The Neurodegenerative and Neuroprotective Effects of Ketamine
It is well known that the NMDA receptor, a primary mediator of excitatory transmission, plays a role in both neuronal generation and excitotoxicity. Since many states of neurological disease are associated with NMDA receptors and because NMDA blockers are currently used in a variety of medical treatments, there has been a great deal of research on exogenous NMDA inhibiting agents. The extant literature clarifies that:
- When NMDA antagonists are administered in high enough doses to mature brain cells, sufficient disruption of homeostatic glutamate transmission results in irreversible excitotoxic degeneration via necrosis.
- When NMDA antagonists are administered during synaptogenesis, disruption of important glutamate signaling causes significant neuronal apoptosis and long term brain deficit. This result has implications for the use of anesthetics on neonatal surgery and the use of PCP, ketamine or ethanol by pregnant mothers.
- When NMDA antagonists are administered in significantly lower doses than suggested above, they promote robust and rapid neurogeneration. It is this action which reverses the neuronal atrophy and loss of connectivity that characterizes psychiatric disorders such as depression and bipolar disorder.
Olney, Labruyere and Price (1989) determined that progressive and single doses of MK-801 (0.05-0.1mg/kg sc) and PCP (0.5-5.0mg/kg sc) given to healthy adult rats produced either no morphological change or a dose dependent vacuolar reaction in large population of diverse neurons. For those dosing scenarios in which vacuolization occurred, an increase was seen from 2 to 12 hours followed by an almost complete spontaneous resolution between 18 by 24 hours. Induction doses of vacuolization occurred at 0.18mg/kg sc MK-801 and 2.83 mg/kg sc PCP.[i]
In 1993, Fix, Horn, Wightman et al determined that, while the previously observed vacuolization reactions resolved, at higher doses: 5 or 10mg/kg sc MK-801, but not at the lower dose of 1 mg/kg, a delayed progression to necrosis occurred in a similarly distributed but smaller number of neurons.[ii] In 1995, Farber, Wozniak Price et al extended the investigation to determine whether the neurotoxicity observed in prior studies was age-specific. Unexpectedly, they found that neither fetal nor infant rats suffered the same NMDA antagonized injury. Increasing sensitivity to the neurotoxic syndrome begins with adolescence and plateaus in adulthood. During this study, 0.5 and 1.0 month old rats were not even given the lowest dose of MK-801 (0.5mg/kg) since it had previously been determined that they were completely insensitive to neurotoxicity at higher doses.[iii]
It might seem that the documented neurotoxicity that results from NMDA blockade brings important clinical consequences. However it is important to grasp the significance of the gap that exists between the degree of NMDA antagonism used in these studies and that which is currently used for sedation, pain management, and reversal of status epilepticus and the symptoms of mood disorder.
Most studies which investigate the neurodegenerative effect of NMDA inhibition use MK-801 or PCP as the antagonist. Ketamine is usually mentioned as an “also ran” while not actually used in the dosing protocols. While MK-801, PCP and ketamine are all non-competitive open-channel NMDA ligands, they have very different affinities for the binding site.[iv],[v] This typically leaves one to extrapolate as to the commensurate ED50 of ketamine to achieve the investigated neurotoxic events.
There are two studies which provide a basis for that comparison. The study by Olney et al which documented vacuolization also reported that, at the doses tested (5, 10, 20 and 40mg/kg sc), ketamine induced vacuolization at 40mg/kg sc. but did not induce vacuolization at the lower doses.[vi] This means that the ED50 ketamine is over 200 times the 0.18 mg/kg sc induction dose by MK-801. Even if one presumes that induction by ketamine could have occurred at some level between the 20 and 40 mg/kg. and backs down the induction point to 21mg/kg ketamine; it is still 100 times the MK-801 dose.
