Gas chromatographic analysis of urinary tyrosine and phenylalanine metabolites in patients with gastrointestinal disorders

Show preview

Hide preview

CLINICA CHIMICA ACTA 289
GAS CHROMATOGRAPHIC ANALYSIS OF URINARY TYROSINE AND
PHENYLALANINE METAIJOLITES IN PATIENTS WITH GASTRO-
INTESTINAL DISORDERS
C. VAN DBli HBIDBN, E. A. K. WAUTBIIS, D. KETTING, M. DURAN AND S. K. WADMAN
Department of Pediatrics. Wilhelmina Children’s Hospital. Nieuwe Gracht 137. Utvecht (The Netherlands)
SUMMARY
Main urinary bacterial metabolites of phenylalanine (total benzoic and phenyl- acetic acids) and of tyrosine (total p-hydroxybenzoic acid and P-hydroxyphenylacetic acid) were determined by gas chromatography in controls and patients with cystic fibrosis of the pancreas, coeliac disease, intestinal resection and unclassified enteritis. In various patients, especially in the untreated coeliacs, high amounts of one or more of the abovementioned metabolites were found. In this paper results in controls and patients are presented and discussed.
INTRODUCTION
In a preceding paper1 we described an abnormal urinary excretion of typical phenolic and phenyl acids in a patient with severely impaired intestinal amino acid resorption. Presumably the greater part of these metabolites are bacterial degradation products of non-absorbed tyrosine and phenylalanine. From tyrosine are thought to originate P-hydroxyphenylacetic acid, p-hydroxybenzoic acid (excreted as p-hydroxy- hippuric acid), ‘p-hydroxyphenylpropionic acid and p-hydroxyphenylacrylic acid; from phenylalanine mainly benzoic and phenylacetic acids (both conjugated) are formed. Small amounts of p-hydroxyphenylpropionic, P-hydroxyphenylacetic, p-hydroxybenzoic, phenylacetic and benzoic acids were found in the patient’s faeces, but the faecal excretion was small compared with the urinary excretion. In the above mentioned paper only one patient has been described. It would be desirable to examine other patients with an impaired intestinal resorption. Here, we describe the urinary excretion of bacterial phenylalanine and tyrosine metabolites in patients with cystic fibrosis, coeliac disease, resection of the small intestine and unclassified diarrhoea.
METHODS
The urine samples were hydrolysed with II N NaOH. Extraction of phenolic and phenyl acids and trimethyl silylation were performed as described earlier*.
Clin. Chim. Acta, 34 (1971) z8g-zg6

I’ig. I. (;as chromntogrnm of the trimcthplsilyl dcrivativcs of some phenyl and phenolic acids and their rctcntion timrs rclnti\rc to phcnylbutyric acid (internal standard). BA: benzoic acid; PAA: phcwylawtlc acid; I’l’:\ : phcnylpropionic acid; PU.1: phenylbutyric acid; PXcrA: phenylacrylic acitl and I’LjZ: phcnyllnctic acltl.
Gas-chromatographic separation was improved by using a high performance 5% SE-52 chromosorb W AW DMCS; IOO-IZO mesh. Temperatures: oven 100-220~ (2’/min), injection port 190’; detector 220’. Gas flows N,: 27 ml/min; H,: 28 ml/min; air: 450 ml/min. Other parameters remained unchanged.
Methodical results A chromatogram is shown in Fig. I. At present a better separation of p-hydroxy-
benzoic from p-hydroxyphenylacetic acid can be obtained than previously’. Rt values relative to phenylbutyric acid (internal standard), are given.
Table I shows the reproducibility of the complete urinary analysis and in Table II recoveries are given. Only the reproducibility of p-hydroxyphenylacrylic acid is out of the range.
RESULTS OF URINARY ANALYSIS
The Fig. 2, 3, 4, and 5 show the excretion of benzoic, phenylacetic, P-hydroxy- benzoic and P-hydroxyphenylacetic acids in 27 controls and 53 patients with gastro-
TABLE 1 REPRODUCIBILITIES OF PHENYL- AND PHENOLIC ACIDS RELATED TO BACTERIAL METABOLISM OF PHENYLALANINE AND TYROSINE
Compounds Urine + addition I (N = 5) Mean Standard deviation
wil mail %
Benzoic acid 940 27 Phenylacetic acid 481 16 Phenylpropionic acid 840 16 Phenylacrylic acid 883 34 p-hydroxyphenylbenzoic acid 720 I2 p-hydroxyphcnylacetic acid 1475 35 p-hydroxyphenylpropionic acid 879 28 0.hydroxyphenylacrylic acid 990 123
2.9 3.4 I.9 3.9
1.7 2.4 3.2
12.4
Clin. Chim. Acta, 34 (1971) 289-296

