August 2009  

Mal's Musings

DISEASES & LITHIUM

EXPLAINED

 

Prelude Summary

 

 

The Enigma

Lithium has been in clinical use (mainly Psychiatry) for about 50 years. Despite it being involved directly in at least one key pathway for gene regulation, and also affecting the quality of the Heparan sulphate chains for cell-to-cell signalling, there are no significant adverse structural changes recorded, with the only exception being the cardiac defects derived by in utero exposure.

The explanation for this enigma is provided within the general term "redundancy" as applied in biological systems: when a key component in a biological system fails, there are usually a number of back-up participants possessing overlapping functions, and these can replace the missing component, at least in part, with associated negative &/or positive feedback. That there is a loss may not be revealed until the organism is placed under an appropriate stress. A topical example is the enzyme Heparanase; it splits Heparan sulphate chains and also participates in signalling. Mice have been bred with no functional Heparanase, yet, to the surprised workers, the mice seemed essentially normal, with normal growth and reproduction. (Some mild changes were revealed by biochemical testing.) The explanation was that there was extensive redundancy of alternative enzymes and factors, and these could nearly fully compensate for the deficiency. Such a defect was established at conception and the organism had plenty of time for biological adjustment; redundant capacity with feed-back loops could be utilized. Likewise, Psychiatric patients generally take Lithium on a long-term basis, with an attempt to keep the blood level within a narrow range. Accordingly, there would be time for biochemical processes to adjust, relying upon the general redundancy of inherent capacity and feed-back controls.

The experiences and dissertations to be presented here relate to bolus Lithium administered on an alternate day basis. This can be expected to frustrate those adjustments utilizing redundant capacity: imagine the state when Lithium may be in a therapeutically relevant range for one key enzyme (say GSK3β of the Wnt pathway) for only 8 hour out of 48 hour. Compensatory changes would, no doubt commence (with an activation lag and production delay times) but, as the corrective changes start building up, the Lithium level falls below the therapeutic threshold for that pathway, and the changes that were underway would now slow down and cease, perhaps leaving some biochemical pathways "in limbo." Likewise, changes to Heparan sulphate and chondroitin sulphate would be changing cyclically, creating disturbances for the various  growth factors that bind to them, all with their differing thresholds; a degree of biochemical disorganization. The view is, then, that the intermittent, bolus Lithium creates a biochemical environment very different from that of the long-term, approximately constant, daily doses typical of Psychiatry. (In this dissertation there are also documented other examples where an intermittent dosage regime of a compound produces results differing from the usual daily dose. The explanations may be the same.)

Basic Issues  

  

Many disease processes involve the production of excessive cell-to-cell signalling molecules (viz cytokines) or viruses. In many cases, the reception of these molecules or viruses involves molecules that can bind loosely to those signalling molecules or viruses and help them latch onto the receptor molecules. Chief amongst these loose-binding molecules are the long sugar chains of Heparan sulphate on proteoglycan core protein molecules (HSPG), the production of which in the Golgi, Lithium can disturb. Provided that biochemical redundancy-associated readjustments can be reduced, all medical conditions that involve cell-to-cell signalling with cytokines can be expected to be modified, in some way and, to some degree, by Lithium (bolus/intermittent regime). Diseases that come to mind in this context are Rheumatoid arthritis, Lupus erythematosus, Multiple sclerosis, Sarcoidosis, Cancers (of all types) and Hepatitis C.  Personal experience has been with these. There are certain to be more.

This presentation is not prophetic; it is an attempt to explain observed effects identified in laboratory testing of animals or in actual clinical use.

 

Mode of Action

 

This section will present an heuristic overview of the topics in a somewhat didactic form. Hypotheses are involved, with the belief that there may be amendments, deletions and corrections from time to time, particularly as new findings are reported and the knowledge base widens.

 

Lymphocyte and NK cell Changes

 

                                                                                                      GRAPH 1

                                            


Lithium and the cancer

 

 

Specific Conditions

 

  1.  Cancer: Two cancers have been documented for which the intermittent bolus does of Lithium seem to have had beneficial effects. As noted above, Lithium can stimulate the Wnt pathway initially, but later, to leave it unsustained when the Lithium level falls. The Lithium changes to the quantity and quality of the Heparan sulphate will impede the growth factor signalling (eg Wnt & FGF), depriving the cells of stimulus. This can lead to apoptosis (organized cell death). Similar processes can affect stimuli (such as HPSE & VEGF) for endothelial cells, resulting in disturbed new capillary growth, leading to an inability to support the cancer's blood supply requirements.
  2.  Rheumatoid arthritis: The synovial joint lining and the fluid within the joint, have high levels of inflammatory signalling molecules, amongst them HPSE. The effects of intermittent Lithium in this condition are probably similar to those outlined to date; the disturbed Heparan sulphate not supporting adequately the signalling by the cytokines, including HPSE. In practical terms, the Lithium dosing can be limited to 2-3 treatment days per 4-5 week, the protocol depending upon severity. (If mild, 750 mg Lithium carbonate for 2 alternate days may be adequate; for more severe relapses, two doses of 500 mg separated by 6 h every alternate day for 3 treatment days may be considered.) Clinical response is slow and not usual for some days to a week or more, and patients need to be persuaded to persevere. (An elderly patient with an initial attack of moderate severity was commenced on a Lithium protocol. She seemed to have a low pain threshold and was not happy waiting. Relatives demanded that she have a referral to a Rheumatologist, following which, the latter started treatment promptly with Prednisolone [of course]. This made them all happier for some days until severe pain developed in both hips - so, have you guessed it ? - she had developed bilateral avascular necrosis of her femoral heads - an acknowledged complication of Prednisolone therapy. After this, she was effectively immobilized, and was lost to follow-up !) The ability for the Lithium Rx to induce remissions for 5-6 week is puzzling. Such an interval is not easily explained by biochemical processes. The conclusion is that hypothetical Rheumatoid progenitor cells are eliminated, (as may happen with cancer cells, see above), and that the interval of 4-5 weeks represents the time for stem cell/progenitors to proliferate and regenerate inflammation.     
  3.  Lupus Erythematosus: As with Rheumatoid arthritis, a Lithium-induced change of the HSPG would be expected to modify the disease progress in some way. Experience with this condition has been limited but, at a subjective level, Lithium may have aided induction of remissions, and for a patient on long-term Prednisolone prophylaxis, it seemed able to assist in reducing the maintenance dose of Prednisolone to levels not achieved for a long time (years).  
  4.  Multiple sclerosis: Experience with this condition has been limited largely to one patient. Attempts to have others try Lithium were largely abandoned in the mid 1990s, when Interferon-β became available for therapy. Prior to that, there were the problems of explaining the nature of the treatment to the disbelieving patients (who inevitably rushed off to consult their Neurologists; they usually described the use of Lithium in terms such as "unscientific quackery.") However, there are now good reasons to believe that there are valid scientific grounds to support the use of Lithium, at least in principle (see later). Unlike with the other conditions, the patient, through her own trials, came to believe that the Lithium worked best aborting acute relapses. At the first sign of a relapse, she would take 1 g of Lithium, followed by a lot of water (to aid renal excretion in order to limit the blood peak duration). She would then take another 1 g the next day. This seemed to abort relapses, with symptomatic relief of symptoms, such as nystagmus and lethargy, noticeable within hours. Trials of prophylactic Lithium (as for Rheumatoid arthritis) did not seem to assist or were not tried reliably.
  5.  Sarcoidosis: Patients with this diagnosis found that they obtained symptomatic relief from Lithium administered in much the same protocol to those for Rheumatoid arthritis.
  6.  Hepatitis C: Most patients with this condition were former or current drug addicts - not ideal patients ! The most important patient had been investigated in the early 1990s at an infectious diseases hospital. He had been rejected from an Interferon trial because he had mild cirrhosis on liver biopsy. His response to Lithium, as assessed by the serum ALT level was gratifying. However, at about day 10, he developed a self-limited T3 thyrotoxicosis, followed by an episode of paranoid ideation (possibly brought on by the thyrotoxicosis). After a partial improvement, he was lost to follow-up. However, 3 year later, he turned up again. The Lithium treatment result was similar to the first trial (but had a lower starting value). After a limited course, he moved interstate and was lost. All other patients had levels of the ALT enzyme much closer to the normal range, and there were problems with assessing compliance and response. However, some seemed to show falling ALT values. There are now scientific reasons to believe that Lithium may interfere with the Hepatitis C virus invasion of liver cells (see elsewhere[3]).
  7.  Polymyalgia rheumatica & Temporal arteritis. Lest there might be the impression that Lithium is the panacea for all conditions, known and unknown, reference can be to these related conditions. A number of patients with Polymyalgia rheumatica were treated with the Lithium Rx. No appreciable benefits could be identified, but there was no worsening either. When the Lithium Rx was given to patients carrying the diagnosis of Temporal arteritis and receiving tapering doses of Prednisolone, there was no appreciable hastening of the Prednisolone reduction (and no deterioration). Taking these observations, with the assumption then, that Lithium-induced Heparan sulphate perturbations do not modify the pathological bases of these conditions - the activation of vascular dendritic cells[4] may involve more direct or alternate stimulation (possibly non-Heparan-dependent receptors).  The (lack of) response to the Lithium Rx may have some use as a test to distinguish some diseases.

 

ELABORATION

Heparanase, (HPSE) the enzyme that splits Heparan chains, and produced in quantity by many cancer cells, inflammatory cells, and some other cells, has a rather complicated metabolic life[5]. It is assembled as an inactive pro-enzyme of 65 kDa in the Golgi apparatus. In cells that are active producers of it, it makes its way to the exterior of the cell, then to become attached, as the 65 kDa pro-enzyme, to the heparan chains on the Heparan sulphate proteoglycans* (HSPG) on the cell surface[6], the core proteins being chiefly Syndecans and Glypicans (Heparan sulphate [HS] proteoglycans; HSPGs[7]). The cell surface binding and endocytosis of Heparanase are far from clear :

  2008  Ben-Zaken et al. ". . heparanase is subjected to rapid and efficient cellular   uptake  mediated by cell surface HSPG of the syndecan family." The     reference is to 2004 :

  2004  Gingis-Velitski et al. "Syndecan" is not mentioned until the "Discussion:"   "Integrins, possibly cooperating with syndecan family members such as   syndecan -4 are likely to be involved and are currently under investigation."   Two references are quoted :

  2002[8] This is a review dealing with Syndecan-4 and its relationship with signalling   and structural proteins. "Heparanase" is not mentioned.

 1999[9] This is a research paper dealing with Syndecan-4, integrins, Rho, focal   adhesions and actin stress fibres. "Heparanase" is not mentioned.

So, based upon reference sources provided to date, there is little more than speculation that Syndecans might be involved ! (See later regarding Glypicans.) There is more general agreement that HSPGs are involved variously with  the LRP receptor[10] and Mannose-6-phosphate[11],[12] receptor [fibroblasts*], now identified as the Cation-independent form[13] (CIMPR; CD222; MPRCI; MPR300; MPR/IGFRII; but acting independently of the ligand's specific Mannose-6-Phosphate molecular group), and others[14] [non-endothelial]), possibly as a complex with lipid rafts[15](non-endothelial), which also play an important part in the local FGF/HSPG concentration and PKCα activation[16]. "Classical" Clathrin-based endocytosis tends to favour signalling pathways, whereas the lipid raft/Caveolin pathways[17] are more for degradation although, as for TGFβ, both pathways may be involved in proportions set by variable factors[18]. Since Glypicans lack an intracytoplasmic signalling domain (tail), other receptors need to be involved in order to initiate intracellular signalling. The integrins drawn into the lipid microdomains may have the role of signalling to the cytoplasmic tyrosine kinases[19]. Caveolae and Glypicans are usually associated with lipid rafts[20] on which Glypican-cytokine complexes may move over the cell surface, even activating receptors on adjacent cells[21]. Their involvement seems probable and the association of Heparanase, Glypican-1 and Caveolin-1 in late endosomes of human bladder cancer cells has been demonstrated[22]; Heparanase commences cleaving the Heparan chains whilst in the caveolae or caveosomes, before further degradation in the early and late endosomes, with transformed cells showing less control of the subsequent degradation[23] :    [*The Cell-type studied]

  Invagination Caveolae Endosomes Multivesicular endosomes Lysosomes

 

                        Caveosomes ( endoplasmic reticulum/other)

The above scheme provides a flexibility for cargo dispersal[24]; but note that disruption to these functions occurs with development of the transformed (malignant) state, in which expression of (say) the Mannose 6-phosphate receptor sets the level of maturity of the endosomal/lysosomal systems[25]; as if, in the "normal quiescent state" Heparanase is endocytosed and is processed into an active state, then enters a degradation stream in mature, late endosomes/lysosomes, whereas, in the transformed cells, Heparanase goes little further than the immature early endosomes, but is activated, and is then able to be secreted as such from the cells. The sugar receptors are of particular interest because rat lymphocytes carrying these can adhere to surfaces incorporating Fucoidan[26] [F. vesiculosus]) or Polyphosphomannan ester (which, in binding features that probably involve a Cation-dependent Mannose 6-Phosphate Receptor, has some similarities to the stronger, yet different, Fucoidan, which had a different receptor). Heparin did not prevent binding, but dextran sulphate blocked the Fucoidan binding, indicating specificities for charge and shape (see later, with respect to SREC, which may be involved).        

