STEM CELLS, CANCER and LITHIUM
Presented July 20, 2011
STEM CELLS, NORMAL ADULT TISSUES
Tissues are believed to be supplied with, and maintained by stem cells (eg General, Rosen 2009; Skin, Muffler 2008, Harper 2010, Legg 2010, Pincelli 2010, Snippert 2010; Gut, Anderson 2011, Buske 2011, Khalek 2009, May 2009, Shaker 2010; Ovary, Stewart 2011; Brain, Lim 2011). These cells occupy special niches surrounded and supported by stromal cells and tissues, the latter providing congenial physical surroundings and appropriate chemical stimuli (by cytokines). The cells are pluripotent, being able, not only to reproduce and maintain their own stem cell line, but split off from them daughter cells (progenitors) that can differentiate towards, and ultimately attain, the mature cellular characteristics of the tissue involved (terminal differentiation). In the traditional paradigm, the cells most closely derived from the stem cells are progenitor or transit amplifying (TA) cells. This has been challenged, with the stem cells of mice tails maintaining cellular homeostasis by dividing into either stem cells, stem cell and a committed and differentiated line, or both committed and differentiated cells within the basal layer (Clayton 2007; Jones et al. 2007).The mature cells are referred-to here as post mitotic progeny cells or those that have terminally differentiated. A possible explanation for this finding is that, in very thin skinned regions, there may be only two layers of cells and, for the most part, eliminate a need for (or greatly attenuate) a TA stage.
The cellular kinetics are not easy to discover, but stem cells are believed to contribute ~1% to the cellularity of the tissue and to divide slowly. Some progenitors multiple more rapidly and yet may still retain the ability to self-perpetuate, but once in the transit amplifying stream, they may be able to divide 4-5 times to produce clones of less than 40 cells (based upon culture studies; Janes 2009). Disruption of stem cells’ associations at the basal membrane (niches) result in the loss of stem cell characteristics, with either rapid apoptosis, or the irreversible change to towards terminal differentiation (Muffler 2008) which may occur after about 3-4 weeks. The terminally differentiated progeny cease to divide further.
All this seems simple and attractive, but there are difficulties identifying cells that may truly be regarded as stem cells according to this paradigm ; there are reasons to believe that the true ‘stem cell’ may be better regarded as an organoid collection of cells (with mild variations within small limits) that are able to change from one form to another form within the group in response to physical or chemical stimuli (Buske 2011). Such an organoid basis for a tissue may, in part, explain the heterogeneity found in tumours arising within tissue.
A recent review summarizes and provides clarification (Li & Clevers 2010); within the sites of hair follicles, gut and bone marrow, they proposed zoned stem cell types, one type being primitive, long lived, and slowly replicating (‘reserved’, perhaps dividing once in 145 d), the other being more actively cycling (self-replicating lineage-producing), (‘primed’). The relative activity of the two groups is controlled by the stimulatory actions of such cytokines as Wnt, noggin, Fibroblast Growth Factor [FGF], Transforming Growth Factor-ß (TGFß; Moustakas 09) and gremlin, being countered by the inhibitory actions of BMP, DKK, Wif, sFRP related to the respective stem cell groups within their niches.
The important roles for bFGF are complex, with differences between species (Greber 3010, Böttcher 2005); in particular, mice have more primitive (naïve) embryonal stem cells which are encouraged to mature to an epiblast stem cell by bFGF, where they are held by the bFGF. The human embryonal stem cell is ‘primed’ and more like the murine epiblast stem cell. Withdrawal of bFGF from the latter will bring about differentiation (loss of stem cell capacity; Lanner 2010).
In effect, within the stem cell group in each tissue, there is a transit amplifying step (sTA) from the most primitive, slowly dividing type to the higher developed, faster growing type, prior to switching into the terminal differentiation pathway.
