Journal of Applied Physiology

What does glycolysis make and why is it important?

George A. Brooks

classic references (e.g., Ref. 6) and Wikipedia ( are alike in confusing readers on the process, regulation, and physiological roles of glycolysis. The reference sources assert that glycolysis produces pyruvic acid (i.e., pyruvate and protons), and that under anaerobic conditions, glycolysis produces lactic acid. In their thorough review of the stoichiometry of glycolysis, Robergs et al. (10), among others, including authors of papers in the Journal of Applied Physiology, have forced us to reconsider the matters involved. In this issue of the journal, Marcinek et al. (8) take a major, important step on a renewed journey to understanding.

Why does it matter what glycolysis makes? Minimally, it matters from the standpoint of understanding a fundamental process in biology: it matters because it is essential to know how oxygenation and metabolism affect acid/base chemistry in muscle and blood, and it matters because it is important to understand the energetics of working muscles and other tissues.

Neglecting glycolytic flux from glycogen, consider the classical presentation of glycolysis, asserting that glucose degradation makes pyruvic acid. Then, classic resources go on to state that under anaerobic conditions, glycolysis progresses to make lactic acid. However, while all sources agree that the reduction of pyruvate to lactate by lactate dehydrogenase utilizes NADH and a proton as substrates, not all authors have appreciated that lactate production from pyruvic acid is an alkalinizing reaction that, in effect, buffers acid production from glycolysis. Nonetheless, even if it is understood that the pK of lactic acid is 3.8, and hence almost completely dissociated at physiological pH, it has often been stated that rapid glycolysis in muscle and other tissues results in the accumulation of “lactic acid.”

Diverse observations cause contemporary physiologists to have problems with long-standing beliefs that “anaerobic glycolysis” makes “lactic acid.” For instance, the lactate-to-pyruvate ratio (L/P) in resting muscle and its venous drainage is typically 10 at rest and can rise an order of magnitude during submaximal exercise (3); all the while, ample oxygen exists to fully support cell respiration (9). Hence, the assumption of classic authors that lactate production means oxygen lack proves to be inconsistent with observations that lactate production occurs continuously and under fully aerobic conditions (1). As well, contemporary physiologists have observed that while working human muscles release both lactate and protons, the H+/L stoichiometry of release can range from 1 to 2 in vivo (4).

To reiterate from above, for those interested in muscle energetics, there is more than esoteric interest in knowing the H+/L stoichiometry in muscle and other tissues. In short, if one can relate H+ to L, and if one can measure H+, then one can relate H+ to a glycolytically produced ATP with certainty. The importance of knowing the stoichiometry cannot be understated, but, if the stoichiometry varies, then uncertainty arises and problems arise in determining muscle energetics.

It is clear that resting and working muscles release both lactate and protons, and contemporary knowledge is that the mediators of exchange include not only the monocarboxylate transporters (i.e., MCT1 and MCT4) that are symporters facilitating the movement of protons and monocarboxylate anions (1), but also a variety of other transporters, such as the Na+/H+ (4) and Na+/K+ (2) exchangers, as well as the Na+/HCO3 symport (5) and carbonic anhydrase (11), also play roles in cellular hydrogen ion exchanges. Thus there are many factors that affect pH in muscle, its venous drainage, and the systemic circulation (7).

Now, using a combination of contemporary 31P-magnetic resonance spectroscopy (MRS) and biochemical assays, Marcinek and colleagues (8) provide novel data on an ischemic mouse model system and show a tight 1:1 H+/L over a significant range of ischemia durations, yielding La concentrations up to 25 mM and decrements in pH from 7.0 to 6.7. From chemical analyses, there were no significant changes in ATP content, so H+ production from net ATP degradation could be discounted from the analysis, thus allowing ATP use to be determined from decrements in PCr and corresponding increments in Pi, as measured by 31P-MRS. The use of ischemia ensured that neither protons nor lactate anions could escape detection, and that neither cell H+ nor CO2 production from oxidative phosphorylation could affect H+ or La accounting. In this way, the authors concluded that glycolysis produces lactic acid and that the acidosis from contraction is a lactic acidosis in vivo. For them, this result is important, as they use the phosphate peak separation seen in 31P-MRS to calculate glycolytic ATP production, a key factor in determining muscle energetics.

As important as the results are, Marcinek and colleagues (8) are encouraged to continue their efforts, moving beyond correlational analysis to show that the protons accumulated during muscle ischemia stimulation are indeed from glycolysis and not some other process. Perhaps prelabeling with [3H]glucose and using 1H-MRS could be helpful in this regard? Additionally, perfusing with 13C-labeled substrates in conjunction with 13C-MRS might prove to be useful in identifying the pathways of intramuscular glucose disposal during repeated contractions of graded intensities and durations. Also, perhaps more importantly, the investigators are encouraged to move from studying stopped-flow to free-flow conditions, with gradations in hypoxemia that are more typical of both normal and pathological conditions. In this way, we can know if glycolysis makes lactic acid or lactate, the extent to which the acidosis of exercise is attributable to lactic acidosis, and if lactate and proton accumulations can be used interchangeably in determining the contribution of nonoxidative (“anaerobic”) glycolysis to muscle energetics.


No conflicts of interest, financial or otherwise, are declared by the author.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.