BIOLOGICAL MASS SPECTROMETRY, VOL. 21, 560-566 (1992)
Gas Chromatographic/Mass Spectrometric Analysis of Stable Isotopes of 3-Methylhistidine in Biological Fluids : Application to Plasma Kinetics in Vivot J. A. Rathmacher, G. A. Link, P. J. Flakoll and S. L. Nissent Department of Animal Science, Iowa State University, Ames, Iowa 50011, USA
A simple and rapid method for measuring fmetbylhistidine (3MH) in plasma and urine is described. Internal standard, 1-methylhistidine(IMH), was added to plasma, acidified and absorbed onto cation-excbange columns It was then eluted from columns, dried, and derivatized for gas chromatography/mass spectrometry. A major fragment of 3MH was monitored at 238 u and fmethyl+nethyl-2H3)histidine (d3-3MH) (used for in vivo kinetics) at 241 u, whereas 1MH was monitored at 340 u and eluted 0.5 min later than 3MH. Standard curves for plasma analysis were linear and nanamole amounts of 3MH in plasma were determined with a precision of 3.5%. 3MH was also quantitated in urine; however, because of substantial amounts of IMH, ("0,)IMH was used as the internal standard. Nanamok amounts of 3MH were determined in urine with a precision of 2.7%. Application of the 3MH analytical method was used to develop a kinetic compartmental model by using the stable isotope of 3MH, d33MH. Cattle, like humans, quantitatively excrete 3MH in the urine. A young bovine was injected with 4 3 M H and the enrichment curve in plasma was evaluated in order to obtain a steady-state production rate of 3MH. Tbe decay curve was modeled through the use of NIH-SAAM modeling program. The kinetics of d3-3MH from plasma were adequately described by a three-pool compartmental model. The de now production rate of 3MH estimated in the calf was 665 pmol per day. This corresponded to an estimated fractional turnover rate of 1.56% per day, which was similar to estimates obtained from urine collections. These data suggest that 4-3MH can be used to model 3MH production in viva
INTR ODUC TI O N
The primary sequence of actin and myosin white fibers in skeletal muscle contain the unique amino acid 3methylhistidine (3MH3).' During degradation of muscle proteins, free 3MH is released. Yet, 3MH is not reutilized for protein synthesis because it does not have a specific tRNA.' 3MH is quantitatively excreted in the urine of man, rat, cattle and rabbit;'-' therefore, it is thought to be a marker of skeletal muscle protein breakdown. Urine 3MH estimation of muscle proteolysis depends on quantitative collection and accurate measurement of urinary 3MH. It is assumed that no metabolism of 3MH occurs in uiuo, which is true in most species.6 However, in sheep' and pigs,' 3MH is not quantitatively excreted in urine but is retained in muscle as a dipeptide, balenine.6 Hence, urinary 3MH excretion cannot be used to estimate muscle protein breakdown in these species. An alternative method of quantitating 3MH production in uiuo would be to isotopically measure the de novo production rather than the excretion. An isotopic model for measuring 3MH production would have the
t Journal Paper J-14443 of the Iowa Agriculture and Home Economics Experiment Station, Ames, Iowa 50011, project no. 2928. $ Author to whom correspondence should be addressed. 1052-9306/92/110560-07 $08.50
0 1992 by John Wiley & Sons, Ltd.
advantages of eliminating the need for urine collections ; it also may be applicable to species which do not quantitatively excrete 3MH. The objective of this research was first to develop a simple method to quantitate 3MH and the enrichment of a stable isotope tracer, 3-methyl-(methyl-2H3)histidine (d3-3MH), in plasma through the use of gas chromatography/mass spectrometry (GC/MS). Secondly, the method to quantitate the enrichment of the stable isotope of 3MH was applied to a kinetic model. This model will be used to estimate the de nouo production rate of 3MH in cattle, which are thought to excrete 3MH quantitatively, as do humans. EXPERIMENTAL Materials
3-Methyl-(methyl-2H,)histidine was purchased from MSD Isotopes (99.5% isotopic purity, St Louis, Missouri). 3MH, 1-methylhistidine (lMH), Jack bean type IX urease (61000 units g-'), and Dowex-SOW in the hydrogen form were obtained from Sigma Chemical Co. (St Louis, Missouri). A urease solution was prepared by adding 3.5 mg of the crystalline urease to 100 ml of a 20 mM phosphate buffer (adjusted to pH 7.0). Acetonitrile and N-methyl-N-(t-butyldimethylsilyl) trifluoro- acetamide (MTBSTFA), were purchased from Received 3 April 1992 Revised I8 June I992
3-METHYLHISTIDINE IN PLASMA
Regis Chemical Co. (Morton Grove, Illinois). Ammonium hydroxide, perchloric acid (PCA), hydrochloric acid and filter columns were procured from Fisher Scientific (Fair Lawn, New Jersey). The cation-exchange columns were prepared by adding 3 ml of Dowex-SOW cation exchange resin (50 :50 v/v) in the hydrogen form to filtering columns and washing with 4 x l ml aliquots of 0.01 N HCl. Plastic microinjection sample vials were obtained from Sun Brokers, Inc. (Wilmington, North Carolina).
Internal standards One of the internal standards, (1,l -la0 2 ) 1-methyl- histidine ((1802)1MH),was prepared by exchanging the carboxyl oxygens in l80water: 1 ml of H2180 (96% l 8 0 , Cambridge Isotopes, Woburn, Massachusetts) was added to 50 mg of dry crystalline 1MH. The sample was gassed with HCl for 1-2 min until a pH of much less than 1.0 was achieved. The sample was capped and incubated at 100°C for 30 min, then it was neutralized with 3 N NaOH. The resulting neutral amino acid solution was diluted to 100 ml. Eighty-six per cent of this final product was (180,)1MH (m + 4) as assessed by GC/MS. (180,)1MH (40 pl of a 3 mM stock solution) was used as the internal standard for the analysis of 3MH in urine because a significant quantity of 1MH accumulates in urine. At the normal concentrations of 1MH in urine there was no signal overlap of natural 1MH in the signal of (180,)1MH (ion 344). The '802-labeled standard was quite stable under assay conditions used. (l8O2)1MH had remained stable after being stored at - 70 "C for two years, being treated with perchloric acid or being incubated at 37 "C, at pH 7 for 2 h. LMH was used as an internal standard (50 pl of 3.5 mM stock solution) in the analysis of 3MH from plasma because of its structural similarity to 3MH and its undetectability in the plasma of animals fed meatfree diets.