TYROSINE AND PHENYLALANINE STUDIES 291
TABLE II
RECOVERIES OF PHENYL- AND PHENOLIC ACIDS RELATED TO BACTERIAL METABOLISM OF PHENYL-
ALANINE AND TYROSINE
Compounds Urine + addition I Urine t addition II Recovery of addition (N=5) (N =. 4) Mean: mgll Mean : mg 11 IL Found Found Calculated Mean Range
Benzoic acid Phenylacetic acid Phenylpropionic acid Phenylacrylic acid
p-hydroxyphenylbenzoic acid p-hydroxyphenylacetic acid p-hydroxyphenylpropionic acid p-hydroxyphenylacrylic acid
940 481 840 883
720 1475 879 990
1398 861
1620 1694
1560 2400
1704 2120
‘334 902
1662
‘654
7477 2287 1659 2150
105 IOO-II0 ‘$5 88-102
97 94-100 IO2 93-I I I
106 99-113 ‘OS 97-113 ‘03 97-109 99 73-125
mggcr / T 1 700 -
600 -
500 -
LOO-
300 -
200-
100 -
coNr..oLs
(N= 27)
URINARY EXCRET/ON OF BfNZO/C AC/L,
,NL%-7,NAL UNff ASS/FED RESECT/ON ENrER/T/S
(N=S) (N = ZOj
NORMAL VALUE
Fig. 2. Total benzoic acid excretion in patients with various gastrointestinal disorders: untreated (0) ; after treatment during a period > Q year (0) ; < h year (~3) ; pancreatine substituted (A).
intestinal disorders; the first two metabolitts originate from phenylalanine and the latter two from tyrosine. In order to exclude urinary phenolic acids of exogenous origin as much as possible, the 27 controls (18 children aged o-7 years and 9 adults) were given a special diet, with normal protein content, but no vegetables, fruits, coffee, tea, spices and beverages. The mean values and standard deviations of the above mentioned metabolites, expressed as mg/g creatinine, were as follows: total benzoic acid: 115 * 44; total phenylacetic acid: 116 & 60; total $-hydroxybenzoic acid: 19 f II and p-hydroxyphenylacetic acid: 36 * 23. The mean value of P-hydroxyphenylacetic acid equalled that of phenylacetic acid, whereas the mean value of P-hydroxyphenyl- acetic acid was nearly twice as large as that of p-hydroxybenzoic acid. From a total
Clin. Chim. Acta, 34 (1971) 289-296

292 VAN DER HEIDEN et al.
UR/h!ARb’ EXCRET/ON OF PHENYLACET/C AC/D
501
LO(
3oc
2oc
100
CONTROLS
(N= 27)
P
. 1 .
0 I
* ’ 0
------
.
*
CYSNC CO.61 UC /NTESTfNAL ffNCfA.WF/EO NBROSIS OXEASE RESECUON ENTERfT/S (N= 7) (Nz IO) (N= 7J) /N= 5) (NC 20)
4EAN VALUE
Fig. 3. Total phenylacetic acid excretion in patients with various gastrointestinal disorders: untreated ( 0) ; after treatment during a period > i year ( l ) ; < + year (a) ; pancreatine sub- stituted (A).
UR/NARY EXCRET/ON OF p- OH BEN.?O/C AC/D
6C
(N= /o/(N= N,
/NTEST,NM
.QE.SECTfON
(N=S)
IEAN VALUE

