Distribution of FcRn Across Species and Tissues

Abstract

The neonatal Fc receptor (FcRn) is a major histocompatibility complex class I type molecule that binds to, transports, and recycles immunoglobulin G (IgG) and albumin, thereby protecting them from lysosomal degradation. Therefore, besides the knowledge of FcRn affinity, FcRn protein expression is critical in understanding the pharmacokinetic behavior of Fc-containing biotherapeutics such as monoclonal antibodies. The goal of this investigation was to achieve for the first time a comparative assessment of FcRn distribution across a variety of tissues and species. FcRn was mapped in about 20 tissues including placenta from human and the most frequently used species in non-clinical safety testing of monoclonal antibodies (mouse, rat, cynomolgus monkey). In addition, the FcRn expression pattern was characterized in two humanized transgenic mouse lines (Tg32 and Tg276) expressing human FcRn under different promoters, and in the severe combined immunodeficient (SCID) mouse. Consecutive sections were stained with specific markers, namely, anti-CD68 for macrophages and anti–von Willebrand Factor for endothelial cells. Overall, the FcRn expression pattern was comparable across species and tissues with consistent expression of FcRn in endothelial cells and interstitial macrophages, Kupffer cells, alveolar macrophages, enterocytes, and choroid plexus epithelium. The human FcRn transgenic mouse Tg276 showed a different and much more widespread staining pattern of FcRn. In addition, immunodeficiency and lack of IgG in SCID mice had no negative effect on FcRn expression compared with wild-type mice.

Introduction

The neonatal Fc receptor, FcRn, is a major histocompatibility complex (MHC) class I type molecule, a heterodimer consisting of an α-chain and β-2-microglobulin. FcRn binds to, transports, and recycles immunoglobulin G (IgG) and albumin, thereby protecting them from lysosomal degradation. It was first discovered in the intestine of neonatal rats, where it confers passive humoral immunity to the neonate by transporting maternal IgG from ingested milk through the intestinal epithelium and into the newborn’s bloodstream.^1,2 A similar mode of function was later discovered in the human placenta, where FcRn transfers maternal IgG to the fetus across the syncytiotrophoblasts.³

Since its initial discovery, FcRn has been shown to encompass and uphold a variety of functions long after the fetal and neonatal periods.⁴ One of the main functions of FcRn throughout life is the regulation of IgG homeostasis by protecting endocytosed IgG from lysosomal degradation, with recycling of internalized IgG molecules back to the cell surface. Due to this selective recycling, IgGs have exceptionally long serum half-lives compared with other antibody classes (in humans ~21 days instead of 6–8 days). Mice deficient in either one of the FcRn subunits, the β-2-microglobulin or the α-chain, have been shown to have significantly lower endogenous IgG levels, with half-lives similar to those of other Ig classes.^4–7 This recycling mechanism was later found to extend the serum half-life of albumin as well.⁸

So far, FcRn has been detected by immunohistochemistry (IHC) or immunofluorescence in most tissues of all species examined. It is most consistently localized in endothelial cells and antigen-presenting cells (APCs), such as macrophages/monocytes, dendritic cells, and B cells.^9–11 In addition, several mucosal surfaces and epithelial cells have been found to express the receptor, such as enterocytes in the intestine,¹² tubular cells in the kidney,¹³ bronchial and alveolar epithelium in the lung, epithelial cells of the choroid plexus and ciliary body,^14,15 and keratinocytes in the skin.¹⁶

One of the most established locations of FcRn expression is endothelial cells, especially in the large vascular beds of skin and muscle, and to a slightly lesser extent in liver,¹⁷ where FcRn-mediated IgG recycling has been demonstrated both in vivo and in vitro. In each of these tissues, endothelial cells have access to high levels of circulating IgGs and they have been demonstrated to express functional FcRn, capable of intercepting internalized IgG and recycling it back to the cell surface.¹⁷ High and rather ubiquitous levels of FcRn expression have also been detected in professional APCs, where FcRn directs IgG-containing immune complexes toward MHC.¹⁸ The importance of bone marrow–derived cells for IgG homeostasis has been shown in chimeric mouse studies, in which mice with FcRn-deficient bone marrow had faster clearance of injected IgG tracer.⁹ Selective deletion of FcRn in endothelial and hematopoietic cells in a mouse model effectively removed all capacity for IgG salvage, demonstrating that endothelial and hematopoietic cells are the principal sites for IgG regulation. For albumin, a similar increased clearance was detected, although not as drastic due to its higher synthesis rate.¹⁹