Plasma 3MH analysis One milliliter of plasma and 50 pl of 1MH internal standard were added to a plastic sample tube (12 ml); 3 ml of 1.5 N PCA was then added. The tube was vortexed and centrifuged at 2300 x g for 15 min at 5°C. The supernatant was poured directly onto a prepared cation-exchange column. After the supernatant drained through the column, it was rinsed with four 1 ml aliquots of 0.01 N HCl and the rinses were discarded. The compounds of interest were eluted from the column with four 1 ml washes of 25% NH,OH into a 20 ml scintillation vial. The eluate containing 3MH was heated to 65°C in an aluminium block under a stream of nitrogen gas. When the eluate appeared dry, two sequential 100 p1 aliquots of methylene chloride were added to azeotrope off any additional water. To derivatize the dried sample and prepare it for GC/MS analysis, 100 pl of acetonitrile and 100 pl MTBSTFA were added. The vial was capped and allowed to incubate overnight at room temperature. The derivatized
561
sample was transferred into a plastic autosampler vial and capped. GC/MS analysis of the samples was accomplished using a Hewlett-Packard gas chromatograph/mass selective detector (Model 5890/ 5970B). The column used was a 25 m x 0.22 mm i.d. x 0.11 pm film thickness, cross-linked methyl silicone gum phase capillary column (HP-1, Hewlett Packard, Avondale, Pennsylvania). The injection port was set at 285 "C, the transfer line was set at 285 "C and the initial oven temperature was set at 50 "C. After splitless injection, the oven was programmed to hold for 0.5 min. It was then ramped (ramp 1) up to 240°C at 50°C min-' and held at 240°C for 4.1 min; a second ramp (ramp 2) to 300°C at 50°C min-' followed and the oven temperature was held at 300°C for 1 min. The retention times were between 6.0 and 7.5 for 3MH and 6.5 and 8.0 min for lMH, depending one the age and length of the column. The major ion fragments for 3MH and d3-3MH were monitored using selected ion monitoring (SIM). 3MH was monitored at 238 u and its stable isotope, d3-3MH, at 241 u, whereas 1MH was detected at 340 u. 3MH in plasma was quantified from a linear peak area standard curve. When d3-3MH was used as tracer, as described below, the enrichment was quantitated in plasma as the ratio of d3-3MH/3MH (ion abundance of 241 :238). The natural background enrichment of d3-3MH/3MH was subtracted from detected d,-3MH ion abundance. The ion overlap of d3-3MH into the signal of 3MH was not backed out because the interference was negligible. Standards were prepared by pipetting known quantities of 3MH from a 50 pM stock solution and 50 pl of 1MH internal standard into plastic sample tubes (12 ml). One milliliter of double-deionized water was added to each standard tube. As the standards had little neutralizing capacity compared to plasma, they were acidified with only 0.25 ml of 1.5 N perchloric acid then treated in the same manner as the plasma sample.
Urine 3MH analysis Unlike plasma, urine contains relatively large amounts of lMH, making 1MH an impractical internal standard. Therefore, di-labeled (I8O2)1MH was used as the internal standard to quantitate 3MH in urine. A urine sample was prepared by transferring 1 ml to a microfuge tube and removing particulate matter by spinning the tube for 3 min. From this tube, 100 pl of urine was pipetted into a plastic sample tube (12 ml). To this 40 p1 of internal standard was added. One milliliter of double-deionized water was also added to the tube, and the contents were acidified with 5 p1 of 3 N HCl. The urine sample was vortexed, poured over a cationexchange column (prepared as already outlined) and allowed to drain. The column was rinsed with four 1 ml aliquots of 0.01 N HCl. Subsequently eluates from four 1 ml aliquots of 25% ammonium hydroxide were collected in a scintillation vial and dried with a stream of nitrogen gas on a heating block at 65 "C. The high level of urea in urine must be removed prior to derivatization of 3MH. To accomplish this, the dry sample was incubated with 1 ml of urease solution for 2 h at 37°C. The same was dried again on the heating block, derivatized
562
J. A. RATHMACHER. G. A. LINK, P. J. FLAKOLL A N D S. L. NlSSEN
with 100 pl acetonitrile and 100 pl MTBSTFA, and incubated overnight at room temperature. The derivatized sample was transferred into an autosampler injection vial, with care being taken to leave precipitated material behind in the scintillation vial. The sample was injected in a Hewlett-Packard gas chromatograph/mass selective detector by using the above-described procedure. ("02)1MH was monitored at M 4 or 344 u. Peak areas of natural 3MH and ("02)1MH were measured in urine using SIM and 3MH was quantitated from a linear peak area standard curve.
+
squares between observed and calculated data points. The model consisted of three compartments or pools of 3MH; the only exit from model was out of pool 1. A steady-state solution was also obtained and initialized by setting the mass of compartment 1 equal to the mean concentration of natural 3MH in the plasma multiplied by the space of distribution in compartment 1. Compartmental masses and flux of 3MH between compartments were also acquired.
RESULTS
Measurement of 3MH production rate in vivo
Application of the 3MH method was accomplished in a castrated male calf (216 kg), which is representative of animals, such as humans and rats, which quantitatively excrete 3MH. The animal was placed in a metabolism cage, catheterized in the external jugular vein on the morning of the study, and injected with a bolus of d,-3MH (30 pmol). After the injection, serial blood samples were taken from 0 to 4320 min. Total urine was also collected for three consecutive 24 h periods. The enrichment of d,-3MH was then quantitated in blood over time. d,-3MH decay from plasma was evaluated using a compartmental model developed through the use of the Simulation, Analysis and Modeling program (SAAM).'.'' The model, illustrated in Fig. 6, was configured by entering the isotope ratio in plasma of the d3-3MH:natural 3MH into compartment 1 over time. The ratio was standardized by dividing the isotope ratio by the dose of tracer in the bolus injection. The model was solved using SAAM and allowed to converge on the observed tracer data, produced transfer coefficients that minimized the weighted total sum of
The electron impact (EI) ionization mass spectra of tertbutyldimethylsilylated 3MH is presented in Fig. 1 and the mass spectra of tert-butyldimethylsilylated 1MH is shown in Fig. 2. The mass spectra of these two tertbutyldimethylsilylated amino acids produced a characteristic fragmentation pattern exhibited by other amino acids." The largest suitable fragments for SIM of 3MH (Fig. 1) were the [M - 571 and the [M - 851 ions 340 and 238 u, respectively. The ion at 238 u was used for quantitation of 3MH as we often found other amino acids or non-specific contaminants producing interference at 340 u. The largest suitable fragment for SIM of lMH, in our assay conditions (Fig. 2), was [M - 571, ion 340. The 302 fragment found in both the mass spectra of 3MH and 1MH does not contain the imidazole ring structure and therefore the 302 fragment was not used. Plasma and standards exhibited similar chromatograms that contained baseline resolution of both the base peak 238 u and the tracer 241 u (Fig. 3). Also, the baseline resolution was obtained for 1MH at 340 u.
y
3- METHYLHlSTlDlNE
NH
3
/si CH3
y3
-c-CH3
c,H3
dHB
.c;
-Si -C-CH3 dH3
dH3
340
54 3 / 400
500
600
V'
ilq 0 .\
?@El
MASS/CHARGE Figure 1. El ionization mass spectra of ?eft-butyldirnethylsilylated3MH derivative.
3-METHYLHISTIDINE I N PLASMA
563
1-METHYLHISTIDINE
qli"
3132
I
W
0
34
I
340
BQO
;3 i 0
180
200
340
560
400
600
700
800
MASSICHARGE Figure 2. El ionization mass spectra of tert-butyldimethylsilylated 1MH derivative.
