Bioconjugate Chem. 1994, 5, 577-582
577
Temperature-Responsive Bioconjugates. 3. Antibody-Poly(Wisopropylacry1amide) Conjugates for Temperature-ModulatedPrecipitations and Affinity Bioseparations? Yoshiyuki G. Takei, Miki Matsukata, Takashi Aoki, Kohei Sanui, Naoya Ogata, Akihiko Kikuchi,* Yasuhisa Sakurai,*a n d Teruo Okano",' Department of Chemistry, Faculty of Science and Technology, Sophia University, 7-1 Kioi-cho, Chiyoda, Tokyo 102, Japan, and Institute of Biomedical Engineering, Tokyo Women's Medical College, 8-1 Kawada-cho, Shinjuku, Tokyo 162, Japan. Received July 22, 1994@
Immunoglobulin G (IgG) has been modified by poly(N-isopropylacrylamide)(PIPAAm)to create a novel bioconjugate which exhibits reversible phase transition behavior at 32 "C in aqueous media. A terminal carboxyl group introduced into PIPAAm molecule by polymerization of IPAAm with 3-mercaptopropionic acid was used for conjugation to IgG via coupling reaction of activated ester with protein amino group. These conjugates exhibited rapid response to changes in solution temperature and significant phase separation above a critical solution temperature corresponding to that for the original PIPAAm. These conjugates bound to antigen quantitatively in aqueous system, and antigen-bound complex also demonstrated phase separation and precipitation above a critical temperature. Precipitate was reversibly redissolved in cold buffer. Though particular conjugate which includes 12 molecules of PIPAAm with 6,100 molecular weight suppressed more than 95% of Fc-dependent binding with protein A, it retained approximately 60% of original specific antigen binding activity. It was manifested that polymer content of conjugate was 20-30 wt% for the case of 6,100 molecular weight of PIPAAm to demonstrate specific antigen binding activity most effectively and to reduce Fc-dependent binding with protein A. IgG-PIPAAm conjugates were soluble in water and formed antigen-bound complex in homogeneous solution system below a critical temperature. These conjugates were separated from solution and other solutes corresponding to PIPAAm nature and scarcely bound to antigen above a critical temperature. It is revealed that temperature-responsive PIPAAm conjugated to biomolecule operated as a switching molecule. These phenomena are attractive for not only reversible bioreactors and protein separations but also carrier substrate to localize biomolecules such as drugs, peptides and hormones in a living body.
INTRODUCTION
In 1977, Abuchowski and coworkers demonstrated that covalent attachment of poly(ethy1ene glycol) (PEG) to a protein leads to minimal loss of activity and decreases protein immunogenicity and antigenicity entirely (1,2). Since this time, many investigators have used PEGmodified biomolecules for chemical, biotechnological and biomedical applications, termed PEG-ylation (3). Bioactivity of PEG-modified proteins has been investigated and correlated to their structures. Bioactivities were strongly affected by the conjugated polymer chain mobility, corresponding to PEG polymer chain length (4-6). While PEG-ylation modifies biomolecule interface, it does not introduce any stimuli-response into these biomolecules to control the activity of biomolecules using external stimuli. It is well-known that poly(N4sopropylacrylamide) (PIPAAm)' exhibits a remarkable phase transition in aque-
* Author to whom correspondence should be addressed. Part 1: see ref 23. Part 2: ref 24.
* Tokyo Women's Medical College.
Abstract published in Advance ACS Abstracts, November 1, 1994. Abbreviations used: IPAAm, N-isopropylacrylamide; PIPAAm, poly-IPAAm; DMF, N,N-dimethylformamide; IgG, immunoglobulin G; HSA, human serum albumin; FITC -HSA, fluorescein isothiocyanate-labeled HSA; G, goat IgG; AG, antihuman serum albumin goat IgG; G-X, G-PIPAAm conjugate; AG-X, AG-PIPAAm conjugate. @
1043-1802/94/2905-0577$04.50/0
ous media in response to changes in temperature, demonstrating a lower critical solution temperature (LCST) (7-9). The temperature-responsive phase transition behaviors of these networks have been investigated and utilized for drug delivery systems (10-13), cell culture substrates (14, 15), and immobilized enzymes (16, 17). Hoffman and co-workers have reported the synthesis and characterization of PIPAAm-biomolecule conjugates as devices for affinity immunoassay and bioseparation (1822). In these biomolecule conjugates, N-(acry1oxy)succinimide was used for conjugation of biomolecules to PIPAAm. Therefore, the characteristic of the conjugate is multipoint binding between biomolecule and the copolymer. It is sometimes difficult to control the solubility because of the formation of the crosslinking of the conjugate, and multipoint conjugation may also lead denaturalization of biomolecules. We previously reported the synthesis of temperatureresponsive PIPAAm with a carboxyl group a t one end (semitelechelic PIPAAm) by polymerization of IPAAm using 3-mercaptopropionic acid as a telogen (23). We demonstrated that the number of conjugated PIPAAm per biomolecule affected the temperature-responsive behavior of the conjugate and the energy required to dehydrate the whole conjugate (24). PIPAAm-biomolecule conjugates using semitelechelic PIPAAm as a phase transition inducer achieved both rapid responses to changes in temperature and drastic phase separation. These conjugates were conveniently separated from reactive products and other solutes with small temperature 1994 American Chemical Society
578 Bioconjugate Chem., Vol. 5, No. 6, 1994
Takei et al.
