- Email: [email protected]
Improved synthesis and iodination of a cleavable photoactivated probe
ANALYTICAL
BIOCHEMISTRY
Improved
16588-95
(1987)
Synthesis and lodination
of a Cleavable
Photoactivated
Probe
HELEN R. MURPHY* AND H. WILLIAM HARRIS, JR.**? *Laboratory of Kidney and Electrolyte Metabolism, National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, Maryland 20982, and TDivision of NephrorogY, The Childrenk Hospital7 Boston, Massachusetts 02115 Received November 3, 1986 J. B. Denny and G. Blobel(l984, Proc. Natl. Acad. Sci. USA 81,5286-5290) have described the synthesis of a water-soluble novel heterobifunctional ‘Z51-labeled photocrosslinking reagent, N-[4-@-azido-m-[‘25I]iodophenylazo)benzoyl]-3-aminopropionyl-N’-oxysulfosuccinimide, which contains a cleavable internal azo bond. We report several modifications of their synthesis which greatly increase the yield of a synthetic intermediate, N-[4-(paminophenylazo)benzoyl]3-aminopropionic acid (compound VII). Evidence is presented that direct iodination of compound VII with chloramine-T produces low yields of ‘Z51-labeled compound VII in a reaction which is difficult to control. Alternatively, the yield of ‘251-labeled reagent can be greatly improved if the iodination is performed following the derivatization of N-[C(p-axidophenylazo)benzoyl]-3-aminopropionyl-N’-oxysulfosu~nimide (compound IX) to a ligand such as dextran. We have demonstrated the transfer of ‘25I label from the reagent when derivatized to aminodextran to proteins in solution after photolysis and cleavage of this r2’1-labeled reagent. 0 1987 Academic F’res, Inc. KEY WORDS: organic
synthesis-radioactive; tein structure; endocytosis.
affinity crosslinking; receptors-general;
Photoactivatable reagents have proven extremely useful in the study of both protein and plasma membrane interactions and topography(l-7). Jaffe et al. (8) synthesized a series of novel heterobifunctional photoaffinity reagents each of which contains a cleavable internal azo bond. Cleavage of the azo bond after photolysis bisects the reagent and if the half containing the aryl azide is iodinated, these reagents transfer their ‘*%label from the species originally derivatized with the reagent to molecules of interest. Use of these reagents, however, has been restricted because of their poor water solubility. Denny and Blobel (3) overcame this problem with the design and synthesis of the iodinated, water-soluble N-[4-b-axido-m-[ 1251]iodophenylazo)benzoyl] - 3 - aminopropionyl -N oxysulfosuccinimide ester, (c) an internal azo bond which upon cleavage transfers (a), and (d) an ‘25I label. We report here several key modifications in their synthesis which greatly increase the yield of both the iodinated and 0003-2697187
$3.00
Copyright Q 1987 by Academic F’ms, Inc. AU rights of reproduction in any form reserved.
pro-
uniodinated forms of this reagent. In contrast to the published synthesis (3), we obtained low overall yields of reagent which are probably due to damage during the chloramine-T iodination step. We have circumvented this problem by derivatization of unlabeled reagent to the desired ligand followed by its iodination with chloramine-T. MATERIALS
AND METHODS
Reagents. All reagents and solvents were of the highest grade obtainable. Aniline, formaldehyde sodium bisulfite addition compound, sodium nitrite, tetrahydrofuran, triethylamine, dioxane, sodium azide, sodium bisulfite, and dimethyl sulfoxide were purchased from Aldrich (Milwaukee, WI). Dicyclohexylcarbodiimide and N-hydroxysuccinimide were from Pierce Chemical (Rockford, IL). Chloroform and methanol were from Baker (Phillipsburg, NJ). Ovalbumin, dextran (Mr 73,400), p-aminobenzoic 88
REVISED
SYNTHESIS
OF PHOTOACTIVATED
HOOC~N=N~NHCH,SO,-Ne’
34
VI
1. NaOH. 2. I-ICI
90°C
I-$N~N=N-@QOH 1. DCC, N-hydroxysuccinimide 2. /?-alanine. pH 9.5
VII
50
6*
89
Reagent synthesis.The modified reaction scheme for the uniodinated reagent (IX) is shown in Fig. 1. In Step 3, the orange sodium salt of compound VI coprecipitated with the desired brick red acid form and reduced its yield by as much as 50%. This coprecipitation was prevented by the following procedure, which includes increases in the volume of the working solutions. Compound V was dissolved in 450 ml of 1 M NaOH, diluted to
acid, and @-alanine were purchased from Sigma Chemical (St. Louis, MO). Na’251 was from DuPont/New England Nuclear (Billerica, MA) and bovine serum albumin was purchased from Reheis (Phoenix, AZ). Sodium dithionite and chloramine-T were from Fisher (Richmond, VA) and dithiothreitol was from Boehringer-Mannheim (Indianapolis, IN). Protein A-Sepharose CL4B was from Pharmacia (Uppsala, Sweden).
V
PROBE
1. H,SO,, 2. NaN,
NaNO*
1
N-hydroxysulfosuccinimide DCC
HG. 1. Synthesis scheme for the crosslinking reagent. Each synthetic step is numbered from 1 to 6 immediately to the left of the arrows in the center of the page. Asterisks indicate steps in which the original synthesis was modified (see text). The roman numerals designate selected compounds in each synthetic step.
