Embryo-fetal distribution of a biopharmaceutical IgG2 during rat organogenesis

Reproductive Toxicology 34 (2012) 66–72

Contents lists available at SciVerse ScienceDirect

Reproductive Toxicology journal homepage: www.elsevier.com/locate/reprotox

Embryo-fetal distribution of a biopharmaceutical IgG2 during rat organogenesis C.J. Bowman a,∗ , L.E. King b , D.B. Stedman a a b

Drug Safety Research & Development, Pfizer Worldwide Research & Development, Groton, CT, United States Pharmacokinetics, Dynamics & Metabolism, Pfizer Worldwide Research & Development, Groton, CT, United States

a r t i c l e

i n f o

Article history: Received 16 February 2012 Received in revised form 20 March 2012 Accepted 12 April 2012 Available online 20 April 2012 Keywords: Placental transfer Biodistribution Monoclonal antibodies Biopharmaceutical IgG

a b s t r a c t Embryo-fetal biodistribution of a maternally administered humanized IgG2 in rats was evaluated by enzyme-linked immunosorbent assay in dose response and time course studies. Fetal and maternal plasma IgG2 levels increased with dose from 10 to 300 mg/kg but fetal:maternal ratio decreased with increasing dose. Plasma IgG2 levels decreased in fetal rat with increasing time post-dose but more slowly than maternal levels. This difference in post-dose kinetics resulted in an increased fetal:maternal ratio with increasing days since last dose. Lastly, IgG2 in embryo-fetal tissue was detected at very low levels on gestation day (GD) 10–12 and levels increased over 100-fold by GD 17. The profile of increasing IgG2 levels as gestation progressed continued in extra-embryonic fluid (GD 12–19) until the end of gestation in fetal plasma (GD 19–21). Based on the current study, there is a potential for direct effects on rat embryo-fetal development following maternal administration of a biopharmaceutical IgG2. © 2012 Elsevier Inc. All rights reserved.

1. Introduction In primates (humans and monkeys) the majority of maternal IgG are transferred prenatally to offspring starting with low amounts mid-gestation rising to maternal amounts by birth with very little postnatal transfer [1]. In contrast, maternally derived IgG is transferred to offspring both pre- and postnatally in rodents [2]. Early reports indicate that only 6% of maternally derived IgG is present at birth in neonates but this rises substantially in lactation suggesting that the majority of maternally derived IgG is transmitted postnatally in rodents [3]. These data indicate that the rodent appears to underestimate the prenatal transfer of IgGs in humans and thus use as a nonclinical species in developmental toxicity safety assessment for humans would be limited [4]. There are very few data on maternal–fetal transfer of IgGs in any species but general patterns have emerged and have been summarized in recent review papers [4,5]. In general, data within species have been consistent, with the exception of data in rodents during gestation. Specifically, there is evidence that human IgG administered to pregnant mice during gestation transfers to embryos as early as GD 11 and increases rapidly thereafter [6]. Increasing fetal levels of a human Fc fusion protein in rats on GD 18 and 21 [7] and a mouse IgG2a in mice toward the end of organogenesis (GD 14) is consistent with these data [8]. These data indicate rat embryos may

∗ Corresponding author at: Pfizer Worldwide Research & Development, Eastern Point Road, MS 8274-1260, Groton, CT 06340, United States. Tel.: +1 860 686 2361; fax: +1 860 686 7943. E-mail address: christopher.j.bowman@pfizer.com (C.J. Bowman). 0890-6238/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.reprotox.2012.04.004

also be exposed to human and/or biopharmaceutical IgGs administered to pregnant rats earlier in gestation, during organogenesis. If this holds true the rodent may overestimate rather than underestimate the prenatal transfer of IgGs in humans and thus their use as a nonclinical model species in developmental toxicity safety assessment for humans may serve as a relevant and conservative screen for human developmental toxicity [5,7,9]. The overall objective of this work was to evaluate the embryo-fetal biodistribution of a maternally administered biopharmaceutical humanized IgG2 in rats. The first study used multiple dose levels to illustrate potential saturation of maternal–fetal transfer at the end of gestation and effects on fetal:maternal ratio. The second study analyzed IgG2 levels following different postdose intervals to understand differences between maternal and fetal levels over time. The third study evaluated embryo IgG2 levels at different intervals 24 h post dose during mid-late gestation, including organogenesis. 2. Materials and methods 2.1. hIgG2a hIgG2a is a humanized IgG2 kappa isotype with 2 genetically modified residues in the Fc region to minimize immune effector function, specifically, hIgG2a does not bind to Fc␥RI or complement. This effectively decreases monocyte activation, complement activation, and antibody-dependent cell-mediated cytotoxicity [10,11]. These mutations do not alter the binding to the human, NHP, or mouse FcRn (data not shown) and thus it is assumed that functional FcRnmediated activity (placental transcytosis) would be similar to the wild-type IgG2 kappa isotype. The specific hIgG2a used in these rat studies is a candidate human biopharmaceutical that binds a soluble target in the rat and has pharmacological activity in the rat at doses used in this study (data not shown). The specific

