Inulin-Inulin- 125125 I-Tyramine, an Improved Residualizing Label for Studies I-Tyramine, an Improved Residualizing Label for Studies on Sites of Catabolism of Circulating Proteins on Sites of Catabolism of Circulating Proteins

Residualizing labels for protein, such as dilactitol- '261-tyramine (lZ6I-DLT) and cellobiitol-'261-tyramine, have been used to identify the tissue and cellular sites of catabolism of long-lived plasma proteins, such as albumin, immunoglobulins, and lipoproteins. The radioactive degradation products formed from labeled proteins are relatively large, hydrophilic, resistant to lysosomal hydrolases, and accumulate in lysosomes in the cells involved in degradation of the carrier protein. However, the gradual loss of the catabolites from cells (tn - 2 days) has limited the usefulness of residualizing labels in studies on longer lived proteins. We describe here a higher molecular weight (Mr - 5000), more efficient residualizing glycoconjugate label, inulin- '261-tyramine (lZ6I-InTn). Attachment of "'1-InTn had no effect on the plasma half-life or tissue sites of catabolism of asialofetuin, fetuin, or rat serum albumin in the rat. The half-life for hepatic retention of degrada- tion products from '261-InTn-labeled asialofetuin was 5 days, compared to 2.3 days for '261-DLT-labeled asialofetuin. The whole body half-lives for radioactivity from '261-InTn-, "'1-DLT-,


In~lin-~~~I-Tyrarnine, an
Improved Residualizing Label for Studies on Sites of Catabolism of Circulating Proteins* (Received for publication, February 4, 1988) Janet L. Maxwell$, John W. BaynesSQ, and Suzanne R. ThorpeSlI Residualizing labels for protein, such as dilactitol-'261-tyramine (lZ6I-DLT) and cellobiitol-'261-tyramine, have been used to identify the tissue and cellular sites of catabolism of long-lived plasma proteins, such as albumin, immunoglobulins, and lipoproteins. The radioactive degradation products formed from labeled proteins are relatively large, hydrophilic, resistant to lysosomal hydrolases, and accumulate in lysosomes in the cells involved in degradation of the carrier protein.
However, the gradual loss of the catabolites from cells ( t n -2 days) has limited the usefulness of residualizing labels in studies on longer lived proteins. We describe here a higher molecular weight (Mr -5000), more efficient residualizing glycoconjugate label, inulin-'261-tyramine (lZ6I-InTn). Attachment of "'1-InTn had no effect on the plasma half-life or tissue sites of catabolism of asialofetuin, fetuin, or rat serum albumin in the rat. The half-life for hepatic retention of degradation products from '261-InTn-labeled asialofetuin was 5 days, compared to 2.3 days for '261-DLT-labeled asialofetuin. The whole body half-lives for radioactivity from '261-InTn-, "'1-DLT-, and '261-labeled rat serum albumin were 7.5, 4.3, and 2.2 days, respectively. The tissue distribution of degradation products from '261-InTn-labeled proteins agreed with results of previous studies using lZ6I-DLT, except that a greater fraction of total degradation products was recovered in tissues. Kinetic analyses indicated that the average half-life for retention of lZ6I-InTn degradation products in tissues is -5 days and suggested that in vivo there are both slow and rapid routes for release of degradation products from cells. Overall, these experiments indicate that '261-InTn should provide greater sensitivity and more accurate quantitative information on the sites of catabolism of long-lived circulating proteins in vivo.
Residualking labels are biologically inert radioactive tags used for studies on the sites of protein catabolism in uiuo. These labels are designed to yield limit, hydrophilic degradation products of a sufficient size that they are retained in lysosomes following catabolism of the carrier protein. The sites of degradation of the labeled. protein may then be determined either by measuring acid-soluble radioactivity in various tissues and cells or by autoradiography. Residualizing labels, such as [3H]raffin~~e (2), dilactitol-"51-tyramine (lZ5I-* This work was supported by National Institutes of Health Research Grant DK 25373. A preliminary report of this work has been presented (1). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
T To whom correspondence should be addressed. DLT)' (3), and cellobiitol-"51-tyramine (4), have been used to identify the tissue and cellular sites of catabolism of plasma proteins, such as albumin (5-7), lipoproteins (4,8), and immunoglobulins (9), and are also being increasingly used in studies on the uptake and catabolism of proteins by cells in culture (8,10). One of the limitations of the use of these labels, however, is that whereas their rate of loss from cells is slow ( t~ -2 days for '251-DLT in rat tissues), the rates of catabolism of plasma proteins are often equally slow or slower.
Thus, a substantial fraction of degradation products is lost from tissues by the time significant amounts of the protein have been catabolized. Under these circumstances, it is not possible to assess rigorously the quantitative role of various tissues in catabolism of a protein since the distribution of degradation products in the body could be biased by differences in the rate of loss of the label from various cell types. Because of the limited residualization of tetrasaccharide derivatives of tyramine and the fact that residualization is improved with increasing saccharide content or molecular weight of the label (3), we set out to design a higher molecular weight oligosaccharide derivative of tyramine, with the expectation that this label would be retained more efficiently in cells. We describe here the biological properties of in~lin-"~Ityramine ( lZ5I-InTn), a residualizing glycoconjugate label derived from the inert fructan polymer inulin ( M , -5000). The retention of protein degradation products containing the 1251-InTn label results from both their size and the absence of lysosomal fructofuranosidase activity (11). As shown below, lZ5I-InTn, the largest residualizing label for protein described thus far, has negligible effects on the kinetics or tissue sites of plasma protein catabolism and is retained in lysosomes more efficiently than is lz5I-DLT. The data indicate that lz5I-InTn should be widely applicable in studies on the catabolism of long-lived circulating proteins.

