Generation and characterization of a novel tetravalent anti-CD22 antibody with improved antitumor activity and pharmacokinetics

International Immunopharmacology 6 (2006) 791 – 799 www.elsevier.com/locate/intimp

Generation and characterization of a novel tetravalent anti-CD22 antibody with improved antitumor activity and pharmacokinetics B Xiao-yun Liu a,1, Laurentiu M. Pop a,1, Derry C. Roopenian b, Victor Ghetie a, Ellen S. Vitetta a,*, Joan E. Smallshaw a a

The Cancer Immunobiology Center, University of Texas Southwestern Medical Center at Dallas, 6000 Harry Hines Blvd, NB9.210, Dallas, Texas 75390-8576, USA b The Jackson Laboratory, Bar Harbor, 600 Main Street, Maine 04609, USA Received 9 August 2005; received in revised form 2 October 2005; accepted 22 November 2005

Abstract The purpose of this study was to prepare a tetravalent anti-human CD22 recombinant antibody with improved antitumor activity and a half life longer than that of its divalent counterpart. We compared the ability of tetravalent vs. divalent antibody to associate/ dissociate to/from CD22-positive Daudi cells, to interact with murine and human Fcg receptors (FcgR), to bind human complement component C1q, to inhibit the growth of tumor cells, to diffuse into various tissues, to be internalized by Daudi cells, to react with human neonatal Fc receptors (FcRn), and to persist in the circulation of normal mice. As compared to the murine or chimeric divalent antibodies, the chimeric tetravalent counterpart has a longer half life in mice. It also has an affinity for FcRns that is identical to that of human IgG. The tetravalent antibody has increased antitumor activity in vitro and completely conserved effector functions (binding to FcgR-positive cells and to C1q) in vitro. Despite its 33% higher molecular weight, it penetrates mouse tissues as well as its divalent antibody counterpart. Based on the improved in vitro performance and pharmacokinetics of the tetravalent antibody it will now be tested for its antitumor activity in vivo. D 2005 Elsevier B.V. All rights reserved. Keywords: Monoclonal antibodies; Tetravalent; Pharmacokinetics; Antitumor; Lymphoma

1. Introduction Tetravalent monoclonal antibodies (TetraMAbs) have been generated either chemically by crosslinking two molecules of IgG MAbs with the same specificB Supported by: The Cancer Immunobiology Center, UT Southwestern Medical Center, Dallas, Texas. * Corresponding author. Tel.: +1 214 648 1200; fax: +1 214 648 1204. E-mail address: [email protected] (E.S. Vitetta). 1 These authors contributed equally to this study.

1567-5769/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.intimp.2005.11.022

ity (two Fc regions) [1–3] or by genetic engineering [4] to construct a single IgG antibody molecule with four identical Fv regions and a single Fc region [5,6]. TetraMAbs have higher functional affinity than their divalent counterparts and they inhibit the growth of various tumor cells in vitro. They have increased antitumor activity in vivo [3]. These TetraMAbs also retain all of their effector functions in vitro [1,2,6]. The only drawback of the recombinant TetraMAbs vs. divalent MAbs has been their shorter half life [6] and 33% larger size. Although the in vivo half life of a TetraMAb was considerably longer than

792

X. Liu et al. / International Immunopharmacology 6 (2006) 791–799

that of chemical homodimers [6], it was still significantly shorter than that of its divalent antibody counterpart [5,6]. The shorter half life of the previously described TetraMAb was probably due to a modification in the structure of the hinge region since Cys-229 was changed to Ser-229 [6]. Since the half life of an IgG molecule in the circulation depends on the Cys–Cys bond in the hinge region [7], we postulated that this alteration in the hinge region of the TetraMAb might account, at least in part, for its shorter half life. The enhancement of crosslinking and resulting antitumor activity by TetraMAbs might be offset by their shorter half life in vivo, hence decreasing their diffusion gradient into a tumor. Therefore, we reengineered the TetraMAb by changing the Ser residue back to Cys and by lengthening the hinge region to increase the flexibility of the four Fv regions. To avoid any mispairing due to Cys-220, this residue was changed to proline. To improve the stability of the tandem Fv construct, the linker which was used to join the two Fv portions in the previous constructs (Gly4Ser) was replaced with a more stable and flexible peptide linker (Ala–Ser–Thr– Gly–Ser). This new RFB4 anti-CD22 recombinant TetraMAb binds more avidly on CD22+ cells and retains all its effector functions. Importantly, it has a longer half life than both the murine and chimeric divalent MAb in vivo. 2. Materials and methods 2.1. Construction, expression and purification of chimeric TetraMAb and divalent MAb (see Fig. 1) To produce the chimeric RFB4 (cRFB4) TetraMAb, the pDZFc vector which is derived from pSecTag2A (Invitrogen, Carlsbad, CA), and pSV2/dhfr vectors (purchased from American Type Culture Collection, ATCC, Manassas, VA) and containing the cDNA encoding native human IgG1 Fc under the control of cytomegalovirus promoter, a Zeocin resistant gene (Sh ble), and a dihydrofolate reductase (DHFR) gene expression cassette were used to construct the cRFB4 TetraMAb expression vector (Fig. 1). The hinge sequence of Fc portion was changed to AAAEPKSPDKTHTCPPCP by using primers A, 5VAAGGAAAAAAGCGGCCGCAGAGCCCAAATCTCCCGACAAAACTCACACATG-3V; and B, 5V-CATGATTCTAGATCATTTACCCGGAGACAGGG-3V. In this manner, the pDZFc-Pro220 vector containing a Cys-220 to Pro-220 mutation in Fc portion was obtained. To assemble the tandem single chain Fv (scFv) construct, cDNA of scFv of RFB4 in pCR2.1-TOPO vector was used as the PCR template. The following PCR primers, designated C through F, were used to obtain two scFv constructs, respectively: primer C, 5V-GGGCGGCCCAGCCGGCCAG-