Besides its roles in conferred maternal immunity in early neonatal life and maintaining high protective IgG levels, FcRn has other significant functions in host immune responses, especially in the intestinal mucosa. FcRn has been shown to participate in initiating a specific immune response against bacterial pathogens by capturing luminal IgG-complexed antigens and transporting them across the epithelium to lamina propria dendritic cells, which further activate a CD4⁺ T cell response.²⁰ In the case of colorectal cancer, an even wider role in antitumor immunosurveillance has been detected, with FcRn expression in dendritic cells shown to be critical for the activation of tumor-reactive CD8⁺ T cells and cytokine secretion.²¹

A quantitative understanding of the affinity of FcRn interactions is opening up opportunities for engineered therapeutic IgGs with desired pharmacokinetic (PK) properties.^22–25 Detailed knowledge of FcRn expression patterns and species-specific differences are necessary to better extrapolate preclinical PK data from the model animal species to the human. It has been demonstrated that the binding affinity is greater between murine FcRn and human IgGs than between murine FcRn and endogenous murine IgGs.^26,27 Thus, when a human IgG is administered to a mouse, it is preferentially recycled over the endogenous IgGs, greatly complicating allometric-based predictions of human PK. The development of transgenic mouse lines expressing human FcRn may provide a translational animal model for human PK prediction as an alternative to non-human primates. A series of mouse lines have been generated, including Tg32 expressing hFcRn under the control of its endogenous human promoter sequence and Tg276 expressing hFcRn under a chicken β-actin (CAG) promoter that results in a ubiquitous expression pattern.^23,28,29 Both lines have a depleted α-chain murine FcRn and contain either hemizygous or homozygous transgenes of human FcRn. Various PK profiles of IgG have been observed,²³ raising the question about selection of the best model for PK prediction in humans.^30,31 Recent attempts have been made to quantify the amount of FcRn protein levels in tissue of transgenic mice.³² Considering the heterogeneity of tissue and the inability of this method to distinguish between cell populations, the conclusion from measurement of total protein content which can be derived for a quantitative physiological PK model is limited.

To bring together and expand the currently scattered information on FcRn expression in different tissues and species, FcRn was mapped in about 20 tissues of several species, including the human and the most frequently used species in non-clinical safety testing of monoclonal antibodies (MAbs), that is, mouse, rat and cynomolgus monkey. In addition, we report the FcRn expression in two humanized transgenic mouse lines, Tg32 and Tg276. Furthermore, the severe combined immunodeficient (SCID) mouse was included as this model is used often in tumor biology testing the effect of therapeutic MAb.

Materials and Methods

Sample Acquisition and Preparation

In this study, the distribution of FcRn was mapped across a set of species and selected genetically modified mouse lines. Included were human, cynomolgus monkey, wild-type mouse C57BL/6J, FcRn knock-out mouse B6.129X1-Fcgrt^tm1Dcr/DcrJ (FcRn^−/−), humanized transgenic mouse line B6.Cg-Fcgrt^tm1Dcr Tg(FCGRT)32Dcr/DcrJ (Tg32), humanized transgenic mouse line B6.Cg-Fcgrt^tm1Dcr Tg(CAG-FCGRT)276Dcr/DcrJ (Tg276), SCID mouse CB17.Cg-Prkdc^scidLyst^bg-J/Crl, and Wistar rat (Crl:WI (Han)). All rodents were obtained from Charles River Laboratories (Sulzfeld, Germany), whereas the cynomolgus monkey of Mauritius origin had been a part of a long running colony housed at the Roche animal facility. All work concerning live animals was conducted according to national and European animal health and welfare guidelines (Decree n° 2001-464; Decree n° 2001-486; Directive 2010/63/EU). The human tissues were ordered either from Asterand Bioscience (Detroit, MI) or AnaBios (San Diego, CA) as ready-to-use paraffin-embedded tissue blocks with a 48-hr formalin fixation.

From each species and mouse line, the following tissues were collected: liver, kidney, heart, lung, skin, skeletal muscle, eye, brain, sciatic nerve, small and large intestine, thyroid, spleen, thymus, lymph node, bone marrow, adrenal gland, salivary gland, prostate, testis, and placenta (except for FcRn knock-out mouse and SCID mouse). Tissues from rodents and cynomolgus monkey were fixed in 10% neutral buffered formalin for either 24 hr (rodents) or 48 hr (cynomolgus monkey), after which the tissues were dehydrated using series of graded alcohols and routinely embedded in paraffin. To better preserve the histomorphology of eyes and testes, Davidson’s fixative from AppliChem (Darmstadt, Germany) was used as a fixative (24 hr for rodents and 48 hr for cynomolgus monkey) instead of formalin,³³ followed by postfixation in 70% ethanol for 24 hr. To access bone marrow, pieces of sternum were decalcified and fixed by immersion in 10% formic acid/formalin solution until the tissue became soft enough to cut (~5 days). After fixation, eyes, testes, and bone marrow were dehydrated and paraffin embedded as previously described.

