Further Studies on Acetamidination as a Technique for Preparation of a Biologically Valid 3H-Labeled Tracer for Parathyroid Hormone*

SUMMARY The technique of acetamidination of amino groups in parathyroid hormone (PTH) for the purpose of preparation of a tritiated, biologically valid tracer for this hormone, has been more extensively studied. It was found that eight of the ten amino groups in PTH are readily reactive but that two appear unreactive. The kinetics of labeling suggest that preparations of PTH which are labeled to this level of 80% of theoretical consist of a homogeneous population of molecules in which each PTH molecule contains eight tritiated acetamidino groups. There is no question of the presence of unlabeled hormone in such preparations. Eighty per cent labeled acetamidino-PTH is identical, qualitatively and quantitatively, in its biological activity with native PTH as shown by three accepted bioassays: serum calcium elevation, urine phosphate excretion, and activation of kidney cortex plasma membrane adenylate cyclase in vitro. The dose-response relationships are identical for labeled and native hormone in all three systems. Tritiated acetamidino-PTH tends to lose biological activity on storage but full activity can be regained by reduction of the hormone with excess cysteine. However, cysteine reduction conducted 80” 2 hours causes some loss of tritium


Further
Studies on Acetamidination as a Technique

SUMMARY
The technique of acetamidination of amino groups in parathyroid hormone (PTH) for the purpose of preparation of a tritiated, biologically valid tracer for this hormone, has been more extensively studied.
It was found that eight of the ten amino groups in PTH are readily reactive but that two appear unreactive.
The kinetics of labeling suggest that preparations of PTH which are labeled to this level of 80% of theoretical consist of a homogeneous population of molecules in which each PTH molecule contains eight tritiated acetamidino groups.
There is no question of the presence of unlabeled hormone in such preparations. Eighty per cent labeled acetamidino-PTH is identical, qualitatively and quantitatively, in its biological activity with native PTH as shown by three accepted bioassays: serum calcium elevation, urine phosphate excretion, and activation of kidney cortex plasma membrane adenylate cyclase in vitro. The dose-response relationships are identical for labeled and native hormone in all three systems.
Tritiated acetamidino-PTH tends to lose biological activity on storage but full activity can be regained by reduction of the hormone with excess cysteine. However, cysteine reduction conducted at 80" for 2 hours causes some loss of tritium from the hormone. Therefore, dithiothreitol reduction at room temperature was utilized to maintain biological activity.
Eighty per cent acetamidino-PTH can be further purified by ion exchange chromatography on carboxymethylcellulose using a continuous gradient of sodium acetate in 8 M urea. Such chromatography reveals the presence of isohormones in both cold and radioactive PTH. Co-chromatography of tritiated acetamidino-PTH with a cold isohormone of PTH shows that while the isohormones can be separated from one another, the acetamidino derivatives of each isohormone elute in a virtually identical position with their parent unlabeled PTH molecule.
* This work was supported by Grant AM 14496 from the National Institutes of Health and Grant P3B2813 from the National Science Foundation. 4 Supported by Career Development Award AM 70031; to whom correspondence should be sent.
In 1972, we reported the preparation of a biologically active derivative of parathyroid hormone labeled with high specific activity tritium (1,2).
This derivative, acetamidino-PTH,' was active in the in vitro adenylate cyclase assay system described by Marcus and Aurbach (3), and in elevating the serum calcium of thyroparathyroidectomized rats. It was also successfully utilized in establishing the existence of specific binding of PTH to kidney cortex plasma membranes in vitro (4). The radioactive hormone was readily prepared under mild conditions, with generation of no detectable side products.
The labeling procedure was also utilized to prepare tritiated insulin, and the acetamidino derivatives of insulin were extensively characterized (5) ' However, in this previous work we utilized acetamidino-PTH preparations of a variable extent of labeling, some preparations containing as much as an average of seven labeled ammo groups per hormone molecule, and some with as little as one to two labeled groups.
We also observed a considerable variability in the biological potencies of the tritiated hormone preparations although qualitatively the labeled hormone appeared very similar to native PTH. Therefore, we have undertaken a more extensive investigation of the labeling of PTH with acetimidate, and of the biological and chemical characteristics of the labeled hormone.
