The MHC Class I-Like IgG Receptor Controls Perinatal IgG Transport, IgG Homeostasis, and Fate of IgG-Fc-Coupled Drugs

Production of FcRn-deficient and human FcRn-transgenic mice

To produce a standard FcRn targeting vector, a 2.7-kb 129/J-derived genomic BamHI fragment carrying sequences 5′ of the FcRn transcriptional start site was inserted upstream of PGK-Neo^r into a pGTN/28 targeting vector (New England Biolabs, Beverly, MA), and a 5.2-kb fragment starting at the ApaI site in exon 2 was inserted 3′. The 129/SvJ-derived ESV/J-1182 embryonic stem cells (Genome Systems, St. Louis, MO) were electroporated with the NotI-linearized targeting construct. DNAs from clones surviving G418 selection were digested with EcoRV for Southern blot screening for homologous recombination of the 5′ targeting vector arm and with EcoRV + HindIII for recombination of the 3′ arm. PCR amplicons based on DNA sequences immediately 5′ and 3′ of the targeting vector arm were then used as probes for Southern blotting, and identified embryonic stem cell clone 573 carried the predicted 5′ and 3′ restriction fragment. After genotypic confirmation, clone 573 was injected into C57BL/6J (B6) blastocysts and implanted into pseudopregnant B6 mice by The Jackson Laboratory Cell Biology and Microinjection Service. Chimeric mice and their germline transmitting progeny were typed by PCR using primers designed to distinguish the targeted from the wild-type (wt) FcRn alleles: FcRn o393 F-GGGATGCCACTGCCCTG, FcRn vs o394 R-CGAGCCTGAGATTGTCAAGTGTATT, 248-bp wt allele; targeting vector o395 F-GGAATTCCCAGTGAAGGGC vs o394, 378-bp targeted allele. Protein expression of both mouse FcRn and the human FcRn transgene (see below) was determined by Western blot analysis using an antiserum produced by immunizing FcRn-deficient mice with recombinant, bacterially expressed human FcRn in CFA, followed by boosts with the same protein in IFA. Protein extracts (25 μg) from neonatal small intestine were separated on a 4–20% gradient gel, transferred to polyvinylidene difluoride membrane, and incubated with antiserum against FcRn as a primary Ab and HRP-conjugated goat anti-mouse Ig (BD Biosciences, San Diego, CA) as a secondary Ab. β-actin was revealed using a rabbit anti-mouse β-actin antiserum (Sigma-Aldrich, St. Louis, MO) and a HRP-conjugated donkey anti-rabbit F(ab′)₂ (Amersham Biosciences, Piscataway, NJ) as a secondary Ab. Bands were visualized by ECL detection. To produce mice transgenically expressing a complete human FcRn gene, a 34-kb XhoI fragment was excised from a human-derived bacterial artificial chromosome (Genome Systems) and cloned into a SuperCos vector (Stratagene, La Jolla, CA). This fragment, which encompasses the complete ∼11-kb FcRn gene plus ∼10-kb 5′ and 3′, was microinjected into zygotes from C57BL/6J female mice, and implanted into pseudopregnant female mice. A chimeric, germline transmitting male founder was used to establish the C57BL/6J-human FcRn transgenic line 44. Expression of human FcRn was confirmed by RT-PCR (not shown) and by Western blot analysis. Unless otherwise indicated, mice ranged from 6 to 12 wk of age and were age and sex matched within each experiment. FcRn^−/− mice used as indicated were either mixed 129:B6 stocks, congenic B6 background (generated after five backcross generations), or isogenic 129/J background stocks. FcRn^−/− human FCRN transgenic mice were on a mixed 129:B6 genetic background.

Ig transfer and clearance

For maternal Ig transfer, a female FcRn^+/− mouse was mated to a male FcRn^−/− mouse. At 17 days of gestation, the pregnant female was injected i.v. with 500 μg IgG1 anti-trinitrophenyl (TNP) (mAb 1B7.11) tracer Ab. Sera from her 3-day-old pups were assayed for anti-TNP Ab activity by ELISA, and their liver DNA was used to determine their FcRn genotypes. Ig clearance in adult mice was determined as described (16). Briefly, 100 μg of mouse IgG1 anti-TNP, mouse IgA anti-TNP (mAb 2F.11.15), purified human IgG (GammaGuard; Baxter Healthcare, Deerfield, IL), or directly biotinylated mouse IgG1, IgG2a, IgG2b, and IgG3 (Sigma-Aldrich) tracer was injected into mice i.p., and sera prepared from retro-orbital plexus blood were assayed by ELISA. β-phase Ig tracer clearance was based on the amount of Ig retained normalized to that present 24 h after injection.