EXPERIMENTAL PROCEDURES AND RESULTS'
The chemistry of synthesis and coupling of lZ5I-InTn to protein is outlined in Fig. 1 and described in detail under "Experimental Procedures." Also described in the Miniprint are a series of preliminary experiments validating the usefulness of InTn in studies on catabolism of circulating proteins such as asialofetuin and fetuin. The effectiveness of InTn as a residualizing label is clearly illustrated in Fig. 5, which shows both the plasma and whole body kinetics of clearance of Iz5I-, lz5I-DLT-, and "51-InTn-labeled RSA. Notably, as The abbreviations used are: lZ5I-DLT, dila~titol-~~~I-tyramine; lz5I-InTN, in~lin-'~~I-tyrarnine; RSA, rat serum albumin; *I, lZ5I. * Portions of this paper (including "Experimental Procedures," part of "Results," Figs, 2-4 and 6, and Tables 1-111) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are included in the microfilm edition of the Journal that is available from Waverly Press.

FIG. 1.
Reaction scheme for synthesis of lZSI-InTn and its coupling to protein.
Inulin is reduced with NaBH4, and the terminal alditol is oxidized with a limiting amount of periodate to generate an aldehyde. Tyramine is coupled to inulin aldehyde by reductive amination using NaBH3CN, yielding InTn. InTn is then labeled with radioactive iodine using IODO-GEN and coupled to protein using cyanuric chloride.

INULIFI-O-CHz
I with asialofetuin and fetuin, the plasma half-life of RSA was unaffected by the attachment of InTn. The lZ5I-InTn degradation products also residualized more efficiently, with a whole body half-life of about 7.5 days, compared to 4-5 days for lZ5I-DLT ( Fig. 5 and Ref. 7). The tissue distribution of radioactivity at 4 days after injection of '251-InTn-labeled RSA, shown in Table 11, confirms previous evidence using lZ5I-DLT-labeled RSA that catabolism of RSA takes place primarily in muscle and skin (7). However, in these experiments, a significantly greater fraction of total degradation products was retained in the body. Thus, the circulating halflife of this preparation of '251-InTn-labeled RSA was 1.8 days, i.e. 79% catabolism at 4 days; the loss of 27% of radioactivity from the body by 4 days indicates that about 66% (2779) of the theoretical yield of '251-InTn-labeled RSA degradation products was recovered, compared to 45% for '"I-DLT-labeled RSA (7). Overall, with all three proteins studied, the lZ5I-InTn label yielded results consistent with previous studies using other labels, but with substantially improved retention of degradation products in the body.
Kinetic Modeling of Plasma Protein Catabolism-As a first step toward understanding the biological behavior of residualizing labels, we have attempted to develop kinetic models for quantitative comparison of the rates of loss of the various labels from tissues. For this purpose, we have used SCoP and SCoPfit programs, which are simulation control and optimization programs developed for the IBM-PC/AT computer by the National Biomedical Simulation Resource at Duke University (Durham, NC). SCoP generates a graphical simulation of a kinetic process, given a series of differential equations (the kinetic model) and specified kinetic constants. The SCoPfit program accepts actual experimental data and, using the SCoP model, develops kinetic constants to optimize the  Fig. 7, showing the kinetics of whole body release of degradation products from lZ5I-DLTlabeled RSA (Fig. 7, upper) and '251-InTn-labeled RSA (Fig.  7, lower). The various lines drawn on the graph represent different fits to the data using the SCoP or SCoPfit program with various models and assumptions, as described in detail in the Miniprint. In summary, the dotted lines are derived from the SCoP program using the three-compartment model described in Fig. 6 and assuming that RSA degradation products leak from all tissues at the same rate at which asialofetuin degradation products leak from liver. Because of the poor fit to the experimental data, SCoPfit was used with the same model to optimize the kinetic constants and to improve the fit to the data. However, this SCoPfit optimization (dashed lines) was also unsatisfactory, and systematic error was apparent, suggesting that the model was inappropriate. Since recent work by Buktenica et al. (30) indicated that degradation products were routed through both slow and fast release compartments in cells in vitro, the three-compartment model was revised to allow for a fraction of degradation products to be released rapidly from cells in vivo. The solid lines in Fig. 7  (upper and lower) are the results of SCoPfit optimizations to this revised model and yield good and consistent fits to the experimental data. The development, mathematical description and assumptions, and the kinetic constants obtained with the various models are described in detail in the Miniprint.