Table 1. Preparation and Analysis of IgG-PIPAAm Conjugates
code G-0 G-3 G-13 G-21 AG-0 AG-4 AG-12 AG-20
IeG nonspecific
activated PIPAAmU (moVmol IgG)
anti-HSA
molecular w t of conjugates ( x lo3)
PIPAAm content
25 50
150 167 220 264
titrationC 150 170 223 276
moVmol IgGC 0 3.2 13.1 20.6
0 10 25 50
150 169 212 245
150 177 223 273
4.4 12.0 20.1
0 10
HPSECb
wt
%d
LCST ("C)
10.7 31.6 43.2
34.4 33.8 33.6
11.2 29.3 38.8
34.2 34.0 33.6
0
a PIPAAm with M , = 6100 in ref 23. Determined by HPSEC. Estimated by consuming residual primary amino groups using fluorescamine. (wt P I P W w t conjugate) 100.
increases (23-25). It is reasoned therefore that conjugation of semitelechelic PIPAAm to IgG molecules would not only maintain high IgG-antigen binding activity associated with IgG conjugation to one reactive end group per polymer but also decrease protein immunogenicity with local and large steric hindrance due to the inherently high mobile nature of polymer free-end. In this paper, the preparation of immunoglobulin G (1gG)-PIPAAm conjugates using semitelechelic PIPAAm with a carboxyl end group is described. Effect of the number of conjugated PIPAAm molecules on both temperature-responsive phase transition behavior and specific antigen binding activity of these conjugates is reported. In addition, the IgG Fc region demonstrates biological activity characteristic of immunoglobulins and their subclasses such as complement fixation. One of the major problems is that cellular Fc receptors exhibit undesirable entrapment of conjugate and interfere with binding specificity with particular antigens in vivo. Fluorescein isothiocyanate-labeled human serum albumin (FITC-HSA) was used as a model antigen, and specific antigen binding activity of the conjugate was estimated by measuring FITC fluorescence. Antigen binding activity was investigated by fluorescence spectroscopy, and the Fc-dependent binding activity was also estimated from the suppression of binding to protein A. The potential value of PIPAAm with a carboxyl end group as switching molecule for reversible bioreactors and temperature-modulated biochemotherapy is indicated. EXPERIMENTAL PROCEDURES
Preparation of Polymers and Biochemicals. Semitelechelic poly(N-isopropylacrylamide)(PIPAAm) with a carboxyl end group was synthesized by polymerization of IPAAm with 3-mercaptopropionic acid as a chain transfer agent in N,N-dimethylformamide (DMF) as described in our previous work (23). PIPAAm molecules of mol wt ca. 6100 ( M J M , = 1.22), one carboxyl group per polymer chain, and exhibiting LCST near 32 "C were used for conjugation with IgG. Anti-human serum albumin goat IgG (AG) and fluorescein isothiocyanatelabeled human serum albumin (FITC-HSA) were purchased from Cappel Research Products, USA, and used as received. Goat IgG reagent grade (G)and fluorescamine were obtained from Sigma Chemical Co. Ultrapure water used for sample solutions was provided by a commercial water purification device (LV-lOT, Toray, Japan). All other reagents were from Wako Pure Chemicals Co., Japan. Solvents were reagent grade and purified by conventional methods. Conjugation of PIPAAm to IgG. PIPAAm with a carboxyl end group was activated by N-hydroxysuccinimide with dicyclohexylcarbodiimide in dry ethyl acetate
a t 4 "C for 16 h in a molar ratio of 1:2:2, respectively. After filtration and concentration, the reactant was poured into diethyl ether to precipitate activated-PIPAAm. Activated PIPAAm was purified by reprecipitation with ethyl acetate/diethyl ether twice. Activated ester group was confirmed by infrared and ultraviolet spectroscopy (26). IgG-PIPAAm conjugates were synthesized following the protocol of Biickman et al. with a weight feed ratio of PIPAAm to IgG of 10-50 (27). The reaction solution was adjusted to and maintained a t pH 8.5 during the course of the reaction, and all steps were carried out a t 4 "C. The IgG was dissolved in 0.1 M carbonate-bicarbonate buffer (0.15 M NaCl, pH 8.51, and the IgG concentration of the solution was adjusted to 2 mg/mL. Activated-PIPAAm was dissolved in 4 mL of