STEM CELLS in CANCERS
The concept that cancers have stem cells, progenitor cells and progeny cells seems attractive, but there are problems (Rosen 2009):
a) As with normal tissues, there have been difficulties identifying the stem cells (which, in some reports, are referred-to as tumour-initiating cells, TIC) thereby producing conflicting findings; a developing concept is that there are probably multiple stem cell ‘pools’ that are normally lineage-restricted as a result of specific environmental cues - cells that can interchange roles within a restricted range. This seems to apply to skin (Jensen 2008, Watt 2009, White 2011) and ovary (Stewart 2011 and probably most others.
b) Different groups have found that the stem cells in cancers may contribute a lot to the cellular bulk - some may be over 25%; the examination of fresh human head and neck squamous cell cancers indicate that the ‘stem’ cells (CD44high) constitute some 3.9-11% (means of 5) of the tumour cellular bulk, compared with of 3-5% of normal oral mucosal origin (Harper et al., 2010).
c) What the make-up of individual tumours may be is difficult to determine; the level of stem cell development and proliferation is probably unknown and the contribution by, and kinetics of the transit amplifying cells are not well characterized.
d) Stem cells are considered to be quite resistant to chemotherapy and radiotherapy (Harper 2010). In part, this may be due to the expression of melanoma chondroitin sulphate proteoglycan (human MCSP, CSPG4, HMW-MAA, rat NG2, mouse AN2), which is associated with integrin-PI3K/Akt signalling (Chekenya 2008), activation of FAK/ERK (Yang 2004), and which may be retained by progenitor (TA) or even progeny cells (terminally differentiated), thereby resulting in reduced apoptosis. This means that the impressive ‘tumour crashing’ that may be seen after such treatments is simply an illusion - the sensitive progenitors and progeny might have been convincingly destroyed, but the stem cells simply repopulate the tumour; the net benefits being marginal. Just how the stem cells re-populate the new tumour bulk with resistant cells is unclear. One suggestion is that the resistant ‘stem’ cells proliferate and replace the formerly transit amplifying cells that have been destroyed (Anderson 2011)
e) Stem cells in the skin, squamous cell cancer, and malignant melanoma have a MCSP coating. This seems to match the ß1 integrin marker distribution with stem cells as assessed at the cellular level in vitro (Legg 2003, Yang 2004), although the distributions of MCSP and ß1 integrin within the basal layer of normal (but variably ageing) human skins do not seem to correspond, as assessed by medium microscopic magnification; MCSP being found between the rete ridges, the ß1 integrin distribution being more general (Giangreco 2010).
MCSP has a core protein with an essential cytoplasmic tail, and may have variable glycosaminoglycan side chains. Its core protein (studied with MG2) can act as a receptor for bFGF (a role enhanced by the glycosaminoglycan) and PDGF-AA (being slightly inhibited by the glycosaminoglycan), but not PDGF-BB, VEGF, TGFß1 and EGF (Goretzki 1999). Sadly, the effects of soluble or bound heparan within their protocols were not studied, and nor were other heparan-binding ligands (eg Bone Morphogenic Protein [BMP], HB-EGF-like). Ligand interactions in binding affinities also were not addressed. MCSP is up-regulated in malignant stem cells, indicating a difference with the normal stem cells (Jensen 2008) but does not seem to be involved in differentiation directly, appearing to influence positively cell adhesion, possibly involving fibronectin (etc.) binding and as a co-receptor for α4ß1 integrin and, consequently, maintaining stem cell nesting within the presumed niches. Movement out of these select niche environments may expose stem cells to differentiation-inducing and hostile conditions, and these may lead to either apoptosis or a switch to the differentiation pathway, which may not involve cell division. In this way, MCSP perturbation (in quantity &/or quantity) may result in a depletion of the stem cell pool, being a slow loss that may have late effects. Once out of the protective niche (as in cell culture), MCSP may encourage migration and aid metastasis (Wang 2010), with the protein core and its intracellular component crucial (Yang 2009). The importance of MCSP may be in that, having chondroitin sulphate as a component, the glycosaminoglycan component’s formation and configuration may be influenced by Lithium, in ways similar to those involving the heparan molecules (involving Lithium’s inhibition of the gPAPP Golgi enzyme; Frederick 2008) as by a reduction in quantity &/or quality. Since MCSP can act as a co-receptor for PDGF and bFGF, it may play a part in stem cell growth kinetics.