90
MURPHY
1 liter with water, heated at 90°C for 80 min, diluted to 3 liters with water at room temperature, and then allowed to cool. In order to precipitate the acid form of VI, 65 ml of 6 M HCl was quickly added to the cooled solution with vigorous stirring, bringing the pH to 2.9. The resulting brick red crystals were collected by centrifugation and washed with several volumes of water. When this material was used directly in Step 4, a phase separation of the tetrahydrofuran/water solution occurred and compound VII did not precipitate. This phase separation was prevented by chromatography of compound VI on Sephadex G- 10 equilibrated in tetrahydrofuran:H20 (2:l). The desalted material was lyophilized and stored at -20°C. The yield based on the amount ofp-aminobenzoic acid used was 50% as compared to the reported 33% (3). The product has an Rf of 0.62 on thin-layer chromatography (TLC) performed on silica gel 60 plates in chloroform:methano1 (2: 1). Step 4 was similar to Ref. (3) but included important revisions in the precipitation of compound VII. Four hundred milligrams of desalted compound VI was dissolved in 25 ml of tetrahydrofuran:H,O (2: 1). To this solution, 1.63 ml of 1 M dicyclohexylcarbodiimide in tetrahydrofuran and 0.19 g of N-hydroxysuccinimide were added and the solution was stirred at 4°C for 2 h. The dicyclohexylurea formed was removed by filtration and 4 ml of 0.56 M /3-alanine was then added. After the pH was adjusted to 9.5 with triethylamine, the solution was stirred for 2 h at room temperature. The pH was then decreased to 4.5 with 6 M HCl, 50 ml of water was added, and the solution was incubated for 2 h at room temperature. The crystals were collected by centrifugation and washed with water. Compound VII (Rf0.45) was separated from residual compound VI by TLC. The average yield of compound VII was 37% using this procedure as compared to the reported 13% (3). The lyophilized material was stored at -70°C. Purified compound VII had a h,,, 380 nm (6 12,000). An NMR
AND
HARRIS
spectrum of the material in dimethyl-& sulfoxide showed 8.74 ppm (NH lH, s), 7.94 (aromatic 2H, d), 7.76 (aromatic 2H, d), 7.70 (aromatic 2H, d), 6.72 (aromatic 2H, d), 6.25 (aromatic NH2 2H, s), 3.45 (aliphatic CH2 2H, q), 2.50 (aliphatic CH2 2H, t). The remaining steps in our synthesis were modified to produce large quantities of uniodinated reagent IX. In Step 5 compound VII was directly converted to compound VIII. Ninety and one-half milligrams of compound VII was dissolved in 0.5 ml of dioxane:H,O (2: 1) and chilled on ice with 0.29mlof3MH2S04.Next,0.181 mlof2M NaN02 in water at 2°C was added, the solution was chilled for an additional 5 mitt, and, under subdued light, 0.18 1 ml of NaN3 at 2°C was added. The resulting solution was incubated on ice for 20 min. After addition of 1.3 ml of dioxane, the material was desalted on a Sephadex G- 10 column in dioxane:HzO (2:l). The product was lyophilized and immediately taken up in 1.2 ml of dry dimethyl sulfoxide (DMSO).’ In Step 6 compound VIII was converted to the sulfosuccinimide ester by adding 79 mg of N-hydroxysulfosuccinimide sodium salt (9) in 0.7 ml of dry DMSO and 90 mg of dicyclohexylcarbodiimide in 0.2 ml of tetrahydrofuran (THF). After overnight incubation, the unwanted dicyclohexylurea was removed from the mixture by centrifugation, the DMSO was evaporated, and the residue was washed with ether. The resultant product was contaminated with unreacted compound VIII and N-hydroxysulfosuccinimide but was not purified to avoid possible hydrolysis of the active ester (see below). These contaminants were innocuous in the subsequent derivatization reactions described below. The material was stored at -70°C in dry DMSO. Compound IX migrated near the solvent front on ’ Abbreviations used: DMSO, dimethyl sulfoxide; THF, tetrahydrofuran; amino-dextran, N-(2-aminoethyl)carbamoylmethyl derivative of a dextran fraction from L. mesenteroides; BSA, bovine Serum albumin; OVA, ovalbumin; TCA, trichloroacetic acid; SDS, sodium dodecyl sulfate.
REVISED
SYNTHESIS
OF
TLC. Twenty percent of the compound VIII used was found to be unreacted as determined by uv absorbance after its isolation on TLC. Compound IX has uv h,,, 360 nm (t 10,000). Dextran studies. The N-(2-aminoethyl)carbamoylmethyl derivative of a dextran fraction from Leuconostoc mesenteroides with a weight average molecular weight of 73,400 (Sigma Chemical Co., St. Louis, MO), abbreviated as amino-dextran, was prepared by the method of Inman ( 10). Dextran concentrations were determined according to Dubois et al. (11) and amino groups were quantitated by the method of Fields ( 12). Amino-dextran was derivatized with compound IX using 5 ~1 of 0.18 M unlabeled reagent IX in DMSO which was added to 20 mg of amino-dextran in 0.5 ml of 0.1 M NaHC03 at pH 8.5. The mixture was incubated at room temperature for 90 min. Uncoupled compound IX was removed by chromatography of the mixture on a Sephadex G-50 column. The void volume of the column containing the dextran derivatized with unlabeled reagent IX (IX-dextran) was lyophilized and taken up in 500 ~1 of 50 mM phosphate buffer, pH 7.5, and stored at -70°C. IX-dextran ( 1.6 mg) was then iodinated by addition of 1 mCi of carrier-free Na’251 and 20 ~1 of chloramine-T (12 mg/ml) in 50 mM phosphate buffer, pH 7.5. After a 20-s incubation, 20 ~1 of NaHS03 (10 mg/ml) in 50 mM phosphate, pH 7.5, was added to terminate the reaction. The reaction mixture was then immediately desalted on a Sephadex G-50 gel filtration column equilibrated in 5 mM phosphate buffer, pH 7.5, to remove unwanted reactants. Transfer of 12’1 label from ‘251-IX-dextran with photolysis and cleavage was performed in the following manner. Tubes. labeled 1 through 5 were used and each received 2.5 ~1 of 1 M NaHC03 and 10 ~1 of 1251-IX-dextran solution of a concentration of 1.5 mg/ml ( 15 pg; 2.89 X lo6 cpm/tube). Tubes 2 through 5
PHOTOACTIVATED
PROBE
91
received 6 ~1 of bovine serum albumin (BSA) and 4 ~1 of ovalbumin (OVA) solutions, each of which was 5 mg/ml. Ten microliters of 0.1 M NaHC03 was added to all the tubes. Tubes 1, 3, and 4 were photolyzed for 5 min. Samples 1, 4, and 5 were cleaved for 45 min by adding 3 ~1 of 1 M sodium dithionite and the remaining tubes received 3 ~1 of 3 M NaCl. All samples were then treated by addition of 1 ml of 30% trichloroacetic acid (TCA) and incubated on ice for 30 min. The precipitates were centrifuged at 7000g for 10 min and the pellets were washed with acetone and 10% TCA, dissolved in 20 ~1 of Laemmli sample buffer (13) which was 80 mM in dithiothreitol, and incubated for 4 min at 100°C. Miscellaneous methods. All work involving photoactivatable azides was done under a Kodak safelight lamp with a GBX-2 filter. Desalting and fractionation were carried out on Sephadex gel filtration columns (Pharmacia). Sodium-dodecyl sulfate (SDS)-gel electrophoresis (13) and autoradiography (14) were performed as previously described. Autoradiograms were scanned by laser densitometer (LKB 2202 ultroscan with a 2220 recording integrator). Ultraviolet spectra were obtained on either a Beckman Model 25 spectrophotometer (Beckman Instruments) or a Perkin-Elmer Lambda 3B spectrophotometer. ‘H NMR spectra were taken on a Nicolet NT 360 nuclear magnetic spectrometer operating at 36 1 MHz. Concentrations of aryl azide reagents were calculated using the absorbance and extinction coefficient at the h,,, for the reagent. Photolysis was carried out at 360 nm using a Model B- 100 A black-ray uv lamp (Ultra-Violet Products, San Gabriel, CA). Irradiation with this lamp was 7000 hW/cm2 at 38 1 mm with a negligible output below 300 nm. Open 1.5ml Eppendorf microfuge tubes containing samples were placed underneath a Pyrex dish over which was mounted the photolysis lamp. The reagent’s azo bond was cleaved by treating with 0.1 or 0.3 M sodium dithionite for 45 min. Protein concentrations were measured using the method of Bradford (15)
92
MURPHY
AND HARRIS
or the BCA protein assay (Pierce Chemical Co., Rockford, IL) and bovine y-globulin as a standard. RESULTS
Modifications of Reagent Synthesis The overall yield of compound VII was increased from 4.3% as reported in Ref. (3) to 18.5% with alterations of Steps 3 and 4 (Fig. 1) as described under Materials and Methods. As the next step in the synthesis, Denny and Blobel (3) chose to iodinate excess compound VII with chloramine-T and subsequently converted the rzS1-labeled product to lz51-labeled compound IX. They reported that 10% of the original lzsI added to the iodination reaction was incorporated into compound VII and only one 1251-labeled spot was observed on TLC. In contrast, we found that direct iodination of an excess of compound VII with chloramine-T consistently incorporated a maximum of 2% of the total lz51 added and produced multiple other iodinated azo compounds (Fig. 2A). This reaction was difficult to control and the yield of 12SI-labeled compound VII decreased dramatically in iodination reactions longer than 30 s. Lower concentrations of chloramine-T produced higher yields of iodinated compound VII (Fig. 2B), but were far from ideal. Furthermore, when the resulting ‘251-labeled compound VII was purified by TLC and converted to the ‘251-labeled final radioactive crosslinker as previously described, we were able to successfully derivatize only 1.1 -+ 0.69% (n = 4) of the iodinated reagent to protein A under conditions identical to those described in Ref. (3). Our evidence suggests that exposure of compound VII to the iodination conditions with chloramine-T may result in its oxidative destruction/alteration and may account for our inability to obtain results similar to those previously described (3). One alternative approach is to iodinate other synthetic intermediates in the hope that they are more stable, but choices are
FIG. 2. Autoradiogram of ‘2sI-labeled products from chloramine-T iodination of compound VII. Excess compound VII was iodinated with chloramine-T according to Ref. (3) and the resulting reaction mixture applied to the origin of thin-layer plates. These were developed by ascending chromatography using a chloroform:methanol(2: 1) mixture. Lane A shows ‘zsI-labeled products resulting from an iodination performed according to Ref. (3). Lane B displays ‘251-labeIed products formed under conditions identical to those of lane A except using one-tenth the amount of chloramine-T. The spot migrating at an R/of 0.48 is 1251-labeled compound VII.