GCGCGACATCCAGATGACACAGAC-3V; primer D, 5VATGACACAGAC-3V; and primer F, 5V-ATAGTTAGCGGCCGCTGCAGAGACAGTGACCAG-3V. The scFv constructs were inserted in frame into the pDZFc-Pro220 vector by three-way ligation to create the cRFB4 TetraMAb expression vector. The pcRFB4 TetraMAb vector was transfected into CHO cells using Lipofectaminek 2000 reagent (Invitrogen). Stable transfectants were selected in IMDM medium (Sigma-Aldrich, St. Louis, MO) containing 10% dialyzed fetal bovine serum (FBS) (Hyclone, Logan, UT), 200 Ag/mL Zeocin (Invitrogen) and 20 nM Methotrexate (Calbiochem, La Jolla, CA). The highest yield achieved was 10–16 mg/L as determined by ELISA. Purification of the cRFB4 TetraMAb was carried out by precipitation with 50% ammonium sulfate followed by affinity chromatography on Protein G-Sepharose (Amersham Biosciences, Piscataway, NJ). The bound protein was eluted with 0.1 M glycine–HC1 buffer, pH 2.8. To remove some high molecular weight aggregates from the cRFB4 TetraMAb, size exclusion high performance liquid chromatography (HPLC) was performed using a preparative TSK-3000SW column (TosoHaas, Montgomeryville, PA). The construction and expression of the divalent cRFB4 and murine RFB4 has been previously described [6]. 2.2. Cells Cells (Daudi, U937 and WEHI-274.1) were maintained in culture by serial passages in RPMI-1640 (ICN Biomedicals, Aurora, OH) as previously described [6]. 2.3. Pharmacokinetic (PK) analysis Swiss Webster mice (Taconic, Germantown, NY) and B6.mFcRn / - .hFcRn Tg line 276 + / - transgenic mice (The Jackson Laboratory, Ben Harbor, ME) were used. The radiolabeled proteins (5–10  107 cpm/150 AL) were injected into the tail vein, and whole body radioactivity was measured daily [6]. The PK parameters were determined using a noncompartimental model with the PKCALC program from data collected daily between 24 and 168 h after injection [6]. 2.4. Inhibition of cell growth Daudi cells were suspended in complete RPMI-1640 medium at a concentration of 104/mL. 5 mL of the cell suspension was added to each 25-cm2 tissue culture flask. 50 AL of MAbs was then added to the flasks. The cells were grown for 7 days at 37 8C in a CO2 incubator. An aliquot of the cell suspension was removed on days 1, 2, 3, 4, 5, 6, 7 and cells were stained with Trypan Blue. The cell viability was determined on each day and the results were expressed as the percentage of cell number relative to that of the control without antibody.

X. Liu et al. / International Immunopharmacology 6 (2006) 791–799 A SfiI/NotI

A.

793

B XbaI

DHFR CMV Hi-CH2-CH3 Zeo pDZFc PCR

SfiI/NotI

XbaI

DHFR CMV Hi-CH2-CH3 Zeo pDZFc-Pro220

B. C

VL

VH

D

(GGGGS)3 E

VL

VH

F

(GGGGS)3 PCR

VL SfiI

PCR

VH (GGGGS)3

BamHI ASTGS VL GS

VH (GGGGS)3

NotI

SfiI/NotI DHFR CMV Hi-CH2-CH3 Zeo pDZFc-Pro220

Ligation

DHFR CMV scFv-ASTGS-scFv Hi-CH2-CH3 Zeo pcRFB4 TetraMAb

Fig. 1. Generation of the cRFB4 TetraMAb expression vector. A. Mutation of the human IgG1-Fc gene. The mutation was performed by PCR using primers A and B. The PCR product digested by NotI/XbaI was inserted in frame in the same restriction sites in the pDZFc vector, and then pDZFcPro220 vector was obtained; B. The RFB4 scFv gene was modified by PCR using Primer C/D and E/F to introduce a ASTGS peptide linker between the scFv genes. Thereafter, the PCR products were digested by either SfiI/BamHI or BamHI/NotI. The three-way ligation was then performed between the digested PCR products and the pDZFc-Pro220 vector linearized by SfiI/NotI to obtain the pcRFB4 TetraMAb vector. CMV, the cytomegalovirus promoter; DHFR, the dihydrofolate reductase gene expression cassette; Hi-CH2–CH3, the cDNA encoding human IgG1 Fc; Zeocin, the Zeocin resistance gene (Sh ble) expression cassette.