C.J. Bowman et al. / Reproductive Toxicology 34 (2012) 66–72 Table 1 Plasma hIgG2a concentration dose response on gestation day 21. Dose (mg/kg)a

hIgG2a concentration (ng/mL)b

Fetal:maternal ratiod

Maternal plasma 10 30 100 300 a b c d

1042 8775 36,125 77,463

± ± ± ±

256 5519 23,079 28,359

1897 1983 4073 5778

± ± ± ±

366 380 551 743

1.9 0.6 0.14 0.08

Maternal dose administered on gestation days 6 and 12. Mean ± standard deviation. Fetal samples pooled by litter. Mean of ratios calculated from individual maternal–fetal samples.

hIgG2a used in these studies was a high purity protein solution of hIgG2a at a concentration of 52 mg/mL in 20 mM histidine, 84 mg/mL trehalose dehydrate, 0.2 mg/mL polysorbate-80, 0.05 mg/mL EDTA dehydrate, 0.1 mg/mL l-methione, pH 5.5 provided by Pfizer Inc., Chesterfield, MO. The dosing solution was used as provided and the intravenous dose volume (mL/kg) was adjusted to deliver the target mg/kg dose per rat. 2.2. Animal husbandry and maintenance Time-mated female Sprague-Dawley rats [Crl:CD(SD)] were acquired from Charles River Laboratories (Kingston, NY). GD 0 was defined as the day a sperm plug was observed. Animals were between 10 and 12 weeks old at the time of receipt. The rats were housed individually in polycarbonate cages with contact bedding in the form of heat-treated hardwood chips and acclimated for at least 24 h prior to the start of each study. Certified Rodent Diet 5002 (PMI Feeds Inc.) and municipal drinking water further purified by reverse osmosis was provided ad libitum. Environmental room conditions had a minimum of 12 air changes per hour, relative humidity of 50% +10/−15%, temperature of 70◦ +7/−4 ◦ F, and a 12-light/dark cycle. The animal care and experimental procedures were reviewed and approved by the Pfizer Groton Animal Care and Use Committee and were performed in accordance with the standards of the Institute of Laboratory Animal Resources Guide. The facility in which this study was conducted is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International. 2.3. Study design 2.3.1. Dose response evaluation of embryo-fetal biodistribution hIgG2a was administered to 4 groups of 8 time-mated rats by intravenous injection once on GD 6 and again on GD 12 (1 week apart). Dose levels of 10, 30, 100 and 300 mg/kg were achieved by administration of the 52 mg/mL solution at a dose volume of 0.2, 0.6, 2 and 5.8 mL/kg. A similar group of 8 animals received the 5.8 mL/kg vehicle following the same dosing route and regimen and served as a control. Clinical signs, body weights, and food consumption were monitored. Cesarean sections were performed on all rats on GD 21, including gross placental evaluation, collection of gravid uterine weights, numbers of corpora lutea, implantation sites, early and late resorptions, viable and dead fetuses and collection of individual fetal sex and weights. Blood samples were collected and analyzed for plasma hIgG2a concentration from all maternal animals on GD 13 (∼24 h following the second dose) and at necropsy on GD 21. Samples were also collected from the 300 mg/kg females on GD 7, 9, 10, 12, 15, 16 and 18 (no more than 4 survival samples/animal). Blood samples were also collected and analyzed for the presence of anti-drug antibodies (ADA) from all maternal animals on GD 6 (predose) and at necropsy on GD 21. Pooled fetal blood samples (within each litter) were also collected at necropsy on GD 21 and analyzed for plasma hIgG2a concentration and ADA. 2.3.2. Kinetics of embryo-fetal biodistribution post-dose hIgG2a was administered at 300 mg/kg once per animal on GD 10, 12, 14, 18, and 20 to 3 time-mated rats per day as listed in Table 2. Clinical signs and body weights were monitored. Maternal blood was collected from all rats on GD 21. Following euthanasia on GD 21, viable fetuses were removed from the uterus and fetal blood (pooled by litter) was collected. Maternal and fetal samples were analyzed for plasma hIgG2a concentration. The dosing regimen and necropsy on GD 21 correspond to 1, 3, 7, 9 and 11 days post-dose. 2.3.3. Embryo-fetal biodistribution during mid to late gestation hIgG2a was administered at 300 mg/kg to 3 time-mated rats (once per animal) on GD 9, 10, 11, 12, 13, 14, 16, 18, and 20. Clinical signs and body weights were monitored. Three mated rats were euthanized 24 h after each dose. Maternal blood was collected from all rats on the day of euthanasia. Following euthanasia, the uterus was removed, opened, and viable embryo-fetuses removed. All rats were pregnant except one on GD 11, therefore all data are from 3 litters per time point except GD 11 with 2 litters. Pooled (by litter) fetal blood was collected on GD 19 and 21. Maternal and fetal blood samples were analyzed for plasma hIgG2a concentration. Samples