limited. For example, iodination of compound IX is difficult because active esters hydrolyze in the aqueous media (2) present in both the iodination and purification (gel filtration or thin-layer chromatography) steps. As a practical solution to this problem, we chose to iodinate compound IX after its derivatization to amino groups present on a model ligand, amino-dextran.
Synthesisand Iodination of IX-Dextran As prepared (see Materials and Methods section), amino-dextran contained an average of 52 amino groups per dextran molecule. Amino-dextran was then derivatized with unlabeled compound IX to form IXdextran with an average of two derivatized reagent molecules per dextran molecule. Unlabeled compound IX was then iodinated with chloramine-T under conditions similar to those described earlier. Using a 90-fold molar excess of derivatized compound IX over l*%, 12.1 -t_ 1.6% (n = 3) ofthe total 1251
REVISED
SYNTHESIS
OF PHOTOACTIVATED
incorporated into reagent IX and recovery of ‘251-IX-dextran was greater than 94%. The specific activity of the resultants ‘25I-IXdextran averaged 1.92 X lo5 5’0.49 X lo5 cpmhrg (n = 3). Storage of the 1251-IX-dextran at -70°C in the dark was problematic. Rechromatography of the 1251-IX-dextran after storage consistently resulted in the appearance of lowmolecular-weight iodinated material after 48 h which increased as the interval of storage was lengthened. In contrast, we did not observe a similar phenomenon upon storage of unlabeled IX-dextran.
93
PROBE
was
Transfer of 125Ifrom 12’1-IX-Dextran with Photolysis and Cleavage Radioactive label was transferred from the ‘251-labeled reagent linked to dextran to neighboring proteins in solution upon photolysis and cleavage (Fig. 3). ‘251-IX-dextran was added to a solution containing BSA and OVA at final concentrations of 0.9 and 0.6 mg/ml, respectively. Aliquots of this mixture were subjected to various protocols involving no treatment, photolysis, or cleavage, and then the protein was precipitated by the addition of 30% TCA. The precipitates were centrifuged, washed with 10% TCA and then acetone, and then subjected to SDS-gel electrophoresis and autoradiography. Photolysis and cleavage of ‘251-IX-dextran in the absence of protein produced no protein bands as detected by Coomassie blue staining or autoradiography (Fig. 3, lane 1). Addition of protein to ‘251-IX-dextran in the absence of photolysis or cleavage resulted in recovery of unlabeled protein (lane 2). Photolysis (lane 3) in the absence of cleavage produced ‘25I labeling of both BSA and OVA, which was increased after cleavage of the photolyzed mixture (lane 4). Cleavage in the absence of photolysis (lane 5) resulted in minimal labeling. The appearance of minimal ‘25I label in lane 5 of Fig. 3 was probably the result of photolysis from background irradiation. As demonstrated in lane 3, photolysis alone
94K, 67K’ 7:.
“L
43K’
@I
autoradiogram
FIG. 3. Transfer of 12’1 label from ‘Z51-IX-dextran to proteins in solution. After a series of experimental manipulations detailed below, samples were precipitated with TCA, washed with acetone and 10% TCA, and dissolved by addition of SDS. Selected samples were then cleaved by addition of sodium dithionite to a final concentration of 0.1 M and incubated for 45 min. The samples were then dissolved in Laemmli sample buffer containing SDS and dithiothreitol, heated, and electrophoresed on 9% SDS-polyacrylamide gels. The Coomake blue-stained gel is shown on the left. Its corresponding autoradiogram on the right was exposed for a total of 48 h. Lane 1 shows photolysis and cleavage of ‘2SI-IX-dextran alone. 12SI-IX-dextran was combined with BSA, M, 67,000, and OVA, M, 43,000, followed by no treatment (lane 2) or photolysis at 360 nm for 5 min (lane 3). The remaining lanes display the results of dithionite cleavage of the mixture of ‘251-IX-dextran, BSA, and OVA with (lane 4) and without (lane 5) prior photolysis treatment (see text for details).
caused the ‘25I label to appear in both the BSA and OVA gel bands. There was approximately 40% of the labeling of both BSA and OVA as compared to identical conditions where the reagent was cleaved. The apparent transfer of 1251 label from the photolyzed ‘251-IX-dextran as visualized in lane 3 of Fig. 3 did not result from (i) small molecules of dextran-containing reagent since the 1251IX-dextran eluted as a single peak of approximately 80,000 Da; (ii) electrostatic binding of free reagent to derivatized dextran since repeated column chromatography and incubation of 0.6 M sodium phosphate, pH 7.5, did not elute any free reagent; (iii) TCA cleavage of the internal azo bond since incubation of ‘251-IX-dextran with TCA for prolonged periods produced no measurable cleavage; or (iv) originate from the time dependent breakdown of “‘1-IX-dextran since the results from the experiment described in
94
MURPHY
AND HARRIS
Fig. 2 were identical whether photolysis was performed with ‘251-IX-dextran either prepared immediately prior to the experiment or stored for as long as 14 days (data not shown). The foregoing results suggest that under these conditions, L25I label is transferred to proteins from a fraction of photo1yzed ‘251-IX-dextran independent of dithionite addition. Labeling of these proteins via iodine radicals cannot be excluded since there is some evidence for photodissociation of the triplet state of iodoaromatics by excitation at 365 nm (16).
