Plasmalogen Precursor Design and Development (C101)

Dr. Goodenowe was the first to discover and patent that low blood plasmalogens caused Alzheimer’s and he was the first to design and patent plasmalogen supplements that are effective neuroprotectants.    

To access the FREE seminars with full presentations and videos please visit Dr. Goodenowe’s resource site here. This is the article for seminar C101, Supplements (Series C).

The plasmalogen story began in 2005 when I initially sourced clinical blood and cerebral spinal fluid samples from persons with confirmed cognitive impairment and confirmed healthy cognition. The study included 68 persons with normal cognition, 78 with mild cognitive impairment, 112 persons with moderate cognitive impairment and 66 persons with severe cognitive impairment. Using my patented non-targeted metabolomics technology platform, my laboratory performed a comprehensive analysis of the biochemistry of each of the 324 participants. Thousands of metabolites were detected, and approximately 200 were observed to be statistically altered in patients suffering from cognitive impairment. One biochemical system that had a particularly robust association with not only the presence of dementia but also with the severity of dementia was a class of metabolites called plasmalogens.  To be sure that the observation was real, I obtained additional clinical samples from three separate academic research institutes. One set of samples was collected at the time of death in persons diagnosed by post-mortem autopsy to have definitive dementia of the Alzheimer’s type. Another set of samples were collected from cognitively impaired persons presumed to have Alzheimer’s disease six months to 12 years before they died and who were later confirmed after death by post-mortem examination to have Alzheimer’s disease. Another set of samples was collected from cognitively normal and cognitively impaired subjects in Japan. The results of these analyses confirmed that low levels of plasmalogens were present in blood of persons with cognitive impairment. Based upon the average annualized rate of cognitive decline in Alzheimer’s it was predicted that low plasmalogen levels began approximately seven years prior to clinical symptoms. These results strongly supported the hypothesis that low plasmalogen levels preceded and were part of the causation cascade in Alzheimer’s and cognitive impairment. These results along with a detailed analysis of plasmalogen biochemistry and mechanisms was published in my historic 2007 paper titled “Peripheral Ethanolamine Plasmalogen Deficiency: A Logical Causative Factor in Alzheimer’s Disease and Dementia”, which also included co-authors from Case-Western Reserve University in Cleveland, Sun Research Institute in Arizona, and Osaka University in Japan. The publication is available here Peripheral ethanolamine plasmalogen deficiency: a logical causative factor in Alzheimer’s disease and dementia – PubMed (  


Based upon these discoveries I filed and was granted patents related to the use of plasmalogens for the diagnosis of Alzheimer’s and the risk of Alzheimer’s. From this discovery, I shifted my research focus to understanding everything there was to know about plasmalogens – how they are made, where they are made, how they are degraded, and what biochemical systems are dependent upon them. However, as a research scientist, the very first thing I needed to be able to do was to alter plasmalogen levels in a specific and controlled manner. To do this, I needed to figure out how to manipulate plasmalogen levels either with fully intact plasmalogen species or with plasmalogen precursors. 


The Making of a Plasmalogen 


Plasmalogens are a special class of glycerophospholipid. All glycerophospholipids are based on a very simple three-carbon molecule called glycerol. Each of the three carbons has a hydroxy group (-OH) which it uses to attach to other metabolites. So, glycerol is very much like a biochemical version of a 3-outlet power strip. Scientists label the three positions on the glycerol backbone as sn-1 (top), sn-2 (middle), or sn-3 (bottom). A typical glycerophospholipid will have two fatty acids plugged in at sn-1 and sn-2 and a polar phosphate group at sn-3 (phospho-choline or phospho-ethanolamine). If another fatty acid is plugged in at sn-3, then it is called a tri-acylglycerol. For example, olive oil is a triacylglycerol with oleic acid plugged in at all three positions. The only thing different about plasmalogens is that instead of a fatty acid plugged in at sn-1, there is a fatty alcohol. Unlike other glycerol attachments, this fatty alcohol cannot be unplugged; it is hard-wired. Although plasmalogens closely resemble regular glycerophospholipids, they are actually very different. They are synthesized from entirely different starting materials, and they are synthesized in an entirely different organelle in your cells, the peroxisome. The process of making a plasmalogen is as follows: 


  1. The glycerol backbone that becomes part of the final plasmalogen comes from glucose. Glycolysis is a multi-step process by which the body breaks down the six-carbon alcohol glucose (C6H12O6) into two smaller three-carbon alcohol metabolites, glyceraldehyde-3-phosphate and dihydroxyacetone phosphate (DHAP), which are then either used as building blocks to make other metabolites, or they enter the mitochondria to be used to create ATP energy. One of these metabolites, DHAP, is used as a building block to make plasmalogens. From this metabolite, there are two obligate peroxisomal enzymes required for plasmalogen synthesis. There is no redundancy and no back-up plan. The first enzyme is called DHAP acyltransferase (DHAP-AT). This enzyme puts a temporary fatty acid at the sn-1 position (Acyl-DHAP). The second obligate enzyme is called alkyl-DHAP synthase. This enzyme replaces the sn-1 fatty acid with a fatty alcohol. This step represents the “birth” of a plasmalogen. 