Further spectral anaIysis of these peaks indicates that standard 3MH ion spectra are identical to that in plasma. Standard curves developed for analysis of 3MH in plasma were linear for 0 pM to 50 pM, with typical
u 4.0ES c
3.0E5
C
2
2.0E5
r
I . WE5 0
6.0
)
.
.
__-
/
1,
-
6.5
7.0
7.5
7.0
7.5
7.8
7.5
Time r m i n . ) I o n 2 4 1 . 10 a m " .
6.0ES 5.0E5
u 4.0E5
i.O
6.5
I a n 340. I0 a m " . 6.0E5j 5.OE5
u 4.0E5 Y
3.0E5 c
2
2.0E5
c 1.WE5 0 G.O
6.5
Figure 3. The El GC/MS selected ion chromatograms from a sample of 1 ml of plasma for 3MH (ion 238, M), d3-3MH (ion 241, M + 3), and 1MH (ion 340, M). 1/200 of the derivatized sample was injected.
regression statistics as follows: observed ratio (238/ 240) = 0.2338 (nmol 3MH), R 2 = 0.9995. Table 1 shows the percentage recovery of 3MH from plasma for two genetic groups of calves. The recovery of added known amounts of 3MH to 1 ml of plasma averaged 96.1 k 2.15% for the two groups of calves. The data in Table 1 illustrate the typical range of variances observed in the 3MH measurement between replicate samples. The coefficient of variation between plasma samples A and B were 5.6% and 1.4% for pooled plasma 1 and 2, respectively. The inter-assay variation of repeated analysis of these plasma samples averaged 5.6%.The assay has previously been sensitive enough to detect 61.25 pmol of 3MH. In a separate study involving five steers given a bolus dose of tracer, the enrichment in plasma was determined. The coefficient of variation from triplicate analyses was determined at two time points when the isotopic enrichment was approximately 1.0 and 0.5. Plasma tracer enrichments from these steers were detected with an average precision of 5.9% at 1 % enriched and an average precision of 10.35% when enrichments were measured at 0.5% (data not shown). 3MH was resolved from baseline in urine (Fig. 4) and exhibited similar chromatographs to those of plasma. Also, baseline resolution was obtained for ("0,)lMH at 344 u. Standard curves developed for analysis of 3MH from urine were linear, with typical regression statistics, ion ratio (238/344) = 0.2804 (nmol 3MH), R2 = 0.9994. Table 2 presents the recovery of 3MH added to calf urine. Known amounts of 3MH were added to 100 pl of urine of two calves. Coefficients of variation between urine samples A and B were 1.25% and 4.21% for calves 1 and 2, respectively. The interassay variation of the 3MH in urine averaged 5.0%, which was based on three repeated analysis of samples A and B of the two calves. The percentage recovery, which averaged 98%, is also presented in Table 2.
J. A. RATHMACHER, G. A. LINK, P. J. FLAKOLL A N D S. L. NISSEN
564
Table 1. Recovery of 3MH added to calf plasma Pooled pladma 2
Pooled plasma 1 Samples'
Additional 3MH added (nmol)
A
0
B
0
3MH found (nmol)
7.74 7.48 8.37 8.61 8.05 0.229 10.23 9.98 12.96 13.33 17.81 17.69
3MH found (nmol)
Percentage recovery
*
x*SE C D E F G H xfSE
2 2 5 5 10 10
Percentagerecovery
6.34 6.11 6.32 6.23 6.25 f 0.046 '8.37 8.04 10.98 10.48 15.45' 14.81
108.9 96.6 98.2 105.6 97.6' 96.4 100.6 f 1.99
105.7 89.7 92.5 84.6 91.9 85.6 91.7 + 2.83
a Internal standard (1MH) was added to 1 ml aliquots of plasma collected from two growing groups of calves differing in frame size; pooled plasma 1 was from large framed calves and pooled plasma 2 was from small framed calves. A 'spike' of 3MH was added to aliquots C, D, E, F, G and H.
Ian 2 3 8 . 1 0
ISOTOPE RATIO -BY DIVIDING BY DOSE
amu
9.0ES
J
3MH
h
6.5
7.0
Tirna ( n r n 1
I n n 3 4 4 . 1 8 amu.
j l ' 0 - 1 ~ ~
l.0E-06
1.2ES 1.0ES
m
t1
B.0E4 6.0E4
i4
0E4
R
I I
UI
-
1 OE-07'
1
0
0 5
1
15
2
2.5
3
35
4
TIME (min x 1000) Tina
lnin
Figure 5. The decay curve of ('HJ3MH : 3MH isotopic ratio in plasma in a calf given a bolus injection of tracer. The symbols represent the observed data points and the line is calculated data produced by SAAM.
I
Figure 4. The El GC/MS selected ion chromatograms from sample of 100 pl of calf urine for 3MH (ion 238, M) and (l80,)1 MH (ion 344, M + 4). 1/200 of the derivatized sample was injected, with ion 238 representing 20.65 nmol of 3MH and ion 344 representing 175 nmol of ('80,)1 MH.
Table 2. Recovery of 3MH added to calf urine Calf 2
Calf 1 Samples.
Additional 3MH added (nmol)
A
0
B
0
x*SE C D E F G
H xiSE
5 5 10 10 20 20
3MH found (nmol)
9.16 9.23 9.35 9.46 9.30 f 0.058 14.87 14.95 20.1 9 19.84 30.34 30.42
Percentage recovery
111.5 112.7 108.8 105.4 105.2 105.6 108.2 f 1.24
3MH found (nmol)
20.65 22.34 19.92 21.23 21.03 f 0.443 25.45 24.87 30.24 30.0 38.53 39.5
Percentagerecovery
88.9 76.7 92.1 88.0 87.5 97.7 88.5 f 2.80
"Internal standard ((1802)1MH) was added to 100 PI aliquots of urine collected from two growing calves. A 'spike' of 3MH was added to aliquots C, D, E, F, G and H.
3-METHYLHISTIDINE IN PLASMA
,/---\,
\
[COMPARTMENT
I\
357
DE NOVO PRODUCTION
COMPARTMENT 3
13809 n r-
Figure 6. This diagram represents the steady-state compartment mass size (pmol) and transfer (pmol min-’) of 3MH between compartments for a calf and the production rate of 3MH into pool 3.
Figure 5 presents the decay curve for d,-3MH : 3MH ratio in plasma obtained from the in uiuo kinetic study. This is intended to document the value of this method for estimating the de nouo production rate of 3MH from a three-pool compartmental model. After the bolus injection of d,-3MH, the tracer was ‘rapidly distributed throughout plasma, with the isotopic enrichment rapidly decreasing during the first 6 h. This was in contrast to the last 96 h, where the rate of isotopic disappearance was significantly less. The model utilized was a three-pool compartmental model with one primary exit from the system. The one exit from the model was from pool 1, a plasma compartment of 3MH. The de nouo production of 3MH was into pool 3, which is thought to be an intramuscular compartment of free 3MH. Figure 6 displays the values obtained at steady state for the mass of the compartments (pmol) and the rate of flow (pmol min- ’) of 3MH between compartments. This method, in addition to estimating the production rate of 3MH per day (660 pmol per day), allows for the calculation of the fractional breakdown rate of skeletal muscle protein, which was 1.56% per day for this example. This calculation compares favorably to that of urinary 3MH, which was 720 pmol per day and 1.69% per day for the urinary production rate and fractional breakdown rate, respectively.
