bp) encoding HCA112-HA

(EC)YIWKRFF222AIAAAAA. The PCR product was ligated into pGEM-T-easy

(Promega) and then sub-cloned into pEF-IRES-puro6, utilising the XhoI and BamHI

restriction sites (underlined in primer).

2.4.17.11 pEF-LR8-FLAG-IRES-puro6

To generate a mammalian expression construct encoding mouse LR8 with a C-terminal

FLAG tag, mouse LR8 was first amplified by PCR from mouse L929 cell cDNA, using the

LR8 specific forward primer 5’-CTCGAGCACCATGGTCCAGAGCACAGTGACTG-3’

which incorporates a Kozak consensus sequence (bold) and the reverse primer 5’-

GGATCCTCACTTGTCATCGTCGTCCTTGTAGTCCAGGATAGCAGGGATCTTCTC-

3’ which incorporates a FLAG tag (italicised). The PCR product was ligated into pGEM-T-

Easy (Promega) and then sub-cloned into pEF-IRES-puro6, utilising the XhoI and BamHI

restriction sites (underlined in primer).

2.4.17.12 pCaveolin-1-EGFP and pcDNA3-FAT/CD36

The plasmids pCaveolin-1-EGFP and pcDNA3-FAT/CD36, which encode a fusion protein

of caveolin-1 and EGFP, and rat FAT/CD36, respectively, were a kind gift from Dr Nick

Eyre, School of Molecular and Biomedical Science, University of Adelaide, Adelaide,

Australia (Eyre et al., 2007).

2.5 Production of antibodies

2.5.1 Production of anti-peptide antibodies in rabbits

The peptide KRFFTKAETEEKKLLGA from the mouse HCA112 amino acid sequence

was chosen to raise anti-peptide antibodies in rabbits. It has high similarity with the

orthologous rat sequence (KRFFTKAETE_KKLLGA) and it was expected that antibodies

raised against it would cross-react with the rat protein. Furthermore, the peptide is

predicted to consist of � helix, there are no cysteines present, and it is not expected to have

rigid secondary structure or tertiary structure. The peptide is not similar to any other

proteins in the NCBI database, as determined using BLAST analysis.

54

The peptide was produced by the Biomolecular Resource Facility at the John Curtin

School of Medical Research, Australian National University, Canberra. A cysteine was

added to the N-terminus (CKRFFTKAETEEKKLLGA), thus allowing conjugation with

keyhole limpet hemocyanin (KLH) to improve immunogenicity of the peptide.

2.5.1.1 Antigen Preparation

The peptide–KLH conjugate was prepared for immunisation by emulsification with

Freund’s Complete (primary immunisation) or Incomplete (subsequent immunisations)

Adjuvant (CFA and IFA respectively). Peptide (dissolved in PBS) and adjuvant were

mixed at a ratio of 45:55, and emulsified using two syringes joined by a stop-cock.

2.5.1.2 Immunisation

Immunisations were performed by Laboratory Animal Services under the ethical

guidelines of the University of Adelaide Animal Ethics Committee. Rabbits were

immunised with 1 mg peptide in adjuvant subcutaneously at 10 sites (100 µl per site) on

days 0 (CFA), 14, 42 and 70 (IFA). Test bleeds were taken via the ear vein 10 days after

immunisation to assess serum antibodies against the immunising peptide. The animals

were bled terminally by cardiac puncture on day 83.

2.5.1.3 Analysis of serum antibodies against the peptide - direct ELISA

Wells in 96 well Costar EIA/RIA flat bottomed high binding trays were coated with un-

conjugated peptide diluted in 100 mM sodium bicarbonate (200 ng/well) by incubation at

4oC overnight. Following washes with PBS/0.05% Tween 20, wells were incubated with

PBS containing 3% BSA for 2 hours at room temperature to block non-specific protein

binding sites. After washing, serum diluted in PBS containing 1% BSA (1:3000 -

1:100,000) was added to the wells and incubated at room temperature for 90 minutes.

Rabbit antibodies were detected by incubation with donkey anti-rabbit Ig conjugated with

HRP (Rockland) (1:10,000) for 45 minutes at room temperature. After several washes,

bound donkey antibody was detected by addition of o-Phenylenediamine dihydrochloride

(SigmaFAST OPD) (Sigma, country), and allowing colour to develop for 5-10 minutes

before stopping with 50 µl of 3 M HCl per well. Absorbance was measured at 490 OD on a

plate reader (Amersham Biosciences).

55

2.5.2 Production of anti-peptide antibodies in rats

2.5.2.1 Antigen preparation

A recombinant peptide for use in immunisation of rats was produced by transforming

CodonPlus BL21(DE3) E. Coli (Stratagene) with an expression plasmid (pET-21a-

HCA112140-191) encoding amino acids 140-191 of the mouse HCA112 protein. Details of

the plasmid are described in section 2.4.17.6. Protein expression was induced with 0.1 mM

IPTG (Astral Scientific, Australia) for 5 hours at 37oC. Following induction, the bacteria

were harvested by centrifugation at 11,000 x g and lysed by gentle rotation for 1 hour at

room temperature in buffer containing 8 M urea, 100 mM NaH2PO4, 10 mM Tris-Cl and 1

mM PMSF, pH 8. The HCA112 peptide was purified from cleared lysate (centrifugation at

20,800 x g for 30 minutes) using nickel-coated beads – His-Link (Promega). Twenty mls

of the cleared lysate was incubated with 0.5 mls of beads for 60 minutes at room

temperature with rotation, to allow binding. The mixture was then added to a 10 ml column

(Pierce Biotechnology), the beads were allowed to settle and the lysate was drained. The

beads were washed with 20 ml of lysis buffer containing 25 mM imadazole, before the

peptide was eluted with lysis buffer containing 1 M imadazole. Purified protein was

precipitated (Mundy et al., 2002), by mixing the eluted protein with TCA (7% final

concentration, w/v) and sodium deoxycholate (0.015% final concentration, w/v). The

mixture was incubated on ice for 2 hours or overnight at -80oC, before centrifugation at

20,800 x g. The pellet was washed in ice cold acetone, followed by centrifugation at

20,800 x g. The resulting pellet was resuspended in PBS and stored at -20oC until required.

2.5.2.2 Immunisation of rats

Eight week old female Hooded Wistar rats were immunised under isoflurane anaesthesia

with 50 µg of recombinant peptide in 0.15 ml of PBS mixed with 25 µg muramyl dipeptide

(Bachem). Half was delivered by intra-peritoneal injection and the remainder at three

subcutaneous sites. The rats received booster immunisation at 3 week intervals, in each

case at three subcutaneous sites. Ten days after each immunisation, approximately 500 µl

of blood was obtained via the tail vein while the rats were under isoflurane anaesthesia.

56

2.6 Antibodies used in this study

See Table 2-3 and Table 2-4 for the primary and secondary antibodies used in this study.

2.7 Protein techniques

2.7.1 Preparation of whole cell lysates

To prepare whole cell lysates, cells were grown to confluency in tissue culture dishes,

washed twice with ice cold PBS and lysed in an appropriate volume of lysis buffer

containing protease inhibitor cocktail (1:100) (Sigma). Cells were harvested by scraping

with a plastic cell scraper, transferred to a microcentrifuge tube and homogenised by 10

passages through a 28g needle. Lysates were incubated on ice for 30 minutes before

centrifugation at 1000 x g to remove the nuclear debris. Protein concentration was

quantified using a Bradford assay (reagent from Bio-Rad) and 25 µg was subjected to

SDS-PAGE and immunoblotting.

Whole cell lysates of mouse tissues were prepared from tissues harvested from an 8

week old male C57BL/6 mouse. The tissues were placed in a microcentrifuge tube and

snap frozen in liquid nitrogen. Frozen tissue was ground to a fine power under liquid

nitrogen using a mortar and pestle cooled by dry ice. Lysis buffer (25 mM Tris-Cl, pH 7.5,

150 mM NaCl, 5 mM EDTA, 1% Triton X-100 and protease inhibitors) was added to the

tissue powder at a concentration of 100 mg/ml and the mixture was passaged through a 28g

needle 10 times. After incubation on ice for 30 minutes, lysates were centrifuged at 1000 x

g to remove nuclear debris. Protein concentration was measured using a Bradford assay

and a volume containing 25 µg was subjected to 12% SDS-PAGE and immunoblotting.