So, Lithium, through its action on the Golgi gPAPP enzyme, may reduce the quantity &/or quality of both stem cell MCSP and heparan sulphate glycosaminoglycan, resulting in
i) a loss of FGF cytokine stem cell stimulus, FGF being one of the key cytokines that maintain the stem cells in status quo,
ii) a loss of the quantity &/or quality of the MCSP stem cell coat, leading to a greater tendency for the stem cells to wander outside the protective stem cell niche, and
iii) enhancement of BMP signaling to the stem cells, either directly involving the stem cell BMP receptors &/or internalization of BMP (Kuo 2010; Jiao 2007, Hu 2009) &/or
iv) reduced binding of the BMP inhibitors noggin &/or chordin to the stem cell surface, noggin being another heparan-binding cytokine (Hu 2009).
Lithium’s capability in activating the Wnt/ß-catenin pathway and switching off the deactivating phosphorylation of SMAD1 by GSK3ß, allows SMAD1 to suppress the promoter of the gene Nanog and drive the stem cells towards differentiation (Sumi 2008; Xu 2008). This could happen after the bolus dose on alternate days.
Since these factors may result in either apoptosis or the irreversible switch to differentiation (possibly without cell division), there is a stripping of stem cells from the stem cell niches. In cancers, there may be a delay before this loss of stem cell reserve may be noticed (days-weeks). In non-malignant conditions, such as rheumatoid arthritis (RA), this delay may explain an observed remission of about 5-6 weeks following a short course of Lithium treatment (over 3-5 days); the relapse probably coming after the depleted RA stem cell pool builds up to some critical level. A possible reason for the Lithium treatment not affecting normal tissues may be that there is sufficient redundancy of adhesion factors (eg integrins, cadherins) to provide the necessary level of stem cell anchors to maintain their position. In diseased tissues, the effectiveness of the usual anchoring and the anchoring redundancy decline, allowing the stem cells to wander.
MESOTHELIOMA and LITHIUM
Preamble: Malignant mesothelioma is a nasty malignancy arising from mesothelium, the surface lining over the pleural and peritoneal cavities; it is typically from the lining of the lung. It has been associated with previous exposure and breathing-in of asbestos fibres. Its clinical course is usually quite rapid, often with tumour masses coming through the chest wall and appearing under the skin. It is typically quite resistant to standard forms of therapy, namely chemotherapy and radiotherapy. This Chapter reports the findings from a 48 year old man with mesothelioma who, in addition to other therapies, took intermittent bolus lithium over two intervals. Fortunately, he had an initial elevated level of the tumour marker CA125 in his blood, so that this could be measured and its changes assessed, since it is more objective than a large subcutaneous tumour mass coming through his chest wall.
No clear changes attributable to other therapies (other than chemotherapy) can be identified. Accordingly, these other therapies will be regarded as background constants. The graph of his tumour marker is presented (Figure 1) and its features will be discussed.
This report cannot prove anything, but it may provide a stimulus for discussion and further studies.
There must be many assumptions. Many will be listed (A1, A2, A3 . . . etc.) in the Appendix.
To avoid expressing caution repeatedly in the presentation, the issues will be presented in a positive, seemingly definite format. Issues may be questioned, corrected or dismissed.
THE GRAPH of CA125 CHANGES
Before Lithium therapy, this man had chemotherapy with cisplatin and gemcitabine. The pre-treatment CA125 was 1190 U/mL (point A), falling after the chemotherapy to 479 U/mL (point B), a partial response, which, in the case of mesothelioma, is a reasonable response. Treatment using bolus Lithium in an intermittent protocol was, as outlined earlier :
commenced at the end of the last dose of chemotherapy.
Thereafter, the CA125 level showed a very slight fall (point C), and not the expected recommencement of rise usually seen shortly after chemotherapy. (The venepuncture was probably done when he was mildly haemoconcentrated, which would have made the reading higher than would be otherwise.)
After about 64 d, for unclear reasons, he ceased the treatment. Shortly after this, his CA125 was found to have risen swiftly (point D), and the subcutaneous lump in his chest wall had hardened and enlarged. The estimated CA125 doubling time approached 18 day (fast).
After feeling the changes to his lump and believing that he had done the wrong thing in ceasing the Lithium treatment, he recommenced it. Because he felt overwhelmed by the number of tablets he was taking, he discontinued the treatment again after about 18 d.
However, the CA125 continued to rise, although there seemed to be a slow reduction in the rate of rise (points E, F & G).