At the time of necropsy, the mice were between 9 and 12 weeks, the rat 10 weeks, and the cynomolgus monkey 19 years old, with human samples varying mostly between ages 50 and 75 at the time of collection. The placenta originating from different individuals were collected at the following gestation dates (GD): human approximately GD125, cynomolgus monkey GD150, wild-type mouse GD15, humanized transgenic mouse line Tg32 GD14, humanized transgenic mouse line Tg276 GD14, and Wistar rat GD17.

Immunohistochemistry

Immunostaining was carried out in a Ventana Discovery XT immunostainer (Ventana Medical Systems, Inc., Tucson, AZ) using the avidin–biotin peroxidase complex method and a Ventana DAB Map Detection Kit or the Ventana RedMap™ kit (eyes only). Antigen- and species-specific staining protocols including information on antibody suppliers are detailed in Table 1. The two antibodies used for FcRn detection were developed against the following proteins: The antibody from Novus Biologicals NBP1-89128 (Novus Biologicals, LLC, Littleton, CO) was developed against recombinant human FcRn protein corresponding to amino acids Lys175 to Ser294. The antibody from R&D Systems AF6775 (R&D Systems, Inc., Minneapolis, MN) was developed against mouse myeloma cell line NS0 derived mouse FcRn from Ser22 to Ser301. Both antibodies were tested in Western blot and reacted with primate and rodent recombinant FcRn-B2M heterodimer proteins purchased from Sino Biological (human: CT009-H08H, cynomolgus monkey: CT031-C08H, wild-type mouse: CT029-M08H, and rat: CT030-R08H; Sino Biological, Inc., Beijing, China) with specific bands on the SDS-PAGE at 35 kDa for human and cynomolgus monkey, and 45 kDa for wild-type mouse and rat (data not shown).

Table 1.

Used Antibodies and Protocol Specificities for Immunohistochemical Detection of FcRn, von Willebrand Factor, and CD68.

Primary Antibody	Source	Host	Reactive Species	Concentration, Incubation Time	Pretreatmenta and Blocking Stepsb	Secondary Antibodyc, Concentration, and Incubation Time
FcRn	Novus Biologicals; NBP1-89128	Rabbit polyclonal	Human, non-human primate, Tg32, Tg276	4 µg/mL, 20 min (primates) 1 µg/mL, 20 min (rodents)	CC1 mild 40% NGS, 16 min; AB 12 min	A, 1 µg/mL, 8min
	R&D Systems; AF6775	Goat polyclonal	WT mouse, SCID mouse	4 µg/mL, 20 min	CC1 mild 20% NMS, 12 min	B, 3 µg/mL, 8 min
			Rat	2 µg/mL, 12 hr	CC1 mild	B, 15 µg/mL, 8 min
von Willebrand Factor	Dako; A0082	Rabbit polyclonal	Human	10,3 µg/mL, 20 min	CC1 standard 20% NGS, 12 min; AB, 4 min	A, 1 µg/mL, 8 min
CD68	Dako; M0814	Mouse monoclonal	Human	0.23 µg/mL, 44 min	CC1 mild 20% NGS, 16 min; AB, 4 min	C, 5 µg/mL, 24 min

Abbreviations: FcRn, neonatal Fc receptor; WT, wild type; SCID, severe combined immunodeficient.

^aPretreatment conditions: CC1, Ventana Cell Conditioning 1 (Tris-based buffer, pH 8.5); mild 32 min, standard 64 min.

^bBlocking reagents: NGS, Goat Serum (Normal); NMS, Mouse Serum (Normal), Protein block serum free, all from Dako (Agilent Technologies, Santa Clara, CA); AB, Endogenous Biotin Blocking Kit from Ventana Medical Systems, Inc. (Tucson, AZ).

^cSecondary antibodies: (A) biotinylated donkey anti-rabbit IgG, (Jackson Immunoresearch, Inc., West Grove, PA); (B) biotinylated rabbit anti-goat IgG (Vector Laboratories, Inc., Burlingame, CA); (C) biotinylated horse anti-mouse IgG (Vector Laboratories Inc., Burlingame, CA). All dilutions of primary and secondary antibodies and blocking reagents were made using Discovery Antibody Diluent from Ventana Medical Systems, Inc., except for primary antibody CD68, which was diluted using Antibody Diluent from Dako (Agilent Technologies).