Another strong indicator of commensurate dosing can be found in the comparative study of NMDA ligands in which Smith et al (2011) use doses of 0.025, 0.05 and 0.1 mg/kg MK-801, 1,2 and 3mg/kg PCP and 2.5, 5 and 10mg/kg (S)-(+)Ketamine to achieve similar effects.[vii] Again, the order of magnitude between Mk-810 doses and ketamine doses is 100 times.
So, if necrosis occurs at doses of 5 or 10 mg/kg sc MK-801, but not at 1mg/kg sc MI-801,[viii] it is reasonable to extrapolate that necrosis would occur at 500 mg/kg sc ketamine but not at 100mg/kg sc ketamine.
Olney, Wozniak, Jevtovic-Todorovic et al, (2002) clarify that there are strikingly different ultrastructural changes, both in type and sequence, between the excitotoxic cell death described above and the natural apoptotic deletion of neurons which occurs as part of normal CNS development .[ix] These differences make the nature of cell degeneration readily distinguishable.
Ikonomidou et al (1999) assert that the period of synaptogenesis, or rapid brain growth, coincides with a period of glutamate hypersensitivity.[x] This makes neurons susceptible to injury from NMDA blockade. A large number of studies describe significant depletion of neurons in various brain areas when ketamine or another NMDA antagonist is introduced during the synaptogenic window.[xi],[xii],[xiii],[xiv] The result of the insult is permanent behavioral and/or cognitive impairment due to widespread cell death which resembles apoptosis. Importantly, the window of vulnerability closes with the end of synaptogenesis: exogenous agents that would otherwise spur apoptosis are no longer effective.[xv]
Olney et al (2002) delineate the period of synaptogenesis for the rat as occurring from 1 day pre-birth to approximately 14 days post-birth.[xvi] Most studies which examine apoptotic cell death through NMDA blockade do so using fetal or neonatal rats. 7day old rat pups represent a benchmark in the synaptic development of that species and are commonly used. [xvii],[xviii],[xix],[xx],[xxi] Yet each day of the vulnerable window brings profound variations of effect.
In order to translate this information to human clinical relevance, it is necessary to define the time of human synaptogenesis. Cross-species neural development does not progress equally and there is not a significant amount of research available on the subject.
Relying upon a frequently cited 1979 article by Dobbing and Sands, Olney et al equate the period of synaptogenesis in the rat to a period starting in the last 3 months of human gestation and extending through the first several postnatal years.[xxii]
Jomijn, Hofman and Gramsbergen (1991) equate the 12-13 day old rat pup to a full term human infant.[xxiii]
A 2006 review by Watson et al compares neurogenesis across species. Their review includes studies which, in general, support a similar window of vulnerability as Dobbing and Sands. Those studies purport that, for the human, peak overall brain synaptogenesis occurs between 34 weeks gestation and 2 to 3 years after birth and that synaptogenesis per se continues until app. the age of 3.5. Of particular note for this discussion, Watson cites that the peak developmental concentration of NMDA receptors in the full-term human compares to that of the rat at post natal days 6–9.
Clancy et al (2007) created a data-base driven website which uses statistical-based algorithms to integrate hundreds of empirically derived neural developmental events across 10 mammalian species.[xxiv] According to the interactive translator, an event in a 7-day rat pup translates to post-conception (PC) day 156.4 in human cortical events, PC Day 114 for human limbic events and PC Day 123.2 for general events.[xxv] The translation would place the 14 day rat pup, the neuronal developmental age at which Dobbing and Sand, and by extension Olney et al, suggest the window of vulnerability closes, at 152 to 201 PC Days for a human:18 a significantly earlier concluding date than suggested in previous studies.
We can infer from the apoptotic studies and the cross species developmental comparisons that the period during which a human is vulnerable to apoptotic insult from exogenous NMDA blockade, ranges from a minimum of 152 day post conception to a maximum of 3.5 years.