The data presented here strongly support our earlier conclusions that with the proper precautions in its preparation and handling, acetamidino-PTH is a fully valid and useful tracer for native parathyroid hormone. (1). One to two milligrams of [zH]imidste was incubated for 2 hours at 0" in a NY atmosnhere with 100 to 500 YP of PTH at nH 9.0, in Ne-saturated 0.025.ru borate buffer. The ieaction volume was 100 to '200 ~1. The labeled hormone was then separated from the reaction mixture as described below. Albino Rats-Male animals (129 a) were nurchased from Zivia-Miller and immediately placedbn acalcium-free diet (Nutritional Biochemicals).
All animals used for PTH bioassay were kept on this diet for no less than 4 days and no longer than 10 days prior to use.
In Viva Bioassay+-Rats were surgically thyroparathyroidectomized under ether anesthesia.
After recovery from the anesthetic (10 to 15 min), intraperitoneal injection of hormone or carrier solutions (1 to 10 rnM acetic acid) were given in 0.1 to 0.3~ml volume. A second injection of hormone or carrier was given 2% hours after the first. During the B-hour period following surgery, urine was collected, and at the end of this period, the animals were killed by decapitation and the blood was collected for serum calcium analysis. Serum calcium was determined calorimetrically using the calcium-stat kit purchased from Pierce Chemical Company, Rockford, Ill. Urine phosphate was determined by the Fiske and SubbaRow technique as described for urine by Allport and Keyser (6).
Chromatography-Experimental details for the individual chromatographic systems are given below. All columns were run in 8 M ureaunless otherwise indicated.
Hormone Concentrations-Unless otherwise indicated. all hormone concentrations are based on analysis of total nitrogen by the indanetrione hydrate procedure. Samples were sealed in glass tubes and digested in concentrated HeSOd at 480" for 90 min. Aliquots were then analyzed as described by Jacobs (7).

RESULTS
All of the experiments reported here were done with commercially prepared PTH.2 PTH prepared by gradient elution from CMC has been shown to be homogenous in a variety of systems, although small amounts of isohormones have recently been detected by Keutman et al. (8). We directly investigated the purity and the biological activity of the Wilson "highly purified" material which is prepared by CMC chromatography. Fig. 1 shows the results of a gel electrophoresis experiment of the hormone run in 15% acrylamide-8 M urea as described by Hawker et al. (9), illustrating the presence of a single band of peptide material in this system. This result, of course, does not unequivocally rule out the presence of trace contaminants, or of isohormones, which indeed are present as is demonstrated later in this paper. Nonetheless, it appeared that the dominant material in this commercial preparation was a peptide very similar to PTH in its electrophoretic behavior.
On the other hand, we found that the standard, I-mg lots of this "highly purified" material appeared to be highly contaminated with salt since analysis of total nitrogen by the indanetrione hydrate method always yielded an actual protein content varying from 500 to 700 pg per mg of total material, based on 18.4% nitrogen in PTH. We established that the nitrogen present in these preparations was protein and not ammonium or other nitrogen salts by precipitation of the hormone with trichloroacetic acid prior to the nitrogen analysis.
This treatment always totally removed nitrogen from the sample.
The biological activity of the highly purified hormone was assayed on a microgram of protein basis using thyroparathyroidectomized rats. The results of these assays are shown in Fig. 2 The hormone is very active in elevating serum calcium, and a full return to normal calcium levels (10 mg/lOO) in thyroparathyroidectomized rats can be achieved in 5 hours with injection of 8 to 10 pg of this material under the conditions of our bioassay.
From these results it appeared that the highly purified hormone would be satisfactory for further studies of the amidinating reaction and the properties of the labeled hormone. We therefore proceeded to investigate more fully the maximum extent of labeling of hormone which could be achieved, and the biological activity of such extensively labeled hormone. Earlier we had noted that up to 70% of the theoretical value for labeling PTH (7 out of 10 amino groups) could be achieved with retention of biological activity (1,2), so it appeared possible that under appropriate conditions, complete modification might be possible.