2.7.2 (Co-) Immunoprecipitation

To immunoprecipitate proteins, cells were grown to confluency, washed 3 times with cold

PBS and lysed in lysis buffer (1% Triton X-100, 50 mM Tris, pH 7.5, 200 mM NaCl and 1

mM EDTA) containing protease inhibitors (1:100) (Sigma). Cells were harvested by

scraping, transferred to a microcentrifuge tube and lysed by 10 passages through a 28g

needle. Lysates were incubated on ice for 30 minutes, before centrifugation at 1000 x g to

remove nuclear debris. Protein concentration was measured using a Bradford assay

57

(reagent from BioRad) and equal amounts were pre-cleared by rotation for 1 hour at 4ºC

with 5 µl/ml of magnetic protein G beads (Dynal/Invitrogen). Beads were captured using a

magnet (Dynal) and the supernatant was transferred to a fresh microcentrifuge tube and

incubated with antibody at 4ºC overnight with rotation. Magnetic protein G beads (10

µl/ml) were then added and incubated with rotation for 1 hour at 4ºC. The beads were then

washed four times with wash buffer (50 mM Tris, pH 7.5, 200 mM NaCl and 1 mM

EDTA), resuspended in 1x SDS-PAGE loading buffer, boiled and subjected to SDS-

PAGE.

2.7.3 Preparation of crude membrane fractions

To prepare crude membrane fractions, cells were grown to confluency in 100 mm tissue

culture dishes and washed twice with ice cold PBS. They were then lysed in 1 ml of

homogenisation buffer (20 mM Tris pH 7.5, 2 mM MgCl2, 0.2 M sucrose and protease

inhibitor cocktail (1:100)), passaged through a 28g needle 10 times and incubated on ice

for 30 minutes. Samples were centrifuged at 1000 x g to remove the nuclear debris and the

post-nuclear supernatants were transferred to polycarbonate tubes for centrifugation at

100,000 x g for 45 minutes at 4oC. The supernatant was collected and the pellet was rinsed

gently with homogenisation buffer before being harvested in homogenisation buffer. The

protein concentrations of the soluble and particulate fractions were estimated using a

Bradford assay (reagent from BioRad), according to the manufacturer’s directions. Equal

amounts of protein from each fraction were subjected to SDS-PAGE and immunoblotting.

2.7.4 Preparation of detergent-resistant membranes

Detergent-resistant membranes (DRMs) were prepared essentially as described (Peng et

al., 2004). Briefly, cells were grown to confluency in 100 mm tissue culture dishes,

washed twice in ice cold PBS and lysed in 0.7 ml of TNE (25 mM Tris-Cl, pH 7.5, 150

mM NaCl, 5 mM EDTA) buffer containing 1% Triton X-100 and protease inhibitors.

Lysates were transferred to microcentrifuge tubes and passed 10 times through a 28g

needle before incubation on ice for 30 minutes. After removal of the nuclear debris by

centrifugation at 1000 x g, 0.5 ml of the cleared lysate was transferred to a clean

microcentrifuge tube and mixed with an equal volume of 80% (w/v) sucrose in TNE

buffer. The mixture was then transferred to a 4.5 ml SW60 centrifuge tube and the sample

58

was overlaid with 2.5 ml of 38% (w/v) sucrose followed by 1 ml 5% (w/v) sucrose (both in

TNE buffer). After centrifugation at 38,000 rpm (148,305 x g) for 15 hours at 4oC, twelve

equal fractions were collect commencing from the top of the tube. Equal volumes of each

fraction were subjected to SDS-PAGE and immunoblotting. Caveolin-1 was used as a

marker to distinguish both raft- and non-raft-containing fractions.

2.7.5 Fractionation of organelles using Optiprep

Golgi membranes, lysosomes, mitochondria and peroxisomes were fractionated using

Optiprep (Axis-Shield PoC As, Oslo, Norway), according to the manufacture’s

instructions, method of (Graham et al., 1994). Briefly, cells were grown to confluency in

150 mm tissue culture dishes and lysed in ice cold homogenisation buffer containing 0.25

M sucrose, 1 mM EDTA, 10 mM Hepes-NaOH, pH 7.4 and protease inhibitors. Lysates

were transferred to a microcentrifuge tube, passed 10 times through a 28g needle and

incubated on ice for 30 minutes. To pellet the nuclei and heavy mitochondria, the lysate

was centrifuged at 3000 x g. Following this, the supernatant was centrifuged at 17,000 x g

for 12 minutes. The pellet was resuspended in homogenisation buffer and mixed with

Optiprep to a final concentration of 17.5% (w/v) iodixanol. Nine ml of the mixture was

transferred to a 10 ml ultracentrifuge tube and overlaid with 1 ml of homogenisation

buffer. After centrifugation at 270,000 x g for 3 hours at 4oC in an 80Ti (fixed angle) rotor,

nine 1.1 ml fractions were collected from the top of the tube. Equal volumes of each

fraction were subjected to SDS-PAGE and immunoblotting.

2.7.6 SDS-PAGE

To separate proteins based on their molecular weight, samples were boiled for 5 minutes at

100oC in loading buffer and then subjected to gel electrophoresis. Twelve percent (unless

otherwise indicated) SDS-polyacrylamide mini-gels were prepared using the Protean II

Dual Slab Cell gel electrophoesis apparatus (BioRad). After de-gassing, the resolving gel

mixture (for components see section 2-1) was poured and overlayed with water-saturated

butanol to prevent drying, during polymerisation for approximately 30 minutes at room

temperature. After removing the butanol with water and blotting paper, a comb was placed

on the top of the resolving gel and a 4% acrylamide stacking gel mixture (see section 2-1)

was poured. After polymerisation of the stacking gel the complete gels were placed in the

59

electrophoresis tank containing running buffer (GTS) and the samples were loaded in the

wells left by the comb. The proteins in the sample were separated by electrophoresis at 100

volts through the stacking gel and 200 volts through the resolving gel, until the dye front

reached the bottom of the gel (approximately 45 minutes).

2.7.7 Transfer of proteins from SDS-PAGE gel to PVDF membrane

After equilibrating the SDS-polyacrylamide gels in cold transfer buffer (see section 2-1)

for approximately 10 minutes, the proteins were transferred to PVDF membranes

(Hybond-P, Amersham Biosciences) using a Mini Trans Blot Electrophoretic Transfer Cell

(BioRad) at 100 volts for 70 minutes in cold transfer buffer. The gel and PVDF membrane

were sandwiched between Whatman filter paper and sponges, placed in a transfer cassette

and submerged in transfer buffer within the transfer apparatus, with the gel on the side of

the negative terminal. To monitor transfer and loading of proteins, the gel was stained

following transfer with Coomassie Brilliant Blue R-250 (0.025% (w/v) in 40% methanol

(v/v), 10% acetic acid (v/v)) for approximately 2 hours, followed by de-staining in 10%

methanol (v/v), 5% acetic acid (v/v) for several hours.

2.7.8 Immunoblotting

Following transfer, membranes were blocked for one hour at room temperature with 6%

skim milk (or 5% BSA in the case of anti-FAT/CD36 mAb Mo25) in TBS containing 0.1%

Tween-20 (v/v) (TBS-T). They were then incubated overnight at 4oC with primary

antibody (see Table 2-3) diluted in TBS-T containing 0.5% BSA and 0.05% sodium azide.

After thorough washing in TBS-T (3 changes, 15 minutes), membranes were incubated for

two hours at room temperature with horseradish peroxidase-conjugated secondary antibody

diluted in TBS-T containing 3% skim milk (or 0.5% BSA in the case of mAb Mo25). The

membranes were then washed thoroughly with TBS-T before detection of bound conjugate

using chemiluminescent substrate (SuperSignal West Femto Maximum sensitivity

Substrate, Pierce Biotechnology) and X-ray film (Curix Ortho Ht-G Medical X-ray Film,

AGFA). For serial detection of another protein on the same membrane, the membrane was

washed in TBS-T and stripped of bound antibodies by incubation at 55oC for 25 minutes in

Western blot stripping buffer. The membranes were then washed thoroughly with TBS-T,

60

blocked with for 1 hour at room temperature with 5% skim milk/TBST and re-probed with

another antibody as described above.

2.8 Immunofluorescence techniques

2.8.1 Preparation of glass cover-slips for growth of cell monolayers

To enhance cell attachment, glass cover-slips were soaked overnight in 10% (v/v) HCl,

followed by four washes in RO water. The cover-slips were then autoclaved in RO water

and washed in sterile PBS before placing in tissue culture trays.