360 nm after complete photolysis results from the derivatized reagent’s azo bond (8,18). However, we cannot exclude the possibility that azo bond cleavage occurred during photolysis. We attempted to estimate the cleavage of the azo bond in derivatized 1251IX after chemical quenching of the aryl azide by addition of 40 mM dithiothreitol (19). Azo bond cleavage was initiated by addition of sodium dithionite to a final concentration of 0.1 M and samples were incubated for 45 min. Under these conditions, approximately 70% of the absorbance at 360 nm disappeared in both the derivatized and underivaQuantitation of Photolysis and Cleavage of tized forms of ‘25I-IX reagent. Increases in Derivatized IX Reagent the final dithionite concentration to 0.3 M or Photolysis of derivatized IX reagent was longer incubation periods did not result in quantitated by following the initial decrease increased cleavage. We conclude that azo in absorbance of 360 nm (Fig. 4). Photolysis bond cleavage was at least 70% complete of the aryl azide moiety reduces the absor- under these conditions. bance of 360 nm light by the derivatized IX DISCUSSION reagent. This is characterized by an initial decrease in absorbance followed by a plateau 1251-labeled heterobifunctional photoaffinas photolysis nears completion (8,17). Irra- ity reagents which are derivatized to a ligand diation for 1 min caused greater than 80% may be used to identify neighboring molephotolysis and photolysis was complete cules after any number of experimental mawithin 2 min. The absorbance remaining at nipulations have been performed. Introduction of a more water-solubie form of this class of reagent (3) should expand its use. We have introduced several modifications in the .6 initial steps of the published synthesis of one 5 .5 t ? such reagent as described by Denny and Blo8 ” be1 (3) which increases the overall yield of a 4- ‘iL,_-----A *--z synthetic intermediate compound VII. Re.34c, gardless of whether compound VII was prez5 9m .2. pared according to Ref. (3) or with these .lmodifications, we observed that direct iodination of compound VII produced low yields 0’. 1 0 12 3 4 5 6 7 8 9101112 of ‘251-labeled compound VII in a reaction Time of Photolysls [min.] which was difficult to control in a reproducFIG. 4. Time course of photolysis for the ‘251-labeled ible manner. Furthermore, storage of the dederivatizcd reagent. One-milliliter samples containing 5.6 X IO-’ M derivatized compound IX in 20 mM so- rivatized ‘2SI-labeled reagent resulted in the appearance of low-molecular-weight ‘25I spedium bicarbonate, pH 8.5, were photolyzed for various intervals with 360 nm light as detailed under Materials cies, suggesting that there is spontaneous and Methods. Immediately thereafter, the absorbance of breakdown of the iodinated reagent. Tothe samples at 360 nm was determined spectrophotogether, these results suggest that once iodinmetrically. Although the sample was irradiated with ated, this reagent may be unstable and, de360~nm light during the measurement, its absorbance pending on conditions, produce widely variwas not reduced significantly during this short time interval. able results.
REVISED SYNTHESIS
OF PHOTOACTIVATED
We found that the stability of this photoaffinity reagent to both chloramine-T iodination and storage was increased by its derivatization as an unlabeled species to amino-dextran. The derivatized reagent was then iodinated and used for photolysis and cleavage studies. This approach has several theoretical advantages in that large amounts of unlabeled reagent may be stored, derivatized to ligands in a controlled manner, and tested for specificity prior to high-specific-activity iodination. However, chloramine-T iodination of the reagent-ligand complex will iodinate tyrosyl residues in protein ligands (20) and may also alter the structure of the ligand itself (21). These potential drawbacks may be directly tested since (i) the specific activity of the intact and cleaved forms of the iodinated reagent-ligand complex can be measured and the extent of azo bond cleavage can be determined spectrophotometrically; and (ii) the specificity and biological activity of the 1251-labeled ligand-reagent complex may be directly compared to an unlabeled ligand complex containing an identical number of derivatized reagents. 1251-IX-dextran itself has multiple applications which are the subject of our current studies. Dextrans have been widely used as electron microscopic and fluorescent markers for fluid-phase endocytosis in a variety of cell types and are available with a variety of different size and charge characteristics (22). Studies analyzing the composition of specific cell compartments occupied by ‘251IX-dextran using a combination of 129 labeling of proteins after photolysis and cleavage in combination with transmission (23) and freeze-fracture (24) electron microscopy may allow further refinement of our understanding of intracellular vesicle traffic. ACKNOWLEDGMENTS The authors gratefully acknowledge the helpful discussions provided by Drs. J. S. Handler and J. V. Stares. Dr. Robert J. Highet performed the NMR spectroscopy. H. William Harris, Jr., was supported by Clinician-Sci-
PROBE
95
entist Award 84-428 from The American Heart Association and with funds contributed in part by AHA-MA Atfi,iate,
REFERENCES 1. Ji, I., and Ji, T. H. (1980) Proc. Natl. Acad. Sci. USA 77,7 167-7 170. 2. Ji, T. H., and Ji, I. (1982) Anal. Biochem. 121, 286-289.
Denny, J. B., and Blobel, G. (1984) Proc. Natl. Acad. Sci. USA 81,5286-5290. 4. Schmitt, M., Painter, R. G., Jesaitis, A. J., Preissner, K., Sklar, L. A., and Cockrane, C. G. (1983) J. Biol. Chem. 258,649-654. 5. Baenziger, J. U., and Fiete, D. (1982) J. Biol. Chem. 3.
257,442
I-4425.
6. Ji, T. H. (1979) Biochim. Biophys. Acta 559,39-69. 7. Peters, K., and Richards, F. M. (1977) Annu. Rev. Biochem. 46,523-55 1. 8. Jaffe, C. L., Lis, H., and Sharon, N. (1980) Biochemistry 19, 4423-4429. 9. Staros, J. V. (1982) Biochemistry 21, 3950-3955. 10. Inman, J. K. (1975) J. Immunol. 114, 704. 11. Dubois, B., Gilles, K. A., Hamilton, J. H., Rebers, P. A., and Smith, F. (1956) Anal. Chem. 28, 350. 12. Fields, R. (1972) in Methods in Enzymology (Hirs, C. H. W., and Timasheff, S. N., Eds.), Vol. 25, p. 464, Academic Press, New York. 13. Laemmli, U. K. (1970) Nature (London) 227, 680-685.
14. Bonner, W. (1984) in Methods in Enzymology (Jakoby, W. B., Ed.), Vol. 104, pp. 460-466, Academic Press, New York. 15. Bradford, M. (1976) Anal. Biochem. 72,248-252. 16. Levy, A., Meyerstein, D., and Ottolengi, M. (1973) J. Phys. Chem. 77,3044-3047. 17. Ji, T. H. (1977) J. Biol. Chem. 252, 1566-1570. 18. Freeman, S. K. (1965) Interpretive Spectroscopy, pp. 23-25, Reinhold, New York. 19. Stares, J. V., Bayley, H., Strandring, D. N., and Knowles, J. R. (1978) Biochem. Biophys. Res. Commun. g&568-572. LU. Greenwood, F. C., Hunter, W. M., and Glover, J. S. (1963) Biochem. J. 89, 114-123. 2 1. Buckle, R. M., and Potts, J. T. (1970) J. Lab. Clin. Med. 76,46-53. 22. Herzog, V., and Farquhar, M. G. (1983) in Methods in Enzymology (Fleischer, S., and Fleischer, B., Eds.), Vol. 98, pp. 203-224, Academic Press, New York. 23. Herzog, V., and Miller, F. (1981) Eur. J. Cell Biol. 24,74-86. 24.
Carpentier, J. L., Brown, D., Iacopetta, B., and Orci, L. (1985) J. Cell Biol. 101, 887-890.