TYROSINE AND PHENYLALANINE STUDIES 293
UMNARY EXCRETION OF R-Off PHENYLACETK AC/D
300-
250 -
200 -
150-
100-
SO-
4
. 0
.
.
.
.
__4__
0 D
.-
0 _.
I:1; . UNCUSSlflEO ENTERITIS
(N 1 20)
2 I SD.
MEAN VALUE
Fig. 5. Excretion of p-hydroxyphenylacetic acid in patients with various gastrointestinal disorders: untreated (0) ; after treatment during a period > 1 year ( l ) ; < & year (a) ; pancreatine sub- stituted (A).
of 53 patients, untreated and treated, 37 showed an increased benzoic acid excretion; 24, 15 and 28 had elevated phenylacetic acid, P-hydroxybenzoic acid and p-hydroxy- phenyl acid excretions respectively.
In 7 patients with cystic fibrosis, 6 receiving pancreatine substitution therapy, p-hydroxyphenylacetic acid was increased while in 4 out of 7 the phenylacetic acid excretion was elevated. Only in 2 out of 7 patients an increased excretion of benzoic and p-hydroxybenzoic acids was observed.
Within the group of patients with coeliac disease, benzoic acid was increased in 9 out of IO untreated cases and in 6 out of II patients given a gluten-free diet for some time. For phenylacetic acid these numbers were 7 out of IO and 2 out of II, for p-hydroxybenzoic acid 6 and I and for p-hydroxyphenylacetic acid 8 and 2 in un- treated and treated patients respectively. In untreated patients very high excretions of phenylacetic acid were found up to 2300 mg/g creatinine, even exceeding benzoic acid.
Varying quantities of these metabolites were found in 5 patients with intestinal resection. The excretion of phenylacetic acid in the urine of one patient with a sub- total intestinal resection was exceptionally high: 4518 mg/g of creatinine. In this patient the excretion of the other compounds were also excessive: 1626,187o and 1066 mg/g creatinine for benzoic acid, p-hydroxybenzoic acid and p-hydroxyphenylacetic acid respectively.
In patients with unclassified enteritis, phenylalanine metabolites were more frequently increased than tyrosine metabolites. In 15 out of 20 patients benzoic acid and in 8 out of 20 phenylacetic acid was abnormal, the excretory level of benzoic acid being highest.
Cl&. Chim. Ada, 34 (1971) 289-296

294 VAN DER HEIDEN et d.
In all patients, the intermediate metabolites of tyrosine: fi-hydroxyphenyl- acrylic acid and ~-l~ydroxypl~enylpropionic acid were occasionally present, but only in very small amounts compared wit11 the main metabolites. The same can be said of the correspondingphenylalanine metabolites: phenylacrylic and phenylpropionic acids.
In Table III, all 53 patients were classified according to the number of increased main mctabolites of tyrosine and phenylalanine. Only 3 patients, treated for coeliac disease, were completely normal. Of the patients with most abnormal excretions (4 main metabolites increased) 3 belonged to the group of untreated coeliac patients, I to the intestinal resection group and 2 had unclassified enteritis. From the table it can be concluded that in the group of coeliac patients treatment resulted in an improve- ment of the intestinal resorption of phenylalanine and tyrosine.
‘f.\l31~11 I I1
CLASSIFICATION Of’ THE PATIENTS ACCORDING TO THE NUMBER OF INCREASED M.IIN METABOLITES
OF PfIENYL.4LANIXE A?JD TYROSINE
Incrcascd = ;a mean - 2 S.U. of the normal population.
Jlazn metabolites zncvrased Number of patients
I’henylalaniwc TylYkW Cystzc Coeliac disease Intestinal Unclassi$ed Total
fib rosis untv. tr. resection enteritis
2 L 3 _ I 2 6 1 2 I 2 _ _ 3 2 I 2 3 _ I 2 8
I I I _ 2 2 2 7 2 0 I I I 3 6 0 z I* _ _ _ _ I
I 0 _ _ 4 _ 7 II 0 I 2 I I _ 4 8
0 0 _ _ 3 _ 3
* Pancreatine substituted.
In a preceding paper’ we described the urinary excretion pattern of phenyl- and phenolic acids in a patient with a severely impaired amino acid absorption. Very high excretions of benzoic, phenylacetic, p-hydroxybenzoic and@-hydroxyphenylacetic acids were found. These metabolites probably originate from bacterial degradation of phenylalanine and tyrosine in the intestinal lumen. In order to confirm our concepts, we examined the excretion of increased amounts of the above mentioned main tyro- sine and phenylalanine metabolites in various patients with gastrointestinal disorders. As expected, the excretory abnormalities appeared to be present in many of the pa- tients investigated. A matter of importance is the definition of the normal ranges. We considered it necessary to establish firmly the variation of the excretions in a large group of healthy individuals. Data given by other authors and obtained by various methods, are widely divergent. In our control group benzoic acid amounted to 115 f- 41 mg/g creatinine. This is much lower than the date of Stein et aL2: 680-1700mg/zq h, of Williams et aL3: 450 mg/g creatinine with a range of 218-506 and of Hoffman4: 340 mg/g creatinine. The range found by Sunderman et a1.5 was 68-680 mg/q h. All the values mentioned were calculated from those given for hippuric acid.
In our controls we found 116 5 60 mg/g creatinine for the excretion of phenyl-
Clin. Chim. Acta, 34 (1971) z8g--296)