2.8.2 Preparation of tissues for labelling

C57BL/6 mice were sacrificed by CO2 inhalation prior to harvesting tissues. Tissues were

excised, embedded in OCT (Tissue-Tek, Sakura Finetek, CA, USA) and snap-frozen in

isopentane cooled with liquid nitrogen. Tissue blocks were stored at -80oC in a sealed

container until required. Frozen sections (7 µm) were cut using a cryostat, lifted onto a

glass slide and air-dried for approximately 1 hour at room temperature. Sections were

stored in an airtight box with silica gel at -20oC for up to 2 weeks before use.

2.8.3 Preparation of cell smears

Cells were harvested by centrifugation at 1,000 x g for 5 minutes. The supernatant was

removed and the cells resuspended in 50% FCS in PBS followed by centrifugation at 1,000

x g for 5 minutes. All but approximately 50 µl of supernatant was removed and the cells

resuspended in the remaining 50% FCS. Approximately 5 µl of cell suspension was

dropped on a glass slide and smeared out to a circle of approximately 1 cm in diameter.

The slide was then flicked vigorously and allowed to air dry for about 1 hour at room

temperature. Smears were stored in an airtight box with silica gel at -20oC for up to 2

weeks before use.

2.8.4 Labelling of cell monolayers by indirect immunofluorescence

Cells were grown on sterile acid-washed glass cover-slips in 24-well tissue culture trays

for 24-48 hours prior to labelling for immunofluorescence. Where applicable, the cells

were transfected 24 hours prior to labelling. To label lipid rafts, cells were cultured for one

61

hour at 37°C in the presence of 0.5 µg/ml of Alexa594-Cholera toxin B subunit (Molecular

Probes) in DMEM containing 5% FCS. To label the endosomal recycling compartment,

cells were grown overnight in DMEM containing 5% FCS before incubating at 37oC with

human transferrin conjugated to Texas Red (Invitrogen) (10 µg/ml, diluted in culture

medium containing 5% FCS). The cells were washed with PBS, fixed with 3%

formaldehyde in PBS for 6 minutes on ice and rinsed with PBS containing 1% FCS

(PBS/FCS).

For indirect immunofluorescence, the cells were labelled with primary antibody diluted

in PBS/FCS for one hour on ice. After washing twice in PBS/FCS, the cells were labelled

with fluorochrome-conjugated secondary antibody diluted in PBS/FCS by incubation for

one hour in the dark on ice. After three washes in PBS, the cells were fixed with cold 1%

formalin (v/v) in PBS containing 2% glucose (w/v) and 0.02% azide (w/v) for 15 minutes,

rinsed in PBS and the cover-slips were then mounted on glass slides with Vectashield

mounting medium containing DAPI (Vector Laboratories). For detection of intracellular

antigens in fixed cells, 0.1% (w/v) saponin was included in wash buffers and antibody

preparations.

2.8.5 Indirect immunofluorescent labelling of mouse tissues

A liquid blocker pen (Daido Sangyo Co Ltd, Tokyo, Japan) was used to mark a circle of

approximately 1.5 cm in diameter on the glass slide around the tissue section or cell smear

to eliminate spreading of antibody. Tissue sections or cell smears were then fixed in ice

cold 100% acetone for 10 minutes at 4oC before being washed 3 times in cold PBS.

Sections were incubated in 40 µl of primary antibody diluted in cold PBS containing 10%

serum (NMS for mouse tissue sections or FCS for COS-7 cell smears) in a humidifier box

at 4oC for 1 hour. After washing 3 times in cold PBS, sections were incubated in secondary

antibody in a humidifier box at 4oC for 1 hour. Following washing, smears or sections

were fixed in 1% formalin (v/v) in PBS containing 2% glucose (w/v) and 0.02% azide

(w/v) for 15 minutes at 4oC. Finally the sections were washed in cold PBS and mounted

with cover-slips using Vectashield mounting medium containing DAPI (Vector

Laboratories) and sealed with nail polish. Sections were visualised using confocal

microscopy as described below.

62

2.8.6 Tracking internalisation of HCA112

Studies to track the internalisation of HCA112 from the cell surface were performed in

COS-7 cells grown on sterile cover-slips in 24-well tissue culture trays. The cells were

grown for 24 hours prior to transfection with a construct encoding HCA112 with an

extracellular HA tag (pEGFP-HCA112-HA or pEF-HCA112-HA (EC)-IRES-puro6, or

mutant HCA112-HA (EC) constructs described above (see section 2.4.17)). Twenty four to

48 hours after transfection, the cells were incubated with anti-HA antibody (Santa Cruz,

diluted 1:150, unless otherwise specified; purified 12CA5 hybridoma supernatant at 2

µg/ml or 12CA5 fab fragment at 65 µg/ml) for 1 hour on ice. After incubation with anti-

HA, 500µl of warm DMEM was added to the cells and they were incubated again at either

37oC or 4oC for the times indicated. The cells were then washed with PBS and fixed with

5% buffered formalin in PBS for 6 minutes on ice and rinsed with PBS containing 1% FBS

(PBS/FCS). Following fixation, cells were incubated with secondary antibody diluted in

PBS/FCS containing 0.1% (w/v) saponin, washed and fixed. The cover-slips were then

mounted and examined by confocal microscopy as described below.

2.8.7 Flow cytometry

For cell surface analysis of HCA112 expression, cells were grown to confluency in 100

mm tissue culture trays and recovered using trypsin. The cells were washed by

resuspending in approximately 3 ml of PBS and centrifugation at 200 x g. Aliquots of

approximately 1 x 106 cells were centrifuged and the pelleted cells were resuspended in 50

µl of primary antibody diluted PBS/FCS/Az supplemented with 10% normal mouse serum

(NMS) in the case of mouse cell line MA-10, or with 10% foetal calf serum (FCS) in the

case of COS-7 cells. The cells were incubated with the antibody on ice for 1 hour, and then

washed twice with PBS/FCS/Az. The cells were then labelled with 50 µl of fluorochrome-

conjugated secondary antibody diluted in PBS/FCS/Az, supplemented with 10% NMS or

FCS in the dark for 1 hour on ice. After a further two washes in PBS/FCS/Az, the cells

were fixed in 1% formalin (v/v) in PBS containing 2% glucose (w/v) and 0.02% sodium

azide (w/v). Labelled cells were analysed using the FACSCanto and CellQuest

(Biosciences).

63

2.8.8 Conventional fluorescence microscopy

Fluorescence microscopy was performed using an Olympus BX40 microscope, equipped

with an Olympus U-RFL-T burner. Images were acquired using an Olympus DP70 camera

and Olympus DP Controller software.

2.8.9 Confocal fluorescence microscopy

Confocal fluorescence microscopy was performed at Adelaide Microscopy (Adelaide,

Australia) using a Leica SP5 Spectral Scanning Confocal Microscope System. Images were

acquired for a single focal plane using a Kalman setting of at least 3. Images were

examined and manipulated using the Leica Application Suite (LAS) AF software.

2.8.10 Calcium immobilisation techniques

2.8.10.1 Patch clamping

H4IIE rat hepatoma cells were grown on sterile acid-washed glass cover-slips and co-

transfected with cDNA encoding HCA112 (pCMVSport6-HCA112, purchased from RZPD

gene consortium) and enhanced green fluorescent protein (pEGFP-N1, Clontech) at a ratio

of 4:1. Control cells were transfected with pEGFP-N1 alone. Four hours after transfection

the cover-slip cultures were washed with culture medium and incubated in fresh warm

culture medium for 50 hours prior to patch clamping. Transfected cells were identified by

their expression of EGFP.

Whole cell patch clamping (Hamill et al., 1981) was performed at room temperature by

Dr Grigori Rychkov, using a computer-based patch clamp amplifier (EPC-9, HEKA

Electronics, Germany) and PULSE software (HEKA Electronics). The bath solution

contained 140 mM NaCl, 4 mM CsCl, 10 mM CaCl2, 2 mM MgCl2, 10 mM glucose and

10 mM Hepes, pH 7.4. The internal solution contained 120 mM cesium glutamate, 5 mM

CaCl2, 5 mM MgCl2, 1 mM MgATP, 10 mM EGTA and 10 mM Hepes, pH 7.2. Depletion

of intracellular Ca2+ stores was achieved by addition of 20 µM IP3 (Amersham) to the

internal solution. Patch pipettes were pulled from borosilicate glass and fire-polished.