2.5. Binding to Daudi, U937 and WEHI-274.1 cells Cells were suspended in complete RPMI-1640 medium and treated with different concentrations of radiolabeled ligands as described [6]. 2.6. Dissociation of MAbs from Daudi cells The cells were treated with the radiolabeled MAb, centrifuged and suspended in medium +/ cold murine RFB4.

They were then incubated at 37 8C for various intervals of time as described [6]. 2.7. Internalization of MAbs by Daudi cells Daudi cells were incubated with 125I-labeled MAbs on ice for 1 h. After washing out the excess MAb, the cells were resuspended at 106/mL and incubated at 37 8C for 4 and 8 h to measure the radioactivity that was internalized, membrane bound, or released into the medium as described previously

794

X. Liu et al. / International Immunopharmacology 6 (2006) 791–799

[8]. The results were expressed as percentage of trichloroacetic acid (TCA)-precipitable radioactivity. 2.8. Diffusion into tissues 0.01% NaI was added to the drinking water of female Swiss Webster mice (Taconic) 6–8 weeks of age one day prior to injection. The animals were injected intravenously (i.v.) with 5  107 cpm/15 Ag/150 Al radiolabeled cRFB4 TetraMAb and murine RFB4. Groups of four mice were sacrificed following perfusion with heparinized saline solution after 70 h. Samples of skin, muscle and small intestine were harvested, washed in saline, dry-blotted, weighed and counted in a gamma counter. The results were expressed as a percentage of the injected dose per gram of tissue (%ID/g of tissue).

the cRFB4 TetraMAb is: VL–(Gly4Ser)3–VH–Ala–Ser–Thr– Gly–Ser–VL–(Gly4Ser)3–VH–CH2–CH3. Thus, following self-assembly of two chains through the formation of the Fc domain the MAb contains four antigen-binding sites of mouse origin and a single human Fc. The purification and characterization of divalent cRFB4 and murine RFB4 have been described [6]. The cRFB4 TetraMAb was purified to homogeneity by precipitation with ammonium sulfate followed by affinity chromatography on Protein G-Sepharose and HPLC on TSKG 3000SW column. The purity and the molecular weight of the cRFB4 TetraMAb were determined by SDS-PAGE (Fig. 2). One major band at 200 kDa (nonreduced) and one major band at 100 kDa (reduced) were observed. These results verified that the cRFB4 TetraMab was homogeneous and that there was a disulphide bond in the hinge region.

2.9. Binding of human C1q by Daudi cells coated with MAbs 3.2. Pharmacokinetics Daudi cells at 107 cells/mL were incubated with 0.5 Ag/mL MAb for 1 h at 4 8C in complete medium containing 0.01% sodium azide. After washing the cells with ice cold barbitalbuffered saline (Sigma) they were incubated with radiolabeled human C1q (Sigma) (1 Ag/mL), at 4 8C for 30 min, followed by centrifugation. After washing out the unbound ligand, the radioactivity in the pellet was measured in a gamma-counter and the values (after subtracting the binding to the negative control) were expressed as a percentage relative to the binding of cRFB4, which was taken as 100%. 2.10. Measurement of the affinity of the MAbs for the human FcRn

The elimination curves of the radiolabeled MAbs in Swiss Webster mice are shown in Fig. 3. From these curves, PK parameters were calculated. Values are presented in Table 1. As shown, the PK parameters of the cRFB4 TetraMAb were better than those of either cRFB4 or murine RFB4. This was confirmed by using transgenic mice with hFcRns [9]. All the PK parameters of the TetraMAb were significantly better than that of recombinant divalent MAb (Table 2). The shorter half lives of both MAbs in transgenic mice with hFcRn vs. normal mice were likely due to the genetic background of these homozygous animals. 3.3. Interaction of the MAb and human IgG with the hFcRn

The MAbs were incubated with radiolabeled human FcRn (hFcRn) and the mixture was added to human IgG-Sepharose and further processed as described [8]. The affinity was expressed as a fraction relative to the binding of human IgG which was taken as 1.00. 2.11. Radioiodination

The ability of human IgG and tetravalent cRFB4 to react with the hFcRn was measured using a competitive inhibition assay. The relative affinities of cRFB4 TetraMAb and human IgG for hFcRn were evaluated by measuring their ability to inhibit the interaction of hFcRn with human IgG-Sepharose. The human IgG and the tetravalent cRFB4 MAb had similar

MAbs, human FcRn, and human C1q (Sigma, Louisville, MA) were radiolabeled with Na125I (Amersham) and purified as described [6].

200 kDa 150 kDa

2.12. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE)

100 kDa 50 kDa

Proteins were analyzed in a PhastSystem (Amersham) using 4–15% gels as described [6].