67

of extra-embryonic fluid were collected on GD 12, 13, 14, 15, 17, and 19. Following careful removal of all extra-embryonic cellular layers, intact embryos were collected from 2 to 3 litters each on GD 10, 11, 12, 13, 14, 15, and 17 and placed into a tube of chilled Dulbecco’s phosphate buffered saline (PBS) containing 5 mM EDTA, protease inhibitor cocktail (Sigma P8340), and no calcium or magnesium. The number of embryos per tube and total volume of specimen and buffer were documented. The samples were homogenized on ice at low speed to disrupt the intact tissue using a hand-held plastic pellet homogenizer (GD 10–15) or a loose-fitting Teflon probe homogenizer (GD 17). The duration and speed of the homogenization was not intended to break cellular membranes. The homogenate was centrifuged and supernatant collected for analysis. Following total protein determination (Pierce BCA Protein Assay, Rockford, IL) of the extra-embryonic fluid samples and tissue homogenates, these samples were analyzed specifically for hIgG2a concentration. 2.4. Bioanalysis hIgG2a concentration was determined by enzyme-linked immunosorbent assay (ELISA) from maternal and fetal plasma (K2 EDTA anticoagulant), extraembryonic fluid, and embryo-fetal homogenate supernatant. In addition, presence or absence of immunogenicity (ADA) was evaluated in the first study. 2.4.1. hIgG2a in plasma A previously fully validated pharmacokinetic (PK) ELISA for the quantification of a hIgG2a antibody was used for measurement in maternal and fetal rat plasma. The assay utilized an anti-idiotype-specific antibody to capture the hIgG2a antibody from samples. Standards of hIgG2a were prepared in 100% rat plasma at over a range from 50 to 10,000 ng/mL. Quality controls (QCs) at 5 levels from 100 to 7500 ng/mL were prepared independently from the standards in 100% rat plasma. A minimal required dilution of 1:5 in PK Assay buffer (PBS, 2% [w/v] BSA, 0.05% [w/v] Tween 20, 5 mM EDTA) was used to minimize nonspecific binding and matrix interference. Captured drug was detected with a mouse anti-human IgG2 monoclonal antibody conjugated to horseradish peroxidase (Clone HP6014 Southern Biotech) and was visualized as a chemiluminescent signal using luminol as substrate. A standard curve of hIgG2a concentration versus chemiluminescence was generated using a four-parameter logistic fit without weighting. The hIgG2a antibody binds to a soluble target and with this assay format chemiluminescence units were directly proportional to the unbound and partially bound concentration of hIgG2a antibody in samples. 2.4.2. hIgG2a in embryo-fetal tissue homogenate The rat plasma PK ELISA was modified to measure hIgG2a in rat embryo-fetal homogenates from GD 10 to 17 and a “fit for purpose” assessment of assay performance was conducted. The same assay buffer used to homogenize rat embryos was also used as a surrogate matrix to prepare standards and QCs. To improve assay sensitivity microtitre plates were coated with streptavidin (2 ␮g/mL in PBS) overnight at 4 ◦ C, washed and then incubated with biotinylated anti-idiotype specific antibody (5 ␮g/mL in PK Assay Buffer). After washing, the standards, QC and samples were incubated on the plates for 1 h. The captured drug was detected in the same way as the plasma PK assay (described above). All incubations except for the streptavidin coating step were conducted at room temperature (RT) on a high speed plate shaker. hIgG2a standards were prepared in homogenization buffer as a surrogate matrix at 11 concentrations from 10 to 10,000 ng/mL. QCs were prepared independently from the standards at 6 concentrations from 30 to 7000 ng/mL in homogenization buffer. Additionally, a GD 16 matrix pooled control sample was included on every run to monitor non-specific matrix related-signal and this was subtracted from the counts observed in individual samples. Sample concentrations were calculated in the same manner as the plasma PK assay. The combination of a minimal required dilution of 1:10 with PK Assay buffer with 0.5% Tween and normalization of all method development samples with protein concentrations of more than 5 mg/mL by dilution with homogenization assay buffer to a final protein concentration of 5 mg/mL was required reduce the variable non-specific background. Rat embryo-fetal homogenate from GD 16 not exposed to hIgG2a and diluted with homogenization assay buffer to a final protein concentration of 5 mg/mL was used as a blank matrix control and spiked with 100, 300 and 1000 ng/mL hIgG2a as QC for method development and surrogate matrix assessment. Assay pre-study performance was qualified in a single precision and accuracy run. The assessment of precision and accuracy was continued during sample analysis. Additionally, sample dilution linearity and incurred sample reanalysis was assessed to provide increased confidence in the assay robustness and analysis of samples over the gestation period tested. After initial sample analysis was complete, the dilution linearity (parallelism) was assessed for 10 samples (GD 10–17) subjected to 1 freeze thaw from 10 different litters. 2.4.3. hIgG2a in extra-embryonic fluid The rat embryo-fetal tissue homogenate PK ELISA method was modified slightly for use in the measurement of rat extra-embryonic fluid from embryos at GD 12 to 19; standards were prepared in PBS with 2% BSA as a surrogate matrix at 11 concentrations from 10 to 10,000 ng/mL. Sample concentrations were determined in the same manner as the other assays. Extra-embryonic fluid collected from 6