Pipette resistance varied from 3 to 5 MΩ. Series resistance, for which no compensation

was made, did not exceed 25 MΩ. In order to monitor the membrane conductance, voltage

64

ramps between –138 and +102 mV were applied every 2 s, starting immediately after

achieving the whole-cell configuration. The holding potential was -18 mV throughout. Cell

capacitance was compensated automatically by the EPC9 amplifier. All voltages shown

have been corrected for the liquid junction potential of -18 mV between the bath and

electrode solutions, estimated by JPCalc (Barry, 1994).

2.8.10.2 Measurements of [Ca2+]cyt using Fura-2.

COS-7 cells were grown for 24 hours on acid-washed glass cover-slips in a 6-well tray

before transfection. Forty eight hours prior to imaging, the cells were transfected with

pEGFP-HCA112-HA. The cells were loaded with fura-2/AM by incubating the cover-slips

with Eagle’s salt solution (135 mM NaCl, 1.2 mM CaCl2, 0.8 mM MgSO4, 4 mM KCl,

1mM Na2HPO4, 5.5 mM glucose and 10 mM Hepes, pH 7.4) containing 2 M fura-2/AM

(acetoxymethyl ester) and 0.02% (v/v) pluronic acid for 30 minutes at 37oC. The cover-

slips were then washed 3 times in Eagle’s salt solution, and the fluorescence emission of

fura-2/AM was recorded at 340 and 380 nm by selecting fields of interest under the 20×

objective of a Nikon Diaphot inverted microscope in conjunction with a Sutter filter wheel

(model Lambda 10-C), a Photonic Science intensified CCD camera (ISIS-3/S20) and Axon

Imaging Workbench software (v. 5).

2.9 Proteomics techniques

Protein samples for proteomic analysis were prepared as described in (co-)

immunoprecipitation, above (Section 2.7.2). Four 15 cm dishes of L929 cells stably

transfected with either pEF-HCA112-HA-IRES-puro6 or pEF-IRES-puro6 were lysed in

buffer (1% Triton X-100, 50 mM Tris, pH 7.5, 200 mM NaCl and 1 mM EDTA)

containing protease inhibitors (1:100) (Sigma). After pre-clearing 75 mg of post-nuclear

supernatant with 15 µl protein G conjugated magnetic beads (1 hour at 4ºC with rotation),

anti-HA (1.5 µg, overnight at 4ºC with rotation) and 50 µl protein G conjugated magnetic

beads (1 hour at 4ºC with rotation) were used to co-immunoprecipitate HCA112-HA and

interacting proteins. Following washing of the beads, immunoprecipitates were solubilised

in non-reducing LDS sample buffer (Invitrogen) and subjected to 12% SDS-PAGE using

65

MES running buffer (Invitrogen). Bands were visualised by silver staining, method of the

manufacturer (Invitogen).

Proteomics analysis of silver-stained SDS-PAGE bands was carried out at the Adelaide

Proteomics Centre, University of Adelaide, Australia using standard techniques. Briefly,

bands were excised from the gel, de-stained and digested with 100 ng of trypsin per

sample, according to the “low salt” protocol. One µl of each sample was applied to a 600

µl AnchorChip (Bruker Daltonik GmbH, Bremen, Germany) according to the �-cyano-4-

hydroxycinnamic acid (HCCA) thin-layer method. Data was acquired by liquid

chromatography-ESI mass spectrometry (MS and MS/MS) and MS and MS/MS spectra

were subjected to peak detection using DataAnalysis (version 3.4, Bruker Daltonik

GmbH), imported into BioTools (version 3.1, Bruker Daltonik GmbH) and then the spectra

were submitted to a Mascot database-search engine (version 2.2, Matrix Science) using the

SwissProt mammalian protein database.

66

Table 2-2. Oligonucleotides used in this study.

Name Sequence (5’ � 3’) ApplicationHCA112 F1 ATTTTTCACAAAGGCGGAAACAGA PCR HCA112 R1 TAGGACGGTGAAGGATGGAGAGAC PCR HCA112 human F1

GCCCTCAAACTTTGGAATGA PCR

HCA112 Human R1

TCAGCATGTCCATGAAGGAG PCR

HCA112 F3 TTAGCCATGTATTCCGAAGG PCR HCA112 R4 GTGGACCATCCATATGAAGA PCR s27a F CAAGGATAAGGAAGGAATTCCTCCTG PCR s27a R CCAGCACCACATTCATCAGAAGG PCR HCA112 140-191 F

TATACATATGGCTAGCATGATCGTTATTGGGTCTCGTG Clon.

HCA112 140-191 R

GGTGGTGGTGCTCGAGGCTTGTGTAGTATATGCACAAGG Clon.

HCA112-HA R GAATTCCTAAGCCTAGTCTGGGACGTCGTATGGGTAGATCACAGCTGCACCCAGCAG

Clon.

HCA112-cmyc R

GAATTCCTACAGATCTTCTTCAGAAATAAGTTTTTGTTCGATCACAGCTGCACCCAGCAG

Clon.

HCA112 P1 GGATCCCACCATGTCCACAGACATGGAGACTGCAG Clon. HCA112 EGFP-N1 F

CTCGAGCACCATGTCCACAGACATGGAGACT Clon.

HCA112-EGFP-C1 F

CTCGAGCTATGTCCACAGACATGGAGACT Clon.

HCA112-EGFP-C1 R

GGATCCCTAGATCACAGCTGCACCC Clon.

HCA112del2-56 F

CTCGAGCACCATGAGCAGCAGAGTGCTGGT Clon.

HCA112del2-60 F

CTCGAGCACCATGCTGGTGGCCTCCTG Clon.

HCA112del 223-244 R

GGATCCCTAGATGTAGACACATACAGGGGT Clon.

LL238AA R GGAATCCCTAGATCACAGCTGCACCCGCCGCTTTCTT Clon. 222mut IF GTATGTGTCGCCATCGCGGCAGCAGCTGCCACAAAGGCGGAAAC Clon. 222mut IR GTTTCCGCCTTTGTGGCAGCTGCTGCCGCGATGGCGACACATAC Clon. LR8-Flag F CTCGAGCACCATGGTCCAGAGCACAGTGACTG Clon. LR8-Flag R GGATCCTCACTTGTCATCGTCGTCCTTGTAGTCCAGGATAGCAG

GGATCTTCTC Clon.

PEF F CCACTCCCAGTTCAATTACAGCTCTTAAGGCTAG Seq. PEF R GGGAGAGGGGCGGAATTGGGC Seq. M13 F GTAAAACGACGGCCAGT Seq. M13 R CAGGAAACAGCTATGAC Seq. SP6 ATTTAGGTGACACTA Seq. T7 promoter TAATACGACTCACTATAGGG Seq. T7 terminator GCTAGTTATTGCTCAGCGG Seq.

Abbreviations: Seq., sequencing; Clon., cloning; PCR, polymerase chain reaction.

67

Table 2-3. Primary antibodies used throughout this study.

Antigen Isotype Dilution Manufacturer Calnexin Mouse IgG1 IF 1:200 abcam Catalase Rabbit IgG1 IF 1:100; WB 1:500 Calbiochem Caveolin-1 (Clone C060)

Mouse IgM WB 0.5 µg/ml BD Transduction Laboratories

CD36 (Clone Mo25)

Mouse IgG1 WB 1:2,000 (Tandon et al., 1989)

CD36 (Clone UA009)

Mouse IgG1 IF 1 µg/ml (Zhang et al., 2003)

c-myc Rabbit IgG WB 1:500 Santa Cruz Cytochrome c Mouse IgG1 IF 1:100; WB 1:500 Santa Cruz EEA1 Mouse IgG1 IF 1:50 BD Transduction Laboratories FLAG Mouse IgG1 WB 1 µg/ml; IF 1 µg/ml Sigma GFP Goat IgG - biotin

conjugated WB 1:5,000 Rockland

GM130 Mouse IgG1 IF 1:50 BD Transduction Laboratories HA Mouse IgG2a WB 1:500; IF 1:150 Santa Cruz HA (Clone 12CA5)

Mouse IgG2b IF 1.5 ug/ml

HA (Clone 12CA5)

Mouse IgG2b – fab IF 65 ug/ml fab fragments produced by MabSA

HCA112140-191 Rat Ig IF, FACS, WB 1:50 Wendy Parker, University of Adelaide

Human MHC-I (Clone W6/32)

Mouse IgG2a IF 1:2 (hybridoma supernatant)

(Barnstable et al., 1978)

LAMP-1 Sheep IgG WB 1:500 * LAMP-1 (Clone H4A3)

Mouse IgG1 IF 1:100 **

Transferrin Receptor

IgG1 WB 1 µg/ml; IF 5 µg/ml Zymed Laboratories

�-actin (Clone AC-15)

Mouse IgG1 WB 1:10,000 Sigma

* - A kind gift of Dr Peter Meikle, Children, Youth and Women’s Health Service,

Adelaide, Australia.

