The role of the Brambell receptor (FcRB) in liver: protection of endocytosed immunoglobulin G (IgG) from catabolism in hepatocytes rather than transport of IgG to bile

Reflecting the central role of immunoglobulin G (IgG) in the immune system, mammals have evolved complex mechanisms to transmit immunity from mother to young and to protect IgG from catabolism. 1 It was recognized early on that these two processes of IgG transport and IgG protection involve receptors that share many features, e.g. IgG saturation and transendosomal transport, 1–5 acid-enhanced binding, 6–8 and, much later, a shared site on the Fc for binding. 9 Although long suspected to be identical, 1,3–6 it was only recently conclusively demonstrated that the receptor protein that mediates the functions of the IgG neonatal transport receptor (FcRn) and the IgG protection receptor (FcRp) are one and the same: 10–12 unified under the term, Brambell receptor (FcRB), 11,13 in honour of their discoverer (Prof. F. W. Rogers Brambell, 1901–70) who originally described and correlated both of these activities (for reviews, see refs. 13–15). The timing and tissues of FcRB expression highlight its dual functions. Reflecting its role in IgG transport, FcRB is expressed in tissues involved in immunity transmission. To transfer immunity from ingested mothers’ milk, FcRB is expressed at high levels in the intestinal epithelium of neonatal mice and rats; 1,16–18 in its transport expression, FcRB was originally termed ‘FcRn’, for neonatal rat intestine, the tissue from which it was initially cloned. 17 However, FcRB is also present in tissues involved in the antenatal transmission of IgG: fetal yolk sac of rabbit, mouse and rat, 1,18,19 correlating with the exclusively antenatal transmission in rabbits, and the ante- plus neonatal transmission in mice and rats. 1,13 In humans, IgG transmission from mother to young is solely antenatal, 1,5,13 and FcRB is accordingly expressed in human placental syncytiotrophoblasts without prominent expression in newborn intestinal epithelium. 20,21 Reflecting its dominant role in mature animals as FcRp to protect IgG from catabolism, FcRB is highly expressed in the vascular endothelial cells of adult rats, mice and humans (D. D. Sedmak, B. M. Radhill, J. M. Osborne, et al., submitted). 10,22 Logically, to protect a protein from catabolism, a protection receptor should be expressed precisely at those sites of catabolism, and, for the majority of proteins, catabolism follows directly from endocytosis. Vascular endothelium is the most endocytically active tissue in adult animals, correlating with its role to mediate the bulk pinocytosis and processing of plasma fluids, 23 and the high expression of FcRB (as FcRp) in this tissue makes sense. By the action of this unique protection receptor, FcRB makes IgG the longest-surviving of all plasma proteins, increasing blood levels of IgG up to 10-fold under conditions of constant IgG biosynthesis. 5,11,13,24 In contrast, most non-vascular tissue components express FcRB at lower levels, paralleling their lower endocytic activities (D. D. Sedmak, B. M. Radhill, J. M. Osborne, et al., submitted). 17,20,22 FcRB is a heterodimer of a β2-microglobulin (β2m) light chain and a major histocompatibility complex (MHC) class I-like heavy chain 17,25 that binds IgG via Fc histidine residues in a pH-dependent manner that is essential to its function. 9,26 Endocytosis of IgG – whether by fetal or neonatal tissues in transmission or by catabolic tissues in the body – is followed by binding of IgG to FcRB in the acidic endosome environment (pH 6), directed trafficking through cellular conduits to bypass lysosomal activities, and finally releasing of IgG in the neutral physiological pH (7·4) of the extracellular fluids. In mice knocked-out for β2m, both the FcRn and the FcRp activities of the Brambell receptor are deleted. 10−12,27 The heavy chain (like the β2m light chain) is a single copy gene, 17,18 confirmed by cloning from vascular endothelium as well as from transport tissues (D. D. Sedmak, B. M. Radhill, J. M. Osborne, et al., submitted). Blumberg et al. (1995) documented expression of FcRB in rodent hepatocytes, which led to the speculation that the receptor might transport IgG to bile, in analogy to its role in perinatal transport. 