Quantitative cumulative biodistribution of antibodies in mice

Abstract

The neonatal Fc receptor (FcRn) plays an important and well-known role in antibody recycling in endothelial and hematopoietic cells and thus it influences the systemic pharmacokinetics (PK) of immunoglobulin G (IgG). However, considerably less is known about FcRn’s role in the metabolism of IgG within individual tissues after intravenous administration. To elucidate the organ distribution and gain insight into the metabolism of humanized IgG1 antibodies with different binding affinities FcRn, comparative biodistribution studies in normal CD-1 mice were conducted. Here, we generated variants of herpes simplex virus glycoprotein D-specific antibody (humanized anti-gD) with increased and decreased FcRn binding affinity by genetic engineering without affecting antigen specificity. These antibodies were expressed in Chinese hamster ovary cell lines, purified and paired radiolabeled with iodine-125 and indium-111. Equal amounts of I-125-labeled and In-111-labeled antibodies were mixed and intravenously administered into mice at 5 mg/kg. This approach allowed us to measure both the real-time IgG uptake (I-125) and cumulative uptake of IgG and catabolites (In-111) in individual tissues up to 1 week post-injection. The PK and distribution of the wild-type IgG and the variant with enhanced binding for FcRn were largely similar to each other, but vastly different for the rapidly cleared low-FcRn-binding variant. Uptake in individual tissues varied across time, FcRn binding affinity, and radiolabeling method. The liver and spleen emerged as the most concentrated sites of IgG catabolism in the absence of FcRn protection. These data provide an increased understanding of FcRn’s role in antibody PK and catabolism at the tissue level.

Keywords: :

• FcRn
• biodistribution
• pharmacokinetics
• radiolabeled
• metabolism
• indium

Introduction

Advances in monoclonal antibody (mAb) engineering, including phage-display technologies, have enabled the development of humanized, i.e., mouse complementarity determining region (CDR) residues grafted on to a human framework residues, as well as fully human antibodies with desirable pharmacokinetic (PK) and biodistribution properties.¹ These engineered mAbs have been mass-produced as highly specific reagents used across many clinical settings, including oncology, inflammation, infectious diseases, transplantation and cardiovascular medicine.² The unique CDR usually defines an antibody’s binding site, i.e., antigen specificity, and resides in the variable fragment (Fv) portion of the molecule.

In immunoglobulin G (IgG), the Fc domain regulates antibody-based therapy by binding to Fc gamma receptors (FcγR) and eliciting immune effector functions.³ Another characteristic of the Fc portion, specifically the CH2-CH3 domain, has the ability to bind to the neonatal Fc receptor (FcRn) with high affinity at acidic pH, but with very low affinity at neutral, i.e., plasma, pH.^4,5 This scenario enables IgG to be bound by FcRn in the acidic environment of the sorting endosome (pH ≤ 6.0) in endothelial cells and to be recycled back into circulation where dissociation occurs under more alkaline (pH > 7) conditions.⁶ The population of IgG that is not bound to FcRn in the endosome undergoes transport to, and proteolysis in, the lysosomes.⁵ It has been well-established that the strict pH dependence of this IgG/FcRn interaction is responsible for protecting mAbs from degradation.^7,8 Consequently, FcRn plays an important role in IgG catabolism^5,9 and accounts for the persistence of IgG in the circulation of rodents and higher species.^10,11 Also known as the major histocompatibility complex class I-related receptor, FcRn possesses additional roles for protecting serum albumin from degradation¹² and in the passive delivery of IgG from mother to young, i.e., IgG in milk absorbed through gut.¹³ Accordingly, FcRn is expressed widely in endothelial, epithelial and hematopoietic cells including organs such as skin, muscle, kidney, liver, and placenta.^5,14-16