** - Monoclonal antibody developed by August, J.T. and J.E.K. Hildreth, Department of

Pharmacology and Molecular Sciences, John Hopkins University School of Medicine,

Baltimore, USA was obtained from the Developmental Studies Hybridoma Bank,

Department of Biological Sciences, The University of Iowa, Iowa City, USA.

Abbreviations: IF, immunofluorescence; FACS, fluorescence activated cell scanning; WB,

Western blot.

68

Table 2-4. Conjugated secondary antibodies used throughout this study

Conjugate Specificity Host Dilution Manufacturer Biotin Sheep IgG (H+L) Donkey WB 1:500 Rockland Cy3 Mouse IgG (H+L) Donkey IF 3.75 µg/ml Jackson ImmunoResearch Cy3 Rabbit Ig Donkey IF 1:150 Jackson ImmunoResearch FITC Biotin N/A Strept-avidin IF 5 µg/ml Rockland FITC Mouse Ig Goat IF, FACS 1:100 BD Biosciences FITC Rat IgG (H+L) Donkey IF 30 µg/ml Jackson ImmunoResearch HRP Biotin N/A Strept-avidin WB 1:10,000 Amersham Biosciences HRP Mouse Ig Sheep WB 1:10,000 Amersham BiosciencesHRP Rabbit IgG (H+L) Donkey ELISA 1:10,000 Rockland HRP Rabbit Ig Donkey WB 1:40,000 Jackson ImmunoResearch APC Mouse Ig Goat FACS 1:100 BD Pharmingen

Abbreviations: HRP, horseradish peroxidise; FITC, fluorescein isothiocynate; APC,

allophycocyanin; Cy, cyanine dyes; H+L, heavy and light Ig chains, IF,

immunofluorescence; FACS, fluorescence activated cell scanning; WB, Western blot; N/A,

not applicable.

69

Chapter 3: Bioinformatics Analysis of HCA112 and its Gene Product

3.1 Introduction

Ivell and colleagues performed microarray analysis of rat testis cDNA, with the objective

of detecting genes that are expressed specifically by Leydig cells, and using this

information to discover genes that might play an important role in Leydig cell

differentiation, steroidogenesis and male fertility (unpublished, 2004). RNA was extracted

from the testes of control rats and rats treated with the Leydig cell-specific cytotoxin

ethane dimethanesulphonate (EDS). Approximately 200 genes were identified for which

levels of transcripts differed significantly between the control and EDS-treated groups. Of

these, approximately half were known to be involved in Leydig cell functions, with the

remainder undescribed previously, with respect to expression in the testis. Quantitative

reverse transcriptase polymerase chain reaction (qRT-PCR) revealed that transcripts from

ten of the previously undescribed genes were significantly less abundant in the EDS-

treated group compared to the control (p<0.00005), suggesting that expression of these

genes in the testis was confined to Leydig cells. These ten genes were chosen for further

characterisation, using the mouse Leydig cell line MA-10 to examine transcription in

response to a variety of factors known to activate or inhibit intracellular signalling

pathways.

A BLAST search of the ten expressed sequence tags revealed that one had 100%

similarity with sequences in a rat gene with the accession number (acc#) NM_001039008,

and also with an orthologous mouse gene with acc# BC010831. This gene was chosen for

further characterisation for the following reasons. Firstly, preliminary data showed that

transcription of the gene was up-regulated by the cytokine tumour necrosis factor-alpha

(TNF�). This was an unusual finding, because in Leydig cells, TNF� is thought to be an

inhibitory cytokine as it inhibits testosterone production and the 8-bromo-cAMP induced

increase in P450scc and P450c17 mRNA and protein levels (Xiong and Hales, 1993).

Secondly, the deduced amino acid sequence of BC010831 had a predicted protein

conformation similar to members of the tetraspanin family. Members of this family

70

characteristically have four-transmembrane (TM) domains and many play important roles

in regulating cell functions by facilitating protein-protein interactions (reviewed by (Levy

and Shoham, 2005). This led to the hypothesis that the protein with accession number

(acc#) NM_001039008/BC010831 might play an important regulatory role in Leydig cells,

for instance in regulating steroidogenesis and that it might, therefore, be a target for

contraceptive drugs or drugs to otherwise modify male fertility.

3.2 Bioinformatics

Orthologues of the rat gene with acc# NM_001039008 and its protein product have been

described only recently in mice and humans (Nakajima et al., 2002; Wang et al., 2002). To

assist in identifying important clues as to the function of the molecule, a bioinformatics

study was performed on the mouse orthologue of this gene and its protein product.

3.2.1 Orthologues of HCA112 and its gene product

In the human genome, HCA112 is encoded by 7 exons that span approximately 4 kb of

DNA on chromosome 7q36. The gene encoding the mouse orthologue has a similar

arrangement of coding sequence and is located on a syntenic chromosome (chromosome

6). The orthologous genes and the proteins encoded by them have been assigned several

names; including hepatocellular carcinoma-associated antigen 112 (HCA112/HCA112) in

humans (Wang et al., 2002), GS188/KEG2/GS188/KEG2 in mice (Nakajima et al., 2002)

and TM protein 176A (TM176A/TM176A) in the NCBI database (on the basis of the

predicted four TM topology of the protein). In this thesis, HCA112 will be used to identify

the gene and HCA112 to identify the protein of all orthologues, irrespective of species.

BLAST analysis of the nucleotide sequence of HCA112 cDNA, and the deduced amino

acid sequence of HCA112, predicts the presence of orthologous proteins in mammals and

possible homologues in invertebrates. The open reading frames (ORFs) of cDNAs (and

predicted cDNAs) encode proteins ranging from 197 to 352 amino acids in length. The

ORFs of the genes have start sites similar to the Kozak consensus sequence (Kozak, 1991)

(CCACCAUG), as shown for HCA112 in mice (Figure 3-1). HCA112 in mouse, rat,

human, dog, bovine and horse show the greatest homology, with proteins ranging from 232

to 244 amino acids in length. Alignment of the mouse and rat amino acid sequences

71

indicates about 85% identity, while comparison of the mouse and human proteins shows

about 53% identity. Amino acid identity is greater within the predicted cytoplasmic N-

terminal region and the TM domains, with 95% identity between the mouse and rat

proteins, and ~72% identity between the mouse and human proteins in these regions (see

Figure 3-2).

Bioinformatics methods were used to examine the predicted efficiency of translation of

the ORF of HCA112 mRNA, and thus the likelihood of the ORF being genuine. Frequency

of optimal pairs (Fop) and codon bias index (CBI) scores were calculated using the

program codonw1. Fop and CBI scores are good indicators of mRNA translation. A CBI or

a Fop score of one represents most common (preferred) codon usage, with a score of zero

indicating totally random choice of codon usage and thus the mRNA is unlikely to to be

translated (Ikemura, 1981; Bennetzen and Hall, 1982). Fop and CBI scores for the mouse,

rat and human homologues of HCA112 are shown in Table 3-1. The Fop and CBI scores of

the HCA112 ORFs are similar to those of cDNAs encoding a group of better studied mouse

proteins, which are inturn significantly different to the scores of the +1 and +2 reading

frames. This indicates that the putative ORF of each of the orthologues is translatable and

subject to natural selection for optimal codon usage.

3.2.2 Similarity of HCA112 with other proteins

Searches using BLAST show that HCA112 has greatest similarity (approximately 30%

amino acid identity) with a human protein LR8, and with orthologues of this protein in

mice (Clast1) and rats (TORID) (see Figure 3-3, A). This cluster of orthologues is

catalogued as THEM176B in the NCBI database. Interestingly, in all species that have

been examined (human, mouse, rat, chimp, rhesus monkey, cow, dog and horse), the

orthologues of HCA112 and LR8 are encoded on the same chromosome, lying close to

each other and in opposite orientations, with the 5’ ends facing each other (see Figure 3-3,

B). The proximity of the two similar genes suggests that they may have arisen from gene

duplication in an ancient common ancestor. It is possible that HCA112 and LR8 may have

similar functions and that they might be regulated by similar factors.

