Effect of intravenous immunoglobulin (IVIg) on primate complement-dependent cytotoxicity of genetically engineered pig cells: relevance to clinical xenotransplantation

Abstract

Triple-knockout (TKO) pigs may be ideal sources of organs for clinical xenotransplantation because many humans have no preformed antibody to TKO pig cells. Intravenous immunoglobulin (IVIg) is widely used for severe infection or the treatment/prevention of antibody-mediated rejection in allotransplantation. Anti-pig antibodies in IVIg could be harmful in clinical xenotransplantation. It is unknown whether anti-TKO pig antibodies are present in IVIg. The main aim of this study was to investigate in vitro whether IVIg contains anti-TKO pig antibodies with cytotoxic effect to pig cells. Undiluted pooled human serum (HS) and five different commercial preparations of IVIg were tested for IgM and IgG binding to red blood cells (RBCs) from wild-type (WT), α1,3-galactosyltransferase gene-knockout (GTKO), and TKO pigs by flow cytometry. Complement-dependent lysis of IVIg against these pig pRBCs was measured by hemolytic assay. Pooled HS and 4 of 5 IVIg commercial preparations contained anti-pig IgG that bound to WT and GTKO pRBCs, but not to TKO pRBCs. One preparation of IVIg contained antibodies that bound to TKO pRBCs, but there was no cytotoxicity of IVIg to TKO pRBCs. The results suggest that IVIg administration to human recipients of TKO pig grafts would be safe. However, the specific preparation of IVIg would need to be screened before its administration.

Figure 1

(A) IgG and IgM binding of pooled human serum and five different brands of commercial IVIg to WT, GTKO, and TKO pRBCs. (Note the different scales on the Y axis between WT, GTKO, and TKO pRBCs.) There was significant human serum IgG binding to WT and GTKO pRBCs, but not to TKO pRBCs. There was no IgG binding to TKO pRBCs by pooled human serum or by any IVIg, except for GAMUNEX-C. On the y axis, the dotted line represents the lowest measurable limit of binding, below which there is considered to be no binding (relative GM: IgM 1.2, IgG 1.1). Results are expressed as mean + /− SD (*p < 0.05, **p < 0.01). There was significant human serum IgM binding to WT and GTKO pRBCs, but not to TKO pRBCs. There was no IgM binding to TKO pRBCs by pooled human serum or by any IVIg. See M“Materials and methods” section (In vitro binding of IgG/IgM in human serum and IVIg to pRBCs). Heat-inactivated serum was used in this assay. On the y axis, the dotted line represents the lowest measurable limit of binding, below which there is considered to be no binding (relative GM: IgM 1.2, IgG 1.1). Results are expressed as mean + /− SD (*p < 0.05, **p < 0.01). (B) IgG/IgM binding of pooled human serum and IVIg (FLEBOGAMMA) to WT and GTKO pAECs. There was no difference in IgG and IgM binding to WT pAECs between pooled human serum and high-dose IVIg. There was no IgG binding of IVIg to GTKO pAECs (bottom). (TKO pAECs were not available to us.) See “Materials and methods” section (In vitro binding of IgG/IgM in human serum and IVIg to pAECs). Heat-inactivated serum was used in this assay. On the y axis, the dotted line represents the lowest measurable limit of binding, below which there is considered to be no binding (relative GM: IgM 1.2, IgG 1.1). Results are expressed as mean + /− SD (*p < 0.05).