On the cell surface, Heparanase is bound to high affinity, low abundance receptors, such as the Mannose-6-phosphate receptor (CIMPR), and low affinity, high abundance receptors, such as HSPG; the HSPG being the most important at high HPSE levels[27], with the HSPG being saturated ultimately[28] (and probably incapacitated thereby, as far as being a co-receptor for ligands such as FGF). (See later with respect to rafts and endothelial cells). Fibroblasts may be able to convert the enzyme to the active form before internalization[29], an option not favoured, and bypassed for the (generally) immortalized cell lines used by others, perhaps because of unrecognized interactions with Hyaluronan/CD44 and the consequences (see later). HPSE, one or more receptors and HSPG are then internalized by endocytosis (probably by caveolae, but possibly also clathrin-mediated), together with tumour-secreted procathepsin L (at least, which may need a co-receptor[30]) and, when activated, favours membrane-associated substrates[31]. They are transferred to the late endosome/lysosome compartments (see later, with respect to Varicella). In the more acidic environment, enzymes of the cathepsin L & D type[32],[33] are activated and split the HPSE into three - a 50 kDa component, which combines with an 8 kDa second fragment, thereby forming an heterodimer with full enzyme activity, and a third 6 kDa fragment, which is expended. The active enzyme can be stored in perinuclear vesicles for 16-20 h until it then finds its way to the cell exterior in response to receptor stimuli, (such as nucleotides), that activate PKA &/or PKC. In macrophages stimulated with PMA, it moves with the aid of microtubules from intracellular stores  and concentrates in a "cap" on the cell membrane at the leading edge, the location being influenced by chemoattractants[34] (see later regarding TM1-MMP); the arrangement being consistent with an active enzyme patch positioned to dissolve the apposed ECM at the cell's leading edge, producing ~5,000 Da fragments of HSPG. How the HPSE attaches is unclear, but known receptors seem probable. It also finds its way into the nucleus (as do heparan fragments a core protein[35],[36],[37],[38],[39],[40] component).  

This pathway is considered by some to represent an overflow; that, in cells with little HPSE production, the pro-enzyme makes its way directly to the late endosome/lysosome organelles. The inactive HPSE on the cell surface may be activated directly there by proteases (refer to Hyaluronan & TM1-MMP later).

The active (50 kDa) HPSE can have two main actions :

  1.  Enzymatic, in which it splits HSPG into 2-3 large fragments of ~5-7 kDa, 10-20 sugar units. For FGF-2,  a ~pentasaccharide/hexasaccharide is needed for factor binding, but a dodecasaccharide is required for signalling[41],[42]; with greatest FGF2-receptor-assisted mitogenicity obtained with fragments 12-16 units long. This has similar activity to that obtained after Heparinase III action[43], which has a preferential cleavage point for linkages between the units GlcNS/NAc-GlcUA in the transition [T] zones between lowly and highly sulphated domains; the soluble low-sulphated fragments considered inhibitory, the high sulphated fragments stimulatory (the authors had what seems a simplistic approach to the sulphation issue).  HPSE enzyme action is inhibited by Heparin or, to a lesser degree, similar sulphated polysaccharides. The N-Sulphated group is important, but if replaced by N-Acetylation, O-Sulphation maintains an inhibitory function[44]. HPSE, in turn, has little effect on Heparin[45], with few -N-Ac present being suggested as a reason. (The enzyme roles may be contrasted with the Heparin augmentation of some non-enzymatic functions, see later.) The HPSE domain for adhesion to the substrate and the scission points is determined by the more highly sulphated domains* of the HSPG, with a -2OS (absent from the transition zones[46]) important for larger disaccharide chains[47] (of relevance, because the formation of the -2OS is the sulphation step most compromised by the gPAPP enzyme deficiency - see later), but with less well-defined criteria necessary for shorter (trisaccharide) fragments[48], which could have a blocking action. Studies on a cell line with defective 2OS formation showed no overall difference in disaccharide catabolism, but the cells deficient in 2OS lacked intermediate-sized fragments of ~4-7 kDa and ~10-20 kDa found in the wild type and attributed to the inability of Heparanase to utilize the chains as substrate within the endosomal compartments; and some chains were untouched[49]. (Once in the lysosomal compartment, other enzymes take over.) The Heparanase scission point has disaccharide group patterns on each side :

                                                                                                                                          scission

                          -GlcNS-IdoA2OS-GlcNS-HexUA-GlcNAc-GlcUA -GlcNS-IdoA2OS-X-X


    The resultant fragments would be expected to have a sulphate-rich domain at one or near both ends: sulphated groups at one end may block binding associations, whereas, at both ends, the molecules may link binding groups. If growth factors are internalized by attachment to the Heparan, their release in the endosomal compartment may be compromised if the chains are modified or defective. The active enzyme stimulates Syndecan (an HSPG core protein) synthesis partly by increased shedding[50] from the membrane, yet maintains the membrane Syndecan attachment at a similar level. Associated with these effects, there is inhibition of the movement of cell-surface Syndecan to the nucleus[51], which may influence gene expression. However, the HPSE enzyme action on Heparan (on separated tissues as opposed to action on Heparin) may not be as simple as has been given above - a trisulphated disaccharide may be produced that is optimally inhibitory to activated lymphocytes, and suppresses delayed-type hypersensitivity reactions. The conversion of inactive TNF-α to the active form may be inhibited[52] (see later under HSPG Heparan chain fragments). The HPSE-HS scission may not be "clean," and the ramifications of sulphation modulations are difficult to assess.

  2. Non-enzymatic, where it acts as a ligand for one or more receptors that are not well characterized. By this, it activates Akt phosphorylation (a "survival" [anti-apoptosis] factor[53]) and cell motility in endothelial cells by a process that is augmented by Heparin or another sulphated polysaccharide[54] such as Laminarin sulphate, potentially mimicking a role for Fucoidan (see later) or the Heparan sulphate fragments* released by heparanase enzyme action. Surprisingly, mice homozygous for the inactive HPSE genes (HPSE-/-) appear normal phenotypically. However, Heparan chain lengths were increased, and there was an increase in the production of Metalloproteinases (especially MMP-2 & MMP-14),with augmented breast and endothelial responses to cytokines. Direct binding of β-catenin to the MMP-14 gene regulatory sequences (at least) was detected, thought to provide a link between the control of HPSE and control of some of the MMPs (see later)[55].  

    High levels of HPSE may increase malignant cell adhesion at the expense of proliferation[56], and secreted forms promote the vascularity of tumours and metastasis[57]. Induced HPSE production with transgenic mice produced effects in the liver that differed to those in  other tissues[58]: in the liver, HSPG turnover was increased, as was N-sulphation and O-sulphation, with smudging of the sulphated and non-sulphated domains, thereby resembling heparin. In other tissues and in the few human tumours studied, the domains remained delineated, but there was increased 6-OS. The heavily sulphated domains, including those in the fragments produced by HPSE at the specific HPSE scission-point in the sulphated domain[59], promoted FGF signalling strongly. Interestingly, the Golgi was found to contain Heparan fragments. The cause for this was not clear (the authors considered that it was an HPSE enzyme effect). There is the possibility that stressed production of HSPG may result in disorganized disaccharide chain formation, because of the competition within the "GAGosome" in the Golgi[60], with resulting fragments. The increased HPSE gene activity, with increased HPSE and Syndecan production[61], may stimulate and disrupt HSPG production and sulphation, and could account for some of the differing findings relating to HSPG structures and sulphation in animal and human cancers.


    Sulfs[62] The enzymes Sulf 1 and Sulf 2 reside on the exterior of the cell plasma membrane. They split the -6O-Sulphate bond from the trisulphated disaccharides IdoA2S-GlcNS6S, and are believed to have a "fine adjustment" role. In embryology development (quail), Sulf1 responded to Sonic hedgehog* signalling, and had a selective, positive role in Wnt1 signalling[63] with Heparan HSPG* required as a co-factor. The suggestion was that Wnt1 is widely dispersed in the ECM, but was activated locally by liberation from the HSPG binding by Sulf. If the Heparan carried by the core proteins is deficient in trisulphated disaccharides, as occurs with gPAPP deficiency (see later), the signals arriving via Sonic hedgehog signals may be impeded, and the Heparan chains attached to the core proteins now poor binders of Wnt, with the result that the Sulf function of releasing Wnt locally for signalling may be impeded. Expression of the Sulfs in inflammatory and cancer tissues is variable. Sulf1 has been shown to be increased appreciably in pancreatic cancer, and confers some survival advantage[64]. Both Sulfs seem to have a negative role on cancer multiple myeloma), with their action limited to the cell membrane, leaving an increased matrix filled with plentiful 6OS still[65] (see later, with respect to Syndecan).

       

      FIGURE 1

                                                           

      Figure1: The changes to the sulphation status of various disaccharides following Sulf1 expression. There is a decrease of the ΛUA-GlcNS6S, with a corresponding increase of the ΛUA2S-GlcNS; the Sulf2 action generally less (eg ΛUA-GlcAc6S). (From Dai et al. 2005)

       

      Expression of Sulf1 (to reduce -6OS) reduced the signalling by FGF2, but not that by EGF, HB-EGF or IGF-1. The suggestion was that its expression may be linked to MYC expression, which may have relevance with respect to the gene transcription factors TCF and LEF (see later). When over-expressed in hepatic cell carcinoma cells, there was reduced growth of the malignant cells, increased histone acetylation and enhanced apoptosis induced by HDAC inhibitors. Enzyme knockdown was associated with up-regulated Akt (which may sensitize the β-catenin Wnt pathway) and ERK[66]. A contrary outcome was identified for Sulf2, with increased levels identified in patients having hepatocellular cancer, and the over-expression of Sulf2 resulted in increased tumour growth, migration and increased binding of FGF2. Knockdown reversed these effects. The results on Akt and ERK were similar to those with Sulf1. The core HSPG protein Glypican, was found to be up-regulated also[67] (sadly, there was no study of the HS chain make-up). The potential interest lies in the decreased growth stimulation  produced by the increase of Sulf1, and the (presumed) specific reduction of -6OS; and a decreased Akt activation with Sulf1 reduction by knockdown. As will be shown later, the Lithium action on the enzyme gPAPP results in a relative increase of -6OS (see later), particularly on the  Glucosamine-N-sulphate saccharide residue (and to a lesser degree, Glucosamine-N-acetate). Just why the Sulfs have opposing actions is not clear. Perhaps the ability of Sulf2 to up-regulate Glypican expression may compensate for the -6OS loss; the Glypican providing more binding sites for HPSE, probably involving -2OS and not appreciably affected by Sulf2 enzyme action, but possibly a target for reduction by the Lithium effect.


      Transmembrane Metalloproteinase 

      At the leading edge of invading malignant cells there are specialized projections referred-to as invadopodia. Attached to, and drawn to the cell membrane there, in association with caveolae[68], are a particular forms of metalloproteinase (TM1-MMP)[69]. They bind to a linking protein TIM-2, and the complex then splits MMPs that are free or bound, to the ECM, such as MMP-2 (Gelatinase A), letting the ECM be degraded both by TM1-MMP and MMP-2, opening the ECM for the cellular invasion[70]. As with HPSE, there may be a non-enzymatic role, with MT1-MMP up-regulating NF-κB, which then up-regulates COX-2, and thence Prostaglandin E2, providing tumour growth stimuli[71]. These transmembrane enzymes require the middle part of their cytoplasmic tails to be mobilized to the leading edge, by a mechanism that is unclear. There is negative control by auto-degradation[72].

      To date there has been no mention of the involvement of HSPG. The close co-localization of the MT1-MMP and HPSE provide reason to believe that there is some cytoskeletal association over and above the gene transcription link that has been mentioned already derived from the HPSE-/- mice (see earlier).  

       

      Endothelial Cells 

      Much of the work with Heparanase has been done with fibroblasts or epithelial cells (eg CHO) that do not naturally express receptors for the systems under examination. Throughout, Heparanase production has been associated with vascular proliferation, with imputed agents being the VEGF family in particular[73]. Relevant receptors have been found that associate with other receptors, forming complexes. A cluster type with a "hub" comprises Neuropilin-1 (NRP1), VEGFR, Semaphorin, plexin[74],[75] and the Scavenger receptor of endothelial cells (SREC1)[76],[77],[78]. SREC, which is not an endocytic Advanced Glycation Endproduct's (AGE) receptor[79], had binding with a test agonist inhibited by Acetylated-LDL, Dextran sulphate, polyinosinic acid, and partially, by Ox-LDL (ie rather like SR-AI/II). Heparin was not inhibitory (ie does not bind) but it does bind to NRP1, VEGF and VEGFR. NRP1 expression has been found in bone marrow stem cells and it can act as a receptor leading to release of (at least) two cytokines for proliferation. (It was also found in haematopoietic supporting cells and stromal cells, but not in progenitor cells[80].) These complexes bind various agonists singly or in combination, endocytose and escort them to the endosomes for dispersal or breakdown. In this way one receptor pathway may inhibit and take dominance over another. VEGFR2 associates with NRP1 and, with VEGFA binding, all are endocytosed by a clathrin-dependent process, migrating initially along actin guiding paths, then later along microtubular tracts. Semaphorin also attaches to the VEGFR and NRP1 association, but all enter the cell by means of lipid rafts (which may involve caveolae and Glypican). Whilst HPSE was found to enter cells in association with lipid rafts, there has been some uncertainty about receptor specificity and the type of endocytosis.  If the HPSE entry is similar to that of Semaphorin, the VEGF drawn into the cell by the HPSE's entry, could explain the microvascular proliferation that HPSE stimulates. Both NRP1 and VEGF can bind Heparin, which augments the combination and, given the known co-factor role for Heparan sulphate for VEGF, Heparan may be assumed to be involved. There will be more on SREC later.

