Here we show that the three PYCRs make distinct contributions to proline biosynthesis in melanoma cells. PYCR1 contributes primarily to production of proline from glutamate, but under some conditions (no extracellular proline and high ornithine) it can also function along the ornithine route. However, it is unlikely that in these cell line PYCR1 participates in this route in physiological conditions. On the other hand, we cannot exclude the possibility that it may happen in other cell lines. PYCR2 is used exclusively for biosynthesis of proline from glutamate, and PYCRL participates only in production of proline from ornithine. Based on our findings, we propose a working model of proline biosynthesis . The model illustrates the contribution of each PYCR to the two biosynthetic routes to proline in the context of the sub-cellular localization and enzymatic properties of each enzyme.
Model of proline biogenesis in melanoma
One implication of this study, which is illustrated in the working model, is that P5C exists in separate unmixed pools. Although the steady-state level of P5C is too low to be measured directly, the inference of separate P5C pools follows from the results of gene silencing experiments. Thus, knockdown of PYCRL reduces the isotopic enrichment in proline from ornithine, but not from glutamate. However, if P5C, which is the common intermediate along both routes, existed as a single pool, then knockdown of PYCRL would alter the isotopic enrichment in proline in the same way from both precursors. This observation is consistent with the idea that the observed P5C pool separation may reflect its channeling within multi-enzyme complexes that contain the PYCRs, as has been observed in other systems [9,23,24]. The fact that PYCRL functions exclusively in the cytoplasm to convert P5C to proline, raises questions about how P5C is generated in the cytosol. This remains an open issue because OAT is in the mitochondrial matrix, but we cannot exclude the possibility that some form of OAT exists in the cytosol and is coupled to PYCRL.
The PYCRs are functionally tied to proline dehydrogenase (PRODH), which catalyzes the conversion of proline to P5C in the mitochondria. In essence then, PRODH reverses the action of PYCR activity. Superoxide generated by PRODH makes cells more sensitive to stress [25,26,27,28], probably explaining the pro-apoptotic and tumor suppressor function of PRODH . Given their mitochondrial localization, PYCR1 and PYCR2 could potentially partner with PRODH in this process. However, PRODH is generally down-regulated in tumors , and without this reaction the end point of PYCR activity is probably production of proline. Since our results show that PYCR2 is inhibited by proline at the lower end of the physiologic concentration range, PYCR1 is likely to be the dominant enzyme in proline biosynthesis.
Three enzymes involved in the biosynthetic route from glutamate to proline (P5CS, PYCR1 and PYCR2) are up-regulated in melanoma compared to melanocytes. The 13C enrichment in proline from glutamate is also much higher in melanoma cell lines. Together these observations point to a potential role for this biosynthetic route in progression of melanoma. This is not the first association between PYCR and tumors. A recent study found that expression of PYCR1 is up-regulated in prostate cancer . Another study found that PYCR1 is causally linked to growth of breast cancer . The role of PYCR1 in cancer is not entirely clear, but it probably relates to the biosynthesis of proline and potentially production of collagen for the extracellular matrix. The latter possibility may be especially relevant to the tumor extracellular matrix because individuals with mutations in PYCR1 have abnormal collagen fibrils . While it is not up-regulated in all melanoma cell lines, PYCRL could still play an important role in cell growth. PYCRL could couple to the pentose phosphate pathway (PPP) because it is localized in the cytosol, and because it produces NADP, a key cofactor for the PPP. This idea is supported by the fact that addition of extracellular P5C, which is the substrate for PYCRs, is immediately converted to proline in cells and activates the PPP [7,8].
This study also illustrates an important technical point on the silencing of metabolic enzymes with siRNA. In the case of these targets, one cannot always expect the knockdown with siRNA to produce a proportional inhibition of 13C enrichment from a substrate to a product. Indeed, the efficiency of gene knockdown by siRNA is usually 85–95%, so in almost all cases some enzyme is expressed. In fact, others have also observed less than complete inhibition of isotopic enrichment of a product from a substrate when silencing metabolic enzymes with shRNA [2,32]. If the enzyme in question is not the rate-limiting step, then even 5-10% remaining enzyme could support substantial conversion of a substrate into a product. This is likely to be part of the reason that knockdown of P5CS reduces the isotopic enrichment in proline from glutamate by nearly 90%, but knockdown of PYCRs fails to reach this degree. Furthermore, in this instance, where two PYCRs function along a particular path, knockdown of one isozyme is expected to only partially reduce the 13C enrichment in proline from the respective precursor.
In summary, the biogenesis of proline is regulated by three PYCRs that have distinct sub-cellular localization and enzymatic properties. PYCR1 and PYCR2 are localized in the mitochondria, and primarily involved in the conversion of glutamate to proline, and are subject to product inhibition. PYCRL is a cytoplasmic enzyme, exclusively involved in conversion or ornithine to proline. This enzyme is not inhibited by proline. Now that we have established the role of each PYCR in proline biosynthesis, and have illuminated distinctions in their enzymatic properties, it will be possible to probe their role in melanoma and other diseases in an informed manner.