The fact that half-life of an IgG in vivo can be influenced by altering its binding affinity to FcRn at different pH is well-established.⁵ Several studies have shown a correlation between the serum half-life and binding affinity of IgGs for FcRn.^10,17-19 However, this correlation is not always observed, as different human IgG subclasses exhibit different in vivo half-lives, but appear to bind human FcRn with K_D values that are within the same order of magnitude.^20,21 A possible explanation of these discrepancies is that the ratio of IgG protected/destroyed depends not only on the affinity of FcRn, but also on the competition between the binding of IgG to FcRn and its intracellular degradation prior to its interaction with FcRn.⁹ It has also been reported that several IgG Fc variants that bind stronger to FcRn at pH > 7 have shorter in vivo half-lives compared with the corresponding wild-type (WT) IgG because they cannot dissociate from FcRn.^22,23 An ideal IgG Fc variant should possess increased FcRn affinity at acidic pH compared with a WT antibody, thus retaining binding while in the endosome, and having weaker or equivalent affinity at pH > 7 to facilitate release back into systemic circulation.^10,22,23

PK and tissue biodistribution studies play an important role in determining specific and non-specific tissue accumulation of radiolabeled IgGs in rodent models. Disposition characteristics are important parameters to consider when designing and testing novel mAbs because they can help predict or explain target-mediated disposition, toxicity, or off-target effects. Although the effects of FcRn binding on systemic PK are well-established, considerably less is known about the catabolic fates of FcRn binding variants within individual tissues. Instead of directly affecting IgG distribution from blood to tissues, i.e., target-mediated clearance, cellular FcRn is expected to influence the intracellular fates of the endocytosed antibodies based on their relative abilities to bind FcRn and thus be rescued from lysosomal degradation. To examine the potential effects of altering FcRn binding affinity on IgG tissue PK and metabolism in more detail, we engineered two humanized anti-glycoprotein D Fc variants (H310Q and T307Q/N434A, EU numbering) with different binding affinities to mFcRn and compared their biodistributions to that of the WT antibody in mice. Furthermore, we chose to evaluate these antibodies radiolabeled with both iodine-125 (non-residualizing) and indium-111 (residualizing) radionuclides to obtain information regarding both distribution and metabolism. The differences in biodistribution patterns of these two labeling methods is driven by their cellular metabolisms, after which I-125-containing catabolites undergo cellular efflux and systemic (primarily urinary) clearance whereas the more polar In-111 catabolites (mostly metal-chelate-amino acid adducts) have an enhanced retention within cells. Given the importance of pH-dependent interactions to the efficient recycling of an IgG, these studies have been designed to gain a better understanding of the relationships among FcRn binding affinity, systemic PK, biodistribution, and cellular metabolism of antibodies in tissues.

Results

mFcRn affinity assessment by surface plasmon resonance

The T307Q/N434A variant (designated henceforth as FcRn⁺), showed a roughly 5-fold increase in binding affinity to murine FcRn relative to the wild-type anti-gD (IgG1 WT), consistent with previous findings for another antibody bearing the same mutation (Table 1).¹⁰

Table 1. Affinity measurements of anti-gD variant binding to mFcRn by surface plasmon resonance

Variant	kon (M^−1s^−1)	koff (s^−1)	KD (nM)
IgG1 WT	6160	0.0037	601
IgG1 FcRn^+	19100	0.0021	108
IgG1 FcRn^-	1230	0.1146	93000

Radiochemistry

The radiochemical yields of the ¹²⁵I-labeled antibodies used in the studies were 68% for WT, 66% for FcRn⁺ and 61% for FcRn^-. The specific activities of the ¹²⁵I-labeled antibodies used in the studies were 10.00 μCi/μg for WT, 10.49 μCi/μg for FcRn⁺ and 9.43 μCi/μg for FcRn^-. The radiochemical yields of the ¹¹¹In-labeled antibodies used in the studies were 72% for WT, 76% for FcRn⁺ and 79% for FcRn^-. The specific activities of the ¹¹¹In-labeled antibodies used in the studies were 17.0 μCi/μg for WT, 14.0 μCi/μg for FcRn⁺ and 20.8 μCi/μg for FcRn^-. Size-exclusion HPLCs of all radiolabeled antibodies are included in the Supporting Information.

