Search

In the present study we describe a simple and fast method to measure the concentration of total free amino acids in very small amounts of biological tissues. The procedure described here is based on the reaction of free amino acids with o-phthaldialdehyde (OPA) in the presence of a reducing agent, beta-mercaptoethanol (MET), to give a complex which can be measured by fluorescence. It is a very rapid process and has the same reliability as the conventional ninhydrin method of Moore and Stein but is about 500 times more sensitive. The sensitivity of the new protocol is such to permit the determination with high reliability of very small amounts of free amino acids at picomole levels, either in a standard amino acid mixture or in biological tissues, without chromatographic separation of the amino acids. It is particularly useful when the amount of the sample is very low, e.g. on a single pituitary or pineal gland of small animals or on single cells, such as oocytes or eggs, as well as single ganglions or axons of marine invertebrates.

In 1939 Kögl and Erxleben [1-4] reported that tumor proteins contain appreciable amounts of D-amino acids, specifically glutamic acid, valine, leucine and lysine, implying that both the initiation and autonomous character of tumors depends on the formation and maintenance of these D-amino acids in the cell proteins. This postulate remained highly controversial for over 10 years, during which time several papers both supporting and refuting this hypothesis were published. The dispute existed almost entirely between Kögl, a vigorous and able protagonist at the University of Utrecht, Netherlands, and an impressive array of equally vigorous and able dissenters in the United Kingdom and Germany. An excellent review of both sides of this controversy was written by Miller in 1950 [5]. After many years and much effort the controversy then seemed to be put to rest. However, more than 40 years later the development of much more refined analytical techniques for the resolution and detection of amino acid enantiomers provided more definitive evidence that D-amino acids are not common to all tumor tissues and probably are not integral to the cancer process. This is not surprising when one considers that a tumor consists of fast-growing cells. Thus, there would not be sufficient time for any L-amino acid to racemize to the D isomer. Some D-amino acids may originate in foods consumed, but it is uncertain whether enzyme systems are able to incorporate D-amino acids into tumor proteins during growth. Nevertheless, if significant levels of D-amino acids were to be found in tumor proteins, the implications could be far-reaching. Confirmation of the presence of D-amino acids at any concentration in tumors would provide new insights into the mechanism for autogenesis and maintenance of tumors.

This is the first report of the presence of free D-amino acids in lumbar and ventricular human cerebrospinal fluid (CSF) of individuals with Alzheimer disease (AD) compared with CSF of normal control subjects and with individuals affected by multiple sclerosis, as an unrelated neurologic disorder. Free D-amino acids are present at significantly higher levels in AD CSF than normal CSF, whereas in the CSF of patients affected by multiple sclerosis, D-amino acids occurs at the same level as in the normal controls. The total D-amino acid content in ventricular CSF was 1.48 times higher in the AD than controls (26.4 vs 17.9 nmol/mL, p = 0.025). The total D-amino acid content was 1.43 times higher in AD lumbar CSF than controls (1.89 vs 1.32 nmol/mL, p = 0.001). D-Aspartate in particular was 2.74 times higher in AD ventricular CSF compared to normal ventricular CSF (3.34 vs 1.22 nmol/mL, p = 0.029). In lumbar CSF, D-aspartate was 1.5 times higher in AD than controls (0.054 vs 0.036 nmol/mL, p = 0.041). Previously we reported that D-amino acids are elevated in AD brain proteins associated with neurofibrillary tangles compared to normal brain proteins (D'Aniello et al., 1992c; Fisher et al., 1992a,b). Thus, the D-amino acids present in CSF may originate from degradation of brain proteins.

A previously described procedure for determining the enantiomeric ratios of amino acids has produced inconsistent results when determining relatively low (less than or equal to 0.110) D/L ratios. The method involves synthesis of diastereomeric N-trifluoroacetyl-L-prolyl-D/L-amino acid ester dipeptides which are resolved by gas chromatography (GC). We have found that triethylamine, which is added to maintain a basic pH during the coupling reaction, racemizes the chiral reagent N-trifluoroacetyl-L-prolyl chloride (TPC). Coupling of partially racemized TPC to D/L-amino acid esters results in the formation of four dipeptides (two pairs of enantiomers) instead of the expected two diastereomeric dipeptides. The enantiomeric dipeptides coelute on an achiral GC column, resulting in erroneous D/L ratios. More accurate D/L ratios are obtained by preparing the volatile N-trifluoroacetyl-D/L-amino acid isopropyl ester derivative which can be separated into its enantiomers on a chiral GC column such as the Chirasil-Val III (registered trademark of Applied Science Laboratories).

Further to explore the functions of carboxypeptidase N (CPN) in vivo, we undertook two studies to find CPN inhibitors of high potency and relatively long duration of action. In each study we examined for inhibition of hydrolysis of [3H]benzoyl-Ala-Arg using pure bovine serum CPN or human serum. In the first such study we synthesized a series of acyl amino acids and acyl di - and tripeptides containing arginine, lysine or both. All proved to be weak inhibitors (Ki = 10(-3) to 10(-4) M). N alpha-carbamoyl-Arg was the strongest: Ki = 3.5 X 10(-5) M. In the second study we prepared S-acyl (thio ester) derivatives of the highly potent CPN inhibitor 2-mercaptomethyl-3-guanidinoethylthiopropionic acid (2-MGP), as certain S-acyl groups markedly increase the duration of captopril, another mercapto-containing compound. Acetyl-, Boc-phenylalanyl-, phenylalanyl-, benzoyl-alanyl-, alanyl-, and Boc-alanyl-2-MGP retained the high potency of 2-MGP in vitro. Although Ala-2-MGP exerted maximum effects in vivo, like those of 2-MGP, the duration of action of Ala-2-MGP was slightly shorter than that of 2-MGP. These results indicate that the mercapto group of 2-MGP can be taken up in some forms of thioester linkage and still remain virtually the full potency of 2-MGP itself. Thus, it appears that a free mercapto function is not essential for the action of 2-MGP.