In brief, tissues were sectioned at 4 µm, mounted onto glass slides, dried overnight at 37C, deparaffinized, subjected to protocol-specific pretreatment and blocking steps, and incubated with either anti-FcRn, anti–von Willebrand Factor (vWF), or anti-CD68 primary antibody. This was followed by incubation with a biotinylated secondary antibody raised against the host species of the primary antibody, DAB and RedMap detection, and counterstaining with hematoxylin II and bluing reagent from Ventana.

As an isotype control, a matching concentration of normal serum was applied to control sections instead of the primary antibody. FcRn knock-out mice, stained with the anti-hFcRn antibody, were also used as negative controls when staining the humanized transgenic mouse lines. One slide completely omitting primary antibody or serum was applied to each staining round as well as a purely negative control.

Stained slides were evaluated with a light microscope (Zeiss Axio Imager 2; Carl Zeiss AG, Oberkochen, Germany), and staining intensity and abundance were semiquantitatively assessed either with + (mild intensity/expressed only in a subset of cells of this type), ++ (strong intensity/expressed in most cells of this type), or − (no staining in this cell type). A fully quantitative read out was not performed due to species-specific adaptions of the protocol reflecting differences in binding affinity of the antibody. Representative images were taken from scanned slides (Aperio ScanScope AT Slide Scanner; Leica Microsystems GmbH, Wetzlar, Germany).

Results

Distribution of FcRn Within the Placenta of Different Species

FcRn expression was present in the placenta of all species and strains tested showing differences in distribution between primates and rodents (Fig. 1). In the hemomonochorial placenta of human and cynomolgus monkey,³⁴ FcRn was detected in the syncytiotrophoblasts and in endothelial cells of fetal vessels of the placental villi. In the hemotrichorial placenta of rats,³⁴ FcRn was detected in the epithelial cells of the yolk sac endoderm and in endothelial cells of fetal vessels in the labyrinth zone. In the hemotrichorial placenta of wild-type mouse,³⁴ FcRn was detected in the epithelial cells of the yolk sac endoderm only. In the placenta of the humanized mouse line Tg32, FcRn was detected in the epithelial cells of the yolk sac endoderm, in trophoblastic giant cells of the basal zone, and in endothelial cells of fetal vessels in the labyrinth zone. In the placenta of the humanized mouse line Tg276, FcRn was detected in the epithelial cells of the yolk sac endoderm, in all cells of the basal zone, and in endothelial cells of fetal vessels and in syncytiotrophoblasts in the labyrinth zone.

Figure 1.

Immunohistochemical localization of FcRn in the placenta of different species. In human (A) and cynomolgus monkey (B), FcRn was detected in syncytiotrophoblasts (dashed arrow) and in endothelial cells of fetal vessels (arrow). In rat (C), FcRn was detected in epithelial cells of the yolk sac endoderm (asterisk) and in endothelial cells of fetal vessels in the LZ (arrow). In wild-type mice (D), FcRn was detected in epithelial cells of the yolk sac endoderm (asterisk). In humanized mouse line Tg32 (E), FcRn was detected in trophoblastic giant cells of the BZ (dashed arrow) and in endothelial cells of fetal vessels in the LZ (arrow). In humanized mouse line Tg276 (F), FcRn was detected in all cells of the BZ, and in endothelial cells of fetal vessels (arrow) and in syncytiotrophoblasts (dashed arrow) in the LZ. Scale bar = 50 µm. Abbreviations: FcRn, neonatal Fc receptor; LZ, labyrinth zone; BZ, basal zone.

Distribution of FcRn Across Different Tissue Types

The FcRn localization pattern was established across our selected panel of 20 tissues, including intestine, muscle, brain, eye, skin, lung, thyroid, spleen, lymph node, thymus, bone marrow, adrenal, salivary glands, and testis. Photographs of the immunostaining for most tissues from the cynomolgus monkey are shown in Fig. 2. Similar results were obtained in other species. In all tissues, FcRn was consistently detected in interstitial and endothelial cells, with slightly more variation in staining intensity and abundance in the endothelial cells.

Figure 2.

Immunohistochemical localization of neonatal Fc receptor (FcRn) across tissues of cynomolgus monkey. Tissues included in this study were liver (A), small intestine (jejunum; B), large intestine (colon; C), skeletal muscle (D), heart muscle (E), sciatic nerve (F), brain (G), eye (H), skin (I), lung (J), thyroid (K), spleen (L), lymph node (M), bone marrow (N), adrenal gland (O), and testis (P). Scale bar = 50 µm.