28 If shown to be true, this would be the first FcRn-like transport role for FcRB established beyond the perinatal period. The present study addresses whether FcRB in liver functions as FcRn IgG transporter, as postulated. 28 The null hypothesis is that its role is not as transporter but merely as protector against IgG catabolism, as FcRp, in a tissue endowed with endocytic activity. We show that serum-to-bile transport of IgG is unaffected by the presence of a functional FcRB molecule. The hypothesis that FcRB transports IgG from serum to bile is therefore rejected. We accordingly favour the null hypothesis, that FcRB expression in hepatocytes parallels instead the endocytic activity in those cells, protecting IgG from catabolism in hepatocytes just as it does in the more endocytically active vascular endothelium. This adds evidence to the principle that FcRB will be expressed in adult tissues wherever there is endocytic activity, as FcRp, to protect adventitiously endocytosed IgG from catabolism. 13 Because the distinct roles of transport (FcRn) versus protection (FcRp) of this receptor protein are what is in question, we refer to the receptor generically, without prejudice, as FcRB or Brambell receptor, as previously applied. 11,13,25,29–31 Wild-type and β2m knockout (β2m–/–) C57BL/6J mice, 10–12 weeks old, were from Jackson Laboratory (Bar Harbor, ME). The β2m–/– mice are functionally FcRB–/–. 10−12,27 Murine anti-Tac IgG2a,κ antibody (MAT) (from T.A. Waldmann, National Institutes of Health, Bethesda, MD) against human interleukin-2 receptor does not react with any mouse proteins or tissues, and was used as mouse IgG for protein labelling. Murine serum albumin (MSA) was from Inter-Cell (Hopewell, NJ). Proteins were radioiodinated and mixed prior to injection. Samples were counted with decay and cross-over corrections as previously. 11 Endogenous IgG and IgA were measured by enzyme-linked immunosorbent assay (ELISA) relative to purified mouse IgG (UPC, Sigma, St Louis, MO) and bulk IgA (Sigma). 11 The β2m knockout and wild-type mice (five each) were injected intraperitoneally with 200 µl phosphate-buffered saline (PBS) containing 460 kBq 125I-labelled MAT (referred to as ‘IgG’ in the text) and 500 kBq 131I-labelled MSA. Blood samples were collected at the times indicated and normalized to plasma activities based on measured haematocrits, as previously. 11 (To avoid confusion with the abbreviation of ‘p’ for peritoneum, we employ the term ‘serum’ instead of ‘plasma’ in the text, although serum was not actually prepared; values would be expected to be the same for serum and plasma.) Twenty hours after injection, the gall bladder was cleared manually and the ciliary bile duct was ligated. The gall bladder was harvested for bile after 4 hr, during the linear filling period (not shown). Mean weights of control (32·3 g) and knockout (33·4 g) mice were similar and dilutional weight corrections were not made. Samples were processed for protein-bound counts as previously. 32 Rate constant and compartment definitions are shown below. Kinetic constants were derived in part by curve fitting with pcnonlin (SCI Software, Lexington, KY), using a two-compartment, first-order model, with first-order input representing the transfer of protein from the peritoneum to serum (plasma). The third (bile) compartment was not included in the modelling, as it is formally part of catabolism, included in k10. The fraction of total catabolism (k10) due to biliary loss (k1b) is < 1% (see the Results). Instead, it can be shown that k1b can be derived as: in which C and V are concentrations and volumes of serum (1) and bile (b), and Δt is the bile collection interval. Bile volume increases under a constant flow, Vb = a Δt, with a≈ 0·28 ml/day by our measurement (data not shown). V1, the blood serum (plasma) volume, is 48·8 ml/kg in mice, 33 or 1·6 ml for our mean animal weights (above). Cb/C1 is the bile-to-serum (B : S) ratio used in the text. Equation (1) is approximate for kinetic experiments and exact for steady-state experiments. Errors are standard deviations unless standard errors are stated. The difference between two means was considered significant by two-tailed Student’s