dry DMF and added to the IgG solution. This protocol was repeated four more times a t 30 min intervals. Total reaction time was 8 h with gentle stirring a t 4 "C. The solution was then dialyzed against phosphate buffer solution (PBS; 0.15 M NaC1, pH 7.4) using cellulose porous membrane tubing (50 000 molecular weight cutoff, Spectrum Medical Industries, USA) for 24 h a t 4 "C and then lyophilized. The IgG-PIPAAm conjugates were stored a t -20 "C in a biofreezer, and the IgG concentration was determined by the biuret method (28). Similar methods were also used to prepare the conjugates using anti-human serum albumin goat IgG for the antigen binding-activity measurement. The number of PIPAAm conjugated to IgG was estimated by measuring the amount of amino groups consumed during the conjugation process using fluorescamine-based assay as follows (29, 30). A series of conjugate solutions was prepared with PBS a t the following concentrations: 0,0.7, 1.2, 1.6, and 2.5 pg/mL. Fluorescamine (0.3 mg/mL) in acetone (0.5mL) was added to each 1.5 mL Eppendorf tube containing 1mL of sample solution while vortexing, and then solutions were incubated for 10 min a t 4 "C. The fluorescence of the solution was measured on a spectrofluorometer using a microvolume observation cell (0.8mL maximum) with an excitation wavelength of 390 nm and emission a t 475 nm. Preparation and analysis of IgGPIPAAm conjugates are summarized in Table 1. Molecular Weight Measurement. Molecular weights of the resulting IgG-PIPAAm conjugates were measured by high-performance size-exclusion chromatography (HPSEC; HLC-802, equipped with TSK-G-3000SW,Toso, Japan) using sodium azide-containing PBS (0.15 M NaC1, 3 mM NaN3, pH 7.4) as a mobile phase at 4 "C. A calibration curve for globular proteins was obtained by the retention of standard proteins, obtained from Boehringer Mannheim Biochemica, including /3-galactosidase (MI = 465 OOO), IgG ( M , = 150 000), IgG-Fab fragment ( M , = 50 000), myoglobin (MI = 17 OOO), and Gly-Tyr ( M , = 238).
Temperature-Responsive Bioconjugates
Transmittance Measurements. Optical transmittance of IgG-PIPAAm conjugate aqueous solutions in PBS (pH 7.4, 1 mg/mL) a t various temperatures was measured a t 500 nm using a spectrophotometer (UV-240, Shimadzu, Japan). The observation cell was thermostated using a circular water jacket. AG-PIPAAm and HSA (used as antigen) mixture solutions were also measured. The procedure was performed as follows: AG-PIPAAm conjugate solution was prepared in PBS (pH 7.4) a t the concentration of 1 mg/mL. FITC-HSA was dissolved in PBS to the concentration of 100 pg/mL. FITC-HSA solution (500 pL) was added to 500 pL of AG-PIPAAm conjugate solution in a 1.5 mL Eppendorf tube and incubated for 1 h a t 4 "C to allow specific complexation. Then optical transmittance of mixture was measured a t various temperatures using a microvolume observation cell (0.8 mL maximum). Complexation of AG-PIPAAm conjugate with FITC-HSA was determined by the method described in the following section. Elutability Measurement of IgG-PIPAAm Conjugates. Measurement of Fc-dependent binding activity of IgG-PIPAAm conjugates was performed by complexation with protein A. Five hundred p L of I g G - P I P M PBS solution (10 pg/mL) was applied to a protein A-immobilized prepacked column (approximately 2.5 mL bed volume, Protein A Sepharose CLdB, Pharmacia LKB Biotechnology, Sweden). One hundred mL of PBS was then passed through the column to elute unbound conjugates. Bound conjugates were then eluted using 100 mL of PBS containing 8 M urea. UV absorbance of the elution a t 280 nm was continuously recorded on a chart as a function of elution volume. All steps were carried out a t 4 "C. Elutability was defined by the following equation: elutability (%) = amount of unbound conjugates (100) total amount of protein