In the liver (Fig. 2A), strong staining was detected in Kupffer cells with slightly less prominent staining in hepatic sinusoids and potentially in the adjacent sinusoidal hepatocyte membranes. In small and large intestine (Fig. 2B and C), apical enterocytes and enterocytes of the crypts and goblet cells were strongly stained as well as endothelial and interstitial cells in the underlying lamina propria. In skeletal muscle (Fig. 2D), sciatic nerve (Fig. 2F), and skin (Fig. 2I), interstitial cells and endothelial cells showed prominent staining. In heart muscle (Fig. 2E) and thyroid (Fig. 2K), only a smaller subset of endothelial cells showed FcRn expression, with stronger staining and wider abundance in interstitial cells. In the brain, FcRn was present in the choroid plexus epithelium (not shown) and endothelial cells of the microvasculature (Fig. 2G). In the lung, alveolar macrophages and a subset of endothelial cells were lightly stained (Fig. 2J). In spleen (Fig. 2L), lymph node (Fig. 2M), thymus (not shown), and bone marrow (Fig. 2N), strong staining was detected in interstitial cells, with macrophage-like morphology. Adrenal gland (Fig. 2O) and testis (Fig. 2P) both had staining of endothelial cells and interstitial cells. The least abundant staining and lowest intensity was detected in the salivary gland (not shown), where only a subset of interstitial cells had FcRn staining.

Identification of FcRn-Expressing Cell Types via Staining of Consecutive Sections

To confirm FcRn staining in endothelial cells and macrophages, consecutive sections were stained with anti-FcRn, anti-CD68 as a macrophage marker, and anti-vWF as a marker for endothelial cells. An example is shown of the staining of consecutive slides from human small intestine (Fig. 3). FcRn staining colocalized with a subset of both vWF-positive blood vessels and CD68-positive macrophages, confirming FcRn expression in both of these cell types. Nevertheless, mismatches were also present where FcRn staining was detected in vWF-negative vessels or CD68-negative interstitial cells. Similar staining patterns were observed for the other species and other tissues (data not shown).

Figure 3.

Identification of neonatal Fc receptor (FcRn)–positive cell types. Consecutive sections from human small intestine were stained for FcRn (A), macrophage marker CD68 (B), and endothelial marker von Willebrand Factor (vWF; C). FcRn staining colocalized with a subset of macrophages (arrow) and with a subset of vWF-positive blood vessels (asterisk). Scale bar = 50 µm.

Comparison of the FcRn Expression Pattern Between Human, Wild-Type Mouse, and the Humanized Transgenic FcRn Mouse Lines

To better characterize the two FcRn-humanized transgenic mouse lines, Tg32 and Tg276, their pattern of FcRn expression was compared with that of the wild-type mouse and that of the human. Photographs of selected tissues are presented in Fig. 4.

Figure 4.

Comparison of neonatal Fc receptor (FcRn) localization between human (A, E, I), humanized transgenic mouse lines Tg32 (B, F, J) and Tg276 (C, G, K), and wild-type mouse (D, H, L). Stained sections of brain (A–D), skin (E–H), and lung (I–L) are presented. In brain sections, endothelial staining of vessels is marked with an asterisk. Tg276 specific staining in each tissue is marked with arrows (subset of neuronal cells (C), keratinocytes (G, dashed arrow), cells of the sebaceous gland (G, arrow), alveolar epithelial cells (K, arrow), and less intense bronchiolar epithelial cells (K dashed arrow)). Additional pictures of staining in choroid plexus are marked with an “i” in the bottom right corner of the brain sections. Scale bar = 50 µm.

Typical staining was found in the brain, where all species express FcRn in the vascular endothelial cells and choroid plexus epithelium (Fig. 4A–D). In the skin, most prominent FcRn expression was found in interstitial cells, with lower intensity in the endothelium of blood vessels (Fig. 4E–H). In the lung, consistent staining in alveolar macrophages was present across species, as well as in vascular endothelial cells in the human (Fig. 4I), Tg32 (Fig. 4J), and Tg276 (Fig. 4K); this staining was less intense in the wild-type mouse (Fig. 4L). The Tg276 mouse showed much more widespread staining of FcRn, in cell types and tissues which do not express FcRn in other mouse lines and other species, including a subset of neuronal cells (Fig. 4C), keratinocytes and cells of the sebaceous gland (Fig. 4G), and alveolar epithelial cells and at a lower level bronchiolar epithelial cells (Fig. 4K). In addition and in contrast to the other species tested, FcRn is widely expressed as shown in Table 2. Tg276 mice express FcRn in cardiomyocytes, neuroretina of the eye, lymphocytes in spleen, thymus and lymph nodes, Sertoli cells, epithelial cells in the prostate, chromaffin cells of the adrenal medulla, and secretory cells in the salivary gland.

Table 2.