DISCUSSION Histidines are difficult amino acids to derivatize for GC because of the polarity introduced by the imidazole ring.” Therefore, if an internal standard is to be used it must be nearly chemically similar to 3MH to be useful. Norleucine, for example, would be a poor internal standard for 3MH quantitation. In plasma, 1MH appears to be a good internal standard for the analysis of 3MH. It is almost chemically identical and does not occur in significant quantities in normal plasma. As demonstrated in the chromatograms of Fig. 3, 3MH is completely resolved from 1MH; therefore, there is no ion overlap
565
between the two methylated amino acids. However, if the subject under study has consumed a diet containing meat, 1MH could be elevated in the blood and the internal standard used would have to be ( 180,)1MH. Urine cannot be assayed with 1MH as the internal standard because substantial 1MH is present in urine. For this reason, the (“0,)l MH internal standard was used in this study. In addition, an ‘80-labeled 3MH could not be used because the label would be lost upon ionization (ion 238, [M - 851). This method of measuring 3MH in plasma and urine presents many advantages over other methods. First, this GC/MS method can be used for stable isotope analysis, Most methods for analysis of 3MH use highperformance liquid chromatography or cation-exchange chromatography, although both are unsuitable when using stable isotope^.^^' ,-16 Second, this method is fairly rapid and requires few steps. Matthews et a1.” described a sensitive GC/MS assay for 3MH that used an N-acetyl, n-propyl ester (NAP) derivative, which is difficult and time consuming to prepare. He also used a mixture of deuterated 3MHs as an internal standard; this could not be used to measure d3-3MH. Lastly, GC/MS analysis can be completed in 15 min compared with ion-exchange methods, which require longer analysis time. A major benefit of this tracer methodology is its potential use in sheep and pigs, where 3MH is not completely recovered in the urine. The preliminary data presented here suggest that it may be possible to measure the de nouo production of 3MH. Utilizing the decay curve of d,-3MH :3MH ratio from plasma this threecompartment model could estimate the de nouo production rate of 3MH from sheep and pigs. The rate of 3MH production is an important tool in understanding the regulation of muscle protein degradation. The fractional breakdown rate of 1.56% per day demonstrated with this method is similar to values for growing cattle: 1.41% per day: 1.8% per day” and 1.22% per day.” The advantages of this model are: (i) it does not necessitate quantitative urine collection; (ii) error due to the frequency of plasma sampling versus the infrequency of urine collection in the other models; (iii) it measures the total production rate; therefore, it is not dependent on the determination of free or conjugated forms; (iv) it gives information about compartment size and transfer rates; and (v) it does not require restraint of the animals for long periods. With the use of this tracer methodology, it may be possible to perform repeated measures of 3MH production by giving repeated doses of the tracer, if the washout period is sufficient. The model described is very preliminary in nature and is only included to illustrate the potential of the technique. The model is elementary in construction, with only one route for 3MH to leave the system. Identical production rates are obtained when the exit is placed at compartment 2 or compartment 3. The threecompartment model was chosen over a twocompartment model because the three-compartment model resulted in a lower residual sum of squares. There was no advantage of using a four-compartment model over a three-compartment model. The primary form by which 3MH could leave the model, in the calf
566
J. A. RATHMACHER, G. A. LINK, P. J. F L A K O L L A N D S. L. NISSEN
example, is as free 3MH in the urine with a small proportion possibly exiting as free 3MH conjugated with 1-alanine forming baler~ine.~ This is also thought to be the case in rats and humans. Conversely, sheep and pigs excrete only a small proportion of 3MH produced in the urine, with the remaining 3MH accumulating in muscle as balenine.637 In summary, a rapid and precise method for measuring 3MH in plasma and urine has been presented. This analytical method has also been applied to measuring the decay of a stable isotope of 3MH in plasma. Finally,
the decay of the isotope can be compartmentally modeled and appears to be reflective of 3MH production in vivo.
Acknowledgements This work was supported in part by a grant in aid from Eli Lilly & Co. and the Iowa Agriculture and Home Economic Experiment Station. The authors are grateful for the skillful assistance of Rod Berryman, Debra Webb, Brennan Smith, Becky Sperling, Dwayne Bechtol and Jerry Joyce.
REFERENCES 1. P. Johnson, C. 1. Harris and S. V. Perry, Biochern. J . 105, 361
2.
3. 4. 5. 6. 7.
8. 9.
10.
(1967). V. R. Young, S. D. Alex, B. S. Baliga, H. N. Munro and W. Muecke, J. Biol. Chem. 217,3592 (1972). C. L. Long, L. N. Haverberg, V. R. Young, J. M. Kinney, H. N. Munro and J. W. Geiger, Metabolism 24,929 (1975). C. I. Harris and G. Milne, Br. J. Nufr. 45, 41 1 (1981 ). C. I. Harris, G. Milne, G. E. Lobley and G. A. Nicholas, Biochem. SOC.Trans. 5, 706 (1977). C. I. Harris and G. Milne, Comp. Biochem. Physiol. 868(2), 273 (1987). C. I. Harrisand G. Milne, 6r.J. Nutr. 45, 423 (1981). C. 1. Harris and G. Milne, Br. J. Nub. 44, 129 (1980). M. Berman and M. F. Weiss, SAAM Manual, US Department of Health, Education and Welfare Publication No. (NIH) 78-180. US Government Printing Office, Washington, DC (1978). R. C. Boston, P. C. Grief and M. Berman, Comp. Prog. Biomed. 13, 111 (1981).
11. T. P. Mawhinney, R. S. R. Robinett, A. Atalay and M. A. Madson, J. Chromatogr. 358, 231 (1986). 12. D. E. Matthews, J. B. Starren, A. J. Drexler, D. M. Kipnis and D. M. Bier, Anal. Biochem. 110, 308 (1981 ). 13. S. J. Jones, D. L. Starkey, C. R. Calkins and J. D. Crouse, J. Anirn. Sci. 68,2707 (1990). 14. C. I. Harris, G. Milne and R. McDiarrnid, Br. J. Nutr. 57, 467 (1987). 15. H. M. ,H. Van Eijk, N. E. P. Deutz, A. J. M. Wagenmakers and P. 8. Soeters, Clin. Chern. 36, 556 (1990). 16. Y. Tsuruta, K. Kohashi and Y. Ohkura, J. Chromatogr. 224, 105 (1981). 17. P. E. V. Williams, L. Pagliani, G. M. Innes, K. Pennie, C. 1. Harris and P. Garthwaite, 61.J. Nutr. 57,417 (1987). 18. N. Nishizawa, Y. Yoyoda, T. Noguchi, S. Hareyama, H. Itabashi and R. Funabiki, Br. J. Nutr. 42, 247 (1979).