Fluorescence Measurement for Antigen Binding Activity. Fluorescence spectra were recorded on a spectrofluorometer (FP-770, JASCO, Japan). The temperature of the water-jacketed cell holder was controlled with a thermostated circulating bath to 20 "C. An excitation wavelength, 490 nm, and an emission wavelength, 520 nm, were used for the FITC emission measurement. Antigen binding activity was measured as follows: FITC-labeled HSA (FITC-HSA) was used as antigen to AGs. A series of FITC-HSA solutions was prepared with PBS a t the following concentrations: 0, 4.1, 12.0,20.2,50.2,100.0, and 480 pglmL. AG-PIPAAm conjugate was dissolved in PBS a t a concentration of 1 mg/mL. Five hundred pL of FITC-HSA solution was added to 500 pL of AG-PIPAAm conjugate solution in a 1.5 mL Eppendorf tube and incubated for 1 h at 4 "C to allow specific complexation. The mixture was then heated to 37 "C for 8 min to precipitate the polymer. The precipitate was collected by centrifugation a t 4000g for 12 min a t 37 "C. The supernatant was withdrawn, and the precipitate was redissolved in 1mL of cold PBS. The temperature was heated to 37 "Cagain to precipitate the polymer, the precipitate was collected by centrifugation, the supernatant was withdrawn, and finally the precipitate was redissolved in 200 pL of cold PBS. One hundred pL of the solution was diluted into 900 p L of PBS, and the fluorescence was measured in a fluorophotometer. Fluorescence measurements were made on both supernatant and the precipitate for the assay of the conjugates. The complex formation of AG-0, i.e, native anti-human serum albumin goat IgG, with FITC-HSA was deter-
Bioconjugate Chem., Vol. 5, No. 6,1994 579
1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
Relative retention volume (V/Vo) Figure 1. High-performance size exclusion chromatography elution patterns of IgG-PIPAAm conjugates.
mined by high-performance liquid chromatography. AGO was incubated with FITC-HSA solution in a same manner with other AGs, and then the mixture solution was applied in HPLC equipped with TSK gel-G-3000-SW (Toso, Japan). PBS (pH 7.4,0.15 M) was used as mobile phase a t a flow rate of 1.0 mL/min. Fluorescence intensity of FITC-HSA was detected by fluorimeter (FS8010, Toso, Japan) with excitation wavelength a t 490 nm and emission wavelength a t 520 nm. Percentage of immune complex formation was defined by the ratio of two areas corresponding to fluorescence from immune complex-formed HSA and free HSA. All fluorescence measurements were performed a t a concentration low enough so that the fluorescence intensity of the FITC was proportional to its concentration. Antigen binfing activity of IgG-PIPAAm conjugate was expressed by the ratio of fluorescence intensity corresponding to IgG-PIPAAm/HSA complex and that of native IgGMSA complex. RESULTS AND DISCUSSION
Structure of IgG-PIPAAm Conjugates. IgGPIPAAm conjugates were prepared by coupling the activated ester of semitelechelic PIPAAm with amino group of IgG. Preparation and analysis of resulting bioconjugates are summarized in Table 1. The number of conjugated PIPAAm molecules was controlled by changing the molar ratio of IgG and semitelechelic PIPAAm in each preparation. IgG-PIPAAm conjugates are described by two sequential codes, such as AG-12, where AG refers to anti-HSA goat IgG and 12 to the number of PIPAAm molecules grafted per conjugate. The number of grafted PIPAAm molecules was estimated by measurement of primary amino group content using fluorescamine (29, 30). PIPAAm content was also determined by the data from HPSEC, which were in agreement with the results obtained by fluorescaminebased assay. All conjugates were soluble in water and saline solution (physiological pH and ionic strength) a t room temperature. The results of molecular weight measurement by HPSEC are shown in Figure 1. As can be seen in Figure 1, IgG-PIPAAm conjugates were observed to elute earlier through the column than IgG (G-0). Elution time retarded with decreasing the amount of PIPAAm conjugated. Fc-dependent binding of IgG-PIPAAm conjugates was estimated by elutability from the column corresponding to the complexation of the IgG Fc region with columnbound protein A a t 4 "C. The result is shown in Figure
580 Bioconjugafe Chem., Vol. 5, No. 6,1994
80
-* 8
60
P Q *
3
40 20
Takei et al.