Highly purified Wilson PTH was incubated with a lOa molar excess of methyl[3H] acetimidate for varying periods of time up to 2 hours. The reaction was stopped by passage of the incubation mixture through a column of Bio-Gel P-2 which results in nearly total separation of the labeled hormone from the reaction mixture in about 10 min. Fig. 3 shows the excellent separation of the hormone from the smaller radioactive molecules (imidate and hydrolyzed imidate) by this column and describes the conditions for optimal use of this P-2 column for this purpose. We found  Procedures", in 0.4 ml of borate buffer, was placed on a Bio-Gel P-2 column (60 X 0.9 cm), (50 to 100 mesh) equilibratedwith 50 mM acetic acid. Fraction size is 1.0 ml.
that it was essential to use the large size Bio-Gel beads (50 to 100 mesh), since the 200 to 400 mesh Bio-Gel P-2 showed a significant retardation of the hormone, probably due to nonspecific binding. Use of 1.0 mM acetic acid rather than 50 mM also resulted in significant retardation of the hormone and poorer separation. In addition, it was found that column age was critical and that a given P-2 column in 50 rnM acetic acid could only be used for 1 to 2 weeks before nonspecific binding of the hormone increased to an unacceptable level. Utilizing the correct conditions, however, this method is much superior to the use of dialysis, which we described earlier (l), since during dialysis there is always significant loss of PTH.
The result of the labeling experiment is shown in Fig. 4. The specific activities were calculated utilizing both the absorbance at, 277 nm (Emn,ioo = 7.0) and a total nitrogen analysis. As can be seen, the hormone is rather quickly labeled to about 80% of the theoretical value, and subsequently no further reaction occurs. This result was further explored by attempting to label more extensively an isolated sample of 80% amidinated PTH by addition of a second large molar excess of imidate. After a 2hour reaction period the labeled hormone was again passed through Bio-Gel P-2 and its specific radioactivity was determined. As indicated by the triangles in Fig. 4, this treatment produced no further labeling of the hormone, since the specific activity is identical with that of the initial material. Thus, 80% of theoretical, that is, a total of eight amino groups per parathyroid hormone molecule, appear to represent the maximum extent to which the hormone can be labeled with the imidate reagent under the conditions utilized.
The biological activity of 80% labeled PTH was investigated initially by comparison with cold hormone and with an active 50 ye labeled PTH sample, prepared at an earlier date, in a single dose assay. As shown in Table I  Ten milligrams of imidate were incubated with 1 mg of highly purified Wilson PTH (weight basis). Samples were removed at 15, 30, and 60 min and were passed through a Bio-Gel P-2 column (50 to 100 mesh). Specific activity was determined and per cent of labeling was calculated.
In a second experiment (A, A) 170rg of PTH (nitrogen basis) labeled to 80% of theoretical were reincubated with 2 mg of methyl[3H]acetamidate for an additional a-hour period, then isolated, and the specific activity was determined.  labeled hormone preparations were less active than the native hormone, and in particular that the older tritiated hormone which was active at an earlier date had lost considerable biological activity. Since secondary radiation damage frequently can occur in highly radioactive chemicals due to generation of free radicals (particularly oxygen radicals in aqueous solution), and since PTH is particularly sensitive to oxidation of methionine by oxygen radicals (HsOt), we investigated whether a reversible oxidation of methionine might have occurred in the tritiated hormone preparations. Separate samples of active hormone were oxidized by HzOz, and reduced by heating with excess cysteine as described by Tashjian and Munson (10). Eighty per cent labeled [3H]PTH was treated similarly, as was a sample of "old" labeled hormone, and all samples were bioassayed. Table  I also shows the data from all of these single dose bioassays. It is clear from these data that the activity of the 80O/,-labeled PTH and the "old" [3H]PTH was elevated to that of the native hormone by reduction with cysteine. Oxidation of all of the samples eliminated biological activity. The data, therefore, support the belief that oxidation of methionine catalyzed by radiationgenerated free radicals may occur to some extent during extensive Data are expressed as the elevation of serum calcium over that determined for a control group of animals run on the same day. Each point represents the mean of 7 to 10 animals, fS.D.
labeling and during storage of labeled hormone. Therefore, it is important to reduce fully such hormone preparations prior to biological experiments, and to maintain the hormone in a reducing environment.