TYROSINE AND PHENYLALA~IN~ STUDIES 295
acetic acid, which is also lower than found by Stein%: rag-258 mg/z4 h (calculated from phenylacetylglutamine) but higher than found by Vavich et aL6: 60 + 20 mglg creatinine.
For ~-hydroxybenzoic acid values found in a normal population are not avail- able in the literature. In our control group 19 & II mg/g creatinine was found.
For the excretion of p-hydroxyphenylacetic acid we found 36 5 23 mg/g of creatinine in accordance with Thompsett7: 15-31 “g/a4 h, Williams3: rg mg/g creatinine with a range of 8-42, RugeB: 24.5 mg/g creatinine, range 6.3-46.0, and Horning et al.@: 31.3 mg/z4 h.
In the literature benzoic acid has hardly been correlated with bacterial metabo- lism of phenylalanine in the intestinal lumen. Only Young’* pointed to such a relation.
In normals, benzoic acid may arise from an excess of dietary protein, phenyl- alanine and phenylalanine derived from proteins of (a) gastrointestinal secretory fluid (b) sloughed gastrointestinal mucosal cells or (c) intestinal bacteria undergoing lysis. Possibly, benzoic acid can also be formed by the intestinal bacteria from phenyl- alanine transported from the serosal to the mucosal side of the intestine and finally secreted into the intestinal lumen IX. The same can be said of the origin of phenylacetic acid.
In the patients benzoic acid was elevated more frequently than phenylacetic acid but the excretory levels of the latter were higher than those of the former. In the treated coeliac patients significantly lower excretions of phenylacetic acid were found than in the untreated ones. However, this difference was less pronounced for benzoic acid. Also, a striking difference between the excretory level of p-hydroxyphenylacetic acid in treated and untreated coeliac patients existed. The same can be said of p- hydroxybenzoic acid. In all untreated coeliacs a more or less severe steatorrhea was present, whereas in the treated ones the steatorrhea was greatly improved or normal- ized. Because an increased p-hydroxyphenylacetic was also found by Boscott and Cooke12 in patients with idiopathic steatorrhea, there may be a close relation between the occurrence of steatorrhea and the urinary excretion of abnormally high amounts of bacterial amino acid metabolites. Such a relation has already been demonstrated for urinary indican, a well-known bacterial metabolite of tryptophan13.
All 7 patients with cystic fibrosis showed an increased p-hydroxyphenylacetic acid excretion, despite pancreatine therapy. In this disease ~-I~ydroxypl~eIlylacetic acid has been found increased by several authors 13-16. With respect to the bacterial origin of P-hydroxyphenylacetic acid in cystic fibrosis of the pancreas Gjessinglj demonstrated that this compound decreased in the urine, when neomycin was ad- ministered.
Contrary to P-hydroxyphenylacetic acid in most patients the P-hydrosybenzoic acid was not strikingly abnormal for both excretory level and frequency. This com- pound has been considered to be a product of baterial degradation of benzenoid structures different from phenylalanine and tyrosine17. Shawls described a high urinary excretion of this compound after ingestion of coffee or bananas. In our patients the increased p-hydroxybenzoic acid excretion could not be explained by such a dietary source, because urine collection was started 3 days after a coffee- and fruit-free diet. Booth and ~011.~~ found ~-hydro~ybenzoic acid as an ethereal sulfate in urines of rats. Contrary to our theory, they apparently believed that its formation does not depend on intestinal micro-organisms.
Clin. Chim. Acta, 34 (1971) 289-296