I"
:/
0 0
10 20 30 PlPAAm content (wtoh)
40
Figure 2. Elutability of IgG-PIPAAm from protein A-immobilized column as a function of PIPAAm content (0,G series; 0 , AG series). 2. Fc-dependent complexation with protein A decreased significantly with increasing amounts of PIPAAm molecules conjugated with IgG, with the highest conjugate showing less than 10% of the binding which unconjugated IgG demonstrates. When PIPAAm content exceeds 30 wt % in the conjugate, free chain end mobility of the terminally grafted PIPAAm suggested prevention of Fcdependent binding to protein A. Many investigations have described the conjugation of polymers coupled with primary amines corresponding to lysine residue in biomolecules. Alterations in secondary and tertiary structures of biomolecule were often evaluated by circular dichroic spectra. Polymer-protein conjugation in mild conditions leads to no disruption of IgG native structure, and IgG molecules maintain their antigenic properties to some degree through the conjugation (5,6 , 3 1 ) . These results and our own data support our contention that PIPAAm molecules are bound in the outer, exposed surface of globular IgG molecules and little disorganization of IgG structure occurred through conjugation. The Fc region of IgG strongly relates to the complement activation a s well as the binding t o Fc receptor on plasma membrane surface of leukocytes in vivo. The suppression of conjugate binding to protein A observed in Figure 2 suggests that Fc-dependent immunogenicity of conjugates decreases significantly when the conjugates are administered in vivo. Temperature-ResponsivePolymers for Modulating Soluble-Insoluble Changes of the Protein Conjugates. Transmittance changes in PBS solutions of IgG-PIPconjugates are shown in Figure 3. Intact IgG in PBS solution is transparent a t temperatures up t o 50 "C (data not shown in Figure 3). IgG-PIPAAm conjugates exhibit reversible phase transitions; soluble a t lower temperature and insoluble a t higher temperature. Semitelechelic PIPAAm bound to outer surfaces of IgG molecules collapse with increasing solution temperature. It is thought that the biomolecules and PIPAAm molecules in the conjugate form an immiscible structure. Therefore, the segregated conformation of the conjugates provides rapid responses to changes in temperature and a corresponding, complete phase separation due to the highly mobile nature of polymer free-end. IgG-PIPAAm conjugates exhibit LCSTs ranging from 33.6 to 34.4 "C. It has already shown that conjugates constructed by water-soluble biomolecules and temperature-responsive polymers alter the PIPAAm dehydration mechanism and increase the energy required for precipitation (24). Con-
32
33
34
35
36
Temperature ("C)
Figure 3. Temperature dependence of optical transmittance for IgG-PIPAAm aqueous solutions (0,G-3; A, G-13; W, G-21; 0, AG-4; A , AG-12; 0 , AG-20; /z = 500 nm).
32
33
34
35
36
Temperature ("C)
Figure 4. Temperature dependence of optical transmittance for IgG-PIPAAm conjugate/HSA complex in PBS (0, AG-0; 0 , AG-4; A, AG-12; 0 , AG-20).
jugate dehydration increases with an increase in the amount of PIPAAm molecules in the conjugate, and precipitated conjugates strongly aggregate with each other in aqueous media above the LCST. The amount of PIPAAm molecules is likely to produce slight shift in the LCST. Temperature-Modulated Precipitations and Affinity Separation. Antigen binding activity of AGPIPAAm conjugates using FITC-labeled HSA as a model antigen was studied. First, the ability of AG-PIPAAm conjugates to precipitate and to separate antigenbioconjugate complexes from solution was investigated. Optical transmittance changes of PBS solution mixtures of A G - P I P h and HSA as a function of temperature are shown in Figure 4. Specific antigen-antibody complexation is confirmed by the data from HPSEC elution volume. While specific antigen-AG complexation still occurred, the sample AG-0 showed no transmittance changes in all temperatures examined because it lacks molecules of PIPAAm. A G - P I P h conjugates, however, reacted with antigen and showed transmittance changes. Sample AG-4 exhibited, however, only slight transmittance changes compared t o AG-12 and AG-20 and was unable to collapse completely with temperature. As mentioned above, the amount of P I P h molecules in the conjugate dominates the aggregation behavior of the entire conjugates. AG-4 contained an average of 4
Temperature-ResponsiveBioconjugates
Bioconjugate Chem., Vol. 5, No. 6,1994 581
.-0
0
L
E
I-
v)
I-
5
.-c
*
0.8
- 80
-
h
8
.L
c a
v
v)
c a 0.6 c
.L
.-
E
c)
a
I-
a f
0 C
0
a
0 v)
0
0
2
2
2
h
a
2 (0 L c I-
0.4
-
-40
0.2
-
- 20
v)
3
-60
E
:
ii 0.0 0
10
20
30
40
PIPAAm content (wt%) Figure 5. Relative antigen binding activity of IgG-PIPAAm conjugates as a function of PIPAAm content. (HSA concentra7.29 pmol/L). tion: 0 , 1.52 ,umol/L;0,