25 kDa

3. Results 1

3.1. Purification and characterization of the cMAbs Two recombinant cRFB4 anti-CD22 MAbs were obtained, one divalent and one tetravalent. The structure of

2

3

4

5

6

Fig. 2. SDS-PAGE analysis of the MAbs. Samples were electrophoresed under nonreducing (NR) and reducing (R) conditions. 1. cRFB4 (NR); 2. murine RFB4 (NR); 3. cRFB4 TetraMAb (NR); 4. cRFB4 (R); 5. murine RFB4 (R); 6. cRFB4 TetraMAb (R).

Radioactivity remainig in the body (%)

X. Liu et al. / International Immunopharmacology 6 (2006) 791–799 100

795

Table 2 PKs of the cRFB4 MAbs in transgenic mice with hFcRn MAb

T 1/2 (h)a

AUC

FCR (day

1

) MRT (h)

cRFB4 65.3 F 3.3 5,633 F 279 0.254 F 0.013 84.8 F 5.8 cRFB4 90.8 F 3.2 11,973 F 858 0.182 F 0.006 129.4 F 4.8 TetraMAb Groups of four mice were used; one experiment was performed. Abbreviations: see Table 1. a The half life of cRFB4 TetraMAb vs. cRFB4 is significantly different ( p b 0.001). 50 0

20

40

60

80

100

120

140

160

3.5. Dissociation of MAbs from Daudi cells

Time (hours)

.

Fig. 3. Elimination curves of the MAbs in mice. ( ) murine RFB4; (E) cRFB4; (n) cRFB4 TetraMAb. This is one representative experiment.

relative affinities for the hFcRn (1.00 F 0.21 vs. 0.91 F 0.12) suggesting that the cRFB4 TetraMAb should have the same half life in humans as endogenous IgG. 3.4. Binding of the MAbs to Daudi, U937 and WEHI-274.1 cells

The dissociation of the radiolabeled TetraMAbs vs. divalent MAbs from Daudi cells in the presence of cold RFB4 was compared. From the dissociation curves (Fig. 6) the persistence of the MAbs on the cell surface was calculated and expressed as the T 1/2 of dissociation. This value was 4-fold longer for the TetraMAb as compared to the divalent MAbs, suggesting that more than two binding sites of the tetravalent MAb are able to bind simultaneously. 3.6. Internalization of the MAbs by Daudi cells

The functional affinity of the cRFB4 TetraMAb was compared to that of the cRFB4 and the murine RFB4 using Scatchard plots of radiolabeled MAbs binding CD22+ Daudi cells. The results (Fig. 4) demonstrate that there was no significant difference between the binding affinity of the cRFB4 TetraMAb and its divalent counterpart. The binding of MAbs to human U937 and murine WEHI274.1 cells was also determined. U937 cells express FcgR I/ III and WEHI-274.1 cells express FcgR I. MAb binding to both FcgRs is dependent upon the presence of an intact Fc region. The results presented in Fig. 5 demonstrate that there was no significant difference in the binding of the cRFB4 TetraMAb vs. the divalent cRFB4 to either cell line. As expected, the murine RFB4 bound poorly to human U937 cells due to the fact that mouse Fc binds poorly to human FcgRs. These results clearly indicate that the cRFB4 TetraMAb should interact with FcgR-bearing effector cells and mediate antibody-dependent cellular cytotoxicity (ADCC) in both humans and mice.

The intracellular distribution of radiolabeled MAbs in Daudi cells is shown in Table 3. The internalization of membrane bound RFB4 MAbs with time indicates that, as compared to the divalent MAb, a larger proportion of the tetravalent MAb is retained on the cell membrane. The percentage of radioactivity released into the medium was increased for the murine MAb due to its higher rate of dissociation. 3.7. Inhibition of the growth of Daudi cells by the MAbs The growth of Daudi cells cultured in the presence of increasing concentrations of tetravalent and divalent MAb was followed for seven days. The results presented in Table 4 demonstrate that while the cRFB4 and the murine RFB4 had no effect even at the high concentration (667 nM), the cRFB4 TetraMAb reduced the number of viable cells by 25% even at a concentration 10-fold lower (66.7 nM).

Table 1 PKs of the RFB4 MAbs in Swiss Webster mice MAb

T 1/2 (h)a

AUC

FCR (day

1

Murine RFB4 cRFB4 cRFB4 TetraMAb

211.2 F 3.8 186.4 F 22.5 278.7 F 26.0

29,565 F 3672 15,543 F 1504 36,973 F 2654

0.078 F 0.001 0.090 F 0.010 0.060 F 0.006

)

MRT (h) 303.2 F 5.4 259.6 F 32.0 401.0 F 37.2

Groups of four mice were used; two experiments were performed. Abbreviations: T 1/2 = half life (beta); AUC = area under the curve; FCR = fractional catabolic rate; MRT = mean residence time. a The half life of cRFB4 TetraMAb vs. RFB4 and cRFB4 is significantly different ( p b 0.002); the half life of cRFB4 vs. RFB4 is not significantly different ( p N 0.125).