68

C.J. Bowman et al. / Reproductive Toxicology 34 (2012) 66–72

10000000.00

hIgG2Δa concentration (ng/mL)

IgG2Δa concentration (ng/mL)

10000000.00

1000000.00

100000.00

10 mg/kg

10000.00

30 mg/kg 100 mg/kg 300 mg/kg

1000000.00

100000.00

10000.00

GD 13-maternal

1000.00

GD 21-maternal

1000.00

GD 21-fetal 6

7

8

9 10 11 12 13 14 15 16 17 18 19 20 21

100.00

Gestation Day

0

Fig. 1. Mean concentration of maternal plasma hIgG2a at different gestation ages following 10, 30, 100, or 300 mg/kg intravenous doses on GD 6 and 12.

different litters not exposed to hIgG2a with protein concentrations ranging from 1.3 to 4. 5 mg/mL were used to develop the method. A pool of GD 14 extra-embryonic fluid was used to prepare QCs at 6 concentrations from 30 to 7000 ng/mL. Assay performance was qualified in a single run based on standard and matrix QC %CV and %RE as well as the level of interference from 22 individual GD 14 extra-embryonic fluid samples unexposed to hIgG2a. The assessment of precision and accuracy was continued during sample analysis. Additionally, sample dilution linearity was evaluated since only GD 14 samples were evaluated for matrix interference during method development. After initial sample analysis was complete, the dilution linearity of 13 samples subjected to 1 freeze thaw (from 6 different litters) was assessed at dilutions of 1:5, 10, and 20. The observed individual back-calculated results were corrected for dilution factor and expressed as a mean percent. 2.4.4. ADA assay Samples were tested for ADA using a validated electrochemiluminescent assay. In brief: biotinylated IgG2a, ruthenylated IgG2a, and serum samples were combined and incubated to form an antibody-drug complex. The complexes were then transferred to the wells of a Meso Scale Discovery streptavidin plate and incubated with shaking. Biotinylated IgG2a in the complex was then captured and unbound material washed away. Tripolylamine-containing buffer was added to the plate and the chemiluminescent signal generated was detected with a Meso Scale Discovery PR400 plate reader. Data were reported as positive or negative based on respective cut points. Prior to measurement in pregnant rat serum samples, the assay was determined to behave similar to non-pregnant serum negative control. In the absence of control fetal serum, vehicle control samples were analyzed to confirm that they also behaved the same as the negative control.

3. Results 3.1. Dose response evaluation of placental transfer There was no evidence of maternal or fetal toxicity based on the endpoints evaluated (data not shown). Maternal hIgG2a concentrations increased with increasing dose and decreased over time post dose (Fig. 1). Fetal hIgG2a concentrations as measured

50

100

150

200

250

300

350

dose (mg/kg) Fig. 2. Following maternal dose administration on GD 6 and 12, mean concentration of maternal plasma hIgG2a on GD 13 at different dose levels (24 h post dose; filled circle with dashed line) and concentration of maternal (open diamond with solid line) and fetal (filled triangle with solid line) plasma hIgG2a on GD 21 at different dose levels (9 days post dose).

on GD 21 (9 days since last maternal dose) also increased with increasing dose, but not to the same extent (Fig. 2). Specifically at 10 mg/kg, fetal concentrations were higher than maternal levels but at ≥30 mg/kg maternal levels were higher than fetal levels (Table 1). This resulted in GD 21 fetal:maternal ratios that appeared to decrease with increasing maternal dose (Table 1 and Fig. 3). ADA was only detectable in maternal animals on GD 21 in 4 of 8 animals at 10 mg/kg and 1 of 8 animals at 100 mg/kg and was not detected in fetal samples (data not shown). Although hIgG2a present in the sample has the potential to interfere with accurate ADA detection, there was no detectable interference of ADA on hIgG2a concentration from those animals with evidence of ADA.

3.2. Kinetics of placental transfer post-dose Both maternal and fetal hIgG2a concentrations decreased as the number of days increased from 1 to 11 days post single intravenous injection (Table 2). The decrease in fetal concentrations was not as rapid as the maternal decrease and even appeared relatively constant between 7 and 11 days post-dose (Fig. 4A). This different maternal and fetal post-dose concentration profile at the end of gestation resulted in an increase in fetal:maternal ratio between 1 and 7 days post-dose, with similar ratio between 7 and 11 days post-dose (Fig. 4B). This increase in fetal:maternal ratios the first 7 days post-dose was primarily attributed to the relatively faster

Table 2 Post-dose plasma hIgG2a concentration time course. Dose day (GD)

Collection day (GD)

Days post-dose

hIgG2a concentration (ng/mL)a Maternal plasma

20 18 14 12 10

21 21 21 21 21

1 3 7 9 11

GD, gestation day. a Mean ± standard deviation, n = 3 per time point. b Fetal samples pooled by litter. c Mean of ratios calculated from individual maternal–fetal samples.