BIOCHEMISTRY
Improved
16588-95
(1987)
Synthesis and lodination
of a Cleavable
Photoactivated
Probe
HELEN R. MURPHY* AND H. WILLIAM HARRIS, JR.**? *Laboratory of Kidney and Electrolyte Metabolism, National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, Maryland 20982, and TDivision of NephrorogY, The Childrenk Hospital7 Boston, Massachusetts 02115 Received November 3, 1986 J. B. Denny and G. Blobel(l984, Proc. Natl. Acad. Sci. USA 81,5286-5290) have described the synthesis of a water-soluble novel heterobifunctional ‘Z51-labeled photocrosslinking reagent, N-[4-@-azido-m-[‘25I]iodophenylazo)benzoyl]-3-aminopropionyl-N’-oxysulfosuccinimide, which contains a cleavable internal azo bond. We report several modifications of their synthesis which greatly increase the yield of a synthetic intermediate, N-[4-(paminophenylazo)benzoyl]3-aminopropionic acid (compound VII). Evidence is presented that direct iodination of compound VII with chloramine-T produces low yields of ‘Z51-labeled compound VII in a reaction which is difficult to control. Alternatively, the yield of ‘251-labeled reagent can be greatly improved if the iodination is performed following the derivatization of N-[C(p-axidophenylazo)benzoyl]-3-aminopropionyl-N’-oxysulfosu~nimide (compound IX) to a ligand such as dextran. We have demonstrated the transfer of ‘25I label from the reagent when derivatized to aminodextran to proteins in solution after photolysis and cleavage of this r2’1-labeled reagent. 0 1987 Academic F’res, Inc. KEY WORDS: organic
synthesis-radioactive; tein structure; endocytosis.
affinity crosslinking; receptors-general;
Photoactivatable reagents have proven extremely useful in the study of both protein and plasma membrane interactions and topography(l-7). Jaffe et al. (8) synthesized a series of novel heterobifunctional photoaffinity reagents each of which contains a cleavable internal azo bond. Cleavage of the azo bond after photolysis bisects the reagent and if the half containing the aryl azide is iodinated, these reagents transfer their ‘*%label from the species originally derivatized with the reagent to molecules of interest. Use of these reagents, however, has been restricted because of their poor water solubility. Denny and Blobel (3) overcame this problem with the design and synthesis of the iodinated, water-soluble N-[4-b-axido-m-[ 1251]iodophenylazo)benzoyl] - 3 - aminopropionyl -N oxysulfosuccinimide ester, (c) an internal azo bond which upon cleavage transfers (a), and (d) an ‘25I label. We report here several key modifications in their synthesis which greatly increase the yield of both the iodinated and 0003-2697187
$3.00
Copyright Q 1987 by Academic F’ms, Inc. AU rights of reproduction in any form reserved.
pro-
uniodinated forms of this reagent. In contrast to the published synthesis (3), we obtained low overall yields of reagent which are probably due to damage during the chloramine-T iodination step. We have circumvented this problem by derivatization of unlabeled reagent to the desired ligand followed by its iodination with chloramine-T. MATERIALS
AND METHODS
Reagents. All reagents and solvents were of the highest grade obtainable. Aniline, formaldehyde sodium bisulfite addition compound, sodium nitrite, tetrahydrofuran, triethylamine, dioxane, sodium azide, sodium bisulfite, and dimethyl sulfoxide were purchased from Aldrich (Milwaukee, WI). Dicyclohexylcarbodiimide and N-hydroxysuccinimide were from Pierce Chemical (Rockford, IL). Chloroform and methanol were from Baker (Phillipsburg, NJ). Ovalbumin, dextran (Mr 73,400), p-aminobenzoic 88
REVISED
SYNTHESIS
OF PHOTOACTIVATED
HOOC~N=N~NHCH,SO,-Ne’
34
VI
1. NaOH. 2. I-ICI
90°C
I-$N~N=N-@QOH 1. DCC, N-hydroxysuccinimide 2. /?-alanine. pH 9.5
VII
50
6*
89
Reagent synthesis.The modified reaction scheme for the uniodinated reagent (IX) is shown in Fig. 1. In Step 3, the orange sodium salt of compound VI coprecipitated with the desired brick red acid form and reduced its yield by as much as 50%. This coprecipitation was prevented by the following procedure, which includes increases in the volume of the working solutions. Compound V was dissolved in 450 ml of 1 M NaOH, diluted to
acid, and @-alanine were purchased from Sigma Chemical (St. Louis, MO). Na’251 was from DuPont/New England Nuclear (Billerica, MA) and bovine serum albumin was purchased from Reheis (Phoenix, AZ). Sodium dithionite and chloramine-T were from Fisher (Richmond, VA) and dithiothreitol was from Boehringer-Mannheim (Indianapolis, IN). Protein A-Sepharose CL4B was from Pharmacia (Uppsala, Sweden).
V
PROBE
1. H,SO,, 2. NaN,
NaNO*
1
N-hydroxysulfosuccinimide DCC
HG. 1. Synthesis scheme for the crosslinking reagent. Each synthetic step is numbered from 1 to 6 immediately to the left of the arrows in the center of the page. Asterisks indicate steps in which the original synthesis was modified (see text). The roman numerals designate selected compounds in each synthetic step.