Serum Ig concentrations and flow cytometry

Total serum Ig and specific mouse isotypes were quantified as described (16, 22). Total Ig and isotype quantification was based on duplicate serum dilutions in which at least two dilutions hit the exponential curve of serially diluted Ig standards: Ig (Calbiochem, San Diego, CA), IgG1 (BD PharMingen, San Diego, CA), IgG2a (PharMingen), IgG2b (Southern Biotechnology, Birmingham, AL), IgG3 (Southern Biotechnology), IgM (BD PharMingen), and IgA (Southern Biotechnology). Data are expressed as equivalents of Ig compared with the Ab standards. For flow cytometry, splenocytes or bone marrow cells were isolated and incubated with propidium iodide and specific conjugated Abs at 4°C for 50 min in staining buffer (1× PBS with 0.1% BSA and 0.1% sodium azide), washed twice, and then analyzed with live cell gating using a FACSCalibur equipped with CellQuest software (BD Biosciences). Conjugated mAbs, CD4 CY3⁺, CD8a PE, CD45R/B220 APC, CD45R/B220 Cy5, CD19 FITC, and CD138 APC, were obtained from BD Pharmingen.

Quantitative PCR analysis and 5′ RACE

Tissues were collected into RNALater (Ambion, Austin, TX), and total RNA was purified using the RNAqueous 4-PCR kit (Ambion) and DNase treated. Synthesis of cDNA from 5–10 μl total RNA was conducted using the Retroscript kit (Ambion). For FcRn expression analysis, PCR primers MuFcRn.5F, 5′-CCTCAAGACCCTGGAGAAGATATTAA-3′; MuFcRn.5R, 5′-CCGTGGGCACTGAGGAATTA-3′; and a TaqMan probe MuFcRn.5P, 5′-6FAM-CTGCTGGGCTGTGAACTGGCCTC-TAMRA-3′, were used to amplify and detect a 94-bp cDNA fragment spanning FcRn exons 3 and 4. Reactions (in triplicate) containing the equivalent of 0.5 μl cDNA in the presence of TaqMan PCR Master Mix were run in duplex with the TaqMan 18S Ribosomal RNA Control (Applied Biosystems, Foster City, CA), and the ABI Prism 7700 was used for real-time detection. Mean Ct were calculated; the FcRn expression levels were determined via the ddCt method (Applied Biosystems User Bulletin 2); and expression levels are represented relative to skeletal muscle. The ImmunoQuantArray customized gene expression platform, including the list of genes analyzed and the Global Pattern Recognition (GPR) expression analytical algorithm, has been described⁴ (http://www.jax.org/research/roop/). The 5′ RACE of liver cDNA was performed using Ambion 5′ mRNA oligonucleotide adaptors and exon 5 FcRn outer and inner exon 4/5 junction reverse primers. No fragments were detected after 38 amplification cycles using the outer oligonucleotide primers, and only fragments with sizes inconsistent with the expected normal FcRn transcript size were detected with FcRn^−/− mice cDNAs after the 38-cycle amplification using the nested primers. These fragments were cloned, sequenced, and proved to be from transcripts unrelated to FcRn.

Specific Ab responses after immunization and ELISPOT

Mice were immunized and challenged with 100 μg DNP-keyhole limpet hemocyanin (KLH) (Calbiochem, San Diego, CA), and sera from retro-orbital plexus blood were assayed for anti-DNP activity, as described (16). For ELISPOT, MultiScreen-HA plates (Millipore, Bedford, MA) were coated with 10 μg/ml of either DNP-KLH, goat anti-mouse IgG, or 100 μg/ml BSA in ELISA coating buffer at 4°C overnight. The plates were then washed three times with PBS and blocked with DMEM medium containing 5% FBS for 1 h at room temperature. Serial dilutions (starting at 1 × 10⁶ cells/well) of spleen and bone marrow cells were added to the wells. The plates were incubated at 37°C with 5% CO₂ overnight and washed three times with PBS-Tween; Goat anti-mouseκ-alkaline phosphatase (0.5 μg/ml; Southern Biotechnology) was then added to each well. After a 1-h incubation at 37°C, the plates were washed three times with PBS-Tween, followed by PBS and chromogen substrate 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium (Sigma-Aldrich). The plates were then incubated in the presence of chromogen substrate at room temperature for 1 h, and the reaction was terminated by a water wash. Data are reported as the number of spots enumerated in the Ab-coated wells minus those obtained with the corresponding BSA-coated plates.