318,333 123,000 57,633 48,900 43,233

± ± ± ± ±

41,585 20,952 11,940 18,016 3139

Fetal:maternal ratioc b

Fetal plasma 27,433 21,267 15,667 11,457 13,800

± ± ± ± ±

1358 1305 666 3367 361

0.09 0.18 0.28 0.24 0.32

± ± ± ± ±

0.01 0.02 0.05 0.03 0.02

C.J. Bowman et al. / Reproductive Toxicology 34 (2012) 66–72

10

69

1000000 10 mg/kg 100000

100 mg/kg

1

hIgG2Δa levels

GD 21 fetal:maternal ratio (9 days since last dose)

30 mg/kg

300 mg/kg

0.1

10000

1000

100

Maternal plasma (ng/mL) Fetal plasma (ng/mL)

10

Extra-embryonic fluid (ng/mL) Embryo-fetal tissue (ng/embryo)

0.01 5

50

500

1 9

dose (mg/kg)

10

11 12 13 14 15 16 17 18 19 20 21 22

Gestation Day Fig. 3. Fetal:maternal ratios of plasma hIgG2a on GD 21 at different doses 9 days following the last dose. Ratio from each individual fetal–maternal pair is plotted. Fetal:maternal ratios appear to decrease with increasing maternal dose.

hIgG2Δa concentration (ng/mL)

A

1000000

post-dose decrease in maternal concentrations compared to the relatively slow decrease in fetal concentrations post-dose.

Maternal plasma Fetal plasma

3.3. Biodistribution during mid to late gestation

100000

10000 0

1

2

3

4

5

6

7

8

9

10 11 12

Days following single IV dose

B

0.40 0.35

Fetal:Maternal Ratio

Fig. 5. Biodistribution of hIgG2a among the maternal–fetal compartments 24 h following single maternal dose administration of 300 mg/kg on GDs 9, 10, 11, 12, 13, 14, 16, 18 and 20. Mean ± SD is plotted for maternal and fetal plasma (ng/mL), extra embryonic fluid (ng/mL), and embryo-fetal tissue (ng/embryo). Embryo-fetal tissue samples were pooled by litter but normalized for the number of embryos.

0.30 0.25 0.20 0.15 0.10 0.05 0.00 0

1

2

3

4

5

6

7

8

9

10

11 12

Maternal–fetal biodistribution of hIgG2a 24 h following a single 300 mg/kg dose between GD 9 and 20 resulted in increasing hIgG2a levels in embryos from GD 10 until an apparent plateau toward the end of gestation between GD 19 and 21 (Table 3). Concentrations of plasma hIgG2a in maternal animals 24 h post-dose at each gestation age from GD 10 to 21 were similar. hIgG2a levels in embryo homogenates increased from GD 10 (10.1 ng/embryo) to GD 17 (6266 ng/embryo). There was a large variability of total protein in GD 10 embryo homogenates. This was not unexpected since these animals were time-mated by the supplier overnight so variability in several hours between litters for conception and implantation is possible and would be most obvious at the earliest gestation ages evaluated. This variability in total protein did not necessarily correspond to variability in hIgG2a levels, but confidence in the data on GD 10 may be lower than other ages despite relatively consistent hIgG2a levels and alignment with data at later gestation ages/developmental stages. The increasing hIgG2a levels in embryos was consistent with correspondingly higher concentrations in extra-embryonic fluid that slowly increased from GD 12 (2651 ng/mL) to GD 19 (8632 ng/mL). The increasing hIgG2a levels in embryo homogenate and extra-embryonic fluid as gestation progressed continued with even higher concentrations in fetal plasma on GD 19 and 21. Overall, the hIgG2a levels in the embryo between GD 10 and 12 was very low (greater than 4 orders of magnitude lower than maternal levels) 24 h post-dose, with levels increasing throughout gestation to within 1 order of magnitude lower than maternal levels by GDs 19 and 21 (Fig. 5).

Days following single IV dose 4. Discussion Fig. 4. (A) Maternal (open circles with dashed line) and fetal (open triangles with solid line) plasma hIgG2a concentration (mean ± SD) on GD 21 following single maternal dose administration of 300 mg/kg on GD 10, 12, 14, 18, or 20 for different post dose intervals of 11, 9, 7, 3, and 1 days, respectively. (B) Fetal:maternal ratio of plasma hIgG2a on GD 21 following post-dose intervals of 11, 9, 7, 3, and 1 days. Ratio from each individual fetal–maternal pair is plotted.