Plasma pharmacokinetics

After intravenous administration of human IgG1with various binding affinities to murine FcRn, the plasma concentrations of the WT and FcRn⁺ variant declined in a biphasic manner, whereas the FcRn^- variant’s clearance was more rapid and at least biphasic (possibly triphasic). The plasma concentrations, expressed in percentage of injected dose per milliliter (%ID/mL), at 15 min post-dose were similar across antibodies. Specifically, concentrations of the ¹²⁵I-labeled antibodies were 78 ± 8, 74 ± 8, and 68 ± 13%ID/mL for IgG1 WT, FcRn⁺, and FcRn^-, respectively (Fig. 1A). Similarly, concentrations of the respective ¹¹¹In-labeled antibodies were 76 ± 6, 70 ± 7, and 68 ± 13%ID/mL (Fig. 1B). By 24 h post-dose, plasma concentrations of the FcRn^- variant had decreased more rapidly than IgG1 WT and FcRn⁺, with a concentration of 8.8 ± 0.5%ID/mL for ¹²⁵I-labeled FcRn^- compared with 27 ± 2 and 26 ± 3%ID/mL for IgG1 WT and FcRn⁺, respectively (Fig. 1A). A similar trend was observed in the plasma concentrations of ¹¹¹In-labeled variants and continued to the end of the study (Fig. 1B). The differences between IgG1 WT and FcRn⁺ were insignificant as the plasma concentration at 14 d post-dose showed similar values. For instance, the concentrations of ¹²⁵I-labeled IgG1 WT and FcRn⁺ were 6.2 ± 5.2 and 6.6 ± 3.2%ID/mL, respectively (Fig. 1A). The respective values for the ¹¹¹In-labeled variants were 7.3 ± 6.5 and 7.9 ± 2.6%ID/mL (Fig. 1B).

Figure 1. Plasma pharmacokinetic curves of the IgG1 WT, FcRn⁺, and FcRn^- variants radiolabeled with (A) iodine-125 and (B) indium-111.

The plasma exposures were generally similar between IgG WT and FcRn⁺, but different for FcRn^-. Specifically, the plasma AUC_0–7 values from ¹²⁵I data were 164 ± 5, 169 ± 14 and 43 ± 1%ID/mL for IgG1 WT, FcRn⁺, and FcRn^-, respectively (Table 2). In a similar trend, plasma AUC_0–7 values derived from ¹¹¹In data were 182 ± 3, 182 ± 10 and 95 ± 1%ID/mL for IgG1 WT, FcRn⁺, and FcRn^-, respectively. The PK of ¹²⁵I- and ¹¹¹In-labeled variants corresponded generally well with each other except in the case of the FcRn^- variant, for which the ¹¹¹In-labeled variant appeared to clear more slowly than the ¹²⁵I-labeled variant.