1 http://bioweb.pasteur.fr/cgi-bin/seqanal/codonw.pl

72

The functions of LR8 and its orthologues remain unknown. However, Lurton and

colleagues have shown that LR8 is differentially expressed in subsets of human lung

fibroblasts (Lurton et al., 1999). In mice, Clast1 mRNA is expressed in various adult

mouse tissues, with highest expression observed in the lung, liver, kidney and colon. It has

also been found that Clast1 expression is up-regulated following stimulation of splenic B

cells via CD40 ligand. The function of Clast1 is non-essential, as Clast1 gene knock-out

(GKO) mice grow to adulthood. Nevertheless, approximately 65% of Clast1 GKO mice

have severe ataxia. The protein is thought, therefore, to be required for development of

granule cells of the cerebellum (Maeda et al., 2006). In rats, TORID (tolerance related and

induced transcript) is thought to be involved in the control of dendritic cell maturation

(Louvet et al., 2005). For convenience in this thesis, the LR8/TORID/Clast1 orthologues

will be referred to collectively as “LR8”.

3.2.3 Analysis of the amino acid sequence of mouse HCA112

Examination of the amino acid sequence of HCA112 reveals that the protein has four

regions of strongly hydrophobic amino acids. Using programs such as TMHMM (Krogh et

al., 2001) (see Figure 3-4), PSORT II server (Nakai and Horton, 1999) and TMpred

(Hofmann and Stoffel, 1993) to construct hydrophobicity plots, the topology of HCA112

has been predicted. The topology of HCA112 predicted with strongest preference was one

of four TM domains, with cytoplasmic N- and C-termini, a small intracellular loop plus

one small and one larger extracellular loop. HCA112 was also predicted as having four TM

domains with extracellular N-and C-termini, and also with 5 TM domains with an

intracellular N-terminal tail; however these topologies were not consistently predicted with

high probability by the different programs. The homologous LR8 proteins have been

predicted also as having 4 TM domains with cytoplasmic tails (Lurton et al., 1999; Louvet

et al., 2005), and hence it is likely that HCA112 also has this topology.

No putative signal sequences or motifs (such as RNA/DNA binding motifs) were

identified within the amino acid sequence of HCA112 by the prediction server PSORT II

(Nakai and Horton, 1999), giving no clues as to the subcellular localisation or function of

HCA112. However, the server does predict that human, but not mouse or rat, HCA112 is

myristylated at the N-terminus, indicating that it might be anchored to the membrane.

73

Mouse and rat HCA112 have predicted start sites 4 amino acids before that of human

HCA112, which may be why myristylation is not predicted for HCA112 in these species.

Two putative phosphorylation sites (THID, Casein kinase II and SSR, Protein kinase C)

have been predicted within the cytoplasmic N-terminal tail of HCA112 by the server

NetPhos 2.0 (Blom et al., 1999), which are conserved between HCA112 orthologues in

humans, rats and mice (Figure 3-5). The server NetNGlyc 1.02 (Gupta et al., 2004, in

preparation) predicts no N-glycosylation sites in mouse, rat or human HCA112, however

the server OGPET v1.03 predicts an O-glycoslyation site in the large extracellular loop of

mouse (but not rat or human) HCA112 (Thr174). Two conserved cysteine residues are also

found within the large putative extracellular loop and these may allow formation of a

disulphide bridge. It is noteworthy that HCA112 is rich in leucine residues (~14% of the

amino acid residues), particularly in the N-terminal 50 amino acids of the protein (~25%

leucine). The functional significance of this is not known because they do not appear to

conform to the leucine rich repeat (LRR) consensus motif, which is involved in mediating

protein-protein interactions (Kobe and Deisenhofer, 1994; Kobe and Kajava, 2001). A

cartoon depicting the major features of the HCA112 orthologues is shown in Figure 3-6.

3.2.4 HCA112 is not a member of the tetraspanin family

Proteins of the tetraspanin family consist of four TM domains, giving rise to one small and

one large extracellular loop. Superficially, this topology resembles the predicted topology

of HCA112. However, tetraspanin proteins contain several canonical conserved amino acid

residues and motifs. Firstly, in tetraspanins there is a conserved CCG motif, plus 2-6

additional cysteine residues in the large extracellular domain. These allow the formation of

disulphide bridges. Furthermore, the TM domains of tetraspanins contain conserved polar

residues (TM1 - N, TM3 – E, TM4 – E), and the intracellular domains contain several

conserved cysteine residues (Levy and Shoham, 2005). The only common feature between

these conserved motifs in tetraspanins and the amino acid sequence of HCA112 is the

presence of two conserved cysteine residues in the predicted large extracellular domain of

the latter. It appears unlikely, therefore, that HCA112 is part of the tetraspanin family.

2 http://www.cbs.dtu.dk/services/NetNGlyc/ 3 http://ogpet.utep.edu/OGPET/index.php

74

3.2.5 HCA112 is a member of the MS4A family

The predicted four TM topologies of HCA112 and LR8 are similar to those of the newly

described membrane-spanning 4-domains, subfamily A (MS4A) family of proteins. This

family is characterised by structural similarity (small, four TM domain proteins), and

regions of amino acid identity, especially in the first and second TM domains. The genes

encoding these proteins also have similar intron / exon organisation and they are clustered

on chromosome 11q12 in humans (Ishibashi et al., 2001; Liang and Tedder, 2001). This is

a relatively large family, with the B-cell-specific antigen CD20, the high affinity IgE Fc

receptor � chain (Fc�RI�), and a hematopoietic-cell-specific protein HTm4 as founding

members. The family contains at least 23 other proteins in humans, mice and rats. The

functions of members of the MS4A family of proteins remain to be fully elucidated.

However, at least some appear to have signalling or ion channel functions (Ishibashi et al.,

2001; Liang and Tedder, 2001).

In addition to similarities between the amino acid sequences and topologies of HCA112

and LR8, the genes encoding these proteins have similarities in the organisation of introns

and exons. Furthermore, these similarities are shared with other members of the MS4A

family and there is 10-20% amino acid identity with the 12 known MS4A proteins for

which data is available (Louvet et al., 2005). The MS4A family members share several

amino acid motifs within the TM domains. The motifs VLGAIQIL (Ishibashi et al., 2001),

LGAXQI and LSLG (Liang and Tedder, 2001) have been identified within the first TM

domain. The second TM domain contains the motifs GYPFWG (Ishibashi et al., 2001;

Liang and Tedder, 2001) and FIISGSLS, while the third TM domain contains the motifs

SLX2NX2 and SX3AX2G (Liang and Tedder, 2001). Similarities between amino acids

present in HCA112 and these MS4A motifs are shown in Figure 3-7. Another area of

similarity is the presence of a tyrosine- and proline-rich area in the predicted second

extracellular loop in the MS4A family proteins (Ishibashi et al., 2001) and also in HCA112

(see Figure 3-7). Because of these similarities, HCA112 and LR8 should be considered

MS4A family members, despite being located on a different chromosome to the rest of the

family (Ishibashi et al., 2001; Liang and Tedder, 2001).

75

3.3 Discussion

These investigations identify HCA112 orthologues in a number of other mammals and

possibly in invertebrates. LR8 has been identified as a homologue of HCA112. HCA112

and LR8 share approximately 30% amino acid sequence identity and the HCA112 and LR8

orthologues, in all species that have been examined, are located together on the same

chromosome, with opposite orientation and their 5’ ends facing each other. This proximity

of two similar genes suggests that HCA112 and LR8 may have similar or complementary

cellular functions that are co-regulated. However, the functions of neither protein are

known. LR8 is believed to play a role in the development of granule cells of the

cerebellum in mice (Maeda et al., 2006) and may also be involved in the control of

dendritic cell maturation in rats (Louvet et al., 2005). HCA112 and LR8 have a putative

four TM topology, which is similar to that of the MS4A family of proteins. The two

proteins also have amino acid sequence identity with, and contain amino acid motifs

present in the MS4A family of proteins, suggesting that HCA112 and LR8 belong to the

MS4A family of four TM domain proteins and not to the tetraspanin family.

