Dr. Young

[Jeremy]

I hope I'm not asking a question that you've answered before but in the spinal cord below the injury site when the axons die back does the myelin surrounding them remain or does it degenerate over time?

[Wise Young]

It is a great question. In animals and in humans, one sees myelin remaining for a time after the axons have died. It is eerie... there are all these myelin cylinders with no axons inside them. However, over time, especially if many axons have been destroyed, the oligodendroglial cells (that provide the myelin) will undergo apoptosis of programmed cell death. The degenerating tracts become filled with macrophages that then clean up all the debris. This is called Wallerian degeneration and occurs over a period of weeks or even months after injury.

Wise.

[Jeremy]

I have a follow up question - when this has all taken place, does the spinal cord below the injury now provide a sutable enviroment for new axons to grow in when a regeneration therapy comes about? Or will Nogo still be a factor?

[Wise Young]

The dogma use to be that axons cannot grow. Now, the dogma is that axons cannot grow in white matter due to Nogo. However, there is plenty of evidence now from many laboratories (Jerry Silver's, in particular) showing that axons will grow in white matter despite Nogo. So, attention is now shifting to the possibility that axons require another signal in order to grow in white matter. Marie Filbin has shown that axons that have high levels of an intracellular messenger called cAMP will grow in white matter even in the presence of Nogo. So, there is interest in using a drug called Rolipram which increases cAMP levels in axons and, in tissue culture so far, can remarkably increase axonal growth in the presence of Nogo. Several laboratories are now testing Rolipram in animals and we should expect some studies to come out soon. In the meantime, Rolipram has already started clinical trials in MS (see the Trials Forum) and this should help propel Rolipram into spinal cord injury clinical trials, should the animal studies show positive results.

Now, I am going to say something heretical. After several years of looking at chronically injured spinal cords and axons, I believe that some axons may grow across the injury site spontaneously. This is very hard to show in contusion models of spinal cord injury because everybody will claim that the contusion did not eliminate all axons and that the axons that we see were surviving axons rather than regenerated axons.

When Cheng did his peripheral nerve bridging experiments, he deliberately connected the peripheral nerves from white matter to gray matter because people believed that Nogo was not present in gray matter. Indeed, his experiments did show that the axons grew from white matter to gray matter and then grew in gray matter to the lower spinal cord. Unfortunately, nobody has been able to replicate the results that he has had. This does not mean that his idea does not work... the surgery is very difficult to do and I think that it must be done just right or else the axons will stop at the injury site.

Jerry Silver and others have now shown that a chemical called chondroitin-6-sulfate proteoglycan (CSPG) stops axonal growth. This chemical is produced by proliferating glial cells and may also comes into the central nervous system from blood. They found that if they inject neurons into the spinal cord using a very fine needle that causes little or no tissue damage at the site of injection, the neurons will send axons out that will grow long distances in the spinal cord. On the other hand, if they damage the injection site and some material from the blood were to enter the site, the neurons stop growing.

Fortunately, there may be a solution to this problem as well. Recently, several laboratories have reported that an enzyme called chondroitinase will break down CSPG in the tissue. When this enzyme (which incidentally is made by bacteria so that they can invade through tissues) is applied to the brain or spinal cord, it allows axons to grow across areas of damage. I have no doubt that there will also be treatments in the future that would simply block the axonal receptors to CSPG (if the axons don't sense the CSPG, they will not be stopped by it).

In answer to your question of what happens to the distal spinal cord... I must expand the question to the proximal spinal cord as well. There are changes in both the proximal and distal spinal cord after injury. To date, most of the evidence suggest that the changes there are not sufficient to prevent functional regeneration if the axons can be enticed to grow across the injury site and through the spinal cord for some distance.

Note that the cell bodies of neurons generally remain intact in the spinal cord above and below the injury site. This is true of a majority of people. However, many people have injuries to the cervical spinal cord and the lumbosacral spinal cord. In such cases, some of the neurons may be gone and it may be necessary to replace these neurons. That is one of the reasons why I think embyronic stem cell research is so important because several studies now suggest that these cells can replace neurons in the brain and spinal cord.

Finally, you might be saying to yourself that this is too much, that there are too many obstacles to functional regeneration. In answer to that, I want to point out that one does not protect, remyelinate, or regenerate many axons to get functional recovery. That is why so many papers have been published in recent years showing animals that have recovered walking after treatment.

The sheer variety of therapies that have been reported to improve regeneration and recovery is absolutely stunning. They range from IN-1, inosine, and chondroitinases to activated macrophages and olfactory ensheathing glia. Probably none of these treatments regenerate more than a small percentage of axons in the spinal cord but they are resulting in functional improvements.