Gas Chromatographic/Mass Spectrometric Analysis of Stable Isotopes of 3-Methylhistidine in Biological Fluids : Application to Plasma Kinetics in Vivot J. A. Rathmacher, G. A. Link, P. J. Flakoll and S. L. Nissent Department of Animal Science, Iowa State University, Ames, Iowa 50011, USA
A simple and rapid method for measuring fmetbylhistidine (3MH) in plasma and urine is described. Internal standard, 1-methylhistidine(IMH), was added to plasma, acidified and absorbed onto cation-excbange columns It was then eluted from columns, dried, and derivatized for gas chromatography/mass spectrometry. A major fragment of 3MH was monitored at 238 u and fmethyl+nethyl-2H3)histidine (d3-3MH) (used for in vivo kinetics) at 241 u, whereas 1MH was monitored at 340 u and eluted 0.5 min later than 3MH. Standard curves for plasma analysis were linear and nanamole amounts of 3MH in plasma were determined with a precision of 3.5%. 3MH was also quantitated in urine; however, because of substantial amounts of IMH, ("0,)IMH was used as the internal standard. Nanamok amounts of 3MH were determined in urine with a precision of 2.7%. Application of the 3MH analytical method was used to develop a kinetic compartmental model by using the stable isotope of 3MH, d33MH. Cattle, like humans, quantitatively excrete 3MH in the urine. A young bovine was injected with 4 3 M H and the enrichment curve in plasma was evaluated in order to obtain a steady-state production rate of 3MH. Tbe decay curve was modeled through the use of NIH-SAAM modeling program. The kinetics of d3-3MH from plasma were adequately described by a three-pool compartmental model. The de now production rate of 3MH estimated in the calf was 665 pmol per day. This corresponded to an estimated fractional turnover rate of 1.56% per day, which was similar to estimates obtained from urine collections. These data suggest that 4-3MH can be used to model 3MH production in viva
INTR ODUC TI O N
The primary sequence of actin and myosin white fibers in skeletal muscle contain the unique amino acid 3methylhistidine (3MH3).' During degradation of muscle proteins, free 3MH is released. Yet, 3MH is not reutilized for protein synthesis because it does not have a specific tRNA.' 3MH is quantitatively excreted in the urine of man, rat, cattle and rabbit;'-' therefore, it is thought to be a marker of skeletal muscle protein breakdown. Urine 3MH estimation of muscle proteolysis depends on quantitative collection and accurate measurement of urinary 3MH. It is assumed that no metabolism of 3MH occurs in uiuo, which is true in most species.6 However, in sheep' and pigs,' 3MH is not quantitatively excreted in urine but is retained in muscle as a dipeptide, balenine.6 Hence, urinary 3MH excretion cannot be used to estimate muscle protein breakdown in these species. An alternative method of quantitating 3MH production in uiuo would be to isotopically measure the de novo production rather than the excretion. An isotopic model for measuring 3MH production would have the
t Journal Paper J-14443 of the Iowa Agriculture and Home Economics Experiment Station, Ames, Iowa 50011, project no. 2928. $ Author to whom correspondence should be addressed. 1052-9306/92/110560-07 $08.50
0 1992 by John Wiley & Sons, Ltd.
advantages of eliminating the need for urine collections ; it also may be applicable to species which do not quantitatively excrete 3MH. The objective of this research was first to develop a simple method to quantitate 3MH and the enrichment of a stable isotope tracer, 3-methyl-(methyl-2H3)histidine (d3-3MH), in plasma through the use of gas chromatography/mass spectrometry (GC/MS). Secondly, the method to quantitate the enrichment of the stable isotope of 3MH was applied to a kinetic model. This model will be used to estimate the de nouo production rate of 3MH in cattle, which are thought to excrete 3MH quantitatively, as do humans. EXPERIMENTAL Materials
3-Methyl-(methyl-2H,)histidine was purchased from MSD Isotopes (99.5% isotopic purity, St Louis, Missouri). 3MH, 1-methylhistidine (lMH), Jack bean type IX urease (61000 units g-'), and Dowex-SOW in the hydrogen form were obtained from Sigma Chemical Co. (St Louis, Missouri). A urease solution was prepared by adding 3.5 mg of the crystalline urease to 100 ml of a 20 mM phosphate buffer (adjusted to pH 7.0). Acetonitrile and N-methyl-N-(t-butyldimethylsilyl) trifluoro- acetamide (MTBSTFA), were purchased from Received 3 April 1992 Revised I8 June I992
3-METHYLHISTIDINE IN PLASMA
Regis Chemical Co. (Morton Grove, Illinois). Ammonium hydroxide, perchloric acid (PCA), hydrochloric acid and filter columns were procured from Fisher Scientific (Fair Lawn, New Jersey). The cation-exchange columns were prepared by adding 3 ml of Dowex-SOW cation exchange resin (50 :50 v/v) in the hydrogen form to filtering columns and washing with 4 x l ml aliquots of 0.01 N HCl. Plastic microinjection sample vials were obtained from Sun Brokers, Inc. (Wilmington, North Carolina).
Internal standards One of the internal standards, (1,l -la0 2 ) 1-methyl- histidine ((1802)1MH),was prepared by exchanging the carboxyl oxygens in l80water: 1 ml of H2180 (96% l 8 0 , Cambridge Isotopes, Woburn, Massachusetts) was added to 50 mg of dry crystalline 1MH. The sample was gassed with HCl for 1-2 min until a pH of much less than 1.0 was achieved. The sample was capped and incubated at 100°C for 30 min, then it was neutralized with 3 N NaOH. The resulting neutral amino acid solution was diluted to 100 ml. Eighty-six per cent of this final product was (180,)1MH (m + 4) as assessed by GC/MS. (180,)1MH (40 pl of a 3 mM stock solution) was used as the internal standard for the analysis of 3MH in urine because a significant quantity of 1MH accumulates in urine. At the normal concentrations of 1MH in urine there was no signal overlap of natural 1MH in the signal of (180,)1MH (ion 344). The '802-labeled standard was quite stable under assay conditions used. (l8O2)1MH had remained stable after being stored at - 70 "C for two years, being treated with perchloric acid or being incubated at 37 "C, at pH 7 for 2 h. LMH was used as an internal standard (50 pl of 3.5 mM stock solution) in the analysis of 3MH from plasma because of its structural similarity to 3MH and its undetectability in the plasma of animals fed meatfree diets.