76

gtgagctct ggccttccctgctgcacacttccccggtcccagaggaatcagacg 55 atgtccacagacatggagactgcagtcgtcggcaaggtggaccct M S T D M E T A V V G K V D P 100 gaggctccacaacccacccacattgatgtgcacatccaccaggag E A P Q P T H I D V H I H Q E 145 tctgctctggccaaacttctgctggccggatgctcattgctaagg S A L A K L L L A G C S L L R 190 attccagcatccgcttccacccagagccagggcagcagcagagtg I P A S A S T Q S Q G S S R V 235 ctggtggcctcctgggtggtgcagactgtgctgggggctctgagt L V A S W V V Q T V L G A L S 280 gtggttctgggtggaaccctctacataggccattatttagccatg V V L G G T L Y I G H Y L A M 325 tattccgaaggcgcccccttctggactgggatcgtggctatgctg Y S E G A P F W T G I V A M L 370 gctggagctgttgccttccttcacaagaaacggggtggtacctgc A G A V A F L H K K R G G T C 415 tgggccctgatgaggacccttcttgtgctggcaagtttctgcacc W A L M R T L L V L A S F C T 460 gctgtggctgccatcgttattgggtctcgtgagttgaatttttac A V A A I V I G S R E L N F Y 505 tggtattttctcggagatgatgtctgtcaaagagactcttcatat W Y F L G D D V C Q R D S S Y 550 ggatggtccaccatgcctagaaccactccagttcccgaagaagct G W S T M P R T T P V P E E A 595 gataggattgccttgtgcatatactacacaagcatgctaaagacc D R I A L C I Y Y T S M L K T 640 ctgctcatgagcctccaagctatgctcttgggtatctgggtgctg L L M S L Q A M L L G I W V L 685 ctgctcctggcttctctcacccctgtatgtgtctacatctggaaa L L L A S L T P V C V Y I W K 730 agatttttcacaaaggcggaaacagaggagaagaaactgctgggt R F F T K A E T E E K K L L G 775 gcagctgtgatctagcctttcctcttgctccgggcgtccctccta A A V I * ctgaagcctgaaagaagaatcaggcaggactaagaagaccctccc ccactagcagggccatggccactgcctggttctgcccagcaccac agcagctctcagcagcacttgcttgtctctccatccttcaccgtc ctatatccctcctcaggcagcaacttgataataaactctcctgtt attgctaaaaaaaaaaaaaaaaaaa

Figure 3-1. Mouse HCA112 sequence.

Nucleotide sequence of mouse HCA112 cDNA and the deduced amino acid sequence (shown by one letter code) of HCA112. The translational start site (italics) of the HCA112 transcript forms an atypical Kozak consensus sequence (bold). Amino acid residues are designated according to the single letter code; see legend to Table 1-1.

77

Mouse MSTDMETAVVGKVDPEAPQPTHIDVHIHQESALAKLLLAGCSLLRIP----ASASTQSQG

Rat MSTDMGTADVGEVDPEAPQPTNIEVHIHQESVLAKLLLAGCSFLRVP----ASASTQSQG

Human ----MGTADSDEMAPEAPQHTHIDVHIHQESALAKLLLTCCSALRP-------RATQARG

Dog ----METVDSGEVAPGAPQPMHINVHIHQESALAKLLTSGCSLLRS---HFPGATFQTWS

Horse ----MGMVDGGVVAPGGPQPTCIDVHIHQESALGKLLLSGCSLLRP--------SSQTLG

Bovine ----METVDCGEAAPRAPQPASIQVHFHHESGLAKLLLGGCSLLQPLLLPRPRATSRALG

* . . * .** *:**:*:** *.*** ** *: : :: .

Mouse SSRVLVASWVVQTVLGALSVVLGGTLYIGHYLAMYSEGAPFWTGIVAMLAGAVAFLHKKR

Rat SSRVLVASWVVQIVLGILSVVLGGILYICHYLAMNTQGAPFWTGIVAMLAGAVAFLQKKR

Human SSRLLVASWVMQIVLGILSAVLGGFFYIRDYTLLVTSGAAIWTGAVAVLAGAAAFIYEKR

Dog RSRLLLASWVIQIVLGVSSGVLGGFLYIFYCSTLCSSGAAIWTGAVASLAGAVAFIHQKR

Horse SRRLLVASWAVQIVLGVLSGVLGGFLHFFYYSPVRNSGAAIWTGVVAVLAGTIAFIYEKR

Bovine RHRLLATSWVMQIVLGLLSGVLGGFLYIFSSTTLRNSGAPIWTGAVAVLAGAVAFIYEKR

*:* :**.:* *** * **** ::: : ..**.:*** ** ***: **: :**

Mouse GGTCWALMRTLLVLASFCTAVAAIVIGSRELNFYWYFLGDDVCQRDSSY-GWSTMPRTTP

Rat GGTCWALMRILLVLASFCTAVAAIVIGSREFNNYWYYLRDDVCKSDTSY-RWSTMPSITP

Human GGTYWALLRTLLALAAFSTAIAALKLWNEDF-RYGYSYYNSACRISSSS-DWNTPAP-TQ

Dog GGICWALLRILLALAAFSTATAAIVIGASNFYRHRFYLRDFICDVSSKAWSWAPLSPSTP

Horse GGFYLAQLRTLLALAAFSTATAAVVIGARNFYEYRFE-SEDICDISPSG-SWPTSAPHTP

Bovine GGIYWALLRTLLALAAFSTATAATIIGAGRFYEYHFIFYKGICNVSP---SWRPTGAPTL

** * :* **.**:*.** ** : : : : . * .. * . *

Mouse VPEEADRIALCIYYTSMLKTLLMSLQAMLLGIWVLLLLASLTPVCVYIWKRFFTKAETEE

Rat VPEEANRIGLCKYYTSMLKTLLISLQAMLLGVWVLLLLASLIPVCVYLWKRFFTKAET-E

Human SPEEVRRLHLCTSFMDMLKALFRTLQAMLLGVWILLLLASLAPLWLYCWRMFPTKGKRDQ

Dog SPEEATRLHLCLSYLSMLEALFISFQVMLLGIWVLLLLASLVPLCLFCWSRSRHKKID-Q

Horse SPEEVRRLHLCLSYLNMLKALFISIRAMLLGIWILLLLASLAPLCLYCWRRLRPKE---W

Bovine SP-DLERLQQCTAYVNMLKALFISINAMLLGVWVLLLLASLLPLCLCCWRRYRRKEKR-D

* : *: * : .**::*: ::..****:*:******* *: : * *

Mouse KKLLGAAVI-

Rat KKLLGAAVI-

Human KEMLEVSG--

Dog KKLLEANGI-

Horse WPSPESPHP-

Bovine LPLEETVRSE

Figure 3-2. Sequence alignment of HCA112 and its orthologues.

CLUSTAL X (1.81) multiple sequence alignment of human HCA112 (AAF68667) and its orthologues in mouse (BC010831), rat (NM_001039008), dog (XM_532758), horse (XP_001495020) and bovine (AY243098). “*” indicates amino acid identity, “:” indicates strong group conservation and “.” indicates weak group conservation.

78

Table 3-1. Codon bias index (CBI) and frequency of optimal pairs (Fop) scores for the

mouse, rat and human orthologues of HCA112 cDNA.

A CBI or a Fop score of one represents most common (preferred) codon usage, with a score of zero indicating totally random choice. Nucleotide sequences of a group of known mouse genes were subjected to the program and these showed significant differences between the correctly used ORFs (row 4) and of the +1 and +2 reading frames (row 5). The CBI and Fop scores of mouse, rat and human HCA112 cDNA are similar to those shown in row 4, which suggest that the proteins are translatable.

Species CBI Fop

Mouse HCA112 0.143 0.487

Rat HCA112 0.161 0.501

Human HCA112 0.185 0.516

Mouse random ORF 0.226 ± 0.129 0.509 ± 0.091

Mouse random ORF +1/+2 -0.011 ± 0.096 0.319 ± 0.060

79

AHCA112 MSTDMETAVVGKVDPEAPQPTHIDVHIHQESALAKLLLAGCSLLRIPASASTQSQGSSRV

LR8/Clast1 MVQSTVTVNGVKVASTHPQSAHISIHIHQKSALEQLLGAVGSLKKFLSWPQARVHYG---

* . *. ** . **.:**.:****:*** :** * ** :: : ..:: : .

HCA112 LVASWVVQTVLGALSVVLGGTLYIGHYLAMYSEGAPFWTGIVAMLAGAVAFLHKKRGGTC

LR8/Clast1 QLSLGVTQILLGLVSCALGVCLYFGPWTELCAFGCAFWSGSVAILAGVGTIVHEKRQGKL

:: *.* :** :* .** **:* : : : *..**:* **:***. :::*:** *.

HCA112 WALMRTLLVLASFCTAVAAIVIGS----RELNFYWYFLGDDVCQRDSSYGWSTMPRTTPV

LR8/Clast1 SGQVSCLLLLACIATAAAATVLGVNSLIRQTSVPYYVEIFSTCNPLQSSMDPGYGTVRYS

. : **:**.:.**.** *:* *: .. :*. ..*: .* . .