molecules of PIPAAm per IgG; the amount of PIPAAm molecules was 11.2 wt % of the conjugate as shown in Table 1. PIPAAm content of the resulting AG-4/HSA complex was calculated to be 8.1 wt %. Therefore, weak aggregation force of the AG-4MSA complex produces insufficient precipitation. Complexes of antigen and AGs, AG-12 and AG-20, showed drastic changes in transmittance within a narrow temperature range. This result indicates that the transmittance change of the antigenantibody complex can also be controlled by the amount of PIPAAm molecules conjugated. Specific antigen binding activity of AG-PIPAAm conjugates against HSA as the model antigen was examined as a function of PIPAAm content. The results are shown in Figure 5. AG-PIPAAm conjugates in this experiment formed antigen-antibody complexes with FITC -HSA. At HSA concentration of 1.52 pmollL, and with increasing PIPAAm content, antigen binding activity of AG-PIPAAm conjugates decreases. A significant decrease in antigen biding activity was observed for over 30 wt % of PIPAAm conjugated. A G - P I P h conjugates with PIPAAm contents of about 4-12 molecules per IgG retained about 60% antigen binding activity a t this concentration (recall that Fc-dependent binding activity was abolished a t the same PIPAAm content). At higher HSA concentration (7.29 pmol/L), AG-4 retains almost same antigen binding activity and other AGs also keep relatively high antigen binding. Notable decreases in specific antigen binding activity observed in AG-20 might be due to the fact that excessive attachment of PIPAAm molecules yielded a steric hindrance of the IgG active binding site due to the free chain end mobility of the grafted polymers. Increasing polymer chain length and degree of substitution are likely to enhance the solution temperatureresponse due to the increase in the temperatureresponsive sequences of PIPAAm associated with the conjugates. Although there remains the possibility that long polymer chain mobility and high degree of substitution may denature biomolecule structure or destroy bioactivity due to steric hindrance, there should exist an optimum amount of polymer grafts for each biomolecule. We have collected data for temperature-responsive bioconjugates such a s atelo collagen ( M , = 300 000) (231, bovine serum albumin ( M , = 66 OOO), bovine plasma fibrinogen ( M , = 340 000) (24) and pseudomonas lipase ( M , = 33 000) (25) using PIPAAm with a carboxyl end group as a modifier. These investigations suggest that temperature-responsive behavior of bioconjugates depends on the molecular size of the biomolecules and their
0 032
33 34 35 Temperature ("C)
36
Figure 6. Relationship between antigen binding activity and phase transition of IgG-PIPAAm conjugate, AG-12.
respective hydration correlated to the substitution degree with polymer. Moreover, the mobility of the long polymer chains in aqueous medium may also elicit some effect on the bioactivity of the biomolecule conjugate. Chiu et al. reported substrate-size-dependent enzyme specificity corresponding to the ability of the polymer chains to exclude proteins and ligands from its surroundings using PEGmodified chymotrypsin (32). From these results, a n optimal amount of PIPAAm grafts of approximately 6100 molecular weight is predicted to be 20-30 w t % for temperature-modulated precipitations and affinity bioseparations to reduce undesirable or conflicting biological reactions. Temperature-Responsive PIPAAm as Thermal Switching Sequences. AG-12 was used to compare the antigen binding activity of AG-PIP& conjugates with HSA to the phase transition of AG-PIPAAm a t each temperature. Figure 6 shows the relationship between relative fluorescence changes upon precipitation and optical transmittance changes for mixed aqueous solutions as a function of temperature. Transmittance changes for solution mixtures are observed a t temperatures ranging from 32.8 to 34.0 "C. On the other hand, relative fluorescence changes upon precipitation are observed a t temperatures ranging from 32.8 to 33.4 "C. Once phase transition of bioconjugate occurred in response to changes in solution temperature, specific antigen binding activity decreases drastically to less than 10% of that observed below the critical temperature. IgG-PIPAAm conjugates are soluble in aqueous milieu and readily form specific antigen-bound complexes in homogeneous solution below the critical temperature. These conjugates can be separated from solution and other solutes, depending on PIPAAm content and size, and remain unable to bind antigen above the critical temperature. It is proposed that grafted PIPAAm molecules contribute two features to the IgG bioconjugates: (1)as a thermally induced phase transition inducer and (2) as a switching molecule to control bioactivity and affinity in response to changes in temperature. CONCLUSIONS
IgG-PIPAAm conjugates were prepared by the coupling reaction of the activated ester of semitelechelic PIPAAm with the amino group of IgG. These conjugates exhibit a rapid response to changes in temperature and significant phase separation above the critical solution temperature. Amounts of PIPAAm in the conjugates determine the magnitude of bioconjugate aggregation and
582 Bioconjugate Chem., Vol. 5, No. 6, 1994
precipitation in response to temperature changes. It is conceivable from the results presented in this paper that the conjugation of PIPAAm to IgG leads the reduced Fcdependent immunogenicity in vivo, while conjugates retained specific antigen binding activity. Temperatureresponsive behavior of PIPAAm imparts solubility changes to the conjugate in response to small temperature changes in aqueous media. Semitelechelic PIPAAm is highly efficient for temperature-modulated affinity bioseparations due to the highly intrinsic mobile nature of the polymer free-end. ACKNOWLEDGMENT
The author is grateful to Prof. David W. Grainger, Colorado State University, for his valuable comments and discussion and Hideki Morikawa, M.D., School of Medicine, Keio University, for his cooperation in the research presented in this paper. LITERATURE CITED
(1) Abuchowski, A., van Es, T., Palczuk, N. C., and Davis, F. F. (1977) Alteration of immunological properties of bovine serum albumin by covalent attachment of polyethylene glycol.
