August 2014



1 Lithium stimulates monocytes/macrophages to produce TNF. This may start early (Giambelluca et al. 2014) by a mechanism that may involve the GSK-3α of neutrophils activating latent TNF mRNA, to then increase (Kleinerman,1989), possibly by an autocrine positive feedback loop via the TNF receptors. Both TNF and FasL show similar responses (Kaufmann L et al, 2011) - and TRAIL is probably similar. Full production of TNF is probably 20-30 h after the Li stimulus.This provides for roughly a day to ‘prime’ a patient, when it may potentiate cytotoxicity (Beyaert et al. 1989).

2 The TNF receptors lead to two main pathways :

(a) The ‘survival/inflammatory’ pathway driven by NF-κB leading to the activation of any of up to ~500 genes.Their outputs generally rely on ribosomal function to proceed. (Laboratory testing not wanting this pathway add Cycloheximide to block ribosomal function; eg Juric V et al, 2012.)

(b) The programmed cell death (PCD) pathway(s) - apoptosis (Caspase-3), necroptosis (no Caspase-3) etc.

3 Lithium-sensitive enzymes: Three seem important currently -

(a) GSK-3β/Wnt pathway inhibition by Lithium has the potential to induce apoptosis, and it may have been so involved  (Madiehe et al. 1995; also van Gijn et al. 2001). It is considered important in neuropathology. Its Ki ~2.0 mmol/L, too high for therapy without sensitization by other kinases (eg Akt), making study difficult. Its isoform, GSK-3α, generally considered to be of lesser importance, may now be important in the initiation of the TNF response, releasing mRNA for TNF (see Giambelluca et al. 2014). An inhibitor of this enzyme is said to induce breast cancer cell death through Cyclin D1 and p53 activation (Kim et al. 2013), and inhibition within glioma cells can induce death with a marked decrease in NF-κB (Kotliarova, 2008), but its actions are complex (McCubrey et al. 2014). Whilst intranuclear GSK-3, if inactivated by Lithium shortly after a dose, may augment HSF1's actions (Xavier et al. 2000), this action may be over-shadowed by the Bpnt1 inhibition and RNA/ribosome suppression (Santagata et al. 2013) that may occur also at about the same time (see below).

(Note that almost all in vitro testing of GSK-3 involves non-physiological, high concentrations of Lithium that would seem certain to block, almost completely, the enzymes to be discussed below. Such experiments are difficult to relate to clinical medicine.)

(b) The Golgi gPAP, which detoxifies PAP (Frederick et al. 2008). With a Ki ~0.175 mmol/L, it should be inhibited by almost any dose of Lithium. Inhibition perturbs chondroitin sulphate & Heparan sulphate (HS) formation, quantity &/or quality (Vissers et al.2011), and hence the co-receptor functioning of those cytokines that require it (eg FGF & BMP, Lanner et al. 2010; Kraushaar et al. 2012) and the cell-available potency of heparin-binding stromal cytokines, such as TGF-β and SDF1, which activate the transcription factor HSF1 in patients having thermotherapy (Scherz-Shouval et al. 2014). Perturbed chondroitin sulphate production may decrease the supply for MCSP, a stem cell marker that helps tethering of these cells in their niches - without MCSP, the stem cells can move from their niches and lose stem cell characteristics (Legg et al. 2003).

There is uncertainty as to how PAP build-up may upset intra-Golgi sulphation. There are grounds for believing that it suppresses transporter function - in the case of the PAPS transporter, such an action would be a negative feedback to PAPS entry into the Golgi.The effect may be more general, with another transporter (at least) being inhibited - the GDP-fucose transporter (Hellbusch et al.) in particular, which may be essential to maintain stem cells (Sasaki et al. 2009). This latter effect could account (at least in part) for the leukocytosis, decreased leukocyte ‘stickiness’  (probably L-selectin deficiency; MacGregor & Dyson 1980) and diarrhoea (Vestergaard et al. 1988) noted with Lithium treatments; cf LADS II mutation). If this is the case, then stem cell growth and proliferation may be compromised by a second mechanism (see MCSP above).