DISCUSSION
The need for residualizing labels which are more completely retained in tissues has been apparent since the earliest experiments using this technology to identify the sites of plasma protein catabolism. Because of the gradual loss of degradation products from tissues, it has been necessary to terminate experiments at times when only a fraction of the protein has been catabolized and then to apply corrections for intact protein remaining in tissues, for example, by acid precipitation of the intact protein (3, 7) or by injection of a second, nonresidualizing tracer to estimate the amount of intact protein in the tissue (31). These manipulations are not only inconvenient, but they also ultimately affect the precision of estimates of protein degradation in tissues. Our previous work had revealed a relationship between the number of carbohydrate units in the label and its efficiency of residualization (3), so that the synthesis of a larger glycoconjugate label seemed a reasonable route for improving residualization. There are obvious limits to this approach, however, since at some point the size or properties of the label itself will affect the mechanisms and sites of catabolism of the carrier protein.
An inulin derivative was chosen as a reasonably sized target since the resulting molecular weight of the label would be, at most, about 10% of the mass of the smallest plasma protein.
The synthesis of InTn was straightforward, and its iodination and coupling to protein proceeded with good efficiency, 30 and 70%, respectively. Thus, only nanomolar quantities are required to label proteins with high specific radioactivity. The inertness of underivatized inulin in the coupling reaction is also convenient because this inulin serves as a carrier to  Table 111. decrease losses during handling, does not appear to interfere with labeling of the protein, and is readily separated from labeled protein by gel exclusion chromatography. In all of the experiments described here, the average substitution of protein was limited to <1 mol of '261-InTn/mol of protein in order to decrease the probability of multiple derivatization of carrier proteins. The addition of 1 mol of 1261-InTn/mol of protein had no detectable effect on the kinetics, mechanisms, or sites of catabolism of asialofetuin, fetuin, or RSA. This result is consistent with recent hypotheses on the regulation of protein catabolism. Thus, the kinetics of protein catabolism appear to be determined by genetically encoded molecular features of the protein molecule, such as the amino-terminal amino acid or a sequence or array of amino acids in the primary or tertiary structure of the protein, rather than by bulk physical characteristics such as hydrophobicity, subunit molecular weight, or isoelectric point (32).
The size of the radioactive products isolated from urine using the [3H]raffinose (2), '261-DLT (3), or '261-InTN labels indicates that residualizing labels are excreted from the body largely in intact form. Thus, following catabolism of the carrier protein, the labeled degradation products appear to be released from cells by the process of exocytosis or regurgitation, rather than by deiodination or eventual hydrolysis to lower molecular weight products. The difference in whole body half-life of radioactivity from asialofetuin labeled with raffinose, DLT, InTn, and other labels (3) indicates that the structure of the label affects its rate of release from cells. In addition, however, kinetic analysis indicates that there are also differences in the routes of transport of these indigestible compounds within the cell. Thus, whereas some residualization was observed with all of the labels, there was a fraction of these labels rapidly released from the body so that a significant lag phase in whole body clearance was not observed (Figs. 4 and 5, lower). Based on kinetic analysis, we have concluded that degradation products may be partitioned between slow and fast release compartments within the cell and that routing of the partially degraded protein or labeled degradation products to the fast release compartment may be an important factor limiting the long-term retention of catabolites in the body. Our model makes no statement regarding the nature of the fast release compartment, although it is likely to be an early endocytic compartment, either prelyso-soma1 or in equilibrium with the lysosomal compartment. Whereas larger residualizing labels could theoretically prove more efficient, there is greater risk that they will affect the catabolism of the carrier protein. For most purposes, the InTn label should be suitable, for example, in studies on the catabolism of IgGs, which are among the longest lived circulating proteins (tH = 3-5 days in the rat). Sone enrichment for reducing inulin was obtained by fractionating inulin on Sephadex G-50 and pooling the fractions with highest ratio Of reducing sugar, measured by the bicinchoninic acid array 115). to total sugar, measured by the anthrone assay (16) (Fig. 211). using fructose as standard in both assays.
Fractions containing InTn (Fig. 28, fractions The desalting step also removed any residual free tyramine and yieldea The procedure for the coupling Of tyramine derivatives to protein using Cycl is similar to that described previously (