90
MURPHY
1 liter with water, heated at 90°C for 80 min, diluted to 3 liters with water at room temperature, and then allowed to cool. In order to precipitate the acid form of VI, 65 ml of 6 M HCl was quickly added to the cooled solution with vigorous stirring, bringing the pH to 2.9. The resulting brick red crystals were collected by centrifugation and washed with several volumes of water. When this material was used directly in Step 4, a phase separation of the tetrahydrofuran/water solution occurred and compound VII did not precipitate. This phase separation was prevented by chromatography of compound VI on Sephadex G- 10 equilibrated in tetrahydrofuran:H20 (2:l). The desalted material was lyophilized and stored at -20°C. The yield based on the amount ofp-aminobenzoic acid used was 50% as compared to the reported 33% (3). The product has an Rf of 0.62 on thin-layer chromatography (TLC) performed on silica gel 60 plates in chloroform:methano1 (2: 1). Step 4 was similar to Ref. (3) but included important revisions in the precipitation of compound VII. Four hundred milligrams of desalted compound VI was dissolved in 25 ml of tetrahydrofuran:H,O (2: 1). To this solution, 1.63 ml of 1 M dicyclohexylcarbodiimide in tetrahydrofuran and 0.19 g of N-hydroxysuccinimide were added and the solution was stirred at 4°C for 2 h. The dicyclohexylurea formed was removed by filtration and 4 ml of 0.56 M /3-alanine was then added. After the pH was adjusted to 9.5 with triethylamine, the solution was stirred for 2 h at room temperature. The pH was then decreased to 4.5 with 6 M HCl, 50 ml of water was added, and the solution was incubated for 2 h at room temperature. The crystals were collected by centrifugation and washed with water. Compound VII (Rf0.45) was separated from residual compound VI by TLC. The average yield of compound VII was 37% using this procedure as compared to the reported 13% (3). The lyophilized material was stored at -70°C. Purified compound VII had a h,,, 380 nm (6 12,000). An NMR
AND
HARRIS
spectrum of the material in dimethyl-& sulfoxide showed 8.74 ppm (NH lH, s), 7.94 (aromatic 2H, d), 7.76 (aromatic 2H, d), 7.70 (aromatic 2H, d), 6.72 (aromatic 2H, d), 6.25 (aromatic NH2 2H, s), 3.45 (aliphatic CH2 2H, q), 2.50 (aliphatic CH2 2H, t). The remaining steps in our synthesis were modified to produce large quantities of uniodinated reagent IX. In Step 5 compound VII was directly converted to compound VIII. Ninety and one-half milligrams of compound VII was dissolved in 0.5 ml of dioxane:H,O (2: 1) and chilled on ice with 0.29mlof3MH2S04.Next,0.181 mlof2M NaN02 in water at 2°C was added, the solution was chilled for an additional 5 mitt, and, under subdued light, 0.18 1 ml of NaN3 at 2°C was added. The resulting solution was incubated on ice for 20 min. After addition of 1.3 ml of dioxane, the material was desalted on a Sephadex G- 10 column in dioxane:HzO (2:l). The product was lyophilized and immediately taken up in 1.2 ml of dry dimethyl sulfoxide (DMSO).’ In Step 6 compound VIII was converted to the sulfosuccinimide ester by adding 79 mg of N-hydroxysulfosuccinimide sodium salt (9) in 0.7 ml of dry DMSO and 90 mg of dicyclohexylcarbodiimide in 0.2 ml of tetrahydrofuran (THF). After overnight incubation, the unwanted dicyclohexylurea was removed from the mixture by centrifugation, the DMSO was evaporated, and the residue was washed with ether. The resultant product was contaminated with unreacted compound VIII and N-hydroxysulfosuccinimide but was not purified to avoid possible hydrolysis of the active ester (see below). These contaminants were innocuous in the subsequent derivatization reactions described below. The material was stored at -70°C in dry DMSO. Compound IX migrated near the solvent front on ’ Abbreviations used: DMSO, dimethyl sulfoxide; THF, tetrahydrofuran; amino-dextran, N-(2-aminoethyl)carbamoylmethyl derivative of a dextran fraction from L. mesenteroides; BSA, bovine Serum albumin; OVA, ovalbumin; TCA, trichloroacetic acid; SDS, sodium dodecyl sulfate.
REVISED
SYNTHESIS
OF
TLC. Twenty percent of the compound VIII used was found to be unreacted as determined by uv absorbance after its isolation on TLC. Compound IX has uv h,,, 360 nm (t 10,000). Dextran studies. The N-(2-aminoethyl)carbamoylmethyl derivative of a dextran fraction from Leuconostoc mesenteroides with a weight average molecular weight of 73,400 (Sigma Chemical Co., St. Louis, MO), abbreviated as amino-dextran, was prepared by the method of Inman ( 10). Dextran concentrations were determined according to Dubois et al. (11) and amino groups were quantitated by the method of Fields ( 12). Amino-dextran was derivatized with compound IX using 5 ~1 of 0.18 M unlabeled reagent IX in DMSO which was added to 20 mg of amino-dextran in 0.5 ml of 0.1 M NaHC03 at pH 8.5. The mixture was incubated at room temperature for 90 min. Uncoupled compound IX was removed by chromatography of the mixture on a Sephadex G-50 column. The void volume of the column containing the dextran derivatized with unlabeled reagent IX (IX-dextran) was lyophilized and taken up in 500 ~1 of 50 mM phosphate buffer, pH 7.5, and stored at -70°C. IX-dextran ( 1.6 mg) was then iodinated by addition of 1 mCi of carrier-free Na’251 and 20 ~1 of chloramine-T (12 mg/ml) in 50 mM phosphate buffer, pH 7.5. After a 20-s incubation, 20 ~1 of NaHS03 (10 mg/ml) in 50 mM phosphate, pH 7.5, was added to terminate the reaction. The reaction mixture was then immediately desalted on a Sephadex G-50 gel filtration column equilibrated in 5 mM phosphate buffer, pH 7.5, to remove unwanted reactants. Transfer of 12’1 label from ‘251-IX-dextran with photolysis and cleavage was performed in the following manner. Tubes. labeled 1 through 5 were used and each received 2.5 ~1 of 1 M NaHC03 and 10 ~1 of 1251-IX-dextran solution of a concentration of 1.5 mg/ml ( 15 pg; 2.89 X lo6 cpm/tube). Tubes 2 through 5
PHOTOACTIVATED
PROBE
91
received 6 ~1 of bovine serum albumin (BSA) and 4 ~1 of ovalbumin (OVA) solutions, each of which was 5 mg/ml. Ten microliters of 0.1 M NaHC03 was added to all the tubes. Tubes 1, 3, and 4 were photolyzed for 5 min. Samples 1, 4, and 5 were cleaved for 45 min by adding 3 ~1 of 1 M sodium dithionite and the remaining tubes received 3 ~1 of 3 M NaCl. All samples were then treated by addition of 1 ml of 30% trichloroacetic acid (TCA) and incubated on ice for 30 min. The precipitates were centrifuged at 7000g for 10 min and the pellets were washed with acetone and 10% TCA, dissolved in 20 ~1 of Laemmli sample buffer (13) which was 80 mM in dithiothreitol, and incubated for 4 min at 100°C. Miscellaneous methods. All work involving photoactivatable azides was done under a Kodak safelight lamp with a GBX-2 filter. Desalting and fractionation were carried out on Sephadex gel filtration columns (Pharmacia). Sodium-dodecyl sulfate (SDS)-gel electrophoresis (13) and autoradiography (14) were performed as previously described. Autoradiograms were scanned by laser densitometer (LKB 2202 ultroscan with a 2220 recording integrator). Ultraviolet spectra were obtained on either a Beckman Model 25 spectrophotometer (Beckman Instruments) or a Perkin-Elmer Lambda 3B spectrophotometer. ‘H NMR spectra were taken on a Nicolet NT 360 nuclear magnetic spectrometer operating at 36 1 MHz. Concentrations of aryl azide reagents were calculated using the absorbance and extinction coefficient at the h,,, for the reagent. Photolysis was carried out at 360 nm using a Model B- 100 A black-ray uv lamp (Ultra-Violet Products, San Gabriel, CA). Irradiation with this lamp was 7000 hW/cm2 at 38 1 mm with a negligible output below 300 nm. Open 1.5ml Eppendorf microfuge tubes containing samples were placed underneath a Pyrex dish over which was mounted the photolysis lamp. The reagent’s azo bond was cleaved by treating with 0.1 or 0.3 M sodium dithionite for 45 min. Protein concentrations were measured using the method of Bradford (15)
92
MURPHY
AND HARRIS
or the BCA protein assay (Pierce Chemical Co., Rockford, IL) and bovine y-globulin as a standard. RESULTS
Modifications of Reagent Synthesis The overall yield of compound VII was increased from 4.3% as reported in Ref. (3) to 18.5% with alterations of Steps 3 and 4 (Fig. 1) as described under Materials and Methods. As the next step in the synthesis, Denny and Blobel (3) chose to iodinate excess compound VII with chloramine-T and subsequently converted the rzS1-labeled product to lz51-labeled compound IX. They reported that 10% of the original lzsI added to the iodination reaction was incorporated into compound VII and only one 1251-labeled spot was observed on TLC. In contrast, we found that direct iodination of an excess of compound VII with chloramine-T consistently incorporated a maximum of 2% of the total lz51 added and produced multiple other iodinated azo compounds (Fig. 2A). This reaction was difficult to control and the yield of 12SI-labeled compound VII decreased dramatically in iodination reactions longer than 30 s. Lower concentrations of chloramine-T produced higher yields of iodinated compound VII (Fig. 2B), but were far from ideal. Furthermore, when the resulting ‘251-labeled compound VII was purified by TLC and converted to the ‘251-labeled final radioactive crosslinker as previously described, we were able to successfully derivatize only 1.1 -+ 0.69% (n = 4) of the iodinated reagent to protein A under conditions identical to those described in Ref. (3). Our evidence suggests that exposure of compound VII to the iodination conditions with chloramine-T may result in its oxidative destruction/alteration and may account for our inability to obtain results similar to those previously described (3). One alternative approach is to iodinate other synthetic intermediates in the hope that they are more stable, but choices are
FIG. 2. Autoradiogram of ‘2sI-labeled products from chloramine-T iodination of compound VII. Excess compound VII was iodinated with chloramine-T according to Ref. (3) and the resulting reaction mixture applied to the origin of thin-layer plates. These were developed by ascending chromatography using a chloroform:methanol(2: 1) mixture. Lane A shows ‘zsI-labeled products resulting from an iodination performed according to Ref. (3). Lane B displays ‘251-labeIed products formed under conditions identical to those of lane A except using one-tenth the amount of chloramine-T. The spot migrating at an R/of 0.48 is 1251-labeled compound VII.