t-test for P < 0·05. Disruption of FcRB function abrogates both IgG protection (FcRp) and neonatal transport (FcRn) in affected mice. 10−12,27 If the FcRB in hepatocytes mediates IgG transport, serum-to-bile transport in FcRB knockout mice will be suppressed relative to wild-type. Lack of difference in these tests would favour the null hypothesis that FcRB is not a transport receptor in hepatocytes, but rather that its presence is attributable to its alternative role to protect IgG from catabolism in an endocytically active tissue. In the following, albumin and IgA are used as controls, whose presence in the bile is due, respectively, to passive filtration 34,35 and selective secretion [secretory (mainly dimeric) IgA transmits to bile via pIgR; small amounts of monomer IgA transmit passively], 36,37 both independent of FcRB. We employed the intraperitoneal (i.p.) route for the administration of radiolabelled proteins to avoid the impact of fast initial (alpha-phase) kinetics of intravenous injections, thereby reducing absolute differences in blood concentrations of rapidly and slowly catabolized proteins over early time intervals. 32,38 The pharmacokinetic model for this design is described in Fig. 1. Pharmacokinetic model for studies of injected proteins. The system is shown here as a transient four-to-three compartment model. (for the formal modelling, see the Materials and Methods.) Labelled proteins are injected intraperitoneally (compartment p) with sampling of serum (compartment 1) and bile (compartment b) at specified times. The peritoneal (injection) compartment (p) exists separately only during the emptying period postinjection, after which the peritoneum is included in the whole body extravascular compartment (2). There is bidirectional transfer of proteins between compartments 1 and 2, but other flows are unidirectional. The boxes for the sampled compartments (1 and b) are shown in bold. The rate constants for protein transfer are according to standard nomenclatures, with subscripts in sequence with the ‘from’ compartment followed by the ‘to’ compartment (e.g. k12 for 1 → 2 transfer). All catabolism is represented as loss from the serum (plasma) compartment, 5,39 with the net fractional catabolism as k10. The protein loss due to bile clearance (k1b) is < 1% of total catabolism, and the separate representation of k1b (which is included in k10) does not alter the overall kinetic picture. Serum concentrations of radiolabelled IgG mouse monoclonal antibody and MSA peaked between 4 and 8 hr after injection ( Fig. 2), levelling off, then gradually declining in a quasi-stable period 16–24 hr after injection. In this interval, the ratio of the radioactivities of IgG to MSA in wild-type mice (2·6 ± 0·5) is two-fold higher than in β2m knockout mice (1·3 ± 0·4), reflecting the higher catabolic rate constants (k10, Fig. 1) for IgG in the knockout mice. Although intravascular catabolism (k10) is nearly 10-fold faster in knockout than in wild-type mice, 11 these blood level differences were moderated in the present setting versus our prior intravenous studies, 11 supporting the design rationale to use the i.p. injection route. Pharmacokinetics of radiolabelled MSA and IgG after intraperitoneal injection. Blood samples were obtained and processed for protein-bound radioactivity as in the Materials and Methods. Wild-type: MSA (●), IgG (▪). FcRB knockout: MSA (○), IgG (□). Intraperitoneal injection is at day 0, bile duct ligation is at 0·83 day, and bile harvest is at 1·0 day. Error bars represent standard error of the mean (SEM); n = 5 for each group. For comparing bile transport ratios, Fig. 3 shows serum and bile radioactivity of MSA and IgG. The serum activity of MSA was slightly higher for the knockout than the wild-type mice, presumed due to injection differences, but the IgG activities in both serum and bile were roughly two-fold lower in the knockout mice, reflecting their accelerated catabolism of IgG. Considering first the wild-type animals, labelled MSA and IgG had bile-to-serum (B : S) ratios of 0·026 ± 0·010 and 0·008 ± 0·005, respectively. The lower B : S ratios for IgG, a protein of 150 000 MW, versus MSA, a protein of 66 000 MW, parallel differences seen in other studies of rodents, dogs and