      Selectins are membrane-bound proteins with lectin domains. They are involved in the rolling of leucocytes against the capillary endothelial-lined walls. Mutated deficiencies result in an increase in circulating leucocytes and haematopoietic progenitor cells with a thrombocytopenia (in mice). Fucoidan exacerbates the leucocytosis and circulating stem cell increase through a process that may, in part, involve the Selectins, having a stimulus that was dependent on sulphation, but not necessarily of Fucoidan (but not with Heparin)[81].  Interestingly, the leucocytosis response was appreciably bester after two doses of Fucoidan 6 h apart (as opposed to 6 doses over 2 days). Details were lacking, but this may have similarities with the bolus dose responses to Heparin fragments (Disaccharides) and Lithium for EAE (see later).


      Chondroitin sulphate and Hyaluronan

      A prominent feature of the gPAPP-/- phenotype in mice was abnormality in cartilage formation attributed to defective chondroitin sulphate production. The growth plates of bones have, as a predominant HSPG core protein, Perlecan. Chondroitin sulphate can form the Glycosaminoglycan (GAG) to varying degrees, from ~nil near endothelial cells, to ~75% in cartilage. Perlecan in the growth plates can modulate the delivery of growth factors such as FGF to their receptors at two levels; first, FGF is bound to the Heparan sulphate chains on the Perlecan, but cannot reach the receptors because the bulk and shape of the Chondroitin sulphate chains obstruct; secondly, the Chondroitin sulphate chains are removed by enzyme action, allowing the freed FGF to bind to the receptor[82]. The bone growth defects of the gPAPP-/- mice may involve either a failure of the HSPG to collect in bound form, the FGF other growth factors, as a cytokine sink, &/or an inability to remove the Chondroitin sulphate chains adequately, because the GAG structure prevents the enzyme action and, thereby, not releasing the growth factor(s). Chondroitin sulphate is also involved in the minor proteoglycan, Versican, which is involved in the cross-linkage of the Hyaluronan chains. In turn, these are involved in cell shape and motility, and provide an extracellular environment for the HSPG/Syndecan/Glypican complexes to operate[83]. Chondroitin sulphate is also associated with the intracellular movements of Ii from the Golgi to the cell surface and then internalized, which seems linked to MHC II/ligand interactions, with immune functioning implications[84]; both dendritic cells and macrophages express appreciable amounts[85], in Glypicans particularly, but also Syndecans, the functions being largely unknown. The gPAPP effects involving Chondroitin sulphate may be subtle, but worth considering, since some Syndecans include Chondroitin sulphate (close to the cell membrane), in addition to the Heparan sulphate chains[86].

      Hyaluronan (hyaluronic acid) is likely to interrelate with the HPSE/HSPG pathways. Whilst Hyaluronan is secreted from enzymes beneath the plasma membrane, rather than the Golgi, there are features that are similar[87] including endocytosis involving lipid rafts and caveolae. Of particular interest is the ability of small fragments of Hyaluronan to stimulate dendritic cell activity, resulting in the production of TNF-α that stimulates surrounding cells[88]. The ability of Hyaluronan to activate its receptor CD44 in association with caveolae and lipid rafts results in a cascade of events, including local acidification, which provides a working environment for Heparan-binding cathepsin B, which may activate the pro-HPSE at the plasma membrane.  


      The Wnt + PKC pathway.

      In Drosophila, the gene wg and its products Wingless (Wg) need specific HSPGs for normal  larval development[89]. The homologous Mammalian ligands, Wnt(s), are cytokines that activates pathways leading to the activation of key genes in the nucleus. There are two main branches[90]:

      1.  The "canonical" pathway involving β-catenin, leading to the transcription factors TCF and LEF. Among the genes that may be activated are GLCE, gene for the epimerase in the Golgi[91], Sdc4, the gene for Syndecan-4, also MYC and cyclin D1, together with Egr1, which plays a key regulatory role in inducible HPSE gene transcription[92], an indirect target[93]. Lithium part-activates this pathway by suppressing GSK3β, a function augmented by the sensitizing effect of the kinases PKC[94], PKA and Akt[95]. This augmentation by PKC may be nullified by "chronic" (viz 24 h) PKC pathway stimulation (a factor to consider in intermittent Lithium dosing). So, the canonical Wnt pathway, when stimulated, may be expected to increase the transcription of epimerase, which may aid competent HSPG production and also the generation of pro-HPSE. Wnt signalling is considered an important factor in osteoblastic metastases[96]. Blocking the pathway may be expected to decrease the production of growth and signalling factors.
      2.  The Wnt/Calcium branch responds to stimulation with activation of PKC, and a rise in cytoplasmic Calcium. The rise in the activated PKC may also suppress GSK3β, augmenting the Wnt signals by  activating the "canonical" Wnt pathway. ThCalcium signal activates the Calcineurin enzymatic removal of the phosphate from NFAT, which then moves into the nucleus for its transcription factor role[97]. The rise in Calcium and PKC may stimulate Cathepsin exocytosis which may, in turn, activate HPSE at the plasma membrane[98]. There may be another Wnt-dependent variant non-canonical pathway that could be activated by Lithium and transfer NFAT to the nucleus[99].

       

      Inflammation 

      Endothelium-derived HPSE was shown to be an important factor in delayed-type hypersensitivity in mice[100], stimulated &/or augmented by TNF-α and Interferon-γ (presumably influenced by the sulphation characteristics of the HSPGs that surround the cells).  Its suggested role was to improve inflammatory cell adhesion and migration, with associated dissolution of the HSPG ECM. However, the importance of HPSE in delayed-type hypersensitivity has been questioned by the HPSE-/- mice (see earlier).

      HSPG (core protein + GAG chains) has been shown to induce TNF-α and NO2  transcription and their production in murine microglial cells. This was taken to reflect inflammatory stimulation[101]. The heat-denatured HSPG and the GAG chains alone produced significantly inferior responses, pointing to the intact, complete HSPG for inflammatory activity - mechanism unclear. Contrasting with this, murine bone marrow immature dendritic cells, when presented with Heparan sulphate chains alone, stimulated the dendritic cells to maturity and the production of effectors such as TNF-α,  IL-1β & IL-6[102].

       

      Inflammation-promoting effects of Glycosaminoglycans on Murine macrophages.

      Murine peritoneal cells, quiescent after initial stimulation (for harvest), were activated to produce TNF-α, IL-1 and IL-6 by Heparan GAGs and full length Hyaluronan chains[103].    

      Further testing, and using mixed lymphocyte cultures[104] with GAG chains that were largely intact, produced the following effects :

       

      TABLE 1

       

                                  IL-1      IL-4       IL-6       IL-12    TNF-α   IFN-γ    TGF-β   PGE2    CTX       NO        ICAM   I-A

      HS

       +

       

       +

       +

       +

       ()

       +

       +

       +

       

       +

       +

      Hep

       

       

       

       +

       

        +

       

       

       

       

       

       

      CS

       

       

       

       

       

       

       

       

       

       +

       

       

      DS

       

       

       

       

       

       

       

       

       

       +

       

       

      HA

       +

       

       +

       

       +

       

       

       

       

       

       

       

      IL-1 includes IL-1α & β; CTX = cytotoxicity; NO = nitrous oxide; HS = Heparan sulphate; Hep = Heparin; CS = Chondroitin sulphate; DS = Dermatan sulphate; HA = Hyaluronan; () to + = stimulation, mild to appreciable; = no appreciable effect; = suppress

      B cells were stimulated to produce IL-1α & β, IL-6 and TNF-α

       

      HSPG core proteins - Syndecan & Glypican. These are constitutively shed (~5%/h for endothelial cells), but the shedding is accelerated (rapidly, 5-30 min; T1/2 = 30 min) by antibody & complement on endothelial cells[105], or by the selective activation of various receptors, often selectively, associated with the activation of PKC and operating by different serine &/or cysteine proteinases or metalloproteinase(s)[106], severing the core protein near the plasma membrane. Whilst detected within human wounds, the core proteins are not (easily) detected in the general circulation[107]. Once the Syndecans and their attached GAGs are soluble and on the loose, they can bind growth factors and block the factors from reaching their receptors, or intercept the factors and lead them to the receptor. Transgenic mice producing mammary Wnt-1 showed an increased incidence of breast cancers that was prevented by deficiency of Syndecan-1, and Syndecan-deficient Drosophila could not respond to Wg; to do so required Syndecan equipped with HS GAG[108].  In both species, congenital deficiency of Syndecan could, in general, be overcome, presumably by other HSPG's, pointing to redundancy and indicating that disturbances to established pathways may cause only temporary upsets. LasA, a protein virulence toxin produced by Pseudomonas aeruginosa, specifically stimulates the cellular Syndecan shedding mechanism[109]. In a murine bleomycin-induced pulmonary inflammation and fibrosis study, the Heparan GAG of pulmonary epithelial cell Syndecan-1 formed a key bridge for the sheddase matrilysin (MMP-7) and KC (a murine homologue of IL-8) in allowing neutrophils to pass from the interstitial space into the alveoli[110]. Syndecan (+GAG) seems necessary to prime leukocytes (that do not express it) to modulate downwards, their adhesiveness towards endothelial cells in the retina of mice[111]. This may occur in the bone marrow. Multiple myeloma cells generally express Syndecan-1 that, when shed into the surrounding stroma, sets up stromal reactions[112].  In a murine model of multiple myeloma, cells that produced a soluble Syndecan-1 (a secreted, truncated form) formed bigger tumours and metastasized more readily than the cells without the HSPG or the wild type. The secreted Syndecan collected in the matrix[113].

      Syndecan-4, found on T cells, can bind for adhesion with integrins through the GAG and possibly the HSPG core protein of activated T cells, to DC-HIL expressed by dendritic cells, a process blocked by Heparin and other sulphated polysaccharides, probably involving multiple sites[114]. This inhibits the lymphocyte response; when the process is blocked, there is proliferation of lymphocytes in, and enlargement of lymph nodes[115],[116]. In mice that were unstimulated, DC-HIL RNA was found mostly in the bone marrow and adipose tissue, with modest amounts in the thymus, and minor amount in the skin, attributed largely to Langerhans cells. Distribution did not match the DC distribution, indicating that production is probably restricted to certain subtypes of dendritic cells.


      Perlecan

      Perlecan is another HSPG of basement membranes and cartilage; being a major matrix component. Heparan sulphate is the principal disaccharide form, but in cartilage in particular, Chondroitin sulphate is prominent also. The core protein is subject to cleavage by various enzymes. The Heparan sulphate chains bind FGF, which can be displaced by a number of enzyme actions[117].                 

        

      HSPG Heparan chain fragments

      As with Hyaluronan fragments, Heparan sulphate chains (bovine kidney, from Sigma; chain length as supplied commercially) proved a powerful stimulus to cause maturation of immature murine bone marrow dendritic cells[118]. These cells then produced TNF-α, IL-1β and IL-6. This response could be involved in many inflammatory and malignant conditions, when HPSE is released.

      Disaccharides with varying degrees of sulphation have different effects; those produced from Heparin included the disulphated UA2S-GlcNS and unsulphated UA-GlcNAc, and  have been shown to reduce TNF-α associated inflammation (induced arthritis) and delayed type hypersensitivity in mice. For inhibiting the latter, the disulphated disaccharides were most effective at low concentration. Even the non-sulphated units showed some effect. Dosing, both temporal and amount, was critical. Weekly bolus dosing was better than daily doses, and the responses fell on either side of the optimum dose. One explanation may be that the disulphated disaccharides bind to the the more avid-binding partner of a ligand/receptor pair until all sites are saturated, and then it starts binding to the other. Whilst on one partner only, binding is blocked; when on the other, proteins, or other molecules able to cross-link, complete the binding, which increases as more disaccharide units attach. On this basis, monosaccharides would block and trisaccharides would show binding, but at a level less than for the disaccharides because of charge or steric hindrance. (The dose could be oral, but needed to be bigger[119].) These observations are of particular significance, because they show some similarity with the features of Lithium bolus dosing used on EAE (see later) and the involvement of HPSE in delayed-type hypersensitivity. In a later study, some disaccharides derived from Heparan and Heparin were found to inhibit the production of IL-8 and IL-1β from transformed human bowel and breast epithelial cells[120] when applied alone or with TNF-α-induction. Generally, those that did not inhibit instead blocked those that did. These results indicated that cell receptors were involved. The inhibition blocked egress from the cells, rather than transcription/translation. The inhibitory Heparan-derived disaccharides included trisulphated disaccharides and also bi-sulphated forms, with Iduronic acid sulphated. The -6OS was necessary for the mono-sulphated units' inhibition. Acetyl groups on the amino of Glucosamine, a -6OS and/or Iduronic -2OS tended to be non-inhibitory

      Later studies using Heparin-derived Disaccharides modified integrins, cell adhesion and chemotaxis by what appears to be by receptor mechanisms[121], initially stimulating adhesion, but later inhibiting adhesion and chemotactic migration through fibronectin. Trisulphated disaccharide (Ido2S-GlcNS6S) derived from the enzyme action on Heparin down-regulated NF-κB of activated lymphocytes, suppressing TNF-α and IFN-γ expression. The monosaccharide (Idu2S-GlcN) was without appreciable effect. CXCL12 (SDF-1)-induced migration was suppressed[122]. So, the effects of Disaccharides may be variable, with the sustained exposure to the more sulphated forms, probably immune suppressive.