Figure 1

(A) IgG and IgM binding of pooled human serum and five different brands of commercial IVIg to WT, GTKO, and TKO pRBCs. (Note the different scales on the Y axis between WT, GTKO, and TKO pRBCs.) There was significant human serum IgG binding to WT and GTKO pRBCs, but not to TKO pRBCs. There was no IgG binding to TKO pRBCs by pooled human serum or by any IVIg, except for GAMUNEX-C. On the y axis, the dotted line represents the lowest measurable limit of binding, below which there is considered to be no binding (relative GM: IgM 1.2, IgG 1.1). Results are expressed as mean + /− SD (*p < 0.05, **p < 0.01). There was significant human serum IgM binding to WT and GTKO pRBCs, but not to TKO pRBCs. There was no IgM binding to TKO pRBCs by pooled human serum or by any IVIg. See M“Materials and methods” section (In vitro binding of IgG/IgM in human serum and IVIg to pRBCs). Heat-inactivated serum was used in this assay. On the y axis, the dotted line represents the lowest measurable limit of binding, below which there is considered to be no binding (relative GM: IgM 1.2, IgG 1.1). Results are expressed as mean + /− SD (*p < 0.05, **p < 0.01). (B) IgG/IgM binding of pooled human serum and IVIg (FLEBOGAMMA) to WT and GTKO pAECs. There was no difference in IgG and IgM binding to WT pAECs between pooled human serum and high-dose IVIg. There was no IgG binding of IVIg to GTKO pAECs (bottom). (TKO pAECs were not available to us.) See “Materials and methods” section (In vitro binding of IgG/IgM in human serum and IVIg to pAECs). Heat-inactivated serum was used in this assay. On the y axis, the dotted line represents the lowest measurable limit of binding, below which there is considered to be no binding (relative GM: IgM 1.2, IgG 1.1). Results are expressed as mean + /− SD (*p < 0.05).

Figure 2

Cytotoxicity of pooled human serum against RBCs. There was no cytotoxicity of pooled human serum against TKO pRBCs or human blood type O RBCs. When the concentration of human serum was 25% or less, the cytotoxicity against GTKO pRBCs was significantly lower than that against WT pRBCs (p < 0.01). See “Materials and methods” section (Antibody-dependent complement-mediated hemolytic assay). Briefly, RBCs (800 × 10⁶/ml in 75 µl) were isolated and placed in 5 ml round-bottom tubes. Titrated non-heat-inactivated serum (i.e. with complement activity; 450 µl) with PBS and 5% sorbitol (375 µl) instead of IVIg was added to each tube (total 900 µl) and incubated at 37 °C for 150min. Control samples consisted of PBS (instead of the RBC solution) + 450 µl of titrated non-heat-inactivated serum with PBS + 5% sorbitol. After centrifugation at 910 g for 5 min, the supernatant was collected, and each 300 µl was transferred into 96-well plates. The released hemoglobin was measured at an optical density (OD) of 541 nm using a SpectraMax M2e plate reader (Molecular Devices Corp). Data were obtained in duplicate. Results are expressed as mean + /− SD. The dotted line represents cut-off value (7%). (**p < 0.01).

Figure 3

The cytotoxicty of IVIg (FLEBOGAMMA) + /− rabbit complement on the lysis of RBCs (A) cytotoxicity of IVIg (FLEBOGAMMA) against RBCs. The cytotoxicity of IVIg against WT, GTKO and TKO pRBCs. As a positive control (PC) for WT and GTKO pRBCs, non-heat-inactivated pooled human (Hu) serum was used. As the PC for TKO pRBCs, a naïve baboon non-heat-inactivated serum was used, because pooled Hu serum had no cytotoxicity to TKO pRBCs. Human blood type O RBCs were used as the negative control. For details of methods see “Materials and methods” section (Antibody-dependent complement-mediated cytotoxicity of IVIg alone to pRBCs). Results are expressed as mean + /− SD. The dotted line represents cut-off value (7%), below which cytotoxicity is considered negative. (B) Cytotoxicity of pRBCs of IVIg (FLEBOGAMMA) (a) without rabbit complement, (b) with rabbit complement, and (c) after washing away soluble factors from IVIg before adding rabbit complement. For details of methods see Supplementary Fig. 3 and “Materials and methods” section (Antibody-dependent complement-mediated cytotoxicity of IVIg alone to pRBCs). . When no rabbit complement was added (a), there was no cytotoxicity of IVIg to WT, GTKO, or TKO pRBCs. Even when adding complement (b), there was almost no cytotoxicity. However, after washing away soluble factors from the IVIg before adding rabbit complement (c), the cytotoxicity of IVIg against WT and GTKO pRBCs was significantly increased [compared to when there was no washing (b)] because the soluble factors in the IVIg had had a protective effect. However, the cytotoxicity of IVIg against TKO pRBCs remained negative. Results are expressed as mean + /− SD. The dotted line represents cut-off value (7%). (**p < 0.01). (C) Cytotoxicity associated with anti-pig antibodies in titrated IVIg (FLEBOGAMMA) against WT, GTKO, and TKO pRBCs. When the concentration of IVIg was high (> 10 mg/ml), the cytotoxicity associated with anti-WT pig antibodies was significantly higher than that associated with antibodies to GTKO or TKO pRBCs. No cytotoxicity was associated with the binding of anti-TKO pig antibodies in the IVIg, even though at high concentration of IVIg. Results are expressed as mean + /− SD. The dotted line represents cut-off value (7%). (**p < 0.01).