J . Biol. Chem. 252, 3578-3581. (2) Abuchowski, A., McCoy, J. R., Palczuk, N. C., van Es, T., and Davis, F. F. (1977) Effect of covalent attachment of polyethylene glycol on immunogenicity and circulating life of bovine liver catalase. J . Biol. Chem. 252, 3582-3586. (3) Harris, J. M. (1992) Introduction to biotechnological and biomedical applications of poly(ethy1ene glycol). Poly(ethy1ene glycol) Chemistry: Biotechnical a n d Biomedical Applications (J. M. Harris, Ed.) pp.1-14, Plenum Press, New York. (4) Suzuki, T., Ikeda, K., and Tomono, T. (1989) Physicochemical and biological properties of poly(ethy1ene glycol)-coupled immunoglobulin G. Part 11. Effect of molecular weight of poly(ethylene glycol). J. Biomater. Sci. Polym. Edn. 1, 71-84. ( 5 ) Pasta, P., Riva, S., and Carrea, G. (1988) Circular dichroism and fluorescence of polyethylene glycol-subtilisin in organic solvents. FEBS Lett. 236, 329-332. (6) Caliceti, P., Schiavon, O., Veronese, F. M., and Chaiken, I. M. (1990) Effects of monomethoxypoly(ethy1ene glycol) modification of ribonuclease on antibody recognition, substrate accessibility and conformational stability. J. Mol. Recognit. 3, 89-93. (7) Heskins, M., Guillet, J. E., and James, E. (1968) Solution properties of poly(N-isopropylacrylamide).J . Macromol. Sci. Chem. A2,1441-1445. (8) Bae, Y. H., Okano, T., Hsu, R., and Kim, S. W., (1987) Thermosensitive polymer as on-off switches for drug release. Makromol. Chem. Rapid Commun. 8, 481-485. (9) Okano, T., Bae, Y. H., Jacobs, H., and Kim, S. W. (1990) Thermally on-off switching polymers for drug permeation and release. J . Controlled Release 11, 255-265. (10) Bae, Y. H., Okano, T., and Kim, S. W. (1990) Temperature dependence of swelling of crosslinked poly(N,N'-alkyl substituted acrylamide) in water. J . Polym. Sci. Polym. Phys. 28, 923-936. (11) Yoshida, R., Sakai, K., Okano, T., Sakurai, Y., Bae, Y. H., and Kim, S. W. (1991) Surface-modulated skin layers of thermal responsive hydrogels as on-off switches; I. Drug release. J . Biomater. Sci. Polym. Edn. 3, 155-162. (12) Yoshida, R., Sakai, K., Okano, T., Sakurai, Y. (1991) Surface-modulated skin layers of thermal responsive hydrogels as on-off switches; 11. Drug permeation. J . Biomater. Sci. Polym. Edn. 3, 243-252. (13) Okuyama, Y., Yoshida, R., Sakai, K., Okano, K., and Sakurai, Y. (1993) Swelling controlled zero order and sigmoidal drug release from thermo-responsive poly(N-isopropyl acrylamide-co-alkyl methacrylate) hydrogel. J. Biomater. Sci. Polym. Edn. 4, 545-556. (14) Yamada, N., Okano, T., Sakai, H., Karikura, F., Sawasaki, Y., and Sakurai, Y. (1990) Thermo-sensitive polymeric surfaces; control of attachment and detachment of cultured cells. Makromol. Chem. Rapid Commun. 11, 571-576.