Costimulator blockade

B6-FcRn^−/− and B6 wt control mice were injected i.p. on day −3 with PBS or 20 μg purified CTLA-4 Ig (27) and 20 μg anti-CD40 ligand (CD40L) IgG1 (28) in PBS (kindly provided by B. Soper, The Jackson Laboratory), grafted 1 day later with allogeneic 129/J tail skin, and injected with 12 μg purified CTLA-4 Ig and 12 μg anti-CD40L IgG1 3 days later. Graft survival was scored according to the method of Bailey and Usama (29), and analyzed by Kaplin-Meier Analysis (StatView; Abacus Concepts, Berkeley, CA).

Production, verification, and overall health of FcRn-deficient mice

To address the role of FcRn in vivo, we used gene targeting to produce mice with a defective FcRn gene. Homologous recombination resulted in the insertion of a neomycin-resistance cassette and the deletion of a 1588-nt fragment that encompassed the promotor and extended through most of exon 2 to the ApaI site (Fig. 1⇓, A and B). Quantitative PCR of cDNA from livers indicated that FcRn^−/− mice produced a minimal signal (Fig. 1⇓C), and extensive 5′ RACE analysis failed to detect any FcRn-specific transcript (Fig. 1⇓D and data not shown). Western blot analysis of neonatal intestine extracts probed with an anti-FcRn antiserum failed to reveal any hybridized product (Fig. 1⇓E).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Production and verification of FcRn-deficient mice. A, FcRn targeting scheme (not to scale). B, PCR-based genotypic confirmation of targeted mice. C, Quantitative PCR of liver cDNA using PCR primers and a TAQman probe that revealed a cDNA fragment spanning FcRn exons 3 and 4. Values are reported relative to skeletal muscle. D, 5′ RACE of liver cDNA using a reverse FcRn primer originating in exon 3 yields the expected 920-bp fragment in wt mice. E, Western blot analysis of neonatal intestine extracts from FcRn-deficient and human FcRn transgenic mice revealed using anti-FcRn antisera or anti-β-actin antisera. Lanes 1–3, Extracts (25 μg/lane) from intestines of 10-day-old mice. Lane 4, Lung extract (25 μg/lane) from human FcRn-transgenic mouse line 44. Molecular weight markers are indicated in kDa.

Unlike β₂m-deficient mice, which are prone to perinatal lethality (D. Roopenian, unpublished data), FcRn^−/− mice showed no evidence of a failure to thrive. Adult FcRn^−/− and wt mice were indistinguishable based on weight and overt health. Histological analysis of liver, lungs, kidneys, and lymphoid organs of adult FcRn^−/− mice was normal (data not shown).

FcRn is the receptor required for perinatal IgG transport

To determine directly whether FcRn is responsible for perinatal transfer of maternal IgG, a female FcRn^+/− mouse was mated to a male FcRn^−/− mouse. In this way, the mother would be FcRn sufficient, while her progeny would be either FcRn^−/− or wt. To track maternal IgG transmission, the pregnant female was injected i.v. at 17 days of gestation with anti-TNP IgG1. Her pups were then nursed on the same FcRn^+/− mother. Her 3-day-old pups were genotyped for FcRn, and their sera were assayed for anti-TNP Ab activity by ELISA. Anti-TNP IgG1 was evident in sera from wt FcRn^+/− neonates, but lowered a minimum of 190-fold in the FcRn^−/− littermates (Fig. 2⇓A). Although human neonates acquire their maternal IgG by transplacental routes, neonatal mice acquire most maternal IgG by transintestinal transport (30). Our results suggest that this FcRn-null allele functionally ablated perinatal IgG transport. However, they do not distinguish between the maternofetal vs transintestinal routes.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