Additionally, one needs to consider another component of the vulnerability. As in the discussion of vacuolization and necrosis above: degree of exposure matters. Studies have clearly identified that apoptosis is dose-dependent. Hyang, Liu, and Jin et al, (2012) administered doses of 25, 50 and 75 mg/kg ip ketamine to 7 day old rat pups and found evidence of hippocampal apoptosis and persistent learning and memory impairment only at the 75 mg/kg dose.[xxvi] A 2011 study by Liu, Paul, Ali and Wang examined the effects of single and multiple 5, 10 and 20 mg/kg sc ketamine on several major brain regions including striatum, hippocampus, thalamus and amygdala. They determined that while six administrations of 20 mg/kg sc ketamine resulted in significant apoptosis in the frontal cortex, lower doses were not sufficient to cause neurodegeneration. Similarly, in a study using two week old neurons differentiated from human embryonic stems cells, Bosnajak, Yan, Canfield et al (2012) conclude that ketamine time and dose-dependently lead to neurodegeneration but only at supraclinical concentrations.[xxvii]
Recent lines of investigation of depression and bipolar disorder point to dysfunction in the glutamatergic system with an emphasis on the NMDA receptor[xxviii],[xxix] The past 6 years have produced a substantial amount of literature regarding the clinical and neurological effects of low-dose ketamine. Several investigative groups have demonstrated that ketamine antagonism of the NMDA receptor provides rapid improvement of mood and decreased suicide ideation for patients who are treatment resistant and struggle with major depression or bipolar disorder.[xxx],[xxxi],[xxxii],[xxxiii],[xxxiv] While the mechanisms for this effect are not clearly understood, evidence of the effect is, at this point, well documented.
The NMDA receptor is extremely complex with a variety of means and conditions by which it mediates inter- and intra- cellular glutamate (Glu) transmission.[xxxv],[xxxvi],[xxxvii],[xxxviii] It is known that activity stimulated by calcium influx through the NMDA receptor (NMDAr) is specific to the molecular environment in which it is located[xxxix],[xl].[xli] While there is still much more detail to learn about this, recent studies point to the location of the receptor on the neuron to explain its paradoxical role in both neuroprotective and excitotoxic cellular events.
The work of Giles Hardingham and colleagues found that NMDA receptors located outside of the synapse lead to cell death whereas receptors located inside of the synapse lead to neurogeneration.[xlii],[xliii]
In 2003 Hardingham and Bading investigated the effects of Glu stimulation to NMDArs in the synapse. They showed that Ca+ influx from NMDArs located in the cleft are potent activators of CREB, the CREB regulated pro-survival gene that encodes BDNF and other CREB-regulated neurotrophins.[xliv]
Previously, Hardingham et al (2002) demonstrated that Glu stimulation of extrasynaptic NMDArs results in the activation of what they called a CREB ‘shut-off’ pathway. They determined that a specific dose-dependent signal is activated which results in the rapid decay of CREB phosphorylation. They also demonstrated that this effect is not stimulated by synaptic receptors and that blockade of the extrasynaptic receptors restores CREB induction. Further they determined that cell death, indicated by mitochondrial membrane breakdown, was limited to extrasynaptic receptor stimulation. [xlv]
Importantly, these studies provide evidence that the extrasynaptic CREB-inhibiting pathway is dominant over the synaptic CREB promoting pathway. Under normal physiological conditions, neurons are not chronically exposed to Glu. However, when Glu homeostasis is disrupted and the likelihood of synaptic spill-over is increased, the dominating and neurodegenerative effects of extrasynaptic stimulation could explain the well documented neuronal atrophy and loss of synaptic connectivity typical of disorders such as depression and bipolar disorder.