man, previously attributed to sieving based on size. 34,35 In the knockout mice, transport ratios for MSA (0·022 ± 0·009) were comparable to those in wild-type mice (0·026 ± 0·010; P = 0·53). Similarly, the IgG transport ratios in knockout mice (0·011 ± 0·005) did not differ significantly from those in wild-type mice (0·008 ± 0·005; P = 0·50). Accordingly, our kinetics analysis shows no impact of FcRB on transport rates of IgG from serum to bile. Bile transport kinetics of injected radiolabelled MSA and IgG. Left panel, serum concentrations are from the 0·7-day blood sample that preceded ligation of the ciliary bile duct, and bile concentrations are from 1·0 day when bile was harvested, allowing a lag time in serum-to-bile-to-gall bladder transport. Values are all protein-bound counts. Right panel, B : S ratios of MSA and IgG. Wild-type: MSA (●), IgG (▪). FcRB knockout: MSA (○), IgG (□); each symbol represents a different mouse. Horizontal line is the mean. SD values are listed in the text. Between wild-type and knockout mice, only the serum concentration of IgG is significantly different (P = 0·001). A corollary of elementary pharmacokinetics is that steady-state B : S concentration ratios should parallel the kinetically determined transport rate constants in a unidirectional flow (k1b, Fig. 1) with constant volume production in the bile compartment (see the Materials and Methods). In contrast to the prior study with injected radiolabelled proteins, endogenously produced IgA and IgG in serum and bile will be at true steady-state at all times during the experiment, whether by passive filtration or by receptor-mediated transport. Figure 4 represents the endogenous, steady-state concentrations of IgA and IgG in serum and bile. levels of IgA and IgG. Left panel, IgA and IgG levels in serum and bile were determined by as in the Materials and Methods. Right panel, B : S ratios of IgA and IgG. Wild-type: IgA IgG FcRB knockout: IgA IgG (□). Horizontal line is the mean of all : S in which one in on was SD values are listed in the text. the serum concentrations and bile concentrations of IgG were significantly different between wild-type and knockout mice (P < but the B : S values were not humans, in mice and IgA is a serum IgA levels in our mice did not differ significantly between wild-type and knockout animals ± versus ± P = ( Fig. B : S concentration ratios ± one versus ± for wild-type and IgA a relative in transport rate versus IgG by the action of the immunoglobulin receptor Because IgA does not with the comparable values for IgA in knockout versus wild-type mice were as In to serum IgG in wild-type animals was at a higher concentration ± ( Fig. expected from the kinetics the steady-state concentration of IgG in the bile of wild-type mice was 1% of that in serum ( Fig. mice had suppressed serum IgG levels ± P < versus lower steady-state blood levels are with pharmacokinetic of catabolism (k10, Fig. 1) due to loss of the FcRB protector function (FcRp) with an IgG IgG in bile was a similar difference was with lower IgG in the knockout mice ± versus ± in In contrast to the differences in the absolute IgG levels in plasma and bile, the B : S ratios of IgG for wild-type and knockout mice were ± and ± P = ( Fig. This that FcRB expression in hepatocytes no significant role in the transport of IgG to bile. 1 the kinetic constants of Fig. 1 determined here and in prior The kinetic study based on injected mouse monoclonal IgG and the study based on mouse IgG There were no significant differences in the rate constants for transport of IgG from serum to bile (k1b) in and only the catabolic rate constant (k10) was By the ratio the loss of IgG to bile for < 1% of the total catabolic loss of IgG in Endocytosis is to a or by all extracellular is by this its are processed through a complex of and of acidic in which the compartment a protein is in the of the endocytic it to catabolism via this the of most proteins, or it is by receptor binding in the If from catabolism, the protein is the in the manner of secretion of IgA via the receptor, or it is to the in the manner of via the receptor all tissues, the vascular endothelium – with its in – is by the most active the plasma volume in mice and 5 in humans, and was long as