The critical nature of these two steps is best illustrated by rare genetic mutations that affect these steps. DHAP-AT and alkyl-DHAP synthase are not made in the peroxisome. They are transported there from the ribosomes in the cytosol (where proteins are made) using a special transport protein called PEX7. Children born with a mutation in this transport protein cannot make plasmalogens because these enzymes cannot make it to the peroxisome. These children have a disease called Rhizomelic Chondrodysplasia Punctata-Type 1 (RCDP1). Children who are born with normal PEX7 but have a mutation in the first enzyme, DHAP-AT, have RCDP2. Children who are born with normal PEX7 but have a mutation in the second enzyme alkyl-DHAP synthase have a disease called RCDP3. Each of the RCDP mutations results in dramatic plasmalogen deficiencies and severe neurological and skeletal abnormalities. Most children born with any one of these mutations will not live past ten years of age. The severe debilitating nature and early mortality of RCDP exemplifies that plasmalogens are required for a normal life. 


  1. The fatty alcohol that replaces the sn-1 fatty acid in the reaction described above is not derived from the diet. Your body makes it. And just like the two enzymes above, this process is performed in the peroxisome. Three parallel biochemical pathways are involved in making the fatty alcohol that becomes the sn-1 ether. First, the peroxisome makes a short-chain fatty acid by sequentially metabolizing a long-chain fatty acid via peroxisomal β-oxidation. Peroxisomal β-oxidation essentially chops up a fatty acid 2-carbons at a time like a chef cutting a carrot into equal slices. But the peroxisome chef has fat fingers and can only chop it down into a stubby 8-carbon fatty acid. This medium-chain 8-carbon fatty acid is then converted to a medium-chain fatty alcohol. Then, this 8-carbon medium-chain fatty alcohol is built up, two carbons at a time (using the acetyl-CoA generated from β-oxidation), to either a 16- or 18-carbon fatty alcohol. Now, it is ready to be made into a plasmalogen. Accordingly, peroxisomal β-oxidation is a key requirement for plasmalogen synthesis. Peroxisomal activity is stimulated by exercise, especially moderate resistance training.  


Side note: The medium-chain fatty acid created from peroxisomal β-oxidation can also go directly into the mitochondria to be chopped into four acetyl-CoA molecules and used for energy. This is an alternate fatty acid entry point into the mitochondria, which does not use carnitine (carnitine is required to transport the primary fatty acid energy source, palmitic acid (16:0) into the mitochondria). So, the peroxisome is the body’s natural source of the fatty acids found in MCT (medium-chain triglyceride) oil. Peroxisomes and mitochondria are very integrated with each other.  


Plasmalogen biochemistry has been studied for many years, and researchers thought that they had it all figured out. It is supposed to be all downhill from Step 2. This alkyl glycerol metabolite is supposed to leave the peroxisome and enter the regular metabolic system in the endoplasmic reticulum (ER) and behave like other glycerophospholipids. Three more groups of steps, and we are done. From here,  


  1. A fatty acid gets plugged into the sn-2 position. 
  1. A phosphoethanolamine or a phosphocholine gets plugged into the sn-3 position. 
  1. A special desaturase performs the final step, the formation of the vinyl ether bond. 


The Discovery, Invention, and Scientific Evaluation of 1-O-Alkyl-2-Acyl-3-Hydroxy Glycerol Molecules for the Selective Supplementation of Plasmalogens in Laboratory Studies 


When I studied all the research papers that investigated the use of alkylglycerols for supplementing plasmalogens, one consistent observation did not make sense. The three simple alkylglycerols (1-O-alkyl-2-hydroxy-3-hydroxy glycerol) that make up >95% of the sn-1 plasmalogen backbone in the human body have been known and studied since the 1960s. They are called batyl alcohol (18:0 at sn-1), chimyl alcohol (16:0 at sn-1), and selachyl alcohol (18:1 at sn-1). These molecules are the product made by the body after Step 2 in the process described above. However, all the supplementation research using these molecules resulted in only minimal plasmalogen elevations. Also, very high doses of these plasmalogen precursors were needed to see any effect. Biochemistry does not work this way. Feeding a biochemical system with a substrate should result in a dose-dependent increase in the product up to the point that the product of that reaction becomes a substrate for another pathway. Since we know that naturally high levels of plasmalogens occur, we should see elevations to at least these levels and beyond when animals or cell cultures are treated with these precursors. But this is not what scientists were observing. This means that Steps 3-5 (above) may be theoretically correct but not empirically correct. Something else is going on in the synthesis pathway that is preventing these precursors from efficiently elevating plasmalogens. 