796

X. Liu et al. / International Immunopharmacology 6 (2006) 791–799 100

1.6

Bound radioactivity (%)

(r/A-X) x 1014

1.4 1.2 1.0 0.8 0.6 0.4

80 60 40 20

0.2 0.0 1.6

1.8

2.0

2.2

2.4

2.6

2.8

3.0

3.2

0 0

.

Fig. 4. Affinity of the MAbs for Daudi cells (Scatchard plot). ( ) murine RFB4 (k = 0.34  109 M 1); (E) cRFB4 (k = 0.30  109 M 1); (n) cRFB4 TetraMAb (k = 0.52  109 M 1). This is one representative experiment of three performed. Abbreviations: A = the initial concentration of the antibody; A–X = the concentration of the unbound antibody; X = the concentration of the immune complex; r = the number of antibody molecules bound per cell.

Amount bound (ng/107 cells)

A.

50 40 30 20

200

250

.

Fig. 6. Dissociation of the MAbs from Daudi cells. ( ) murine RFB4 (T 1/2 of dissociation = 19.3 F 1.2 min); (E) cRFB4 (T 1/2 of dissociation = 20.5 F 1.3 min); (n) cRFB4 TetraMAb (T 1/2 of dissociation = 75.1 F 5.2 min). This is the average of three experiments.

3.8. Diffusion of radiolabeled MAbs into tissues

3.9. The binding of C1q to MAb-coated Daudi cells 2

4

6

8

Amount added (µg/107 cells/mL)

Amount bound (ng/107 cells)

150

10

0

14 12 10

The binding of radiolabeled human C1q to Daudi cells coated with cRFB4 MAbs was carried out to estimate how much of the complement-binding capacity of the divalent MAb was conserved by the TetraMAb. The results presented in Table 6 demonstrate that the TetraMAb retained full complement-binding activity. Due to the species differences, the murine RFB4 did not have the ability to bind to human C1q.

8 6 Table 3 Uptake of RFB4 MAbs into Daudi cells

4 2 0

0

2 4 6 Amount added (µg/107 cells/mL)

8

Fig. 5. Binding of the MAbs to U937 cells (A) and WEHI-274.1 cells (B). A. ( ) murine RFB4; (E) cRFB4; (n) cRFB4 TetraMAb. B. ( ) murine RFB4; (n) cRFB4 TetraMAb. This is one representative experiment of three performed.

.

100

To determine whether the 33% higher molecular weight of the cRFB4 TetraMAb would have decreased diffusion into the tissues, we examined the distribution of these two MAbs in normal mice 70 h after injection of radiolabeled MAbs. We selected skin, muscle and small intestine as target organs for diffusion of MAbs since it has been shown that the distribution of IgG is a diffuse process occurring primarily in these organs [10–12]. From the results presented in Table 5, there were no significant differences in the percentage of the injected dose in various organs ( p N 0.057). The ratios of cRFB4 TetraMAb / cRFB4 / RFB4 were close to 1.0 for all three organs.

0

B.

50

Time (min)

r x 104

.

Parameter measured Uptake of (% TCA precipitated TetraMAb at radioactivity) 4h 8h Internalized Membrane bound Released into the medium

Uptake of murine MAb at 4h

8h

35.7 F 5.9 35.5 F 10.1 38.8 F 13.0 41.9 F 3.2 54.8 F 2.8 56.7 F 8.9 49.2 F 1.7 38.1 F 8.5 9.5 F 3.1 7.8 F 2.5 12.0 F 1.4 20.0 F 6.2

Average of three separate experiments carried out in triplicate.

X. Liu et al. / International Immunopharmacology 6 (2006) 791–799

797

4. Discussion

Table 5 Diffusion of RFB4 MAbs into mouse tissues at 70 h

Over the past two decades several strategies have been tested to improve the ability of MAbs to kill tumor cells in vivo. The first approach was to decrease their size so that they could penetrate tumors more effectively. To this end Fv regions were employed [13]. While the Fv monomers, dimers or polymers did indeed penetrate tumors more effectively [13] and were useful for imaging when they were radiolabeled [14], they lacked effector functions due to the absence of an Fc portion and hence relied entirely on their ability to induce apoptosis or cell cycle arrest in tumor cells. In addition they were rapidly cleared from the circulation resulting in the need for continuous infusions of large amounts. Fv constructs will likely prove most useful for delivering radionuclides or toxins to cells, since they are rapidly cleared and hence have fewer side effects [15]. In addition they are already being used to image tumors in vivo. The second strategy that has been employed is to increase the binding affinity of MAbs by site specific mutagenesis of their hypervariable regions [16]. However, the advantage of this strategy remains controversial since prolonged retention by the tumor cells nearest the vasculature can result in decreased penetration into the remainder of the tumor [16]. Another strategy, and one that we have favored, is to increase the avidity and valency of the MAb such that it crosslinks its target antigen more effectively. Hypercrosslinking can lead to cell cycle arrest or apoptosis [1,6] and since the Fc region is preserved, the MAb has conserved effector function [6]. However, the problem with this approach is the larger size of the MAb and hence its decreased ability to penetrate tumor tissue. It is also more quickly eliminated from the circulation.