Studies by Ronald Duman and colleagues (2010, 2011, 2012) investigated the mechanisms by which low dose ketamine ameliorate the symptoms of depression, chronic stress and bipolar disorder. [xlvi],[xlvii],[xlviii] Their studies show that administration of ketamine results in activation of mTOR signaling pathways and a rapid increase in synapse and spine formation. Further, the robust increase in spines and synaptic protein synthesis was shown to reverse stress- and/or depression mediated deficits.[xlix]
It is widely recognized that disruption of BDNF and other neurotrophic factors result in neuronal atrophy and the loss of synaptic connectivity. Since activation of mTOR is dependent upon the release of BDNF (and CREB activation), the work of Hardingham and colleagues suggests that stimulation of synaptic NMDArs initiated the mTOR derived improvement of synaptic strength and signaling described by Duman and colleagues. Indeed, Duman and Aghajanian (2012) point to the rapid increase in Glu transmission which follows ketamine administration[l] as the catalyst underlying the neurogenic result.[li]
However, since the neuronal deficits associated with conditions such as bipolar disorder and depression are known to be the result of a hyperglutamatergic system, it seems perplexing that an increase of glutamate could help, rather than exacerbate, the condition.
Pittenger, Sanacora and Krystal (2007) hypothesize that the ability of an NMDA inhibitor to reverse the symptoms of depression may derive not primarily from the extra release of Glu into the synapse, but rather from the simultaneous inhibition of the extrasynaptic receptors.
There is a much greater concentration of NMDArs in the synapse than outside of it. Yet, as Pittenger, Sanacora and Krystal point out the concentration of an antagonist is expected to be equal in synaptic and extrasynaptic spaces. Therefore, administration of a low dose of an NMDA antagonist would block a large proportion of extrasynaptic receptors but only a small proportion of the synaptic ones. This would therefore more efficaciously block tonic levels of Glu from extrasynaptic uptake. In a hyperglutamatergic condition in which spill-over can be significant, this could potentially correct the pathologically overactive CREB shut-off pathway.
Additionally, since a low dose of antagonist leaves open a large number of receptors in the synapse and the extrasynaptic domination has been removed, the remaining active synaptic receptors would be able to promote neurogenerative cascades. Even further, the increased Glu release observed with NMDA inhibition optimizes this situation. Hence the robust repair observed by Duman and colleagues.
This hypothesis would explain the contradictory effects between large and small doses of NMDA antagonists.
[i] Olney JW, Latruyere J, Price M.(1989) Pathological changes induced in cerebrocortical neurons by phencyclidine and related drugs. Science; 244:1360-1362.
[ii] Fix AS, Horn JW, Wightman KA, et al. (1993) Neuronal vacuolization and necrosis induced by the noncompetitive N-methyl-D-aspartate (NMDA) antagonist MK(+)801 (dizocilpine maleate): A light and electron microscopic evaluation of the rat retrosplenial cortex. Exp Neurol; 123:204-215.
[iii] Farber NB, Wozniak DF, Price MT, et al. (1995) Age-specific neurotoxicity in the rat associated with NMDA receptor blockade: Potential relevance to schizophrenia? Biol Psychiatry; 38:788-796.
[iv] Olney JW, Latruyere J, Price M.(1989) Pathological changes induced in cerebrocortical neurons by phencyclidine and related drugs. Science; 244:1360-1362.
[v] Smith JW, Gastambide F, Gilmour G, et al.(2011) A comparison of the effects of ketamine and phencyclidine with other antagonists of the NMDA receptor in rodent assays of attention and working memory. Psychopharm; 217:255-269.
[vi] Olney JW, Latruyere J, Price M.(1989) Pathological changes induced in cerebrocortical neurons by phencyclidine and related drugs. Science; 244:1360-1362.
[vii] Smith JW, Gastambide F, Gilmour G, et al.(2011) A comparison of the effects of ketamine and phencyclidine with other antagonists of the NMDA receptor in rodent assays of attention and working memory. Psychopharm; 217:255-269.
[viii] Fix AS, Horn JW, Wightman KA, et al. (1993) Neuronal vacuolization and necrosis induced by the noncompetitive N-methyl-D-aspartate (NMDA) antagonist MK(+)801 (dizocilpine maleate): A light and electron microscopic evaluation of the rat retrosplenial cortex. Exp Neurol; 123:204-215.
[ix] Olney JW, Wozniak DF, Jevtovic-Todorovic V, et al. (2002) Drug-induced apoptotic neurodegeneration in the developing brain. Brain Pathol; 12:488-498.