A dose-response curve for fully reduced 80% labeled PTH was next obtained by measurement of both serum calcium and urine phosphate in thyroparathyroidectomieed rats. Fig. 5 shows that within the error of this assay the [aH]PTH is identical with the control native hormone preparation in its ability to elevate serum calcium. For these experiments a control group of animals was run on the same day as each experimental group so that the data are presented as an increase in serum Ca2+ over those thyroparathyroidectomized controls. Fig. 6 indicates that excretion of phosphate in the urine is also dramatically elevated by the [aH]PTH compared to the native hormone. Thus, these data appear to provide unequivocal evidence that 80% acetamidino-PTH is virtually identical with the native hormone in its ability to elevate serum calcium and urine phosphate, the two most characteristic activities of parathyroid hormone.
A point of some concern to us, however, related to the conditions required for reduction of the hormone. As described above, this procedure requires heating the hormone to 80" in 0.1 M cysteine at acid pH for 2 hours. It seems possible that under these conditions some 3H exchange might occur, or perhaps some hydrolysis of the hormone to smaller fragments, or hydrolysis of the PTH imidate residue itself. PTH is known to be sensitive to mild acid hydrolysis at the Glnzs-Asp,, peptide bond (10) and such a chemical change might be undetected in the bioassay since the l-29 fragment of 1;TH thus produced has biological activity (10). As shown in Fig. 7, gel filtration of the 80% a.cetamidino-FTH on a Bio-Gel P-10, 8 M urea column did indeed suggest that some degradation of the hormone had occurred. The main peak elutes in a position consistent with PTH, but two smaller fragments were also observed, suggesting that some degradation of the hormone may have occurred. An alternative explanation for the appearance of these secondary radioactive peaks was the possibility that trace contaminants existed in the original hormone preparation. Such contaminants The animals were hydrated with 4 ml of water by stomach tube just after surgery. Each point represents the mean of 7 to 10 animals, f S.D.

Fraction
No.
A hormone samnle (0.5 ml) was annlied to a column (60 X 0.9 cm) equilibrated in 8~ urea, Fraction s&e, 0.5 ml. might be undetected in gel filtration of the original hormone, but might be highly reactive with the imidate, thereby generating detectable peaks upon labeling.
9.80 f 1.1 10.10 (5.0-18.0)" a Average in milligrams per 100 of five rats 5 hours after thyroparathyroidectomy f S.D. b Total milligrams of phosphate excreted in 5 hours after thyroparathyroidectomy.
c The variability is shown as a range, rather than f S.D. because of the extreme variation in the urine phosphate. Ten milligrams of PTH dissolved in 30 ml of 0.01 M sodium acetate-8 M urea-l mM DTT, pH 4.9, were applied to a column (0.5 X 6 cm) equilibrated in the same solution.
After the hormone had entered the column, a gradient of 0.01 to 0.2 M (pH 6.0) sodium acetate (8 M urea-l mM DTT, 70 ml in each reservoir) was begun (---). After elution of the two hormone peaks, the remaining material present in the column was eluted with 1 M NH4 acetate (pH 6.9). Fraction size is 0.5 ml.
to our procedures for producing [3H]PTH. First, the use of dithiothreitol as a reducing agent in place of cysteine was investigated.
As indicated in Table II, 80% [3H]acetamidino-PTH kept in 50 mM DTT does indeed retain apparent full biological activity, and we therefore turned to this agent to maintain the hormone in the fully reduced state.
Second, in order to eliminate any question as to the homogeneity of the original hormone we initiated an effort to purify extensively PTH from the less homogenous commercially available Wilson material, by chromatography on CMC in 8 M urea as described by Keutman et al. (8). Each material was rechromatographed on a separate column of the dimensions and with the gradient described for Fig. 8. 10 mg of "purified" PTH and chromatographed it directly on a small CMC column prepared in a Pasteur pipette (0.4 x 6 cm). In order to avoid contamination of the hormone with ammonium ion so that specific activities could be based directly on nitrogen analysis, we utilized sodium acetate in place of ammonium acetate in the gradient. Fig. 8 shows the chromatographic profile obtained by this procedure.
Large amounts of ultraviolet absorbing material did not stick to the column and this material was washed completely out of the column before starting the gradient, as indicated.
Initiation of the gradient resulted in elution of a broad band of ultraviolet absorbing material, spread over the entire gradient, several minor peaks, and two major components (I and II).
The remaining ultraviolet absorbing material was then eluted with 1 M ammonium acetate. The broad base-line of ultraviolet absorption is believed to result from the oxidation of sulfhydryl groups of DTT to disulfide as reported earlier by Keutman (8). This interpretation is supported by the observation that the base-line drops back to its original value when the column was washed with 1 M ammonium acetate which contains no DTT.