limited. For example, iodination of compound IX is difficult because active esters hydrolyze in the aqueous media (2) present in both the iodination and purification (gel filtration or thin-layer chromatography) steps. As a practical solution to this problem, we chose to iodinate compound IX after its derivatization to amino groups present on a model ligand, amino-dextran.
Synthesisand Iodination of IX-Dextran As prepared (see Materials and Methods section), amino-dextran contained an average of 52 amino groups per dextran molecule. Amino-dextran was then derivatized with unlabeled compound IX to form IXdextran with an average of two derivatized reagent molecules per dextran molecule. Unlabeled compound IX was then iodinated with chloramine-T under conditions similar to those described earlier. Using a 90-fold molar excess of derivatized compound IX over l*%, 12.1 -t_ 1.6% (n = 3) ofthe total 1251
REVISED
SYNTHESIS
OF PHOTOACTIVATED
incorporated into reagent IX and recovery of ‘251-IX-dextran was greater than 94%. The specific activity of the resultants ‘25I-IXdextran averaged 1.92 X lo5 5’0.49 X lo5 cpmhrg (n = 3). Storage of the 1251-IX-dextran at -70°C in the dark was problematic. Rechromatography of the 1251-IX-dextran after storage consistently resulted in the appearance of lowmolecular-weight iodinated material after 48 h which increased as the interval of storage was lengthened. In contrast, we did not observe a similar phenomenon upon storage of unlabeled IX-dextran.
93
PROBE
was
Transfer of 125Ifrom 12’1-IX-Dextran with Photolysis and Cleavage Radioactive label was transferred from the ‘251-labeled reagent linked to dextran to neighboring proteins in solution upon photolysis and cleavage (Fig. 3). ‘251-IX-dextran was added to a solution containing BSA and OVA at final concentrations of 0.9 and 0.6 mg/ml, respectively. Aliquots of this mixture were subjected to various protocols involving no treatment, photolysis, or cleavage, and then the protein was precipitated by the addition of 30% TCA. The precipitates were centrifuged, washed with 10% TCA and then acetone, and then subjected to SDS-gel electrophoresis and autoradiography. Photolysis and cleavage of ‘251-IX-dextran in the absence of protein produced no protein bands as detected by Coomassie blue staining or autoradiography (Fig. 3, lane 1). Addition of protein to ‘251-IX-dextran in the absence of photolysis or cleavage resulted in recovery of unlabeled protein (lane 2). Photolysis (lane 3) in the absence of cleavage produced ‘25I labeling of both BSA and OVA, which was increased after cleavage of the photolyzed mixture (lane 4). Cleavage in the absence of photolysis (lane 5) resulted in minimal labeling. The appearance of minimal ‘25I label in lane 5 of Fig. 3 was probably the result of photolysis from background irradiation. As demonstrated in lane 3, photolysis alone
94K, 67K’ 7:.
“L
43K’
@I
autoradiogram
FIG. 3. Transfer of 12’1 label from ‘Z51-IX-dextran to proteins in solution. After a series of experimental manipulations detailed below, samples were precipitated with TCA, washed with acetone and 10% TCA, and dissolved by addition of SDS. Selected samples were then cleaved by addition of sodium dithionite to a final concentration of 0.1 M and incubated for 45 min. The samples were then dissolved in Laemmli sample buffer containing SDS and dithiothreitol, heated, and electrophoresed on 9% SDS-polyacrylamide gels. The Coomake blue-stained gel is shown on the left. Its corresponding autoradiogram on the right was exposed for a total of 48 h. Lane 1 shows photolysis and cleavage of ‘2SI-IX-dextran alone. 12SI-IX-dextran was combined with BSA, M, 67,000, and OVA, M, 43,000, followed by no treatment (lane 2) or photolysis at 360 nm for 5 min (lane 3). The remaining lanes display the results of dithionite cleavage of the mixture of ‘251-IX-dextran, BSA, and OVA with (lane 4) and without (lane 5) prior photolysis treatment (see text for details).
caused the ‘25I label to appear in both the BSA and OVA gel bands. There was approximately 40% of the labeling of both BSA and OVA as compared to identical conditions where the reagent was cleaved. The apparent transfer of 1251 label from the photolyzed ‘251-IX-dextran as visualized in lane 3 of Fig. 3 did not result from (i) small molecules of dextran-containing reagent since the 1251IX-dextran eluted as a single peak of approximately 80,000 Da; (ii) electrostatic binding of free reagent to derivatized dextran since repeated column chromatography and incubation of 0.6 M sodium phosphate, pH 7.5, did not elute any free reagent; (iii) TCA cleavage of the internal azo bond since incubation of ‘251-IX-dextran with TCA for prolonged periods produced no measurable cleavage; or (iv) originate from the time dependent breakdown of “‘1-IX-dextran since the results from the experiment described in
94
MURPHY
AND HARRIS
Fig. 2 were identical whether photolysis was performed with ‘251-IX-dextran either prepared immediately prior to the experiment or stored for as long as 14 days (data not shown). The foregoing results suggest that under these conditions, L25I label is transferred to proteins from a fraction of photo1yzed ‘251-IX-dextran independent of dithionite addition. Labeling of these proteins via iodine radicals cannot be excluded since there is some evidence for photodissociation of the triplet state of iodoaromatics by excitation at 365 nm (16).