Embryo-fetal biodistribution of maternal IgGs (endogenous maternally derived antigenic response or exogenously delivered biopharmaceutical IgG) is primarily thought to occur by FcRnmediated transcytosis [12]. The anatomy and physiology of this transfer is different among different animal models (rodents,

70

C.J. Bowman et al. / Reproductive Toxicology 34 (2012) 66–72

Table 3 hIgG2a biodistribution during gestation. GD (24 h post dose)

hIgG2a levelsa Maternal plasma (ng/mL)

10 11 12 13 14 15 17 19 21

347,667 381,000 373,333 327,333 364,000 412,000 403,333 340,667 318,333

± ± ± ± ± ± ± ± ±

13,796 77,782 22,855 11,240 37,162 85,159 38,837 44,060 41,585

Extra-embryonic fluid (ng/mL)

Embryo-fetusb

– – 2651 ± 584 3005 ± 331 5028 ± 1059 5899 ± 2072 8234 ± 2037 8632 ± 2645 –

10.1 24.4 15.6 97.8 313 1474 6266 26,667 27,433

± ± ± ± ± ± ± ± ±

3.2 0.3 10.7 3.9 87.1 595 831 451 1358

GD, gestation day; −, not collected. a Mean ± standard deviation. b GD 10–17 samples are embryo-fetal tissue homogenate levels (ng/embryo) and GD 19–21 samples are fetal plasma concentration (ng/mL). Samples pooled by litter.

rabbits and primates). In rodents and rabbits, FcRn-mediated transcytosis occurs across the yolk sac placenta whereas in primates it is across the chorioallantoic placenta, but both mechanisms result in delivery to the developing conceptus [4,5]. Importantly, rodent IgGs have less affinity than human IgG to human FcRn but human IgGs bind rodent FcRn with greater affinity than rodent IgGs [13,14]. This increased binding in rodent relative to human may result in increased placental transfer in rodents relative to primates, but is unlikely to change the timing of transfer during gestation. For rodents and primates, there is also FcRn-mediated IgG uptake across the brush border of enterocytes of the gut by prenatal swallowing of IgG-containing amniotic fluid [2,15]. Regardless of the mechanism of maternal–fetal transfer of IgGs, it is difficult to quantify embryo-fetal biodistribution in the earlier stages of embryogenesis due to challenges in obtaining sufficient samples for analysis. This is exemplified by the limited data available on IgGs in embryos in all species and, as described in the introduction, inconsistencies in that data. Another challenge to understanding the limited data on embryo-fetal IgG biodistribution in the literature is the routine reporting of these data primarily as a fetal:maternal ratio. Since the available data in the literature indicate that distribution of maternal IgG to the embryo in a healthy pregnancy is mediated by FcRn-mediated transfer in all species [12], it is likely that IgG distribution to the conceptus would follow first-order kinetics and have the potential for receptor saturation at high maternal blood levels. In addition, placental FcRn expression (and presumed transfer capacity) is thought to transition from low to high as gestation progresses. Lastly, for some pharmacologic targets there is the potential for target-mediated biodistribution and clearance in addition to FcR binding pharmacokinetics. These concepts, combined with potentially different maternal and fetal post-dose pharmacokinetic profiles, add up to multiple confounding variables contributing to the fetal:maternal ratio at any particular point in time. In the absence of reporting actual fetal concentrations at specific days of gestation, it is difficult to use fetal:maternal ratios in nonclinical animal models to understand potential exposure at the target in the developing embryo and subsequent cross-species extrapolation of embryo-fetal developmental risk to human. In the first study of the current investigation, maternal and fetal plasma hIgG2a levels were measured over a large dose range (10–300 mg/kg) to evaluate the potential for saturation of fetal plasma distribution at a fixed gestation age (end of gestation) and time since last dose. Over this dose range, both maternal and fetal plasma hIgG2a levels increased with increasing dose (Fig. 2), but fetal:maternal ratio decreased with increasing dose (Fig. 3), consistent with the beginnings of receptor-mediated transfer saturation. This example illustrates the importance of analyzing the fetal and maternal levels individually at higher dose levels as this data set

suggests that the decreasing fetal:maternal ratio was more a factor of maternal levels increasing with dose without concomitant proportional increase in fetal levels, likely due to the beginnings of receptor-mediated transfer saturation. In the second study of the current investigation the post-dose profiles of maternal and fetal hIgG2a levels were evaluated at a single dose level with gestation age again held constant. GD 21 fetal levels decreased with increasing time since single dose, but not as markedly decreased as maternal levels (Fig. 4A). Thus in this experiment, the fetus did not accumulate hIgG2a with increasing time since last dose (no evidence for the fetus acting as a sink for accumulating hIgG2a). The difference in maternal and fetal postdose profiles of hIgG2a resulted in an increase in fetal:maternal ratio with increasing days since last dose (Fig. 4B). The third study of the current investigation characterized the biodistribution of a biopharmaceutical hIgG2a in the embryo, extra-embryonic fluid and fetal plasma 24 h following single maternal doses at various gestation ages between GD 9 and GD 20 (including most of organogenesis and later fetal development). hIgG2a levels were present at the earliest age evaluated in the embryo (GD 10), increasing more than 100-fold by GD 17. Notably, the hIgG2a levels in the embryo-fetal tissue were not adjusted for the size of the conceptus (were simply reported as ng/embryo) due to technical challenges weighing the early age specimens and the difficulty adjusting for the biological change in embryo composition (e.g., percent water in the embryo) throughout organogenesis. Therefore, the increasing size of the specimens likely had a corresponding impact on the increasing levels of hIgG2a in these specimens and thus it is not possible to accurately determine the actual hIgG2a concentration in tissue at these different gestation ages. Regardless, the consistent profile of increasing embryonic levels as gestation progressed was also observed in extra-embryonic fluid and fetal plasma hIgG2a concentration (ng/mL). The appearance of very low levels (10 ng/embryo) of hIgG2a in the embryo on GD 10 is coincident with 2 aspects of developmental biology that are likely important for distribution to the embryo. The first is expression of FcRn in the mouse yolk sac on GD 10.5 [16] and the second is evidence of the first circulation in the rat embryo on GD 10–10.5 [17]. Although the embryo tissue levels of hIgG2a between GD 10 and 12 (10–24 ng/embryo) were approximately 4 orders of magnitude lower than maternal plasma concentrations (ng/mL), a direct comparison (ratio) to maternal plasma concentration (ng/mL) was not calculated because these different matrices have different concentration units making direct reconciliation of these data relative to each other problematic for direct comparison. Regardless of the actual concentration or ratio, an important point to be made is that for some biopharmaceuticals with high potency even very low levels present in the developing