Takei et al. (15) Okano, T., Yamada, N., Sakai, H., and Sakurai, Y. (1993) A novel recovery system for cultured cells using plasmatreated polystyrene dishes grafted with poly(N-isopropylacrylamide). J . Biomed. Mater. Res. 27, 1243-1251. (16) Kokufuta, E. (1992) Functional immobilized biocatalysts. Progress in Polymer Science 17, pp 647-697, Pergamon Press, New York. (17) Dong, L. C., and Hoffman, A. S. (1986) Thermally reversible hydrogels: 111. Immobilization of enzymes for feedback reaction control. J . Controlled Release 4, 223-227. (18)Chen, J. P., and Hoffman, A. S. (1990) Polymer-protein conjugates 11. M i n i t y precipitation separation of human immunogammaglobulin by a poly(N-isopropylacrylamide). Biomaterials 11, 631-634. (19) Monji, N., and Hoffman, A. S. (1987)A novel immunoassay system and bioseparation process based on thermal phase separating polymers. Appl. Biochem. Biotech. 14, 107-120. (20) Monji, N., Cole, C.-A., Tam, M., Goldstein, L., Nowinski, R. C., and Hoffman, A. S. (1990) Application of a thermallyreversible polymer-antibody conjugate in a novel membranebased immunoassay. Biochem. Biophys. Res. Commun. 172, 652-660. (21) Cole, C.-A., Schreiner, S. M., Priest, J. H., Monji, N., and Hoffman, A. S. (1987) N-isopropyl acrylamide and N-acryl succinimide copolymers: A thermally reversible water soluble activated polymer for protein conjugation. ACS Symposium Series 350, Reversible Polymeric Gels and Related Systems (P. Russo, Ed.) p 245, American Chemical Society, Washington DC. (22) Park, T. G, and Hoffman, A. S. (1993) Synthesis and characterization of a soluble, temperature-sensitive polymerconjugated enzyme. J. Biomater. Sci. Polym. Edn. 4, 493504. (23) Takei, Y. G., Aoki, Y. G., Sanui, K., Ogata, N., Okano, T., and Sakurai, Y. (1993) Temperature-responsive bioconjugates I. Synthesis of temperature-responsive oligomers with reactive end groups and their coupling to biomolecules. Bioconjugate Chem. 4, 42-46. (24) Takei, Y. G., Aoki, Y. G., Sanui, K., Ogata, N., Okano, T., and Sakurai, Y. (1993) Temperature-responsive bioconjugates 11. Molecular design for temperature-modulated bioseparations. Bioconjugate Chem. 4, 341-346. (25) Matsukata, M., Takei, Y. G., Aoki, T., Sanui, K., Ogata, N., Sakurai, Y., and Okano,T. (1994) Temperature modulated solubility-activity alterations for poly(N-isopropylacry1amide)lipase conjugates. J. Biochem.: Biotechnol. 116, 682-686. (26) Miron, T., and Wilchek, M. (1982) A spectrophotometric assay for soluble and immobilized N-hydroxysuccinimide esters. Anal. Biochem. 126, 433-435. (27) Buckman, A. F., and Morr, M. (1981) Functionalization of poly(ethy1ene glycol) and monomethoxy-poly(ethy1ene glycol). Makromol. Chem. 182, 1379-1384. (28) Gornall, A. G., Bardawill, C. S., and David, M. M. (1948) Determination of serum proteins by means of the biuret reaction. J . Biol. Chem. 177, 751-766. (29) Udenfrienel, S., Stein, S., Bohlen, P., Dairman, W., Leimgruber, W., and Weigele, M. (1972) Fluorescamine: A reagent for assay of amino acid, peptides, proteins, and primary amines in the picomole range. Science 178,871-872. (30) Stocks, S. J., Jones, A. J. M., Ramey, C. W., and Brooks, D. E. (1986) A fluorometric assay of the degree of modification of protein primary amines with polyethylene glycol. Anal. Biochem. 154, 232-234. (31) Suzuki, T., Kanbara, N., Tomono, T., Hayashi, N., and Shinohara, I. (1984) Physicochemical and biological properties of poly(ethy1ene glycol)-coupled immunoglobulin G. Biochim. Biophys. Acta 788, 248-255. (32) Chiu, H.-C., Zalipsky, S., Kopeckova, P., and Kopecek, J. (1993) Enzymatic activity of chymotrypsin and its poly(ethylene glycol) conjugates toward low and high molecular weight substrates. Bioconjugate Chem. 4, 290-295.