FcRn Expression Across Species and Tissues: Overview of the Immunohistochemical Staining Results in Human, Cynomolgus Monkey, Humanized Transgenic Mouse Lines Tg32 and Tg276, Rat, WT Mouse, and SCID Mouse.

Organ	Cell subset	Human	Monkey	32TG	276TG	Rat	WT	SCID
Liver	Hepatocytes	+	+	+	+	+	+	+
	Kupffer cells	++	++	++	++	++	++	++
	Sinusoidal epithelial cells	+	+	+	++	+	+	++
Small intestine	Apical enterocytes	++	++	+	++	++	++	++
	Goblet cells	++	++	+	++	++	+	+
	Enterocytes of crypts	++	++	+	++	++	−	−
Large intestine	Apical enterocytes	++	++	+	++	++	++	++
	Goblet cells	++	++	+	++	++	+	+
	Enterocytes of crypts	++	++	+	++	++	−	−
Brain	Neuronal cells	−	−	−	+	−	−	−
	Choroid plexus	Not assessed	++	++	++	++	++	++
Sciatic nerve	Nerve tissue	−	−	−	−	−	−	−
Lung	Alveolar epithelial cells	−	−	−	++	−	−	−
	Bronchiolar epithelial cells	−	−	−	++	−	−	−
Skin	Keratinocytes	−	−	−	++	−	−	−
Heart	Cardiomyocytes	−	−	−	+	−	−	−
Skeletal muscle	Myocytes	−	−	−	+	−	−	−
Eye	Neuroretina	−	−	−	++	−	−	−
	Ciliary body epithelium	−	−	++	++	−	−	−
Thyroid	Follicular epithelial cells	−	−	−	++	−	−	−
Spleen	Red pulp (macrophages)	++	++	++	++	++	++	++
	White pulp (lymphocytes)	−	−	−	++	−	−	−
Thymus	T lymphocytes	−	−	−	++	−	−	−
Lymph node	Lymphocytes	−	−	−	++	−	−	−
Testis	Sertoli cells	−	−	−	++	−	−	−
Prostate	Epithelial cells	−	−	−	++	−	−	−
Adrenal gland	Chromaffin cells	−	−	−	++	−	−	−
Salivary gland	Secretory cells	−	−	−	++	−	−	−

Staining intensity and abundance indicated either with + (mild intensity/expressed only in a subset of cells of this type), ++ (strong intensity/expressed in most cells of this type) or − (no staining in this cell type). Only tissue-specific cell types are listed; endothelial cells and interstitial cells show FcRn staining throughout tissues and species, and are thus not integrated in this table. Abbreviations: FcRn, neonatal Fc receptor; WT, wild type; SCID, severe combined immunodeficient.

Comparison of FcRn Expression Across All Species and Mouse Lines

FcRn IHC staining across species and all examined tissues is summarized in Table 2. Only tissue-specific cell types are listed. Endothelial cells and interstitial cells show FcRn staining throughout tissues and species, although with slightly variable intensity, and are thus not integrated in this table. Here, the liver is used as an example. IHC staining showed strong FcRn expression in the liver in all of the studied species (Fig. 5). Staining intensity was especially high in Kupffer cells and somewhat less intense in sinusoidal endothelial cells. The precision of the methods used did not allow investigation of FcRn expression in the hepatocyte membrane because of its close anatomical proximity to the sinusoidal endothelium. Thus, FcRn expression on the hepatocyte membrane could only be confirmed for Tg276, which showed staining both around the cell membrane and in the cytosol. FcRn staining was also detected in Kupffer cells and both portal and central vasculature. Only the rat (Fig. 5E) and the humanized mouse lines (Fig. 5C and D) showed FcRn positive staining in the bile duct.

Figure 5.

FcRn expression in liver samples from human (A), cynomolgus monkey (B), humanized transgenic mouse lines Tg32 (C) and Tg276 (D), rat (E), wild-type mouse (F), severe combined immunodeficient mouse (G), and neonatal Fc receptor (FcRn) knock-out mouse (H). All of the species exhibited strong staining of FcRn (excluding the knock-out mouse) in Kupffer cells (dashed arrow), and weak (A–C, E, and F) to strong (D and G) staining in sinusoidal endothelial cells (arrow). FcRn expression was also detected in the bile ducts (asterisk) of both humanized transgenic mouse lines and rat. The FcRn knock-out mouse did not show any staining (internal negative control). Scale bar = 50 µm.

In the kidney, staining in tubular cells as well as for some glomerular cells was detected. As the staining was highly diffuse and similar to the kidney of FcRn knock-out mouse, isotype controls, and antibody-omitting controls, the staining was considered nonspecific in the current study.