I believe that very few axons are required to restore function in people with spinal cord injury. What is the basis for this belief? Three things. First, people with incomplete spinal cord injuries generally recover substantially. Comparisons of the spinal cords between complete and incomplete injuries do not show that much difference in axons that cross the injury site. Second, I have seen people walk out of the hospital after surgery that has destroyed close to 90% of their spinal cords. For example, people who have slow-growing tumors of the spinal cord frequently have destruction of 90% of their spinal cord on MRI scans and also on inspection in the operating room. However, many of these people can walk. Third, many people have substantial function despite horrendous injuries to the spinal cord. For example, it is not easy to tell whether a person has a "complete" spinal cord injury versus a person who has "incomplete" spinal cord injury just on the basis of an MRI scan. In fact, I don't think that one can tell because the margin of difference between a walking person and one who is completely paralyzed is too small to be detected on MRI.

Sorry about such a long posting. My son and daughter use to tease me that they were reluctant to ask me any question about science because I always told them more than they wanted to know.

Wise.

[Steven Edwards]

Actually, good doctor, I wish their were more posts like the one below. It helped to give a good overview of some of the major studies currently underway.

A few questions, though. In the studies where very few axons crossed the site of injury but still showed functional recovery, were those axons remyelinated? Can you actually see axons on an MRI? If so, how do you tell the difference between good and bad axons?

-Steven

[Wise Young]

Modern MRI machines have a resolution of about ±1 mm. Let me explain further. Axons are about 1-2 micra thick within their myelin and 3-4 micro thick when myelinated. As you know, one mm has 1000 micra. So, MRI is not suitable for viewing individual axons.

What can an MRI show? MRI is quite quite good at visualizing myelinated tissue, because myelin is mostly fat compared to gray matter which has more water content. That is why MRI is very good at showing myelin plaques (areas of demyelination).

MRI is also pretty good at detecting acute injury. Magnetic resonance images of the acutely injured spinal cord usually show bright signal intensity at the injury site, due to the increased water content at the injury site (this is due to loss of cells and edema).

Cysts that develop in the spinal cord can be seen with MRI because the cysts are filled with water. If the cysts exceed 1 mm in depth, one can usually see them on MRI. There are several variants of MRI images (called T1 and T2-weighted images) that show cysts and edema particularly well.

Finally, MRI can also be used to detect blood flow in the spinal cord. This requires a special MRI machine with special pulsing hardware and software to detect movement. Called functional MRI or fMRI, they are useful for showing activity in the brain and spinal cord.

What can't MRI show? First, MRI cannot detect tissue features that are much smaller than a 1 mm because signals from surrounding tissues will interfere. Second, MRI's don't show bone particularly well since MRI is not sensitive to calcium. Therefore, to see bone, clinicians usually use CT scans which uses x-rays. Third, MRI uses powerful magnets to polarize water molecules so that they provide a coherent resonance signal when stimulated by radiofrequency pulses. Any ferromagnetic substance in the field will interfere with the magnetic field and that is the main reason MRI cannot be used in people with steel Harrington rods. However, most surgeons use titanium rods and plates these days and these do not interfere with MRI.

---------

Regarding myelination of axons... under a microscope, it is very easy to see a myelinated and non-myelinated axon. The latter is in fact is at the limits of optical microscopy under very high magnification (400-1000x) on a microscope. Injury damages oligodendroglia which myelinate the axons. Thus, most of the axons that survive at the injury become demyelinated at the beginning. They tend to remyelinate and this takes several weeks. Very often, the myelination is incomplete or abnormal. Called dysmyelination, abnormal myelination is hard to see even on a light microscope and most scientists who study myelination of axons use electron microscopes that have magnifications of 5000-50,000x.

Andrew Blight did a classic study in 1985, characterizing the effect of contusion on axons at the injury site. Published in Neuroscience, this study showed that a majority of the surviving axons in cats have abnormal myelination and this then was what impelled him to try 4-AP for chronic spinal cord injury. This was the study that strongly suggested that some animals who had sufficient axons to walk did not recovery walking because of demyelination or dysmyelination.

I should also point out that regenerated axons are naked and therefore need myelination. By the way, rats and mice (maybe because of the small size of their spinal cord and oligodendroglial precursors do not have to travel far in order to myelinate) usually show pretty good myelination of surviving axons at the injury site. It may be the reason why regenerative therapies of rats can restore function even though the therapies do not stimulate remyelination. It is possible that humans will require concomitant remyelination therapies if regenerative therapies are to restore function.

Wise.