An Oncologist persuaded him to try Sunitinib after point G. There is little evidence (if any) that this new therapy did anything, one way or another; the CA125 fall between points G & H would seem a continuation of a curving pattern already established. His condition at this stage was very poor.
At about point I, he was terminal, and he died shortly afterwards.
MESOTHELIOMA and LITHIUM
Figure 1. This graph has a logarithmic vertical axis. The horizontal axis represents time. Pending correcting the omission of some CA125 values from the graph, they are: B 479; E 1772; F 2098; H 2136; I 2340.
With the commencement of the Lithium, the CA125 level remained essentially unchanged (points B-C). With no CA125 determination between points C & D, the previous line (B-C) is assumed to continue until about the time when he ceased the lithium. The rise in the CA125 is then assumed to start at or about that point, continuing through points D-E, with a peak doubling time of about 18 d. (The thin blue line represents the doubling time of 18 d for reference.) Thereafter there is a lessening of the growth rate, eventually becoming negative between points G & H.
CONCLUSIONS and ELABORATIONS from THE GRAPH (See below for assumptions)
1) The Lithium treatment following the chemotherapy (starting point B) seems to have held in check the expected rise in the CA125 (note the paucity of points, A4). Experience to date has indicated that the Lithium treatment takes days (Rheumatoid arthritis) to weeks (cancer) to result in any appreciable objective change or improvement. The surviving post chemotherapy cancer cells would be expected to have some impairments to their metabolism. The graph would be consistent with Lithium being able to exert a beneficial effect, as by limiting tumour growth, commencing at an earlier stage than would otherwise have been expected.
2) He ceased the Lithium sometime after point C. The graph line joining C and D is very likely not a true representation of the events. The line B-C, when extrapolated to the date of the Lithium cessation, corresponds closely with the extrapolation of the line D-E. Accordingly, the rapid growth surge may well have started at or about that point. This would be consistent with the belief that the Lithium had held the tumour in a non-proliferating, dormant state; the withdrawal of the Lithium resulting in a surge of cells entering into or completing cell division - rapid growth. The conclusion from this is that those with tumours and who are taking Lithium, should not cease taking it.
3) Although he restarted taking the Lithium after point D, the damage had been done, it was as if he had not been taking the earlier Lithium course at all, and the expected delay of weeks would be expected before any appreciable advantage.
4) After point E, there seemed to be a fall in the rate of growth. The factors operating are difficult to assess - it may be a late Lithium effect, following the stripping of the stem cell population, coupled with the block of heparan-dependent growth factor cytokines such as VEGF (which stimulates capillary growth) &/or secondary to the typical central necrosis and cavitation of large tumour masses. The degree to which these contributed cannot be assessed.
APPENDIX - ASSUMPTIONS
A1) Lithium, given as intermittent bolus therapy, can have effects attributable to depletion of stem cells and diminution of cytokine, stimulant actions
(see http://www.malsmusings.info/index_files/EXPLAINED.htm) and above.
A2) The serum level of the tumour markers tend to reflect the number of cells producing it, rather than the volume of cells. Stem cells produce no tumour marker and the most immature ~25% of progenitor (TA) cells produce none.
A3) For a poorly differentiated tumour, such as a mesothelioma, there will be few (if any) cells fully reaching terminal differentiation. In this study, they will be insignificant and disregarded.
A4) The man was treated by physicians who may have had limited knowledge of the Lithium treatment and an appropriate level of monitoring for sound conclusions, and made only infrequent serum level determinations of the CA125.
A5) There are assumed to be no major changes in the level of the CA125 between the presented points.
A6) The line B-C is assumed to continue as such until about the date when the Lithium treatment (1) ceased.
CELLULAR COMPARTMENT SHIFTS - Hypothetical
Below is an histogram which presents an hypothetical analysis of the tumour cell groups on the date that each blood level for the CA125 was determined. The horizontal axis is not an accurate time representation. Some assumptions are listed above.
MESOTHELIOMA and LITHIUM - Hypothetical
The chemotherapy would have eliminated about 50% of the rapidly dividing cells, being the CA125__ and the CA125+ pools, the stem cells pool remained untouched.