          

      The bone marrow, cancer and Lithium

      There are good reasons to believe that HPSE may be a major factor in the observed (cancer + Lithium) marrow response (Patient PC):

      • Prostate cancer cells have been documented to produce and release HPSE[123],[124], as have the cells of many cancers[125],[126],[127]
      • HPSE in the plasma of patients with cancer (comprising cancers, leukaemias, lymphomas and sarcomata) can be measured[128] and related to the disease state
      • HPSE from a non-osseous location can influence, from afar, the activity of osteoclasts and produce an associated bone remodeling[129]. How this occurs is unclear; the suggestion being that IL-8 (&/or another), also produced by the tumour, binds to, and is carried to the bone marrow environment by fragments of heparan split off from the HSPG by the active HPSE. This assumes that the IL-8 can retain in vivo effectiveness right up to its receptor in the bone marrow. The in vitro testing involved peripheral blood monocytes (see below). (Sadly, the effects upon other bone morrow progenitors were not examined.)
      • Osteoclasts are derived from the myelo-monocytic series and have close links with bone marrow progenitors, dendritic cells and activated CD4 lymphocytes[130],[131], probably involving M-CSF and RANKL
      • All relevant cell types, myeloid and lymphocytic, are derived from stem cells and progenitors arising in the bone marrow[132], with the pleuripotential genetic expression capability for those progenitors that reach the thymus retained until the DN2 stage[133]. In the adult bone marrow, the haematological stem cells produce progeny with leanings towards various further lineage differentiation; the CD34+ Lin- CD10+ CD45RA+ representing the progenitor line for T and B lymphocytes, NK cells and dendritic cells[134]. Further work indicates that the CD34+ Lin- CD10+ group can be more accurately characterized[135]; whilst those that are CD24+ are B cell committed, those that are CD24 develop into B, T and NK cells. They can be found in peripheral blood throughout life (1 in 5,700 mononuclear cells of adults) and provide a progenitor line for the thymus.  They showed virtually no erythroid and minimal myeloid potential. Those representing the CD10+ group in the thymus could still differentiate towards B cells, but this seemed limited. Comparing the differentiation responsiveness of the CD34+ CD10+ CD14 group, there is considerable similarity with the noted lymphocyte subtype responses with the Lithium Rx tumour :


      TABLE 2

       

       Progenitor cell differentiation and the marrow response to Lithium Rx+tumour

       

      Erythroid

      Myeloid

      Lymphocyte T

      Lymphocyte B

      NK Cells

      CD34+ CD10+ CD24

      Nil

      Small

      Strong

      Weak

      Strong

      Patient PC

      Not done

      Not done

      Strong

      CD8>CD4

      Weak

      Strong

      Patient LC*

      ~Nil

      Weak  

      Weak  

      CD8>CD4 

      Weak Moderate

      *Patient LC. Most of her measurements at day zero read higher than the earlier and subsequent readings - she may have been dehydrated, and results have been so interpreted. As with patient PC, the CD8 and NK cell subtypes showed the best responses.

       

      The CD34+ CD10+ CD14cell line is likely to be the one that is involved in the Lithium/cancer lymphocyte/NK cell responses that have been observed and presented here.

      • HSPGs must play an important role in the maintenance of bone marrow stem cells in the bone marrow niches; cells with a reduction in Heparan sulphate chain size (length) from ~50-150 to ~35 kDa proliferate slowly and are unable to continue stem cell status[136] (in mice), with signaling by FGF and Wnt reduced. HSPG seems important in haematological progenitor cell-endothelial cell relationships in the bone marrow, particularly in relation to cell homing[137], and the 6-OS sulphation pattern is important[138] to continue stem cell maintenance. This is in addition to the requirement for HPSE to permit movement of marrow cells into the blood vessels in order to leave the marrow environment. (See later in relation to Fucoidan)
      • There may be differences between the ternary complex requirements for different ligands, their receptors (including their subtypes). Based upon the in vitro study of a range of oligopolysaccharides[139] and their binding to the FGFs and FGFRs, the most important parameters are the oligosaccharide abundance, the length of the chain and the degree of -O-sulphation within the Glucosamine-N-sulphate domains. This study has particular relevance in considering the roles for the Heparan sulphate fragments produced by HPSE - how reduced sulphation in fragments may fail to signal, or may block signalling coming usually via the more complete HSPG chains: the HPSE produced by cancers and conditions such as rheumatoid arthritis[140] and (presumably) multiple sclerosis (based upon studies of experimental allergic encephalomyelitis[141]) in their active states, will be expected to split off (generally) well-sulphated oligosaccharides, which may be expected to aid stimulation of signalling via the ternary receptor complex. When the HSPG chains have defective sulphation - poor domain delineation, decreased frequency of sulphation - as a result of Lithium inhibiting gPAPP[142], and there may be implied back-flow effects that may inhibit the enzyme action of NDST due to restricted  deacetyl group replacement with a sulphate group[143], GAGs may lack the sulphation features for HPSE to identify correctly the scission points. Consequently, those HPSE-produced fragments that are formed (if they can be) may be more variable in size and with reduced -2-O-sulphation, implying a considerably reduced molecular sugar ring (Iduronic chair/skew-boat) flexibility[144]. Being soluble and free, they may block signalling because of defective binding to ligands &/or receptors. Since the production of HSPG in the Golgi and their release to the cell surface occurs within minutes[145] (10-15 min.), a Lithium effect may occur within an hour of a dose. A subsequent progressive modification of the preexisting HSPG chains on the cell surface would be expected to be increasing over time until the Lithium level falls, and the HSPG chains may show some return to normal.
      • During the murine embryo brain maturation occurring between day 10 (E10) and E12, there is a reduction in cell division and a change in GAG structure, which correlates with a change in FGF subtype influence[146]. The main maturation changes include lengthening of the GAG chains, with an increase in the sulphated domains and an increase in 6-O-Sulphate groups. Since cancers generally show a regression to more embryonal features, finding such GAG changes in reverse would generally be expected. However, when comparing the GAG's of HSPG associated with human colon cancer cells and adenoma cell, the chain lengths were essentially similar (45-50 disaccharides, ~20 kDa) with ~2 sulphated domains (the target for Heparinase I and HPSE) reasonably evenly spaced which, when cleaved, produced fragments ~7 kDa (~15 disaccharides[147]). However, the cancers showed overall sulphate reduction of 20%, particularly of the iduronic acid-2-OS, indicating loss in the sulphated domain, whereas 6OS was increased, probably in the intermediate, transitional and mixed sulphated segments on either side of the sulphated domains. (Contrasting with this are two studies of transformed murine mammary cells, the first not using heparinases[148]. The GAG chains were found to be shorter [~20%] and less sulphated, both -NS and -OS. Also, a study of testosterone-induced malignant transformation in a mouse mammary[149] cell line found reduced sulphation, but mainly of the GlcN6S units, with little change in the 2-OS or NS unit. These differences may reflect Heparanase effects, species, organ  &/or stimulus differences.)
      • More recent studies of the (neuroectodermal) stem cell HSPGs and their maturation to progenitor stage, found that the stem cell GAG chains were long and poorly sulphated. Sulphation, especially of -6OS, came with cell maturation. The hypothesis derived from this, (and study of EXT-deficient stem cells) is that stem cells protect themselves from growth factors by having a shield of low-sulphated HSPG surrounding them[150]. Since many cancers are now believed to have cancer stem cells, the potential for Lithium to reduce the level of sulphation by inhibiting gPAPP may cause a reduction of some stimuli for the cancer stem cells to progress to progenitor status, yet may allow other stimuli through to the cells. (See later with regard to Fucoidan)
      • Examining the effects of oligopolysaccharides of heparin obtained by the use of different forms of Heparinase (I & III) on the growth of tumours in vivo, found that the degradation products using Heparinase III inhibited tumour growth (good), whereas those from Heparinase I encouraged or promoted growth (bad)[151].
      •  The sulphation patterns of these are presented, and are compared to the patterns from the foetal mice with gPAPP deficiency, heterozygous[152] and homozygous NDST deficiency (x2)[153], epimerase deficiency and 2-O-Sulphotransferase deficiency[154] (see Figure 1 below).


        FIGURE 2

                                                            

      Figure: The compositions of the main disaccharide groups in the lungs of foetal mice are presented as above or below the levels found in the wild types; chart bar groups gPAPP (left) to Hs2ST (right of centre). The figures for DUA2S-GlcNS6S were not done. However, these were measured for whole embryos, with wild type 4.9% to gPAPP-/- 3.8%, a fall to 77.6% of the wild type level. The 2 bar groups to the right of centre (Hep I & III) present the measured components in the enzyme digests of heparin used to test tumour growth in vivo and in vitro. These have been estimated from the published charts, reading peak heights. (There was no peak labeled DUA-GlcNAc.) The two bar groups on the right present the measured components expressed as percentages. (The percentage of DUA-GlcNS6S III/I, 1781%, has been truncated to 390% to fit.) [gPAPP=Golgi 3'-phosphoadenosine-5'-phosphate phosphatase, or −/−  = homozygous null genotype; NDST= N-deacetyl-sulphotransferase; a= figures from Grobe et al; b=figures from Ledin et al.; Hsep = Heparan sulphate epimerase; Hep = Heparinase product, see text.]

       

      In order to understand the changes, the outline of the enzymatic steps in Heparan chain formation is presented :

       

      The main steps in the HSPG chain construction pathway :

       

      Disaccharide chain polymerase

       ↓

                      GlcNAc N-deacetylase/N-sulphotransferase (NDST) + PAPS (S donor to -N- groups)

      GlcAc C5 Epimerase (Hsep)-(Changes orientation of C5 -OH)

      2-O-Sulphotransferase (Hs2ST) + PAPS (Sulphate donor to C2-OH)

      6-O-Sulphotransferase (Hs6ST) + PAPS (Sulphate donor to C6-OH)

      3-O-Sulphotransferase (Hs3ST) + PAPS (Sulphate donor to C3-OH)

       

      Highlighted are the sulphation steps that seem most affected by the gPAPP-/- genotype[155]. Of particular significance is the affect that PAPS (the sulphate donor) has at the NDST level, because deficiency of the sulphate donor results in poor sequential sulphate domain delineation along the developing Heparan chains[156], a disorganization that is expected to influence all later enzyme steps in the formation of the chains. In addition, there are examples where downstream perturbations influence upstream events[157],[158], supporting the concept of a "GAGosome."

      The steps in the formation of Chondroitin sulphate are less well understood. However, the outcomes of the gPAPP deficiency are reasonably clear-cut (sadly, the authors did not measure Heparan & Chondroitin chain lengths, a measure that would provide more information of chain domain structure):

       

      TABLE 3

       

           CHONDROITIN SULPHATE - SULPHATION CHANGE TO WILD TYPE

       

      Mouse Tissue                           ΔUA-GalNAc                                          ΔUA-GalNAc6S                                      ΔUA-GalNAc4S   

      WHOLE

            Wild type   

            gPAPP-/-

      10.4

      38.4

            Change = 369%

      1.7

      2.5

            Change = 147%

      87.9

      59.0

              Change = 67%

      LUNG

           Wild type

           gPAPP-/-

      21.9

      35.1

            Change = 160%

      12.4

      18.5

            Change = 149%

      65.7

      46.5

              Change = 71%

      Figures are from Frederick et al. 2008. Measurements from each tissue represent % of total.

       

       

      TABLE 4

       

      Correlations between the gPAPP effects and other enzyme effects:

        

      Enzyme/genetic status

      Correlation coefficient to gPAPP

      NDST+/-

      0.58291

      NDST-/-(a)

      0.91054

      NDST-/-(b)

      0.93746

      Hsep-/-

      0.04304

      Hs2ST-/-

      0.07916

      Hep I

      −0.6695

      Hep III

      0.69413

      Hep III/I %

      0.81198

      Hep I/III %

      −0.7946

       

      Examination of the disaccharide patterns may allow us to draw some conclusions :

      •  The phenotype changes produced by the mouse gPAPP-/- genotype,when compared to the phenotype changes associated with some other Golgi enzyme deficiencies, most closely resembles the changes resulting from lack of the NDST enzyme - with a build-up of the unsubstituted disaccharides (DUA-GlcNAc) and a lack of certain sulphated disaccharides (DUA-GlcNS & DUA2S-GlcNS)
      •  The reduction in DUA-GlcNS may be expected to affect the length and possibly the uniformity of the sulphated domains. The transitional domains on either side may be shortened, poorly developed and not bind proteins well[159].
      •  The particular relative deficiency of DUA2S-GlcNS would be consistent with a considerable disruption of the normal pattern within the sulphated domains[160], with an inability to maintain the Iduronic acid conformation there and hence provide the usual flexibility within that domain[161].
      •  The relative increases of the -6OS groups would be expect to be in the sulphated and transitional domains (at least) and can be regarded as an adaption means to compensate and maintain the required molecular charge.
      •  Trisulphated disaccharides (a feature of Heparin, but much less frequent in HSPGs), would be reduced.

       

      Heparin fragments on tumours

      Also shown are the analysis results on the Heparin breakdown products produced by the two enzymes Heparinase I & III. These enzymes split the molecular chains at different points, resulting in fragments with different structures. When they were administered to animals with cancers, the products from Heparinase III [split between GlcNAc/NS(6S)IdoA/GlcA, tending to leave intact the sulphated domain] retarded the tumour progression (good) whereas, those from the Heparinase I [split generally between GlcNAc/NS(6S)IdoA(2S), and limited to breaking up the sulphated domain, where the HPSE scission point is], stimulated the tumour growth (bad).  When the sulphation characteristics of the fragments are compared with the HSPG disaccharides from the enzyme-deficient mice, the ratio Heparinase III/I showed a positive correlation with the gPAPP-/- deficiency values, whereas Heparinase I/III showed a negative correlation. This is a quite artificial study, but at least it shows that the basic features of the gPAPP deficiency might translate into tumour retardation - one would be concerned if it were the other way around. Similar conclusions were derived from Heparanase III treatment of mice carrying human-derived multiple myeloma[162]. The conclusion was that the expressed Syndecan-1, with its Heparan chains subjected to Heparanase (which split the chains at different sites), provided a stimulus for tumours, and that treatments aimed at the GAGs of Syndecan could offer prospects.