Figure 3

The cytotoxicty of IVIg (FLEBOGAMMA) + /− rabbit complement on the lysis of RBCs (A) cytotoxicity of IVIg (FLEBOGAMMA) against RBCs. The cytotoxicity of IVIg against WT, GTKO and TKO pRBCs. As a positive control (PC) for WT and GTKO pRBCs, non-heat-inactivated pooled human (Hu) serum was used. As the PC for TKO pRBCs, a naïve baboon non-heat-inactivated serum was used, because pooled Hu serum had no cytotoxicity to TKO pRBCs. Human blood type O RBCs were used as the negative control. For details of methods see “Materials and methods” section (Antibody-dependent complement-mediated cytotoxicity of IVIg alone to pRBCs). Results are expressed as mean + /− SD. The dotted line represents cut-off value (7%), below which cytotoxicity is considered negative. (B) Cytotoxicity of pRBCs of IVIg (FLEBOGAMMA) (a) without rabbit complement, (b) with rabbit complement, and (c) after washing away soluble factors from IVIg before adding rabbit complement. For details of methods see Supplementary Fig. 3 and “Materials and methods” section (Antibody-dependent complement-mediated cytotoxicity of IVIg alone to pRBCs). . When no rabbit complement was added (a), there was no cytotoxicity of IVIg to WT, GTKO, or TKO pRBCs. Even when adding complement (b), there was almost no cytotoxicity. However, after washing away soluble factors from the IVIg before adding rabbit complement (c), the cytotoxicity of IVIg against WT and GTKO pRBCs was significantly increased [compared to when there was no washing (b)] because the soluble factors in the IVIg had had a protective effect. However, the cytotoxicity of IVIg against TKO pRBCs remained negative. Results are expressed as mean + /− SD. The dotted line represents cut-off value (7%). (**p < 0.01). (C) Cytotoxicity associated with anti-pig antibodies in titrated IVIg (FLEBOGAMMA) against WT, GTKO, and TKO pRBCs. When the concentration of IVIg was high (> 10 mg/ml), the cytotoxicity associated with anti-WT pig antibodies was significantly higher than that associated with antibodies to GTKO or TKO pRBCs. No cytotoxicity was associated with the binding of anti-TKO pig antibodies in the IVIg, even though at high concentration of IVIg. Results are expressed as mean + /− SD. The dotted line represents cut-off value (7%). (**p < 0.01).

Figure 4

(A) The competitive effect of IVIg (FLEBOGAMMA) on IgG/IgM binding of pooled human serum to WT/GTKO pRBCs. High-dose (40 mg/ml) IVIg attenuated human serum IgG binding (but not IgM binding) to WT pRBCs (p < 0.01). IVIg increased human serum IgG binding to GTKO pRBCs, and increased IgM binding to both WT and GTKO pRBCs. See “Materials and methods” section (Competitive binding to pRBCs of IVIg with IgG and IgM from pooled human serum). Heat-inactivated serum was used in this assay. On the y axis, the dotted line represents the lowest measurable limit of binding, below which there is considered to be no binding (relative GM: IgM 1.2, IgG 1.1). Results are expressed as mean + /− SD (**p < 0.01). (B) The competitive effects of IVIg (FLEBOGAMMA) on IgG/IgM binding of pooled human serum to WT/GTKO pAECs. IVIg did not attenuate human serum IgG binding to either WT or GTKO pAECs and increased IgM binding. There was no IgG binding of pooled human serum or IVIg to GTKO pAECs. See “Materials and methods” section (Competitive binding to pAECs of IVIg with IgG and IgM from pooled human serum). Heat-inactivated serum was used in this assay. On the y axis, the dotted line represents the lowest measurable limit of binding, below which there is considered to be no binding (relative GM: IgM 1.2, IgG 1.1). Results are expressed as mean + /− SD.