The first treatment phase with bolus, intermittent Lithium has a fall in the contribution of the stem cells; the pool having been stripped, as outlined in the text presentation above, with some embarking upon the differentiation for the TA pool. The more immature TA cells will have matured and moved into the CA125+, and some of the latter may have expired by aging.The Lithium, by blocking cytokine growth factor access to receptors by perturbations of heparan quantity &/or structure, has produced a block in proliferation, with a pent up supply of CA125+ cells.
After ceasing the Lithium course (1), the pent up cells enter a vigorous proliferation phase; but the stem cell pool supporting their numbers has been depleted and is unable to repopulate the CA125-- pool. The stem cell pool then starts a slow recovery.
Recommencement of the Lithium (2) saw a slowing of the stem cell recovery, a decline in the CA125-- numbers relative to the CA125+ numbers, and the total cellular growth slowed.
Thereafter, there was a slow increase in stem cell numbers again, but still at levels remain too small to support fully the total TA population, in which the the CA125-- pool (most reliant upon supplies from the stem cell pool) declined.
At the end, there may have been a slight recovery in the CA125+ TA cell pool and proliferation, well after the cessation of the Lithium treatment, with the stems cells representing ~7% of the total, down on the level pre-chemotherapy.
Copyright © MA Traill 2011
Abdul Khalek FJ, Gallicano GI et al. Gastrointest Cancer Res. 2010; Supp 1: S16-S23
Anderson EC, Hessman C et al. Cancers (Basle) 2011; 3(1):319-339
Böttcher RT & Niehrs C. Endocrine Rev 2005; 26(1):63-77
Buske P, Galle J et al. PLoS Comput Biol 2011; 7(1): e1001045
Chekenya M, Krakstad C et al. Oncogene 2008; 27(39): 5182-94
Clayton E, Doupé DP et al. Nature 2007; 446:185-189
Frederick JP, Tafari AT et al. Proc Natl Acad Sci USA 2008;105(33):11605-12
Giangreco A, Goldie SJ et al. J Invest Dermatol. 2010
Goretzki L, Burg MA et al. J Biol Chem. 1999
Greber B, Wu G et al. Cell Stem Cell 2010; 6:215-26
Harper LJ, Costea DE et al. BMC Cancer 2010; 10:166-82
Hu Z, Wang C et al. J Cell Sci 2008; 122:1145-54
Janes SM, Ofstad TA et al. Cell Res 2009; 19(3):328-39
Jiao X, Billings PC et al. J Biol Chem. 2007; 282(2):1080-6
Jensen KB, Jones J et al. Cancer Lett 2008; 272(1):23-31
Jones PH, Simons BD et al. Cell Stem Cell 2007; 1:371-81
Kuo W-J, Digman MA et al. Mol Biol Cell 2010; 21:4028-4041
Lanner F & Rossant J. Development 2010; 137:3351-3360
Legg J, Jensen UB et al. Development 2003;6049-63
Li L & Clevers H. Nature 2010; 327:542
Lim SK, Alcantara Llaguno SR et al. BMB Rep. 2011;44(3):158-164
May R, Sureban SM et al. Stem Cells 2009; 27(10):2571-9
Moustakas A & Heldin C-H. Development 2009; 136:3699-714
Muffler S, Stark H-J et al. Stem Cells 2008; 26:2506-15
Pincelli C & Marconi A. J Cell Physiol. 2010; 225:310-5
Rosen JM & Jordan CT. Science 2009; 324(5935):1670-3
Shaker A & Rubin DC. Transl Res 2010; 156:180-187
Snippert HJ, Haegebarth A et al. Science 2010; 327:1385-9
Stewart JM, Shaw PA et al. Proc Nat Acad Sci USA 2011; 108(16):6468-73
Sumi T, Tsuneyoshi N et al. Development 2008; 135:2969-79
Wang X, Wang Y et al. Curr Mol Med 2010; 10:419-429
Watt FM & Jensen JB. EMBO Mol Med 2009; 1(5):260-267
White AC, Tran K et al. Proc Nat Acad Sci USA 2011
Yang J, Price MA et al. J Cell Biol. 2004; 165(6):881-91
Yang J, Price MA et al. Cancer Res. 2009; 69(19):7538-47