HCA112 PEEADRIALCIYYTSMLKTLLMSLQAMLLGIWVLLLLASLTPVCVYIWKRFFT-----KA

LR8/Clast1 DDSDWKTERCREYLNMMMNLFLAFCIMLTVVCILEIVVSVASLGLSLRSMYGRSSQALNE

:. : * * .*: .*:::: ** : :* ::.*::.: : : . : :

HCA112 ETEEKKLLG-------------AAVI

LR8/Clast1 EESERKLLDGHPAPASPAKEKISAIL

* .*:***. :*::

B

HCA112 LR8HCA112 LR8

Figure 3-3. Alignment comparison of HCA112 and LR8, and chromosomal locations

of the orthologous genes in rat, mouse and human.

A. CLUSTAL X (1.81) multiple sequence alignment of the amino acid sequences of mouse HCA112 and mouse LR8/TORID/Clast1. “*” indicates amino acid identity, “:” indicates strong group conservation and “.” indicates weak group conservation. Amino acid residues are designated according to the single letter code; see legend to Table 1-1.

B. Gene map showing the location and direction of coding of the orthologous HCA112 andLR8 genes in rat, mouse and human (NCBI Map Viewer).

80

Figure 3-4. Hydrophobicity plot of the amino acid sequence of HCA112.

TMHMM hydrophobicity plot of the mouse HCA112 gene product, showing the presence of four TM regions and N- and C-terminal cytoplasmic tails. Hydrophobicity plots of the predicted rat and human proteins are very similar (not shown) (Krogh, Larsson et al. 2001).

81

Figure 3-5. Conserved features of human, mouse and rat HCA112.

CLUSTAL X (1.81) multiple sequence alignment of the amino acid sequences of mouse, rat and human HCA112, showing the putative TM domains predicted from the TMHMM hydrophobicity plot (underlined). Shown also are putative phosphorylation sites (boxes); THID, Casein kinase II phosphorylation; SSR Protein kinase C. ‘*’ indicates amino acid identity, ‘:’ indicates strong group conservation and ‘.’ indicates weak group conservation. Amino acid residues are designated according to the single letter code; see legend to Table 1-1.

Rat MSTDMGTADVGEVDPEAPQPTNIEVHIHQESVLAKLLLAGCSFLRVPASASTQSQGSSRVMouse MSTDMETAVVGKVDPEAPQPTHIDVHIHQESALAKLLLAGCSLLRIPASASTQSQGSSRVHuman ----MGTADSDEMAPEAPQHTHIDVHIHQESALAKLLLTCCSALRP---RATQARGSSRL

* ** .:: ***** *:*:*******.******: ** ** :**::****:N- terminal tail

Rat LVASWVVQIVLGILSVVLGGILYICHYLAMNTQGAPFWTGIVAMLAGAVAFLQKKRGGTCMouse LVASWVVQTVLGALSVVLGGTLYIGHYLAMYSEGAPFWTGIVAMLAGAVAFLHKKRGGTCHuman LVASWVMQIVLGILSAVLGGFFYIRDYTLLVTSGAAIWTGAVAVLAGAAAFIYEKRGGTY

******:* *** **.**** :** .* : :.**.:*** **:****.**: :***** TM1 EC1 TM2 IC

Rat WALMRILLVLASFCTAVAAIVIGSREFNNYWYYLRDDVCKSDTSYRWSTMPSITPVPEEAMouse WALMRTLLVLASFCTAVAAIVIGSRELNFYWYFLGDDVCQRDSSYGWSTMPRTTPVPEEAHuman WALLRTLLALAAFSTAIAALKLWNEDFR-YGYSYYNSACRISSSSDWNT-PAPTQSPEEV

***:* **.**:*.**:**: : ..::. * * :..*: .:* *.* * * ***.TM3 EC2

Rat NRIGLCKYYTSMLKTLLISLQAMLLGVWVLLLLASLIPVCVYLWKRFFTKAET-EKKLLGMouse DRIALCIYYTSMLKTLLMSLQAMLLGIWVLLLLASLTPVCVYIWKRFFTKAETEEKKLLGHuman RRLHLCTSFMDMLKALFRTLQAMLLGVWILLLLASLAPLWLYCWRMFPTKGKRDQKEMLE

*: ** : .***:*: :*******:*:******* *: :* *: * **.: :*::* TM4 C-terminal tail

Rat AAVIMouse AAVIHuman VSGI

.: *

82

Figure 3-6. Cartoon showing the putative membrane topology of the HCA112 protein.

The protein has four predicted TM domains, with putative intracellular N- and C-termini and two putative extracellular loops. Two conserved cysteine residues are found in the second of these loops. The putative intracellular N-terminal domain contains two predicted phosphorylation sites and these are conserved between species.

83

A Rat MSTDMGTADVGEVDPEAPQPTNIEVHIHQESVLAKLLLAGCSFLRVPASASTQSQGSSRVMouse MSTDMETAVVGKVDPEAPQPTHIDVHIHQESALAKLLLAGCSLLRIPASASTQSQGSSRVHuman ----MGTADSDEMAPEAPQHTHIDVHIHQESALAKLLLTCCSALRP---RATQARGSSRL

* ** .:: ***** *:*:*******.******: ** ** :**::****:N-tail

Rat LVASWVVQIVLGILSVVLGGILYICHYLAMNTQGAPFWTGIVAMLAGAVAFLQKKRGGTCMouse LVASWVVQTVLGALSVVLGGTLYIGHYLAMYSEGAPFWTGIVAMLAGAVAFLHKKRGGTCHuman LVASWVMQIVLGILSAVLGGFFYIRDYTLLVTSGAAIWTGAVAVLAGAAAFIYEKRGGTY

******:* *** **.**** :** .* : :.**.:*** **:****.**: :***** TM1 TM2

Rat WALMRILLVLASFCTAVAAIVIGSREFNNYWYYLRDDVCKSDTSYRWSTMPSITPVPEEAMouse WALMRTLLVLASFCTAVAAIVIGSRELNFYWYFLGDDVCQRDSSYGWSTMPRTTPVPEEAHuman WALLRTLLALAAFSTAIAALKLWNEDFR-YGYSYYNSACRISSSSDWNT-PAPTQSPEEV

***:* **.**:*.**:**: : ..::. * * :..*: .:* *.* * * ***.TM3

Rat NRIGLCKYYTSMLKTLLISLQAMLLGVWVLLLLASLIPVCVYLWKRFFTKAET-EKKLLGMouse DRIALCIYYTSMLKTLLMSLQAMLLGIWVLLLLASLTPVCVYIWKRFFTKAETEEKKLLGHuman RRLHLCTSFMDMLKALFRTLQAMLLGVWILLLLASLAPLWLYCWRMFPTKGKRDQKEMLE

*: ** : .***:*: :*******:*:******* *: :* *: * **.: :*::* TM4 C-tail

Rat AAVIMouse AAVIHuman VSGI

VLGAIQIL GYPFWGLSLG

Y + P rich ~15%

B

34

SX3AX6G SX2AX3G

3SX3AX2G

4SL 3SLX2NX2

2FIISGSLS

2GAPFWTG2GYPFWG

1LSVVLG1LSLG

1QTVLGIL1LGAXQI

1QTVLGIL1VLGAIQIL

TMHCA112 (mouse)TMMotif - M4SA

34

SX3AX6G SX2AX3G

3SX3AX2G

4SL 3SLX2NX2

2FIISGSLS

2GAPFWTG2GYPFWG

1LSVVLG1LSLG

1QTVLGIL1LGAXQI

1QTVLGIL1VLGAIQIL

TMHCA112 (mouse)TMMotif - M4SA

Figure 3-7. HCA112 is a member of the MS4A family.

A. Alignment of the amino acid sequences of rat, mouse and human HCA112 showing TM domains (underlined), and the positions of amino acids and amino acid motifs that are similar to those found in the MS4A proteins (Ishibashi, Suzuki et al. 2001; Liang and Tedder 2001) (boxes and highlighted). . “*” indicates amino acid identity between the HCA112 orthologues compared, “:” indicates strong group conservation and “.” indicates weak group conservation.

B. Similarity of amino acid sequences in mouse HCA112 with the common motifs that have been identified in MS4A family proteins (Ishibashi, Suzuki et al. 2001; Liang and Tedder 2001). TM indicates the TM domain where the motif is found in MS4A proteins (column 2) and HCA112 (column 4).

Amino acid residues are designated according to the single letter code; see legend to Table 1-1.