      NK cells and their cancericidal function

      NK cells have innate immune activation which appears to involve, at least to a considerable degree, the NKR-P1 receptor. Saccharides, olig-, di- and mono-, bind to the receptor, with the Heparin-derived trisulphated disaccharide DUA2S-GlcNS6S strongly active. An -6OS seemed frequently involved. Ligand binding to cells could be stimulatory or inhibitory; ligands that were stimulatory included Heparin-derived or Chondroitin sulphate-derived and lipid-associated disaccharides, whereas, the free disaccharides tended to be inhibitory[163].        


      Fucoidin and Laminarin effects 

      Sulphated fucans, or, as they are more usually known, Fucoidans, are an heterogeneous group of sulphated polysaccharides (~200-500 kDa) that can be extracted from marine algae and some invertebrates[164]. The forms extracted from Ascophyllum nodosum, Fucus vesiculosus and Fucus evanescens have the L-fucose sugar groups forming, by linkage and sulphation, alternating units, with (potentially) tri-sulphated ) pairs (S on C2; C2, C3). So, as with Heparin, there is considerable sulphation, with no clear domain pattern. As supplied commercially, there may be impurities[165] and, together with lack of standardization, difficulties may be incurred relating the outcomes of various experiments. Many forms show an anti-viral effect in vitro (at least) with an history of disappointing results in vivo. In general terms, the larger the molecule and the degree of sulphation determine effectiveness[166], with the full size molecular form producing effects somewhat similar to those of Heparin. Those from L. saccharina, L. digitata, F. vesiculosus, F. serratus and F, distichus can inhibit tumour cell adhesion to platelets appreciably. When Fucoidan (origin uncertain) was administered intravenously to monkeys (100 mg/kg), there was produced an impressive cytokine response at between 3-6 h. Of those measured, and compared with the baseline, the peak readings were :


       ~ MCP-1  >  IL-6  >  M-CSF  >  IL-8  >  G-CSF  >  sTNFR1


      and MMP-9 was also increased. There was mobilization of stem/progenitors from the marrow, particularly of the myeloid series; a mobilization that was not reproduced by Heparan, but partially reproduced by dextran sulphate and accentuated by linear Fucoidan from the sea urchin Lytechinus variegatus. The conclusion was that the stem/progenitor mobilization from the marrow was not dominated by the Selectins, but other factors were involved[167], hinting that the bone marrow had special features. Another group concluded that Fucoidan (? F. vesiculosus, based upon Sigma being quoted generally as the supplier of those items in doubt; in turn, the Sigma product having been described as "crude.") mobilized stem/progenitors independently of integrin, and that the process was "cell intrinsic and does not result from altered micro-environment," but was associated with the release, from inactive myeloid cells, of proteases - elastase and Cathepsin G. Mobilization was not reproduced by Heparin[168]. Fucoidan (as Galactofucan; from Undaria pinnatifida; Mekabu, Japan) had basic structural features studied; the purified molecular weight was ~9,000 Da (which could be referred-to as low molecular weight Fucoidan - see later), with uniform charge, no protein, a sulphate substitution of 0.72 (meaning ~7 Sulphate substitutions per 10 sugar residues - medium degree), fucose:galactose of 1.0:1.1. It showed an anti-Herpes effect in vitro[169]. Pursuing this with in vivo infections of HSV-1 in mice (~22-25 g) treated with oral Fucoidan (U. pinnatifida; 5 mg x 2 or 3 /day) indicated that it had an anti-Herpes effect in vivo, also showing heightened neutralizing antibodies, with macrophage, CTL and B cell activation; effects that transcend a mere blocking of viral glycoprotein for cell attachment[170]. A preparation from Tasmania was able to stimulate, in vivo, human bone marrow precursors (CD34+) to express CXCR4, the receptor for SDF-1[171] (CXCL12). Over 12 treatment days, with subjects taking three oral doses per day, there was minimal/no appreciable change in the peripheral blood neutrophil or lymphocyte counts. Despite the molecular size, it can be taken orally and measured in the blood, rising roughly 1 mg/L/day when taking 3g of 75% Fucoidin per day. This seems promising, because useful blood levels have not generally been claimed for the Fucoidans, but transport to the tissue sites is yet to be established; the suggestion having been that immune effects are produced by lymphoid tissue in the bowel wall (see later regarding HSV). Fucoidan (F. vesiculosus) on murine cells in vitro modulated TNF-α, IL-12 and the Bcl-2 family apoptosis-related factors and afforded some protection from irradiation[172]. Further murine in vitro testing, also including the polysaccharide Arabinogalactan, found that, with exposure to these agents, spleen lymphocytes became cytotoxic to tumour cells, lymphocytes and peripheral macrophages were stimulated to divide, and the latter became more tumouricidal and phagocytic, producing lysosomal enzymes, nitrite, H2O2, TNF-α and IL-6[173]. Another in vitro study with dendritic cells (DCs) derived from peripheral blood demonstrated that Fucoidan (F. vesiculosus; 100 mg/mL[100 mg/L]) drove the immature DCs to maturity with the production of Th1 cytokines[174], having a potency less than that provided by 1 mg/mL (1 mg/L) of bacterial lipopolysaccharides - all features that would be considered desirable to combat cancer, but probably undesirable for treating certain chronic inflammatory diseases such as Rheumatoid arthritis. Another study found that murine in vitro dendritic cell viability is increased, with increased expression of MHC classes I & II, CD56, CD86 and IL-12, and it probably activated the Scavenger receptor-A[175],[176]. The Scavenger receptor of endothelial cells-1 (SREC-1) seems important[177], because it is linked with Neuropilin-1/VEGFR/plexin/Semaphorin on the endothelial (and other) cells, and can endocytose the receptors by stimulation with sulphated polysaccharides, such as Fucoidan (? F. vesiculosus) and Dextran sulphate (MW 66,410). They have been termed "internalization inducers[178]," and may be down-regulated by inflammatory cytokines such as TNF-α[179]. SREC also occurs on macrophages where (with murine cells) it can be up-regulated by LPS at the expense of other scavenger receptors, by an unknown process[180]. With in vitro studies, it can bind and endocytose endoplasmic reticulum-derived chaperones such as Calreticulin with some 50% maximal inhibition provided by Fucoidan (? F. vesiculosus) at 1 mg/L. The binding was strong, attributed to the molecular size[181]. This process has similar features to those for the internalization of HPSE in association with LRP, Mannose-6-phosphate receptor and lipid rafts, with the complexed, traveling molecules ending up in the endosomes for processing by a pathway that, currently, appears unclear, but possibly involving caveolae. The human LRP receptor (= CD91) and CED-1 (CEll Death) of the earthworm, Caenorhabditis  elegans[182] are similar[183], whereas SREC, having 24% identity with CED-1 only has considerable similarity in the extracellular domain, but was reported to have "no homology in the cytoplasmic tail;" the tail being required for cell corpse engulfment, but not apoptosis-recognition or receptor-clustering around the target. CED-1 has a NPxY motif in the cytoplasmic tail to bind to a PTB domain in CED-6[184] and an YxxL of uncertain significance. The site-matching motif in SREC-1 is GLAS, which would not appear functionally equivalent. A study of the minor extracellular domains concluded that the human ortholog for CED-1 is MEGF10[185], with later functional studies support this; also the distinction of MEGFT10 from LRP[186] which, however, share via their NPxY motifs, the downstream GULP. MEGFT10, CED-1 & LRP are highly conserved from the earthworm to humans. In the earthworm, CED-1 functions early in one of the (probably) three pathways, in which it senses the earliest changes of apoptosis in nearby cells revealed by the lipid translocator CD-7/ABCA1, subsequently linking with CED-6 (~human GULP; enGULfment adaPter protein. = hCED-6) and locally organized actin to engulf the dying cell :

       

      WORM/HUMAN

       CED-7/ABCA1

       CED-1/MEGF10(human)/LRP/CD91/MEGF6(Rat)

       CED-6/GULP

       DYN-1/Dynamin

       

      CED-7 is shared by both the apoptotic cell and the phagocytosing cell and provides the earliest trigger. CED-1 is thought to be involved in the recruitment of membrane in order to surround the material to be engulfed, with CED-6 shared by at least one other engulfing pathway, that of Stabilin-2, which is a receptor for phosphatidylserine and, when activated, causes release of the anti-inflammatory cytokine TGF-β[187]. Whilst there is considerable conservation of the basic pathway from the earthworm to humans, there are important differences in application: the earthworm does not have macrophages; those cells developing apoptosis are engulfed by obliging neighbours. In mammals and humans, there are professional phagocytes, particularly macrophages. As expected, they are the cells equipped with the pathways to enable large pieces of debris or cells to be engulfed into large vacuoles. Within the intracytoplasmic domain, CED-1/MEFG10 can engage the link to AP50, a component of clathrin-coated pits[188] and produce large vacuoles. LRP/CD91 also has this ability, acting with Calreticulin as an intermediary to detect apoptotic cells for phagocytosis[189], engulfing fluid and debris by a process termed macropinocytosis. SREC-I has a reduced minor domain size in the extracellular N-terminus but, despite lacking the NPxY motif in the intracytoplasmic domain, can still act as a receptor for the endocytosis of the scavenger ligands such as AcLDL and OxLDL. The mechanism seems poorly understood. The shorter isoform SREC-II cannot do that. Instead, when cells over-expressing these two receptors are placed together, the two SREC's (I & II) cross-link, and the cells aggregate in a process independent of intracellular domains. This aggregation can be blocked by ligands that SREC-I can endocytose[190]. This artificial in vitro observation is of interest, because in vivo, the two receptors in close proximity on a cell's surface, need to be kept from rushing together (except when detecting foreign material). Of the possible contenders, HSPG would be high on the list, as may be Heparin. Fucoidan, Dextran sulphate and Polyinosinic acid would very likely counter the normal separation maintenance and accelerate cross-linking. Other sulphated polysaccharides may also break down normal separation. These may include the abnormal Heparans produced by the Lithium action of gPAPP, and the defective degradation fragments produced from the abnormal Heparin by HPSE.

      The phagocytosis pathways are inhibited by Abl [better known through its inhibitor Imatinib {Gleevec}, which, in turn, is inhibited by Abi[191].]) Participating with the other receptors and Fucoidan (and/or abnormal Heparan), it may internalize the receptors of HPSE and leave the HPSE stranded outside the cell without a functional receptor. Of some other interest has been the claim that, because Fucoidin was considered an inhibitor of L-Selectin function, then intravenous Fucoidin given at the times of TNCB sensitizing and challenging applications for delayed skin contact hypersensitivity reactions in mice reflected the inhibition by L-Selectin[192]. There was no direct evidence to support this possibility, and the oedema produced would be difficult to explain by reference to changes in leukocyte rolling on endothelial cells when the initial inflammatory response would be expected to be from within the local tissues. A more recent report of BCG-induced delayed hypersensitivity induced in the pleural cavities of mice noted that Fucoidin (F. vesiculosus) reduced the migration of CD3 T cells (lymphocytes) at day one, but not later[193]; the subsequent inflammation (hypersensitivity) involving cells already within the stroma:

       

      TABLE 5

       

                                                      Control (saline)                             BCG+Saline                          BCG+Saline+Fucoidan

      Total cell count x106

      ~1.83

      ~6.16

      ~4.23  (68.7% of BCG+Saline)

      Neutrophils x106

      Hinted ~~1.43

      2.69

      1.43    (53.2% of BCG+Saline)

      Lymphocytes x106

      ~0.17

      ~0.31

      ~0.20  (64.5% of BCG+Saline)

      Other cell types x106

      ~~0.23(?)

      3.16

      2.6      (82.3% of BCG+Saline)

      Table: The figures have been read both from the histograms and taken from the text. Other cell types were derived by subtraction from the total. The cell counts relate to the numbers in the pleural washings taken at one day after intra-pleural infusions. In the text, the Fucoidan was noted to maintain the neutrophil cell count after the BCG at "baseline." Accordingly, this figure has been applied in the "Control" column.

       

      This cell movement would be derived from the "irritation-induced" acute inflammation, not hypersensitivity. The chloride of the saline may contribute in part (a more physiological solution would have been preferable), making the "Control" counts of uncertain significance. (The "Control" should have had its "control.") The conclusion from all this is, that, whilst the Fucoidan used seemed to reduce neutrophil migration (partially - as had been reported, in particular) at day one, the main hypersensitivity response that developed later did not involve appreciable numbers of migrating inflammatory cells, supporting the concept that the Fucoidan  inhibited other effectors, of which Heparanase has now been implicated. Work on HPSE found that endothelial cell HPSE release seemed to be a key feature of delayed-type skin hypersensitivity (without reference to the earlier study with Fucoidan)[194]. This response was inhibited by glycol-split non-anticoagulant Heparin (MW 11.2 kDa), a sulphated oligosaccharide. In regard to this, the main effect of the Fucoidan may be more to counter the action of the HPSE - as the glycol-split Heparin and Laminarin sulphate do, rather than to inhibit L-Selectin. The reported inhibition of hypersensitivity reactions (in vivo) by Fucoidin contrasts sharply with the immune stimulatory effects that have been reported, generally from in vitro studies. These may not have the extracellular matrix/stroma with HSPGs to modulate events physiologically, making their relevance questionable. The HPSE involvement in delayed-type hypersensitivity may need to be reconsidered, following the study of the HPSE-/- mice. Nevertheless, redundancy substitutions, such as the MMPs, may still show similar HSPG binding and endocytosis pathways. There still seem many features of receptor function and endocytosis that need elucidation.