Table 2. Tissue AUC_0–7 (day*%ID/mL) for anti-gD variants

Variant	Label	Plasma	Lungs	Liver	Kidney	Heart	Spleen	Stomach	Large Intestine	Small Intestine	Brain	Muscle	Skin
IgG1 WT	^125I	164 ± 5.38	21.5 ± 1.20	15.5 ± 0.575	15.4 ± 2.14	15.5 ± 0.692	14.6 ± 0.340	8.01 ± 0.425	7.09 ± 0.591	6.36 ± 0.268	1.55 ± 0.0900	5.43 ± 0.0946	21.5 ± 1.53
	^111In	182 ± 2.76	28.3 ± 1.34	37.2 ± 1.60	32.1 ± 2.45	18.5 ± 0.758	31.6 ± 2.00	9.79 ± 0.575	10.1 ± 0.649	9.16 ± 0.502	2.09 ± 0.0923	6.85 ± 0.243	26.6 ± 1.15
IgG1 FcRn^+	^125I	169 ± 14.0	22.8 ± 3.30	22.1 ± 1.30*	16.1 ± 2.10	16.3 ± 1.25	18.8 ± 1.32*	8.18 ± 0.401	6.92 ± 0.636	6.55 ± 0.626	2.00 ± 0.122*	5.72 ± 0.353	25.2 ± 1.82
	^111In	182 ± 9.97	28.3 ± 2.85	37.8 ± 1.48	31.2 ± 2.02	18.3 ± 1.09	38.1 ± 1.95*	9.63 ± 0.323	10.0 ± 0.628	9.35 ± 0.971	2.59 ± 0.139*	6.67 ± 0.449*	30.7 ± 1.26*
IgG1 FcRn^-	^125I	43.1 ± 1.16*	6.63 ± 0.500*	3.77 ± 0.220*	5.46 ± 0.131*	4.46 ± 0.407*	4.60 ± 0.544*	3.63 ± 0.100*	1.76 ± 0.132*	2.00 ± 0.104*	0.58 ± 0.0485*	1.52 ± 0.124*	5.89 ± 0.490*
	^111In	94.5 ± 1.46*	17.8 ± 0.821*	60.0 ± 2.41*	25.6 ± 1.02*	16.8 ± 0.923	53.3 ± 7.16*	7.58 ± 0.278*	11.1 ± 1.00	10.7 ± 0.598*	1.30 ± 0.121*	4.01 ± 0.423*	22.6 ± 3.41

Significant (P < 0.05) differences by unpaired t test are indicated by asterisks and are relative to IgG WT for the corresponding radionuclide.

Tissue distribution

The radioactivity measured from each tissue was calculated into the percentage of injected dose normalized by the tissue weight in grams (%ID/g). The tissue distribution of the non-residualizing ¹²⁵I-labeled antibodies at 6 h post-dose revealed distribution to multiple blood rich tissues such as liver, lungs, kidneys and heart (Fig. 2). However, there were slight differences between the three variants, especially for the FcRn^- variant that also showed significantly faster blood clearance. The tissue distributions of the ¹¹¹In-labeled antibodies at 6 h post-dose were similar to that of the ¹²⁵I-labeled antibodies for all variants (Fig. 2A and B). By 24 h post-dose, all of the collected tissues for the ¹²⁵I-labeled FcRn^- variant contained significantly lower radioactive uptake compared with the IgG1 WT variant (Fig. 2C). This same observation for the ¹²⁵I-labeled FcRn^- variant persisted at 168 h post-dose (Fig. 2E). Fewer significant differences were observed between the ¹¹¹In-labeled WT and FcRn^- variant relative to the ¹²⁵I-labeled counterparts, especially at 24 h (Fig. 2D); this reflects the fact that, unlike ¹²⁵I, ¹¹¹In is a residualizing label that accumulates within tissues long after the lysosomal degradation of the original radiolabeled molecule. Because of the residualizing effect of the ¹¹¹In-DOTA complex, most tissue levels of ¹¹¹In remained relatively constant between 6 and 24 h (Fig. 2A-D). Splenic uptake of the ¹¹¹In-labeled FcRn^- variant increased from 6.0 ± 1.6%ID/g at 6 h to 9.7 ± 3.5%ID/g at 24 h (Fig. 2B and D). Considerably higher uptake of ¹¹¹In than ¹²⁵I was observed in many tissues, particularly the liver and spleen, at 168 h post-dose, and this was especially true for the FcRn^- variant (Fig. 2E and F). Overall, the trends in tissue distribution between the FcRn⁺ and IgG1 WT were largely similar irrespective of labeling method (Fig. 2). An interesting exception, however, was that the radioactive uptake of both ¹²⁵I and ¹¹¹In in brain at 168 h was roughly 2-fold higher for the FcRn⁺ variant (¹²⁵I: 0.25 ± 0.24%ID/g and ¹¹¹In: 0.41 ± 0.05%ID/g) than for the IgG WT (¹²⁵I: 0.12 ± 0.06%ID/g and ¹¹¹In: 0.22 ± 0.06) (Fig. 2E and F, insets). However, no such difference in brain uptake between the IgG WT and FcRn⁺ variants was observed at 24 h (Fig. 2C and D, insets).