Discussion

The objective of this study was to achieve for the first time a comprehensive assessment of the distribution of the FcRn across a variety of tissues and species used for toxicology and PK studies of IgG-based biotherapeutics. The FcRn expression pattern was also characterized in two transgenic mouse lines expressing human FcRn under different promoters. As significant differences in IgG–FcRn interaction depending on the species the IgG and the FcRn originate from have been reported,²⁷ wild-type rodent models have been found to be unreliable for the PK scaling of Fc-containing biopharmaceuticals in humans. Humanized transgenic FcRn mouse lines may be more appropriate for preclinical studies of Fc-based therapeutics.⁷

The expression of FcRn in the placenta has been described for several species, and the distribution and expression pattern could be confirmed in the current study for the epithelial cells of the yolk sac endoderm in the hemotrichorial placenta of wild-type mice and rats.^35,36 In addition, FcRn expression was detectable in endothelial cells of the fetal vessels in the labyrinth zone in rats and not in wild-type mice. The biological impact of these differences could not be elucidated. In the hemomonochorial placenta of human and cynomolgus monkey, FcRn was detected in the syncytiotrophoblasts as described³⁷ and in addition in endothelial cells of the fetal vessels. In the placenta of the humanized mouse line Tg32, FcRn was detected in the epithelial cells of the yolk sac endoderm, in trophoblastic giant cells of the basal zone, and in endothelial cells of fetal vessels in the labyrinth zone. By this, the placenta of this transgenic mouse shows a different FcRn staining pattern than the wild-type mouse having only the epithelial cells of the yolk sac endoderm stained. In the placenta of the humanized mouse line Tg276, FcRn was detected in the epithelial cells of the yolk sac endoderm, in all cells of the basal zone, and in endothelial cells of fetal vessels and in syncytiotrophoblasts in the labyrinth zone. These results are in accordance with the different genetic backgrounds of these mouse lines; Tg276 expresses hFcRn under the CAG promoter, causing ubiquitous high-level gene expression, whereas in Tg32, hFcRn is expressed under its endogenous human promoter.²⁸ However, FcRn being expressed by different cells within the transgenic mouse placenta might have an impact on prenatal IgG transfer. The expression pattern of FcRn in placenta in these transgenic mouse lines is reported here for the first time.

In line with previous studies pinpointing endothelial cells as the major site of IgG catabolism and recycling,⁹ FcRn expression was confirmed in endothelial cells of the tissues examined for each species. When consecutive sections were stained with anti-FcRn and endothelial marker anti-vWF (as shown for the intestine in Fig. 3), some vessel-like structures were only positive for FcRn and not for vWF. This heterogenic distribution of FcRn in endothelial cells seemed to span across most tissues and species. FcRn-positive vessels not expressing vWF could be lymphatic, or simply follow from the known heterogeneity in endothelial cells lining blood vessels.³⁸ The same phenomenon of only partial colocalization of FcRn and endothelial marker staining has been reported previously for vessels in skeletal muscle¹⁷ and eye¹⁵ and hypothesized to be caused by differences in endothelial cell populations.

In addition, across examined tissues and species, interstitial cells with often macrophage-like morphology showed consistent, abundant, and intense staining, confirmed in this study by the colocalization of FcRn and macrophage marker CD68 staining as shown for the human small intestine (Fig. 3). Different APCs have been shown to participate in both extending IgG half-life⁹ and activating local immune responses on mucosal surfaces.²¹ FcRn expressed in dendritic cells has been found to participate in the activation of antitumor CD8⁺ T cells and cytokine secretion in human and murine large intestine, and is thus instrumental in maintaining protective immunity against mucosal tumors.²¹ In line with these results, a strong FcRn staining was detected both on the mucosal epithelium in the intestine and in the interstitial cells of the lamina propria. Some of the interstitial cells could be identified as CD68-positive macrophages and monocytes, but a large number of other interstitial cells were not colocalized with CD68 staining and probably included dendritic cells.