Figure 5

The effect of IVIg (FLEBOGAMMA) on the cytotoxicity of pooled human serum (50%) against pRBCs. (A) The cytotoxicity of pooled human serum (50%) against WT pRBCs was not inhibited by IVIg. (B) The cytotoxicity of pooled human serum (50%) against GTKO pRBCs was inhibited by high-dose IVIg (*p < 0.05). (C) There was no cytotoxicity of pooled human serum (50%) against TKO pRBCs with/without IVIg. The cytotoxicity of pooled human serum (50%) against (D) WT and (E) GTKO pAECs was significantly inhibited by high-dose (> 10 mg/ml) IVIg. See “Materials and methods” section [(A)–(C) Antibody-dependent complement-mediated hemolytic assay, and (D, E) Antibody-dependent complement-mediated cytotoxicity (CDC) of pAECs]. (A)–(C) Non-heat-inactivated serum (i.e., with complement activity) was used in the hemolytic assay. (D, E) Heat-inactivated serum + rabbit complement (i.e., exogenous complement) was used in the CDC assay of pAECs. The dotted line represents cut-off value (7%), below which cytotoxicity is considered negative. Results are expressed as mean + /− SD (*p < 0.05, **p < 0.01).

Figure 6

IgG and IgM binding to pRBCs after IVIg (FLEBOGAMMA) (2 g/kg) administration to a naive baboon. (A) In a naïve baboon, IgG binding to WT pRBCs was significantly increased after i.v. administration of 2 g/kg IVIg. IgM binding to WT pRBCs was significantly increased 1 and 6 days after IVIg. There was no significant difference in IgG or IgM binding to (B) GTKO, and (C) TKO pRBCs after i.v. administration of 2 g/kg IVIg. See “Materials and methods” section [Binding of anti-pig IgG and IgM, hemolytic assay, and serum complement levels in a baboon administered 2 g/kg of IVIg (FLEBOGAMMA)]. Heat-inactivated serum was used in this assay. On the y axis, the dotted line represents the lowest measurable limit of binding, below which there is considered to be no binding (relative GM: IgM 1.2, IgG 1.1). Results are expressed as mean + /− SD (*p < 0.05, **p < 0.01).

Figure 7

Complement activity after IVIg (FLEBOGAMMA) administration in a naïve baboon. The plasma C3a levels (A) immediately (day 0) after i.v. IVIg administration were significantly decreased [p < 0.05]. By days 6 and 13, however, they were significantly increased (p < 0.01). In contrast, the plasma Bb level (B) immediately after IVIg was significantly increased (p < 0.01). See “Materials and methods” section [Binding of anti-pig IgG and IgM, hemolytic assay, and serum complement levels in a baboon administered 2 g/kg of IVIg (FLEBOGAMMA)]. Results are expressed as mean + /− SD (*p < 0.05, **p < 0.01).

Figure 8

Serum cytotoxicity of baboon serum against WT pRBCs after i.v. administration of IVIg (FLEBOGAMMA) (2 g/kg). (A) The cytotoxicity of baboon serum (final concentration 50%) immediately (day 0) and on day 1 after IVIg (2 mg/kg) against WT pRBCs was significantly decreased. The dotted line represents cut-off value (7%), below which cytotoxicity is considered negative. See “Materials and methods” section [Binding of anti-pig IgG and IgM, hemolytic assay, and serum complement levels in a baboon administered 2 g/kg of IVIg (FLEBOGAMMA)]. Non-heat-inactivated serum (i.e. complement activity +) was used in this assay. Results are expressed as mean + /− SD (*p < 0.05, **p < 0.01). (B) Immediately after IVIg, lysis of WT pRBCs was completely inhibited.