Figure 2. Tissue distributions at 6 (A and B), 24 (C and D), and 168 (E and F) hours of the IgG1 WT, FcRn⁺, and FcRn^- variants radiolabeled with iodine-125 (A, C and E) and indium-111 (B, D and F). Asterisks represent significant (P < 0.05) differences relative to WT by unpaired t test.

The distribution data was further analyzed to estimate the area under the tissue concentration–time curve from zero to seven days (AUC_0–7) for each tissue. The AUC_0–7 values generated by the ¹²⁵I data were similar between IgG WT and FcRn⁺ in all tissues while FcRn^- was significantly lower (Table 2). A similar trend was observed for the ¹¹¹In data, except that uptake of radioactivity in the liver and spleen was much higher for the FcRn^- compared with the IgG WT and FcRn⁺ (Table 2). This is a critical finding since the liver and spleen are both FcRn-expressing tissues that would afford protection from metabolism to FcRn-binding antibodies. This opposite trend between ¹²⁵I and ¹¹¹In uptake data is further highlighted from the ratios of AUC_0–7 from the liver and spleen compared with that of the plasma; these ratios were 0.64 and 0.56, respectively, whereas ratios of < 0.3 were observed for all other tissues (Fig. 3). Furthermore, kidneys, heart, large intestine, small intestine, and skin are also FcRn-expressing organs. They would metabolize the FcRn^- variant much more readily than WT IgG or the FcRn⁺ variant, and this is indeed confirmed by the larger tissue:plasma (AUC_0–7) ratios in terms of the cumulative ¹¹¹In signal (Fig. 3B).

Figure 3. Ratios of tissue AUC_0–7 to plasma AUC_0–7 of the IgG1 WT, FcRn⁺, and FcRn^- variants radiolabeled with (A) iodine-125 and (B) indium-111. The abbreviation AUC_0–7 denotes the area under the concentration-time curve from 0 to 7 d. Note that AUC values for indium-111-labeled antibodies do not reflect true exposure of intact antibody since indium-111 is a residualizing radiometal label (Fig. 4).

Discussion

Characterization of the tissue PK and metabolism of IgGs is essential for understanding the mechanisms regulating the steady-state concentrations of antibodies in the circulation.²⁴ Accordingly, our experiments using indium-111-labeled IgGs provide a more direct approach for identifying the sites of cumulative distribution and potential catabolism of IgG under physiological conditions. It is difficult to argue from the ¹²⁵I-labeled IgG data alone that any one tissue has a dominant role in the metabolism of IgG, since it is simultaneously proposed that ¹²⁵I-labeled degradation products are being rapidly exocytosed or diffused from this tissue.²⁵ As such, alternative residualizing labels for studying IgG distribution, which are more efficiently retained in tissues, should permit a more definitive assessment of the role of various tissues in IgG tissue catabolism.²⁶ Indium-111 labeling via the chelating agent, DOTA, has been widely used in tissue distribution studies of proteins.²⁷ The ¹¹¹In-DOTA label is entirely satisfactory as a residualizing radiolabel; therefore PK analyses of tissue distribution of ¹¹¹In-DOTA-IgG complex have been developed. In this approach, after catabolism, radioactive metabolites that are charged²⁸ and hydrophilic are unable to pass through biological barriers such as the plasma and lysosomal membranes, promising the retention of radioactivity within the cells that have taken up the radiolabeled compound.