Detailed comparative mapping of FcRn expression in humanized transgenic mouse lines Tg32 and Tg276 has not previously been reported. Our data demonstrated a clear difference in FcRn localization between the two homozygous mouse lines, with Tg32 generally showing an expression pattern similar to that seen in human and monkey tissues, whereas Tg276 expressed FcRn in a wide variety of additional cell types and with more intense staining. These results are in accordance with the different genetic backgrounds of these mouse lines; Tg276 expresses hFcRn under the CAG promoter, causing ubiquitous high-level gene expression, whereas in Tg32, hFcRn is expressed under its endogenous human promoter.²⁸ In a recent study, tissue-specific hFcRn levels were quantified in both hemizygous and homozygous Tg32 mice.³⁹ The highest tissue levels of hFcRn across 14 tissues in the homozygous Tg32 mouse were found in lymph node, lung, and skin. The liver, spleen, and intestine showed medium-level expression, whereas heart, kidney, skeletal muscle, and especially brain had the lowest level of hFcRn expression. Our qualitative results for FcRn expression in Tg32 showed an almost similar trend, with lymph node and spleen showing the strongest and most abundant staining of interstitial and endothelial cells. Skeletal and heart muscle had only low levels of FcRn expression, which was mainly situated in interstitial cells with more faint and scattered staining in the endothelial cells. The brain also had a low level of staining, primarily in a small subset of endothelial cells lining capillaries. In our data, skin and lung showed less staining than, for example, small and large intestine. In general, strong staining of interstitial cells was detected throughout tissues, with generally fainter staining in endothelial cells of the capillaries. Taken together, these data indicate that caution should be applied in selecting the most relevant human FcRn transgenic mouse models for preclinical PK studies of Fc-containing biotherapeutics. Tg human FcRn mice which closely mimic the FcRn-huIgG affinity in combination of physiological PK models are often discussed for prediction of human PK properties in drug discovery as an alternative for use of non-human primates.³⁰ Our results suggest that disposition studies of Fc-containing biotherapeutics should be carried out rather in Tg32 than in Tg276 mice. In addition, physiological models require also a quantitative knowledge of FcRn levels in various tissues to allow scaling of PK to humans. Further advancements need to be done to combine information from FcRn protein quantification in homogenized tissue³⁹ and IHC of FcRn expression in different cell population of tissue. A characterization of transgenic mouse model is also relevant for establishing this species for early safety testing of novel therapeutic modalities such as antibody–drug conjugates which are transported by FcRn and may lead to organ toxicities.

The lung tissue is of particular interest regarding alternative transport methods of IgG-based therapeutics. We found a strong FcRn staining in alveolar macrophages as reported previously,⁹ but we found no staining in the alveolar or bronchiolar epithelium except in the mouse line Tg276. In humans, FcRn expression has been mainly detected in upper airways, where FcRn-mediated transport of IgG therapeutics into the bloodstream has also been demonstrated.⁴⁰ Our results in the rat are in contrast to a previous report, where FcRn was detected in both alveolar and bronchiolar epithelium.⁴¹ Similarly, contrary to previous reports,⁴² we found extensive FcRn expression in the adult intestine. The reasons for these differences could not be elucidated. However, the absence of FcRn staining in alveolar and bronchial epithelium and the extensive FcRn expression in the adult intestine in the rat in the current study are in line with the results from the other species tested. We detected FcRn staining in both small and large intestine in the adult rat and at a lower level in the wild-type mouse. Similar staining in these tissue types was not detected in isotype controls or in the FcRn knock-out mouse using the murine antibody, so the staining was considered to be specific.

We also compared the FcRn distribution between the wild-type mouse and the SCID mouse. As SCID mice lack functional T and B lymphocytes (and in this case natural killer cells due to an additional mutation in Lyst gene), they have very low immunoglobulin production,⁴³ making it a suitable model to test whether endogenous IgG levels have an effect on FcRn expression. The SCID mouse showed the same FcRn distribution as the wild-type mouse, but the staining intensity was generally slightly stronger. The more intense staining in the SCID mouse could be indicative of an unknown technical difference during the tissue processing or immunostaining, but would appear to suggest that immunodeficiency and lack of IgG at least has no negative effect on FcRn expression.

FcRn expression in the kidney has been previously reported in human proximal tubules and glomerular cells.¹³ Unfortunately, despite protocol optimization and different antibodies tested, a reliable staining pattern for FcRn could not be established in this tissue. Similar to the previous studies, staining in tubular cells as well as for some glomerular cells in the human and humanized transgenic mice was detected. As the staining was highly diffuse and similar to the kidney of FcRn knock-out mouse, isotype controls, and antibody-omitting controls, the staining was considered nonspecific in the current study.

Another challenge was experienced with minipigs, where a functional staining protocol could not be established. In a previous study from this laboratory, FcRn was detected in formalin-fixed and paraffin-embedded minipig placenta and fetal jejunum using IHC.⁴⁴ These results could be reproduced, but we were not able to establish specific staining of the tissue panel (formalin fixed and paraffin embedded, or frozen) from adult minipigs using the same and a set of other antibodies.

This study provides for the first time a comprehensive overview of the FcRn expression across tissues of the human and the species most often used in non-clinical safety testing of MAbs. This study confirms the consistent and comparable FcRn staining pattern across tissues and species examined with consistent expression of FcRn in endothelial cells and interstitial macrophages, Kupffer cells, alveolar macrophages, enterocytes, and choroid plexus epithelium.