Plasma 3MH analysis One milliliter of plasma and 50 pl of 1MH internal standard were added to a plastic sample tube (12 ml); 3 ml of 1.5 N PCA was then added. The tube was vortexed and centrifuged at 2300 x g for 15 min at 5°C. The supernatant was poured directly onto a prepared cation-exchange column. After the supernatant drained through the column, it was rinsed with four 1 ml aliquots of 0.01 N HCl and the rinses were discarded. The compounds of interest were eluted from the column with four 1 ml washes of 25% NH,OH into a 20 ml scintillation vial. The eluate containing 3MH was heated to 65°C in an aluminium block under a stream of nitrogen gas. When the eluate appeared dry, two sequential 100 p1 aliquots of methylene chloride were added to azeotrope off any additional water. To derivatize the dried sample and prepare it for GC/MS analysis, 100 pl of acetonitrile and 100 pl MTBSTFA were added. The vial was capped and allowed to incubate overnight at room temperature. The derivatized
561
sample was transferred into a plastic autosampler vial and capped. GC/MS analysis of the samples was accomplished using a Hewlett-Packard gas chromatograph/mass selective detector (Model 5890/ 5970B). The column used was a 25 m x 0.22 mm i.d. x 0.11 pm film thickness, cross-linked methyl silicone gum phase capillary column (HP-1, Hewlett Packard, Avondale, Pennsylvania). The injection port was set at 285 "C, the transfer line was set at 285 "C and the initial oven temperature was set at 50 "C. After splitless injection, the oven was programmed to hold for 0.5 min. It was then ramped (ramp 1) up to 240°C at 50°C min-' and held at 240°C for 4.1 min; a second ramp (ramp 2) to 300°C at 50°C min-' followed and the oven temperature was held at 300°C for 1 min. The retention times were between 6.0 and 7.5 for 3MH and 6.5 and 8.0 min for lMH, depending one the age and length of the column. The major ion fragments for 3MH and d3-3MH were monitored using selected ion monitoring (SIM). 3MH was monitored at 238 u and its stable isotope, d3-3MH, at 241 u, whereas 1MH was detected at 340 u. 3MH in plasma was quantified from a linear peak area standard curve. When d3-3MH was used as tracer, as described below, the enrichment was quantitated in plasma as the ratio of d3-3MH/3MH (ion abundance of 241 :238). The natural background enrichment of d3-3MH/3MH was subtracted from detected d,-3MH ion abundance. The ion overlap of d3-3MH into the signal of 3MH was not backed out because the interference was negligible. Standards were prepared by pipetting known quantities of 3MH from a 50 pM stock solution and 50 pl of 1MH internal standard into plastic sample tubes (12 ml). One milliliter of double-deionized water was added to each standard tube. As the standards had little neutralizing capacity compared to plasma, they were acidified with only 0.25 ml of 1.5 N perchloric acid then treated in the same manner as the plasma sample.
Urine 3MH analysis Unlike plasma, urine contains relatively large amounts of lMH, making 1MH an impractical internal standard. Therefore, di-labeled (I8O2)1MH was used as the internal standard to quantitate 3MH in urine. A urine sample was prepared by transferring 1 ml to a microfuge tube and removing particulate matter by spinning the tube for 3 min. From this tube, 100 pl of urine was pipetted into a plastic sample tube (12 ml). To this 40 p1 of internal standard was added. One milliliter of double-deionized water was also added to the tube, and the contents were acidified with 5 p1 of 3 N HCl. The urine sample was vortexed, poured over a cationexchange column (prepared as already outlined) and allowed to drain. The column was rinsed with four 1 ml aliquots of 0.01 N HCl. Subsequently eluates from four 1 ml aliquots of 25% ammonium hydroxide were collected in a scintillation vial and dried with a stream of nitrogen gas on a heating block at 65 "C. The high level of urea in urine must be removed prior to derivatization of 3MH. To accomplish this, the dry sample was incubated with 1 ml of urease solution for 2 h at 37°C. The same was dried again on the heating block, derivatized
562
J. A. RATHMACHER. G. A. LINK, P. J. FLAKOLL A N D S. L. NlSSEN
with 100 pl acetonitrile and 100 pl MTBSTFA, and incubated overnight at room temperature. The derivatized sample was transferred into an autosampler injection vial, with care being taken to leave precipitated material behind in the scintillation vial. The sample was injected in a Hewlett-Packard gas chromatograph/mass selective detector by using the above-described procedure. ("02)1MH was monitored at M 4 or 344 u. Peak areas of natural 3MH and ("02)1MH were measured in urine using SIM and 3MH was quantitated from a linear peak area standard curve.
+
squares between observed and calculated data points. The model consisted of three compartments or pools of 3MH; the only exit from model was out of pool 1. A steady-state solution was also obtained and initialized by setting the mass of compartment 1 equal to the mean concentration of natural 3MH in the plasma multiplied by the space of distribution in compartment 1. Compartmental masses and flux of 3MH between compartments were also acquired.
RESULTS
Measurement of 3MH production rate in vivo
Application of the 3MH method was accomplished in a castrated male calf (216 kg), which is representative of animals, such as humans and rats, which quantitatively excrete 3MH. The animal was placed in a metabolism cage, catheterized in the external jugular vein on the morning of the study, and injected with a bolus of d,-3MH (30 pmol). After the injection, serial blood samples were taken from 0 to 4320 min. Total urine was also collected for three consecutive 24 h periods. The enrichment of d,-3MH was then quantitated in blood over time. d,-3MH decay from plasma was evaluated using a compartmental model developed through the use of the Simulation, Analysis and Modeling program (SAAM).'.'' The model, illustrated in Fig. 6, was configured by entering the isotope ratio in plasma of the d3-3MH:natural 3MH into compartment 1 over time. The ratio was standardized by dividing the isotope ratio by the dose of tracer in the bolus injection. The model was solved using SAAM and allowed to converge on the observed tracer data, produced transfer coefficients that minimized the weighted total sum of
The electron impact (EI) ionization mass spectra of tertbutyldimethylsilylated 3MH is presented in Fig. 1 and the mass spectra of tert-butyldimethylsilylated 1MH is shown in Fig. 2. The mass spectra of these two tertbutyldimethylsilylated amino acids produced a characteristic fragmentation pattern exhibited by other amino acids." The largest suitable fragments for SIM of 3MH (Fig. 1) were the [M - 571 and the [M - 851 ions 340 and 238 u, respectively. The ion at 238 u was used for quantitation of 3MH as we often found other amino acids or non-specific contaminants producing interference at 340 u. The largest suitable fragment for SIM of lMH, in our assay conditions (Fig. 2), was [M - 571, ion 340. The 302 fragment found in both the mass spectra of 3MH and 1MH does not contain the imidazole ring structure and therefore the 302 fragment was not used. Plasma and standards exhibited similar chromatograms that contained baseline resolution of both the base peak 238 u and the tracer 241 u (Fig. 3). Also, the baseline resolution was obtained for 1MH at 340 u.
y
3- METHYLHlSTlDlNE
NH
3
/si CH3
y3
-c-CH3
c,H3
dHB
.c;
-Si -C-CH3 dH3
dH3
340
54 3 / 400
500
600
V'
ilq 0 .\
?@El
MASS/CHARGE Figure 1. El ionization mass spectra of ?eft-butyldirnethylsilylated3MH derivative.
3-METHYLHISTIDINE I N PLASMA
563
1-METHYLHISTIDINE
qli"
3132
I
W
0
34
I
340
BQO
;3 i 0
180
200
340
560
400
600
700
800
MASSICHARGE Figure 2. El ionization mass spectra of tert-butyldimethylsilylated 1MH derivative.