An important aspect of FcRn is that its biological function, at least in terms of IgG interaction, is essentially limited to the intracellular environment due a strict pH dependence. Even though the role of FcRn as a recycling receptor suggests that it does indeed come in contact with the extracellular fluid, it is only to release IgG (pH > 7), whereas binding can only occur in lysosomes or other endocytic compartments (pH ≤ 6) (Fig. 4). Several studies have now demonstrated that maintaining low binding affinity of IgG to FcRn at pH 7.4 is as essential as the increased binding affinity at pH ≤ 6.0 for efficient recycling and serum persistence of IgG.^10,22,23 FcRn protects IgG from intracellular catabolism and has been shown to be capable of transporting IgG across cell monolayers in both the apical-to-basolateral and basolateral-to-apical direction.^5,29 Accordingly, unlike many cell surface receptors recognized by antibody complementarity-determining regions (CDRs), cellular FcRn should not be expected to influence the targeting of IgG to most tissues. As such, one should assume that all three IgG1 variants, WT, FcRn⁺, and FcRn^-, should possess equal abilities to distribute to tissues and enter cells, via blood vessel extravasation and fluid pinocytosis, respectively. Any differential readout in radioactivity should therefore be driven by the intracellular fates of the endocytosed antibodies based on their relative abilities to bind FcRn and their ensuing cellular trafficking.

Figure 4. Schematic diagram of the cellular processing of radiolabeled IgG in the presence and absence of FcRn binding affinity. Both ¹²⁵I- and ¹¹¹In-DOTA-labeled antibodies enter the cell by pinocytosis. In the sorting endosome, FcRn-protected IgG is recycled back to the extracellular space, while non-protected IgG proceeds to lysosomal proteolytic degradation. The radiolabeled catabolite, ¹¹¹In-DOTA-lysine, is residualizing due to its polarity or charge and tends to accumulate within cells. In contrast, radioiodinated protein catabolites are rapidly effluxed from cells.

The plasma PK data for the IgG1 WT, FcRn⁺, and FcRn^- variants (Fig. 1, Table 2) were consistent with previously reported data for similar variants.^4,30 The lack of improvement in PK for the FcRn⁺ IgG variant relative to WT could be due to the species-specific nature of FcRn binding.^31,32 Specifically, human IgG is bound by murine FcRn with higher affinity than by human FcRn. As a result, the enhancement of receptor binding for the FcRn⁺ variant is more modest for mFcRn than for hFcRn (6-fold and 10-fold, respectively).¹⁰ In addition, the possibility exists that the FcRn⁺ variant exhibits increased binding to mFcRn at both acidic and basic pH, resulting in incomplete dissociation of mFcRn-IgG complexes at the cell surface following recycling.

The ¹¹¹In-labeled FcRn^- variant appeared to clear more slowly from plasma than the corresponding ¹²⁵I-labeled variant. We speculate that this phenomenon is due to transchelation of In(III), having similar coordination properties as Fe(III), from DOTA into transferrin (or other iron-containing plasma proteins) during the period of rapid distribution to tissues and subsequent pinocytosis. Although the thermodynamic stability of In(III)-DOTA is higher than In(III)-transferrin,³³ cellular metabolism of the extremely rapidly clearing FcRn^- variant may produce sufficiently high local concentrations of ¹¹¹In-containing catabolites to kinetically favor partial transchelation from DOTA to transferrin, especially within acidic lysosomes. However, for the vast majority of antibodies having typical PK, we have observed very similar blood clearance data for ¹¹¹In- and ¹²⁵I-derived molecules.^27,34,35

As for the tissue distribution data, our major finding that liver and spleen are the major sites of IgG catabolism in rodents is consistent with earlier work.²⁴ It should be noted, however, that although liver and spleen are important on a tissue mass-normalized basis, the skin and muscle are actually more important since they comprise a greater percentage of body weight.³⁶ Relatively few studies have been reported on the effect of FcRn on tissue distribution. Ferl and colleagues described the use of ¹¹¹In- and ¹²⁵I-labeled single-chain Fv-Fc antibody fragments, concluding that liver, skin, and muscle play pivotal roles in IgG catabolism.³⁷ Jaggi and coworkers explored the tissue distribution in a limited number of tissues (liver, kidney, spleen) of an IgG in both wild-type and FcRn^−/− mice, reporting a trend consistent with our findings of decreased renal uptake and increased hepatic uptake of radioactivity in the absence of FcRn protection at a comparable time point (160 h).³⁸ Garg and Balthasar reported the tissue concentrations of an IgG in wild-type and FcRn^−/− mice, but the study was limited to a non-residualizing ¹²⁵I labeling method.³⁹

