Could MSM be a useful tracer for determining CSF flux in the human brain?

 A few years ago I was involved in a project to develop a better chemical shift reference for in vivo MR spectroscopy (Kaiser et al. 2020).  As often happens in science, life, logistics and money conspired to change the directions of those involved and this project got put on the shelf to gather dust. We no longer have either the people or the capabilities to pursue it further. Perhaps someone else would be able to take it on and see whether there are more uses than as a chemical shift reference.

One of the angles we were considering in 2020 was the possibility of using MSM as a tracer for measuring CSF flux in the brain. Various approaches have been developed using MRI, but they are all rather difficult. One involves an intrathecal injection of a gadolinium contrast agent and then looking for signal losses depicting where the Gd contrast diffuses to (Iliff et al. 2013). Negative contrast is always a complication for MRI because signal voids often arise from imperfections in the magnetic field. Another method uses an arterial spin label (ASL) and long post-labeling delays to assess the amount of water passing from the vascular compartment to the tissue compartment, i.e. through the blood-brain barrier, as an index of what is assumed to represent the inflowing part of the glymphatic system (Gregori et al. 2013, Ohene et al. 2019). These methods are low sensitivity and highly prone to motion. A third approach uses low diffusion-weighted imaging to try to differentiate CSF from other water compartments (Harrison et al. 2018). But again the method is inherently sensitive to bulk motion and it's not entirely clear to me how well the signals represent the CSF to interstitial fluid flux versus other microscopic compartments. So, would MSM offer simultaneously positive contrast and improved sensitivity? And would its clearance give an indication of the CSF flux through brain tissue?

MSM is methysulfonylmethane, the trade name for what a chemist would call DMSO2. It is labeled by the FDA as GRAS: "generally regarded as safe." As such, there are few regulations for its use and so you can find it in everything from dog food to ointments for a bad knee. You may well be consuming it and not have a clue. But the good news is you can buy pills of MSM for your experiments. There's no special permission needed, you can get these at your local pharmacy. (A word of caution: the amount listed on the package may not match what is actually in the pills! Do your own assay!) Then, once you've got this past your IRB, you can dose subjects with acute or chronic doses and see what happens to the MSM level in the brain.

MSM is a small, polar molecule which probably distributes throughout biological tissues with approximately the same concentration profile as water. The more water content in the tissue, the higher the MSM concentration is likely to be after a few hours. But this is a guess. What we do know is that entry into the brain is rapid. We can see MSM in a brain spectrum within 10 minutes following an oral dose. The MSM signal then remains fairly stable for several hours, which is a property we wanted for our chemical shift reference. 

But what is driving the clearance rate? In our early tests, we observed a half life in normal brain of about 3 days. This was for a single acute dose. In later tests (not included in the 2020 paper) we saw about the same washout time for a single 6 g dose as for a single 2 g dose. We also had a subject (me!) take a 1 g dose every day for 30 days to ensure steady state concentration, then observed the washout. Again, a half life of about 3 days. 

 

For clearance, we assume the MSM partitions and clears down its concentration gradient. Presumably the MSM distributes into the brain via the blood. Once we stop giving new oral MSM the blood concentration falls to near zero, and presumably clearance of the MSM in the body then occurs via the kidneys. If the routes out of brain tissue include the blood and perhaps CSF clearance, then what matters is the concentration gradients between brain tissue and blood and, perhaps, brain tissue and CSF.

This is where the idea of using MSM as a CSF flux (glymphatic system) tracer comes in. If the half life is around 3 days in normal brain, does the rate of clearance change with sleep deprivation, bouts of vigorous exercise or other challenges to an individual? What about differences between individuals? Do older subjects clear MSM more slowly than younger subjects on average? Women faster than men? Is the density of aquaporin channels a prime determinant of the clearance rate from brain, or is MSM able to diffuse across all membranes with approximately the same rate? And is CSF flux through brain tissue an important determinant of the clearance rate, or incidental to it? We were never able to test these ideas. 

As a practical matter, MSM can be observed easily in a 1H MR spectrum. Its chemical shift of pi (3.142 ppm) and sharp line makes it easy to fit separately from brain metabolites. We also never tested the ability of chemical shift imaging (CSI) to observe MSM, but there's every reason to think that a CSI method which can reliably image the NAA, creatine and choline singlet peaks will be able to map MSM perfectly well, too.

So, there you have it. A free idea for someone to explore and perhaps exploit for the purposes of assessing CSF clearance, sleep, dementias and so on.

_______________