Sprague-Dawley rats, 130-220 9. Methods for animal care and maintenance,
In vivo Exosiments. Experiments were carried out in male and female iniection Of Droteins. and measurement Of ~1 a s m a and whole bod" rahioastivity &e bee" described previously (irl). Unless etheruise indicated, a l l data points for plasma and whale body clearance curves are averages for at least 3 animals and absence of error bars indicates that the coefficient Of variation was 15%. Plasma Dind whole body half-lives were estimated by linear regression analysis.
Total and acid Soluble radioactivity in tissues were distinguished by precipitation with 2 0 % TCA as described previously (3,7). In control experlnent *I-InTn. *I-lnTn-PET or *I-InTn-RSA vas added to nonradioactive liver or skin, and processed for determination of acid Soluble radioactivity. Recoveries Of soluble radioactivity were 94-91) far .I-InTn. and 3 . 9 % for *I-InTn-FET and *I-InTn-RSA. Subcellular fractionation of liver by differential centrifumation was nerformed accordinm t D Crerroriadis aind Sourkes 1191. This seemed label there was also the possibility that the protein could be modified unlikely because Of the much higher pH required for reaction between cycl aind carbohydrates (22). but the inertness of the carrier inulin was also confirmed experimentally. Table I In the absence of competitor, more than 901 of injected radioactivitv from .I-I~T~-ASF was recovered in liver at 15 min after injection following injection Of labeled ASF.

AS 6ho-n in
-. In a11 cases, <IO1 Of radioactivity and enzyme activity were found in the post-lysosomal supernatant fraction, whereas when Intact *I-InTn-ASF was homogenized with liver and fractionated, 9 6 t of the radioactivity. but only 12t of bet.-N-.cetylhexOs.minidaae activity, was recovered in the supernatant.

Kinetics of plasma and whole body clearance
In general. the results of these experiments With *I-InTn-ASF agreed Closely with Tflulte Obtained preVi4USly with Other residualiring labels, such e . * ! w]raffinose (2) and *I-DLT (3). Thus, despite its sire, the attachment Of *I-InTn did not detectably affect the uptake. compartmentation or catabolism Of S F in liver. dies With Mna-liv& proteins. The *I-InTn label Was next applied t0 %dies on the catabolism of native fetuin. Fig. 4 shows that the whole body half-life Of radioactivity from *I-FET is about 1 d, which itr equal +n its n l a c n a half-life in the r a t 1 2 4 ) . while the whole bod" half-lives __ ." ."" ..~" "" ". ~~ ~ ~~~  Table 11, confirms previous evidence using *I-DLT-RSA that catabolism of RSA taker place primarily in muscle and s k i n ( 1 1 .
While this model represents a11 cells involved in the Catabolism Of RSA or FET as a Single compartment and as8ume9 that the distribution Of degradation products between slow and fast release compartments and their rates Of transport through the cell are identical for all cell types. there is not a sufficient amount Of experimental data with various proteins and labels to justify a more eophistisatsd treatment at this time. However, the internal consistency and reasonableness Of the Linetic constants obtalned from the optimization. as well as the quality Of the fit to experimental data for both FET and RSA, lend some support to the compartmental model and justify continued experimentation to test the model in vivo and i n simpler model systems in vitro.