Of the two principal peaks eluted by the gradient, Peak II elutes in a position consistent with that expected for the dominant form of PTH reported by Keutman et al. and Peak I elutes in a position compatible with a major isohormone described by Keutman. This latter material is present in quite significant amounts in this preparation.
It should also be noted that we did not observe the third isohormone reported by Keutman et al. which elutes after the dominant form of the hormone.
However, small amounts of this isohormone may be present in some of the tritiated hormone preparations prepared earlier from highly purified PTH (e.g. see Fig. 11).
Peaks I and II were rechromatographed on the sodium acetateurea gradient and each eluted as a single symmetrical peak of ultraviolet absorbing material (Fig. 9). Both materials were then bioassayed in the rat to confirm that these peaks did indeed 1.8 f 0.9 3 6 pg of Peak II.
. 9.1 f 0. c The variability is shown as a range, rather than f S.D. be-6 cause of the extreme variation in the urine phosphate. 10) on Bio-Gel P-10. Furthermore, there are no secondary peaks as were observed with cysteine-reduced PTH (Fig. 7). We next investigated the chromatographic behavior of PTH, '1 i 1 its isohormones and their acetamidino forms, on the high resolution CMC-urea column described above. CMC chromatography of 60% acetamidino-PTH prepared from highly purified Wilson PTH, utilizing the same column and gradient system as described 8 16 24 32 40 48 56 64 72 SO in Fig. 8, gave the results shown in Fig. 11. In contrast to Fig. 8, hormone preparations probably vary in their content of isohormones since this prepa.ration was from a different batch of commercial hormone. It was of considerable interest to observe represent PTH. As shown in Table III both Peaks I and II were biologically active in elevating the serum calcium of thyro-that the dominant radioactive peak was eluted from this column in precisely the same position as was the dominant form of the parathyroidectomiaed rats. In fact, Peak I, tentatively identified as an isohormone, appears more active than Peak II, which hormone described in Fig. 8, suggesting that acetamidino-PTH is the dominant form of the hormone in this preparation. From and cold PTH are very similar in their chromatographic behavior these data, it is clear that further work will be needed to identify This was further confirmed by the co-chromatography experithe precise nature of these two forms of PTH, and their relation-ment described in Fig. 12, in which Peak I from Fig. 9 was mixed ship to the isohormones described by Keutman et al. (8).
with 60% acetamidino-PTH and chromatographed on the CMCurea column. Further studies were conducted on the dominant hormone It is quite clear from the elution pattern that the form, Peak II. This material was labeled to the maximum acetamidino-PTH is chromatographically indistinguishable from extent with acctimidate and the specific activity calculated.
native PTH even on this high resolution system which can As separate PTH isohormones and labeled isohormones from one with the highly purified hormone, we observed a maximal labeling another. near 80% of theoretical. The homogeneity of this material was then further investigated on &o-Gel P-10 in 8 M urea; the hor-DISCUSSION mone was kept in the reduced form by addition of 1 mM DTT to This paper confirms and considerably extends our earlier report all of the solutions utilized in this work. As shown in Fig. 10, that acetamidino-PTH is a biologically active derivative of the hormone, and that introduction of tritium into the hormone by the amidinating reaction produces a valid and highly useful tracer for PTH. The generation of fully active, 80% labeled PTH is of fundamental importance to these conclusions, since this result eliminates the possibility that any biological activity is due to unlabeled PTH in the preparations. Furthermore, the striking parallel between the biological activity of labeled and unlabeled hormone and the kinetics of the labeling reaction, taken together strongly suggest that the 8Oa/, labeled material is the active species, and not some more lightly labeled species present in the preparations.
The presence of a significant number of less labeled molecules in 80% acetamidino-PTH, would require an equal number of more extensively labeled molecules ( >SO$&) in order to achieve an average of 80%. However, if such more extensively labeled molecules are present, one would expect that an increase to an average >800/, would eventually result, with time. The fact that the labeling quickly reaches its maximum value, and no additional labeling can be achieved (Fig. 3) beyond this point thus argues strongly that the dominant species present in these preparations is 80% acetamidino-PTH.