296 VAN I)ER HEIDlrN et al.
convincing evidence for our concepts was obtained from the urinary analysis
of a patient with a sub-total intestinal resection. Very high values were found for
benzoic acid: 1626 mg/g creatinine, phenylacetic acid: 4518 mg/g creatinine, fi-
hydroxybenzoic acid: 1807 mg/g creatinine and p-hydroxyphenylacetic acid: 1066
mg/g creatinine. In this patient the excretion of $-hydroxybenzoic acid exceeded that
of P-hydroxyphenylacetic acid. In most patients P-hydroxyphenylacetic acid ex-
cretion surpassed that of p-hydroxybenzoic acid. Possibly, the type of the excretion
pattern is determined by the bacterial flora, present in the patient’s intestinal lumen.
The pathways of bacterial amino acid catabolism were already discussed in a pre-
ceding paperl. They were deduced from the excretion pattern of the patient described,
characterized by the occurrence of minor intermediate metabolites as p-hydroxy-
phenylpropionic, p-hydroxyphenylacrylic and phenylacrylic acids. These minor
metabolites occurred only occasionally in the patients here described.
REFERENCES
I
2
3 4
5
6
; 9
C. v. D. HEIDEN, S. K. WADMAN, D. KETTINC AND P. K. DE BREE. Clin. Chim. Acta, 31 (1971)
‘33. W. STEIN, A. C. PALADINI, C. H. W. HIRS AND S. MOORE, J. Amer. Chem. Sac., 76 (1954) 2848. C. M. WILLIAMS AND C. C. SWEELEY, J. Clin. Endocrinol. Metab., 21 (1961) 1500. W. S. HOFFMAN, The Biochemistry of Clinical Medicine, Year Book Publishers, Chicago, 1959. P. 350. F. W. SUNDERMAN AND F. BOERNER, Normal Values in Clinical Medicine. Saunders, Phila- phia, 1949. P. 351. J. M. VAVICH AND R. R. HOWELL, J. Lab. C&z. Med., 77 (1971) 159. S. L. TOMPSETT, Clin. Chim. Acta, 3 (1958) 149. W. RUGE, Z. Klin. Chew. Klin. Biochem., 5 (1968) 448. M. G. MORNING, K. L. KNOX, C. E. DALCLIESH AND E. C. HORNING, Anal. Biochem., 17 (1966)
244. 10 D. S. YOUNG, Clin. Chem., 16 (1970) 681. II F. A. JACOBS, Federation PYOC., 24 (1965) 946. 12 R. J. BOSCOTT AND W. T. COOKE, Quart. J. Med., XXIII (91) (1954) 307. ‘3 N. J. GREENBERGER, S. SAEGH AND R. D. RUPPERT, Gastroenterology, 55 (1968) 204. ‘4 R. ROBINSOK, Clin. Chim. Acta, 14 (1966) 166. 15 L. R. GJESSING AND R. LINDEMAS, Lancet, i (1967) 47. 16 I. S. E. GIBBOUS, J. W. T. SEAKINS AND R. S. ERSS~R, Lancet, i (1967) 877.
17 R. J. HOSCOTT AND H. BICKEL, Biochem. J,, 56 (1953) i. 18 K. N. F. SHAW AND J. TREVARTHEN, Nature, 182 (1958) 797. 19 A. W. BOOTH, F. T. Jor~s AND F. DEEDS, J. Riot. Chem., 233 (1958) 280.
Clin. Chim. Acta. 34 (1971) 289-296