      Limited testing of Fucoidan (F. evanescens; 20-40 kDa) on a murine Lewis lung cancer model found that 25 mg/kg (injected via an undisclosed route) was toxic and with an appreciable mortality. When 10 mg/kg was given on day 5, 8 and 12 after tumour injection, there was a significant reduction in tumour growth (to 67.4% of control group) and metastases (to 71.1% of control group) at 20 day. Tumour cathepsin levels were measured, with cathepsins B and D being little changed, but cathepsin L had a highly significant reduction to 30 % of the control group[195]. This is of interest, given that the splitting of HPSE has been attributed (in part) to cathepsin L and both have similar cell re-entry-to-endosomal routes. This study should give some comfort for those who may try Fucoidan administration on humans - at least the Fucoidan did not make the murine tumours grow faster or metastasize more. Low molecular weight Fucoidan (~20 kDa) has a size not very dissimilar to that of the Heparan fragments produced by HPSE. The effects may be similar, with enhancement of the binding of VEGF165 to VGEFR2/NRP1 for (human endothelial cell) differentiation and migration[196]. Laminarin sulphate (β-1,3 Glucan sulphate), as has been assessed[197], and has a MW~ 4,000 kDa with ~2.3 sulphated groups per Glucan unit. It inhibits Heparanase and, with murine in vivo, intraperitoneal dosing, reduced metastases appreciably, but had only a mild tumour anti-proliferative effect. The opinion was that the Heparanase-inhibiting feature was largely related to the degree of sulphation. A linking between the SREC and the receptor(s) for HPSE would not be unexpected.

              

      α-Difluoromethyl ornithine (DFMO)

      Proliferating cells require polyamines, such as putrescine (NH2-CH2-CH2-CH2-CH2-NH2), spermine [NH2-(CH2)3-NH-(CH2)4-NH-(CH2)3-NH2] and spermidine  [NH2-(CH2)3-NH-(CH2)4-NH2] to stabilize DNA at mitosis. These components either enter the body through/from the gut/contents, with the charged molecules binding to, and being assisted into the cell, by the Heparan proteoglycans (exogenous source), or are made within the cells using the enzyme ornithine decarboxylase (endogenous source). The cellular uptake of polyamines was dependent upon their binding to the charged Heparan sulphate groups. DFMO was originally developed as an anti-cancer drug, but the results were disappointing; it only inhibited the endogenous production of the polyamines. Now, better known as a treatment of Trypanosomiasis ("sleeping sickness"), it is considered to be reasonably safe and available (although expensive and difficult to access). In vitro studies demonstrated the involvement of GAGs in the cellular uptake of the polyamine spermine, with the uptake poor if the level of sulphation was low and, in the presence of DFMO, there was an up-regulation of GAG production, including a highly sulphated strong spermine-binding fraction[198]. When the effect of supplying DFMO, in addition to methods inhibiting the cellular uptake of exogenous polyamines, were applied to cancer cells, there were prospects for more success in cancer treatment[199].

      The effect of Lithium on the gPAPP enzyme results in the formation of Heparan sulphate with defective and  reduced sulphation. This may make cells less able to respond to the DFMO-induced adaptation methods of increasing both GAG output and ionic binding strength. Accordingly, Lithium may well both inhibit the cellular uptake of the polyamines, and heighten the sensitivity of proliferating cells to DFMO, a drug that is already on the market. Fucoidan may also have an effect, to be assessed.


      Combined anti-cancer Treatments.

      The most appropriate protocol for Lithium, when used to treat any non-Psychiatric condition, is unknown. Adding another medication will compound complexity. The addition of Fucoidan (oral) would seem a logical step; the Fucoidan would be expected to perturb further the characteristics of the HSPG layer around the cells. Based upon the likelihood that redundancy with adaptions involving positive and negative feedback operate, logic might support the Fucoidan being administered on an alternate day basis, on the same days as the Lithium. With such an administration, at the time when the Lithium is at its peak, and the Heparan chains are most disturbed and displacing the formerly-bound growth factors, the Fucoidan would be there to bind the factors and withhold them from receptors. However, such logic may not prevail, and such a protocol may not prove best; better results may occur if the Fucoidan peaks at the time that the Heparan chains are showing partial reversion to normal. Only by empirical testing may the best protocol be established.

      The DFMO also has to compete with adaption processes of redundancy-like type. A daily routine may not be the most appropriate. Again, to administer it at a time when the Lithium may have just perturbed the Heparan chains may give DFMO the most appropriate time to block the endogenous polyamine production. Fucoidan at that stage may block the exogenous source further. Again, empirical testing may be required.     


      Specific Conditions

       

      1.  Rheumatoid arthritis: The synovial (joint) fluid seems to contain components which, when treated with hyaluronidase, result in an appreciably increased dentine-resorbing activity of the DC-derived osteoclasts produced in vitro by stimulation with M-CSF & RANKL. This would indicate the involvement of chondroitin sulphate-like fragments hyaluronic acid proteoglycans[200]. This feature is of interest because, by the breakdown of the chondroitin-sulphate-like compounds, a block to activity is removed, reminiscent of the Lithium effect on the lymphocyte/NK cells in patient PC. (Perhaps PC's adenocarcinoma produced, with others, a component which, when modified by the Lithium Rx, ceases to suppress the bone marrow micro-environment, or that fragments such as soluble hyaluronic acid stimulate monocyte-derived DCs.) Note has already been made of an appreciable production of HPSE in Rheumatoid joints. The Lithium effect on the HSPG may divert the pro-HPSE from the activation pathway and the inflammation-inducing effectors.
      2.  Lupus Erythematosus: As with Rheumatoid arthritis, a Lithium-induced change of the HSPG would be expected to modify the disease progress in some way. Experience with this condition has been limited but, at a subjective level, Lithium may have aided induction of remissions, and it a patient on long-term Prednisolone prophylaxis seemed able to assist in reducing the maintenance dose to levels not achieved for a long time (years).
      3.  Multiple sclerosis: A prominent feature of the infiltrating CD4 cells is their production of HPSE[201]. Early work indicated that, by modifying the HPSE action using low doses of heparin (or modification of it) could reduce the severity of experimental allergic  encephalomyelitis (EAE), a laboratory model for Multiple Sclerosis (MS) and also allergic arthritis and allergenic skin grafts[202]. There was an optimum dose noted, with the effectiveness falling off with lower and higher doses. Lithium, by modifying the HSPG chains, as by blurring the sulphated domain landmarks and NS/2OS sulphation, may reduce the HSPG substrate suitability for HPSE, with either poor degradation and/or fragments that are heterogeneous with respect to length and sulphation. These, in turn, would be expected to perturb signal transmission either by disturbing the binding/displacement of cytokines in/from the HSPG in the ECM &/or interfere with the ternary receptor complexes (as probably occurs with Heparin without anticoagulant activity on passive EAE in mice[203]). In addition, Lithium inhibits the GSK3β of microglial cells, thereby inhibiting motility and chemotaxis, expression of CD11b and the production of IL-6 and NO[204]. In these ways, Lithium may be able to abort acute relapses within hours (reported earlier[205]). The recent study of the use of Lithium in mice that had EAE induced[206] confirmed the observation with human MS, that bolus Lithium can abort the acute relapses. The mouse study produced results analogous to other human conditions treated, namely, that Lithium could induce remissions in Rheumatoid arthritis. (The protocols had appreciable differences, with the human treatments based upon the alternate day protocols used in rats[207].) The more recent study used Lithium-laced food, supplemented after immunizing injections, by two daily injections of Lithium administered via an unspecified route. Sadly, they ignored one of the key points of the earlier work, namely, that alternate-day treatments were far superior to daily treatments and did not explore that conclusion. Also, they attributed a lack of follow-up interest in the earlier work upon the belief that "the immunosuppression was a toxic effect." It is true that the Lithium serum levels attained shortly after the intraperitoneal injections of the Lithium chloride could be considered quite toxic, particularly after the two doses on one day !



      4. TABLE 6

             Time post Lithium Rx - Rats

        Administration of Lithium                        2 h                                     4 h                                        1 d                                     2 d    

      200mg/kg LiCl i.p.

      3.67 mM/L

      2.11 mM/L

      4h:2h= 57.5%

      0.28 mM/L

      1d:2h= 7.6%

      0.11 mM/L

      2d:2h= 3.0%

      200mg/kg LiCl i.p. x 2, 6h apart

      4.75 mM/L

      3.6  mM/L

      4h:2h= 75.8%

      Not done Not done

      The figures are means of two determinations for each point.

      Table: Serum levels of lithium from rats as dosed according to the published account, with time. The ratios of the levels compared to the 2 h levels are given as percentages.

      These may be contrasted with the serum levels on patient PC taking 500 mg of Lithium  carbonate 6 h apart on alternate days (~15.4 mg/kg/day) :

       

      TABLE 7


         Time post Lithium Rx - Patient PC

       

      Baseline (tail of last cycle) Post 1st 500 mg dose, 45 min Pre 2nd 500 mg dose (at ~6 h) Post 2nd 500 mg dose, 45 min 24 h post 1st dose 48 h post 1st dose
      0.2 mmol/L 0.6 mmol/L 0.4 mmol/L 0.7 mmol/L 0.4 mmol/L 0.2 mmol/L

       

      Given that the peak level after an oral dose is probably ~1.25 h after the dose, adding this time to the abscissa times on a pharmacokinetic Lithium graph[208] and reading off the likely serum values, the equivalent 2 h human serum level after 31.4 mEq Lithium citrate syrup (1.16 g Lithium carbonate) was ~0.72 mmol/L and that at the equivalent 4 h point, ~0.5 mmol/L, being ~69% of the former. The level at 24 h was ~0.21 mmol/L (~29%) and at 48 h, 0.11 mmol/L (~15.3%).

      Clearly, there are appreciable discrepancies, with the serum Lithium levels in the rats at 2 h and 4 h much higher (and, seemingly without appreciable toxic symptoms) than the roughly equivalent levels (as related by time) found in the human patient, and the decline in the rats after the second dose of the day was very much slower, for reasons that are not clear. If we assume that there was no technical problem in making the determinations, the other likely explanation could, in part, be a species variation in the plasma clearance of Lithium into the cellular compartment. Elsewhere on the website, there have been noted some inconsistencies in the claimed relationship between the intracellular Lithium level versus the extracellular level. The authors make no mention of the mental status of the intact rats, only mentioning that the adrenalectomized rats "became weak and lethargic after i.p. injection of 200 mg/kg LiCl." There is the possibility that, for intact rats, mental function is not appreciably altered by such high serum levels because the lithium has not entered the intracellular compartment appreciably. When it does equilibrate at ~1 day post dose, the serum level is not high (0.28 mmol/L) - it is barely at a therapeutically relevant level for humans !  The Lithium seems to have vanished ! We may wonder if the intracellular level was ever "high."                            

      In humans, such high post-injection levels would be impossible to achieve with equivalent  oral, bolus doses of Lithium carbonate, because everyone would vomit the doses up !  Patients start tending towards vomiting when the bolus dose rises above ~750 mg. To  attain a serum level of ~4.75 mmol/L would require ~4.75 gram - an excellent emetic !  Accordingly, hyperacute toxicity has never been witnessed. Acute toxicity from an  excessive dose taken over some days resulted in a patient with a serum Lithium of ~2.2  mmol/L. She was incoherent, confused and unable to sit up. [She was never given lithium again  but, for a patient having breast cancer with marrow infiltration and a pancytopenic, leukoerythroblastic  peripheral blood picture, she subsequently went on to survive for a remarkable time !] To attempt to  attain such high Lithium levels recorded in rats, in humans by injection or infusion, would  very likely kill the patients. How the rats survived is an enigma.

      The authors had an unfortunate choice of wording in attributing the Lithium effects to  toxicity. "Toxicity" implies "poisonous," or something harmful to metabolic processes.  From the rats' viewpoint, the lithium was not toxic; it was most beneficial; it prevented or  alleviated the EAE disease process. The authors noted a pronounced thymic cortical  atrophy in the intact rats. This may indicate a stress-related response, because it was not a  feature with the adrenalectomized rats. However, the latter had appreciably increased  thymic weights, which probably surprised the authors sufficiently for them to ignore those  figures.

       

      TABLE 8

       

      Rats - Lithium and Thymus weight

       Rats (180-200 g)                  Dose LiCl              Schedule                                                                                              Thymus weight

      Adrenalectomy 200 mg/kg Once per day x10 562 27 mg
      Intact 200 mg/kg Once per day x10 283 23 mg
      Intact 200 mg/kg Twice per day on alternate days (x10 doses) 228 5 mg
      Intact Control Saline   424 34 mg

       

      Today, and in the current discourse, the increased thymic weights in the adrenalectomized  rats may be caused by the Lithium-induced, disrupted GAG chain sulphation within the  thymus, removing the DC-HIL-Syndecan-4 inhibition over thymocyte proliferation.  Other factors that may be relevant to the Multiple sclerosis disease process may include  anti-phospholipid antibodies and cerebral endothelial cell abnormalities. The significance  of their participation remains to be established[209].