the site of catabolism for all plasma proteins. In this the FcRB two both originally described and by Brambell, that are by tissue and timing of The of the of this receptor is to the central of IgG to the one the FcRB (as transmits immunity that from their immune can this and, on the other FcRB (as FcRp) up to 10-fold increases in IgG levels the levels significant to all the immunoglobulin only of IgG is with in the of However, the of the expression of FcRB in hepatocytes is it transmit IgG to bile, to the FcRn transport, as previously 28 or its role be in the protection of IgG from the endocytic activity of this as FcRB was shown in hepatocytes in 28 only the FcRn transport function had with the receptor, and it was to solely in of its transport activity. However, with the that the FcRp is the same receptor as previously 1,3–6 it of to between these alternative roles for FcRB in FcRB heavy chain studies of FcRB function have employed mice, 10−12,27 which expression of other as This is based on the functions and of impact on IgG or protein catabolism in β2m–/– mice in refs. and based on the to of any evidence for via loss of other that might have a on IgG transport and catabolism. By our study the hypothesis is that FcRB mediates IgG transport as by transport rates in knockout the null hypothesis is that IgG transport to bile is unaffected by FcRB. to an FcRn-like transport the alternative role of the receptor in hepatocytes is as that FcRB expression merely parallels the endocytic activity of the IgG to the and protecting it from catabolism. by Brambell 4 would that the catabolism of in the manner in any part of the body in which pinocytosis of plasma proteins Serum and bile concentrations for MSA and total IgA and their B : S ratios were the same for wild-type and knockout mice. the processes of passive filtration of albumin and selective secretion of IgA as by FcRB In contrast, IgG was by the loss of FcRB as indicated by the accelerated clearance of injected radiolabelled IgG and by the lower concentrations of IgG in serum and bile in the setting of constant IgG loss of FcRB function in β2m knockout mice, no significant differences were in the IgG B : S ratios and serum-to-bile transport rate constants 1) in the kinetic or the steady-state experiments. Accordingly, we the hypothesis that FcRB transports IgG from serum to bile and instead favour the null hypothesis that FcRB in hepatocytes is to its alternative as on in the bile of different have shown selective secretion of IgA from serum to bile. In contrast, IgG in bile is not more represented than other proteins relative to it in mice and these by that the only transport receptor for IgG, the present in does not a transport function in this IgG to bile by passive In the FcRB role as FcRp, IgG protection from receptor to the same In its FcRn transport of IgG in shown to be to any IgG concentration in In the of evidence for a transport function in hepatocytes, our also makes the speculation of a role for FcRB in biliary sampling and immune 28 Although many of biliary to be the passive transport of proteins to bile to be via two one is termed the the other an termed the or 34,35 The of proteins in shows the same to lower as in bile. Accordingly, these and processes have to at the level of the endothelial and not at the level of the or which is with our that FcRB in hepatocytes is not a transport function. It is that these proteins then from to bile in an between cellular or as passive in which bypass the and in which are and The net of bile production and protein transfer from blood is that bile is to in total termed a of The to the for of plasma proteins to bile via trafficking is or receptor-mediated endocytosis. is also the major route for catabolism of extracellular proteins by hepatocytes and other For the same that cells of the vascular endothelium express FcRB at high levels to protect endocytosed IgG from catabolism, one that hepatocytes in the which are active in will also express the FcRB receptor, to of endocytosed IgG. This is with that in its role as FcRp, is in parallel with the endocytic activity of 13 We are to for of time and for in the We to our and for their on the We and for with the This was by to from the and to from the and