The results from the biomarker studies clearly indicated that specific plasmalogens were selectively depleted in Alzheimer’s disease and dementia. My medicinal chemistry and synthetic chemistry training immediately kicked in, and I knew that I had two problems to solve. First, I had to be able to selectively modulate individual plasmalogen species. Second, I had to be able to elevate these plasmalogens in a significant way. The first place I started was on Step 3. What happens if we synthetize precursors, not after Step 2, the mono-alkylglycerols, but after Step 3, the mono-alkyl, 2-acyl-glycerols? So, I had my chemistry team synthesize a whole series of both plasmalogen and non-plasmalogen phospholipid precursors with different fatty acids pre-installed at the sn-2 position. Next, I had my biochemistry team treat various cell culture models that included normal healthy cells and cells with plasmalogen biosynthesis mutations with these different precursors, and voila! My strategy worked! Pre-installing the desired fatty acid at the sn-2 position solved both problems. Not only did this result in selective plasmalogen supplementation, but it was dose-dependent, and significant elevations above normal could be achieved – much more so than the old alkylglycerols. This discovery and invention is the basis of my first series of plasmalogen precursor patents. These observations also meant that Steps 2 and 3 were not isolated biochemistry events but connected and that without the endogenous formation of the sn-2 fatty acid, the rest of the pathway could not proceed efficiently.  


The Discovery, Invention, and Scientific Evaluation of 1-O-alkyl-2,3-Diacylglycerol Molecules for the Selective Supplementation of Plasmalogens  


After proving the biochemical mechanisms and biological activities of selective plasmalogen species supplementation, the next step was to develop an orally bioavailable version of these precursors to see if these laboratory observations would hold up in animals. From a medicinal chemistry perspective, moving from cell culture studies to animal studies has significant practical issues concerning the purity of synthesis, cost of synthesis, stability, and bioavailability. The cellular mechanistic studies that used the 1-O-alkyl-2-acylglycerol species preserved the sn-3 position as the free hydroxy. Making these specialty precursors requires several extra chemistry synthesis and purification steps that are not practical for large-scale studies. There are also numerous other synthetic chemistry reasons not to develop a pure 1-O-alkyl-2-acylglycerol as a final drug (or supplement).  


Since I had a free hydroxy group at sn-3 to work with, and since any carboxylic acid (not just a fatty acid) could be plugged into the glycerol backbone at sn-3, I had my chemistry team synthetize a whole range of molecules with naturally occurring carboxylic acids at the sn-3 position, like alpha-lipoic acid and N-acetylcysteine. Although these carboxylic acids occur naturally, they do not naturally occur attached to the sn-3 position on a 1-O-alkyl-2-acylglycerol. Therefore, although these molecules were a medicinal chemistry parlor trick, they were novel compositions of matter and could be patented. This series of molecules formed the basis of my second set of plasmalogen precursor patents. The best 1-O-alkyl-2,3-diacylglycerol molecule design from a synthesis, stability, and cost perspective was the one that contained lipoic acid at sn-3. The most important sn-2 fatty acid from a disease treatment perspective was DHA. Therefore, the lead molecule for animal studies had DHA at sn-2 and alpha-lipoic acid at sn-3. This molecule was code-named PPI-1011. Since the molecule was designed for oral administration, it did not matter what was on the sn-3 position as long as it had a good safety profile. I already knew from extensive historical research studies on dietary glycerols that as soon as the molecule reached the gut, the gut lipases would cleave off the sn-3 metabolite, and we would be left with the same molecule we studied in the cell culture experiments being absorbed into the blood supply. Now we had our molecule ready for animal testing.  


The first set of animal studies we did was on healthy rabbits. There were several questions that I wanted the team to address.  

  1. What is the time course of a one-time dose? Using a single high dose of 200 mg/kg, we did a time-course to see the conversion of the precursor into the target plasmalogen. It was observed that full conversion occurred at 12 hours and stayed constant for up to 48 hours. Blood levels of the final target plasmalogen were five times higher than baseline. Oral bioavailability was established. 
  1. What was the effect of dose on plasmalogen elevation after chronic treatment? Animals were then treated at either 10 mg/kg/day or 50 mg/kg/day for two weeks. After two weeks, the target plasmalogens levels in the blood were 200% of baseline for the 10 mg/kg dose and 350% of controls at the 50 mg/kg dose. However, retinal levels did not increase after two weeks at the 10 mg/kg dose. Retinal DHA-plasmalogen levels were doubled at the 50 mg/kg dose after two weeks. Modest oral dosing of this molecule class resulted in a highly significant elevation of blood plasmalogens.  
  1. How were the different parts of the precursor incorporated into the naturally circulating plasmalogens? To measure the different parts of the molecule, we used stable isotope labeling of three different parts of the molecule: the fatty alcohol, the glycerol backbone, and the DHA. All three sections of the precursor were incorporated into the endogenous target plasmalogen pool. It was also observed that the labeled DHA was rapidly incorporated into other phospholipids. Maximum labeling was observed at 24 hours.  
  1. We also measured the time course of the sn-3 cleavage of the lipoic acid and observed that this occurred 100% in the gut within three hours. 


These results indicated that modest dosing of 1-O-alkyl-2,3-diacylglycerol plasmalogen precursors significantly elevated blood and tissue levels after only two weeks of administration.  


Dr. Goodenowe explains the relevant research and literature regarding plasmalogen precursors in seminar C101 – Plasmalogen Precursor Design and Development.