Trompe-l'œil

Blog entry posted by anciendaze, Jun 3, 2015.

This is a term in graphic arts for a trick that fools the eye, a deliberate optical illusion. Famous examples include recesses with statues, or windows with distant vistas, which turn out to be flat images on a wall, or huge, complicated domes or high vaults which turn out to be painted on flat ceilings or low vaults.

The practice is often said to have begun with Renaissance discoveries of true perspective, but this is mistaken. While the perspective in some murals at Pompeii is defective there are some definite examples of true central perspective from the the early Roman Empire.

Literary references suggest that the discovery of these techniques took place about two or three centuries earlier, in Hellenistic culture, but complete examples are hard to find. It would be over 1,500 years before painters again reached the same level of skill in this technique. It now appears that some of the distortions which helped to link the pictures found in Nero's domus aurea with the adjective grotesque were deliberate distortions by people who knew how to create true perspective, which Nero may have regarded as boring by the time they were painted.

In medicine we have just discovered the opposite illusion. Instead of making a flat 2-D object appear to have depth, we have seen how a complicated three-dimensional structure can appear 2-D and blank in the case of lymphatic ducts around the brain, as reported in Nature Immunology.

This lymphatic system is not apparent to the naked eye, and separating the dura from the meninges for examination destroys the system. It was only after researchers looking for leukocytes noticed that these were found positioned as if in linear and branching structures that they tested the adjacent material to see if it was endothelial tissue resembling similar tissue which lines most blood vessels and lymphatic ducts. It is very unusual to find new discoveries in anatomy in the 21st century. Simply looking at the meninges and dura would not have found this -- it tricks the eye. This illustrates an historical bias which can be found all through medicine.

Please note that the leukocytes had no trouble following these paths. They have no eyes or ears, depending instead on what you might call smell, taste and touch (chemotaxis). One aspect of this is their preference for certain laminins on the surface of membranes. This has been observed to guide them along paths across membranes in the lungs which show no distinct markings to human observers. Supplying a different laminin, or blocking the receptor with a free molecule, keeps them away, which can be useful in preventing damage from an overenthusiastic immune response. Lymphatic vessels do not use the integrins, etc. which act in other vascular systems. This makes it relatively easy for leukocytes to tell the difference. Entry to the vessel may require special portals.

I want to point out here that I'm talking about receptors for molecules which are generally part of a membrane instead of floating free. This is where I prefer to talk about cells "feeling" the shape of these molecules rather than undergoing any chemical reaction. Similar activity keeps Schwann cells out of neuromuscular junctions, or glial cells out of synapses in the brain. When this mechanism breaks down, the corresponding cells enter the synaptic cleft, destroying the synapse. This is called synaptic stripping. In reading about activity involving receptors in the brain it helps to keep in mind that the vast majority, though present, are blocked by glial cells sitting on top of them. Glial cells are another group which "get no respect", but formation of new synapses can't take place without their cooperation.

While I've already had feedback from a couple of medical experts indicating this will not have any profound effect on medical research or practice I feel this is more a commentary on the dismal state of medical research and practice than an accurate assessment of implications. For one thing it ought to put an end to a great deal of simplistic nonsense about "the blood-brain barrier". It is true that some molecules do not move easily from peripheral blood to the cells of the brain, but these molecules are not living. Even a small virus is many times the size of a large molecule, and even a small living cell is larger than the vast majority of viruses. This is as if we found that shark nets allowed supertankers to pass through unhindered. Expect more discoveries about this system.

If cells which can be seen under an ordinary light microscope can move out of the brain on such a regular basis that they need a system of established pathways, it follows that these cells, which do not originate inside the brain, must be moving into the brain on a regular basis. These movements are not restricted to exceptional pathologies.

If the lymph nodes where this system drains are more accessible than the inside of the skull this discovery should open a new window on immune activity within the brain which offers a very different view from that found by samples taken from peripheral blood drawn from any old part of the body. Until we tap this source of information we simply will not know how immune activity inside the brain differs from general immune activity.

I have previously written about my frustration with investigations of immune activity which completely ignore evidence that this is highly dynamic, (something which should be apparent to anyone who has experienced the sweats and chills of a fever.) One frustration I don't remember expressing was with reports of increases or decreases in numbers of cells of specific types in peripheral blood without any awareness that they might be missing there because they moved to another physiological compartment.

If a patient has an obvious boil we do not talk about a decrease in leukocytes in peripheral blood because we can see where those lymphocytes are going. They are present in such numbers that we can even drain pus from the local site of infection. We don't often look for leukocytes in cerebrospinal fluid because there are definite serious risks associated with obtaining samples. How often are movements of leukocytes from peripheral blood to the brain to deep cervical lymphatic nodes regarded as decreases in numbers of active leukocytes of a particular class? This kind of quantitative reasoning resembles accounting when a corporation has offshore branches in the Cayman Islands.

Particular inaccessible anatomical structures are even less likely to be investigated. This is behind the wild variety of estimates for incidence of pituitary hypophysitis with associated leukocyte infiltration, ranging from 1 in 10,000,000 to about 1 in 20. That last estimate is for pituitary damage found at autopsy, which may include results of trauma, vascular accidents, etc. If I make a wild guess there might be leukocyte infiltration in about half of those cases found at autopsy, I'm still left with a range of estimates from 1 in 40 to 1 in 10,000,000. A reasonable inference is that we scarcely have a clue about what the immune system is doing to the pituitary. Last time I checked the pituitary was said to be a fairly important organ.

There is a natural bias in medical literature toward pathology visible at the level of gross anatomy. (And, some pathological anatomy is exceptionally gross.) Beyond the possibility of damage to much less accessible organs or structures, there is yet another possibility involving diffuse low-level inflammation spread over tissues like endothelial tissues found throughout the body. Microscopic examination of such tissues does show invasion of apparently healthy tissue by leukocytes in pre-clinical stages of a number of pathologies. This may result in a loss of elasticity in blood vessels or evidence of damage to nerves like nodules of Nageotte. Damage to the pituitary, mentioned above, is an example of a serious endocrine problem which appears to have the same cause. It is also possible invasive leukocytes will cause subtle damage which is even harder to identify. You need to be careful your investigation of pathological processes does not parallel the drunk looking for his keys under a lamppost because the light is better there.

I have more to say about prior ignorance concerning interactions between immunology, endocrinology and neurology, but this is probably enough for one post.
JT1024, garcia, natasa778 and 9 others like this.
anciendaze

About the Author

As the name suggests, I am old and dazed. The avatar illustrates my rule of thumb: "Hang on! This ride isn't over."
  1. anciendaze
    My estimate of incidence of pituitary hypophysitis above involved substantial guesswork. I was using Swedish studies based on autopsies which attempted to adjust for age, and reading between the lines. Not a very reliable procedure. If you do not adjust for age you find that about 17% show pituitary damage, but most are small adenomas. This yields a range of estimates from 1 in 6 to 1 in 10,000,000.

    How often does pituitary damage by leukocytes result in clinical illness? Nobody knows.