      1.  Sarcoidosis. Sarcoidosis is considered to involve an excessive and unbalanced cytokine stimulation[210].  Lithium, by perturbing the ternary ligand/HSPG/receptor function would be expected to influence cytokine activity. Patients have reported subjective improvements when using the Lithium Rx protocol. Fucoidan may be expected to have an effect  also.
      2.  Herpes simplex virus, Varicella and Dengue. HSPGs provide binding platforms for HSV1 & 2, Varicella (Chickenpox) and others of the Herpesvirus group[211]. After Heparan-aided cell binding, the unusual and relatively infrequent, specifically sulphated disaccharide -

                                               -IdoA2S GlcNH23S6S-  or  -IdoA2S GlcNH23S-

      (from the enzyme 3-OST-3A or B isoforms) can provide an additional link to the HSV1 (>HSV2) coat glycoprotein gD, that aids viral cell entry[212],[213]. The documented anti-Herpes type I effects of Lithium may be explained (in part, at least) by the latter disturbing the specific coding of the HSPG. A similar explanation may not apply so readily for Herpes type II and Varicella (Chickenpox; VZV), although the involvement of Heparan has been established[214], and the cellular HSPG, when stripped of its sulphate groups by chlorate, lost its ability to bind or aid infection (ie a not very dissimilar structural change to that of the gPAPP-/- phenotype, though probably more severe). Fucoidan, likewise, can compete with Heparan and block viral entry (see earlier). VZV has the smallest genome amongst herpesviruses and, whilst this virus is said to share the gene for gE[215], it does not have the gene for the gD coat component, instead, relying upon gB and gC for initial binding. Initial attachment to the cells was by HSPG, with the Mannose 6-Phosphate receptor taking over the endocytosis stage. Heparin, but not Chondroitin sulphate, blocks infection[216] by free viruses. The virus is thought to spread through human subjects by cell-to-cells contact, relying upon cell fusion for the virus to gain entry into the next cell[217]. Accordingly, free viral binding to Heparan at the vesicular stage does not seem so important, rather the fusogenic mechanism.  However, Varicella (VZV) still performs best with both glycoprotein coat components gB[218] and gE[219], of which the latter, in the absence of gD, takes on a wider role of functions and may, in the presence of Heparan sulphate (or another binding co-factor) assist the viral-induced fusion process by aggregation on the cell surface and associating with cellular molecules[220]. When there is entry by endocytosis, early processing of the VZV (or CMV[221],[222]) has similarities with those for HPSE[223],[224],[225]: there is (partial) blocking by Heparin (&/or FGF), indicating a Heparan role, and (partial) inhibition by blocking the Mannose-6-phosphate receptor (of lipid rafts) and (partial) inhibition when acidification of the late endosomes &/or lysosomes is prevented. By extension, the Lithium-induced changes may result in an inability of HPSE to gain re-entry into the cells, leaving the pro-enzyme on the loose, and more able to enter the circulation. [A trumpet player was troubled by recurrent Herpes simplex lesions (cold sores) on his lip. He agreed to try a prophylactic regime of Lithium tablets, and received a prescription for 200 tablets. He did not return for a repeat prescription. When he did return about another issue, he claimed that the Lithium had stopped the cold sores and that, on running out of the tablets, the cold sores did not return ! There is the possibility that the Lithium allowed the development of better innate immune mechanisms to prevent reactivation of the virus. See earlier with respect to Undaria pinnatifida Fucoidan.] Clinical experience with Lithium for early Chickenpox is documented[226].         

      If Lithium is able to modify the sulphated polysaccharides that make up the HSPGs and thereby reduce the incidence &/or severity of infections involving the Herpes group of viruses, serious consideration could be given to examining the place for prophylactic Lithium use in renal dialysis, bone marrow and organ transplant units, and patients with severely compromised immune function.

      Dengue, the scourge of urbanized tropical regions, is caused by a virus that causes infected cells to produce, via the Golgi apparatus, a non-structural glycoprotein (NS1). This leaves the infected cells and attaches to the cell surfaces of uninfected cells by means of GAGs, including Heparan and Chondroitin sulphate E (which is the only type with two sulphate groups on one sugar molecule of the disaccharide). Binding is anatomically selective, with particular endothelial and mesenchymal cell types involved, and there must be a high level of sulphation[227]. There is attachment to the infected cells, but the mechanism is unknown. The distribution of NS1 to non-infected cells may contribute to the clinical syndrome of DHF and DSS. The changes that Lithium may bring about in both the Heparan and Chondroitin sulphates may have an appreciable modifying effect on the course of the complications of Dengue. Bolus Lithium could be administered to all the patients considered to have Dengue fever, and who may be at risk of progressing to Dengue haemorrhagic fever or Dengue shock syndrome (the feared complications).



      Epilogue

      The recognition that Lithium may be used to modify the characteristics of the Heparan sulphate on the surface of cells has opened up new vistas in therapy. That is because the basic Pharmacology of Lithium is known and it has been used in clinical practice for about 50 years. Most observations to date are empirical. There is a need to establish appropriate protocols by applying organized testing.

       

      Malcolm Adams Traill

      Copyright  MA Traill 15th August 2009

       


       

       

      [4] Younge BR Weyand CM et al. Trans. Am. Ophthalmol. Soc. 2007; 105:141-145

      [5] Vlodavsky I Ilan N et al. Curr. Pharm. Design. 2007; 13:2057-7

      [6] Bernfield M Gtte M et al. Annu. Rev. Biochem. 1999; 68:729-77

      [7] Esko JD & Selleck SB. Annu. Rev. Biochem. 2002; 71:435-71

      [8] Bass MD & Humphries MJ. Biochem. J. 2002; 368:1-15

      [9] Saoncella S Echtermeyer F et al. Proc. Natl. Acad. Sci. USA. 1999; 96:2805-110

      [10] Herz J & Strickland DJ. J. Clin. Invest. 2001; 108:779-84

      [11] Kornfeld S. J. Clin. Invest. 1986; 77:1-6

      [12] Ben-Zaken O Shafat I et al. Int. J. Biochem. Cell. Biol. 2008; 40(3):530-42

      [13] Wood RJ & Hulett MD. J. biol. Chem. 2008; 238(7):4165-76 

      [14]Vreys V Delande N et al. J. Biol. Chem. 2005; 280(39):33141-8

      [15] Ben-Zaken O Gingis-Velitski S et al. Biochem. Biophys. Res. Commun. 2007; 361(4):829-34

      [16] Chu CL Bukzek-Thomas JA et al. Biochem. J. 2004; 379:331-41

      [17] Parton RG & Richards AA. Traffic 2003; 4:724-38

      [18] LeRoy C & Wrana JL. Nat. Rev. Mol. Cell. Biol. 2005; 6:112-25

      [19] Cheng Z-J Singh RD et al. Mol. Memb. Biol. 2006; 23(1):101-110

      [20] Poon GM & Garipy J. Biochem. Soc. Trans. 2007; 35(4):788-93

      [21] Han C Belenkaya T et al. Development 2004; 131(3):601-11

      [22] Cheng F Mani K et al. J. Biol. Chem. 2002; 277(46):44431-9

      [23] Mani K Cheng F et al. Glycobiology 2006; 16(12):1251-61

      [24] Pelkmans L Brli T et al. Cell 2004; 118:767-80

      [25] Probst OC Puxbaum V et al. Int. J. Cancer 2009; 124:2559-67

      [26] Brandley BK Ross TS et al. J. Cell. Biol. 1987; 105:991-997  

      [27] Ben-Zaken O Shafat I et al. Int. J. Biochem. Cell Biol. 2008; 40(3):530-42

      [28] Zetser A Bashenko Y et al. Cancer Res. 2003; 63:7733-41

      [29] Nadav L Eldor A et al. J. Cell Sci. 2002; 115:2179-2187

      [30] Lazzarino D & Gabel CA. J. Biol. Chem. 1990; 265(20):11864-71

      [31] Takahashi K Ueno T et al. Biol. Pharm. Bull. 2009; 32(3):475-9

      [32] Abboud-Jarrous G Rangini-Guetta Z et al. J. Biol. Chem. 2005; 280(14):13568-75

      [33] ShafatI Vlodavsky I et al. J. Biol. Chem. 2006; 281(33):23804-11

      [34] Sasaki N Higashi N et al. J. Immunol. 2004; 172:3830-5

      [35] Fedarko NS & Conrad HE. J. Cell Biol. 1986; 102:587-99

      [36] Ishihara M Fedarko NS et al. J. Biol. Chem. 1987; 262(10):4708-16

      [37] Richardson TP Trinkaus-Randall V et al. J. Cell Sci. 2001; 114:1613-1623

      [38] Schubert SY Ilan N et al, Lab. Investig. 2004; 84:535-44

      [39] Hsia E Richardson TP et al. J. Cell. Biol. 2003; 88:1214-25

      [40] Nobuhisa T Naomoto Y et al. Cancer Sci. 2007; 98(4):535-40

      [41] Guimond S Maccarana M et al. J, Biol. Chem. 1993; 268(32):23906-14

      [42] Parish CR Freeman C et al. Cancer Res. 1999; 59:3433-41

      [43] Kato M Wang H et al. Nat. Med. 1998; 4(6):691-607 

      [44] Bar-Ner M Eldor A et al. Blood 1987; 70:551-7

      [45] Nakajima M Irimura T et al. J. Biol. Chem. 1984; 259(4):2283-90

      [46] Gallagher JT. Biochem. Soc. Trans. 2006; 34(3):438-41

      [47] Pikas DS Li J-P et al. J. Biol. Chem. 1998; 273(30):18770-77

      [48] Vreys V & David G. J. Cell. Mol.Med. 2007; 11(3):427-52

      [49] Bai X Bame KJ et al. J. Biol. Chem. 1997; 272(37):23172

      [50] Yang Y MacLeod V et al. J. Biol. Chem. 2007; 282(18):13326-33

      [51] Chen L & Sanderson RD. PLoS ONE 2009; 4(3):e4947

      [52] Lider O Cahalon L et al. Proc. Natl. Acad. Sci. USA. 1995; 92:5037-41

      [53] Datta SR Brunet A et al. Genes Dev. 1999; 13:2905-27

      [54] Gingis-Velitski S Zetser A et al. J. Biol. Chem. 2004; 279(22):23536-41

      [55] Zcharia E Jia J et al. PLosONE  2009; 4(4):e5181

      [56] Zetser A Bashenko Y et al. Cancer Res. 2003; 63:7733-41

      [57] Goldshmidt O Zcharia E et al. Proc. Natl. Acad. Sci. USA 2002; 99(15):10031-6