[x] Ikonomidou C, Bosch F, Miksa M, et al. (1999) Blockade of NMDA receptors and apoptotic neurodegeneration in the developing brain. Science; 283:70-74.
[xi] Olney JW, Wozniak DF, Jevtovic-Todorovic V, et al. (2002) Drug-induced apoptotic neurodegeneration in the developing brain. Brain Pathol; 12:488-498.
[xii] Wang C, McInnis J, Ross-sanchez M, et al. (2001) Long-term behavioral and neurodegenerative effects of perinatal phencyclidine administrations: implications for schizophrenia. Neuorscience; 107(4):535-550.
[xiii] Horvath Z, Czopf J, Buzsaki G. (1997) MK-801-induced neuronal damage in rats. Brain Research; 753:181-195.
[xiv] Brambrink AM, Evers AS, Avidan MS, et al. (2012) ketamine-iduced neuroapoptosis in the fetal and neonatal rhesus macaque brain. Anesthesiology. 116(2):372-384.
[xv] Olney JW, Wozniak DF, Jevtovic-Todorovic V, et al. (2002) Drug-induced apoptotic neurodegeneration in the developing brain. Brain Pathol; 12:488-498.
[xvi] Olney JW, Wozniak DF, Jevtovic-Todorovic V, et al. (2002) Drug-induced apoptotic neurodegeneration in the developing brain. Brain Pathol; 12:488-498.
[xvii] Gutierrez S, Carnes A, Finucane GM et al. (2010) Is age-dependent ketamine-induced apoptosis in the rat somatosensory cortex influenced by temperature; Neuroscience, 168:253-262.
[xviii] Olney JW, Wozniak DF, Jevtovic-Todorovic V, et al. (2002) Drug-induced apoptotic neurodegeneration in the developing brain. Brain Patho 12:488-498.
[xix] Ikonomidou C, Bosch F, Miksa M et al. (1999) Blockade of NMDA receptors and apoptotic neurodegeneration in the developing brain. Science 283:70-74.
[xx] Wang C, Anastasio N, Papov V, et al. (2004) Blockade of N-methyl-D-aspartate receptors by phencyclidine causes the loss of corticostriatal neurons. Neuroscience 125:473-483.
[xxi] Wang CZ, Johnson KM (2007) the role of caspase-3 activation in phencyclidine-induced neuronal cell death in postnatal rats. Neuropsychopharmacology 32:1178-1194.
[xxii] Dobbing 1979
[xxiii] Romijn H, Hofman M, Gramsbergen A. (1991) At what age is the developing cerebral cortex of the rat comparable to that of the full-term newborn human baby? Early Hum Dev; 26(1):61-67.
[xxiv] Clancy B, Kersh B, Hyde J, et al.(2007) Web-based method for translating neurodevelopment from laboratory species to humans. Neuroinformatics 5:79-94.
[xxvi] Huang L, Liu Y, Jin W, et al. (2012) Detamine potentiates hippocampal neurodegeneration and persistent learning and memory impairment throught the PKCy-ERK signaling pathway in the developing brain.
[xxvii] Bosnjak ZJ, Yan Y, Canfield S, et al. Ketamine induces toxicity in human neurons differentiated from embryonic stem cells via mitochondrial apoptosis pathway. Curr Drug Saf. 7(2):106-119.
[xxviii] Diazgranados N, Ibrahim L, Brutsche N, et al. (2010) A randomized add-on trial of an N-methyl-D-aspartate antagonist in treatment-resistant bipolar depression. Arch Gen Psychiatry; 67(8):793-802.
[xxix] Pittenger C, Sanacora G, Krystal J. (2007) The NMDA receptor as a therapeutic target in major depressive disorder. CNS and Neurological Disorders-Drug Targets; 6:101-115.
[xxx] Zarate C Jr., Singh J, Carlson P, et al. (2006) A randomized trial of N-methyl-D-aspartate antagonist in threatement-resistant major depression. Arch Gen Psychiatry; 63:856-864.