Further spectral anaIysis of these peaks indicates that standard 3MH ion spectra are identical to that in plasma. Standard curves developed for analysis of 3MH in plasma were linear for 0 pM to 50 pM, with typical
u 4.0ES c
3.0E5
C
2
2.0E5
r
I . WE5 0
6.0
)
.
.
__-
/
1,
-
6.5
7.0
7.5
7.0
7.5
7.8
7.5
Time r m i n . ) I o n 2 4 1 . 10 a m " .
6.0ES 5.0E5
u 4.0E5
i.O
6.5
I a n 340. I0 a m " . 6.0E5j 5.OE5
u 4.0E5 Y
3.0E5 c
2
2.0E5
c 1.WE5 0 G.O
6.5
Figure 3. The El GC/MS selected ion chromatograms from a sample of 1 ml of plasma for 3MH (ion 238, M), d3-3MH (ion 241, M + 3), and 1MH (ion 340, M). 1/200 of the derivatized sample was injected.
regression statistics as follows: observed ratio (238/ 240) = 0.2338 (nmol 3MH), R 2 = 0.9995. Table 1 shows the percentage recovery of 3MH from plasma for two genetic groups of calves. The recovery of added known amounts of 3MH to 1 ml of plasma averaged 96.1 k 2.15% for the two groups of calves. The data in Table 1 illustrate the typical range of variances observed in the 3MH measurement between replicate samples. The coefficient of variation between plasma samples A and B were 5.6% and 1.4% for pooled plasma 1 and 2, respectively. The inter-assay variation of repeated analysis of these plasma samples averaged 5.6%.The assay has previously been sensitive enough to detect 61.25 pmol of 3MH. In a separate study involving five steers given a bolus dose of tracer, the enrichment in plasma was determined. The coefficient of variation from triplicate analyses was determined at two time points when the isotopic enrichment was approximately 1.0 and 0.5. Plasma tracer enrichments from these steers were detected with an average precision of 5.9% at 1 % enriched and an average precision of 10.35% when enrichments were measured at 0.5% (data not shown). 3MH was resolved from baseline in urine (Fig. 4) and exhibited similar chromatographs to those of plasma. Also, baseline resolution was obtained for ("0,)lMH at 344 u. Standard curves developed for analysis of 3MH from urine were linear, with typical regression statistics, ion ratio (238/344) = 0.2804 (nmol 3MH), R2 = 0.9994. Table 2 presents the recovery of 3MH added to calf urine. Known amounts of 3MH were added to 100 pl of urine of two calves. Coefficients of variation between urine samples A and B were 1.25% and 4.21% for calves 1 and 2, respectively. The interassay variation of the 3MH in urine averaged 5.0%, which was based on three repeated analysis of samples A and B of the two calves. The percentage recovery, which averaged 98%, is also presented in Table 2.
J. A. RATHMACHER, G. A. LINK, P. J. FLAKOLL A N D S. L. NISSEN
564
Table 1. Recovery of 3MH added to calf plasma Pooled pladma 2
Pooled plasma 1 Samples'
Additional 3MH added (nmol)
A
0
B
0
3MH found (nmol)
7.74 7.48 8.37 8.61 8.05 0.229 10.23 9.98 12.96 13.33 17.81 17.69
3MH found (nmol)
Percentage recovery
*
x*SE C D E F G H xfSE
2 2 5 5 10 10
Percentagerecovery
6.34 6.11 6.32 6.23 6.25 f 0.046 '8.37 8.04 10.98 10.48 15.45' 14.81
108.9 96.6 98.2 105.6 97.6' 96.4 100.6 f 1.99
105.7 89.7 92.5 84.6 91.9 85.6 91.7 + 2.83
a Internal standard (1MH) was added to 1 ml aliquots of plasma collected from two growing groups of calves differing in frame size; pooled plasma 1 was from large framed calves and pooled plasma 2 was from small framed calves. A 'spike' of 3MH was added to aliquots C, D, E, F, G and H.
Ian 2 3 8 . 1 0
ISOTOPE RATIO -BY DIVIDING BY DOSE
amu
9.0ES
J
3MH
h
6.5
7.0
Tirna ( n r n 1
I n n 3 4 4 . 1 8 amu.
j l ' 0 - 1 ~ ~
l.0E-06
1.2ES 1.0ES
m
t1
B.0E4 6.0E4
i4
0E4
R
I I
UI
-
1 OE-07'
1
0
0 5
1
15
2
2.5
3
35
4
TIME (min x 1000) Tina
lnin
Figure 5. The decay curve of ('HJ3MH : 3MH isotopic ratio in plasma in a calf given a bolus injection of tracer. The symbols represent the observed data points and the line is calculated data produced by SAAM.
I
Figure 4. The El GC/MS selected ion chromatograms from sample of 100 pl of calf urine for 3MH (ion 238, M) and (l80,)1 MH (ion 344, M + 4). 1/200 of the derivatized sample was injected, with ion 238 representing 20.65 nmol of 3MH and ion 344 representing 175 nmol of ('80,)1 MH.
Table 2. Recovery of 3MH added to calf urine Calf 2
Calf 1 Samples.
Additional 3MH added (nmol)
A
0
B
0
x*SE C D E F G
H xiSE
5 5 10 10 20 20
3MH found (nmol)
9.16 9.23 9.35 9.46 9.30 f 0.058 14.87 14.95 20.1 9 19.84 30.34 30.42
Percentage recovery
111.5 112.7 108.8 105.4 105.2 105.6 108.2 f 1.24
3MH found (nmol)
20.65 22.34 19.92 21.23 21.03 f 0.443 25.45 24.87 30.24 30.0 38.53 39.5
Percentagerecovery
88.9 76.7 92.1 88.0 87.5 97.7 88.5 f 2.80
"Internal standard ((1802)1MH) was added to 100 PI aliquots of urine collected from two growing calves. A 'spike' of 3MH was added to aliquots C, D, E, F, G and H.
3-METHYLHISTIDINE IN PLASMA
,/---\,
\
[COMPARTMENT
I\
357
DE NOVO PRODUCTION
COMPARTMENT 3
13809 n r-
Figure 6. This diagram represents the steady-state compartment mass size (pmol) and transfer (pmol min-’) of 3MH between compartments for a calf and the production rate of 3MH into pool 3.
Figure 5 presents the decay curve for d,-3MH : 3MH ratio in plasma obtained from the in uiuo kinetic study. This is intended to document the value of this method for estimating the de nouo production rate of 3MH from a three-pool compartmental model. After the bolus injection of d,-3MH, the tracer was ‘rapidly distributed throughout plasma, with the isotopic enrichment rapidly decreasing during the first 6 h. This was in contrast to the last 96 h, where the rate of isotopic disappearance was significantly less. The model utilized was a three-pool compartmental model with one primary exit from the system. The one exit from the model was from pool 1, a plasma compartment of 3MH. The de nouo production of 3MH was into pool 3, which is thought to be an intramuscular compartment of free 3MH. Figure 6 displays the values obtained at steady state for the mass of the compartments (pmol) and the rate of flow (pmol min- ’) of 3MH between compartments. This method, in addition to estimating the production rate of 3MH per day (660 pmol per day), allows for the calculation of the fractional breakdown rate of skeletal muscle protein, which was 1.56% per day for this example. This calculation compares favorably to that of urinary 3MH, which was 720 pmol per day and 1.69% per day for the urinary production rate and fractional breakdown rate, respectively.