C.J. Bowman et al. / Reproductive Toxicology 34 (2012) 66–72

embryo during organogenesis may translate to activity at a pharmacologic target and with it the potential for unintended biological effects. As described earlier the limited data on maternal–fetal transfer of IgGs in rodents are not consistent, consequently embryo-fetal biodistribution during organogenesis and gestation is not currently predictable in rodents relative to what is known in primates. The position put forth based on maternally derived IgG is that rodents may underestimate exposure in humans [4]. The embryofetal biodistribution data reported in the current investigation is believed to be the first report of early embryonic levels of a humanized biopharmaceutical IgG in the rat and is consistent with data reported in mice using radioactivity [6], and in rats in later gestation [7]. Based on available information on the specific molecule used in the current study, there is no evidence of target-mediated clearance and target-mediated biodistribution is not thought to alter the embryo-fetal biodistribution studies reported in this manuscript (data not shown). The detection of very low levels of biopharmaceutical IgG in rat embryos during organogenesis following maternal dosing may have been predicted based on the limited data on FcRn expression and fetal-placental circulation in rodents [16,17]. These data may raise the question of whether very low levels may be present in primate embryos (humans and monkeys) following maternal dosing during organogenesis. Currently the very limited data on embryo IgG exposure in humans [18,19] and in monkeys [20] report very low levels of IgG at the end of the first trimester and organogenesis, respectively. In addition, different IgG isotypes have different fetal exposure profiles in humans [21], though it is currently unknown if standard nonclinical model species share these differences in isotype profile. Most of the reported data in humans and monkeys to date are limited by access to sufficient samples for analysis. Thus, future research may reveal embryo exposure (very low levels) earlier than expected in primates, coincident with embryo blood circulation and FcRn expression in this period of organogenesis as in rodents. Therefore, these data demonstrating rat embryo hIgG2a levels (albeit relatively low) during the period of major organogenesis are consistent with the hypothesis that rodents overestimate the generally accepted embryo exposure profile in primates [5,7,9]. If these data are representative of other IgG biopharmaceuticals and exposure profiles in primates are as expected, the use of rodents as a nonclinical model species in developmental toxicity safety assessment for IgG biopharmaceuticals may serve as a conservative screen for human developmental toxicity [5,7].

5. Conclusion Fetal and maternal hIgG2a levels at the end of rat gestation increased with increasing dose from 10 to 300 mg/kg but fetal:maternal ratio decreased with increasing dose, consistent with differences in pharmacokinetics in these 2 compartments and likely saturation of maternal–fetal transfer mechanisms. In addition, fetal rat hIgG2a levels decreased with increasing time post-dose but more slowly than the decrease in maternal levels. This difference in post-dose profile resulted in an increased fetal:maternal ratio with increasing days since last dose. These data indicate that fetal:maternal ratio data alone is insufficient to understand the profile of embryo-fetal biodistribution when dosing maternal animals during pregnancy. Finally, embryo-fetal levels of maternally administered hIgG2a in rats was detected at very low levels on GD 10–12 (earliest GDs evaluated) and levels increased over 100-fold by the end of organogenesis on GD 17. A consistent profile of increasing hIgG2a levels in the embryo as gestation progressed was also observed in extra-embryonic fluid (GD 12–19) and