Abuqayyas and Balthasar recently reported a lack of evidence for a role of FcRn in antibody distribution to the brain using an ¹²⁵I-labeled antibody in wild-type vs. FcRn^−/− mice.⁴⁰ This is consistent with an earlier report from the same group.⁴¹ Although our current data showed higher brain uptake at 168 h for the FcRn⁺ variant relative to WT in terms of both ¹²⁵I and ¹¹¹In (see Fig. 2E and F, insets), this trend was absent at other time points and did not translate into higher overall brain exposures (Fig. 3, insets). Others have postulated that higher brain concentrations of variants with enhanced binding to FcRn could result from a role in efflux of IgG, as opposed to influx.^42,43 Further evidence toward this notion is the fact that brain expression of FcRn is co-localized with the Glut1 glucose transporter in the capillary endothelium, suggesting that FcRn is expressed in the proper location to potentially mediate reverse transcytosis of IgG from brain to blood.⁴⁴

A physiologically-based PK (PBPK) model incorporating the protective role of FcRn predicted that the skin is one of the major organs responsible for IgG catabolism in mice, accounting for approximately 33% of the total IgG elimination.³⁹ This finding is also supported by evidence of FcRn expression on the endothelium of the large vascular beds of skin in mice.¹⁵ Indeed, our data for both the wild-type IgG and enhanced FcRn binding variant demonstrate that the skin has roughly triple the uptake of muscle (Fig. 2), despite the two tissues having a similar vascular volume. In the absence of FcRn protection, however, the liver and spleen emerge as the major sites of antibody catabolism.

In conclusion, we augmented the FcRn binding affinity of herpes simplex virus glycoprotein D-specific antibody (humanized anti-gD) by genetic engineering to measure both the real-time IgG uptake (I-125) and cumulative uptake of IgG and catabolites (In-111) in individual tissues up to 7 d post injection. The PK and distribution of the wild-type IgG and the variant with enhanced binding for FcRn were largely similar to each other, but vastly different for the low-FcRn-binding variant due to its very low systemic exposure. Uptake in individual tissues varied across time, FcRn binding affinity, and radiolabeling method. The liver and spleen emerged as the most concentrated sites of IgG catabolism in the absence of FcRn protection. These data provide an increased understanding of FcRn’s role in antibody tissue PK and catabolism.

Materials and Methods

mFcRn affinity assessment by surface plasmon resonance

The binding affinity of the high affinity Soluble murine FcRn (mFcRn) was expressed and purified from Chinese hamster ovary cells as reported previously.¹⁰ The binding affinity of mFcRn for WT IgG1 and the high- and low-affinity FcRn binding variants was measured at pH 5.8 by surface plasmon resonance using a Biacore T-200TM instrument (GE Healthcare). The variants were immobilized onto a CM5 Series S sensor chip and 12 serial dilutions of mFcRn were injected for 60 s at a flow rate of 50 μL/min. Following each mFcRn injection the surface was regenerated by two 30 s injections of pH 8.0 buffer to dissociate bound mFcRn. Raw sensogram data were reduced and referenced using the Scrubber II software package (BioLogic Software) and fit to a simple 1:1 binding model.

Antibody variants

The humanized IgG1 anti-herpes simplex virus glycoprotein D antibody (wild-type anti-gD, designated henceforth as IgG1 WT) was obtained from Genentech, Inc. as previously described.³⁴ Two antibody variants were produced from the wild-type IgG1 based on knowledge that Fc residues Thr307, Asn434 and His310 are all critical for IgG interaction with FcRn in mice.^8,45-47 A variant with enhanced binding affinity for FcRn, T307Q/N434A (designated henceforth as FcRn⁺), was selected based on previous demonstration of a 5-fold increase in binding affinity to murine FcRn.¹⁰ A null binding variant, H310Q (designated henceforth as FcRn^-), was also selected since His310 is highly conserved and important for the pH-dependent binding to FcRn across different species.¹⁰

