点击显示 收起
Food Science and Human Nutrition Department PO Box 110370 Gainesville, FL 32611-0374 E-mail: jfgy{at}ufl.edu
Department of Nutritional Sciences University of California, Berkeley Berkeley, CA 94720-3104 E-mail: bandie{at}socrates.berkeley.edu
Dear Sir,
Baggott raises issues regarding the experimental design and the interpretation of the data presented in our recent paper (1), and he also questions the role of mitochondrial oxidation of serine as a source of cellular one-carbon units. The objective of our paper was to present data showing the feasibility of a serine infusion procedure for investigating one-carbon metabolism involving the acquisition of carbon units from serine and their use in the remethylation of homocysteine. The data presented were from a pilot study conducted to evaluate this basic approach and to provide data on the basis of which the protocol could be modified and optimized for further use. As stated clearly in our paper, "We present a working hypothesis that both cytosolic and mitochondrial compartments yield one-carbon units needed for cellular processes such as homocysteine remethylation." We strongly believe that the data support a growing body of other experimental data leading toward the conclusion that the mitochondrial pathway is indeed a major source of one-carbon units.
It is curious that most of the points raised by Baggott regarding the pros and cons of this protocol were already discussed in our paper. For example, Baggott states that "[2,3,3-2H3]serine is a poor choice of substrate." As we discussed in the paper, we concur that deuterated serine is not the ideal tracer because of the potential for isotopic exchange through several enzymatic reactions in which either serine or tetrahydrofolate-linked one-carbon units react. That is obvious from our presentation of data for plasma and apolipoprotein B-100 in which the presence of [2H1] and [2H2] forms of serine clearly occurred. We believe that these serine isotopomers are formed largely by processes catalyzed by serine hydroxymethyltransferase (SHMT) and 5,10-methylenetetrahydrofolate dehydrogenase as discussed in our paper. We agree that additional reactions proposed by Baggott also could be involved in the formation of [2H1] and [2H2]serine from infused [2H3]serine. A fraction of the infused serine would inevitably serve as a gluconeogenic substrate, but loss of a C-3 deuterium atom from serine during the phosphoenolpyruvate carboxykinase (GTP) reaction can be measured (see below). Racemization of the label due to the tricarboxylic acid cycle is unlikely to have influenced our results because of the large isotopic dilution that would have occurred.
Baggott also reiterates another point made in our paper—that sites of labeling on the serine molecule cannot be determined from gas chromatography–mass spectrometry analysis of this type. Consequently, conclusive identification of the mechanism of this enzyme-catalyzed proton-deuterium exchange cannot be clearly made. We have found that gas chromatography–mass spectrometry monitoring of the labeling patterns of dehydroalanine, which is formed from serine during derivatization in this analysis, provides further information regarding serine labeling. Comparison of the isotopic labeling pattern of serine and methionine yields clear evidence of the source of the methionine methyl group. That comparison of serine and methionine enrichments, as well as the ratios of 2H1 and 2H2 isotopomers of serine and methionine, is the key to the interpretation that mitochondria metabolize serine to form one-carbon units as formate. We agree that kinetic isotope effects complicate the interpretation of data from [2,3,3-2H3]serine infusions. The major value of this method is that it represents an approach in which nutritional, physiologic, or genetic variables affecting one-carbon metabolism can be quantitatively tested. This was illustrated effectively in a study in which vitamin B-6 deficiency caused major changes in one-carbon metabolism in rats (2). As we also discussed in the paper, much of our future use of this procedure will involve 13C-labeled forms of serine.
Baggott then implies that the entire paradigm of the mitochondrial generation of one-carbon units might be erroneous by stating that "the importance of the mitochondrial folate metabolic pathway . . . has been questioned." He bases this contention on a 1993 paper by Yang and MacKenzie (3), and he ignores the large body of evidence from enzymatic and metabolic studies that supports the quantitative importance of the mitochondrial folate-dependent pathway of one-carbon metabolism. The review by Wagner (4) of mammalian folate metabolism strongly supports mitochondrial production of one-carbon units that enter the cytosolic compartment largely as formate, as in the paradigm on which our hypothesis was based, as does a more recent comprehensive review by Cook (5). Indeed, subsequent publications from the MacKenzie laboratory report kinetic and mechanistic data that are consistent with the human mitochondrial NAD-dependent methylenetetrahydrofolate dehydrogenase/methenyltetrahydrofolate cyclohydrolase bifunctional enzyme operating in the direction of 10-formyltetrahydrofolate synthesis (6, 7), leading to the production of formate.
Several lines of metabolic data further support the role of mitochondrial serine metabolism as a source of one-carbon units for cellular metabolism. First, radiolabeling studies clearly showed that the C-3 atom of serine is converted to mitochondrial formate that is available for cytosolic generation of 10-formyltetrahydrofolate (8). An extension of these studies showed that the formation of mitochondrial formate was dependent on the concentrations of ADP and NADP+ and had an absolute requirement for tetrahydrofolate polyglutamate, which is fully consistent with the mitochondrial environment (9). Cell culture studies with cells lacking mitochondrial SHMT provide the second line of metabolic evidence. Chinese hamster ovary (CHO) cells lacking mitochondrial SHMT (glyA cell line) are auxotrophic for glycine and exhibit a rate of serine to glycine flux that is <40 times less than that of normal CHO cells (10). This is strong evidence that the cellular synthesis of glycine, as well as of one-carbon units, from serine strongly requires the mitochondrial folate-dependent pathway. Furthermore, glyA CHO cells transfected with a complementary DNA encoding human mitochondrial SHMT fully alleviated the glycine requirement of the cell line (11). A large body of evidence from yeast cell lines similarly supports the role of mitochondrial production of formate as a source of carbon units for cellular metabolism, as reviewed by Cook (5). Finally, stable-isotope studies with cultured wild-type mammalian cells and mutants defective in mitochondrial folate metabolism conclusively show the essential role of mitochondrial metabolism of serine as a cellular one-carbon source (B Shane, unpublished observations, 2001).
In summary, we agree that human one-carbon metabolism and its subcellular compartmentalization is a very complex system. As discussed in our paper, we plan to use [3-13C]serine in many of our future studies to alleviate some of the technical complexities associated with [2H3]serine. Unfortunately, infusions of [3–13C]serine will allow measurement of only aggregate cellular processing of serine and will provide no evidence of subcellular compartmentalization. Again, the main points of our paper were to report data that show the feasibility of serine infusions for measuring human one-carbon metabolic processes and to show that labeling patterns resulting from homocysteine remethylation are fully consistent with those anticipated from mitochondrial and cytosolic production of one-carbon units by distinct metabolic processes involving cytosolic and mitochondrial SHMT isozymes. Additional isotopic studies will involve variations of this isotopic protocol that will aid in clarifying other aspects of one-carbon metabolism and subcellular compartmentalization of these vital metabolic processes.
REFERENCES