Organ

Table 4 Inhibition of the growth of Daudi cells by RFB4-MAbsa Concentration of MAb ( 109 M

RFB4 cRFB4 cRFB4 TetraMAb a

1

)

6.67

66.7

667b

94.0 F 8.5 97.0 F 5.7 85.5 F 2.1

81.5 F 4.9 83.0 F 5.7 63.0 F 4.9

81.0 F 1.4 79.0 F 4.2 47.0 F 2.8

The data were expressed as the percentage of viable cells relative to the control (without MAb). Average of two separate experiments carried out in duplicate. b The difference between the cRFB4 TetraMAb vs. RFB4 and cRFB4 is significant ( p b 0.012); the difference between the RFB4 and the cRFB4 is not significant ( p N 0.590).

Percent of injected dose per gram of tissue (%ID/g) at 70 h RFB4

Skin 3.97 F 0.38 Muscle 1.13 F 0.41 Small 0.84 F 0.16 intestine

cRFB4

cRFB4 TetraMAb

3.71 F 0.77 1.03 F 0.34 0.73 F 0.06

4.06 F 0.86 1.23 F 0.30 0.90 F 0.28

Groups of four mice were used; two experiments were performed.

Finally, efforts to alter half live [17] and effector functions [18,19] of Fc-containing MAbs are ongoing. The objective of the study described here was to generate a new tetravalent MAb with increased antitumor activity in vitro, conserved effector functions and an in vivo half life similar or longer than that of its divalent counterpart. These three characteristics should lead to better antitumor activity in vivo. The major findings to emerge from this study are as follows: 1. The half life of the cRFB4 TetraMAb was longer than that of divalent murine or recombinant chimeric RFB4. The half life of the cRFB4 TetraMAb was also significantly longer than the previously described cRFB4 TetraMAb [6], apparently due to a change in the hinge region. Hence, by making the Cys–Ser substitution in the lengthened hinge region, we obtained a tetravalent MAb with an in vivo half life longer than that of murine or recombinant cRFB4. Since the PK of cRFB4 TetraMAb cannot be tested in humans, we have used transgenic mice with hFcRns. The results have confirmed that the half life of the TetraMAb is significantly longer than that of the recombinant divalent cRFB4. In agreement with this finding, the affinity of TetraMAb for hFcRn was indistinguishable from that of human IgG. This indicates that the TetraMAb should have a half life in humans that is as long as that of endogenous IgG (~20 days). It is well known that Table 6 C1q binding by MAb-coated Daudi cells RFB4 ligand on Daudi cells

C1q binding (%)a

cRFB4 cRFB4 TetraMAbc Murine RFB4

100b 81.9 F 5.0 4.5 F 4.0

a

Average of three separate experiments carried out in triplicate. The binding of cRFB4 was considered to be 100%; the binding to the negative control (w/o antibody) was subtracted. c The difference between the cRFB4 and the cRFB4 TetraMAb is not significant ( p N 0.5). b

798

X. Liu et al. / International Immunopharmacology 6 (2006) 791–799

FcRn is the receptor that controls IgG homeostasis in mammals [20] and that there is a direct relationship between the half life of an antibody in vivo and its affinity for the FcRn [21]. The very similar hFcRnbinding ability of both human IgG and tetravalent cRFB4 and the comparable PK of tetravalent and divalent MAb clearly indicate that the CH2–CH3 domain interface, involved in the binding of FcRn and controlling the catabolism of IgG [20], is completely functional in the cRFB4 TetraMAb. The exposure of the surface area of the CH2–CH3 cleft is dependent upon the free movement of the CH2 domain, which in turn is controlled by the flexibility of the hinge region [22]. Since the hinge region of the cRFB4 TetraMAb is identical to that of human IgG, the tetravalent construct and human IgG have the same affinity for human FcRns and the cRFB4 TetraMAb has a longer half life than the divalent RFB4 in both normal mice and in hFcRn transgenic mice. 2. The binding (association) affinities of the TetraMAb vs. murine RFB4 were similar as determined by the Scatchard analysis. The fact that no differences were observed is consistent with the behavior of our previously described cRFB4 TetraMAb [6] and is due to the functionally irreversible binding to CD22+ target cells (Daudi) of both the divalent and tetravalent MAbs. This was demonstrated by the virtual lack of dissociation of these ligands in the absence of a competitor. It has been reported by others that the functional affinity (avidity) of MAbs cannot be determined for birreversibleQ antigen–antibody reactions [23] such as those that occur with both the divalent and tetravalent RFB4 MAbs. Hence, dissociation of a MAb from a cell provides a more accurate estimation of the strength of binding [23,24]. By comparing the dissociation rates of both the divalent and tetravalent RFB4 MAbs in the presence of an excess of competitor, we were able to demonstrate the higher binding activity of cRFB4 TetraMAb for CD22. However, by using this procedure, we could not determine the number of combining sites engaged but due to the 4-fold slower dissociation rate of tetravalent MAb, we presume that more than two sites on average are bound. 3. The inhibition of the growth of Daudi cells by the RFB4 TetraMAb but not by the divalent MAbs is consistent with hypercrosslinking by the tetravalent MAb. This confirms our published findings [6] that a TetraMAb inhibits cell growth, due to its ability to hypercrosslink CD22. The longer persistence of cRFB4 TetraMAb on the surface of Daudi cells was confirmed by investigating the intracellular dis-