An indirect effect of Lithium upon stem cell survival may be associated with the need for such cells to respond to the stresses that may cause unfolded proteins within the endoplasmic reticulum (van Galen et al. 2014). Such protein defects may be induced by reactive oxygen species or DNA damage, and human haematological stem cells seem specifically sensitive to such defects, proceeding to selective programmed cell death (eg apoptosis). Lithium may augment this process by a likely inhibition of the N-glycosylation of Golgi proteins (as does Tunicamycin), by interfering with the enzyme NDST, subsequent to the gPAPP inhibition (Traill, 2009).

(c) The cytoplasmic Bpnt1 (López-Coronado et al. 1999), which detoxifies PAP. With a Ki ~0.8 mmol/l, it is likely to be partially inhibited by high therapeutic levels of Lithium. Mutations in mice demonstrate a dramatic and striking disturbance of RNA handling in hepatocytes (but also upper bowel enterocytes & renal tubular cells - cell type specificity seems important). Nuclei swell, nucleoli greatly enlarge, and cytoplasmic RNA virtually disappears (Hudson et al. 2013, 2014). (Some of the more important side effects of Lithium [eg renal damage], are attributed to the PAP build-up following Bpnt1 inhibition.)  The key biochemical targets of PAP include the oligoribonucleases (Mechold 2006). Cells with little demand for sulphate donation would likely be at smaller risk, but those with considerable sulphation (as in liver detoxification and glandular mucous production - adenocarcinomata) may be at particular risk. Little is known about this toxic effect, but it would be easy to speculate that even small rises in the cytoplasmic concentration  of PAP may induce RNA aberrations that could have profound effects on the RNA /Ribosomal/Nucleolus functional performance & structure, with probably a decreased HSF1 genetic stimulation (Santagata, 2013),  and that these may not be apparent at the morphological level. These may occur to the extent that effects may last into the next day. There may be an expectation that, by disturbing RNA processing, the ‘survival/inflammatory’ pathway may be inhibited, just as cycloheximide inhibits it in vitro (see above), and the TNF action redirected into the PCD pathway(s).

Concept: Treatment with Lithium should not be regarded as similar to the treatments with most other medications. The latter tend to have direct effect upon targets; effects which operate for as long as the medication is above the relevant thresholds.

Lithium seems to start a cascade of actions, which may continue after the presence of the initiating Lithium dose may have dissipated - and the continuing presence of Lithium could modify later steps in the cascade.

Stem Cells: The dual effects of defective MCSP and Heparan sulphate may be expected to have a deleterious effect on stem cell maintenance, which may affect cancer proliferation and also ‘rheumatoid stem cell’ growth & maintenance (perhaps by ‘forbidden clones’), which may be revealed in long-term treatments (eg > 6 weeks). There is a possibility that inhibition of the Golgi enzyme NDST may lead to human stem cell loss.


1) Initial Treatment Day If a patient is given an oral dose of 1 g Lithium carbonate, the peak serum level is reached at about 1 h, with the level (mmol/L) roughly the same numerals as the dose (g). Doses >750 mg may induce nausea/vomiting, so the dose is best split with 2-6 h between portions.

The serum level fall is exponential (equilibration) until ~6 h, then there is a steady fall (renal excretion).

Determinations of the intracellular Li+ levels are rare and conflicting. The highest level is probably at about the time when equilibration is completed.

Typical serum values may be (noting that standard laboratory testing may have disappointing accuracy) . . . .               peak (1 g divided)   ~         0.7 mmol/L

                                        24 h                       ~          0.4 mmol/L

                                        48 h                       ~          0.2 mmol/L

a. gPAPP will be inhibited appreciably throughout the 48 h cycles. This may interfere with the intra-Golgi detoxification of PAP, with the possibility of reduced precursors entering the Golgi - PAPS & GDP-Fucose. The latter may affect stem cell growth. The structural integrity of Heparan and Chondroitin sulphate and molecules that incorporate them may affect co-receptor function and stem cell MCSP, the latter also affecting stem cell growth. An inhibition of N-glycation by NDST may specifically impede human stem cell survival.

b. Bpnt1 may be inhibited partially at about the equilibration peak, but for a short time only. In the case of cells with Sulphate processing, there may be a build-up of PAPP, with disturbed RNA metabolism affecting the cell-type concerned. Speculating, the smallish Bpnt1 inhibition may, by disturb only a small portion of the RNA/Ribosomal/Nucleolus structure & function, but this may have profound organizational effects that may last into the second day.

c. GSK-3α/β The sensitivity of the GSKs is dependent upon phosphorylation in vivo. This is difficult to assess reliably.