In our earlier report (2) we showed a dose-response relationship between PTH and activation of rat kidney cortex adenylate cyclase identical with that reported for PTH in other laboratories (3,11). The present paper shows similar bioassays for elevation of serum calcium which is the classic bioassay for PTH, and for elevation of urinary phosphate. Thus, the acetamidino-PTH derivative has now been demonstrated to be quantitatively as active as native hormone in both of the accepted quantitative bioassay systems, and additionally its effect on urine phosphate elevation is identical with that of the native hormone. Thus, there is a firm basis for concluding that the tritiated derivative is a fully valid tracer for PTH. In further support of this con-clusion, it should be noted that we have demonstrated binding of [3H]PTH to isolated kidney plasma membranes in vitro (4) which is specific for only active PTH.
The chromatographic data provide further chemical evidence which illustrates the similarity of the acetamidino and native PTH. Since the free amino groups in PTH are modified by amidination, and since it is these functional groups which are responsible for the binding of hormone to CMC, any significant modification of the ability of the positive lysyl sites to interact electrostatistically with negative sites should be reflected in altered affinity of the hormone for CMC. However, as shown in Figs. 9 to 12, acetamidino-PTH elutes virtually identically with native hormone even on columns which resolve isohormones of PTH from one another. Thus, chromatographically, acetamidino-PTH is more similar to its parent hormone than it is to its naturally occurring isohormone.
As indicated above, it is critical to have a fully reduced PTH derivative in order to retain biological activity. We have been able to substitute 50 mM DTT at room temperature for 100 mM cysteine at 80" to reduce the hormone so that any possible hormone degradation due to reduction is eliminated. As we reported earlier (4), the inverse correlation between membrane binding of PTH to kidney receptor in vitro and oxidation of hormone by Hz02 is an interesting one and probably explains the loss of biological activity in oxidized PTH. It now seems likely that a direct measure of the extent of oxidation of PTH in future hormone preparations should be obtainable by a standardized membrane-binding assay.
It will be of considerable interest to determine which amino groups in PTH are not labeled by acetamidate. Since the NHtterminal group is important in biological function (12,13), it is possible that this residue is protected and not reactive to the imidate. The lack of any modification of biological activity in the labeled hormone supports this suggestion. It is possible that a substantial segment of the NHz-terminal end of PTH is protected or buried within the PTH molecule. Preliminary evidence in our laboratory indicates that the l-34 NHt-terminal fragment of PTH can be labeled with acetimidate but that this labeling leads to some loss of biological activity. Presently it is not known whether this loss is due to oxidation of the peptide, or to the modification of sites more critical to biological function than those labeled in the intact molecule. These and other questions are presently under investigation in our laboratory.
The presence of at least one major isohormone of PTH in commercially available preparations is of interest. This observation confirms the earlier report concerning the existence of such isohormones (8). In one batch of commercially prepared hormone, we found a rather large amount of only one isohormone rather than the trace amounts of two isohormones as reported by Keutman, but it appears from Fig. 11 that other batches of hormone more closely resemble the preparation described by these workers. In addition, it is of interest that the isohormone was more active in the rat in wivo bioassay than was the dominant hormone species in the experiment described in Table III. It seems likely that the amount and nature of such isohormones will vary with the original source (e.g. bovine variety differences), and the subject of PTH isohormones, their origin, and their comparative biological activities will be of interest in future work. The use of [3H]isohormone should facilitate such studies.
The concept of the general utility of acetimidination as a technique for labeling peptides containing free amino groups is considerably strengthened by the data presented in this paper.
The acetamidino-lysine residue is clearly very similar to the lysine free amino group and in many cases where a lysine residue is critical to biological activity it may nonetheless be possible to acetamidinate without loss of such activity. In fact, acetamidino lysine is sterically quite similar to arginine, and it is well known that the conservative replacement of arginine for lysine frequently does not detectably modify biological function in proteins.
In fact, in one of the earliest studies of amidination and its effects on biological function, Wofsey and Singer (14) found that binding of antigens to certain rabbit antibodies was not altered in any way by exhaust.ive amidination of antibody amino groups.
In practice, we have found that both YI'H and insulin can be extensively acetamidinated without detectable alteration in biological activity (1,2,5), and we have likewise found that ACTH can be acetamidinated and still retain biological potency.3