FcRn-deficient neonates are unable to absorb maternal IgG, and adult FcRn-deficient mice clear IgG rapidly. A, Anti-TNP IgG serum concentrations of 3-day-old pups born from an FcRn^+/− mother who had been injected i.v. with IgG1 anti-TNP tracer Abs on day 17 of gestation. −/− vs +/−, p < 0.00001 computed using the one-sided Student’s t test. B, Mouse anti-TNP IgG1 mAb and IgA anti-TNP Abs were injected i.p. into groups of four to five 6- to 7-wk-old mice, and the concentration of retained IgG tracer from serially sampled sera was determined. Data are representative of four independent experiments and are shown as the mean ± SEM of the anti-TNP normalized to the day 1 values. In some cases, the SEM was smaller than the plot symbols. Values of p (two-sided t test) for FcRn^+/− vs FcRn^−/− on days 2–5, p < 0.0001; β₂m^−/− vs FcRn^−/−, p < 0.005, day 5. C, FcRn^−/− mice, FcRn^+/− littermate controls (two to three per group) were injected i.p. with 25 μg directly biotinylated mouse IgG1, IgG2a, IgG2b, or IgG3. D, FcRn^−/− human FcRn transgenic line 44, FcRn^+/+, and FcRn^−/− mice were injected with i.p. purified human IgG tracer. Two to three mice/group were, and the data are representative of, three independent experiments. All mice except β₂m^−/− cohorts were a mixed 129:B6 genetic background.

FcRn is responsible for extending the serum IgG t_1/2 and for maintaining normal IgG levels

To determine whether FcRn is responsible for the extended t_1/2 of IgG, adult FcRn^−/− and FcRn^−/+, as well as β₂m-deficient mice, were injected with tracer anti-TNP IgG1 and IgA Abs, and serial blood samples were tested for the amount of tracer remaining. In FcRn wt mice, the t_1/2 of the IgG1 tracer (9 days) was predictably greater in comparison with the IgA tracer (1.4 days) (Fig. 2⇑B). In contrast, the IgG1 t_1/2 was considerably shortened in both FcRn^−/− (1.4 days) and β₂m^−/− mice (1.0 days). The small difference in FcRn^−/− vs β₂m^−/− t_1/2 was significant (p = 0.004 at day 5), and was observed in multiple experiments. FcRn protected all mouse IgG isotypes, because FcRn^−/− and β₂m^−/− mice demonstrated rapid clearance of biotinylated mouse IgG1, IgG2a, IgG2b, and IgG3 in comparison with wt mice (Fig. 2⇑C). Moreover, human IgG was cleared rapidly in FcRn^−/− mice, but it was protected in FcRn^−/− mice onto which a human FcRn transgene had been crossed (Fig. 2⇑D). These results indicated that both mouse and human FcRn play an important role in extending the t_1/2 of IgG.

To determine the effects of FcRn deficiency on maintaining homeostatic Ab levels, serum Ig concentrations were determined in FcRn^−/− and wt mice on an isogenic 129/J background. FcRn^−/− mice had reduced concentrations of all IgG isotypes compared with their wt sex-matched counterparts, but IgM and IgA concentrations did not differ significantly from wt controls (Fig. 3⇓). These results indicated that FcRn plays an important role in IgG homeostasis.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

FcRn-deficient mice have low serum IgG. Sera from 10-wk-old 129/J FcRn^−/− (n = 5) and FcRn^+/+ (n = 5) female and male mice were analyzed by ELISA for specific Ig isotypes. ∗, p < 0.01; ∗∗, p < 0.001; NS, p > 0.05. Values shown are the mean ± SEM.

An FcRn deficiency does not alter transcriptional and cellular immunological profiles

To determine whether an FcRn deficiency affects the normal immunological status, FcRn-deficient and wt mice were subjected to a battery of phenotypic tests. We first tested whether an FcRn deficiency affects the transcriptional profile of immunologically important genes. RNA samples isolated from spleens and lungs of three replicate 129 mice isogenic for FcRn^−/− and three 129-FcRn wt controls were analyzed using an ImmunoQuantArray, a quantitative PCR-based method that allows one to measure simultaneously the expression of 95 genes selected because they play significant roles in immune homeostasis and function.⁴ Of the genes surveyed, only FcRn showed a significant difference by the global pattern recognition analytical algorithm (Table I⇓). Thus, FcRn deficiency leads to negligible alterations of the transcriptional immunological phenotype.