Radiochemistry

Antibodies were conjugated to 1,4,7,10-tetraazacyclododecane-N,N’,N,”N”’-tetraacetic acid (DOTA) for indium-111 (¹¹¹In) complexation by random modification of lysine residues as previously described.^34,35,48 Antibodies were radioiodinated using an indirect iodogen addition method.⁴⁹ The radiolabeled proteins were purified using NAP5™ columns (GE Healthcare Life Sciences, cat. 17–0853–01) pre-equilibrated in PBS. All radiolabeled antibodies were characterized by size-exclusion high-performance liquid chromatography with in-line radiometric and UV detectors to compare the profiles of radioimmunoconjugates and corresponding unlabeled antibodies.

Tissue distribution and pharmacokinetics

All animals were handled in accordance with IACUC guidelines. Female CD1 mice ranging from 6 to 8 wk old and weighing approximately 20 g at the initiation of the study were randomized into groups. At 1 and 24 h prior to dosing, all mice received a 100 μL intraperitoneal injection of 30 mg/mL sodium iodide to prevent iodine-125 (¹²⁵I) sequestration in the thyroid. Equal amounts of radioactivity for ¹²⁵I-labeled and ¹¹¹In-DOTA-labeled antibodies (5 μCi each) were mixed with the corresponding unlabeled antibody at 5 mg/kg and administered intravenously by bolus tail vein injection. Whole blood was collected via retro-orbital bleed under isoflurane anesthesia at pre-selected time points up to 14 d (n = 3) from staggered mice. Blood samples were processed for plasma and stored at -70 °C until analysis. The following tissues were harvested terminally at 6 h, 1 d and 7 d: liver, lungs, kidneys, heart, spleen, stomach, small intestine, large intestine, brain, skin, and muscle. Tissues were also collected at 14 d, but these data are not reported because accurate counting of ¹¹¹In (2.8 d decay half-life) was difficult in some groups/tissues at this late time point. All tissues were rinsed with phosphate buffered saline, blot-dried, weighed, frozen on dry ice, and stored at -70 °C until radiometric analysis. It should be noted that neither perfusion nor mathematical correction of residual blood from tissues was performed.

Radioactivity counts were obtained from plasma and tissue samples using a Wallac 1480 Wizard 3 automatic gamma counter (EC&G Wallac, cat. 1480–011). Radioactive counts were converted to dose-normalized concentrations by calculating the percentage of injected dose per gram of tissue (or milliliter of blood or plasma). Plasma concentration–time data was used to estimate the area under the curve from zero to seven days (AUC_0–7) by using a non-compartmental analysis method (NCA) using sparse sampling with WinNonlin software (v 5.2.1 Pharsight Corporation). Calculation of AUC_0–7 in lieu of AUC_0–14 was performed for two reasons: (1) low counts of ¹¹¹In (2.8-d decay half-life), with selected groups approaching the detection limit of the gamma counting assay; and (2) likely (but unconfirmed) anti-therapeutic antibody (ATA) formation leading to fast clearance (signified by low levels of radioactivity in plasma and tissues) in selected animals. Tissue concentration–time data was also analyzed by using NCA with sparse sampling with WinNonlin software to estimate the area under the curve from zero to seven days (AUC_0–7). Standard errors in AUC_0–7 were calculated by the method of Bailer.⁵⁰

Disclosure of Potential Conflicts of Interest

All authors are employees of Genentech, a member of the Roche Group, and hold financial interest in Hoffmann-La Roche. All financial support was provided by Genentech.

Acknowledgments

We thank Andy Yeung, Henry Lowman, Rong Deng, Jaime Anguiano and Simon Williams for helpful discussions. We also acknowledge Sheila Ulufatu, Jason Ho, Shannon Stainton, Bernadette Johnstone, Cynthia Young, Kirsten Messick, Jose Imperio, Maria Duran, Mayra Monett, Nina Ljumanovic, Roxanne Vega, Nicole Valle, and Michelle Schweiger for animal studies support.