[xxxi] Diazgranados N, Ibrahim L, Brutsche N, et al. (2010) A randomized add-on trial of an N-methyl-D-aspartate antagonist in treatment-resistant bipolar depression. Arch Gen Psychiatry; 67(8):793-802.
[xxxii] Messer M, Haller I, Larson P, et al. (2010) The use of a series of ketamine infusions in two patients with treatment-resistant depression. J Neuropsychiatry Clin Neursci; 22(4):442-444
[xxxiii] Messer M, Haller I. (2010) Maintenance ketamine treatment produces long-term recovery from depression. Primary Psychiatry; 17(4):48-50
[xxxiv] Price R, Nock M, Charney D, Mathew S. (2009) Effects of intravenous ketamine on explicit and implicit measures of sucidality in treatment-resistant depression. Biol Psychiatry; 66(5):522-526.
[xxxv] Newcomer J, Krystal J. (2001) NMDA receptor regulation of memory and behavior in humans. Hippocampus; 11:529-542.
[xxxvi] Pittenger C, Sanacora G, Krystal J. (2007) The NMDA receptor as a therapeutic target in major depressive disorder. CNS and Neurological Disorders-Drug Targets; 6:101-115.
[xxxvii] Paoletti P, Neyton J. (2007) NMDA receptor subunits: function and pharmacology. Current Opinion in Pharmacology; 7:39-47.
[xxxviii] Gardoni F, Di Luca M. (2006) New targets for pharmacological intervention in the glutamatergic synapse. European Journal of Pharmacology; 545:2-10.
[xxxix] Hardingham G, Bading H. (2003) The yin and yang of MNDA receptor signaling. Trends Neurosci; 2(2):81-89.
[xl] Paoletti P, Neyton J. (2007) NMDA receptor subunits: function and pharmacology. Current Opinion in Pharmacology; 7:39-47.
[xli] Pittenger C, Sanacora G, Krystal J. (2007) The NMDA receptor as a therapeutic target in major depressive disorder. CNS and Neurological Disorders-Drug Targets; 6:101-115.
[xlii] Hardingham G, Fukunaga Y, Bading H. (2002) Extrasynaptic MNDARs oppose synaptic NMDARs by triggering CREB shut-off and cell death pathways. Neuroscience; 5(5):405-414.
[xliii] Hardingham G, Bading H. (2003) The yin and yang of MNDA receptor signaling. Trends Neurosci; 2(2):81-89.
[xliv] Hardingham G, Bading H. (2003) The yin and yang of MNDA receptor signaling. Trends Neurosci; 2(2):81-89.
[xlv] Hardingham G, Fukunaga Y, Bading H. (2002) Extrasynaptic MNDARs oppose synaptic NMDARs by triggering CREB shut-off and cell death pathways. Neuroscience; 5(5):405-414.
[xlvi] Li N, Lee B, Liu R, et al.(2010) mTOR-dependent synapse formation underlies the rapid antidepressant effects of NMDA antagonists. Science; 329:959-964.
[xlvii] Li N, Liu R, Dwyer J, et al. (2011) Glutamate N-methyl-D-aspartate receptor antagonists rapidly reverse behavioral and synaptic deficits caused by chronic stress exposure. Biol Psychatry; 69(754-761).
[xlviii] Duman R, Aghajanian G. (2012) Synaptic dysfunction in depression: potential therapeutic targets.Science; 338:68-72.
[xlix] Li N, Lee B, Liu R, et al.(2010) mTOR-dependent synapse formation underlies the rapid antidepressant effects of NMDA antagonists. Science; 329:959-964.
[l] Pittenger C, Sanacora G, Krystal J. (2007) The NMDA receptor as a therapeutic target in major depressive disorder. CNS and Neurological Disorders-Drug Targets; 6:101-115.
[li] Duman R, Aghajanian G. (2012) Synaptic dysfunction in depression: potential therapeutic targets.Science; 338:68-72.