Splenic total leukocyte numbers and leukocyte subsets, including plasma cell numbers, of spleens and bone marrow cells of FcRn^−/− and wt mice were indistinguishable by multiparameter flow cytometric analysis (Table II⇓ and data not shown). ELISPOT analysis of spleen and bone marrow cells failed to suggest an increase in total or anti-TNP-specific IgG-secreting plasma cells after immunization (Table III⇓). Spleen and lymph node cells from FcRn^−/− mice showed normal in vitro proliferative responses to the T cell mitogens, Con A, and anti-CD3 mAbs. Recall proliferative responses to OVA and hen egg lysozyme were indistinguishable from their wt controls (data not shown). Similarly, FcRn^−/− and wt spleen cells showed superimposable B cell proliferative responses to LPS and anti-IgM + anti-CD40 Abs (data not shown). Thus, despite lowered homeostatic serum IgG levels, FcRn deficiency does not lead to aberrations in the transcriptional status, cellular composition, or in vitro function of peripheral lymphoid cells.

FcRn is required for a robust IgG Ab response

To determine the consequences of an FcRn deficiency on a specific humoral immune response, FcRn^−/− and wt C57BL/6 (B6) and 129 background mice were immunized and then challenged 3 wk later with DNP-KLH, and their serum anti-DNP-specific Ig titers were determined. FcRn deficiency on B6 background mice reduced substantially the specific anti-DNP response regardless of IgG Ig isotype (Fig. 4⇓A). This decrease was not a consequence of numbers of plasma cells because TNP-KLH-immunized FcRn^−/− and wt mice showed similar numbers of anti-IgG- and anti-TNP-secreting cells by ELISPOT analysis (Table III⇑). On the 129/J background mice, the anti-TNP response was heightened overall, and an FcRn deficiency resulted in a reduction of all anti-DNP IgG isotypes, but it was more episodic compared with B6 background mice (Fig. 4⇓B). Moreover, following immunization, the titers of anti-TNP IgG Abs in 129-FcRn wt mice dropped at an unexpectedly rapid rate given functional FcRn-mediated protection. A more sustained lowering in FcRn^−/− mice was observed in total IgG concentrations (Fig. 4⇓C, lower right panel). FcRn protection is known to be saturable at high serum IgG concentrations (half-maximal concentration ∼5 mg/ml) (31). Immunization of 129 background mice may have induced sufficient hypergammaglobulinemia to exceed the FcRn saturation capacity, thus resulting in an attenuation of FcRn-mediated protection. Taken together, the data suggest that FcRn plays an important role in controlling the concentrations of serum IgG Abs in response to specific antigenic stimuli provided that the concentrations do not exceed the FcRn saturation threshold.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

FcRn is required for a robust humoral response. Groups of four to five B6 background (A) and 129 background (B) FcRn^−/− and wt FcRn^+/− or FcRn^+/+ mice were immunized i.p. with DNP-KLH in CFA and challenged 21 days later with DNP-KLH in IFA. Serially sampled sera were analyzed for DNP-specific (A, B) and total Ig (C, 129 only) by ELISA. Values shown are the mean ± SEM.

FcRn is required to achieve efficacy of IgG-Fc-conjugated therapeutics

It is well established that the covalent addition of an IgG-derived Fc moiety prolongs drug t_1/2 (3). However, whether this extension is solely dependent on FcRn as compared with other FcR is less clear. To determine directly whether FcRn is the molecule responsible for enhancing the pharmacological efficacy of IgG-coupled therapeutics, we addressed whether mice lacking FcRn were refractory to the immunosuppressive effects of a costimulatory blockade including CTLA-4 Ig and anti-CD40L (both of which possess IgG Fc moieties). B6-FcRn^−/− and wt controls were treated with a cocktail of CTLA-4 Ig and anti-CD40L IgG1, and grafted with allogeneic 129/J tail skin (Fig. 5⇓). The rate of allograft rejection by FcRn-deficient mice was retarded considerably compared with that of FcRn wt controls, and similar to that observed in mice that received PBS rather than the costimulatory blockade. These results indicated that FcRn is the FcR responsible for the enhanced pharmacological efficacy of IgG-Fc-conjugated drugs.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

FcRn is required for efficacy of IgG-Fc-conjugated therapeutics. Groups of four or five mice were injected i.p. on days −1 and +3 with CTLA-4 Ig and rat anti-CD40L IgG1, and grafted on day 0 with allogeneic 129/J tail skin. Syngeneic allografts were retained throughout the course of the experiments (not shown).