Li+ on GSK-3α seems to be involved in the initiation of TNF production; however full output may not be achieved until ~20 h later (~ the next day) by sequentially evolving mechanisms.

Li+ on GSK-3β (the Wnt pathway) seems to cause PCD, but the mechanism(s) are not clear.

2) Second Day (Usually no Lithium is administered)

a. TNF production will be established and will operate through the receptor(s) TNFR1 (& TNFR2). This may set up a positive feedback loop to generate more TNF. Whilst the receptors do not require Heparan as a co-receptor, TNF can bind to Heparin, meaning that Heparan may influence the cellular availability of TNF (see gPAPP above). There is then the pathway option :

i] Survival/inflammation (via NF-κB). If the gene products are unable to access fully functioning RNA transcription and Ribosomes (see Bpnt1 above), the PCD pathway(s) will be favoured.

ii] Other factors may favour the PCD pathway :

                Curcumin, said to inhibit the NF-κB pathways (Anand, 2008)

                (Possibly) Oxidizing conditions (eg intravenous Vitamin C)

b. gPAPP. The Li+ inhibitory effects upon this and the downstream consequences should still be in evidence.

3) Later days

a. The general cycle of 48 h can be repeated indefinitely

b. There are reasons to suspect that Stem Cells are inhibited :

i] MCSP may be abnormal, and the Stem Cells wander from their niches, a state that leads either to death or differentiation

ii] If the inhibition of gPAPP leads to PAP poisoning the transporters into the Golgi, the perturbations to L-fucose production, the possible suppression of NDST, and the subsequent inhibition of Stem Cell growth and proliferation. This may be a factor in the treatment of ‘auto-immune’ conditions such as Rheumatoid arthritis, where diseased Stem Cells (if lymphoid, then ‘forbidden clones’ ) may be inhibited or suppressed; a state that may last for 4-5 weeks until the transit-amplifying cells can recover.


Anand P, Sundaram C et al. Cancer Lett. 2008; 267:133

Beyaert R, Vanhaesebroeck B et al. Proc Natl Acad Sci USA, 1989(86):9494

Frederick JP, Tafari AT et al. Proc Natl Acad Sci USA. 2008; 105(33):11605

Giambelluca MS, Bertheau-Mailhot G et al, FASEB J, 2014, May 6th

Hellbusch CC, Sperandio M et al. J Biol Chem. 2007; 282(14):10762

Hudson BH, Frederick JP et al. Proc Natl Acad Sci USA. 2013; 110(13):5040

Hudson BH & York JD. Adv Biol Regul. 2014; 54:208

Kim HM, Kim C-S et al. PLos ONE 2013; 8(4):e60383

Kotliarova S, Pastorino S et al. Cancer Res. 2008; 68(16):6643

Kraushaar D, Rai S et al. J Biol Chem. 2012; 287(27):22691

Lanner F, Lee KL et al. Stem Cells 2010; 28:191

Legg J, Jensen UB et al. Development 2003, 130:6049

López-Coronado J, Bellés J et al. J Biol Chem. 1999; 274:16034

MacGregor RR & Dyson WL. Adv Exp Med Biol. 1980; 127:347

Madiehe A, Leseilane J et al. Biochem Biophys Res Commun. 1995; 209(2):768

McCubrey JA, Steelman LS et al. Oncotarget 2014; 5(10):2881

Mechold U, Ogryzko V et al. Nucl Acid Res. 2006; 34(8):2364

Santagata S, Mendillo ML et al. Science 2013; 341(6143):1238303

Sasaki N, Hirano T et al. PLos ONE 2009; 4(12):e8262

Scherz-Shouval R,l Santagata S et al. Cell 2014;158;564

Traill MA,  www.

van Galen P, Kreso A et al. Nature 2014; 510():268

van Gijn ME, Snel F et al. Exp Cell Res. 2001; 265:46

Vestergaard P, Poulstrup I et al. Acta Psychiatr Scand. 1988; 78(4):434

Vissers LE, Lausch E et al. Am J Hum Genet. 2011; 88:608

Xavier IJ, Mercier PA et al. J Biol Chem. 200; 275(37):29147

More to follow

Copyright © MA Traill, August 2014