360 nm after complete photolysis results from the derivatized reagent’s azo bond (8,18). However, we cannot exclude the possibility that azo bond cleavage occurred during photolysis. We attempted to estimate the cleavage of the azo bond in derivatized 1251IX after chemical quenching of the aryl azide by addition of 40 mM dithiothreitol (19). Azo bond cleavage was initiated by addition of sodium dithionite to a final concentration of 0.1 M and samples were incubated for 45 min. Under these conditions, approximately 70% of the absorbance at 360 nm disappeared in both the derivatized and underivaQuantitation of Photolysis and Cleavage of tized forms of ‘25I-IX reagent. Increases in Derivatized IX Reagent the final dithionite concentration to 0.3 M or Photolysis of derivatized IX reagent was longer incubation periods did not result in quantitated by following the initial decrease increased cleavage. We conclude that azo in absorbance of 360 nm (Fig. 4). Photolysis bond cleavage was at least 70% complete of the aryl azide moiety reduces the absor- under these conditions. bance of 360 nm light by the derivatized IX DISCUSSION reagent. This is characterized by an initial decrease in absorbance followed by a plateau 1251-labeled heterobifunctional photoaffinas photolysis nears completion (8,17). Irra- ity reagents which are derivatized to a ligand diation for 1 min caused greater than 80% may be used to identify neighboring molephotolysis and photolysis was complete cules after any number of experimental mawithin 2 min. The absorbance remaining at nipulations have been performed. Introduction of a more water-solubie form of this class of reagent (3) should expand its use. We have introduced several modifications in the .6 initial steps of the published synthesis of one 5 .5 t ? such reagent as described by Denny and Blo8 ” be1 (3) which increases the overall yield of a 4- ‘iL,_-----A *--z synthetic intermediate compound VII. Re.34c, gardless of whether compound VII was prez5 9m .2. pared according to Ref. (3) or with these .lmodifications, we observed that direct iodination of compound VII produced low yields 0’. 1 0 12 3 4 5 6 7 8 9101112 of ‘251-labeled compound VII in a reaction Time of Photolysls [min.] which was difficult to control in a reproducFIG. 4. Time course of photolysis for the ‘251-labeled ible manner. Furthermore, storage of the dederivatizcd reagent. One-milliliter samples containing 5.6 X IO-’ M derivatized compound IX in 20 mM so- rivatized ‘2SI-labeled reagent resulted in the appearance of low-molecular-weight ‘25I spedium bicarbonate, pH 8.5, were photolyzed for various intervals with 360 nm light as detailed under Materials cies, suggesting that there is spontaneous and Methods. Immediately thereafter, the absorbance of breakdown of the iodinated reagent. Tothe samples at 360 nm was determined spectrophotogether, these results suggest that once iodinmetrically. Although the sample was irradiated with ated, this reagent may be unstable and, de360~nm light during the measurement, its absorbance pending on conditions, produce widely variwas not reduced significantly during this short time interval. able results.
REVISED SYNTHESIS
OF PHOTOACTIVATED
We found that the stability of this photoaffinity reagent to both chloramine-T iodination and storage was increased by its derivatization as an unlabeled species to amino-dextran. The derivatized reagent was then iodinated and used for photolysis and cleavage studies. This approach has several theoretical advantages in that large amounts of unlabeled reagent may be stored, derivatized to ligands in a controlled manner, and tested for specificity prior to high-specific-activity iodination. However, chloramine-T iodination of the reagent-ligand complex will iodinate tyrosyl residues in protein ligands (20) and may also alter the structure of the ligand itself (21). These potential drawbacks may be directly tested since (i) the specific activity of the intact and cleaved forms of the iodinated reagent-ligand complex can be measured and the extent of azo bond cleavage can be determined spectrophotometrically; and (ii) the specificity and biological activity of the 1251-labeled ligand-reagent complex may be directly compared to an unlabeled ligand complex containing an identical number of derivatized reagents. 1251-IX-dextran itself has multiple applications which are the subject of our current studies. Dextrans have been widely used as electron microscopic and fluorescent markers for fluid-phase endocytosis in a variety of cell types and are available with a variety of different size and charge characteristics (22). Studies analyzing the composition of specific cell compartments occupied by ‘251IX-dextran using a combination of 129 labeling of proteins after photolysis and cleavage in combination with transmission (23) and freeze-fracture (24) electron microscopy may allow further refinement of our understanding of intracellular vesicle traffic. ACKNOWLEDGMENTS The authors gratefully acknowledge the helpful discussions provided by Drs. J. S. Handler and J. V. Stares. Dr. Robert J. Highet performed the NMR spectroscopy. H. William Harris, Jr., was supported by Clinician-Sci-
PROBE
95
entist Award 84-428 from The American Heart Association and with funds contributed in part by AHA-MA Atfi,iate,
REFERENCES 1. Ji, I., and Ji, T. H. (1980) Proc. Natl. Acad. Sci. USA 77,7 167-7 170. 2. Ji, T. H., and Ji, I. (1982) Anal. Biochem. 121, 286-289.
Denny, J. B., and Blobel, G. (1984) Proc. Natl. Acad. Sci. USA 81,5286-5290. 4. Schmitt, M., Painter, R. G., Jesaitis, A. J., Preissner, K., Sklar, L. A., and Cockrane, C. G. (1983) J. Biol. Chem. 258,649-654. 5. Baenziger, J. U., and Fiete, D. (1982) J. Biol. Chem. 3.
257,442
I-4425.
6. Ji, T. H. (1979) Biochim. Biophys. Acta 559,39-69. 7. Peters, K., and Richards, F. M. (1977) Annu. Rev. Biochem. 46,523-55 1. 8. Jaffe, C. L., Lis, H., and Sharon, N. (1980) Biochemistry 19, 4423-4429. 9. Staros, J. V. (1982) Biochemistry 21, 3950-3955. 10. Inman, J. K. (1975) J. Immunol. 114, 704. 11. Dubois, B., Gilles, K. A., Hamilton, J. H., Rebers, P. A., and Smith, F. (1956) Anal. Chem. 28, 350. 12. Fields, R. (1972) in Methods in Enzymology (Hirs, C. H. W., and Timasheff, S. N., Eds.), Vol. 25, p. 464, Academic Press, New York. 13. Laemmli, U. K. (1970) Nature (London) 227, 680-685.
14. Bonner, W. (1984) in Methods in Enzymology (Jakoby, W. B., Ed.), Vol. 104, pp. 460-466, Academic Press, New York. 15. Bradford, M. (1976) Anal. Biochem. 72,248-252. 16. Levy, A., Meyerstein, D., and Ottolengi, M. (1973) J. Phys. Chem. 77,3044-3047. 17. Ji, T. H. (1977) J. Biol. Chem. 252, 1566-1570. 18. Freeman, S. K. (1965) Interpretive Spectroscopy, pp. 23-25, Reinhold, New York. 19. Stares, J. V., Bayley, H., Strandring, D. N., and Knowles, J. R. (1978) Biochem. Biophys. Res. Commun. g&568-572. LU. Greenwood, F. C., Hunter, W. M., and Glover, J. S. (1963) Biochem. J. 89, 114-123. 2 1. Buckle, R. M., and Potts, J. T. (1970) J. Lab. Clin. Med. 76,46-53. 22. Herzog, V., and Farquhar, M. G. (1983) in Methods in Enzymology (Fleischer, S., and Fleischer, B., Eds.), Vol. 98, pp. 203-224, Academic Press, New York. 23. Herzog, V., and Miller, F. (1981) Eur. J. Cell Biol. 24,74-86. 24.
Carpentier, J. L., Brown, D., Iacopetta, B., and Orci, L. (1985) J. Cell Biol. 101, 887-890.