tribution of these two MAbs on Daudi cells in culture for 4 and 8 h. The percentage of membrane bound radioactivity at 8 h clearly demonstrates that more tetravalent MAb remains bound to the cells. The longer persistence of the cRFB4 TetraMAb on the cells suggests that this MAb will be more effective in mediating ADCC and complement-dependent cytotoxicity in vivo. The penetration of the cRFB4 TetraMAb into several normal mouse tissues was not impaired by its larger size (200 vs. 150 kDa of divalent cRFB4 or murine RFB4). These tissues were chosen since 90% of the extravascular rat IgG is located in the skin, muscle, and adipose tissue [25]. The similar penetration of these two MAbs suggests that they might behave similarly in penetrating tumor tissue in humans. 4. The binding of cRFB4 TetraMAb to FcRs on both human (U937) and mouse (WEHI-274.1) cells was identical to that of the divalent cRFB4 (on U937) and of murine RFB4 (on WEHI-274.1). The identical binding of cRFB4 TetraMAb to mouse WEHI-274.1 cells suggests that this mouse/human chimeric MAb will mediate ADCC in mice with human tumors. Taken together, these results demonstrate that the tetravalent MAb has a completely functional FcRbinding site despite the absence of the CH1 domain and a substitution of Cys-220 with Pro-220 in the upper hinge region which normally forms a disulfide bond with the L chain. This substitution was necessary to avoid the formation of a disulfide bond with the Cys residues of Fv portions. The conserved FcRbinding function in cRFB4 TetraMAb was expected since this binding site is located in the lower hinge region at the N-terminal end of the CH2 domain [26] not affected by the mutation at residue 220. 5. The binding of human C1q by the RFB4 TetraMAb and its divalent counterpart was also similar indicating that this effector function of IgG is conserved in the tetravalent construct. This was expected since the site of C1q-binding of human IgG is located in the lower hinge region and CH2 domain involving amino acid residues which are identical in both the TetraMAb and the divalent MAb [27]. The CH1 domain of IgG is involved in the linking of C3b by a covalent bond [28]. However, deleting the CH1 domain did not alter the ability of IgG to bind to C3 or to activate the complement cascade [29]. This is likely due to the fact that C3b binds not only to the CH1 but also to CH2 domain [30]. Therefore we would predict that the cRFB4 TetraMAb will fix C1q and activate the complement pathway leading to the lysis of the targeted tumor cells.

X. Liu et al. / International Immunopharmacology 6 (2006) 791–799

In summary, this study has demonstrated that a tetravalent recombinant antibody has longer in vivo half life and the same effector functions as its divalent counterpart. Importantly it is more effective than the divalent MAb in killing tumor cells in vitro. We will now determine whether this TetraMAb has improved antitumor activity in SCID mice xenografted with CD22+ lymphomas. Acknowledgments We thank Linda Berry and Erica Garza for assistance in preparing the manuscript. We also thank Drs. Pamela Bjorkman, Anthony West and Devin Tesar (California Institute of Technology) for their generous gift of the recombinant FcRn and Thomas Sproule for excellent technical assistance. References [1] Ghetie MA, Podar EM, Ilgen A, Gordon BE, Uhr JW, Vitetta ES. Homodimerization of tumor-reactive monoclonal antibodies markedly increases their ability to induce growth arrest or apoptosis of tumor cells. Proc Natl Acad Sci U S A 1997;94: 7509 – 14. [2] Ghetie MA, Bright H, Vitetta ES. Homodimers but not monomers of Rituxan (chimeric anti-CD20) induce apoptosis in human B-lymphoma cells and synergize with a chemotherapeutic agent and an immunotoxin. Blood 2001;97:1392 – 8. [3] Wolff EA, Schreiber GJ, Cosand WL, Raff HV. Monoclonal antibody homodimers: enhanced antitumor activity in nude mice. Cancer Res 1993;53:2560 – 5. [4] Shopes B. A genetically engineered human IgG mutant with enhanced cytolytic activity. J Immunol 1992;148:2918 – 22. [5] Miller K, Meng G, Liu J, Hurst A, Hsei V, Wong WL, et al. Design, construction, and in vitro analyses of multivalent antibodies. J Immunol 2003;170:4854 – 61. [6] Meng R, Smallshaw JE, Pop LM, Yen M, Liu X, Le L, et al. The evaluation of recombinant, chimeric, tetravalent anti-human CD22 antibodies. Clin Cancer Res 2004;10:1274 – 81. [7] Kim JK, Tsen MF, Ghetie V, Ward ES. Evidence that the hinge region plays a role in maintaining serum levels of the murine IgG1 molecule. Mol Immunol 1995;32:467 – 75. [8] Pop LM, Liu X, Ghetie V, Vitetta ES. The generation of immunotoxins using chimeric anti-CD22 antibodies containing mutations which alter their serum half-life. Int Immunopharmacol 2005;5:1279 – 90. [9] Roopenian DC, Christianson GJ, Sproule TJ, Brown AC, Akilesh S, Jung N, et al. The MHC class I-like IgG receptor controls perinatal IgG transport, IgG homeostasis, and fate of IgG-Fccoupled drugs. J Immunol 2003;170:3528 – 33. [10] Henderson LA, Baynes JW, Thorpe SR. Identification of the sites of IgG catabolism in the rat. Arch Biochem Biophys 1982;215:1 – 11. [11] Covell DG, Barbet J, Holton OD, Black CD, Parker RJ, Weinstein JN. Pharmacokinetics of monoclonal immunoglobulin G1, F(abV)2, and FabV in mice. Cancer Res 1986;46:3969 – 78.