71

fetal plasma hIgG2a concentration (GD 19–21). Based on the current study, there is a potential for direct effects on rat embryo-fetal development following maternal administration of a biopharmaceutical IgG2. Conflicts of interest statement The authors declare that there are no conflicts of interest. Acknowledgements The authors graciously acknowledge Rebecca Koitz, Scott Davenport and Bill Nowland for their technical expertise, Michael Mays at QPS, LLC for the plasma bioanalysis, Jessica Whritenour for ADA analysis, Lei Lu for organizing the bioanalysis, Dr. Rob Webster for his support and bioanalytical expertise, Drs. Bora Han and Scott Fountain for managing this specific program, and Drs. Mark Hurtt, Gregg Cappon, Bob Chapin, Sarah Campion, and Mitchell Thorn for their helpful discussions and expert review of this manuscript. Funding for all of this work came from Pfizer, Inc. References [1] Martin PL, Weinbauer GF. Developmental toxicity testing of biopharmaceuticals in non-human primates: previous experience and future directions. International Journal of Toxicology 2010;29:552–68. [2] Brambell FWR. The transmission of passive immunity from mother to young. In: Brambell FWR, editor. Frontiers of biology. Amsterdam: North-Holland Publishing Co; 1970. p. 80–141. [3] Halliday R. Prenatal and postnatal transmission of passive immunity to young rats. Proceedings of the Royal Society of London Series B 1955;144:427–30. [4] Pentˇsuk N, van der Laan JW. An interspecies comparison of placental antibody transfer: new insights into developmental toxicity testing of monoclonal antibodies. Birth Defects Research Part B: Developmental and Reproductive Toxicology 2009;86:328–44. [5] DeSesso JM, Williams AL, Ahuja A, Bowman CJ, Hurtt ME. The placenta, transfer of immunoglobulins, and safety assessment of biopharmaceuticals in pregnancy. Critical Reviews in Toxicology 2012;42:185–210. [6] Morphis LG, Gitlin D. Maturation of the maternofoetal transport system for (-globulin in the mouse. Nature 1970;228:573. [7] Martin PL, Sachs C, Hoberman A, Jiao Q, Bugelski PJ. Effects of CNTO530erythropoietin mimetic- IgG4 fusion protein, on embryofetal development in rat and rabbits. Birth Defects Research Part B: Developmental and Reproductive Toxicology 2010;89:87–96. [8] Martin PL, Cornacoff JB, Treacy G, Eirikas E, Marini J, White Jr KL, et al. Effects of administration of a monoclonal antibody against mouse tumor necrosis factor alpha during pregnancy and lactation on the pre- and postnatal development of the mouse immune system. International Journal of Toxicology 2008;27:341–7. [9] Martin PL, Breslin W, Rocca M, Wright D, Cavagnaro J. Considerations in assessing the developmental and reproductive toxicity potential of biopharmaceuticals. Birth Defects Research Part B: Developmental and Reproductive Toxicology 2009;86:176–203. [10] Armour KL, Clark MR, Hadley AG, Williamson LM. Recombinant human IgG molecules lacking Fcgamma receptor I binding and monocyte triggering activities. European Journal of Immunology 1999;29:2613–24. [11] Armour KL, Atherton A, Williamson LM, Clark MR. The contrasting IgG-binding interactions of human and herpes simplex virus Fc receptors. Biochemical Society Transactions 2002;30:495–500. [12] Roopenian DC, Akilesh S. FcRn: the neonatal Fc receptor comes of age. Nature Reviews Immunology 2007;7:715–25. [13] Ober RJ, Radu CG, Gheti V, Ward ES. Differences in promiscuity for antibody–FcRn interactions across species: implications for therapeutic antibodies. International Immunology 2001;13:1551–9. [14] Andersen JT, Daba MB, Berntzen G, Michaelsen TE, Sandlie I. Cross species binding analyses of mouse and human neonatal Fc receptor show dramatic differences in immunoglobulin G and albumin binding. Journal of Biological Chemistry 2010;285:4826–36. [15] Shah U, Dickinson BL, Blumberg RS, Simister NE, Lencer WI, Walker WA. Distribution of the IgG Fc receptor, FcRn, in the human fetal intestine. Pediatric Research 2003;53:295–301. [16] Jaffe L, Jeannote L, Bikoff EK, Robertson EJ. Analysis of beta-2 microglobulin gene expression in the developing mouse embryo and placenta. Journal of Immunology 1990;145:3474–82. [17] DeSesso JM. Comparative features of vertebrate embryology. In: Hood RD, editor. Developmental and reproductive toxicology: a practical approach. 2nd ed. Boca Raton: Taylor & Francis; 2006. p. 147–97. [18] Dancis J, Lind J, Oratz M, Smolens J, Vara P. Placental transfer of proteins in human gestation. American Journal of Obstetrics and Gynecology 1961;82:167–70.

72

C.J. Bowman et al. / Reproductive Toxicology 34 (2012) 66–72

[19] Jauniaux E, Jurkovic D, Gulbis B, Liesnard C, Lees C, Campbell S. Materno-fetal immunoglobulin transfer and passive immunity during the first trimester of human pregnancy. Human Reproduction 1995;10:3297–300. [20] Wang H, Schuetz C, Akihiro A, Woods C, Xiao J, Iyer S, et al. Assessment of placental transfer and the effect on embryo-fetal development of a humanized monoclonal antibody targeting lymphotoxin-alpha in non-human primates.

Late-breaking abstract poster presentation at the 50th annual meeting of the society of toxicology; 2011. p. 2772–844. [21] Malek A, Sager R, Kuhn P, Nicholaides KH, Schneider H. Evolution of maternofetal transport of immunoglobulins during human pregnancy. American Journal of Reproductive Immunology 1996;36:248–55.