      [58] Escobar Galvis ML Jia J et al. Nat. Chem Biol. 2007; 3(12):773-8

      [59] Vlodavsky I Goldshmidt O et al. Cancer Biol. 2002; 12:121-9

      [60] Presto J Thuveson M et al. Proc. Natl. Acad. Sci USA. 2008; 105(12):4751-6

      [61] Mahtouk K Hose D et al. Blood 2007; 109(11):4914-23

      [62] Morimoto-Tomita M Uchimura K et al. J. Biol. Chem. 2002; 277(51):49175-85

      [63] Dhoot GK Gustafsson MK et al. Science 2001; 293:1663-6 

      [64] Li J Kleef J et al. Mol. Cancer 2005; 4:14

      [65] Dai Y Yang Y et al. J. Biol. Chem. 2005; 280(48): 40066-73

      [66] Lai J Yu C et al. Gastroenterology 2006; 130:2130-44

      [67] Lai J Sandhu DS et al. Hepatology 2008; 47(4):1211-22

      [68] Annabi B Lachambre M-P et al. Biochem. J. 2001; 353:547-553

      [69] Sato H Takino T et al. Nature 1994; 370:61-5

      [70] Nakahara H Howard L et al. Proc. Natl. Acad. Sci. USA 1997; 94:7959-64

      [71] Annabi B Laflamme C et al. J. Neuroinflammation 2009; 6:8

      [72] Lehti K Valtanen H et al. J. Biol. Chem. 2000; 275(20):15006-13

      [73] Ilan N Elkin M et al. Int. J. Biochem. Cell Biol. 2006; 38:2018-39

      [74] Narazaki M & Tosato G. Blood 2006; 107(10):3892-901

      [75] Salikhova A Wang W et al. Circ. Res. 2008; 103(6):e71-e79

      [76] Dhaliwal BS & Steinbrecher UP. Clin. Chim. Acta 1999; 286:191-205

      [77] Adachi H Tsujimoto M et al. J. Biol. Chem. 1997; 272(50):31217-20

      [78] Narazaki M Segarra M et al. Blood 2008; 111:4126-36

      [79] Tamura Y Adachi H et al. J. Biol. Chem. 2003; 278(15):12613-17

      [80] Tordjman R Ortga N et al. Blood 1999; 94(7):2301-9

      [81] Frenette PS & Weiss L. Blood 2000; 96:2460-8

      [82] Smith SM West LA et al. Matrix Biology 2007; 26:175-84

      [83] Evanko SP Tammi MI et al. Adv. Drug Del. Rev. 2007; 59(13):1351-65

      [84] Arneson LS & Miller J. Biochem. J. 2007; 406:97-103

      [85] Wegrowski Y Milard A-L et al. Clin. Exp. Immunol. 2006; 144:485-93

      [86] Gtte M. FASEB J. 2003; 17:575-91

      [87] Bourguignon LY Singleton PA et al. J. Biol. Chem. 2004; 279(26):26991-27007

      [88] Termeer CC Hennies J et al. J. Immunol. 2000; 165:1863-70

      [89] Selleck SB. Trends Genet. 2000; 16(5):206-12

      [90] Miller JR Hocking AM et al. Oncogene 1999; 18:7860-72

      [91] Ghiselli G & Agrawal A. Biochem. J. 2005; 390:493-9

      [92] de Mestre AM Khachigian LM et al. J. Biol. Chem. 2003; 278(50):50377-85

      [93] Ziegler S Rhrs S et al. FEBS J. 2005; 272:1600-15

      [94] Chen R-H Ding V et al. J. Biol. Chem. 2000; 275(23):17894-99

      [95] Jope RS & Bijur GN. Mol. Psychiat. 2002; 7:S35-S45 

      [96] Virk MS & Lieberman JR. Arthritis Res. Ther. 2007; 9(Suppl 1):S5

      [97] Sheldahl LC Slusarski DC et al. J. Cell Biol 2003; 161(4):769-77

      [98] Roshy S Sloane BF et al. Cancer Metastasis Rev. 2003; 22:271-86

      [99] Salazar Murphy LL & Hughes CC. J Immunol. 2002; 169:3717-25

      [100] Edovitsky E Lerner I et al. Blood 2006; 107(9):3609-16

      [101] Bussini S Meda L et al. Immun. Ageing. 2005; 2:11

      [102] Kodaira Y Nair SK et al. J. Immunol. 2000; 165:1599-604

      [103] Wrenshall LE Cerra FB et al. J. Immunol. 1995; 154:871-80

      [104] Wrenshall LE Stevens RB et al. J. Leukoc. Biol. 1999; 66:391-400

      [105] Ihrcke NS & Platt JL. J. Cell. Physiol. 1996; 168:625-37

      [106] Fitzgerald ML Wang Z et al. J. Cell Biol. 2000; 148(4):811-24

      [107] Subramanian SV Fitzgerald ML et al. J. Biol. Chem. 1997; 272(23):14713-20

      [108] Alexander CM Reichsman F et al. Nat. Genet. 2000; 25:329-332

      [109] Park PW Pier GB et al. J. Biol Chem. 2000; 275(5):3057-64

      [110] Li Q Park PW et al. Cell 2002; 111:635-46

      [111] Gtte M Joussen AM et al. Invest. Ophthalmol. Vis. Sci. 2002; 43:1135-41

      [112] Bayer-Garner IB Sanderson RD et al. Mod. Pathol. 2001; 14(10):1052-8  

      [113] Yang Y Yaccoby S et al. Blood 2002; 100(2):610-719

      [114] Shikano S Bonkobara M et al. J. Biol. Chem. 2001; 276(11):8125-34

      [115] Chung J-S Sato K et al. Blood 2007; 109:4320-7

      [116] Chung J-S Dougherty I et al. J. Immunol. 2007; 179:5778-84

      [117] Whitelock JM Murdoch AD et al. J. Biol. Chem. 1996; 271(17):10079-86

      [118] Kodaira Y Nair SK et al. J . Immunol. 2000; 165:1599-604

      [119] Cahalon L Lider O et al. Int. Immunol. 1997; 9(10):1517-22

      [120] Chowers Y Lider O et al. Gastroenterology 2001; 120:449-159

      [121] Hershkoviz R Schor H et al. Immunology 2000; 99:87-93

      [122] Hecht I Hershkoviz R et al. J. Leukoc. Biol. 2004; 75:1139-46

      [123] Kosir MA Quinn C et al. J. Surg. Res. 1997; 67(1):98-105

      [124] Lerner I Baraz L et al. Clin. Cancer Res. 2008; 14(3):668-76

      [125] Vlodavsky I Friedmann Y. J. Clin. Invest. 2001; 108:341-347

      [126] Maxhimer JB Quiros RM et al. Surgery 2002; 132:326-33

      [127] Shafat I Pode D et al. Neoplasia 2008; 10(2):125-30

      [128] Shafat I Ben Barak A et al. Neoplasia 2007; 9(11):909-916

      [129] Kelly T Suva LJ et al. Cancer Res. 2005; 65(13):5778-84

      [130] Rivollier A Mazzorana M et al. Blood 2004; 4027-37

      [131] de Vries TJ Schoenmaker T et al. J. Leukoc. Biol. 2009; 85(6):919-27

      [132] Alnaeeli M Penninger JM et al. J. Immunol. 2006; 177:3314-26 

      [133] Tydell CC David-Fung E-S et al. J. Immunol. 2007; 179:421-38 

      [134] Galy A Travis M et al. Immunity 1995; 3:458-73

      [135] Six EM Bonhomme D et al. J. Exp. Med. 2007; 204(13):3085-93

      [136] Sasaki N Okishio K et al. J. Biol. Chem. 2008; 283(6):3594-606

      [137] Netelenbos T van den Born J et al. J. Leucoc. Biol. 2003; 74:1035-44

      [138] Gupta P Oegema TR et al. 1998; Blood 1998; 92:4641-5

      [139] Jastrebova N Vanwildemeersch M et al. J. Biol. Chem. 2006; 281(37):26884-92 

      [140] Li RW Freeman C et al. Arthritis Rheum. 2008; 58(6):1590-1600

      [141] de Mestre AM Staykova MA et al. J. Leucoc. Biol. 2007; 82:1289-1300

      [142] Frederick JP Tafari AT et al. Proc. Natl. Acad. Sci. USA. 2008; 105(33):11605-12

      [143] Riesenfeld J Hk M et al. J. Biol. Chem. 1980; 255(3):922-8

      [144] Mulloy BS & Forster M. Glycobiology 2000; 10(11):1147-56

      [145] Egeberg M Kjeken R et al. Biochim. Biophys. Acta 2001; 1541:135-49

      [146] Brickman YG Ford MD et al. J. Biol. Chem. 1998; 273(8):4350-9

      [147] Jayson GS Lyon M et al. J. Clin. Chem. 1998; 273(1):51-57

      [148] David G & van den Berghe H. J. Biol. Chem. 1983; 258(12):7338-44

      [149] Safaiyan F Lindahl U et al. Eur. J. Biochem. 1998; 252:576-82

      [150] Johnson CE Crawford BE et al. Stem Cells 2007; 25:1913

      [151] Liu D Shriver Z et al. Proc. Natl. Acad. Sci. USA. 2002; 99(2):568-573

      [152] Grobe K Inatani M et al. Development 2005; 132(16):3777-86

      [153] Ledin J Staatz W et al. J. Biol Chem. 2004; 41(8):42732-41

      [154] Merry CL Bullock SL et al. J. Biol. Chem 2001; 276(38):35429-34

      [155] Frederick JP Tafari AT et al. Proc. Natl. Acad. Sci USA. 2008; 105(33):11605-12

      [156] Carlsson P Presto J et al. J. Biol. Chem 2008; 283(29):20008-14

      [157] Rong J Habuchi H et al. Biochem. J. 2000; 346:463-8

      [158] Jia J Maccarana M et al. 2009; 284(23):15942-50

      [159] Kreuger J Spillmann D et al. J. Cell Biol. 2006; 174(3):323-7

      [160] Maccarana M Sakura Y et al. J. Biol. Chem. 1986; 271(30): 17804-10

      [161] Mulloy BS & Forster M. Glycobiology 2000; 10(11):1147-56 

      [162] Yang Y MacLeod V et al. Blood 2007; 110:2041-8

      [163] Bezouska K Yuen C-T et al. Nature 1994; 272;150-7

      [164] Cumashi A Ushakova NA ey al. Glycobiology 2007; 17(5):541-52

      [165] Berteau O & Mulloy B. Glycobiology 2003; 13(6):29R-40R

      [166] Damonte EB Matulewicz MC et al. Curr. Medicinal Chem. 2004; 11:2399-419

      [167] Sweeney EA Priestley GV et al. Proc. Natl. Acad. Sci. USA. 2000; 97(12):6544-49

      [168] Hidalgo A Peired AJ et al. Blood 2004; 104:993-1001

      [169] Lee J-B Hayashi K et al. Chem. Pharm. Bull. 2004; 52(9):1091-4

      [170] Hayashi K Nakano T et al. 2008; 109-16

      [171] Irhimeh MR Fitton JH et al. Exp. Hematol. 2007; 35(6):989-94

      [172] Byon Y-Y Kim M-H et al. J. Vet. Sci. 2008; 9(4):359-65

      [173] Choi E-M Kim A-J et al. J. Med. Food 2005; 8(4):446-53

      [174] Yang M Ma C et al. Int. Immunopharmacol. 2008; 8(13-14):1754-60

      [175] Kim M-H & Joo H-G. Immunol. lett. 2008; 115:138-143

      [176] Platt N & Gordon S. J. Clin. Invest. 2001; 108:649-54

      [177] Adachi H & Tsujimoto M. Prog. Lipid Res. 2006; 45:379-404

      [178] Narazaki M Segarra M et al. Blood 2008; 111:4126-36

      [179] Adachi H & Tsujimoto M. J. Biol. Chem. 2002; 277(27):24014-21

      [180] Tamura Y Osuga J-i et al. J. Biol. Chem. 2004; 279(30):30938-44

      [181] Berwin D Delneste Y et al. J. Biol. Chem. 2004; 279(49):51250-7

      [182] Kinchen JM Cabello J et al. Nature 2005; 434:93-9

      [183] Su HP Nakada-Tsukui K et al. J. Biol. Chem, 2002; 277(14):11772-9

      [184] Zhou Z Hartwieg E et al. Cell 2001; 104:43-56

      [185] Callebaut I Mignotte V et al. Biochem. Biophys. Res. Commun. 2003; 300:619-23

      [186] Hamon Y Trompier D et al. PLosONE 2006; 1(1):e120

      [187] Park S-Y Kang K-B et al. J. Biol. Chem. 2008; 283(16):10593-600

      [188] Suzuki M & Nakayama M. Exp. Cell Rec. 2007; 313:3729-42

      [189] Ogden CA deCathelineau A et al. J. Exp. Med. 2001; 194(4): 781-95

      [190] Ishii J Adachi H et al. J. Biol. Chem. 2002; 277(42):39696-702

      [191] Hurwitz ME Vanderzalm PJ et al. PLoS Biol. 2009; 7(4):e1000099

      [192] Nasu T Fukuda Y et al. Immunol. Lett. 1997; 59:47-51

      [193] Souza MC Penido C et al. Infect. Immun. 2008; 76(12):5686-93

      [194] Edovitsky E Lerner I et al. Blood 2006; 107(9):3609-16

      [195] Alekseyenko TV Zhanayeva SY et al. Bull. Exp. Biol. Med. 2007;143(6):730-2  

      [196] Lake AC Vassy R et al. J. Biol Chem. 2006; 281(49):37844-52

      [197] Miao H-Q Elkin M et al. Int. J. Cancer 1999; 83:424-31

      [198] Belting M Persson S et al. Biochem. J. 1999; 338:317-23

      [199] Belting M Borsig L et al. Proc. Natl. Acad. Sci. USA. 2002; 99(1):371-6

      [200] Rivollier A Mazzorana M et al. Blood 2004; 104:4029-37

      [201] de Mestre AM Staykova MA et al. J. Leukoc. Biol 2007; 82:1289-1300

      [202] Lider O Baharav E et al. J. Clin. Invest. 1989; 83:752-6

      [203] Willenborg DO & Parish CR. J. Immunol. 1988; 140(10):3401-5

      [204] Yuskaitis CJ & Pope RS. Cellular Signalling 2009; 21:264-73

      [205] Traill MA; http://www.malsmusings.info/index_files/UHF.htm

      [206] De Sarno P Axtell RC et al. J. Immunol. 2008; 181:338-45

      [207] Levine S & Saltzman A. Immunopharmacology 1991; 22:207-214

      [208] Reiss RA Haas C et al. Clin. Pharmacol. Ther. 1994; 55:392-8

      [209] Horstman LL Wenche j et al. J. Neuroinflammation 2009; 6:3

      [210] Ziegenhagen MW & Mller-Quernheim J.  J. Intern. Med. 2003; 253(1):18-30

      [211] Shukla D & Spear PG. J. Clin. Invest. 2001; 108:503-10

      [212] Shukla D Liu J et al. Cell 1999; 19:13-22

      [213] Spear PG & Longnecker R. J. Virol. 2004; 77(19):10179-85

      [214] Zhu Z Gershon MD et al. Proc. Natl. Acad. Sci. USA. 1995; 92:3546-50

      [215] Davison AJ & McGeoch DJ. J. Gen. Virol. 1986; 67:597-611

      [216] Zhu Z Gershon MD et al. Proc. Natl. Aca.. Sci. 1995; 92:3546-50

      [217] Chen JJ Zhu Z et al. Cell 2004; 119:915-26

      [218] Jacquet A Haumont M et al. Virus Res. 1998; 53:197-207

      [219] Maresova L Pasieka TJ et al. J. Virol. 2001; 75(19):9483-92

      [220] Cole NL & Grose C. Rev. Med. Virol. 2003; 13:207-22

      [221] Ryckman BJ Jarvis MA et al. J. Virol. 2006; 80(2):710-22

      [222] Compton T Nowlin DM et al. Virology 1993; 193:834-841

      [223] Vreys V Delande N et al. J. Biol. Chem 2005; 280(39):33141-8

      [224] Ben-Zaken O Shafat I et al. Int. J. Biochem. Cell. Biol. 2008; 40(3):530-42

      [225] Zetser A Levy-Adam F et al. J. Cell Sci. 2004; 117:2249-58

      [227] Avirutnan P Zhang L et al. PLoS Pathogens 2007; 3(11):e183