799

[12] Borvak J, Richardson J, Medesan C, Antohe F, Radu C, Simionescu M, et al. Functional expression of the MHC class I-related receptor, FcRn, in endothelial cells of mice. Int Immunol 1998;10:1289 – 98. [13] Kortt AA, Dolezal O, Power BE, Hudson PJ. Dimeric and trimeric antibodies: high avidity scFvs for cancer targeting. Biomol Eng 2001;18:95 – 108. [14] Goel A, Baranowska-Kortylewicz J, Hinrichs SH, Wisecarver J, Pavlinkova G, Augustine S, et al. 99mTc-labeled divalent and tetravalent CC49 single-chain Fv’s: novel imaging agents for rapid in vivo localization of human colon carcinoma. J Nucl Med 2001;42:1519 – 27. [15] Huhalov A, Chester KA. Engineered single chain antibody fragments for radioimmunotherapy. Q J Nucl Med Mol Imaging 2004;48:279 – 88. [16] Weiner LM, Carter P. Tunable antibodies. Nat Biotechnol 2005;23:556 – 7. [17] Ghetie V, Popov S, Borvak J, Radu C, Matesoi D, Medesan C, et al. Increasing the serum persistence of an IgG fragment by random mutagenesis. Nat Biotechnol 1997;15:592 – 8. [18] Shields RL, Namenuk AK, Hong K, Meng YG, Rae J, Briggs J, et al. High resolution mapping of the binding site on human IgG1 for FcgRI, FcgRII, FcgRIII and FcRn and design of IgG1 variants with improved binding to FcR. J Biol Chem 2001;276: 6591 – 604. [19] Idusogie EE, Wong PY, Presta LG, Gazzano-Santoro H, Totpal M, Ultsch M, et al. Engineered antibodies with increased activity to recruit complement. J Immunol 2001;166:2571 – 5. [20] Ghetie V, Ward ES. Multiple roles for the major histocompatibility complex class I-related receptor FcRn. Annu Rev Immunol 2000;18:739 – 60. [21] Medesan C, Matesoi D, Radu C, Ghetie V, Ward ES. Delineation of the amino acid residues involved in transcytosis and catabolism of mouse IgG1. J Immunol 1997;158:2211 – 7. [22] Edmundson AB, Luke VG, Rosauer R.A, Andersen KN, Shan ZC, Fan ZC. Three-dimensional aspects of IgG structure and function. In: Zanetti M, Capra JDThe antibodies, vol. 1. Australia7 Harwood Acad. Publ., 1995. p. 41 – 100. [23] Kyriakos RJ, Shih LB, Ong GL, Patel K, Goldenberg DM, Mattes MJ. The fate of antibodies bound to the surface of tumor cells in vitro. Cancer Res 1992;52:835 – 42. [24] Ong GL, Mattes MJ. Re-evaluation of the concept of functional affinity as applied to bivalent antibody binding to cell surface antigens. Mol Immunol 1993;30:1455 – 62. [25] O’Connor SW, Bale WF. Accessibility of circulating immunoglobulin G to the extravascular compartment of solid rat tumors. Cancer Res 1984;44:3719 – 23. [26] Hulett MD, Hogarth PM. Molecular basis of Fc receptor function. Adv Immunol 1994;57:1 – 127. [27] Nezlin R, Ghetie V. Interactions of immunoglobulins outside the antigen-combining site. Adv Immunol 2004;82:155 – 215. [28] Shohet JM, Pemberton P, Carroll MC. Identification of a major binding site for complement C3 on the IgG1 heavy chain. J Biol Chem 1993;268:5866 – 71. [29] Munoz E, Vidarte L, Casado MT, Pastor C, Vivanco F. The C(H)1 domain of IgG is not essential for C3 covalent binding: importance of the other constant domains as targets for C3. Int Immunol 1998;10:97 – 106. [30] Anton LC, Ruiz S, Barrio E, Marques G, Sanchez A, Vivanco F. C3 binds with similar efficiency to Fab and Fc regions of IgG immune aggregates. Eur J Immunol 1994;24:599 – 604.