DISCUSSION Histidines are difficult amino acids to derivatize for GC because of the polarity introduced by the imidazole ring.” Therefore, if an internal standard is to be used it must be nearly chemically similar to 3MH to be useful. Norleucine, for example, would be a poor internal standard for 3MH quantitation. In plasma, 1MH appears to be a good internal standard for the analysis of 3MH. It is almost chemically identical and does not occur in significant quantities in normal plasma. As demonstrated in the chromatograms of Fig. 3, 3MH is completely resolved from 1MH; therefore, there is no ion overlap
565
between the two methylated amino acids. However, if the subject under study has consumed a diet containing meat, 1MH could be elevated in the blood and the internal standard used would have to be ( 180,)1MH. Urine cannot be assayed with 1MH as the internal standard because substantial 1MH is present in urine. For this reason, the (“0,)l MH internal standard was used in this study. In addition, an ‘80-labeled 3MH could not be used because the label would be lost upon ionization (ion 238, [M - 851). This method of measuring 3MH in plasma and urine presents many advantages over other methods. First, this GC/MS method can be used for stable isotope analysis, Most methods for analysis of 3MH use highperformance liquid chromatography or cation-exchange chromatography, although both are unsuitable when using stable isotope^.^^' ,-16 Second, this method is fairly rapid and requires few steps. Matthews et a1.” described a sensitive GC/MS assay for 3MH that used an N-acetyl, n-propyl ester (NAP) derivative, which is difficult and time consuming to prepare. He also used a mixture of deuterated 3MHs as an internal standard; this could not be used to measure d3-3MH. Lastly, GC/MS analysis can be completed in 15 min compared with ion-exchange methods, which require longer analysis time. A major benefit of this tracer methodology is its potential use in sheep and pigs, where 3MH is not completely recovered in the urine. The preliminary data presented here suggest that it may be possible to measure the de nouo production of 3MH. Utilizing the decay curve of d,-3MH :3MH ratio from plasma this threecompartment model could estimate the de nouo production rate of 3MH from sheep and pigs. The rate of 3MH production is an important tool in understanding the regulation of muscle protein degradation. The fractional breakdown rate of 1.56% per day demonstrated with this method is similar to values for growing cattle: 1.41% per day: 1.8% per day” and 1.22% per day.” The advantages of this model are: (i) it does not necessitate quantitative urine collection; (ii) error due to the frequency of plasma sampling versus the infrequency of urine collection in the other models; (iii) it measures the total production rate; therefore, it is not dependent on the determination of free or conjugated forms; (iv) it gives information about compartment size and transfer rates; and (v) it does not require restraint of the animals for long periods. With the use of this tracer methodology, it may be possible to perform repeated measures of 3MH production by giving repeated doses of the tracer, if the washout period is sufficient. The model described is very preliminary in nature and is only included to illustrate the potential of the technique. The model is elementary in construction, with only one route for 3MH to leave the system. Identical production rates are obtained when the exit is placed at compartment 2 or compartment 3. The threecompartment model was chosen over a twocompartment model because the three-compartment model resulted in a lower residual sum of squares. There was no advantage of using a four-compartment model over a three-compartment model. The primary form by which 3MH could leave the model, in the calf
566
J. A. RATHMACHER, G. A. LINK, P. J. F L A K O L L A N D S. L. NISSEN
example, is as free 3MH in the urine with a small proportion possibly exiting as free 3MH conjugated with 1-alanine forming baler~ine.~ This is also thought to be the case in rats and humans. Conversely, sheep and pigs excrete only a small proportion of 3MH produced in the urine, with the remaining 3MH accumulating in muscle as balenine.637 In summary, a rapid and precise method for measuring 3MH in plasma and urine has been presented. This analytical method has also been applied to measuring the decay of a stable isotope of 3MH in plasma. Finally,
the decay of the isotope can be compartmentally modeled and appears to be reflective of 3MH production in vivo.
Acknowledgements This work was supported in part by a grant in aid from Eli Lilly & Co. and the Iowa Agriculture and Home Economic Experiment Station. The authors are grateful for the skillful assistance of Rod Berryman, Debra Webb, Brennan Smith, Becky Sperling, Dwayne Bechtol and Jerry Joyce.
REFERENCES 1. P. Johnson, C. 1. Harris and S. V. Perry, Biochern. J . 105, 361
2.
3. 4. 5. 6. 7.
8. 9.
10.
(1967). V. R. Young, S. D. Alex, B. S. Baliga, H. N. Munro and W. Muecke, J. Biol. Chem. 217,3592 (1972). C. L. Long, L. N. Haverberg, V. R. Young, J. M. Kinney, H. N. Munro and J. W. Geiger, Metabolism 24,929 (1975). C. I. Harris and G. Milne, Br. J. Nufr. 45, 41 1 (1981 ). C. I. Harris, G. Milne, G. E. Lobley and G. A. Nicholas, Biochem. SOC.Trans. 5, 706 (1977). C. I. Harris and G. Milne, Comp. Biochem. Physiol. 868(2), 273 (1987). C. I. Harrisand G. Milne, 6r.J. Nutr. 45, 423 (1981). C. 1. Harris and G. Milne, Br. J. Nub. 44, 129 (1980). M. Berman and M. F. Weiss, SAAM Manual, US Department of Health, Education and Welfare Publication No. (NIH) 78-180. US Government Printing Office, Washington, DC (1978). R. C. Boston, P. C. Grief and M. Berman, Comp. Prog. Biomed. 13, 111 (1981).
11. T. P. Mawhinney, R. S. R. Robinett, A. Atalay and M. A. Madson, J. Chromatogr. 358, 231 (1986). 12. D. E. Matthews, J. B. Starren, A. J. Drexler, D. M. Kipnis and D. M. Bier, Anal. Biochem. 110, 308 (1981 ). 13. S. J. Jones, D. L. Starkey, C. R. Calkins and J. D. Crouse, J. Anirn. Sci. 68,2707 (1990). 14. C. I. Harris, G. Milne and R. McDiarrnid, Br. J. Nutr. 57, 467 (1987). 15. H. M. ,H. Van Eijk, N. E. P. Deutz, A. J. M. Wagenmakers and P. 8. Soeters, Clin. Chern. 36, 556 (1990). 16. Y. Tsuruta, K. Kohashi and Y. Ohkura, J. Chromatogr. 224, 105 (1981). 17. P. E. V. Williams, L. Pagliani, G. M. Innes, K. Pennie, C. 1. Harris and P. Garthwaite, 61.J. Nutr. 57,417 (1987). 18. N. Nishizawa, Y. Yoyoda, T. Noguchi, S. Hareyama, H. Itabashi and R. Funabiki, Br. J. Nutr. 42, 247 (1979).
